ASHWORTH Andrew, HORDER Jeremy - Principles of Criminal Law - PDFCOFFEE.COM (2024)

Preface

Principles of Criminal Law (7th edn) Andrew Ashworth and Jeremy Horder Publisher: Oxford University Press Print ISBN-13: 9780199672684 DOI: 10.1093/he/9780199672684.001.0001

Print Publication Date: May 2013 Published online: Sep 2013

Preface For this seventh edition I am delighted to be joined by Professor Jeremy Horder. He has undertaken most of the revisions for this edition, and I am grateful to him for bringing his deep knowledge and understanding of the criminal law to bear on the text. There have been considerable developments both in the law and in scholarship in the four years since the last edition. Thus, for example, the homicide provisions of the Coroners and Justice Act 2009 have begun to be interpreted by the courts, and there have been substantial developments in the law on complicity. These and other changes have been taken into account, and it is hoped that statements of the law were correct at 1 December 2012. The general layout and order of chapters remain unchanged for this edition. The context and functions of the criminal law are outlined in Chapter 1, and Chapter 2 on criminalization examines reasons for creating or for not creating criminal laws. Chapter 3 then discusses key principles and policies relevant to the criminal law. In Chapters 4, 5, and 6, the ‘general part’ elements of culpability, justification, and excuse are analysed: Chapter 4 deals generally with actus reus questions, Chapter 5 is devoted to criminal capacity (insanity, infancy, and corporate liability) as well as to mens rea issues, and Chapter 6 deals with excusatory defences. Three areas of the special part of the criminal law are then selected for examination: Chapter 7 deals with homicide, Chapter 8 with non-fatal physical violations

Preface (including sexual offences), and Chapter 9 with offences of dishonesty. The book concludes with Chapter 10 on complicity and Chapter 11 on inchoate offences. As in previous editions, the focus of the book is upon the identification and discussion of issues of principle and policy raised by the statements of the courts, Parliament, the law reform bodies, and academic commentators. The judgments of the courts provide much material for discussion, and the resurgence of criminal law scholarship has continued, with the publication of important new monographs, articles, and essays. The contention is not that English criminal law is grounded in a stable set of established doctrines: on the contrary, there is ample evidence that the arguments and assumptions that influence the development of the law form a disparate group, sometimes conflicting and sometimes invoked selectively. Often there are political factors influencing the shape of legislation or the activities of law enforcement officers, and reference is made to these below. But the aim of the book is to focus on principles, some of which are immanent in existing legal rules and practices, some of which are not recognized (or not fully recognized) and which are commended here on normative grounds. To conduct a full normative argument on many of these points would require greater detail, in discussing elements of moral and political philosophy or of criminology, than is possible within the confines of this book. The same applies to (p. vi) comparative legal material: some references are made, particularly to the American Law Institute's Model Penal Code, but it is not possible to go much further here. The processes of Oxford University Press have been splendidly efficient, and we are grateful to John Carroll and Natasha Flemming for their support. We have retained the gender specific ‘he’ in most parts of the book when referring to defendants and offenders, on the ground that the vast majority of them are male. A.J.A.

New to this Edition

Principles of Criminal Law (7th edn) Andrew Ashworth and Jeremy Horder Publisher: Oxford University Press Print ISBN-13: 9780199672684 DOI: 10.1093/he/9780199672684.001.0001

Print Publication Date: May 2013 Published online: Sep 2013

New to this Edition • Increased coverage of key cases for undergraduates. • Takes full account of the effect of the Coroners and Justice Act 2009 on the law of homicide. • Provides discussion of key new cases including: • R v Clinton, Parker and Evans [2012] EWCA Crim 2 (loss of self-control), and • R v Dowds [2012] EWCA Crim 281 (diminished responsibility). • The chapter on Complicity has also been substantially revised, in part to take account of recent cases such as R v ABCD [2010] EWCA Crim 1622. (p. viii)

Table of cases

Principles of Criminal Law (7th edn) Andrew Ashworth and Jeremy Horder Publisher: Oxford University Press Print ISBN-13: 9780199672684 DOI: 10.1093/he/9780199672684.001.0001

Print Publication Date: May 2013 Published online: Sep 2013

Table of cases A (Conjoined Twins: Surgical Separation), Re [2000] 4 All ER 961 101, 107, 119, 133, 134, 239 A v Home Secretary [2004] UKHL 56 50, 51, 81 A v UK 118 ABCD [2010] EWCA Crim 1622 437 Abbott v R [1977] AC 755 212 Abdul-Hussain et al [1999] Crim LR 570 208 Acott [1997] 2 Cr App R 94 253, 255 Adams [1957] Crim LR 365 132, 281, 282 Adomako [1995] 1 AC 171 290, 293, 294, 299, 301 ADT v United Kingdom (2000) 31 EHRR 33 338, 340 Ahluwalia [1992] 4 All ER 889 (CA) 254, 255 Airedale NHS Trust v Bland [1993] AC 789 100, 101, 133, 239, 282 Aitken (1992) 95 Cr App R 304 196, 324 Albert v Lavin (1981) 72 Cr App R 178 216 Aldo Bordessa (C–358/93 & C–416/93) [1995] ECR I–361 61 Allan [1965] 1 QB 130 426

Table of cases Allen [1988] Crim LR 698 200 Allen v Whitehead [1930] 1 KB 211 149 Alphacell Ltd v Woodward [1972] AC 824 105, 152, 164 Anderson [1986] AC 27 476 Anderton v Ryan [1985] AC 560 466 Andrews v DPP [1937] AC 576 176, 287, 289, 290 Andronicou and Constantinou v Cyprus (1998) 25 EHRR 491 121, 126–8 Arnold [1997] Crim LR 833 385 Arrowsmith [1975] QB 678 222 Associated Octel [1996] 1 WLR 1543 148, 164 Atakpu and Abrahams (1994) 98 Cr App R 254 376 Attorney-General v Able [1984] 1 QB 795 433 Attorney-General for Hong Kong v Reid [1994] 1 AC 324 385 Attorney-General for Jersey v Holley [2005] 2 AC 580 258, 266 Attorney-General for Northern Ireland v Gallagher [1963] AC 349 159, 194 Attorney-General for Northern Ireland's Reference [1977] AC 105 123, 129 Attorney-General's Reference (No. 1 of 1974) [1974] QB 744 403 Attorney-General's Reference (No. 1 of 1975) [1975] QB 773 422, 430, 433, 447, 448 Attorney-General's References (Nos 1 and 2 of 1979) [1980] QB 180 387 Attorney General's Reference (No. 1 of 1980) [1981] QB 715 322 Attorney-General's Reference (No. 4 of 1980) [1981] 1 WLR 705; (1981) 73 Cr App R 40 104, 159 Attorney-General's Reference (No. 6 of 1980) [1981] QB 715 323 Attorney-General's Reference (No. 2 of 1982) [1984] QB 624 383 Attorney-General's Reference (No. 1 of 1983) [1985] QB 182 385 Attorney-General's Reference (No. 2 of 1983) 123 Attorney-General's Reference (No. 1 of 1985) (1986) 83 Cr App R 70 385 Attorney-General's Reference (No. 1 of 1988) (1989) 89 Cr App R 60 68 Attorney-General's Reference (No. 1 of 1992) (1993) 96 Cr App R 298 463, 475 Attorney General's Reference (No. 2 of 1992) (1993) 97 Cr App R 429 89 Attorney-General's Reference (No. 3 of 1992) (1994) 98 Cr App R 383 460 Attorney-General's Reference (No. 3 of 1994) [1998] AC 245 190, 238, 246 Attorney-General's Reference (No. 1 of 1995) [1996] 2 Cr App R 320 222 Attorney-General's Reference (No. 2 of 1999) [2000] 2 Cr App R 207 60, 151 Attorney General's Reference (No. 4 of 2000) [2001] Crim LR 578 87 Attorney-General's Reference No. 4 of 2002 [2004] UKHL 43 72 Attorney-General's Reference No. 3 of 2003 [2004] EWCA Crim 868 332 Attorney General's Reference (No. 3 of 2004) [2005] EWCA Crim 1882 440 Attorney General's Reference (No. 79 of 2006) (Whitta) [2007] 1 Cr App R (S) 752 355 B (Refusal of treatment), Re [2002] 2 All ER 449 277, 278 B v DPP [2000] 2 AC 428 99, 166, 167, 218, 231, 357, 366 B v K [2002] 1 AC 462 99 (p. xx) Badza [2010] EWCA Crim 1363 444 Bagshaw [1988] Crim LR 321 388 Bailey (1983) 77 Cr App R 76 93, 199, 201 Baillie [1995] Crim LR 739 254, 255 Bainbridge [1960] 1 QB 219 435

Page 2 of 15

Table of cases Baker [1994] Crim LR 444 449 Baker and Ward [1999] 2 Cr App R 335 211 Ball [1989] Crim LR 579 121, 289 Barnes [2005] 1 WLR 910 323 Basherdost [2008] EWCA Crim 2883 346 Bateman (1925) 94 LJKB 791 289 Battams (1979) 1 Cr App R (S) 15 402 Baugh v Crago [1976] Crim LR 72 186 Beasley et al (1987) 9 Cr App R (S) 504 329 Beatty v Gillbanks (1882) 9 QBD 308 125 Becerra and Cooper (1975) 62 Cr App R 212 449 Beckford v R [1988] 1 AC 130 116, 125, 216, 217 Bedder v DPP [1954] 2 All ER 201 259, 260 Benham v United Kingdom (1996) 22 EHRR 293 3 Billam (1986) 82 Cr App R 347 336 Bilton, Daily Telegraph, 20 July 2005 91, 143 Bingham [1991] Crim LR 433 91 Bird [1985] 1 WLR 816 124 Birmingham and Gloucester Railway Co (1842) 3 QB 223 147, 149, 152 Black-Clawson International v Papierwerke Waldhof-Aschaffenber AG [1975] AC 591 68 Blakely and Sutton v Chief Constable of West Mercia [1991] Crim LR 763 201, 432, 434 Bland [1988] Crim LR 41 425 Blaue (1975) 61 Cr App R 271 113 Bloxham [1983] 1 AC 109 403 Bollom [2004] 2 Cr App R 6 245, 311 Bonner (1970) 54 Cr App R 257 383 Bonollo [1981] VR 633 393 Booth v CPS (2006) 170 JP 305 178 Bourne (1952) 36 Cr App R 125 109, 132, 133, 430, 445, 446 Bowen [1996] 2 Cr App R 157 206, 212, 229 Bradbury [1996] Crim LR 808 425 Bratty v Attorney-General for Northern Ireland [1963] AC 386 90 Breaks and Huggan [1998] Crim LR 349 385 Bree [2007] EWCA Crim 804 351 Briggs [2004] 1 Cr App R 34 377 British Steel plc [1995] 1 WLR 1356 148, 164 Broad [1997] Crim LR 666 472 Brook [1993] Crim LR 455 403 Broome v Perkins (1987) 85 Cr App R 321 88, 89 Brown [1985] Crim LR 611 364, 399 Brown [1994] 1 AC 212 318, 320, 322, 324, 326, 327 Brown v United States (1921) 256 US 335 129, 132 Bryce [2004] 2 Cr App R 35 431, 433, 436, 451 Bubbins v UK (2005) 41 EHRR 458 127, 128 Burgess [1991] 2 QB 92 91, 143 Byrne [1960] 2 QB 396 269 C The Times, 9 June 2008 363

Page 3 of 15

Table of cases C v DPP 60 C v Eisenhower [1984] QB 331 311 CR and SW v United Kingdom (1996) 21 EHRR 363 81 Caldwell [1982] AC 341 176, 179–83, 195, 232, 290 Calhaem [1985] QB 808 115, 430, 431 Cambridgeshire and Isle of Ely CC v Rust [1972] 1 QB 426 222 Campbell (1991) 93 Cr App R 350 463, 464 Cantoni v France (1997) VIII HRCD 130 220 Carey [2006] Crim LR 842 289 Carter v Richardson [1974] RTR 314 432 Cartledge v Allen [1973] Crim LR 530 312 Cato [1976] 1 WLR 110; (1976) 62 Cr App R 41 105–107, 296 Celaire and Poulton [2003] 1 Cr App R (S) 610 333 Chambers [2008] EWCA Crim 2467 219 Chan Chi-hung v R [1996] AC 442 61 Chan Man-Sin v Attorney-General for Hong Kong (1988) 86 Cr App R 303 377 Chan Wing-Siu [1985] AC 168 438 Chandler v DPP [1964] AC 763 174, 175 Chan-f*ck (1994) 99 Cr App R 147 311 Charles (1976) 63 Cr App R 252 70 Charlson [1955] 1 WLR 317 89 Chase Manhattan Bank NA v Israel-British Bank [1981] Ch 105 385 Cheshire [1991] 1 WLR 844 106, 111 Chisam (1963) 47 Cr App R 130 123, 216 Chiu-cheung [1995] 1 AC 111 476 Christian v R [2006] 2 AC 400 219 Church [1966] 1 QB 59 158, 159, 288 Churchill v Walton [1967] 2 AC 224 473 Clarence (1888) 22 QBD 23 311, 312, 317, 321 Clark [2003] 2 Cr App R 363 64 Clarke (1972) 56 Cr App R 225 142 Clarke (1982) 75 Cr App R 119 70 Clarke (1984) 80 Cr App R 344 434, 450 Clarkson [1971] 1 WLR 1402 425, 426 Clegg [1996] 1 AC 482 (HL) 265 Clingham v Royal Borough of Kensington and Chelsea [2003] 1 AC 787 43, 53 (p. xxi) Clinton, Parker and Evans [2012] EWCA Crim 2 262 Clouden [1987] Crim LR 56 396, 397 co*ckburn (1968) 52 Cr App R 134 386 co*cker [1989] Crim LR 740 (CA) 270, 281 Cogan and Leak [1976] 1 QB 217 109, 346, 354, 445, 447 Collins [1973] QB 100 399, 400 Collins v Willco*ck (1984) 79 Cr App R 229 317, 318 Conally v General Construction Co (1926) 269 US 385 65 Concannon [2002] Crim LR 215 421 Coney (1882) 8 QBD 534 424, 425 Constanza [1997] 2 Cr App R 492 318, 319

Page 4 of 15

Table of cases Conway [1989] 3 All ER 1025; (1988) 88 Cr App R 159 205, 206 Cook v Atchison [1968] Crim LR 266 87 Cooper v Simmons (1862) 7 H and N 707 222 Coppen v Moore (No. 2) [1898] 2 QB 306 149 Corbett [1996] Crim LR 594 112 Corcoran v Whent [1977] Crim LR 52 376 Cotter [2002] Crim LR 824 64 Court [1989] AC 28 344 Courtie [1984] AC 463 398 Cousins [1982] QB 526 123 Craig [2007] EWCA Crim 2913 404 Crossman (1986) 82 Cr App R 333 301 Cumming [2007] 2 Cr App R (S) 20 277 Cunningham [1957] 2 QB 396 176, 186 Cunningham [1982] AC 566 245, 246 Dadson (1850) 4 Cox CC 358 122 Dagnall [2003] EWCA Crim 2441 350 Davies [2004] 1 Cr App R (S) 209 398 Dawson (1985) 81 Cr App R 150 289 Dawson and James (1976) 64 Cr App R 150 396 Dear [1996] Crim LR 595 113 Densu [1998] 1 Cr App R 400 333 Devonald [2008] EWCA Crim 527 346, 347 Deyemi and Edwards [2008] 1 Cr App R 25 99, 167, 168 Dhaliwal [2006] 2 Cr App R 348 287, 311, 314 Dica [2004] QB 1257 321 Dietschmann [2003] 1 AC 1209 271, 321 Donnelly v Jackman (1969) 54 Cr App R 229 318 Donovan [1934] 2 KB 498 322 Doughty (1986) 83 Cr App R 319 256, 257 Dowds [2012] EWCA Crim 281 271, 272 DPP v Camplin [1978] AC 705 (HL) 259, 260 DPP v Collins [2007] 1 Cr App R 5 32 DPP v Dunn [2001] 1 Cr App R 352 328 DPP v H [1997] 1 WLR 1406 143, 161 DPP v Hammond [2004] Crim LR 851 32 DPP v K (1990) 91 Cr App R 23 317 DPP v K and B [1997] 1 Cr App R 36 445, 446 DPP v Kent and Sussex Contractors Ltd [1944] KB 146 150 DPP v Lavender [1994] Crim LR 297 388 DPP v Little (1992) 95 Cr App R 28 316 DPP v Lynch [1977] AC 653 212 DPP v Majewski [1977] AC 443 92, 195, 197, 198 DPP v Morgan [1976] 1 AC 182 195, 215, 342, 354 DPP v Newbury and Jones [1977] AC 500 288 DPP v Redmond-Bate [1999] Crim LR 998 32 DPP v Rogers [1998] Crim LR 202 208

Page 5 of 15

Table of cases DPP v Santana-Bermudez [2004] Crim LR 471 317 DPP v Smith [1961] AC 290 311 DPP v Withers [1975] AC 842 468 DPP for Northern Ireland v Lynch [1975] AC 653 206 DPP for Northern Ireland v Maxwell [1978] 3 All ER 1140 471 Du Cros v Lambourne [1907] 1 KB 40 427 Dudgeon v United Kingdom (1982) 4 EHRR 149 326, 338 Dudley and Stephens (1884) 14 QBD 273 131–3, 212, 213 Duffy [1949] 1 All ER 932 253 Durante [1972] 3 All ER 962 195 Dyke and Munro [2002] 1 Cr App R 30; [2002] Crim LR 153 382, 384 Dytham [1979] QB 722, 332 E.B. [2006] EWCA Crim 2945 351 Easom [1971] 2 QB 315 387 Edwards v Ddin (1976) 63 Cr App R 218 376 Edwards v R [1973] AC 648 159 Elbekkay [1995] Crim LR 163 61, 348 El-Kurd [2001] Crim LR 234 473 Elliott v C (1983) 77 Cr App R 103 180 Emery (1993) 14 Cr App R (S) 394 102, 422 Emmett (1999) The Times 15 October. 326 Engel v Netherlands (1976) 1 EHRR 647 3 English [1999] AC 1 440, 443, 444 Environment Agency v Empress Car Co (Abertillery) [1999] 2 AC 22 110, 149 Evans [2009] EWCA Crim 650 292, 293 Evans v Hughes [1972] 3 All ER 412 123, 333 F, Re [1990] 2 AC 1 133 fa*gan v Metropolitan Police Commissioner [1969] 1 QB 439 157, 158 Fallon [1994] Crim LR 519 310, 312 Feely [1973] QB 530 390, 393 (p. xxii) Fennell [1971] 1 QB 428 121 Ferguson v Weaving [1951] 1 KB 814 423 Fernandes [1996] 1 Cr App R 175 387 Field [1972] Crim LR 435 124 Finau v Department of Labour [1984] 2 NZLR 396 96 Finney (1874) 12 Cox CC 625 289 Flattery (1877) 2 QBD 410 347 Floyd v DPP [2000] Crim LR 411 385 Forbes and Webb (1865) 10 Cox CC 362 315 Forrester [1992] Crim LR 792 396 Fotheringham (1989) 88 Cr App R 206 196 Fowler v Padget (1798) 7 Term Rep 509 157 G [2004] 1 AC 1034 176, 179–81, 188 G [2008] UKHL 37 73, 167, 357 Gallasso (1994) 98 Cr App R 284 375 Gammon v Attorney-General for Hong Kong [1985] AC 1 166 Gateway Foodmarkets Ltd [1997] Crim LR 512 148, 164

Page 6 of 15

Table of cases Geddes [1996] Crim LR 894 463, 464 George [1956] Crim LR 52 345 Ghosh [1982] QB 1053 391, 393, 394, 407 Giannetto [1997] 1 Cr App R 1 421, 423, 451 Gibbins and Proctor (1918) 13 Cr App R 134 332 Gibson [1990] 2 QB 619 469 Gilks (1972) 56 Cr App R 734 385 Gillick v West Norfolk and Wisbech Area Health Authority [1986] AC 112 133, 174–6, 428, 434 Gilmartin (1983) 76 Cr App R 238 406 Gilmour [2000] 2 Cr App R 407 440 Giorgianni v R (1985) 156 CLR 473 432 Gladstone Williams (1984) 78 Cr App R 276 216 Gnango [2011] 1 All ER 153 443 Gnosil (1824) 1 C & P 304 397 Gold and Schifreen [1988] AC 1063 70, 381 Goldstein and Rimmington [2005] UKHL 63 51, 81 Gomez [1993] AC 442 64, 68, 374–9, 383, 384, 393, 413 Goodyear [2005] 3 All ER 117 14 Gotts [1992] 2 AC 412 212 Graham (1982) 74 Cr App R 235 206, 208, 214, 218 Grant v Borg [1982] 1 WLR 638 219, 222 Great North of England Railway Co (1846) 9 QB 315 147 Grenwal [2010] EWCA Crim 2448 356 Gül v Turkey (2002) 34 EHRR 719 121 Gullefer [1987] Crim LR 195 463 H [2005] Crim LR 734 316, 344, 345 Hale (1978) 68 Cr App R 415 158, 376, 396 Hall [1973] QB 126 385 Hanco*ck and Shankland [1986] AC 455 172, 244 Hardie (1985) 80 Cr App R 157 200, 201 Harman's Case (1620) 2 Roll Rep 154 397 Harrow LBC v Shah [2000] 1 WLR 83 166 Hasan [2005] 2 AC 467; [2005] UKHL 22 61, 160, 208, 209, 213, 218, 229, 230, 355 Hashman and Harrup v UK (2000) 30 EHRR 241 63–5, 393 Hatton [2006] 1 Cr App R 16 196 Heard [2007] EWCA Crim 125; [2008] QB 43 195, 204, 356 Hennessy (1989) 89 Cr App R 10 91, 143, 161 Hill (1986) 83 Cr App R 386 331 Hill [1997] Crim LR 459 163 Hilton [1997] 2 Cr App R 445 383 Hinks [2000] 4 All ER 833; [2001] 2 AC 241 64, 68, 321, 375, 380, 393, 413 Hobson [1998] 1 Cr App R 31 270 Horncastle [2006] EWCA Crim 1736 399 Howe [1987] AC 417 61, 130, 212, 226, 233, 423, 444 Hudson and Taylor [1971] 2 QB 202 205, 207, 208, 210 Hughes v Holley (1988) 86 Cr App R 130 64

Page 7 of 15

Table of cases Hyam v DPP [1975] AC 55 175, 247 Hysa [2007] EWCA Crim 2056 352 ICR Haulage Ltd [1944] KB 551 150 Ibrams (1981) 74 Cr App R 154 255 Idress [2011] EWHC 624 (Admin) 406 Impress (Worcester) Ltd v Rees [1971] 2 All ER 357 110 Ireland and Burstow [1998] AC 147 311, 312, 317–19, 328, 335 Isaveya v Russia (2005) 41 EHRR 791 127 Isitt (1978) 67 Cr App R 44 88 Ismail [2005] EWCA Crim 397 342 JF Alford (Transport) Ltd [1997] 2 Cr App R 326 151, 427, 432 Janjua [1999] 1 Cr App R 91 311 Jackson [1985] Crim LR 442 474 Jacobson v US (1992) 112 S Ct 1535 224 James [2006] QB 588 258 James & Son v Smee [1955] 1 QB 78 150, 186 Janjua [1999] 1 Cr App R 91 245, 311 Jefferson et al (1994) 99 Cr App R 13 330, 420, 421, 425 Jheeta [2007] EWCA Crim 1699 347, 348 Jobidon [1991] 2 SCR 714 323 Johnson (1989) 89 Cr App 148 159 Johnson [2008] Crim LR 132 143 Johnson v Youden [1950] 1 KB 544 432, 435 (p. xxiii) Jones (1986) 83 Cr App R 375 324 Jones (1990) 91 Cr App R 351 463, 464 Jones, The Times, 19 August 1994 219 Jones et al. (1974) 59 Cr App R 120 470 Jones v Bedford and Mid-Befordshire Magistrates’ Court [2010] EWHC 523 315 Jones, Milling et al [2007] 1 AC 136 60, 120, 123, 125, 131, 134 Jordan (1956) 40 Cr App R 152 111 Julien [1969] 1 WLR 839 124 Juozaitiene v Lithuania (2008) 47 EHRR 1194 127 K [2002] 2 AC 462; [2002] 1 Cr App R 121 167, 218, 231, 357, 366 K.A. and A.D. v Belgium (judgment of 17 February 2005, App No. 42758/98) 327 Kaitamaki v R [1985] 1 AC 147 158 Kamara v DPP [1973] 2 All ER 1242 468, 469 Kamipeli [1975] 2 NZLR 610 198 Kaur v Chief Constable of Hampshire [1981] 1 WLR 578 378 Kay v Butterworth (1945) 173 LT 191 93 Kelly (1992) 97 Cr App R 245 401 Kelly v UK (1993) 74 DR 139 124 Kelly and Donnelly [2001] 1 Cr App R (S) 341 315, 328 Kemp and Else [1964] 2 QB 341 447 Kendrick and Hopkins [1997] 2 Cr App R 524 375 Kennedy [1999] Crim LR 65 110 Kennedy (No 2) [2005] 2 Cr App R 23; [2007] UKHL 38; [2008] 1 AC 269 103, 104, 110, 288, 293, 296, 331

Page 8 of 15

Table of cases Keys and Sween (1986) 8 Cr App R (S) 444 329 Khaliq v H M Advocate 1983 SCCR 483 58 Khan (1990) 91 Cr App R 29 459, 475 Khan and Khan [1998] Crim LR 830 288, 289 Kimber [1983] 1 WLR 1118; (1983) 77 Cr App R 225 176, 216 Kingston [1995] 2 AC 355 97, 201, 223, 234 Klineberg and Marsden [1999] 1 Cr App R 427 385 Knuller v DPP [1973] AC 435 58, 59, 62, 468, 469 Kohn [1997] 2 Cr App R 445 381 Kokkinakis v Greece (1994) 17 EHRR 397 61, 63, 68 Kolender v Lawson (1983) 103 S Ct 1855 62 Kong Cheuk Kwan (1986) 82 Cr App R 18 290 Konzani [2005] 2 Cr App R 13 321 Korbeley v Hungary (2010) 50 EHRR 1192 62 Lamb [1967] 2 QB 981 286, 290, 319 Lambert [2001] 3 WLR 206; [2002] 2 AC 545 50, 71 Lambert v California (1957) 355 US 225 221 Lane and Lane (1986) 82 Cr App R 5 422 Lankford [1959] Crim LR 209, 484 Larsonneur (1933) 149 LT 542 96 Laskey v UK (1997) 24 EHRR 39 327, 364 Latimer (1886) 17 QBD 359 190 Lawrence [1972] AC 626 373, 374, 379 Lawrence [1982] AC 510 176 Le Brun [1992] QB 61 158, 159 Le Compte, Van Leuven and De Meyere v Belgium (1981) 4 EHRR 1 4 Lester (1976) 63 Cr App R 144 163 Letenock (1917) 12 Cr App R 221 216, 257 Lewis (1988) 87 Cr App R 270 98 Lidar [2000] 4 Archbold News 3 294 Liggins et al. [1995] Crim LR 45 469, 470 Lim Chin Aik v R [1963] AC 160 166, 219 Linekar [1995] 2 Cr App R 49; [1995] QB 250 347, 353 Lipman [1970] 1 QB 152 92, 196, 287 Lloyd [1985] QB 928 388 Logdon v DPP [1976] Crim LR 121 319 Lomas (1913) 9 Cr App R 220 429 Looseley; Attorney-General's Reference No. 3 of 1999 [2002] 1 Cr App R 29 224, 228 Lowe [1973] QB 702 287 Luc Thiet-Thuan v R [1997] AC 131 266 Lynch v DPP [2002] Crim LR 320 488 MC v Bulgaria (2005) 40 EHRR 459 338 McCann and others v UK (1996) 21 EHRR 97 126, 127, 217, 231 McCoy [1953] 2 SA 4 354 McDermott [2007] 1 Cr App R (S) 145 200 M'Naghten's Case (1843) 10 Cl & Fin 200 142, 143 McNamara (1988) 87 Cr App R 246 98

Page 9 of 15

Table of cases McShane (1978) 66 Cr App R 97 277 Milberry [2003] 2 Cr App R (S) 142 336 Miller [1983] 2 AC 161 290, 291 Morrison (1989) 89 Cr App R 17 312 Moses [1991] Crim LR 617 411 Mahroof [1989] Crim LR 72 329 Majewski See DPP v Majewski Malcherek and Steel (1981) 73 Cr App R 173 111, 239 Mandair [1995] 1 AC 208 312 Marcus (1981) 73 Cr App R 49 331 Marshall [1998] 2 Cr App R 282 384, 388 Marshall [2009] EWCA Crim 2076 408 Martin (1881) 8 QBD 54 312 Martin (1989) 88 Cr App R 343 205 (p. xxiv) Martin [2002] 1 Cr App R 27 127 Matthews and Alleyne [2003] 2 Cr App R 30 172, 245 Matudi [2003] EWCA Crim 697 167 Mavji (1987) 84 Cr App R 34 405 Mawgridge (1707) 1 Kel 119 258 Maxwell v DPP for Northern Ireland [1978] 3 All ER 1140; [1979] 1 WLR 1350 421, 435, 436, 439, 451, 483 Mazo [1997] 2 Cr App R 518 375 Meachen [2006] EWCA Crim 2414 322 Meade and Belt (1823) 1 Lew CC 184 318 Melliti [2001] EWCA Crim 1563 348 Mendez [2010] EWCA Crim 516 444 Mercer [2001] All ER (D) 187 421 Meridian Global Funds Management Asia Ltd v Securities Commission [1995] 2 AC 500 151 Metropolitan Police Commissioner v Charles [1977] AC 177 353 Michael (1840) 9 C and P 356 109, 447 Miller [1954] 2 QB 282 314 Miller [1983] 2 AC 161 100, 158, 317 Millward [1994] Crim LR 527 446 Misra [2005] 1 Cr App R 21 293 Mitchell [2008] EWCA Crim 850 386, 387, 395 Mohan [1976] 1 QB 1 169, 170, 458 Moloney [1985] AC 905 172, 180, 244, 252 Moore v I Bresler Ltd [1944] 2 All ER 515 150 Morgan See DPP v Morgan Morhall (1994) 98 Cr App R 108; [1996] 1 AC 90 260 Morris [1984] AC 320 374, 377, 379, 412 Morris [1998] 1 Cr App R 386 328 Mousell Bros v London and North-Western Railway Co [1917] 2 KB 836 147 Mowatt [1968] 1 QB 421 312, 313 Muhamad [2003] QB 1031 166, 167 Mulcahy (1868) LR 3 HL 306 472 NW et al. [2008] 3 All ER 533 404

Page 10 of 15

Table of cases Nachova v Bulgaria (2006) 42 EHRR 933 56, 120–2 Nash [1999] Crim LR 308 463 National Coal Board v Gamble [1959] 1 QB 11 428, 429, 433, 434 National Rivers Authority v Yorkshire Water Services [1995] 1 AC 444 105 Naylor (1987) 84 Crim App R 302 256 Nedrick (1986) 83 Cr App R 267 172, 244 Newbury and Jones [1977] AC 500 286 Newington [1990] Crim LR 593 332 Nicol and Selvanayagam v DPP [1996] Crim LR 318 125 Norris v United States of America [2008] UKHL 16 66, 411 Nunn [1996] 2 Cr App R (S) 136 3 O'Connor (1980) 146 CLR 64 198 O'Connor [1991] Crim LR 135 196 O'Driscoll (1977) 65 Cr App R 50 287 O'Grady (1987) 85 Cr App R 315 196 Osborne [2010] EWCA 547 269 Owino [1996] 2 Cr App R 128 120 Oxford v Moss (1979) 68 Cr App R 183 70, 381 Pagett (1983) 76 Cr App R 279 109 Palmer [1971] AC 814, 128, 129 Parker [1977] 1 WLR 600 178 Pattni et al [2001] Crim LR 570 64 Pearman (1985) 80 Cr App R 259 458 Pembliton (1874) 12 Cox CC 607 190 People v Goetz (1986) 68 NY 2d 96 120 Pepper v Hart [1993] AC 593 68 Percy v DPP [2002] Crim LR 835 51 Perka v R (1984) 13 DLR (4th) 1 210 Philippou (1989) 89 Cr App R 290 383 Piper [2007] EWCA Crim 2131 347, 348 Pitham and Hehl (1977) 65 Cr App R 45 377 Postermobile v Brent LBC [1998] Crim LR 435 223, 228 Powell and Daniels [1999] AC 1 439, 451, 452 Powell v Texas 392 US 514 (1968) 97 Practice Direction [1977] 2 All ER 540 470 Practice Direction (Crime: Mandatory Life Sentences) (No. 2) [2004] 1 WLR 2551 241 Practice Direction: Crown Court (Trial of Children and Young Persons) [2000] 1 Cr App R 483 140 Preddy [1996] AC 815 70, 385 Prentice [1994] QB 302 290, 293 Price (1990) 90 Cr App R 409 391 Quick [1973] QB 910 89, 91–4 R [2008] EWCA Crim 619 489 RJ and M-J D v France [2005] Crim LR 307 118 R v ABCD [2010] EWCA Crim 1622 437 R v DPP, ex parte Jones [2000] Crim LR 858 291 R v R [1992] 1 AC 599 58

Page 11 of 15

Table of cases R v S &H [2011] EWCA Crim 2872 481 R v Wells Street Magistrates’ Court and Martin, ex p Westminster City Council [1986] Crim LR 695 166 R v West London Coroner, ex p Gray [1988] QB 467 290 (p. xxv) R (Bennett) v HM Coroner for Inner London [2006] EWHC Admin 196 127 R (McCann) v Manchester Crown Court; Clingham v Kensington and Chelsea LBC [2003] 1 AC 787 4 R (on application of W) v DPP [2005] EWHC Admin 1333 221 R (on the application of Anderson) v Secretary of State for the Home Department [2003] 1 AC 837 240 R (on the Application of Purdy) v DPP [2009] UKHL 45 278, 281 R (on the application of T) v DPP [2003] Crim LR 622 314 Rabey (1978) 79 DLR (3d) 414 92 Rahman [2008] UKHL 45 442–4 Raphael [2008] Crim LR 995 387 Rashford [2006] Crim LR 546 121 Redford (1989) 89 Cr App R 1 405 Redmond-Bate v DPP [1999] Crim LR 998 125, 330 Reed [1982] Crim LR 819 474 References re Section 94(2) of the Motor Vehicles Act (1986) 48 CR (3d) 289 165 Richards [1974] QB 776 423, 430, 444 Richardson [1998] 2 Cr App R 200 321, 323, 353 Richardson and Irwin [1999] 1 Cr App R 392 196, 324 Rimmington and Goldstein [2005] UKHL 63; [2006] 1 AC 459 58, 60, 64, 221 Rivas v France [2005] Crim LR 305 50 Roberts (1972) 56 Cr App R 95 112, 113 Roberts (1987) 84 Cr App R 117 391 Roberts and George [1997] Crim LR 209 432 Robinson [1977] Crim LR 173 396 Robinson v California 370 US 660 (1962) 97 Rodger and Rose [1998] 1 Cr App R 143 205, 207 Rook [1993] 2 All ER 955 424 Roper v Taylor's Garages Ltd [1951] 2 TLR 284 180 Rose (1884) 15 Cox CC 540 216 Royle [2005] 2 Cr App R (S) 480 363 Rushworth (1992) 95 Cr App R 252 314 Russell and Russell (1987) 85 Cr App R 388 422 S C v United Kingdom (2005) 40 EHRR 226 140 SW and CR v United Kingdom (1995) 21 EHRR 363 59 S v Robinson 1968 (1) SA 666 443 Saik [2007] 1 AC 18 474, 475 Salabiaku v France (1988) 13 EHRR 379 71 Salisbury [1976] VR 452 312 Salomon v Salomon [1897] AC 22 147 Salvo [1980] VR 401 391 Sandhu [1997] Crim LR 288 163 Sandie Smith [1982] Crim LR 531 229

Page 12 of 15

Table of cases Santana-Bermudez [2004] Crim LR 471 321 Satnam S and Kewal S (1984) 78 Cr App R 149 176 Saunders and Archer (1573) 2 Plowd 473 441, 448 Savage and Parmenter [1992] 1 AC 699 70, 186, 312–14, 317, 319 Scarlett (1994) 98 Cr App R 290 120, 287 Schwartz (1979) 29 CCC (2d) 1 143 Scott v Metropolitan Police Commissioner [1975] AC 819 410, 469 Seaboard Off shore Ltd v Secretary of State for Transport [1994] 1 WLR 541 148, 164 Secretary of State for Trade and Industry v Hart [1982] 1 WLR 481 222 Senior [1899] 1 QB 283 287 Seymour [1983] 1 AC 624 290 Shannon [2000] 1 Cr App R 168 225 Shannon v UK [2005] Crim LR 133 225 Sharp [1987] QB 853 208, 209 Shaw v DPP [1962] AC 220 57–60, 62, 468, 469 Shaw v R [2002] 1 Cr App R 10 127 Shayler [2001] 1 WLR 2206 134, 207 Sheehan and O'Mahoney [2007] 1 Cr App R (S) 149 200 Sheldrake v DPP; Attorney General's Reference No. 4 of 2002 [2004] UKHL 43 50, 72, 73 Shepherd (1988) 86 Cr App R 47 209 Sheppard [1981] AC 394 186 Shivpuri [1987] AC 1 466 Silverman (1988) 86 Cr App R 213 408 Siracusa (1990) 90 Cr App R 340 473, 476 Smith (Morgan) [2001] 1 AC 146 229 Smith [1960] 2 QB 423 111, 174, 175, 450 Smith [2011] EWCA Crim 66 382 Smith and Jones (1976) 63 Cr App R 47 399, 400 Smith and Smith [1986] Crim LR 166 387 Smith v Chief Superintendent of Woking Police Station (1983) 76 Cr App R 234 318 Smith v Reynolds et al. [1986] Crim LR 559 425 Speck (1977) 65 Cr App R 161 100 Stafford v United Kingdom (2002) 35 EHRR 1121 240 Stallard v HM Advocate 1989 SCCR 248 58 Stapleton v R (1952) 86 CLR 358 143 State v Bristol (1938) 53 Wyo 304 124 State v Frazer (1936) 98 SW (2d) 707 112 (p. xxvi) State v Tally (1894) 15 So 722 424 Steane [1947] KB 997 174–6, 206 Steel v UK (1999) 28 EHRR 603 64 Stephenson [1979] QB 695 177, 178 Stewart and Schofield [1995] 1 Cr App R 441 440 Stone and Dobinson [1977] QB 354 290, 291 Stripp (1979) 69 Cr App R 318 93 Sullivan [1984] AC 156 91, 142, 143 Sullivan [2005] 1 Cr App R (S) 308 241 Sullivan and Ballion [2002] Crim LR 758 382

Page 13 of 15

Table of cases Sunday Times v UK (1979) 2 EHRR 245 63 Sutherland v UK (1997) 24 EHRR CD22 338, 340, 358 Sweet v Parsley [1970] AC 132 161, 164–6, 215 T [1990] Crim LR 256 92 T.S. [2008] CLW 08/07/1 355 Taaffe [1984] AC 539 466 Tabassum [2000] 2 Cr App R 238 321, 347, 353 Tan [1983] QB 1053 59, 62 Tandy [1989] 1 All ER 267 270 Taylor [1950] 2 KB 368 67 Taylor, Harrison and Taylor [1998] Crim LR 582 (CA) 421 Teixeira de Castro v Portugal (1999) 28 EHRR 101 224, 225 Tesco Supermarkets v Nattrass [1972] AC 153 150–2 Thabo Meli v R [1954] 1 WLR 228 158, 159 Thain [1985] NI 457 122 Thatcher v R (1987) 39 DLR (3d) 275 421 Thomas (1985) 81 Cr App R 331 316 Thomas [2006] Crim LR 71 221 Thomson (1965) 50 Cr App R 1 472 Thorne v Motor Trade Association [1937] AC 797 397 Thornton [1992] 1 All ER 306 254 Thornton v Mitchell [1940] 1 All ER 339 446 Tifa*ga v Department of Labour [1980] 2 NZLR 235 96 Tosti [1997] Crim LR 746 463, 464 Tuck v Robson [1970] 1 WLR 741 427 Tudor [1999] 1 Cr App R (S) 197 120 Turner (No. 2) (1971) 55 Cr App R 336 382 Tyrell [1894] 1 QB 710 449 Uddin The Times, 14 September 1929 263 United States v Holmes (1842) 26 Fed Cas 360 131 V and T v United Kingdom (1999) 30 EHRR 121 140 Valderrama-Vega [1985] Crim LR 220 206, 207 Vane v Yiannopoullos [1965] AC 486 149 Velumyl [1989] Crim LR 299 386 Venna [1976] QB 421 319 Vo v France (2005) 40 EHRR 259 238 Wacker [2003] QB 1203 291 Wai Yu-tsang [1992] 1 AC 269 411 Walkington (1979) 68 Cr App R 427 399 Wardrope [1960] Crim LR 770 216 Warner v Metropolitan Police Commissioner [1969] 2 AC 256 98, 99 Watson [1989] 1 WLR 684 289 Webbe [2002] 1 Cr App R (S) 82 402 Webster [2006] 2 Cr App R 6 433 Weller [2003] Crim LR 724 (CA) 266 Westminster City Council v Croyalgrange Ltd (1986) 83 Cr App R 155 180 Wheatley v Commissioner of Police of the British Virgin Islands [2006] 1 WLR 1683 375

Page 14 of 15

Table of cases White [2010] EWCA Crim 1929 349 Whitefield (1984) 79 Cr App R 36 448 Whitehouse [1977] QB 868 449 Whybrow (1951) 35 Cr App R 141 458, 459 Wilcox v Jeff ery [1951] 1 All ER 464 423, 425 Willer (1986) 83 Cr App R 225 61, 205 Williams [1923] 1 KB 340 347 Willoughby [2005] Crim LR 389 291 Wilson [1984] AC 242 317, 318 Wilson [1996] 2 Cr App R 241 326, 327 Wilson [2007] QB 960 214 Wilson, Jenkins [1984] AC 242 312 Windle [1952] 2 QB 826 143 Winship, Re (1970) 397 US 358 71 Winterwerp v Netherlands (1979) 2 EHRR 387 144 Winzar v Chief Constable of Kent The Times, 28 March 1983 96 Wood [2002] EWCA Crim 832 382 Wood [2008] 2 Cr App R 34 271 Woods (1981) 74 Cr App Rep 312 198 Woollin [1999] 1 AC 82 172, 175, 180, 244, 245, 458 Woolmington v DPP [1935] AC 462 71 X and Y v Netherlands (1986) 8 EHRR 235 338 Yip Chiu-Cheung [1995] 1 AC 111 174, 450 Zecevic v R (1987) 162 CLR 645 217

Page 15 of 15 Subscriber: Oxford University Press - Master Gratis Access; date: 17 January 2015

Table of Treaties and Conventions

Principles of Criminal Law (7th edn) Andrew Ashworth and Jeremy Horder Publisher: Oxford University Press Print ISBN-13: 9780199672684 DOI: 10.1093/he/9780199672684.001.0001

Print Publication Date: May 2013 Published online: Sep 2013

Table of Treaties and Conventions Amsterdam Treaty 1999 48 European Convention on Human Rights 1950 8, 17, 27 Art 2 31, 48, 50, 55, 56, 109, 117, 118, 121, 126, 127, 129, 130, 136, 238 Art 3 31, 48, 50, 118, 136, 140, 278, 309 Art 4 48, 51 Art 4(1) 50 Art 5 49, 51, 63, 118, 125, 145, 309 Art 5(1)(e) 144 Art 6 43, 51, 140, 224, 240 Art 6(2) 49, 50, 71, 73, 142, 167, 488 Art 6(3) 52, 357 Art 6(3)(a) 421 Art 7 49–51, 57, 59–65, 68, 81, 220, 293, 412, 481 Art 7(1) 59 Art 8 31, 49, 51, 167, 278, 309, 326, 327, 357, 362, 365, 471 Art 8(1) 326 Art 8(2) 278

Table of Treaties and Conventions Arts 8–11 63 Art 9 49, 51, 81 Art 10 49, 51, 81, 125, 471 Art 11 49, 51, 331, 425, 471 Art 14 51 Art 15 51, 62 Protocol 1 51 Protocol 6 117, 240 EU Constitutional Treaty 48 Maastricht Treaty 1992 48

Table of legislation

Principles of Criminal Law (7th edn) Andrew Ashworth and Jeremy Horder Publisher: Oxford University Press Print ISBN-13: 9780199672684 DOI: 10.1093/he/9780199672684.001.0001

Print Publication Date: May 2013 Published online: Sep 2013

Table of legislation

United Kingdom Abortion Act 1967 132 Accessories and Abettors Act 1861 419–21 Aggravated Vehicle-Taking Act 1992 394 s 1 300 Anti-Terrorism, Crime and Security Act 2001 s 39 315, 328 s 50 451 Banking Act 1987 405 s 35 370 Bribery Act 2010 s 1(2) 409

Table of legislation Children Act 1989 295 Children and Young Persons Act 1933 s 1 54, 332 Companies Act 1985 s 458 405 Companies Act 1989 s 41 383 Companies Acts 1985–1989 146 Computer Misuse Act 1990 381 Conspiracy and Protection of Property Act 1875 468 Consumer Protection Act 1987 s 12 333 Control of Pollution Act 1974 s 3(4)(a) 222 Coroners and Justice Act 2009 207, 252 s 52 268, 272 s 52(1)(a) 270 s 54 253 s 54(1) 253, 264 s 54(1)(c) 266, 267 s 54(2) 254, 264 s 54(3) 265, 266 s 54(4) 254, 255 s 55 253 s 55(3) 213, 254 s 54(6) 256, 263 s 54(6)(a) 254 s 54(6)(b) 254 s 55(2) 257 s 55(3) 262, 263 s 55(4) 258, 259, 262 s 55(4)(a) 259–61, 263 s 55(4)(b) 257 s 55(6)(c) 261, 262 s 56(6) 159 s 59 277 Corporate Manslaughter and Corporate Homicide Act 2007 47, 153, 155, 294, 483 s 1 154 s 1(2) 295 s 1(4) 154, 296 s 2(2) 295

Page 2 of 12

Table of legislation ss 3–7 295 s 8 154, 304 Sch 1 295 Crime and Disorder Act 1998 11 s 1 5, 43, 53 s 28 315 s 28(1)(a) 315 s 28(1)(b) 315 s 29 315 ss 29–32 30 s 32 328 s 34 140 s 36 240 Criminal Attempts Act 1981 458 s 1 459 s 1(1) 170, 458, 462 s 1(2) 466 s 1(3) 466 s 4(3) 462 s 5 469, 474 Criminal Damage Act 1971 79, 181, 188, 190, 320, 458 s 1(2) 333, 460, 461 s 5(2) 120, 218 Criminal Justice Act 1967 s 8 198 Criminal Justice Act 1988 s 36 284 s 39 316, 317 s 93(2) 475 s 134 332 ss 139–141 333 Criminal Justice Act 1991 398 s 26(1) 370 Criminal Justice Act 2003 12, 301 s 142(1) 19 s 143(1) 20 s 144 14 (p. xxix) s 152(2) 18 s 153(2) 18 s 269 240, 241, 275

Page 3 of 12

Table of legislation s 269(3) 241 Sch 3 14 Sch 21 241, 242, 275, 440 Criminal Justice and Immigration Act 2008 128 s 73 360, 362 s 76 49, 119, 120, 123, 130, 136, 196, 197 s 76(2) 119 s 76(3) 218 s 76(4) 196, 257 s 76(5) 196 s 76(7) 217 s 76(9) 119 Sch 15 360, 362 Criminal Justice and Public Order Act 1994 17, 329, 465 Pt I (ss 1–24) 469 Pt II (ss 25–30) 401 ss 34–37 11 Pt V (ss 61–80) 41 s 68 469 s 70 469 s 77 469 s 142 61, 340 Criminal Law Act 1967 s 3 119 s 4 95, 486 Criminal Law Act 1977 Pt I (ss 1–5) 469 s 1(1) 473, 474, 476 s 1(2) 475 s 2(2)(a) 472 s 2(2)(b) 472 s 2(2)(c) 473 s 4(1) 473 s 5 469 Pt II (ss 6–13) 401 s 7 402 Criminal Procedure (Insanity) Act 1964 s 6 142 Criminal Procedure (Insanity and Unfitness to Plead) Act 1991 90, 91, 94, 141, 143, 144 s 5 145

Page 4 of 12

Table of legislation Domestic Violence, Crime and Victims Act 2004 54, 422 s 5 237, 300, 425 s 24 145, 267 Education Act 2005 34 Energy Act 2011 34 Enterprise Act 2002 405 Financial Services and Markets Act 2000 146, 405 s 397 370 Firearms Act 1968 332, 333 s 5 167 Forgery and Counterfeiting Act 1981 405 Fraud Act 2006 64, 80, 108, 368–70, 392, 454, 486, 489 s 1 405 s 1(2)(b) 407 s 1(2)(c) 407, 409 s 2 405, 407, 408, 410 s 2(1) 405 s 2(2)(b) 407 s 2(2)–(5) 406 s 3 407, 408 s 4(2) 409 s 5 407 s 11 410 Health and Safety at Work Act 1974 153, 294, 297, 333 s 2(1) 148 s 3 148, 149 s 3(1) 148 s 5(2)(a) 406 Health and Safety (Offences) Act 2008 333 Homicide Act 1957 s 2 145, 268 s 3 252, 253, 255, 258, 264 s 4 274–6 Human Rights Act 1998 3, 27, 48, 117 s 2 8, 49 s 3 8, 50, 68, 71 s 4 50 s 6 8, 59 s 10 50 s 12 51 s 13 51

Page 5 of 12

Table of legislation s 19 49 Incitement to Disaffection Act 1934 222 Indecent Displays (Control) Act 1981 38 Infant Life Preservation Act 1929 s 1 239 Knives Act 1997 333 Law Reform (Year and a Day Rule) Act 1996 s 2 239 Legal Aid, Sentencing and Punishment Act 2012 469 s 144 401 s 148 119 Licensing Act 2003 168 s 136 149 (p. xxx) Magistrates’ Courts Act 1980 32 s 20 14 s 101 72 Mental Capacity Act 2005 s 4 101 Mental Health Act 1983 141, 332 s 37 145 Mental Health Act 2007 141 Misuse of Drugs Act 1971 134 s 5(3) 97 s 5(4) 97 s 5(4)(b) 488 Obscene Publications Act 1959 72 Offences Against the Person Act 1861 67, 186, 309, 330, 334 s 16 486 s 18 168, 310–13, 332 s 20 169, 311–13, 315, 324 s 22 331 s 23 288, 331 s 24 331 s 27 332 s 32 333 s 33 333 s 38 314 s 47 138, 314, 315, 317

Page 6 of 12

Table of legislation Offensive Weapons Act 1996 333 Police Act 1996 s 89 315 Police and Criminal Evidence Act 1984 10, 12 s 24(3)(b) 462 s 24(4)(b) 462 s 24(7)(b) 462 s 80 425, 472 Codes of Practice 10, 12 Powers of Criminal Courts (Sentencing) Act 2000 s 128 18 Prevention of Crime Act 1953 97, 333 s 1 488 Prevention of Terrorism Act 2005 17, 51 Proceeds of Crime Act 2002 475 ss 327–329 404 Prosecution of Offences Act 1985 2 Protection from Eviction Act 1977 401 Protection from Harassment Act 1997 335 s 1 134, 328 s 2 328 s 4 134, 328 Protection of Children Act 1978 53 Protection of Freedoms Act 2012 s 111 53 Public Order Act 1986 329–31, 335, 425 ss 1–4 329 s 4 330, 487 s 4(1) 187 s 4A 329 s 5 32, 41, 44, 49, 51, 329, 487 s 6(5) 195, 200, 201 s 8 329 s 19(1) 187 s 29 32 Racial and Religious Hatred Act 2006 32 Road Safety Act 2006 237 s 19 134 s 30 301

Page 7 of 12

Table of legislation Road Traffic Act 1988 305 s 1 300 s 2(3) 301 s 2A(1) 300 s 2A(2) 300 s 2B 300 s 3ZA 301 s 3ZB 300 Road Traffic Regulation Act s 87 134 Serious Crime Act 2007 17, 277, 426 Pt 2 (ss 44–67) 41, 47, 444, 471, 477, 483 s 44 424, 477, 478–80, 486, 490 ss 44–46 419 s 45 479, 480, 490 s 46 480, 481, 490 s 47(2)–(4) 483 s 47(4) 481 s 47(5)(a) 479 s 47(5)(b) 479 s 47(7)(b) 483 s 49(7) 480 s 50 135, 429, 434, 450 s 50(3) 483 s 50(3)(b) 135 s 51 483 s 59 119 s 65 487 s 65(1) 478 s 65(3) 479 Sch 3 480 Sexual Offences Act 1956 57 s 2 350 s 3 348 Sexual Offences Act 1967 38 Sexual Offences Act 2003 47, 49, 72, 80, 166, 170, 185, 188, 217, 221, 307, 335–66, 450, 479 s 1 341 s 1(1)(c) 231 ss 1–4 339, 346, 354 (p. xxxi) s 2 343, 344 s 3 344 s 4 344, 346, 359

Page 8 of 12

Table of legislation s 5 167, 358 ss 5–8 231, 339, 356–8, 364 s 6 356 s 7 356 s 8 357, 359 s 9 358, 359, 459, 460, 483 s 9(1) 359 s 9(1)(c) 231 s 9(2) 359 ss 9–12 358, 361, 363 ss 9–13 360, 489 ss 9–15 339, 358 s 11 359, 365 s 12 365, 490 s 13 138, 358, 359, 490 s 14 135, 365, 489 s 14(3) 360 s 15 360 ss 16–19 138, 360, 361 ss 16–24 339 s 18 138, 365 s 19 365 s 21 360, 408 s 22 408 s 25 361 ss 25–29 339 s 26 362 s 28 362 s 30(2) 351, 363 ss 30–33 363 ss 30–44 339, 350 ss 34–37 363 ss 38–44 363 ss 45–50 53, 363 ss 52–60 339, 363 ss 61–63 339 s 62 363 s 63 363, 401 s 64 339, 362 s 65 339, 362 s 66 363 ss 66–71 339 s 67 363 s 67(1) 319 s 69 363 s 70 364 s 71 38, 364

Page 9 of 12

Table of legislation s 73 134, 365 s 74 348, 349, 351, 352, 355, 360, 364 s 75 352, 354, 364 s 75(2) 349, 354 s 75(2)(d) 355 s 75(2)(f) 355 s 76 348, 352, 354, 364 s 76(2) 347 s 76(2)(a) 348 s 78 344, 345 s 79(2) 158, 342, 376 s 79(3) 342 s 79(8) 343, 344 Sexual Offences Act 2009 s 37 365 Sexual Offences (Amendment) Act 2000 360 Statutory Instruments Act 1946 s 3(2) 219 Street Offences Act 1959 58 Suicide Act 1961 s 2 277, 279 s 2(1) 276–8 s 2A 277 Terrorism Act 2000 s 57 236 Terrorism Act 2006 s 1 477 s 1(1) 42 s 1(2) 41 s 1(3)(b) 42 s 1(5)(b) 41 s 5 42, 454 s 20(2) 42 Theft Act 1968 70, 79, 368–70 s 1 372, 410 s 2 390, 391, 407, 412 s 2(1) 393 s 2(1)(a) 218, 390, 396 s 2(1)(b) 390 s 2(2) 390 s 3(1) 373, 377, 403

Page 10 of 12

Table of legislation s 3(2) 378, 393, 403 s 4 382 s 4(1) 381 s 4(3) 382, 393 s 4(4) 382 s 5 382–4 s 5(1) 382, 383 s 5(2) 384 s 5(3) 384, 385 s 5(4) 380, 385 s 6 387, 389 s 6(1) 388 s 6(2) 388 s 8 395, 397 s 8(2) 313 s 9 363, 398, 399, 417 s 9(1)(a) 400, 401, 412 s 9(1)(b) 400, 401 (p. xxxii) s 10 400 s 12 386, 389, 394 s 13 381, 402 s 15 405 s 16 405 s 17 405 s 19 405 s 21 397 s 21(1) 85, 220 s 22 402 s 25 97 s 34(2) 397 Theft Act 1978 ss 1, 2 405 Theft (Amendment) Act 1996 70, 71 Trade Descriptions Act 1968 150 Trade Disputes Act 1906 468 Water Resources Act 1991 110 s 85 149

France Code Penale Art 121–3 292 Art 221 282 Art 221–4-8 471

Page 11 of 12

Table of legislation Art 221–6 286, 292 Art 222–7 286, 289 Art 222–17 487 Art 223 425 Art 223–15–2 376 Art 311–1 389 Art 345 284

USA California Penal Code s 484–502–9 379 Model Penal Code 138, 160, 175 s 2.01(1) 88 s 2.02(s)(c) 176 s 2.03 108 s 2.04(3) 223 s 2.5.01 462 s 2.06(3) 428, 433 s 2.13 224 s 3.02 131 s 3.04 121 s 3.04(2)(a)(i) 121 s 4.01 145 s 6.02(4) 165 s 210.3.1(b) 283

Criminal Justice and the Criminal Law

Principles of Criminal Law (7th edn) Andrew Ashworth and Jeremy Horder Publisher: Oxford University Press Print ISBN-13: 9780199672684 DOI: 10.1093/he/9780199672684.001.0001

Print Publication Date: May 2013 Published online: Sep 2013

1. Criminal Justice and the Criminal Law Chapter: (p. 1) 1. Criminal Justice and the Criminal Law Author(s): Andrew Ashworth and Jeremy Horder DOI: 10.1093/he/9780199672684.003.0001

1.1 The contours of criminal liability 1.2 The machinery of English criminal law 1.3 The sources of English criminal law 1.4 The criminal law in action 1.5 Outline of the aims and functions of the criminal law 1.6 The criminal law and sentencing Further reading The operation of the criminal law requires little explanation in clear cases. Someone who deliberately kills or rapes another is liable to be prosecuted, convicted, and sentenced. Criminal liability is the strongest formal censure that society can inflict, and it may also result in a sentence which amounts to a severe deprivation of the ordinary liberties of the offender. Of course, there are other official deprivations of our liberties: taxation is one, depriving citizens of a proportion of their income, or adding a compulsory levy to commercial transactions (for

Criminal Justice and the Criminal Law example, Value Added Tax). And taxation, no less than the criminal law, may be seen as justified by the mutual obligations necessary for worthwhile community living. But the taxing of an activity does not carry any implication of ‘ought not to do’, whereas criminal liability carries the strong implication of ‘ought not to do’. It is the censure conveyed by criminal liability which marks out its special social significance, and it is the imposition of this official censure, and the ensuing liability to state punishment ordered by the court, that requires a clear social justification. The chief concern of the criminal law is to prohibit behaviour that represents a serious wrong against an individual or against some fundamental social value or institution.1 This suggests, perhaps, that there are some wrongs that are not serious enough (or appropriate for) any legal liability, such as breaking a promise to a friend without good reason, or divulging information given in confidence by a friend, and there are some (p. 2) wrongs that are serious enough for civil liability—such as breach of contract—but not for criminal liability. But the notion that English criminal law is only concerned with serious wrongs must be abandoned as one considers the broader canvas of criminal liability. There are many offences for which any element of stigma is diluted almost to vanishing point, as with illegal parking, riding a bicycle without lights, or dropping litter. This is not to suggest that all such offences are equally unimportant; there are some situations in which illegal parking can cause danger to others, for example. Yet it remains true that there are many offences for which criminal liability is merely imposed by Parliament as a practical means of regulating an activity, without implying the element of social censure which is characteristic of the major or traditional crimes. There is thus no general dividing line between criminal and non-criminal conduct which corresponds to a distinction between immoral and moral conduct, or between seriously wrongful and other conduct. The boundaries of the criminal law are explicable largely as the result of exercises of political power at particular points in history. The idea of a crime is that it is something that rightly concerns the State, and not just the person(s) affected by the wrongdoing. Many crimes are civil wrongs as well (torts or breaches of contract, for example), and it is for the injured party to decide whether or not to sue for damages. But the decision to make conduct into a crime implies that there is a public interest in ensuring that such conduct does not happen and that, when it does, there is the possibility of State punishment. The police and the Crown Prosecution Service take decisions on whether to prosecute someone who is reasonably suspected of committing an offence: although they should ‘take into account any views expressed by the victim regarding the impact that the offence has had’, ‘prosecutors should take an overall view of the public interest’.2 Moreover, even if an individual citizen is wronged behind closed doors, as in cases of ‘domestic violence’, the State has an interest: But whatever else is unclear about the rights and wrongs of a domestic dispute … such violence should surely not be seen as a matter for negotiation or compromise. It should be condemned by the whole community as an unqualified wrong; and this is done by defining and prosecuting it as a crime.3 This view is sometimes phrased in terms of crimes as ‘attacks on the community as a whole’, but, as Grant Lamond argues,4 a more convincing way of understanding crimes as public wrongs is to regard them not as wrongs to the community but as wrongs that the community is appropriately responsible for punishing. That, in philosophical terms, is what is characteristic

Page 2 of 21

Criminal Justice and the Criminal Law of crimes, at least of fault-based crimes. Thus it is in the public interest to provide for the punishment of the serious wrongs involved in violent (p. 3) acts, wherever they occur and whoever inflicts them. But in practice matters are not so clear-cut: some crimes simply cannot be prosecuted without the victim's testimony, and so the victim's refusal to co-operate with the prosecution may be determinative.5 Victims of crime now have the right to make a Victim Personal Statement about the effects of the offence on them, although the courts should decline to take the further step of taking account of the views of a victim or victim's family on the question of sentence.6 Various pro-victim initiatives, and the advent of forms of restorative justice, have raised further questions about the interface between crimes as ‘offences against the State’ and the involvement of victims in decision-making in criminal justice.7 Given the variety of forms of behaviour that have been criminalized, it is no surprise that Glanville Williams ended his search for a definition of crime without identifying any criterion based on subject-matter. He concluded that only a formal definition is sustainable: ‘in short, a crime is an act capable of being followed by criminal proceedings having a criminal outcome’.8 But one consequence of the Human Rights Act 1998 is that it is no longer for Parliament to stipulate that proceedings should be regarded as civil only. Article 6 of the European Convention on Human Rights confers extra procedural rights on any person ‘charged with a criminal offence’—the presumption of innocence, a right to legal aid, a right to confront witnesses, a right to an interpreter, and so forth. The European Court of Human Rights in Strasbourg has insisted that the question whether a person is ‘charged with a criminal offence’ is for the court to determine by looking at the substance of the situation. This amounts to what one might term an ‘anti-subversion device’, created by the Strasbourg Court to prevent governments from manipulating the criminal/civil boundary and thereby avoiding those extra procedural rights. The leading decisions establish that if (a) the proceedings are brought by a public authority, and (b) there is a culpability requirement, or (c) there are potentially severe consequences (such as imprisonment or a significant financial penalty), the person will be deemed to be ‘charged with a criminal offence’ and will be granted the full Art. 6 protections.9 The effect of deciding that particular proceedings are essentially criminal is not to alter the court in which the case should be heard, but to require the court to ensure that the defendant is accorded the full rights conferred by Art. 6 on any person ‘charged with a criminal offence’.10 This is a development of particular significance, not least because it is the magnitude of the penalty (the possibility of imprisonment or a sizeable fine) that is a major factor inclining the court to declare the proceedings criminal. In this way, the idea of criminal law is firmly linked with the possibility of a significant sentence, which in turn calls for extra procedural protections for defendants.11 (p. 4) Why might governments wish to avoid allowing defendants to have the full protections appropriate to criminal proceedings? One reason is that they may wish to use the criminal law to deal with relatively minor infractions, in a kind of streamlined procedure. Some European countries have instituted a separate system of administrative offences, with low penalties, as a way of dealing swiftly, inexpensively, effectively, and not unfairly with non-serious wrongdoing.12 English law does not have a general separate system dealing with minor ‘infringements’ rather than with serious wrongs, but in many individual areas of activity—such as environmental control, animal welfare, and some aspects of financial service provision—it does permit regulators to impose civil or administrative penalties rather than using criminal prosecution against violators. What has, though, been particularly noteable, as a development in England and Wales, has been the creation of more serious forms of hybrid measure to

Page 3 of 21

Criminal Justice and the Criminal Law regulate and deter behaviour that is considered ‘anti-social’ or that is thought to present an unacceptable risk to others. Examples include anti-social behaviour orders, risk of sexual harm orders, foreign travel restriction orders, and violent offender orders. These new measures, often referred to as civil preventive orders, consist of two stages. The first is the imposition, by a court in civil proceedings, of an order prohibiting D from doing certain things or going to certain places. The second is a criminal offence of failing without reasonable excuse to comply with that order, usually carrying a maximum sentence of five years’ imprisonment. The only element of the civil preventive order that falls within the criminal law, according to the government and the judiciary, is the offence of failure to comply. Even though the contents of that offence (i.e. the prohibitions imposed on D) are determined in civil proceedings without the protections of criminal procedure, the courts have not used their power to declare that the two stages of proceedings taken together are criminal in substance.13 They have accepted the government's device of de-coupling the two stages: the House of Lords recognized the potentially severe consequences for D of a breach of the prohibition(s), but its only concession was the compromise of requiring the criminal (rather than the civil) standard of proof in the civil proceedings.14 The upshot is that a court sitting in civil proceedings can create a kind of personal criminal code for D, breach of which renders him liable to a maximum punishment higher than that for many ordinary criminal offences. The best known example of this is the anti-social behaviour order, or ASBO. It was conceived as a means for tackling social problems that are undeniable in their effect (p. 5) on the quality of others’ lives, and it became the talisman of civil preventive orders because the problem of witness intimidation and the focus of criminal proceedings on a particular event (rather than on aggregate nuisance) was thought to render the criminal law insufficiently effective. A court can only make an ASBO if satisfied that D has acted in a manner likely to cause harassment, alarm, or distress, and if the prohibitions in the order are ‘necessary’ for the purpose of protecting persons from further anti-social acts by D (Crime and Disorder Act 1998, s. 1, as amended). However, the prohibitions may include non-criminal or criminal conduct, they must last for at least two years, and about half of breach prosecutions result in imprisonment. This raises serious questions about whether civil preventive orders, and the way in which they use the criminal law so as to avoid or minimize its procedural protections, constitute a lawful, fair, or proportionate response to the risks concerned.15 More broadly, these orders may be regarded as one manifestation of a more general movement away from the paradigms of the criminal law, and the consequent side-lining of the protections of criminal procedure. Thus the greater use of diversion from the criminal process, of fixed penalties, of summary trials, of hybrid civil– criminal processes, of strict liability offences, of incentives to plead guilty, and of preventive orders—all of these challenge the paradigm of the criminal law, and challenge the way it is traditionally presented.16

1.1 The contours of criminal liability When we refer to criminal liability, what sort of conduct are we talking about? The answer may differ not only from one country to another, but also from one era to another in the same country. Thus in the last fifty years there have been several changes in the boundaries of the law of sexual offences, and (for example) most hom*osexual encounters which were criminal in England before 1967 are not criminal now, whereas some forms of insider trading on the stock market and of the possession of indecent photographs have become criminal. There are

Page 4 of 21

Criminal Justice and the Criminal Law certain serious wrongs which are criminal in most jurisdictions, but in general there is no straightforward moral or social test of whether conduct is criminal. The most reliable test is the formal one: is the conduct prohibited, on pain of conviction and sentence? The contours of criminal liability may be considered under three headings: the range of offences; the scope of criminal liability; and the conditions of criminal liability. The range of criminal offences in England and Wales is enormous. There are violations of the person, including offences of causing death and wounding, sexual offences, (p. 6) certain public order offences, offences relating to safety standards at work and in sports stadiums, offences relating to firearms and other weapons, and serious road traffic offences. Then there are violations of general public interests, including offences against state security, offences against public decency, offences against the administration of justice, and various offences connected with public obligations such as the payment of taxes. A third major sphere of liability comprises violations of the environment and the proper conditions of communal living, including the various pollution offences, offences connected with health and purity standards, and minor offences of public order and public nuisance. Fourthly, there are violations of property interests, from crimes of damage and offences of theft and fraud, to offences of harassment of tenants and crimes of entering residential premises. And then there is a mass of financial, business, and industrial offences, often created in order to enforce a regulatory scheme, but many having maximum penalties as high as seven years’ imprisonment. More will be said about these five major fields of criminal liability in Chapter 2. As in many other legal systems, there is a whole host of miscellaneous criminal prohibitions as well. When we turn to the scope of criminal liability, we raise the question of the circ*mstances in which a person who does not actually cause one of the above harms may, nevertheless, be held criminally liable. In legal terms, the question has two dimensions: inchoate liability and criminal complicity. A crime is described as inchoate when the prohibited harm has not yet occurred. Several of the offences mentioned in the last paragraph are defined in terms of ‘doing an act with intent to cause X’, and they do not therefore require proof that the prohibited harm actually occurred. Additionally, there are the general inchoate offences of attempting to commit a crime (e.g. attempted murder), conspiring with one or more other people to commit a crime (e.g. conspiracy to rob), and encouraging or assisting crime. These offences broaden the scope of criminal liability considerably, by providing for the conviction of persons who merely try or plan to cause harm. Turning to criminal complicity, this doctrine is designed to ensure the conviction of a person who, without actually committing the full offence, plays a significant part in an offence committed by another. Thus a person may be convicted for aiding and abetting, counselling or procuring another to commit a crime, or for participating in a joint criminal venture during which another participant commits a more serious offence than planned. The conditions to be fulfilled before an individual is convicted of an offence vary from one crime to another. There are many crimes which require only minimal fault or no personal fault at all. These are usually termed offences of ‘strict liability’: some of them are aimed at companies, but others (including many road traffic offences) are aimed at individuals. More of the traditional offences, which have been penalized by the common law of England for centuries, are said to require ‘mens rea’. This Latin term indicates, generally, that a person should not be convicted unless it is proved that he intended to cause the harm, or knowingly risked the occurrence of the harm. The emphasis of these requirements has been upon the

Page 5 of 21

Criminal Justice and the Criminal Law defendant's personal awareness of what was being done or omitted. However, neither the legislature nor the courts subscribe to a firm rule that any serious offence should require proof of mens rea: not only (p. 7) are most of the offences tried in magistrates’ courts strict liability offences, but half of the offences triable in the Crown Court have at least one strict liability element.17 Beyond the mens rea or fault requirement, which may differ in its precise form from crime to crime, there is a range of possible defences to criminal liability, so that even people who intentionally inflict harms may be acquitted if they acted in self-defence, whilst insane, whilst under duress, and so on. The contours of the criminal law are thus determined by the interplay between the range of offences, the scope of liability, and the conditions of liability. Inevitably there are times when the discussion focuses on only one of the elements, but the relevance of the other two must always be kept in view if the discussion is not to lose perspective.

1.2 The machinery of english criminal law The criminal courts in England and Wales are the magistrates’ courts and the Crown Court. Those offences considered least serious are summary offences, triable only in the magistrates’ courts. Those offences considered most serious are triable only on indictment, in the Crown Court. A large number of offences, such as theft and most burglaries, are ‘triable either way’, in a magistrates’ court or the Crown Court. If a defendant decides to plead guilty to an ‘either way’ offence, the magistrates can proceed to sentence (or commit to the Crown Court for sentence) without further ado. If the defendant signifies an intention to plead not guilty, proceedings to determine the mode of trial are held before magistrates. The magistrates may decide (having heard representations from the prosecutor) that the case is so serious that it should be committed to the Crown Court for trial. If they decide not to commit it to the Crown Court, the defendant still has an absolute right to elect trial by jury. In practice a majority of ‘either way’ offences are dealt with in magistrates’ courts, since neither the defendant nor the magistrates think Crown Court trial necessary. However, the question of a defendant's ‘right’ to trial by jury is a perennial subject of debate, as we will see in section 1.4. Finally, it should be added that the great majority of prosecutions of persons under 18 are brought in Youth Courts, where hearings are less formal and take place before specially trained magistrates.

1.3 The sources of English criminal law The main source of English criminal law has been the common law, as developed through decisions of the courts and the works of such institutional writers as co*ke and (p. 8) Hale in the seventeenth century, and Hawkins, Foster, and Blackstone in the eighteenth century. The bulk of English criminal law is now to be found in scattered statutes. There was a major consolidation of criminal legislation in 1861, and the Offences Against the Person Act of that year remains the principal statute on that subject. In recent years Parliament has created a range of new crimes, from keeping a dangerous dog to stalking, from failing to comply with an anti-social behaviour order to intimidating witnesses. However, some offences are still governed by the common law and lack a statutory definition—most notably, murder, manslaughter, assault, and conspiracy to defraud. Many of the doctrines that determine the

Page 6 of 21

Criminal Justice and the Criminal Law conditions of criminal liability are also still governed by the common law—not merely defences such as duress, intoxication, insanity, and automatism, but also basic concepts such as intention and recklessness. The judges therefore retain a central place in the development of the criminal law. They seem to bear the major responsibility for developing the conditions and the scope of criminal liability, and also exert considerable influence on the shape of the criminal law through their interpretation of statutory offences. Moreover, the Human Rights Act 1998 bestowed on the judges further powers and duties. Section 6 requires them always to act compatibly with Convention rights; s. 3 requires them to interpret statutory provisions, so far as is possible, in such a way as is compatible with Convention rights; and s. 2 obliges them to take account of the jurisprudence of the European Court of Human Rights. The higher courts do have the power to make a ‘declaration of incompatibility’ if they cannot otherwise bring a statute into conformity with the Convention, but that power is exercised sparingly. Cases in which the courts have used their interpretive powers under the Human Rights Act will be signalled throughout the book.18 The Convention has not had as large an effect on the criminal law as upon criminal procedure and evidence, but there are already several judicial decisions in which the Human Rights Act has made a difference. Also of significance is European Community law, not least because it has direct effect in this country and thus (unlike the Convention) automatically takes precedence over domestic laws. Where a rule of English criminal law unjustifiably curtails a right conferred by Community law (such as the free movement of goods), the domestic law is disapplied and the defendant should not be convicted. European Community law has not yet had great effects on English criminal law, particularly the more serious offences which occupy much of the discussion in this book, but its potential as a source of liability and of defences should not be overlooked.19 One obvious difference between English criminal law and that of most other European jurisdictions is the absence of a criminal code. The somewhat faltering steps taken towards the enactment of an English criminal code are described at Chapter 3.2. (p. 9) For the present, we must record that the absence of a criminal code reduces the internal consistency of English criminal law (because it has to be gathered from judicial decisions, scattered statutes, and occasionally from the old institutional writers) and also makes it harder to locate the applicable law. The attraction of these practical arguments in favour of a code is powerful, but, as we shall see, there has been controversy about which parts of the criminal law should be codified and in what way.

1.4 The criminal law in action It would be unwise to assume that the criminal law as stated in the statutes and the textbooks reflects the way in which it is enforced in actual social situations. The key to answering the question of how the criminal law is likely to impinge on a person's activities lies in the discretion of the police and other law enforcement agents: they are not obliged to go out and look for offenders wherever they suspect that crimes are being committed; they are not obliged to prosecute every person against whom they have sufficient evidence. On the other hand, they cannot prosecute unless the offence charged is actually laid down by statute or at common law. So we must consider the interaction between the law itself and the practical operation of the criminal process if we are to understand the social reality of the criminal law. Page 7 of 21

Criminal Justice and the Criminal Law

Even before the discretion of law enforcement officers comes into play, there is often a decision to be taken by a member of the public as to whether to report a suspected offence. The Crime Survey England and Wales (formerly the British Crime Survey) suggests that at least a half of all offences are not reported to the police,20 often because they are thought to be too trivial, or because it is thought that the police would be unable to do anything constructive, or because it is thought that the police ‘would not be interested’. Even where an assault results in hospital treatment, a significant proportion of victims fail to report the offence or at least to make a formal complaint.21 Thus, if an offence is to have any chance of being recorded, either the victim or a witness must take the decision to report the offence to the authorities. About four-fifths of the offences which come to police attention are reported by the public. This means that people's (sometimes stereotyped) views on what forms of behaviour amount to criminal offences, and also on whether the police should be called, exert considerable influence on the cases entering the criminal justice system. For this and various other reasons, many offences committed at work or in the home remain concealed from official eyes. As for the one-fifth of offences that come to light in other ways, most of (p. 10) these are observed or discovered by the police themselves. There are some crimes, such as drug dealing and other so-called ‘crimes without victims’, which are unlikely to be reported and which the police have to go looking for. And there are other crimes, such as obstructing a police officer and some of the public order offences, which the police may use as a means of controlling situations—charging people who disobey police instructions about moving on, keeping quiet, etc.22 In these contexts the police use the criminal law as a resource to reinforce their authority. It will thus be seen that most police investigations of offences are ‘reactive’, that is, reacting to information from the public about possible offences. Only in a minority of cases do the police operate ‘proactively’. Other law enforcement officials may have a larger proactive role. Her Majesty's Revenue and Customs (HMRC) investigate offences relating to taxation and smuggling. Various inspectorates are required to oversee the observance of legal standards in industry and commerce—the Health and Safety Executive (which includes seven inspectorates: Factory, Agriculture, Nuclear, Offshore, Mines, Railway, and Quarries), the Environment Agency, the Environmental Health Departments of local authorities, and so on. Although these agencies often react to specific complaints or accidents, much of their work involves visits to premises or building sites to check on compliance with the law. It is therefore proactive work: the number of offences coming to an inspectorate's attention is largely a reflection of the number of visits and inspections carried out, and the response depends on the general policies and specific working practices of that inspectorate.23 What happens when an offence has been reported to the police? In most cases the offence is recorded and the police may investigate it. However, the Crime Survey England and Wales suggests that about one-half of all incidents reported to the police as crimes are not recorded as such: sometimes the evidence is thought unconvincing, or the offence too minor, or the incident redefined as lost property rather than theft of property. Of those that are recorded as crimes, the police trace around 28 per cent to an offender or suspected offender. The proportion of offences thus detected is much higher for offences of violence (around one-half) —where the victim often sees and knows the offender—than for the offences where the perpetrator's identity will often be unknown (such as burglary, 13 per cent, and robbery, 20 per cent) and for offences that may be thought not to justify a great investment of police time 24 Page 8 of 21

Criminal Justice and the Criminal Law and resources (e.g. criminal damage, 14 per cent).24 When the police find a suspect, they will invariably try to question this person. The Police and Criminal Evidence Act 1984 and its Codes of Practice require investigators to follow certain procedures before and during any interrogation, including notifying suspects of the right to a free and private consultation with a lawyer. The tape recording (p. 11) of suspects’ statements is a routine feature at police stations, although statements (allegedly) made elsewhere remain admissible in evidence. Many of the miscarriages of justice uncovered in the late 1980s and early 1990s, after wrongly convicted people had spent many years in prison, stemmed from misconduct by the police at this stage of the investigation, including the falsification of notes of interviews. Following the quashing of convictions in the cases of the Guildford Four and the Birmingham Six, the Royal Commission on Criminal Justice was appointed in 1991 to examine the effectiveness of the criminal justice system. In its report the Royal Commission recognized that ‘confessions which are later found to be false have led or contributed to serious miscarriages of justice’,25 but one of its key proposals, on preserving the right of silence, was rejected by the then government in favour of introducing a law that permits adverse inferences from failure to answer police questions.26 When the police have completed their questioning, they should release the suspect if they have insufficient evidence. If they believe they have sufficient evidence, or if the suspect has admitted guilt, there are choices to be made between prosecution, one of the forms of ‘caution’, and no further action. In recent years young offenders have usually received a caution, in the form of either a reprimand or a final warning under the Crime and Disorder Act 1998. For some years the policy had been to delay the entry of young people into the formal criminal justice system, in the belief that cautions were no less likely to be effective in preventing further offences, and that labelling a youth as a delinquent through formal court proceedings could reinforce that person's tendency to behave like a delinquent. The system of reprimands and warnings introduced by the 1998 Act was intended to be more rigorous and more demanding of young offenders (a warning also involves referring the young offender to a youth offending team, who may require the offender to participate in a scheme designed to prevent re-offending).27 Thus, although the emphasis remains on the prevention of future offending, the rhetoric and the method have changed to confronting offenders with their behaviour and helping them to take more responsibility for their actions.28 The emphasis of the Youth Justice Board remains on diverting most young offenders away from court, and making use of alternative approaches such as restorative conferences that may bring the offender face-to-face with the victim.29 The proportion of young offenders receiving reprimands or warnings rather than prosecution is around two-thirds for boys and four-fifths for girls in the 12–14 age group, while in the 15–17 age group the proportions are about 45 per cent for boys and some two-thirds for girls. For adults the police may also decide to caution an offender, but again the use of cautions has declined somewhat from its peak (p. 12) in the early 1990s, so that around 20 per cent of adult male offenders and one-third of adult female offenders receive a caution. National Standards encourage the police to prefer a formal caution to prosecution where the offence is relatively minor, where the offender is old, infirm, or suffering from mental disturbance, and in other situations where there is little blame. Since 2003, the Crown Prosecution Service (CPS) has had the power to offer an offender a conditional caution, in cases where there is both an admission by the offender and objectively sufficient evidence. The conditions specified may include the making of reparation or participation in a restorative justice process. The offender is required to sign a document that spells out the conditions and 30 Page 9 of 21

Criminal Justice and the Criminal Law records his admission of the offence.30 There are also various out-of-court disposals, such as cannabis warnings, Penalty Notices for Disorder, and fixed penalties.31 Standing in contrast to the preferred use of prosecutions for suspected adult offenders (as distinct from juveniles) is the long-standing preference for alternatives to prosecution among the various inspectorates and other public authorities, such as HMRC. Many of these agencies regard their main aim as securing compliance rather than convictions. The Environment Agency, for example, states: ‘We regard prevention as better than cure. Our general approach is to engage with business to educate and enable compliance.’32 Such agencies therefore tend to rely on informal and formal warnings as a means of putting pressure on companies, employers, taxpayers, and the like to conform to the law. Most of these agencies regard prosecution as a last resort: the criminal law remains as a background source of the pressure towards compliance which the agencies are able to exert. Thus, for example, the HMRC Prosecutions Office typically responds to tax evasion through civil procedures and rarely resorts to prosecution.33 In these contexts, then, the criminal law is very much in the background, and the criminal process is experienced by relatively few of those caught breaking the law.34 Where the police are involved, however, prosecution remains the normal response for persons aged 18 and over. The initial decision whether or not to charge is taken under the ‘statutory charging scheme’ introduced by the Criminal Justice Act 2003. This means that police and prosecutors work together at this stage, but it is the CPS that takes the decision whether to charge and, if so, with what offence to charge the suspect. One of the aspects to be considered is evidential sufficiency: is there enough evidence on each of the elements required to prove the offence, so that it can be said that there is a realistic prospect of conviction? This requires the prosecutor to consider both the amount of evidence available and its admissibility in court (e.g. whether there has been a breach of the Police and Criminal Evidence Act 1984 and its Codes of Practice). The (p. 13) second, related, factor is whether a prosecution would be in the public interest. There is a Code for Crown Prosecutors (latest version, 2010) to provide general guidance on this and other decisions which prosecutors must take, and there is detailed guidance on prosecution policy for particular types of offence.35 It will be apparent from the preceding paragraphs that the defendants and offences brought to court form a highly selective sample of all detected crimes. Those convicted in court are certainly a small sample of the whole: the Home Office has estimated that if one takes account of those offences not reported, not recorded, not cleared up, and cautioned rather than prosecuted, only some 2 or 3 per cent of crimes result in a conviction.36 Although this rises to 10 per cent for crimes such as wounding, it would not be accurate to say that the cases brought to court involve the most serious offences and offenders, because: (1) there are crimes for which a person under 18 would not be prosecuted, whereas a person of 18 or over would be; (2) fairly serious crimes committed in the home or in certain workplaces may sometimes not be reported or prosecuted,37 whereas prosecution is often the normal response to less serious offences in the street; and (3) some of the crimes of petty theft which are prosecuted are, by almost any measure, less serious than many crimes which HMRC or other regulatory agencies deal with by

Page 10 of 21

Criminal Justice and the Criminal Law warnings, civil penalties, or other alternative methods. What emerges from this is that adults suspected of committing ‘traditional’ offences outside their own home are much more likely to appear in court than adults known to have committed more regulatory kinds of offence, such as tax evasion, pollution, having an unsafe workplace, and so on. Even leaving young people aside, then, court proceedings are a poor representation of the reality of crime in our society. It will be evident, too, that it is not rules but discretionary decisions which characterize these early stages in the criminal process. Police decision-making is largely discretionary, structured only by the cautioning guidelines, local arrangements for dealing with young defendants, police force orders, and internal police supervision. As research into public order policing confirms, there are considerable variations in policy and practice, not just between police force areas but also among police divisions in the same force,38 and this determines the nature and volume of cases placed before the CPS for consideration for prosecution. The same is largely true of the regulatory and other agencies which have the power to prosecute. Moreover, the elements of discretion do (p. 14) not stop with the decision whether or not to prosecute. A question of particular importance for our present purposes is that of deciding what offence to charge. In some cases there is little choice, but there are other cases where the prosecutor can choose between a more serious and a less serious offence. If there is a prosecution for the higher offence, it is usually possible for a court to convict of a lesser offence if it does not find the higher offence proved. But this does not mean that prosecutors routinely try for the higher offence. If, for example, the lower offence is triable only summarily (i.e. in a magistrates’ court), whereas the higher offence is ‘triable either way’ (i.e. in a magistrates’ court or at the Crown Court), the prosecutor may prefer the lesser charge so as to keep the case in a magistrates’ court—for various reasons, one of which may be the belief that a conviction is more likely if the case is tried by magistrates rather than by a jury.39 Whether the case is set down for trial in a magistrates’ court or the Crown Court, the prosecution may reach an agreement with the defence to accept a plea of not guilty to the offence charged but guilty to a lesser offence, or not guilty to some offences charged and guilty to others. We have also noted that, where a defendant on an indictable charge signifies an intention to plead guilty, the magistrates’ court may proceed to pass sentence (or may commit to the Crown Court for sentence). A guilty plea, even to a lesser offence, is often advantageous to the prosecutor because a conviction is assured and the hazards of a trial (with the possibility that a key witness will not give evidence convincingly) are avoided. A guilty plea may also appear advantageous to the defendant, since the conviction may be for a lesser offence (or for fewer offences). The sentence should also be lower because of the guilty plea, as s. 144 of the Criminal Justice Act 2003 requires. Guidelines issued by the Sentencing Guidelines Council establish a sliding scale of sentence reductions, running from a one-third discount for indicating a guilty plea at the earliest opportunity down to a one-tenth reduction for a guilty plea ‘at the door of the court’.40 These incentives could generate considerable pressure on a defendant to plead guilty even where innocence is maintained, if defence lawyers emphasize the strength of the evidence for the prosecution, etc., and there are clearly some cases in which innocent persons feel driven to plead guilty.41 Average prison sentences for those pleading guilty are around a third below those convicted after a trial—a striking difference. Moreover, there are other procedural incentives to plead guilty, such as the right to ask the court for an indication of whether the sentence will be custodial or non42 Page 11 of 21

Criminal Justice and the Criminal Law custodial (the court is not obliged to give the indication).42 Not only, then, are the cases prosecuted a selective sample of all crimes committed, but the offences for which convictions are recorded may sometimes underestimate (p. 15) the true seriousness of the crimes brought to court. It may be in the apparent best interests of both prosecution and defence to settle for conviction of a less serious offence. The way in which offences are defined may facilitate or constrain many of these decisions, and so the structure of the criminal law may have a greater direct influence at this point than at some earlier stages in the criminal process. However, the predominance of official discretion opens the way for other motivations, including bias and prejudice, to enter in. Although the ratio of male to female known offenders is around five to one, we have already seen that females are cautioned (rather than prosecuted) at a higher rate than men. The picture at the sentencing stage is rather complex, with some women receiving less severe sentences than comparable men, but some receiving disposals which may turn out to be more intrusive and therefore more severe.43 There is also some evidence of racial discrimination in the pre-trial system,44 although some of this is a form of structural bias stemming from the way in which the system imposes disincentives on those who elect Crown Court trial and who maintain a plea of not guilty. This brief outline has, it is hoped, demonstrated some of the ways in which the criminal law in action differs from the law as declared in the statutes and in court decisions. Little has been said about the ways in which defence lawyers may sometimes construct the defence around their own working priorities as well as around reconstructed versions of the defendant's narrative,45 but that, too, is a factor in the presentation and the outcome of cases. So far as official agencies are concerned, each case brought to court is the product of a system which is heavily reliant on victims and other members of the public for the detection of offenders and the provision of evidence,46 and which leaves considerable discretion in the hands of the police, other law enforcement agencies, and the CPS. We have seen that that discretion is exercised unevenly, in the sense that those who commit crimes on the streets and in other public places are likely to be prosecuted, even for relatively minor incidents, whereas offenders of certain kinds (‘white-collar’) are rarely brought to court. This not only emphasizes that the law in practice is different from the law in the books. It also raises questions about priorities and social justice: should we not have a ranking of crimes that makes it clear which are the most serious and which are the least serious, with the greatest efforts directed at enforcement against those who perpetrate the most serious offences, and the strongest measures taken against those offenders?47 (p. 16) 1.5 Outline of the aims and functions of the criminal law Is the criminal law necessary at all? It has been suggested from time to time that law in general and criminal laws in particular are needed only in conflict-ridden societies, and that the establishment of a political system based on consensus would remove the sources of crime and, therefore, the phenomenon itself.48 This view may be attacked for its simplistic assessment of the causes of crime, but it is sufficient for present purposes to state that no modern industrially developed country seems able to dispense with criminal law, and, indeed, that occasional instances of the breakdown of policing have led to increases in certain forms of criminal behaviour.49 This suggests that one fundamental reason for having a criminal law

Page 12 of 21

Criminal Justice and the Criminal Law backed by sanctions is deterrent or preventive: so long as its provisions are enforced with some regularity, it constitutes a standing disincentive to crime and reinforces those social conventions and other inhibitions which are already in place. Fundamental though this proposition is, it should not obscure the importance of two related propositions. One is that criminal prosecutions should not be regarded as a primary means of protecting individual and social interests. In terms of prevention, more can probably be achieved through various techniques of situational crime prevention,50 social crime prevention, and general social and educational policies51—not least because relatively few offences lead to a court appearance and conviction. The other proposition is that the underlying deterrent rationale of the criminal law does not support the notion that changes in particular laws, or changes in sentencing levels, ought to be pursued for their deterrent effects. The evidence on what is called ‘marginal deterrence’ is generally equivocal, and it cannot be assumed that creating a new crime or increasing the maximum punishment will lead—in a kind of hydraulic relationship—to a reduction in the incidence of that conduct.52 A primary justification for criminal law and sentencing is that offenders deserve punishment for their offences, and that it is therefore just (and not merely expedient) to provide that serious wrongs culpably inflicted should lead to censure and sanctions by the State. The justice of punishment for culpable public wrongs may be said to stem (p. 17) from the wrongdoer's disrespect for the value enshrined in the law, in the sense of a ‘demonstrated unwillingness to be guided by that value in acting’.53 But to speak of justice in this context raises wider questions of justification. It assumes that the applicable criminal law has an acceptable moral content and is the product of a sufficiently democratic political process. It also raises questions about the justice of punishing the types of people who tend to be convicted and sentenced— the unemployed and the otherwise disadvantaged. This is often seen as a particular drawback of sentencing systems that aspire to proportionality or ‘just deserts’. Since ‘proportionate’ sentences are merely reinforcing existing social inequalities, this does not achieve justice so much as confirm injustice. The point is an important one, but it is an argument against most other rationales for sentencing, too. A deterrent theory seeks to reinforce the value structure inherent in the criminal law. A rehabilitative theory would attempt to mould offenders’ behaviour towards compliance with the norms of the criminal law. A restorative theory might be concerned to achieve compensation or a reconciliation that restores the status quo ante. Moreover, a system based on desert and proportionality can be operated humanely, without escalation of penalties.54 The justification for criminal law and punishment should therefore be sought in two dimensions —as a deserved response to culpable wrongdoing, and as a necessary institution to deter such wrongdoing. The former suggests that the criminal law, being society's strongest form of official censure and punishment, should be concerned only with major wrongs, affecting central values and causing significant harms. We have already noted that this is not the case in practice: the criminal law is sometimes used against relatively minor kinds of harm, often as a long-stop for regulatory systems. Thus, a perusal of the statute book would reveal that criminal offences are by no means limited to serious wrongs: the quasi-moral content of the criminal law is joined, indeed outnumbered, by the regulatory criminal law. In practice, therefore, the reach of the criminal sanction is determined by a number of conflicting social, political, and historical factors.55 This applies chiefly to the range of offences, enlarged on occasion in response to a wave of political concern and without overall consideration of the proper limits of the criminal sanction (or of the European Convention on Human Rights):

Page 13 of 21

Criminal Justice and the Criminal Law prominent examples are the Criminal Justice and Public Order Act 1994, the Prevention of Terrorism Act 2005, and the Serious Crime Act 2007. The scope of criminal liability also bears witness to similar conflicts (e.g. over the extent of the offence of conspiracy); the conditions of criminal liability barely conceal the conflicts of principle and policy which have shaped them throughout their years of development, with recurrent debate over the merits of fault requirements necessitating subjective awareness by the defendant, as against the justifications for objective standards of liability, based on what the reasonable person should have appreciated in the defendant's position. (p. 18) 1.6 The criminal law and sentencing A person who has been found guilty of a criminal offence is liable to be sentenced by the court. A conviction may be bad enough in itself: it is a form of public censure, and many convictions (at least for non-motoring offences) make it difficult or impossible to obtain certain jobs or enter a profession. For most of the offences covered in this book, the sentence is likely to involve considerable deprivation, either of money or of liberty or both. The criminal law may therefore be said to open the way for coercive official sanctions against an offender. Indeed, we will see below that sentencing has considerable significance for the contours of criminal liability: when Parliament creates a crime it authorizes not merely the affixing of a label of censure on the perpetrator, but also the imposition of certain deprivations by means of sentence. The range of sentences available to English courts, and the actual exercise of judicial discretion in imposing sentences, can be outlined only briefly here.56 The law of sentencing was consolidated in the Powers of Criminal Courts (Sentencing) Act 2000, but has since been altered in major ways by the Criminal Justice Act 2003 and other legislation. In brief, an absolute or conditional discharge may be thought sufficient for the least serious crimes or where the defendant has very strong mitigation. For many offences a fine will be the normal punishment: the size of the fine should reflect the seriousness of the offence, adjusted in accordance with the means of the offender.57 If the offence is serious enough to warrant it, the court may consider imposing a community sentence: that sentence may contain one or more of twelve separate requirements, including a requirement to do unpaid work, a requirement to undergo drug treatment, and a curfew reinforced by electronic monitoring. The most severe sentence is a custodial one, and a custodial sentence should be imposed only where the offence or offences are so serious that neither a fine alone nor a community sentence can be justified.58 The length of any custodial sentence ‘must be for the shortest term … that in the opinion of the court is commensurate with the seriousness of the offence’.59 If the sentence is for up to two years, the court may suspend it and may require the offender to comply with certain requirements during the supervision period. Sentencing guidelines have been handed down in Court of Appeal judgments since the 1980s, but the main source of English guidelines is now the Sentencing Council (replacing the Sentencing Guidelines Council and Sentencing Advisory Panel), and guidelines have recently been issued on drug offences and burglary.60 These are intended to guide the courts in setting the length of sentence for different types of offence. The high use of imprisonment continues to be a contentious (p. 19) issue: England and Wales have one of the highest rates of imprisonment in Europe, and in autumn 2012 the prison population stood at just over 86,000 (compared with 42,000 in early

Page 14 of 21

Criminal Justice and the Criminal Law 1993). How necessary is an understanding of sentencing law and practice to a study of the criminal law? Ideally the criminal law would be studied in conjunction with the other elements of criminal justice with which it is so intimately linked in practice and in theory—prosecution policy and other aspects of pre-trial criminal process, the laws of evidence, and sentencing too.61 This makes the case for two years of teaching in this field, or for four one-semester modules, to cover the ground. Given that this would be difficult to achieve in many institutions, because of other pressures on the curriculum, many criminal law teachers are left with a decision about the amount of sentencing or other matters to put into criminal law modules. At a minimum, the interactions between sentencing and criminal law must be kept in view. It is sentencing, largely, that gives the criminal law its bite, and so decisions on criminal liability should be viewed as decisions about the application of censure and coercion. An example of the impact of sentencing on the criminal law is that, as we shall see in Chapter 7, the shape of the law on murder and manslaughter has been influenced by the existence of the mandatory penalty for murder. Another area of interaction between criminal law and sentencing is where the courts or Parliament have taken a restrictive approach in defining defences to criminal liability, in the belief that circ*mstances which ‘almost’ amount to a defence might result in significant mitigation of sentence (e.g. some cases of duress and entrapment). On the other hand, courts have tended to adopt a much looser notion of responsibility at the sentencing stage than at the liability stage. Thus the criminal law itself proclaims individual responsibility for actions, maintaining strict standards of conduct and setting its face publicly against the idea that social or other circ*mstances can excuse behaviour, whilst at the sentencing stage courts do recognize from time to time the exculpatory force of preceding or surrounding circ*mstances.62 The aims of sentencing are not simply part of the background of the criminal law: they have implications for the shape of the criminal law itself. Thus proportionality should be a key element in the structure of the criminal law. It is a major function of the criminal law not only to divide the criminal from the non-criminal, but also to grade offences and to label them proportionately. As Nils Jareborg expresses it, ‘the threat of punishment is not only a conditional threat of a painful sanction. It is also an official expression of how negatively different kinds of action or omission are judged’.63 At the level of judicial sentencing, the Criminal Justice Act 2003 requires courts to ‘have regard to’ some five different purposes of sentencing—punishment, deterrence, rehabilitation, public protection, and reparation.64 Not only do these conflict among (p. 20) themselves, but government-sponsored research demonstrates, for example, that the evidence for the effectiveness of deterrent sentences is unpromising.65 The sentencing guidelines endeavour to save sentencing practice from the inconsistency that a ‘pick-and-mix’ approach to the purposes of sentencing might produce by emphasizing that s. 143(1) of the Act insists upon proportionality of sentencing.66 Thus proportionality should have a central role, not only at the legislative stage of grading offences in the criminal law, but also at the judicial stage of passing sentence in individual cases. Within ‘desert’ theory there is a distinction between two kinds of proportionality.67 One is cardinal proportionality, which requires that the severity of the punishment be in proportion to the seriousness of the offence. Exactly what the level of sentences should be remains a matter for debate, taking account of criminological research and, it is submitted, of the principle of restraint in the use of custody. The second kind of proportionality, more important for our present purposes, is ordinal proportionality. This requires an assessment of the seriousness of

Page 15 of 21

Criminal Justice and the Criminal Law the crime in relation to other forms of offending, so as to establish acceptable relativities. The precise meaning of ‘seriousness’ here will be explored in Chapters 3 and 5, for it represents a combination of the wrong or harm done or risked and of the culpability of the offender. This book is intended as an exploration of principles of the criminal law. It does not purport to be a textbook, and does not treat all the parts of the criminal law. Nor does it attempt to convey a general description of the criminal law in action. There are perhaps over ten thousand offences in English criminal law, the bulk of them being ‘regulatory’ offences of strict liability with relatively low penalties. Although strict liability offences are discussed in Chapter 5.5(a) and in other places, the focus of the book is upon offences with substantial penalties. In this book it is principles for offences and for defences that provide the vehicle of study. In Chapter 3, many of the principles which do or should inform the criminal law are drawn together for discussion: this is largely a normative exercise, raising questions about what principles should or should not determine the boundaries of criminal liability. It does not purport to be descriptive or historical, although it gives considerable prominence to the European Convention on Human Rights as a source. On the other hand, any meaningful discussion of principles must be connected to the kinds of laws that have been and are introduced, and the book's approach of examining the law through principles begins in Chapter 2, where the approach to creating new crimes is discussed. The idea is to analyse essentially political decisions to create new crimes, in terms of a more principled approach to the reach of the criminal sanction. Principles which have a particular bearing on culpability and the conditions of liability are discussed in Chapters 4 to 6. They are kept in mind when analysing three different groups of offences in Chapters 7 to 9. They are also related to questions about the scope of criminal liability in Chapters 10 and 11. (p. 21) There are also frequent references to research that has a bearing on criminal justice, to give some indication of the social context in which the criminal law operates. Much more coverage could be given to these contextual issues, such as enforcement policy, police powers, the pre-trial construction of cases, and sentencing, but the primary focus of this work is upon the consideration of doctrine. This means that the discussion of the criminal law is largely centred on appellate courts, as opposed to concentrating on the law as it is enforced by the police and others. There is an endeavour to recognize the constitutional responsibilities of the courts in developing the law and interpreting legislation. There is also an endeavour to remain alert to the implications for law enforcement of leaving areas of discretion when formulating laws: we saw in section 1.4 how wide this discretion often is, and how the selective policies of various enforcement agencies lead to a rather skewed sample of defendants appearing before the courts. But the centrepiece of this book is the doctrine of the criminal law, by which is meant the policies and social values which underlie decisions to increase or decrease the range of the criminal law, the principles and values which bear upon decisions about the scope of criminal liability, and the principles, policies, and values which relate to the conditions of criminal liability.

Further reading N. LACEY AND L. ZEDNER , ‘Legal Constructions of Crime’, in M. MAGUIRE , R. MORGAN , AND R. REINER (eds), Oxford Handbook of Criminology (5th edn., 2012), ch 6.

Page 16 of 21

Criminal Justice and the Criminal Law A. ASHWORTH and L. ZEDNER , ‘Defending the Criminal Law: Reflections on the Changing Character of Crime, Procedure and Sanctions’, (2008) 2 Criminal Law and Philosophy 21. G. LAMOND , ‘What is a Crime?’, (2007) 27 OJLS 609. N. LACEY , C. WELLS, AND O. QUICK , Reconstructing Criminal Law (4th edn., 2010 by C. WELLS and O. QUICK), chs 1–3.

Notes: 1 The form of English criminal laws is not usually that of a prohibition, however. Laws are

usually written as if addressed to police, prosecutors, and courts: ‘a person is guilty of x if …’, or ‘any person who does x, y, z shall be liable to imprisonment for a term not exceeding …’ 2 Crown Prosecution Service, Code for Crown Prosecutors (6th edn., 2010), paras. 4.18–19; a

victim does have the right to bring a private prosecution, but under the Prosecution of Offences Act 1985 the Director of Public Prosecutions has the power to take it over and, if appropriate, to drop it. 3 R. A. Duff, Punishment, Communication and Community (2001), 62. 4 G. Lamond, ‘What is a Crime?’ (2007) 27 OJLS 609, at 615–20. 5 But see n 22. 6 Consolidated Criminal Practice Direction (Ministry of Justice, 2011), part III.28; the leading

case is Nunn [1996] 2 Cr App R (S) 136. 7 See further A. Ashworth, ‘Rights, Responsibilities and Restorative Justice’ (2002) 40 B J Crim

578. 8 G. Williams, ‘The Definition of a Crime’ [1955] CLP 107, at 130. 9 See e.g. Engel v Netherlands (1976) 1 EHRR 647. 10 For an English example, see Benham v United Kingdom (1996) 22 EHRR 293. 11 See further A. Ashworth, ‘Is the Criminal Law a Lost Cause?’ (2000) 116 LQR 225. 12 Where there are separate procedures for administrative offences, fixed penalties, etc., it

must be open to the defendant to have the case heard in a criminal court (with all the safeguards) if desired: Le Compte, Van Leuven and De Meyere v Belgium (1981) 4 EHRR 1. 13 This is the power, exercised by the European Court of Human Rights, to treat the term

‘criminal charge’ as having an ‘autonomous meaning’, i.e. the court looks at the substance of the proceedings and not at the label placed upon it by the domestic legislature. See n 9. 14 R (McCann v Manchester Crown Court; Clingham v Kensington and Chelsea London

Borough Council [2003] 1 AC 787. 15 See further E. Burney, Making People Behave: Anti-Social Behaviour, Politics and Policy

Page 17 of 21

Criminal Justice and the Criminal Law (2nd edn., 2009); P. Ramsay, The Insecurity State (2012); A. von Hirsch and A. P. Simester (eds), Incivilities (2007). 16 See further A. Ashworth and L. Zedner, ‘Defending the Criminal Law: Reflections on the

Changing Character of Crime, Procedure and Sanctions’ (2008) 2 Criminal Law and Philosophy 21. 17 A. Ashworth and M. Blake, ‘The Presumption of Innocence in English Criminal Law’ [1996]

Crim LR 306. 18 For more detailed treatment, see B. Emmerson, A. Ashworth, and A. Macdonald (eds),

Human Rights and Criminal Justice (3rd edn., 2012), particularly chs 16, 17, and 18. 19 See further E. Baker, ‘Taking European Criminal Law Seriously’ [1998] Crim LR 361, and S.

Peers, EU Justice and Home Affairs Law (3rd edn., 2010), ch 8. 20 C. Kershaw et al., Crime in England and Wales 2007–08 (Home Office Statistical Bulletin

07/08), 39, estimating that 42 per cent of crimes are reported to the police. The overall figure masks considerable variation: thefts of vehicles are the most likely to be reported (93 per cent), and theft from the person, vandalism, and assault with no injury least likely (32 per cent, 35 per cent, and 34 per cent in 2003–4). See now Crime in England and Wales, year ending June 2012 (Office of National Statistics). 21 See the research by C. Clarkson et al., ‘Assaults: the Relationship between Seriousness,

Criminalisation and Punishment’ [1994] Crim LR 4. 22 See the findings of D. Brown and T. Ellis, Policing Low-Level Disorder: Police Use of section

5 of the Public Order Act 1986 (Home Office Research Study No. 135, 1994). 23 See K. Hawkins, Law as Last Resort (2002); B. Hutter, Compliance: Regulation and

Environment (1997); and B. Hutter, Regulation and Risk: Occupational Health and Safety on the Railways (2001). 24 Kershaw et al, 169 (reference at n 20). 25 Royal Commission on Criminal Justice, Report (Cm 2263 (1993)), 57. 26 Sections 34–7 of the Criminal Justice and Public Order Act 1994, the impact of which is

discussed in A. Ashworth and M. Redmayne, The Criminal Process (4th edn., 2010), ch 4. 27 For discussion, see C. Ball, ‘Youth Justice? Half a Century of Responses to Youth Offending’

[2004] Crim LR 167. 28 Home Office, No More Excuses: a New Approach to Tackling Youth Crime in England and

Wales (Cm 3809 (1997)), 1. 29 See further 〈www.justice.gov.uk/youth-justice〉. 30 For fuller discussion, see A. Ashworth and M. Redmayne, ch 6 (reference at n 26). 31 M. Maguire, R. Morgan, and R. Reiner (eds), Oxford Handbook of Criminology (5th edn.,

2012), ch 32. 32 Page 18 of 21

Criminal Justice and the Criminal Law 32 Environment Agency, Enforcement and Prosecution Policy (issued 4 January 2011), 2, at

〈www.environment-agency.gov.uk〉. 33 See 〈www.hmrc.gov.uk/prosecutions/crim-inv-policy.htm〉. 34 See further Ashworth and Redmayne, 160–3 (reference at n 26). 35 This may be found at 〈www.cps.gov.uk〉. 36 Home Office, Digest 4: Information on the Criminal Justice System in England and Wales

(1999), 29. 37 See Ashworth and Redmayne, 187–90 (reference at n 26). 38 T. Bucke and Z. James, Trespass and Protest: Policing under the Criminal Justice and

Public Order Act 1994 (Home Office Research Study No. 190, 1998). 39 Ashworth and Redmayne, at 298 (reference at n 26); cf. Code for Crown Prosecutors, para.

7. 40 Sentencing Guidelines Council, Reduction in Sentence for a Guilty Plea: Revised Guideline

(2007). 41 For fuller discussion and references, see Ashworth and Redmayne, ch 10 (reference at n

26). 42 See Criminal Justice Act 2003, Sch 3, amending s. 20 of the Magistrates’ Courts Act 1980;

and Goodyear [2005] 3 All ER 117. 43 The complexities are explored in C. Hedderman and L. Gelsthorpe (eds), Understanding the

Sentencing of Women (Home Office Research Study No. 170, 1997). 44 C. Phillips and B. Bowling, ‘Ethnicities, Racism, Crime and Criminal Justice’, in M. Maguire, R.

Morgan, and R. Reiner (eds), Oxford Handbook of Criminology (5th edn., 2012). 45 See generally, M. McConville et al., Standing Accused (1994). 46 See the fascinating study by P. Rock, The Social World of an English Crown Court (1993). 47 See further A. Ashworth, ‘Is the Criminal Law a Lost Cause?’ (2000) 116 LQR 225. 48 Anarchist theorists such as Godwin held a version of this view (see G. Woodco*ck,

Anarchism (1962)), and according to Marxist theory the law would wither away on the attainment of perfect communism (see H. Collins, Marxism and Law (1982)). 49 Usually cited in this connection are the police strikes in Liverpool in 1919, and Melbourne in

1918, and the incarceration of the Danish police force by the Nazis in 1944. Broadly speaking, property crimes (looting) tend to increase whereas sexual and violent crimes do not. 50 See R. V. G. Clarke, Situational Crime Prevention: Successful Case Studies (2nd edn.,

1997). 51 Page 19 of 21

Criminal Justice and the Criminal Law 51 See generally A. Crawford and K. Evans, ‘Crime Prevention and Community Safety’, in M.

Maguire, R. Morgan, and R. Reiner (eds), The Oxford Handbook of Criminology (5th edn., 2012), ch 26. 52 See e.g. A. von Hirsch et al., Criminal Deterrence and Sentence Severity: an Analysis of

Recent Research (1999), and P. Robinson and J. Darley, ‘Does Criminal Law Deter? A Behavioural Science Investigation’ (2004) 24 OJLS 173. 53 G. Lamond, ‘What is a Crime?’ (2007) 27 OJLS 609, at 622. 54 See A. von Hirsch and A. Ashworth, Proportionate Sentencing (2005), chs 5 and 6. 55 N. Lacey and L. Zedner, ‘Legal Constructions of Crime’, in M. Maguire, R. Morgan, and R.

Reiner (eds), Oxford Handbook of Criminology (5th edn., 2012), ch 6. 56 For fuller discussion, see A. Ashworth, Sentencing and Criminal Justice (5th edn., 2010). 57 Powers of Criminal Courts (Sentencing) Act 2000, s. 128. 58 Criminal Justice Act 2003, s. 152(2). 59 Criminal Justice Act 2003, s. 153(2). 60 See 〈www.sentencing-council.org.uk〉. 61 Cf. P. Alldridge, ‘What's Wrong with the Traditional Criminal Law Course?’ (1990) 10 Legal

Studies 38. 62 See Chapter 6.8, and A. Norrie, Crime, Reason and History (2nd edn., 2001), ch 10. 63 N. Jareborg, ‘The Coherence of the Penal System’, in his Essays in Criminal Law (1988). 64 Criminal Justice Act 2003, s. 142(1). 65 See references in n 52. 66 Sentencing Guidelines Council, Overarching Principles (2004). 67 See further A. von Hirsch and A. Ashworth, Proportionate Sentencing (2005), ch 9.

Criminalization

Principles of Criminal Law (7th edn) Andrew Ashworth and Jeremy Horder Publisher: Oxford University Press Print ISBN-13: 9780199672684 DOI: 10.1093/he/9780199672684.001.0001

Print Publication Date: May 2013 Published online: Sep 2013

2. Criminalization Chapter: (p. 22) 2. Criminalization Author(s): Andrew Ashworth and Jeremy Horder DOI: 10.1093/he/9780199672684.003.0002

2.1 The principle of individual autonomy 2.2 The principle of welfare 2.3 The harm principle and public wrongs 2.4 The minimalist approach 2.5 Morally wrong behaviour 2.6 Remote harms 2.7 Conclusions and applications Further reading We have seen (in Chapter 1.5) that a system of criminal law may be justified as a mechanism for the preservation of social order. As a type of law, its technique is censuring: it requires a procedure that provides the defendant with certain safeguards, it may result in a conviction that gives D a criminal record, and, if so, it authorizes the infliction of State punishment. To criminalize a certain kind of conduct is to declare that it is a public wrong that should not be

Criminalization done, to institute a threat of punishment in order to supply a pragmatic reason for not doing it, and to censure those who nevertheless do it. This use of State power calls for justification— justification by reference to democratic principles, and justification in terms of sufficient reasons for invoking this coercive and censuring machinery against individuals and organizations. So serious are the potential consequences, and so significant the public censure, that the decision to criminalize conduct should not simply be made ‘on balance’. It will be argued here, following Doug Husak,1 that in a liberal State there is something equivalent to a right not to be punished, which should place the burden of proof firmly on those who would wish to turn non-criminal activity into an offence. However, it is not argued or assumed here that there exists some objective benchmark of criminality, or some general theory which will enable us to tell whether or not certain conduct ought to be criminalized. The range of actual and potential crimes is so wide and varied that this seems unattainable. We should abandon: ‘attempts to derive the content of the criminal law from a single master principle … [and] accept that debates about its scope will be more piecemeal, gradual affairs, more focused on particular offences (actual or suggested), and informed by a range of values, presumptions and considerations …’2 (p. 23) The chapter's purpose is to identify some general principles and values that, it is submitted, ought to be considered when deciding whether or not to make conduct criminal. Although it is true that the frontiers of criminal liability are not given but are historically and politically contingent, it remains important to strive to identify those interests that warrant the use of the criminal law and to refine notions such as harm and wrongdoing which play so prominent a part even in political discussions of these questions. This chapter will focus on legislative decisions to extend or curtail the criminal law. Yet we must recall, as argued in Chapter 1, that the practical impact of the criminal law on citizens is determined not so much by the legislature as by the workings of the various law enforcement agents—chiefly police officers, but also officials from HMRC, the various statutory inspectorates, and so on. Thus the legislature may be said to provide the tools, resources, or authority for law enforcement agents when it creates a criminal offence, but decisions about when to invoke and when not to invoke the available powers are taken by enforcement officers. The exercise of discretionary power therefore provides the key to practical instances of criminalization. The chapter begins with a discussion of the two fundamental principles of autonomy and of welfare. It then develops the principle of minimalism, centred on the right not be subjected to punishment and also including consequentialist considerations such as not using the criminal law when it would be ineffective or counter productive. Then follows a discussion of the harm principle and the principle of publicly wrongful conduct, leading to an assessment of the arguments for and against punishing ‘immoral’ behaviour, for including paternalistic laws, and for criminalization based on remote harms.

2.1 The principle of individual autonomy One of the fundamental concepts in the justification of criminal laws is the principle of

Page 2 of 22

Criminalization individual autonomy—that each individual should be treated as responsible for his or her own behaviour. This principle has factual and normative elements that must be explored, briefly, in turn. The factual element in autonomy is that individuals in general have the capacity and sufficient free will to make meaningful choices. Whether this is true cannot be demonstrated conclusively. Over the centuries the ‘free will’ argument has been contradicted by the ‘determinist’ claim that all human behaviour is determined by causes that ultimately each individual cannot control. There is an immense literature on these issues, which cannot be examined here.3 Most philosophers arrive at compromise positions which enable them to accept the fundamental proposition that behaviour is not so determined that blame is generally unfair and inappropriate, and yet to accept that, in (p. 24) certain circ*mstances, behaviour may be so strongly determined (e.g. by threats from another) that the normal presumption of free will may be displaced. Similar in many ways is the ‘principle of alternative possibilities’, according to which an individual may properly be held responsible for conduct only if he or she could have done otherwise.4 In support of these approaches is the fact that most of everyday life is conducted on the basis of such beliefs in individual responsibility, and that in the absence of proof of determinism we should not abandon those assumptions of free will that pervade so many of our social practices. However, as Barbara Hudson has warned: the notion of free will that is assumed in ideas of culpability … is a much stronger notion than that usually experienced by the poor and powerless. That individuals have choices is a basic legal assumption: that circ*mstances constrain choices is not. Legal reasoning seems unable to appreciate that the existential view of the world as an arena for acting out free choices is a perspective of the privileged, and that potential for selfactualization is far from apparent to those whose lives are constricted by material or ideological handicaps.5 This point may be conceded without gainsaying the fundamental assumption of free will, so long as the possibility of qualifications is recognized. Thus, for example, the capacities assumed by the law may not be present in those who are too young or who are mentally disordered. These capacities relate to what R. A. Duff terms the ‘preconditions of criminal liability’, preconditions that he goes on to connect with the ability to participate in a trial as a communicative enterprise.6 The general assumption is thus that sane adults may properly be held liable for their conduct and for matters within their control, except in so far as they can point to some excuse for their conduct—for example, duress, mistake, or even social deprivation, as discussed in Chapter 6. No less important a part of the principle of autonomy is its normative element: that individuals should be respected and treated as agents capable of choosing their acts and omissions, and that without recognizing individuals as capable of independent agency they could hardly be regarded as moral persons.7 Some such principles lie at the centre of most liberal political theory, and can be found, for example, in Ronald Dworkin's principle that each individual is entitled to equal concern and respect.8 The principle of autonomy assigns great importance to liberty and individual rights in any discussion of what the State ought to do in a given situation. Indeed, a major part of its thrust is that individuals should be protected from official censure, through the criminal law, unless they can be shown to have chosen the conduct for which they (p. 25) are being held liable.9 This, as we shall see in section 2.4, is a central element in

Page 3 of 22

Criminalization the ‘defensive’ approach to criminalization advanced by Nils Jareborg and others, insisting on the importance of protecting individuals from undue State power. We will also see that H. L. A. Hart's famous principle, that an individual should not be held criminally liable unless he had the capacity and a fair opportunity to do otherwise, is grounded in the primary importance of individual autonomy.10 On the other hand, returning to the scope of criminalization, this emphasis on individual choice militates against creating offences based on paternalistic grounds, as argued in section 2.5. If autonomy is to be respected, the State should leave individuals to decide for themselves and should not take decisions ‘in their best interests’. In liberal theory, the principle of autonomy goes much further than this. Thus Joel Feinberg, towards the end of his discussion of autonomy, states that: the most basic autonomy-right is the right to decide how one is to live one's life, in particular how to make the critical life-decisions—what courses of study to take, what skills and virtues to cultivate, what career to enter, whom or whether to marry, which church if any to join, whether to have children, and so on.11 The difficulty is to decide how far this is to be taken. Whilst the principle of autonomy gives welcome strength to the protection of individual interests against collective and State interests, it seems less convincing in other respects. The question ‘whose autonomy?’ must always be asked: the criminal law is often claimed to be neutral, and yet certain forms of bias—such as gender bias12 —may be evident in the law's assumptions and reasoning. In some of its formulations the principle of autonomy pays little or no attention to the social context in which all of us are brought up (which may both restrict and facilitate the pursuit of certain desired ends) and the context of powerlessness in which many have to live.13 The idea that individuals should be free to choose what to do cannot be sustained without wide-ranging qualifications. A developed autonomy-based theory should find a central place for certain collective goals, seen as creating the necessary conditions for maximum autonomy. Thus Joseph Raz argues that: Three main features characterize the autonomy-based doctrine of freedom. First, its primary concern is the promotion and protection of positive freedom which is understood as the capacity for autonomy, consisting of the availability of an adequate range of options, and of the mental abilities necessary for an autonomous life. Second, the State has the duty not merely to prevent the denial of freedom, but also to promote it by creating the conditions of (p. 26) autonomy. Third, one may not pursue any goal by means which infringe people's autonomy unless such action is justified by the need to protect or promote the autonomy of those people or of others.14 This third feature proposes a minimalist approach to the use of the criminal law, and all three features reappear when we consider the principle of welfare.

2.2 The principle of welfare We have seen that an individualist principle of autonomy is too limited, and that Raz and others have therefore developed an approach which emphasizes the State's obligation to create the social conditions necessary for the exercise of full autonomy by individual citizens.15 Modern communitarian theorists have gone further, often emphasizing the centrality

Page 4 of 22

Criminalization of collective goals. Thus Nicola Lacey describes the principle of welfare as including ‘the fulfilment of certain basic interests such as maintaining one's personal safety, health and capacity to pursue one's chosen life plan’.16 The specification of the interests to be thus protected should be a matter for democratic (participatory) decision-making: this means both that the interests will be objectively determined, not just according to the preference of each individual, and also that individuals whose preferences are at odds with those of the majority will lose out. Both of these are familiar features of social life. But much depends on how the notion of ‘community’ is developed. It ought to respond to the problems of differential power and privilege raised by Hudson.17 The danger is that the notion of ‘community’ may sometimes be invoked in support of policies emphasizing public safety and the need for greater measures to ensure security: much as it is important to pursue collective goals such as environmental protection, public safety, and so on, the result may be to promote an idea of community without any special weighting of individual rights. This may produce harsh and intrusive policies—a tendency against which Lacey, in her development of communitarian perspectives, issues a strong warning.18 Whereas the principle of autonomy seems to suggest that individual rights should be given high priority in the legal structure, the principle of welfare recognizes the social context in which the law must operate and gives weight to collective goals.19 Clearly there are conflicts between the two principles, but that may not always be the case. If the principle of autonomy is taken to require a form of positive liberty (freedom to pursue one's goals, etc.) rather than merely negative liberty (freedom from attack, etc.), then the principle of welfare may work towards the same end by ensuring that citizens (p. 27) benefit from the existence of facilities and structures which are protected, albeit in the last resort, by the criminal law. Some criminalization may therefore be accepted as the only justifiable means of upholding certain social practices as ‘necessary for the general good’. Matters such as the obligation to state one's income accurately for the purpose of taxation or for the receipt of benefits can hardly be analysed convincingly in terms of individual autonomy: once a public decision has been made about the system to be adopted, it may be justifiable for at least egregious departures from these rules to be criminalized. The same may be said of laws relating to industrial safety, food safety, environmental protection, and so on. Although it remains to be decided whether violations of these norms should be criminalized or dealt with in some other way, the legitimacy of some criminalization on the basis of welfare as well as on the basis of autonomy cannot be put in doubt.20 Those versions of the principle of autonomy which suggest that individuals should remain free to decide these matters according to their own preferences are not sustainable. Yet the value of individual autonomy as a restraint upon collective and State action should not be overlooked. Decisions by the wider community may threaten basic interests of individuals, unless there is recognition of a set of protected rights. The significance of the Human Rights Act 1998 is that it imported into English law a set of protected rights—those declared in the European Convention. As we will see in Chapter 3.3, the Convention rights are weighted differently: some are almost absolute, some are strong rights (from which derogation may be permitted in defined circ*mstances), and others are qualified rights, defined in a way that allows certain restrictions on them where this is adjudged ‘necessary in a democratic society’. In a sense, the group of qualified rights (and, to some extent, the interpretation of the other rights) demonstrates both the inevitability of conflicts between autonomy and welfare and the possibility of devising a procedure for ‘resolving’ those conflicts—although the satisfactoriness

Page 5 of 22

Criminalization of the Convention approach will be considered in Chapter 3.3 and at appropriate points in later chapters. On any realistic view, the principles of autonomy and welfare have a degree of mutual interdependence, which should be recognized and structured.21 However, that should not lead to a vague notion of ‘balancing’ the two principles. Rather, it should lead to the development of ways of prioritizing some rights, and of the structuring of public interest arguments so as to ensure that they meet criteria such as urgency, unavoidability, effectiveness, and so forth.22 (p. 28) 2.3 The harm principle and public wrongs The traditional starting point of any discussion of criminalization is the ‘harm principle’. Its essence is that the State is justified in criminalizing any conduct that causes harm to others or creates an unacceptable risk of harm to others. John Stuart Mill's statement of the principle was that ‘the only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others’. Its main thrust is as a negative or limiting principle, having the objective of restricting the criminal law from penalizing conduct that is regarded as immoral or otherwise unacceptable but which is not harmful to others.23 It is developed by Joel Feinberg in his detailed rejection of ‘legal paternalism’ and ‘legal moralism’ as sufficient reasons for criminalization.24 However, one cannot proceed far without adopting a definition of harm: can a satisfactory line be drawn through such things as physical harm, harm to property, harm to feelings, and indirect harm? The question is both fundamental and somewhat intractable. As Neil MacCormick argues: ‘harm’ is itself a morally loaded (and essentially contested) concept … . Nothing … could be more obviously a moral question than the question whether individual interests in private property are always, sometimes, or never legitimate. The issue of the justice of systems of private property is a central one in the great clash of ideologies in the contemporary world.25 Thus if we wish to define harm in terms of violations of people's legitimate interests, we must remain conscious of the moral, cultural, and political nature of the interests recognized in a particular system. Feinberg's definition of harm as ‘those states of set-back interest that are the consequence of wrongful acts or omissions of others’26 does not conclude these wider issues.27 If the harm principle is primarily negative in its impact, by what route can we move forward to consider some positive principles for criminalization? Feinberg's definition includes both harm and wrongfulness, and it is necessary to attend to both of these elements. When it comes to stating a positive version of the harm principle, Feinberg proposes the following definition: It is always a good reason in support of penal legislation that it would probably be effective in preventing (eliminating, reducing) harm to persons other than the actor and there is probably no other means that is equally effective at no greater cost to other values.28 (p. 29) This formulation is an important step away from the apparently exclusionary phraseology used by Mill (‘the only purpose …’)29 and towards the identification of a range of relevant reasons for and against criminalization. For Feinberg, other relevant matters are the

Page 6 of 22

Criminalization gravity of the possible harm, its degree of probability, and the social value of the (otherwise dangerous) conduct.30 His definition of the harm principle also incorporates, through its references to effectiveness and cost, some of the limiting principles encompassed by the minimalist approach described at section 2.4. The other important element, in addition to harm, is wrongfulness. It is not the causing of harm that alone justifies criminalization, but the wrongful causing of harm—wrongful in the sense of culpably assailing a person's interests, or abusing them by using them as a means to another's satisfaction. Thus Feinberg's elaboration of the concept of harm requires, and extends only to, ‘setbacks of interests that are wrongs, and wrongs that are setbacks to interest’.31 Similarly, Andrew Simester and Andreas von Hirsch develop the argument that one necessary prerequisite of criminalization is that the conduct amounts to a moral wrong.32 This does not imply that this is a sufficient condition, merely that it is a prerequisite, since there may be morally wrong conduct (e.g. forms of lying or sexual infidelity) which, for other reasons, should not be criminalized. Nor does this rule out the criminalization of what might be termed coordination offences, i.e. those offences necessary to regulate an activity such as driving, where the law must make certain determinations (e.g. that drivers should drive on the left) which then, through their instrumental value, impart moral force to related requirements.33 However, before criminalization is justified, not only must the conduct be morally wrong, but there must also be no strong countervailing considerations, such as the absence of harm, the creation of unwelcome social consequences, the curtailment of important rights, and so forth.34 Indeed, given the content and consequences of public censure and punishment—in terms of restrictions on, and even deprivations of, basic liberties—there is a strong case for restricting the criminalization of non-serious moral wrongs. That is straightforward in a system that has a lesser category of ‘administrative offences’ with low penalties, which can be used to penalize public wrongs that do not attain the appropriate level of seriousness to be criminalized. A third aspect, in addition to harmfulness and wrongfulness, is the public element in wrongs. One manifestation of this consists of those general obligations of citizens that are so important that the criminal sanction may be justified to reinforce them. A core of offences against State security may be justified on these grounds, as may some offences against the taxation and benefits system, so long as the limiting effect of the (p. 30) minimalist principle (see 2.4) is kept in view. These are public wrongs, inasmuch as the victim is not an individual but the community as a whole, and it is right that the more serious among them are considered suitable for criminalization—not least where the gain or advantage obtained is as great as, or greater than, that obtained in the typical offence with an individual victim. But it is to the latter kind of offence that we must now turn, and this is where the public element becomes problematic. How can we tell which wrongs done to individuals are sufficiently ‘public’ to warrant the condemnation of the criminal law? As Antony Duff argues, the answer lies not in an aspect of the wrong itself, but in the public valuation of the wrong: We should interpret a ‘public’ wrong, not as a wrong that injures the public, but as one that properly concerns the public, ie the polity as a whole … A public wrong is thus a wrong against the polity as a whole, not just against the individual victim: given our identification with the victim as a fellow citizen, and our shared commitment to the values that the rapist violates, we must see the victim's wrong as also being our wrong.35

Page 7 of 22

Criminalization The public element does not have anything to do with location: unkind remarks made to a friend in public would not ‘concern the public’ unless they tended to provoke a breach of the peace, and a very public breaking of a promise to attend a certain event may not be regarded as sufficiently important for the polity as a whole to be required to take action. Contrast those instances with domestic violence (e.g. a substantial beating) which, even if it occurs entirely in the private realm of a home, is a moral and social wrong that the community should regard as a wrong that ought to be pursued through the public channels of prosecution and trial.36 Thus, as Grant Lamond has argued, the question is whether the community is appropriately responsible for punishing these wrongs.37 The supporting argument here is presumably that the State should protect and promote the basic value of security and freedom from physical attack by prosecuting assaults wherever they occur (leaving aside questions of consent and its proper limits),38 and that the fact that an assault occurs in a domestic context should make no difference to this. It does not follow from this that adultery is a good candidate for criminalization, harmful and wrongful though it may be in many instances, since the question is whether the value of marriage as an institution is so central and fundamental to the political community that the State is expected to prosecute through the criminal law those whose conduct threatens it. One example of the public element of a wrong is the creation of racially and religiously aggravated offences of assault, harassment, and so forth.39 Calling someone insulting names is not usually a criminal offence, but the rationale for these crimes (which provide for greater punishment when, inter alia, the offence is accompanied by (p. 31) racist or religious insults) is clearly connected with a belief that it is proper for the State to promote the basic value of racial tolerance and that this value is so significant as to justify criminalization. Thus these offences can be regarded as harmful public wrongs. However, the argument in this section does not suggest that there are sharp dividing lines that tell us whether or not certain conduct is sufficiently harmful, or sufficiently wrong, or has a sufficiently public element, to justify criminalization. Rather, the point is that these should be recognized as the appropriate kinds of argument in support of a decision to criminalize. It is submitted in the next section that those arguing in favour of criminalization should bear the burden of proof, so the contention here is that the three elements identified above—harm, wrongdoing, and the public element—are what should be proved. Finally, there is one respect in which the notion of criminalizing harmful public wrongs is less indeterminate, and that is where the State has a positive obligation to protect certain fundamental rights. In the English context this means positive obligations under the European Convention on Human Rights. Thus the right to life is guaranteed by Art. 2, and states must ensure that they protect this by having, for example, appropriately restricted rules of selfdefence.40 Article 3 declares that no one shall be subjected to torture or to inhuman or degrading treatment, and this requires states to ensure that their laws give adequate protection, for example, to children from physical beatings (of a certain magnitude) by their parents.41 Article 8 declares the right to respect for private life, which includes sexuality and sexual relations, and this means that the criminal law ought not to discriminate against different forms of sexual orientation and that it should provide protection from sexual molestation, particularly for the young and mentally impaired.42

2.4 The minimalist approach

Page 8 of 22

Criminalization The minimalist approach is based on a particular conception of the criminal law and its relationship to the principles of autonomy and welfare and to other forms of social control. Its four main components are a) the principle of respect for human rights, b) the right not to be subjected to State punishment, c) the principle that the criminal law should not be invoked unless other techniques are inappropriate, and d) the principle that conduct should not be criminalized if the effects of doing so would be as bad as, or worse than, not doing so. Each of these is now discussed in turn. (p. 32) (a) The principle of respect for human rights The first point is that a minimalist approach to criminalization should respect human rights protections. Thus, for example, any criminal laws should respect freedom of expression, freedom of assembly and association, freedom of thought and religion, the right of privacy, and the right not to be discriminated against in any of those four rights. Under the Convention this does not mean that no criminal law may curtail or abridge one of those rights: the first four rights (not the right against discrimination) are all qualified rights, which means that interference with them is permissible if it is ‘necessary in a democratic society’ for one of the stated purposes. Thus freedom of expression may be curtailed by an offence of sending a grossly offensive message through a public communication system,43 by offences of speech likely to stir up racial or religious hatred,44 or by offences of inciting violence. There are bound to be difficult borderline decisions to be taken, as where an evangelical Christian was convicted under s. 5 of the Public Order Act 1986 for displaying a sign saying ‘Stop Immorality, Stop hom*osexuality, Stop Lesbianism’, the court concluding that the interference with his rights to freedom of religion and freedom of expression was justified by the disorder and violence it provoked.45 (b) The right not to be punished Husak argues that we should recognize a right not to be subjected to State punishment, and that this flows from the social significance of the public censure involved in conviction and from the sacrifice of human rights usually entailed by the imposition of punishment.46 What this means in practice is that the decision to criminalize, and therefore to authorize punishment, should be recognized as being of a different order from many other legislative decisions. It is different from taxation, or from the creation of a regime of administrative regulation over a certain activity, important as those kinds of decision are. The element of public censure and the overriding of other rights means that strong justifications for criminalizing conduct are called for, and that the burden of proof should lie on those who would impose criminal liability. Moreover, this is not simply a threshold decision, of whether or not to criminalize. Much stronger justifications should be required where the offence is to be punishable with imprisonment,47 in view of the potential deprivation of the right to liberty, and even stronger justifications where the maximum sentence of imprisonment is substantial. A particular concern (p. 33) in this regard is the maximum sentence of five years’ imprisonment typically assigned to offences of breaching the terms of a civil preventive order, such as the anti-social behaviour order.48 (c) Criminalization as a last resort The criminal law is a censuring and preventive mechanism, but there are others. Morality, social convention, and peer pressure are three informal sources of control, and in many Page 9 of 22

Criminalization spheres it seems preferable to leave the regulation of certain unwelcome behaviour to those forces. Within the law itself, there are at least two other major techniques in addition to criminalization: there is civil liability, best exemplified by the laws of tort and contract, and also there is administrative regulation, which includes such measures as licensing and franchising. What considerations should determine the choice of technique? The minimalist's answer, drawing on the considerations in (b), would be that the law's most coercive and censuring technique (criminalization) should be reserved for the most serious invasions of interests.49 Less serious misconduct is more appropriately dealt with by the civil law, by administrative regulation, or even by introducing a new category of non-criminal financial levies. This approach has a straightforward utilitarian foundation, traceable back to Jeremy Bentham's injunction not to punish ‘where it must be inefficacious: where it cannot act so as to prevent the mischief’, and ‘where the mischief may be prevented … without it: that is, at a cheaper rate’.50 The proper approach is therefore to assess whether a particular kind of misconduct is more appropriately dealt with through a regulatory framework, or by civil liability, or by a civil preventive order. The key question of appropriateness will depend on other factors such as the public element in the wrongdoing and the magnitude of the harm or wrong involved. But the thrust of the principle—known sometimes as the principle of subsidiarity, and applied so as to ensure that a right is not infringed where the objective of the interference could be secured in some other (lesser) way—is that the criminal law should be reserved as a legislative technique of last resort, used only for seriously wrongful or harmful conduct. (d) The principle of not criminalizing where this would be counter productive The fourth component of the minimalist approach is the principle of not creating a criminal offence, or set of offences, where this might cause greater social harm than leaving the conduct outside the criminal law, or where the prohibition is unlikely to be effective.51 This view may be challenged on the ground that, if conduct is serious (p. 34) or harmful enough to justify criminalization, there is at least an important symbolic reason for declaring it to be criminal. It was suggested in section 2.1 that the purposes of the criminal law are threefold— declaratory, preventive, and censuring. The declaratory purpose may reassure some and deter others. But even if the preventive purpose is largely unfulfilled (as, perhaps, with the 70 mph speed limit on British motorways), it may achieve some degree of prevention or reduction of the unwanted behaviour. Thus limited efficacy is not necessarily an argument against criminalization, although it provides a good reason to search for supplementary ways of controlling the unwanted conduct. Perhaps the restrictive policy against ‘ineffective’ laws is a version of the argument that the inclusion of unenforceable offences may bring the criminal law into disrepute: if so, it must be established that the patchy enforcement of speed limits, for example, really does diminish people's respect for other parts of the criminal law.52 The other thrust of this restrictive principle is more powerful: if the criminalization of certain conduct, such as the possession of ‘soft’ drugs or various ‘vice’ offences, gives rise to social consequences that are hardly better than the mischief at which the laws aim, this militates strongly in favour of decriminalization. Thus drugs and vice laws may (i) produce active ‘black markets’, (ii) lead the police to adopt intrusive means of enforcement, (iii) allow the police to be selective in their enforcement, and (iv) lead to a degree of police corruption.53 Prohibitions that have these consequences ought to be reconsidered: there is an ongoing debate about the propriety and wisdom of penalizing drug offences at all, and certainly so severely, when it appears that the law has little effect on the scale of drug use, importation, and supply.54 There are other objections against criminalizing drug use, although some may still argue (notably in Page 10 of 22

Criminalization relation to ‘hard’ drugs) that the case for criminalization outweighs the other social consequences.55 This leads on to a more general argument for restraint in criminalization: that, since the enforcement of the criminal law is selective and tends to bear down most heavily on the least advantaged (the enforcement of drug laws is one reason for the disproportionately high number of non-white offenders in prison in Britain and the USA), these injustices should be kept to a minimum. The effectiveness principle has sometimes been turned on its head, so as to produce the argument that where criminalization would be productive and cost-efficient it should be used. This has been an integral element in English criminal legislation for many years: it is rarely spelt out, but underlies the creation and re-enactment every year of scores of offences with low penalties, attached to statutes on sundry matters such as the Education Act 2005 and the Energy Act 2011. There are two arguments (p. 35) of principle against this approach, however. One is that culpability is central to the notion of wrongdoing, and most of these offences contain little or no culpability requirement. The other is the minimalist principle, also expressed in the de minimis limitation, that the criminal law should not be used for minor wrongs. Whilst some of these ‘regulatory’ offences are clearly not minor, others are. As noted in Chapter 1.5, English law lacks a general sanctioning system which does not involve the censure of the criminal law—a system of civil violations, infractions, or administrative wrongs. This makes it even more unlikely that decisions to criminalize are preceded by a vigorous examination of whether some non-criminal sanction would be sufficient. Small wonder that this inverted form of the effectiveness principle tends to lead to over criminalization.

2.5 Morally wrong behaviour If certain behaviour is regarded as morally wrong, is this a sufficient element of wrongfulness to come within the principles set out in section 2.3? This question has been the subject of vigorous debates about the proper ambit of the criminal law in the realms of sexual morality. In the notable exchanges between Lord Devlin and Professor Hart,56 Devlin's argument was that a society is entitled to use the criminal law against behaviour which may threaten its existence; that there is a common morality which ensures the cohesion of society; that any deviation from this common morality is capable of affecting society injuriously; and that therefore it may be justifiable and necessary to penalize immoral behaviour.57 In response, Devlin's opponents have broadly followed the approach of John Stuart Mill58 in proclaiming that the main or only acceptable reason for criminalizing behaviour is that it causes harm to others, and that supposed ‘immorality’ is not a sufficient reason. Lord Devlin's argument relies on an unacceptably loose concept of morality. He assumes that immorality is to be defined and measured according to the strength of feelings of ordinary people. If certain behaviour evokes feelings of intolerance, indignation, and disgust among ordinary members of society, that is a sufficient indication that the behaviour threatens the common morality and is therefore a proper object of the criminal law. The difficulty is that these feelings of ordinary people may be more the expression of prejudice than of moral judgment. If a person's reaction to certain behaviour is to be termed ‘moral’, it ought to be grounded in reasons as well as in feelings, and those reasons ought to be consistent with other standards used by that individual to judge personal behaviour. A theory about morality and the criminal law (p. 36) must be based on a defensible definition of morality, not one which confuses it 59 Page 11 of 22

Criminalization with mere feelings of distaste and disgust.59 Is there, then, such a thing as a common morality? On core matters such as the use of force, fear, and fraud there may well be widely shared moral views, but on sexual matters the divergences may be great. Whose morals are to be a guide? Although Devlin maintained that morals and religion are inextricably joined, he did not argue that the teachings of the established church constitute the common morality. In this he was realistic: British society contains adherents of several religions, with diverse views on abortion, prostitution, euthanasia, and so forth, and there is a large proportion of the population which professes no religion (though its moral codes may bear some traces of religious teachings). Devlin proposed that the common morality could be discovered by assembling a group of ordinary citizens, in the form of a jury, and asking them to reach decisions on certain types of behaviour. However, not only would this method confound prejudices with moral judgments, but it might also fail to elicit agreement on some subjects such as hom*osexual behaviour and abortion. Devlin's opponents have tended to be in the individualistic liberal tradition, linking Mill's harm principle with Kantian ethics. According to this view, the law should respect the autonomy of each individual above all; it should treat persons as individuals and allow each to pursue his or her own conception of the good life subject only to the minimum number of constraints necessary to secure the same freedom to other individuals.60 This is prominent in the minimalist approach to criminalization, and in a ‘defensive’ criminal law policy that treats the protection of individuals from State power as one of its principal objectives. Similarly, Feinberg argues that ‘paternalistic interference is offensive morally because it invades the realm of personal autonomy where each competent, responsible adult should reign supreme’.61 Some liberals, when discussing whether or not to criminalize the nonwearing of seat-belts and crash-helmets, for example, might have recourse to the idea of ‘harm to others’: they might use Mill's principle to argue that the failure to wear seat-belts may result in harm to others, in the sense that the individuals involved may become a burden to others, creating human misery and public expenditure that are easily avoidable. It is doubtful whether this style of argument succeeds. Once the concept of harm is extended to cover indirect hardship to other individuals or to the State, Mill's principle is blunted and the possibilities for criminalization are enormous. Either one must qualify Mill by adopting the strong version of paternalism criticized by Feinberg, or one must recognize frankly that there is a competing principle at work here, such as the welfare principle described earlier. The welfare argument may be that it is strongly in the interests of the community, at a time when resources are limited, (p. 37) to avoid unnecessary expenditure on attending to the injuries of citizens who could protect themselves with little inconvenience. But does even that consideration argue for criminalization, as distinct from requiring them to pay for any medical treatment necessitated by their voluntary assumption of risk? It is widely recognized that some paternalism is appropriate so as to ensure the protection of the young and the mentally disordered. The value of autonomy applies primarily to adults, and there are dangers in persons below the designated age of majority participating in heterosexual or hom*osexual activities, in drinking alcohol, in betting and gaming, etc. (In Britain this age varies according to the activity, and questions may be raised about the justification for this.) This principle of paternalism towards the vulnerable does not imply that all these activities are ‘harmful’: rather, it implies that they may have potentially far-reaching

Page 12 of 22

Criminalization consequences for the individual concerned, and that only persons who have sufficient capacity should be allowed to take their own decisions about the potential risks. An example of the adjustment of conflicting policies and principles is provided by the report of the Wolfenden Committee on hom*osexual Offences and Prostitution in 1957.62 The Committee followed Mill's approach in asserting that ‘there must remain a realm of private morality and immorality which is, in brief and crude terms, not the law's business’, but it maintained also that this principle must interact with the need to protect the vulnerable against exploitation and corruption, and with the policy of protecting the citizen ‘from what is offensive and injurious’. The protection of the vulnerable may be a justifiable form of paternalism, but what about protecting all citizens (adults and the young) from ‘offence’? This goes beyond protection from harm on Feinberg's definition, since individuals can hardly be said to have a stake in not being shocked or offended.63 The idea of offensive behaviour draws a public–private distinction in respect of decency and shock to feelings; what adults do in private is not the law's business, so long as harm is not inflicted non-consensually, but what they do in the public domain may be the law's business if it is likely to give serious offence to the feelings of ordinary members of the public. Feinberg's ‘offence principle’ is that: It is always a good reason in support of a proposed criminal prohibition that it is probably necessary to prevent serious offence to persons other than the actor and would probably be an effective means to that end if enacted.64 This is not just a traditional utilitarian balancing exercise. Feinberg limits his principle to ‘serious offence’, but both of these words suffer from considerable indeterminacy. Andrew von Hirsch and Andrew Simester argue that the Feinberg approach is over-inclusive (many things may cause serious offence), subjective, and not clearly connected with wrongdoing. They argue for a more objective benchmark, and suggest that the essence of the wrong involved in offensiveness lies in treating others (p. 38) with a gross lack of respect or consideration.65 As with other principles, this one must be mediated by allowing other values to restrain criminalization. There should be a margin of social tolerance (being called a rude name should not be sufficient); conduct that is readily avoidable should be excluded (as where an area for nude bathing is indicated or well known); and the conduct should be immediately offensive, and not simply create a risk of subsequent offence.66 How do these criteria apply? hom*osexual acts between adult men in private have not been criminal since the Sexual Offences Act 1967, but hom*osexual acts in public places (e.g. public lavatories) remain criminal67 on a public-decency or ‘offensiveness’ rationale, although it is not clear whether the ‘ready avoidability’ principle is properly applied here. Another example of the public-decency and offensiveness rationale may be found in the Indecent Displays (Control) Act 1981, which criminalizes the display of any indecent matter which is visible from a public place: here, the questions concern the grossness of the lack of consideration, the margin of social tolerance, and of course the possibility that vulnerable people (particularly the young) will be exposed to the display.

2.6 Remote harms One kind of justification offered for criminalization is that certain conduct may create an opportunity for serious harm to be caused subsequently. The preventive function of the

Page 13 of 22

Criminalization criminal law may be interpreted as licensing the State to criminalize conduct that creates the risk of a certain harm: the conduct may not be wrongful or harmful in itself, but it is criminalized because of the consequences that may flow from it. The nature of the risk is explicit when the offence is dangerous driving or careless driving. It is implicit, and more remote, in an offence such as speeding—the prohibition on driving a motor vehicle above the applicable speed limit, a law that aims to reduce the risk of death, injury, and damage to property. Another example is a prohibition of conduct based on what the individual may do subsequently, e.g. criminalizing the possession of knives or firearms on the basis that they may be used to kill, injure, or threaten unlawfully. A further example is a prohibition of conduct based on what others may be led to do subsequently, e.g. criminalizing certain processions or public demonstrations because of what others might be tempted to do in response. Two objections to criminalizing remote harms stand out. One is that normal causal principles appear not to support liability: if conduct is criminalized on account of what it might lead another person to do, such an intervening voluntary act should relieve the original actor of criminal liability, and so it is the person who does that voluntary act who should be penalized. The other objection is that conduct that is not harmful in (p. 39) itself should not attract liability, or (as with the inchoate offences) at least not unless it is accompanied by an intention to encourage, assist, or commit a substantive offence.68 This would rule out most offences of possession, which do not require evidence of any further intent. In particular, cases in which the occurrence of harm depends on a further decision by the actor or by another (e.g. to fire the gun unlawfully) are unsuitable for criminalization.69 In the context of many modern societies, however, it would seem foolish to have no offences of unregistered possession of a firearm or of explosives. The social context could be used as a basis for arguing that abstaining from possession of certain dangerous articles like guns and explosives ought to be recognized as a duty of citizenship. It is, of course, a curtailment of liberty. So are all other criminal prohibitions. The question is whether it is justified as a curtailment of liberty to have a registration system reinforced by some criminal offences, and in answering this question one should have regard to the magnitude of the harm as well as the likelihood of its occurrence. It is considerations of that nature which tend to support offences of speeding and of drunk driving. Of course it would be possible to have only advisory speed limits, or advisory limits of consumption of alcohol or drugs, and to reserve criminal prosecution for cases where damage, injury, or death is negligently caused. But in terms of prevention that may be regarded as disastrous, on the ground that far more people would be likely to exceed the relevant limits and far more preventable victimization would probably occur. Thus the concern for people's welfare, in the context of the great harm that may result if no criminal laws are in place, tells in favour of criminalization.

2.7 Conclusions and applications The main determinants of criminalization continue to be political opportunism and power, both linked to the prevailing political culture of the country. Though an attempt has been made in this chapter to identify some general principles, it remains true that key concepts such as harm, wrongdoing, and offensiveness may tend to melt into the political ideologies of the time, as MacCormick argues: resort to the criminal law is always parasitic on or ancillary to an established legal order Page 14 of 22

Criminalization of rights and duties in the spheres of private law and public law. Such an order of rights and duties (et cetera) has to be founded on some (however muddled and patchwork) conception of a just ordering of society. The interests protected from invasion by criminal laws are interests legitimated by a given conception of a just social order. And the harm principle would be vacuous without some such conception of legitimate interests. Hence, naturally, the laws which are justified by the harm principle on a given interpretation of ‘harm’ do indeed (p. 40) coincide with widely held precepts against ‘harmful’ behaviour. But they do not merely coincide; the criminal law in so far as it is concerned with fending off harmful behaviour is necessarily geared to protection of what are legitimate interests according to a certain dominant political morality.70 Without overlooking the politically contingent nature of much criminal legislation, it is still appropriate to discuss the values and principles that ought to be relevant to criminalization decisions, since such considerations rightly play some part at various stages in the generation and refinement of reform proposals. In these concluding remarks, the approach of the chapter will be summarized, and then the possible application of the approach to certain public order offences, terrorism offences, and to the enforcement of civil preventive orders will be discussed. The approach here has been developed for a broadly liberal democratic society, and for that reason the principles of autonomy and welfare were identified as the leading considerations. No attempt was made to conceal the fact that they may conflict in many situations, and that a key issue will always be the relative weighting of those two principles. Nonetheless, the chapter went on to set out a number of other principles and values, some of them flowing from the nature of criminal law and punishment, some having social derivations. It was argued that the building blocks of criminalization decisions are that the conduct in question must be harmful, wrongful, and of public concern—three key elements which are contestable in their application to given facts, but which are crucial dimensions. Given the censuring purpose of conviction and the probability of punishment, it was argued that the approach to criminalization ought to be minimalist. This means: • respecting human rights when enacting criminal laws; • recognizing a right not to be subjected to State punishment and the rights-deprivations it often involves; • regarding the criminal law as a technique of last resort, after less invasive and stigmatic measures have been dismissed as palpably inappropriate; and • stepping back from criminalization if its effects are likely to be worse, or no better, than adopting some other approach. There may be a limited role for paternalistic offences, to protect the vulnerable. There may also be a limited (preventive) role for the criminalization of conduct that is more or less remotely connected with the occurrence of harmful consequences. Lastly, whenever conduct is held to be a sufficiently harmful public wrong to justify criminalization, the maximum penalty to which it exposes an offender must be proportionate to culpability and to the seriousness of the interests violated. Fundamental to many arguments about criminalization and decriminalization are evidential

Page 15 of 22

Criminalization issues, some of which are empirical, others predictive. Thus, for example, the principled arguments in the debate about drugs and the criminal sanction must be related (p. 41) to empirical evidence of the effects of drug-taking compared with the effects of alcohol, tobacco, and similar substances, and empirical evidence about the nature and volume of drug-related crimes; the debate must also be related to properly founded predictions of the effects of changing the law. This is only one example: the evidential foundations of arguments for criminalization and decriminalization should always be addressed. No less necessary is a properly based prediction of the practical effect of introducing new offences, particularly in terms of selective enforcement and creative adaptation. Selective enforcement may mean that the impact falls disproportionately on certain sections of society: traditional patterns of policing may suggest this, and there is some evidence that s. 5 of the Public Order Act 1986 (which criminalizes disorderly, threatening, abusive, or insulting behaviour likely to cause harassment, alarm, or distress) has been invoked disproportionately against members of racial minorities.71 Creative adaptation has also been apparent as the police have reinterpreted s. 5 in a way not anticipated by the legislators, using it to penalize those who swear at them, and occasions of similar adaptation were discovered in the research into the use of the public order offences introduced by Part V of the Criminal Justice and Public Order Act 1994.72 It may be replied that these are problems for the control of discretion among police and prosecutors, not for the legislature at the stage of criminalization. However, even if the objectionable vagueness of offences such as s. 5 is left aside, the question of enforcement cannot be dismissed too readily. Unless there is a prospect of rapid and significant change in on-the-ground policing, the theoretical possibility of greater control of police discretion cannot be a telling counter-argument to over-broad offences. The Terrorism Act 2006 contains offences that extended the ambit of the criminal law.73 How would the principles set out in this chapter apply to them? Let us first focus on the offence in s. 1(2) of publishing a statement that is likely to be understood as glorifying acts of terrorism, intending to encourage others or reckless as to whether others are encouraged to commit or prepare for such acts. Several features of this offence give grounds for concern, despite the seriousness of the harm against which it is designed to protect. First, it is an inchoate offence aimed at preventing a remote harm. The offence consists of publishing a statement in the knowledge that others may be encouraged: there is no requirement that anyone is encouraged, let alone that anyone actually carries out any of the preparatory acts mentioned (s. 1(5)(b)). Those acts would, on traditional principles, be the responsibility of the person carrying them out (although the encourager would also be liable).74 Secondly, the offence does not require proof of an intention to encourage the commission of these further acts: recklessness (knowing (p. 42) that there is a risk of encouraging) is sufficient. That ties in with the provisions making it clear that it is enough if the statement is ‘likely to be understood by some members of the public’ as an encouragement (s. 1(1)) and that it is enough if members of the public ‘could reasonably be expected to infer that what is being glorified is being glorified as conduct that should be emulated by them in existing circ*mstances’ (s. 1(3) (b)). In other words, unlike almost all other offences in the inchoate mode, proof of intent is not required, and there are objective elements in the definition. Two other concerns are that the key term ‘glorifies’ remains vague, despite the explanation in s. 20(2) that it ‘includes any form of praise or celebration’; and that the significance of this

Page 16 of 22

Criminalization vagueness is enhanced by the maximum penalty for this inchoate and remote offence, which stands at seven years’ imprisonment. The harm against which the offence is designed to guard is a major harm, which could be a terrorist attack causing death; the key question is how much the penalty should be discounted in view of the considerable distance between this offence and the actual taking of steps to cause such a harm. The analysis of this offence also demonstrates how the justifications for criminalization ought to be stronger as the maximum penalty increases. It is one thing to take the decision to criminalize, and quite another to authorize imprisonment, especially for a substantial period. Also introduced by the Terrorism Act 2006 was the offence of preparation of terrorist acts. Section 5 makes engaging in any conduct in preparation for committing acts or terrorism, or assisting another to commit such acts, with intent to commit or assist such acts, punishable with life imprisonment. The effect of this offence is to extend the ambit of attempts liability much further: whereas a criminal attempt requires conduct that is ‘more than merely preparatory’ to the commission of the substantive offence, this offence is committed if any conduct that can be viewed as preparatory is engaged in. The offence does require intention, and neither recklessness nor any other objective element is part of the definition. But, despite the enormous distance between the preparatory act and the causing of any harm, the maximum penalty is the highest in English law, life imprisonment.75 Lastly, we saw that the minimalist approach includes principle (c), which regards the criminal law as a technique of last resort, after less invasive and stigmatic measures have been dismissed as inappropriate. Discussions of this in England and Wales are blunted by the absence of any established alternative form of regulating unwanted conduct. Although a few particular agencies have alternative methods at their disposal, English law knows no general category of ‘infractions’, ‘violations’, ‘civil offences’, or ‘administrative offences’. There is also no unitary machinery for enforcing or adjudicating upon such a category of wrongs. In theory, the criminal law ought to be divided from civil sanctions and administrative regulation by reference to its censuring function, and by the minimalist principle. An important aspect (p. 43) of this is that, if a criminal offence is to be created, then the concomitants are proper safeguards for the defendant, at least those that Art. 6 of the Convention treats as the minimum. If it is decided to control certain conduct by way of either regulatory or civil mechanisms, then the penalties ought to be kept very low; otherwise, the mechanism will rightly be held to be ‘criminal’ in substance, and all the Art. 6 safeguards will have to be respected. However, by enacting more and more ‘civil preventive orders’, the government has exploited a gap in the system of protections: the anti-social behaviour order, introduced by s. 1 of the Crime and Disorder Act 1998, may be made in civil proceedings, on the application of the local authority, a social landlord, or the police, and has been held not to be a ‘criminal charge’.76 Yet breach of the order (whose conditions were set in civil proceedings) is not just a criminal offence, but an offence thought serious enough to warrant a maximum penalty of five years’ imprisonment—an ingenious scheme for imposing harsh punishments yet bypassing the appropriate protections at the crucial stage of the proceedings. There is a criminal offence involved here, that of breaching a civil preventive order (notably, an anti-social behaviour order), and that offence must be justified. The question here is whether this is a sufficiently serious harm and wrong to justify criminalization: one view is that, where the prohibited conduct is not a criminal offence, its inclusion in an ASBO should not be a good enough reason to open the way to a prison sentence, let alone a substantial one. The contrasting view is that, if the court is satisfied that D's conduct was likely to cause

Page 17 of 22

Criminalization harassment, alarm, or distress to others, that is enough to carry it across the threshold of harmfulness, particularly where it was persistent conduct. Even those who accept this, however, must be prepared to defend a maximum sentence of five years—longer than for many offences—as proportionate.

Further reading D. HUSAK, Overcriminalization (2008). A. P. SIMESTER and A. VON HIRSCH, Crimes, Harms and Wrongs (2011). R. A.DUFF, Answering for Crime (2007), chs 4 and 6. N. JAREBORG, ‘What Kind of Criminal Law do we Want?’, in A. SNARE (ed.), Beware of Punishment: On the Utility and Futility of Criminal Law (1995). A. ASHWORTH, Positive Obligations in Criminal Law (2013).

Notes: 1 D. Husak, Overcriminalization (2008). 2 R. A. Duff, Answering for Crime (2007), 142–3. 3 For an accessible discussion see A. Kenny, Freewill and Responsibility (1978). 4 For critical discussion see e.g. J. Fischer, ‘Responsibility and Control’ (1982) 79 Journal of

Philosophy 24. 5 B. Hudson, ‘Punishing the Poor: a Critique of the Dominance of Legal Reasoning in Penal

Policy and Practice’, in A. Duff et al. (eds), Penal Theory and Practice (1994), 302. 6 R. A. Duff, ‘Law, Language and Community: Some Preconditions of Criminal Liability’ (1998)

18 OJLS 189; see also the concept of ‘moral autonomy’, developed by J. Gardner, ‘On the General Part of Criminal Law’, in R. A. Duff (ed.), Philosophy and the Criminal Law (1998), 239–44. 7 See D. N. MacCormick, Legal Right and Social Democracy (1982), 23–4. 8 R. Dworkin, Taking Rights Seriously (1977), 180. 9 For the argument that this aspect of autonomy (which he terms ‘personal autonomy’) does

not presuppose what I term the factual element in autonomy see Gardner, ‘On the General Part’, 241. 10 H. L. A. Hart, Punishment and Responsibility (2nd edn., 2008), discussed in Chapter 6. 11 J. Feinberg, Harm to Self (1986), 54; cf. his extensive discussion of autonomy in chs 17, 18,

and 19 of that work. 12 Page 18 of 22

Criminalization 12 N. Lacey, Unspeakable Subjects (1998), 4–14; for an example, see the partial defence of

loss of control discussed in Chapter 7.4(b). 13 See the quotation from Hudson, n 5. 14 J. Raz, The Morality of Freedom (1986), 425. 15 Raz, The Morality of Freedom, passim. 16 N. Lacey, State Punishment (1988), 104. 17 See n 5. 18 E.g. N. Lacey, ‘Community in Legal Theory: Idea, Ideal or Ideology?’ (1996) 15 Studies in

Law, Politics and Society 105. 19 See Feinberg, Harm to Self, 37–47 for discussion. 20 A. Brudner, ‘Agency and Welfare in the Penal Law’, in S. Shute, J. Gardner, and J. Horder

(eds), Action and Value in Criminal Law (1993). 21 A. Brudner, ‘Agency and Welfare in the Penal Law’, in S. Shute, J. Gardner, and J. Horder

(eds), Action and Value in Criminal Law (1993). 22 For further development in the context of criminal procedure, see A. Ashworth and M.

Redmayne, The Criminal Process (4th edn., 2010), ch 2.4. 23 J. Raz, ‘Autonomy, Toleration and the Harm Principle’, in R. Gavison (ed.), Issues in

Contemporary Legal Philosophy (1987). 24 Feinberg, Harm to Self, and J. Feinberg, Harmless Wrongdoing (1988). 25 D. N. MacCormick, Legal Right and Social Democracy (1982), 29; B. Harcourt, ‘The

Collapse of the Harm Principle’ (1999) 90 J Cr L and Crim 109. 26 J. Feinberg, Harm to Others (1984), 215. 27 Feinberg, Harm to Others, 31–6. 28 Feinberg, Harm to Others, 26. 29 See Duff, Answering for Crime (2007), 123. 30 Feinberg, Harm to Others, ch 5. 31 Feinberg, Harm to Others, 36. 32 Simester and von Hirsch, Crimes, Harms and Wrongs, 22 and ch 2, passim. 33 Simester and von Hirsch, Crimes, Harms and Wrongs, 27. 34 For fuller exploration, see Simester and von Hirsch, Crimes, Harms and Wrongs, ch 11, and

D. Husak, Overcriminalization (2008), chs 2 and 3. 35 Page 19 of 22

Criminalization 35 R. A.DUFF , Answering for Crime (2007), 141–2. 36 Duff, Answering for Crime, 141. See also V. Tadros, ‘The Distinctiveness of Domestic

Abuse’, in R. A. Duff and S. P. Green (eds), Defining Crimes (2005). 37 G. Lamond, ‘What is a Crime?’ (2007) 27 OJLS 609. 38 See Chapter 8.3(f). 39 Crime and Disorder Act 1998, ss. 29–32. 40 See Chapter 4.6 for the argument that English law fails in this respect. 41 See Chapter 4.7. 42 See Chapters 8.5, 8.6, and 8.7, and L. Lazarus, ‘Positive Obligations and Criminal Justice’, in

L. Zedner and J. Roberts (eds), Principles and Values in Criminal Law and Criminal Justice (2012). 43 DPP v Collins [2007] 1 Cr App R 5. 44 But note s. 29J of the Public Order Act 1986 (inserted by the Racial and Religious Hatred Act

2006), re-stating freedom of expression as a value. 45 DPP v Hammond [2004] Crim LR 851; cf. DPP v Redmond-Bate [1999] Crim LR 998, where

the Divisional Court held that the interference with rights was not justified. 46 Husak, Overcriminalization, 95–101. 47 A. Ashworth, ‘Should Strict Criminal Liability be Removed from All Imprisonable Offences?’

(2010) 45 Irish Jurist 1; A. Ashworth, Positive Obligations in Criminal Law (2013), ch 4. 48 Discussed in Chapter 1.1. 49 A. Ashworth, ‘Is the Criminal Law a Lost Cause?’ (2000) 116 LQR 225. 50 J. Bentham, Introduction to the Principles of Morals and Legislation (1789), ch 13. 51 See Husak, Overcriminalization, ch 3, and J. Schonsheck, On Criminalization (1994), ch 3. 52 Cf. the assertions in S. Kadish, Blame and Punishment (1987), 23, 57. 53 See Kadish, Blame and Punishment, 22–8, for elaboration of this argument. 54 See e.g. D. Husak, Drugs and Rights (1992), Schonsheck, On Criminalisation, ch 6, and UK

Drug Policy Commission, Final Report, A Fresh Approach to Drugs (2012). 55 Cf. the argument of Peter Alldridge, in ch 7 of his Relocating Criminal Law (2000), to the

effect that dealers who supply drugs to addicts are committing the wrong of exploitation. His conclusion, however, is that there are powerful counter-arguments to continued criminalization. 56 The principal essays written by the protagonists are collected in H. L. A. Hart, Law, Liberty,

Page 20 of 22

Criminalization and Morality (1963), and P. Devlin, The Enforcement of Morals (1965). 57 See ch 1 of his The Enforcement of Morals. 58 J. S. Mill, On Liberty (1859), passim. 59 Dworkin, Taking Rights Seriously, ch 10. 60 See the discussion in section 2.1, and also W. Wilson, Central Issues in Criminal Theory

(2002), ch 1. 61 Feinberg, Harmless Wrongdoing, p. xvii, summarizing the argument developed in Feinberg,

Harm to Self, chs 18 and 19. For a stimulating discussion in relation to the place of consent in criminal law see P. Roberts, ‘Consent in the Criminal Law’ (1997) 17 OJLS 389; see also Chapter 8.3(g) . 62 Cmnd. 257 (1957). 63 See n 28. 64 J. Feinberg, Offense to Others (1985), ch 7. 65 A. von Hirsch and A. P. Simester, ‘Penalising Offensive Behaviour’, in A. von Hirsch and A. P.

Simester (eds), Incivilities: Regulating Offensive Behaviour (2006), 120. 66 von Hirsch and Simester, ‘Penalising Offensive Behaviour’, 124–30. 67 Sexual Offences Act 2003, s. 71. 68 On the new inchoate offences of encouraging and assisting crime, see Chapter 11.7. 69 For further discussion, see A. Ashworth, ‘The Unfairness of Risk-Based Possession

Offences’ (2011) 5 Criminal Law & Philosophy 237; A. Ashworth, Positive Obligations in Criminal Law (2013), ch 6. 70 MacCormick, Legal Right and Social Democracy, 30. 71 D. Brown and T. Ellis, Policing Low-Level Disorder: Police Use of Section 5 of the Public

Order Act 1986 (1994), 28–34. 72 T. Bucke and Z. James, Trespass and Protest: Policing under the Criminal Justice and

Public Order Act 1994 (1998). 73 See the analysis by V. Tadros, ‘Justice and Terrorism’ (2007) 11 New Crim LR 658. 74 See the offences of ‘encouraging and assisting crime’ introduced by Part 2 of the Serious

Crime Act 2007, and discussed in Chapter 11.7. 75 C. P. Walker, Blackstone's Guide to the Anti-Terrorism Legislation (2nd edn., 2009), 197,

notes that other criminal offences were available in all the s. 5 cases he discusses. 76 Clingham v Royal Borough of Kensington and Chelsea [2003] 1 AC 787, and the criticism in

A. Ashworth, ‘Social Control and “Anti-Social Behaviour”: the Subversion of Human Rights?’

Page 21 of 22

Criminalization (2004) 120 LQR 263.

Principles and Policies

Principles of Criminal Law (7th edn) Andrew Ashworth and Jeremy Horder Publisher: Oxford University Press Print ISBN-13: 9780199672684 DOI: 10.1093/he/9780199672684.001.0001

Print Publication Date: May 2013 Published online: Sep 2013

3. Principles and Policies Chapter: (p. 44) 3. Principles and Policies Author(s): Andrew Ashworth and Jeremy Horder DOI: 10.1093/he/9780199672684.003.0003

3.1 Rules and principles 3.2 Constitutionality and codification 3.3 Human rights and criminal law 3.4 The range of the criminal law 3.5 The rule of law and fair procedures 3.6 Principles relating to the conditions of liability 3.7 Conclusions Further reading

3.1 Rules and principles The criminal law is sometimes presented and discussed as if it were a system of rules. It will already have become apparent from Chapters 1 and 2 that this is not true. Although there are

Principles and Policies rules, and although Parliament often goes through lengthy debates before enacting rules, there is also a great deal of discretion which often enables the police, prosecutors, magistrates, judges, and juries to adopt approaches that cannot be said to have been ‘dictated’ by the law. Even if it is pointed out that over 90 per cent of cases in the magistrates’ courts and some twothirds in the Crown Court involve a plea of guilty and therefore no trial, it remains the case that some of those guilty pleas will have involved negotiation between prosecution and defence; and, more especially, it must be recalled that the police exercise considerable discretion in their daily encounters with citizens.1 There is another sense in which study of the rules is unsatisfactory as the sole or primary approach to understanding the criminal law. English criminal law both is shaped and ought to be shaped by a number of principles, policies, and other standards and doctrines. One of the purposes of this chapter is to draw together and to discuss critically some of the foremost principles that ought to exert an influence on the substance of English criminal law. Examples will be given to show when certain principles have been officially recognized and may therefore have played a part in the development of case law or legislation, but the emphasis is on the normative (p. 45) justifications for upholding each principle. The principle is then followed by a policy or other instrumental goal that may often run counter to the principle in practice. It is not maintained that the principles and policies discussed in this chapter exhaust the range of standards, doctrines, and other arguments that may be relevant to the shaping of English criminal law: indeed, particular considerations relevant to specific offences are discussed in relation to those offences in the remainder of the book. Here, the focus is on a group of defensible principles that are consistent with the principle of autonomy outlined in Chapter 2.1, as modified by a minimal commitment to the principle of welfare in order to ensure that the social arrangements necessary to enable citizens to exercise their autonomy are also supported by the criminal law where necessary.2 For ease of exposition, the discussion is divided into three parts: section 3.4 deals with the range of offences, recalling Chapter 2; section 3.6 deals with principles bearing on the conditions of liability, to some extent anticipating Chapters 4, 5, and 6; and section 3.5 states some procedural principles. Questions of priority will then be discussed in section 3.7.

3.2 Constitutionality and codification We noted in Chapter 1.3 that many offences and defences in English criminal law are still governed by the common law, giving a significant role to the courts. In constitutional theory, decisions about what conduct should be criminal should be taken by the legislature, and these decisions should then be implemented by the executive and applied by the courts. This has led to a movement not only for legislative reform of the criminal law, but also for codification. Thus the Law Commission gave its general support, when putting forward its Draft Criminal Code in 1989, to the following proposition: because a criminal code makes a symbolic statement about the constitutional relationship of Parliament and the courts, it requires a judicial deference to the legislative will greater than that which the courts have often shown to isolated and sporadic pieces of legislation. Far from it being a possible disadvantage of codification that it places limitations upon the ability of the courts to develop the law in directions which might be considered desirable, we believe that for the criminal law this is one of its greatest 3 Page 2 of 39

Principles and Policies merits.3 Thus the enterprise of codifying English criminal law has been seen partly as an exercise in constitutional propriety, subjecting the contours of the criminal law to the democratic process of Parliament rather than leaving them largely to the common (p. 46) law and the judges. One might argue that this change would be more symbolic than practical: the parliamentary process may be democratic merely in theory, since a powerful government may push through measures not directly related to its political mandate, and the judiciary is bound to retain considerable powers through its interpretive role. Indeed, the Law Commission's draft code deliberately left a number of points open for judicial development (e.g. liability for omissions, the recognition and scope of defences), although the significance of these areas is small compared with the bulk of the code. Constitutional propriety apart, the chief aims of codifying the criminal law would be to improve its accessibility (having a large number of offences ‘set out in one well-drafted enactment in place of the present fluctuating mix of statute and case law’4 ), its comprehensibility (adopting a simpler drafting style), its consistency (in the sense of greater uniformity of reasoning and of terminology), and its certainty (settling many issues in advance, rather than leaving judicial decisions to do so after the event).5 The twentieth century codification project was commenced and shaped by a small team of academic lawyers, who produced a draft code for the Law Commission in 1985,6 and the Law Commission, after consultation, published its own Draft Criminal Code in 1989.7 Influential opinion suggested that it would not be practical to try to put the whole code through Parliament as a single Bill, since it was too large for the legislative system to cope with satisfactorily,8 and so the Commission began to come forward with some shorter Bills. The first of these dealt with non-fatal offences against the person,9 and an amended version of it was put out for consultation by the government in 1998.10 Another short Bill was put forward by the Law Commission on manslaughter,11 and that too was followed by a government paper.12 Absolutely nothing happened as a result of either declaration of government intent. The Law Commission's 2008 programme announced the abandonment of the codification project, citing factors such as increased complexity and rapid changes in legislation.13 Since then there has been sporadic legislation based on specific reports of the Law Commission on matters such as partial defences to homicide and bribery, and in the coming years reports on the insanity defence, contempt of court, and non-fatal offences against the person (again) are anticipated.14 (p. 47) Despite leading judges adding their voices to the call for a criminal code,15 it seems that the emphasis now is on accomplishing reforms of particular parts of the criminal law rather than enacting a code and amending it subsequently. This means that the criminal law will continue to be scattered and difficult to find: as Toulson LJ stated, ‘to a worryingly large extent, statutory law is not practically accessible today, even to the courts whose constitutional duty it is to interpret and enforce it’.16 The piecemeal approach to criminal law reform has signally lost some of the key objectives of codification—the largely formal virtues of clarity, certainty, and consistency. Reforming legislation such as the Corporate Manslaughter and Corporate Homicide Act 2007 and Part 2 of the Serious Crime Act 2007 (not to mention the Sexual Offences Act 2003, although that was not a Law Commission product) has become so technical and complex in its style as to make the original draft code proposed by the Law Commission in 1989 read like a Beatrix Potter story. However, although the style of drafting in the 1989 code has much to commend it, its approach to the proper contents of a criminal code 17 Page 3 of 39

Principles and Policies was more controversial.17 There ought to be reconsideration of the decision to confine the code to ‘traditional’ crimes. In so far as the process of codification is intended to make the law more accessible and understandable to the public, is there not an argument for trying to deal with the most serious offences, rather than simply the traditional ones?18 To exclude from the code serious offences carrying maximum penalties of seven, ten, or fourteen years, such as causing death by dangerous driving, while including offences such as kerb-crawling and wearing a military uniform when not entitled, seems questionable on grounds of social symbolism, to say the least. The Scots tradition in criminal law has also placed reliance on common law development. Scottish courts, it has been said, have an inherent power to punish conduct that is grossly immoral or mischievous, or is obviously of a criminal nature.19 This, as we will see below,20 has led to the judicial creation of new crimes even in modern times. The justification often advanced is one of keeping the criminal law in touch with the community, but this begs questions about the judges’ ability to represent or distil community values. On this view, as Lindsay Farmer argues, ‘the community is idealized and free of conflicts and, of course, is not represented by the legislature’.21 However, the Scottish Law Commission has published a Draft Criminal Code, prepared by a group (p. 48) of academic lawyers.22 Its relative brevity is to be commended, but it is open to similar social criticisms as its English counterpart.

3.3 Human rights and criminal law If one constitutional principle is that the reach of the criminal law should be declared by the legislature, leaving the courts to apply and to interpret the legislation, another is that the criminal law should respect fundamental rights and freedoms. There are two sources of fundamental rights relevant to English criminal law—European Community law, and the European Convention on Human Rights. European Community law is potentially more powerful, since it takes priority over domestic law. However, its impact on English criminal law remains somewhat scattered, even if frequently underestimated.23 The ‘third pillar’ of the European Union, as established by the Maastricht Treaty of 1992, relates to co-operation in the fields of justice and home affairs. The Amsterdam Treaty of 1999 defined the objective of the ‘third pillar’ as the creation of an ‘area of freedom, security and justice’. The EU Constitutional Treaty anticipates its further development, and the provisions in Art. III-270 envisage harmonization of criminal laws as well as harmonized procedures, to go with mutual assistance and other cooperation (such as the European arrest warrant) already in place.24 More important in practice has been the change wrought by the Human Rights Act 1998, which may be loosely described as having incorporated into English law the European Convention on Human Rights. Reference has already been made to some Convention rights in Chapter 2.3 and 2.4. For present purposes, it is sufficient to make a broad sketch of the substantive rights guaranteed by the Convention, with some indication of their relevance to English criminal law: 25 • Article 2 (right to life): self-defence and permissible force as exceptions; abortion; the surgical separation of conjoined twins; the right to self-determination, and assisting suicide. • Article 3 (right not to be subjected to torture or inhuman or degrading treatment): protection through laws against sexual and physical violation; extent of defence of parental

Page 4 of 39

Principles and Policies chastisem*nt. • Article 4 (right not to be held in slavery or servitude): protection through laws against forced labour and human trafficking. (p. 49) • Article 5 (right to liberty and security of person): the defence of insanity; arrest for breach of the peace; and the ‘quality of law’ test. • Article 6.2 (presumption of innocence): the burden of proof, and (possibly) offences of strict liability. • Article 7 (prohibition on retroactive criminal laws): judicial lawmaking, and certainty in the definition of criminal offences. • Article 8 (right to respect for private life): sexual offences; consent to physical harm; child abduction. • Article 9 (freedom of religion): blasphemy (also Art. 10). • Article 10 (freedom of expression): obscenity; racial hatred offences; contempt of court; incitement to disaffection; official secrets legislation; breach of the peace, and s. 5 of the Public Order Act 1986. • Article 11 (freedom of assembly): breach of the peace, and various offences under the Public Order Act 1986 and the Criminal Justice and Public Order Act 1994 concerned with processions and demonstrations. This list does not go into detail, but there will be references to the Convention and its jurisprudence at appropriate points in the later chapters. Nor is the list an exhaustive one. What is significant is that the Convention rights operate as a source of ‘higher law’ that can be used as a benchmark of the constitutionality of criminal legislation. Where a court finds that the definition of an offence interferes with one of the defendant's Convention rights and (if it is a right protected by Arts. 8–11) does so either without it being ‘necessary in a democratic society’ or proportionate to such a necessity, it may recognize this as the basis for a defence to liability.26 What has been the impact of the Human Rights Act on the criminal law? The reports and consultation papers issued by the Law Commission have taken considerable care to deal with possible Convention issues. The compatibility of legislation with the Convention should be assured by the procedure whereby the Minister sponsoring a Bill certifies that it is compatible with Convention rights,27 but in fact certificates have been issued for some Bills whose compatibility has been much contested.28 Section 6 of the Human Rights Act requires all public authorities (including the courts) to act in compliance with the Convention: this means that courts are bound to overrule judicial precedents which they find to be inconsistent with the Convention. Courts also have a duty under s. 2 of the Human Rights Act to ‘take into account’ decisions of the European Court of Human Rights. The wording of s. 2 makes it clear that the Strasbourg decisions are not binding: English courts have to interpret the Convention in the light of the Strasbourg jurisprudence, and may also consider other relevant decisions which may be drawn to their attention (e.g. decisions of the Privy Council, or constitutional cases (p. 50) from Canada, New Zealand, the USA, or South Africa). If the English courts were to decide not to follow a decision of the European Court of Human Rights, the right of an individual to petition to Strasbourg is available. In some spheres, where the Convention jurisprudence is weak (such as the burden of proof and Art. 6(2)), the English courts have gone further than the Strasbourg decisions and have followed the lead of other Commonwealth countries.29 Page 5 of 39

Principles and Policies

Most powerful of all is s. 3 of the Act, which requires courts to construe legislation so as to comply with the Convention, ‘so far as it is possible to do so’. This confers on courts a rather different interpretative role from that assumed at common law. Judicial discussions about ‘the intention of Parliament’ should be less frequent in cases where a Convention right is engaged, since the primary task is to reach an interpretation which protects the rights of individuals under the Convention—which may be the rights of defendants or of (potential) victims, for example. However, in some cases the courts have used this interpretative power extravagantly, so as to hold that a legislative provision bears a meaning that seems difficult to reconcile with its wording.30 The courts do have an alternative approach in such situations: if a court is unable to read a statutory provision compatibly with the Convention, it will have to proceed as normal and the defendant will then appeal. An appellate court (Court of Appeal, Divisional Court, House of Lords) has the power under s. 4 to make a ‘declaration of incompatibility’ if it is satisfied that a statutory provision is incompatible with the Convention.31 This may lead the government to take remedial action (s. 10), but the issue of a declaration of incompatibility itself has no effect on the continuing validity of the law or on the outcome of the proceedings in the case. In coming to grips with the Convention and its jurisprudence, it is important to note the difference in patterns of reasoning that it requires. The rights declared in the Convention have different strengths and, where they have exceptions, the structure of the exceptions may differ markedly. One pointer to this is Art. 15, which permits States to derogate from certain rights under the Convention ‘in time of war or other emergency threatening the life of the nation’, ‘to the extent strictly required by the exigencies of the situation’,32 but specifically excludes from derogation the rights in Arts. 2, 3, 4(1), and 7. One might therefore construct the following hierarchy of rights: • Non-derogable rights: the right to life (Art. 2), the right not to be subjected to torture, inhuman, or degrading treatment (Art. 3), the prohibition on slavery and forced labour (Art. 4(1)), and the right not to be convicted of a crime that was not in force at the time of the conduct (Art. 7). Article 2 does provide for certain exceptions, and the same exceptions should apply in some Art. 3 cases.33 But those exceptions, discussed in Chapter 4.6, are narrowly circ*mscribed. (p. 51) • Strong rights: the right to liberty and security of person (Art. 5), the right to a fair trial (Art. 6), and the right to enjoy Convention rights without discrimination on any ground (Art. 14). A State is permitted to derogate from these rights under the strict terms of Art. 15,34 and the Strasbourg court has in some cases been content to afford States some margin of appreciation in respect of these rights. • Qualified rights: the right to a private life (Art. 8), the right to freedom of thought and religion (Art. 9), the right to freedom of expression (Art. 10), the right to freedom of assembly (Art. 11). These are qualified or prima facie rights, their common feature being that the first paragraph of the Article declares the right, and the second paragraph sets out the circ*mstances in which the right may justifiably be interfered with. This affords considerable scope for argument, using the Strasbourg jurisprudence and other sources.35 The right to peaceful enjoyment of possessions, declared by Protocol 1, is also subject to a ‘public interest’ exception which places it within this broad category.

Page 6 of 39

Principles and Policies The grounds for justifying exceptions to the qualified rights under the second paragraphs of Arts. 8 to 11 are fairly broad and wide-ranging, and turn on the two requirements of ‘necessary in a democratic society’ and ‘proportionality’. Although the Strasbourg Court does not have an entirely consistent approach to the question of proportionality, its approach is more rigorous than that of the English courts to their preferred and looser concept of ‘balancing’.36 In summary, the rights declared in the Convention are not extensive and were never intended as a complete statement of the limits of the criminal sanction. As expected, the impact of the Human Rights Act on the substantive criminal law (as distinct from criminal procedure and evidence) has been rather small: the issues surrounding the compatibility of the rules on selfdefence and insanity have not yet come up for decision, and few of the reported cases have necessitated a re-writing of English criminal law.37 However, there remains scope for critical discussion of the certainty of some aspects of English criminal law, as we shall see in paras. 3.5(g) to (j). (p. 52) 3.4 The range of the criminal law The preceding chapter illustrated the difficulties involved in deciding which interests the criminal law should protect (to which the Convention has some relevance), and in ranking harms so as to achieve some kind of proportionality. Clearly a primary aim of the criminal law ought to be to provide for the conviction of those who culpably cause major harms to other citizens or to the community, but it has already been noted that in practice the criminal law contains a myriad of less serious or more controversial offences. What principles and policies are relevant to the decisions to expand or contract the criminal law in these spheres? The enquiry begins by summarizing four principles and policies already discussed in Chapter 2, and then moves on to consider two other relevant principles. (a) The principle of minimum criminalization This principle, which was discussed in Chapter 2.4, is that the ambit of the criminal law should be kept to a minimum. It flows from the principle of autonomy and the minimalist notion of welfare already developed in Chapter 2. As we saw, the point is not so much to reduce criminal law to its absolute minimum as to ensure that resort is had to criminalization only in order to protect individual autonomy or to protect those social arrangements necessary to ensure that individuals have the capacity and facilities to exercise their autonomy. The principle is supported by various evidential and pragmatic conditions, so that even if it appears to be justifiable in theory to criminalize certain conduct, the decision should not be taken without an assessment of the probable impact of criminalization, its efficacy, its side-effects, and the possibility of tackling the problem by other forms of regulation and control. Creation of a criminal offence has the consequence that a defendant accused of that crime has the minimum rights guaranteed by Art. 6(3) of the Convention. (b) The policy of social defence Perhaps the strongest arguments against minimum criminalization are thought to derive from the policy of social defence. According to this view, the criminal law may properly be used against any form of activity which threatens good order or is thought reprehensible. There are, on this view, no limits to the use of the criminal sanction apart from financial ones. It was argued in Chapter 2 that many extensions of the criminal law are examples of political Page 7 of 39

Principles and Policies posturing, a government response to a matter of social concern about which ‘something must be done’. For this reason a sceptical stance should be adopted towards claims of ‘social defence’, which are easy to advance. If it is claimed that the new crime is needed to protect people from certain harms, it must be asked whether the wrong involved is so serious as to justify criminalization, and (p. 53) whether protection cannot be supplied more effectively by other means outside the criminal law. Creating a new crime may have a welcome symbolic effect, in condemning certain activity, but criminalization may still be neither appropriate nor effective in terms of protecting people from harm. One difficulty with the principle of minimum criminalization is, however, that it could be taken to freeze the contours of the criminal law. If it were interpreted as a barrier against further extensions of criminal law, this would be unsatisfactory, as it would take scant account of the many anomalies accumulated in English law, as in other systems, over the years. Rectification of an anomaly (for example, the old rule that a husband could not be convicted of the rape of his wife38 ) may well lead to a new sphere of criminalization; so may the extension of the criminal law to cover a newly arising mischief, such as Internet p*rnography,39 or a newly publicized mischief, such as stalking, that may cause significant harm.40 The extension of the criminal law into areas such as these may be justified on the ground that the wrongs involved in such conduct are no less significant than those involved in many serious crimes already established. These examples are important as a corrective to extreme libertarian arguments deriving from the policy of minimum criminalization. One might well agree that we all prefer our behaviour to be subject to as few constraints as possible, but that preference must be placed in the context of our membership of a community. Certain constraints may be reasonable in the interests of the community at large, even though they restrict particular individuals, as we saw when elaborating the principle of welfare in Chapter 2.2. However, the interaction between the principle of minimum criminalization and the policy of social defence may operate in undesirable ways. It may be decided to deal with significantly anti-social behaviour through the civil law, thereby avoiding the extra protections conferred by English law and by Art. 6 on persons ‘charged with a criminal offence’. This was the strategy behind the creation of the anti-social behaviour order: the House of Lords has confirmed, taking a narrow view of s. 1 of the Crime and Disorder Act 1998, that proceedings for the imposition of an anti-social behaviour order are civil (although holding that a standard of proof indistinguishable from the criminal standard should be applied),41 even though the consequence of a breach of such an order is the commission of a strict liability offence with a maximum penalty of five years’ imprisonment. Such orders operate like a Trojan horse. They pay lip-service to the principle of minimum criminalization, whilst enabling severe punishment with no more than a token gesture towards the normal rights of the defendant.42 (p. 54) (c) The principle of liability for acts not omissions This principle has often been cited, in the courts and elsewhere, as a reason for restricting the ambit of the criminal sanction.43 In fact, Parliament has applied the policy of social defence to produce a great increase in the number of offences which penalize persons for ‘failing to’ fulfil certain requirements, usually concerned with motoring, business, and finance. It now seems to be accepted that there are justifications for imposing positive duties at least on those who engage in potentially dangerous activities, such as handling radioactive substances, selling food, or driving on the roads. There is a long-standing and fundamental duty on a parent to 44 Page 8 of 39

Principles and Policies ensure the health or welfare of her or his child,44 and that has now been extended to a duty to protect members of one's household who are children or vulnerable adults.45 The courts have tended to regard omissions liability as exceptional and in need of special justification. The main reason is that positive duties to act are regarded as an incursion on individual liberty: the principle of autonomy militates against omissions liability, on the ground that public duties restrict one's liberty to pursue one's own ends by requiring one to respond to events whenever they occur (e.g. by throwing a lifebelt, or assisting injured people). The courts have, however, held that familial ties and voluntarily assumed obligations may be acceptable as bases for criminalizing omissions, but there has been no legislative or judicial enthusiasm for a general duty to assist strangers or to take steps towards enforcing the law. As we will see in Chapter 4.4, this viewpoint is grounded in a highly individualistic version of the autonomy principle. (d) The principle of social responsibility This countervailing principle adopts the welfare-based proposition that society requires a certain level of co-operation and mutual assistance between citizens. There are powerful arguments of welfare which support certain duties to act to protect others in dire situations. Many other European countries criminalize the failure to render assistance to another citizen who is in peril, so long as that assistance can be given without danger to oneself.46 Three arguments are often raised against such ‘extensions’ of the criminal law. First, it is objected that it will inevitably be unclear what is expected of the citizen: such laws often use the word ‘reasonable’, and this fails to give fair warning of what should be done and when.47 A second and consequential objection is that the exercise of prosecutorial discretion then becomes a major determinant of criminal liability. This may be criticized as weakening the rule of law, by (p. 55) transferring effective power to officials.48 Thirdly, it is argued that omissions liability calls for much greater justification than the imposition of liability for acts. It is said that this reflects a widely felt moral distinction: ‘we do much more wrong when we kill than when we fail to save, even when such a failure violates a positive duty to prevent death.’49 But even if there is such a distinction, it would only establish that omissions are viewed less seriously than acts, not that they are unsuitable for criminalization—and research into public attitudes suggests otherwise.50 So long as proper attention is paid to ‘rule of law’ protections such as the principle of fair warning, there may be good arguments for criminal liability for omissions, but the requirements and the boundaries of omissions liability need further debate and elucidation.51 The recognition of some social duties is therefore essential if all individuals are to have a proper capacity for autonomy, and further the imposition of duties backed by the criminal sanction may be justifiable to safeguard vital interests (such as life and physical integrity), if this can be done without risk or hardship to the citizen. This principle of social responsibility would therefore support, for example, an offence of failing to render assistance to a citizen in peril, where that assistance can be accomplished without danger to the rescuer. The critical element here is the danger to human life and the (qualified) duty to take certain action: it is not a prescription for making all citizens into their fellow citizens’ keepers, nor need it render D liable for all the consequences.52 These arguments on social responsibility and omissions liability are developed further in Chapter 4.4(c). (e) Conflicting rights and the principle of necessity There may be circ*mstances in which it is a person's right to use force on another, even to

Page 9 of 39

Principles and Policies take another's life, as is evident from the exceptions to the right to life declared in Art. 2 of the Convention. In principle this applies only where it is necessary for the defender to use force in order to prevent the infringement of the right to life or the right to security of person. Similarly, where it is absolutely necessary for the apprehension of a suspected offender, prevention of the escape of a person lawfully detained, or for the protection of an individual from attack, it may be justifiable for one individual to infringe the normal rights of the other (the aggressor). The ambit of the principle is examined more fully in Chapter 4.6. (p. 56) (f) The principle of proportionality This principle operates so as to place limitations on the amount of force that may properly be used in conditions of necessity. No individual, even an offender, should have his or her interests sacrificed except to the extent that it is both absolutely necessary and reasonably proportionate to the harm committed or threatened. This should apply equally to law enforcement officers and to ordinary citizens. A sharper formulation of this principle would be that the principle of necessity, in cases of conflicting rights, grants the authority to inflict only the minimum harm—a version of the view that one is permitted to use force only if it is a lesser evil than allowing events to take their course.53 As we will see in Chapter 4.6, the Strasbourg Court has read a requirement of ‘strict proportionality’ into Art. 2.54 There may also be arguments for differentiating between sudden and instinctive responses and those cases where there is ample time for reflection. Further, the assumption that the user of force is innocent and the other party is the wrongdoer does not apply in all cases, as we will see in Chapter 6.4.

3.5 The rule of law and fair procedures In this section we deal, at greater length, with those principles and policies relating to the function of the criminal law as a means of guiding the conduct of members of society and the conduct of courts and law enforcement officers. In relation to each pair of contrasting precepts, the first-mentioned principle will have the support of the European Convention of Human Rights, whereas the second is usually based on pragmatic and political considerations of the time. Three pairs deal with aspects of the principle of legality, sometimes expressed by the maxim nullum crimen sine lege. This fundamental principle is more frequently rendered in England in terms of ‘the rule of law’: According to the ideal of the rule of law, the law must be such that those subject to it can reliably be guided by it, either to avoid violating it or to build the legal consequences of having violated it into their thinking about what future actions may be open to them. People must be able to find out what the law is and to factor it into their practical deliberations. The law must avoid taking people by surprise, ambushing them, putting them into conflict with its requirements in such a way as to defeat their expectations and to frustrate their plans.55 This is a fundamental principle, with both procedural and substantive implications. It expresses an incontrovertible minimum of respect for the principle of autonomy: citizens must be informed of the law before it can be fair to convict them of an offence (many of the mens rea and culpability doctrines discussed in Chapter 5 are connected (p. 57) to this), and both legislatures and courts must apply the rule of law by not criminalizing conduct that was lawful Page 10 of 39

Principles and Policies when done. (g) The non-retroactivity principle In many other jurisdictions, especially within Europe, it is usual to begin a discussion of general principles of the criminal law by stating the maxim nullum crimen sine lege, sometimes known as the principle of legality. However, the connotations of the principle of legality are so wideranging that it is preferable to divide it into three distinct principles—the principle of nonretroactivity, the principle of maximum certainty, and the principle of strict construction of penal statutes. The essence of the non-retroactivity principle is that a person should never be convicted or punished except in accordance with a previously declared offence governing the conduct in question. The principle is to be found in the European Convention on Human Rights, Art. 7: ‘no one shall be held guilty of any offence on account of any act or omission which did not constitute a criminal offence under national or international law at the time when it was committed.’ The rationale links back to the autonomy principle and to the concept of reliance inherent in the ‘rule of law’ ideal: ‘respect for autonomy involves respect for the ability to plan, which requires respect for the ability to rely on the law’, which in turn generates the principle of non-retroactivity.56 How does it apply to the courts? It may seem obvious to state that they should not invent crimes and then punish people for conduct which falls within the new definition. But how would the common law have developed if such a power had not been exercised? The courts have developed and extended English criminal law over the years, untrammelled by the non-retroactivity principle. To ‘adapt’ the law is a great temptation for a court confronted with a defendant whose conduct it regards as plainly wicked but for which existing offences do not provide. The conflict between the non-retroactivity principle and the functioning of the criminal law as a means of social defence reached its modern apotheosis in Shaw v DPP (1962).57 The prosecution had indicted Shaw with conspiracy to corrupt public morals, in addition to two charges under the Sexual Offences Act 1956 and the Obscene Publications Act 1959. The House of Lords upheld the validity of the indictment, despite the absence of any clear precedents, on the broad ground that conduct intended and calculated to corrupt public morals is indictable at common law. The decision led to an outcry from lawyers and others. One objection to Shaw is that it fails to respect citizens as rational, autonomous individuals: a citizen cannot be sure of avoiding the criminal sanction by refraining from prohibited conduct if it is open to the courts to invent new crimes without warning. What happened in Shaw was that a majority of the House of Lords felt a strong pull towards criminalization because they were convinced of the immoral and anti-social nature of the conduct—thus regarding (p. 58) their particular conceptions of social defence58 as more powerful than the liberty of citizens to plan their lives under the rule of law. But there are two more, interconnected, objections to this decision. First, the new crime was even less defensible, since it concerned a socially controversial realm of conduct (prostitution) rather than behaviour widely accepted as an evil warranting the criminal sanction: if the courts are to legislate, they should at least confine themselves to relatively uncontroversial cases. Secondly, this realm of conduct had only recently been considered by Parliament, which had introduced limited reforms in the Street Offences Act 1959; thus it could

Page 11 of 39

Principles and Policies be argued that since Parliament did not then extend the law to penalize conduct such as Shaw's, the courts were usurping the legislative function when they did so. This constitutional dimension of the decision should not be underestimated. The proper procedure is for a democratically elected legislature to create new offences. What Shaw seems to admit is that the police and prosecution may prefer to press a hitherto unknown charge, and the courts may uphold its validity at common law. This accords great power to the executive and the judiciary, and since an offence thus created operates retrospectively on the defendant, it fails to respect the citizen's basic right that the law be knowable in advance. The criminal law embodies the height of social censure, and its extent should be determined in advance by accountable democratic processes rather than ex post facto by judicial pronouncement.59 It appears that the English courts no longer claim the power to create new criminal offences,60 apparently accepting the force of the principle of non-retroactivity. The Scottish judiciary does still claim this power, as part of a dynamic system of common law which must be adapted to deal with changing social circ*mstances. In 1983 the Scottish courts in effect created a criminal offence of selling glue-sniffing equipment,61 and in 1989 they reached their famous decision to extend the crime of rape to husbands, overturning a long-standing exception.62 Yet the English courts, which (since Knuller) ostensibly adhere to the principle of nonretroactivity, took the same decision in relation to marital rape: in R v R 63 the House of Lords abolished the husband's immunity from liability for rape of his wife. There are many convincing reasons why the old rule should have been abolished,64 but the relevant question here is whether the law should have been changed by a judicial decision which operated retrospectively on the defendant, rather than prospectively by the legislature. (p. 59) That question was taken to Strasbourg, alleging that the House of Lords violated Art. 7 (non-retroactivity) in this case, now referred to as SW and CR v United Kingdom.65 The Strasbourg Court held that the removal of the marital rape exemption by the House of Lords66 did not amount to a retrospective change in the elements of the offence. As the European Commission put it: Article 7(1) excludes that any acts not previously punishable should be held by the courts to entail criminal liability or that existing offences should be extended to cover facts which previously did not clearly constitute a criminal offence. It is, however, compatible with the requirements of Article 7(1) for the existing elements of an offence to be clarified or adapted to new circ*mstances or developments in society in so far as this can reasonably be brought under the original concept of the offence. The constituent elements of an offence may not however be essentially changed to the detriment of an accused and any progressive development by way of interpretation must be reasonably foreseeable to him with the assistance of appropriate legal advice if necessary.67 The majority of the Court went on to hold that the development of the law by the English courts ‘did not go beyond the legitimate adaptation of the ingredients of a criminal offence to reflect the social conditions of the time’, whereas a strong dissenting opinion argued that the abolition of the marital immunity from rape prosecution was not ‘mere clarification’ and could not be brought under the original concept of the offence. The Court's decision was clearly affected by the subject-matter, since it purported to justify its narrow reading of Art. 7 by reference to the incompatibility between ‘the unacceptable idea of a husband being immune against

Page 12 of 39

Principles and Policies prosecution for rape of his wife’ and the ‘respect for human dignity’ that is a fundamental objective of the Convention.68 This decision implants a degree of flexibility into what ought to be a fundamental rule-of-law protection for individuals: it is not that the law ought to exist before the conduct took place, but that it ought to have been foreseeable (if necessary, with legal advice) that the law would be changed in a particular direction. Some may argue that, as a result of this decision, the ordinary development of the common law by the courts is unlikely to be held to breach Art. 7. However, it is arguable that Art. 7 ought to be interpreted as placing some outer limits on judicial creativity. In the light of s. 6 of the Human Rights Act it is no longer lawful for the courts to reach decisions such as those in Shaw v DPP 69 and Knuller v DPP.70 Where, as in Tan,71 the prosecution is described as ‘novel’, there may be good reason for mounting an Art. 7 challenge. It may be expected that questions of retroactivity will arise more frequently in the context of statutory interpretation, since there are few common law crimes remaining. (p. 60) However, the judicial creation of new defences is a possibility, although the courts have sometimes deferred to the legislature on this matter.72 The principle of non-retroactivity did not feature prominently in Lord Lowry's reasoning in C v DPP,73 when he articulated five criteria for judicial lawmaking: (1) if the solution is doubtful, the judges should beware of imposing their own remedy; (2) caution should prevail if Parliament has rejected opportunities for clearing up a known difficulty, or has legislated leaving the difficulty untouched; (3) disputed matters of social policy are less suitable areas for judicial intervention than purely legal problems; (4) fundamental legal doctrines should not lightly be set aside; (5) judges should not make a change unless they can achieve finality and certainty. These are important principles, focusing on the constitutional aspects of judicial lawmaking that had been neglected in Shaw v DPP (which would fall foul of (2) and (3), at least). Lord Lowry's criteria were cited when the Court of Appeal declined to change and to broaden the basis of corporate criminal liability.74 However, the criteria fail to give explicit recognition to the significance for individuals of the principle of non-retroactivity, and Art. 7 should now be given greater weight in this context. As Lord Bingham put it in Jones (2007), ‘it is for those representing the people of the country in Parliament, and not the executive and not the judges, to decide what conduct should be treated as lying so far outside the bounds of what is acceptable in our society as to attract criminal penalties’.75 Even if English law were codified, it seems likely that courts would retain some power to develop defences to liability by creating new rules and extending old ones. Mental states such as insanity and intoxication are inconsistent with the kind of reliance presupposed by the idea of fair warning. These excusatory elements in the criminal law constitute rules of adjudication for the courts rather than rules of conduct to guide citizens, in contrast to the definitions of offences and of the permissive defences (e.g. self-defence, prevention of crime), which may be relied on by citizens in planning their behaviour. It therefore follows that the usual ‘reliance’ arguments against judicial creativity do not apply in the sphere of excusatory defences.76 It may be thought, too, that the constitutional arguments are less troublesome when the courts are dealing with excusatory defences: even if it is not proper for the courts to pursue their own

Page 13 of 39

Principles and Policies conception of social defence, it may be proper for them to exercise creative power in (p. 61) giving effect to considerations of individual culpability.77 An example of this was the judicial creation of a defence of ‘duress of circ*mstances’ in the late 1980s,78 although some years later the House of Lords declined to approve the creation of a defence of involuntary intoxication on the ground that the task was one for Parliament.79 If courts are granted wider powers in relation to excusatory defences, should they be permitted to create and extend them but not to abolish or restrict them at a later stage? One of the criticisms of the decision of the House of Lords in Howe (1987),80 which reversed a previous authority and held that duress could not be a defence to murder either as a principal or as a secondary party, was that it effectively breached Art. 7: what D did was not an offence when he did it, since at that stage duress was a defence and he would have been acquitted. On this view, once a court has created a defence it cannot abrogate it without falling foul of the principle of non-retroactivity.81 The argument is even stronger in relation to Elbekkay,82 where the Court of Appeal held that it was no defence for a man to argue that his impersonation of the victim's boyfriend was insufficient to negate the woman's apparent consent. This decision was all the more remarkable because s. 142 of the Criminal Justice and Public Order Act 1994 had recently redefined rape but had repeated the reference to rape by impersonating a husband (without extending the reference to a partner, etc.), and because the Court commented that no previous decision or statute required it to hold otherwise.83 It can be argued that this development of the law by the courts (as distinct from the legislature) was not reasonably foreseeable: whereas in the case of marital rape there had been a series of lesser decisions suggesting that the courts were moving in the direction of criminalizing all rapes of wives by husbands, there was nothing in the law prior to Elbekkay to suggest that a change might be imminent. It is therefore suggested that this is the type of case in which it can be argued that the contraction of a defence would be contrary to Art. 7.84 The question may occasionally arise whether a purportedly retrospective provision that is favourable to the defendant should be upheld, and both the Strasbourg Court and the Privy Council have held that a defendant should have the benefit of such a law.85 (p. 62) (h) The ‘thin ice’ principle A counterpoint to the non-retroactivity principle is provided by what may be called the ‘thin ice’ principle, following Lord Morris's observation in Knuller v DPP (1973) that ‘those who skate on thin ice can hardly expect to find a sign which will denote the precise spot where he [sic] will fall in’.86 The essence of this principle seems to be that citizens who know that their conduct is on the borderline of illegality take the risk that their behaviour will be held to be criminal. Another popular phrase for this would be ‘sailing close to the wind’. On occasions the courts have applied this principle both to the creation of a new offence and to the extension of an existing offence.87 The arguments in favour of it seem to combine moral/social and political elements. The social element may be that courts should be able to penalize conduct which is widely regarded as, and which D ought to be aware is, on the boundaries of illegality; the political element may be that when citizens indulge in anti-social conduct that lies close to an existing offence, they ought also to know that there is a risk of criminal liability being extended to cover activities on the fringe of illegality. There are obvious counter-arguments. The principle appears to assume that it may be right for courts to extend the criminal law by analogy, whereas that has frequently been held to be contrary to Art. 7 of the Convention.88 Extension (as opposed to interpretation) should constitutionally be the province of the

Page 14 of 39

Principles and Policies legislature. The ‘thin ice’ principle also neglects the role of the criminal law as a censuring institution whose convictions may result in both punishment and considerable stigma and social disadvantage, and overlooks the violation of the principle of autonomy caused when a citizen is convicted on the basis of a law that did not clearly cover the conduct at the time it took place. It is unacceptable for Art. 7 to be trumped by the ‘thin ice’ principle: Art. 7 is an absolute right under the Convention, from which (according to Art. 15) no derogation is possible. The days of new crimes created at common law ought to be long gone. However, the elasticity of the Strasbourg Court's decision in CR and SW v United Kingdom,89 with its notion of the ‘reasonable foreseeability’ of the law continuing its development in a particular direction, leaves some leeway for the ‘thin ice’ principle to exert an influence. (i) The principle of maximum certainty The next principle—maximum certainty in defining offences—embodies what are termed the ‘fair warning’ and ‘void for vagueness’ principles in US law.90 All these (p. 63) principles may be seen as constituents of the principle of legality, and there is a close relationship between the principle of maximum certainty and the non-retroactivity principle. A vague law may in practice operate retroactively, since no one is quite sure whether given conduct is within or outside the rule. Thus Art. 7 of the Convention is relevant here, since it is: not confined to prohibiting the retrospective application of the criminal law to an accused's disadvantage. It also embodies, more generally, the principle that only the law can define a crime and prescribe a penalty (nullum crimen, nulla poena sine lege) and the principle that the criminal law must not be extensively construed to an accused's detriment, for instance by analogy: it follows from this that an offence must be clearly defined in law. This condition is satisfied where the individual can know from the wording of the relevant provision and, if need be, with the assistance of the courts’ interpretation of it, what acts and omissions will make him liable.91 However, the Strasbourg Court has also recognized that some vagueness is inevitable in order ‘to avoid excessive rigidity and to keep pace with changing circ*mstances’, and that a reasonable settled body of case law may suffice to reduce the degree of vagueness to acceptable proportions.92 It is for this reason that the court refers to access to legal advice in order to determine the precise ambit of a law. The test applied under Art. 7 is the same as that applied as the ‘quality of law’ standard elsewhere in the Convention. Whenever a member state seeks to rely on a provision in the Convention in order to justify its actions—whether the arrest or detention of a citizen (Art. 5), or interference with one of the qualified rights in Arts. 8–11—it must establish that its officials acted ‘in accordance with the law’. This means a valid law, and this requires the State to show that the relevant rule satisfies the ‘quality of law’ standard. As the Court stated in the Sunday Times case: Firstly, the law must be adequately accessible: the citizen must be able to have an indication that is adequate in the circ*mstances of the legal rules applicable to a given case. Secondly, a norm cannot be regarded as a ‘law’ unless it is formulated with sufficient precision to enable the citizen to regulate his conduct: he must be able—if

Page 15 of 39

Principles and Policies need be with appropriate advice—to foresee, to a degree that is reasonable in the circ*mstances, the consequences which a given action may entail.93 The standard has been applied in a number of subsequent decisions. In Hashman and Harrup v UK (2000)94 the applicants had been bound over to keep the peace and be of good behaviour after disturbing a fox-hunt by blowing horns. The Strasbourg Court held that their Art. 10 right to freedom of expression had been breached by the binding over, since the interference with their right was not ‘prescribed by law’ inasmuch as (p. 64) the relevant law did not meet the ‘quality of law’ standard. The applicants had been bound over after a finding that they acted contra bonos mores, which was defined as behaviour that is ‘wrong rather than right in the judgment of the majority of contemporary citizens’.95 The Court held that this did not meet the standard because it failed to describe the impugned behaviour at all, whereas other provisions (such as conduct likely to provoke a breach of the peace) are acceptable because they describe behaviour ‘by reference to its effects’.96 It remains unclear how far the ‘quality of law’ standard may be used to challenge various offences under English law. Many offences in the Theft Acts include a requirement that the defendant acted dishonestly, a concept that plainly does not ‘describe behaviour by reference to its effects’. The Court in Hashman and Harrup stated that the offences turning on dishonesty were different because dishonesty ‘is but one element of a more comprehensive definition of the proscribed behaviour’.97 Even if that is true following recent House of Lords decisions,98 it would hardly apply to a new general offence of dishonesty or of deception, as the Law Commission concluded.99 There is considerable uncertainty of definition in common law offences such as cheating100 and perverting the course of justice (although the conduct is defined by reference to its effects),101 and they should be scrutinized urgently in the light of Art. 7's requirements. When the House of Lords examined the offence of public nuisance in Rimmington and Goldstein (2006),102 it narrowed the definition of the offence in order to avoid uncertainty, and Lord Bingham approved the statement (in a case of perverting the course of justice) that, if the ambit of a common law offence is to be enlarged, it ‘must be done step by step on a case by case basis and not with one large leap’.103 This is consistent with the Strasbourg position outlined in paragraph (g). Why should such emphasis be placed on certainty, predictability, and ‘fair warning’? As with the principle of non-retroactivity, a person's ability to know of the existence and extent of a rule is fundamental: respect for the citizen as a rational, autonomous individual and as a person with social and political duties requires fair warning of the criminal law's provisions and no undue difficulty in ascertaining them. The criminal law will also achieve this respect more fully if its provisions keep close to moral distinctions that are both theoretically defensible and widely felt: 104 this suggests a (p. 65) connection between fair warning and fair labelling (on which see 3.6(s)). A connected reason in favour of the principle of maximum certainty is that, if rules are vaguely drafted, they bestow considerable power on the agents of law enforcement: 105 the police or other agencies might use a widely framed offence to criminalize behaviour not envisaged by the legislature, creating the very kind of arbitrariness that rule-oflaw values should guard against. Similarly, when the law gives the court power to make an anti-social behaviour order in response to conduct ‘likely to cause harassment, alarm or distress’, this gives little warning to citizens about the type of conduct that may be prohibited with the threat of criminal conviction for repetition.106 It will be noticed, however, that the principle is stated in a circ*mscribed form—the principle of maximum certainty, not absolute

Page 16 of 39

Principles and Policies certainty—which indicates the compromise already inherent in the principle. In its pure form, the ‘rule of law’ would insist on complete certainty and predictability, but this is unattainable —‘vagueness is ineliminable from a legal system, if a legal system must do such things as to regulate the use of violence …’107 Unless the criminal law occasionally resorts to such openended terms as ‘reasonable’ and ‘dishonest’, it would have to rely on immensely detailed and lengthy definitions which might be so complicated as to restrict the intelligibility of the law. As Timothy Endicott argues, neither vagueness nor discretion is necessarily a deficit in the rule of law, so long as the law can perform its function of guiding behaviour.108 Thus those who adhere to the principle of maximum certainty would insist that the use of such vague terms should be reinforced by other definitional elements, guidelines, or illustrative examples which inform the citizen and structure the court's discretion.109 Any claim that a derogation from maximum certainty is necessary for the practical administration of the law must be scrutinized carefully. As the US Supreme Court put it in Conally v General Construction Co (1926): ‘a statute which either forbids or requires the doing of an act in terms so vague that men of common intelligence must necessarily guess at its meaning and differ as to its application, violates the first essential of due process of law’.110 Whether the Strasbourg Court's interpretation of the non-retrospectivity principle in Art. 7 or the more general ‘quality of law’ requirement has been overly conservative is open to debate, but the decision in Hashman and Harrup demonstrates that the principle can bite, and it will do so in this country if the courts take it as seriously as the Law Commission appears to have done. (p. 66) (j) The policy of social defence The policy of social defence runs counter to the principle of maximum certainty. It maintains that some vagueness in criminal laws is socially beneficial because it enables the police and the courts to deal flexibly with new variations in misconduct without having to await the lumbering response of the legislature. The policy of social defence thus supports the same aims as the ‘thin ice’ principle. It also suffers from similar defects, such as differing opinions of the social interests to be defended by means of the criminal law. The interests of the powerful are thus likely to prevail. The policy of social defence would support the enactment of laws vague enough to leave room for the law enforcement agents to apply them to new forms of anti-social action—for example, the public order offences in the Acts of 1986 and 1994,111 the power to make an anti-social behaviour order, and the common law offence of conspiracy to defraud.112 To the objection that such crimes delegate far too much de facto power over citizens’ lives to law enforcement agents, proponents of social defence would reply that this should be tackled by means of internal guidelines and police disciplinary procedures rather than by depriving the police and courts of the means of invoking the criminal sanction against conduct which arouses social concern. The offences themselves appear to be worded objectively and neutrally—although they suffer from what the Americans call over-breadth—but their use may be selective.113 The policy of social defence therefore favours considerable low-level discretion, conferred (in effect) by broadly phrased offences, and often supported by the use by politicians and journalists of imagery that depicts certain groups as the enemies of society against whom new powers are ‘necessary’ for ‘public safety’.114

Page 17 of 39

Principles and Policies Social defence arguments are sometimes used to support the argument that ignorance of the criminal law should be no excuse. Thus English law authorizes the conviction of persons who were unaware of the existence of a crime, even in circ*mstances where it would have been difficult for them to find out that they were committing it.115 This derogation from the notions of maximum certainty and fair warning is usually supported in terms of social defence by suggesting that, if the defence were allowed, everyone would claim it and there would be large-scale acquittals. Such arguments are unpersuasive in theory and in practice.116 The policy of social defence may be used to point out a distinct social dysfunction of the principle of maximum certainty. If members of society can rely upon criminal laws being drafted precisely and upon enforcement agents and the courts keeping within those boundaries, it is open to resourceful citizens to devise ways of circumventing those laws— conforming to the letter of the law, whilst dishonouring its spirit. Where (p. 67) this kind of activity is pursued in a systematic way, with powerful financial support, it may be regarded as a distinct threat to the values that the criminal law seeks to uphold. It is said that there are those in the financial and business worlds who make their living on these fringes of legality, exploiting the principle of maximum certainty as a shield to protect them from conviction.117 Can these people be distinguished from Shaw, Knuller, Tan,118 and others? It is doubtful whether a distinction between sexual and financial morality would be sufficient to justify a difference in approach. The proper response is that the principle of legality, and in particular of maximum certainty, would accept that there is a distinction between avoidance and evasion, and that mere avoidance must be tackled by legislative amendments to the law rather than by ex post facto stretching by the courts. The fact that the wide common law offence of conspiracy to defraud remains in full vigour, with the result that financial misdealers are not safe from its elastic clutches, should not be viewed with approval: the offence ought to be abolished and replaced with discrete offences that comply in both letter and spirit with the principle of maximum certainty.119 (k) The principle of strict construction Two of the principles which are often brought under the umbrella of the principle of legality have already been discussed (non-retroactivity, maximum certainty); the principle of strict construction is the third. The difference here is that whereas the non-retroactivity principle applies to the lawmaking activities of Parliament and the courts, this principle relates to the courts’ task in interpreting legislation. The formulation of the principle is a matter for debate. In its bald form, it appears to state that any doubt in the meaning of a statutory provision should, by strict construction, be resolved in favour of the defendant. One justification for this may be fair warning: where a person acts on the apparent meaning of a statute but the court gives it a wider meaning, it is unfair to convict that person because that would amount to retroactive lawmaking. Historically speaking, the principle seems to have originated either as a means of softening the effect of statutes requiring capital punishment, through the notion of construction in favorem vitae,120 or as a response to statutory incursions into the common law—which in turn led Parliament to enact more detailed, subdivided offences of the kind that still survive in the Offences Against the Person Act 1861.121 (p. 68) The status of the principle of strict construction is unclear. It has some connection with Art. 7 of the Convention in that, as we saw earlier, the Court has held that the nonretroactivity principle requires that the criminal law must not be extensively construed to an 122 Page 18 of 39

Principles and Policies accused's detriment, for instance by analogy.122 However, it is not clear how and when the Court would apply this principle. References to a principle of strict construction have been fitful both in England and the USA, leading to the claim that it is invoked more to justify decisions reached on other grounds than as a significant principle in its own right.123 There is certainly no difficulty in assembling a list of cases in which it appears to have been ignored.124 But it may be that it was not properly understood in its more sophisticated form in England, since it is relatively recently that a sequence of principles to be applied when interpreting criminal statutes has been established. According to the House of Lords, the proper approach is not to be bound by any particular dictionary definition of a crucial word in a statute, but rather to construe a legislative provision in accordance with the perceived purpose of that statute.125 In order to assist in ascertaining that purpose, a court may consult a Hansard report of proceedings in Parliament, a government White Paper, or the report of a law-reform committee so as to ascertain the gap in the law which the legislation was intended to remedy.126 If a Convention point arises, however, it is not a question of seeking the intention of Parliament but rather of applying s. 3 of the Human Rights Act 1998 and interpreting the statute, so far as possible, so as to comply with the Convention. Those who disagree with the principle have sought to ridicule it by arguing that no system of criminal law can function adequately if absolutely every ambiguity has to be resolved in favour of the defendant.127 But this line of attack misunderstands the true role of the principle, which has now been reasserted in the courts. Its proper place is in a sequence of points to be considered by a judge when construing a statutory offence. It will be an important advance in the development of English criminal law if other courts routinely follow the approach now established by the House of Lords, although the evidence suggests that neither courts nor counsel consider statutory interpretation to be a discrete subject with its own approach and its own precedents.128 However, there are further important questions of interpretation to which no authoritative approach has been established. For example, uncertainty still prevails over the proper approach to interpreting statutory offences which do not include a fault requirement (p. 69) in their definition: the courts are still without a coherent approach to the question of strict liability, and no sooner is a high-sounding (‘constitutional’) principle declared than other courts ignore or circumvent it.129 What is the argument in favour of the more sophisticated version of the principle of strict construction? The ‘fair warning’ argument undoubtedly plays a part, in so far as it respects the idea of citizens as rational, choosing individuals, but the primary argument is constitutional. In terms of interpreting statutes the courts are the authoritative agency. Just as the principles of non-retroactivity and maximum certainty ought to be recognized by the legislature, so they should be recognized by the courts when engaging in interpretation. Indeed, the argument is even stronger for the courts, for they are the ultimate agency for determining the practical limits of the law, and yet they are an unelected group. Parliament should retain the main responsibility for the extent of the criminal law, and, indeed, it has the right to determine the courts’ approach towards the task of interpretation (for example, by including some canons of interpretation in the Criminal Code130 ). The practical implication of this approach is that the courts should exercise restraint in their interpretive role, favouring the defendant where they are left in doubt about the legislative purpose. (l) A broader purposive approach

Page 19 of 39

Principles and Policies Militating against the principle of strict construction is a broader purposive approach which relies on the aims of the criminal law as a whole rather than on a particular legislative purpose. Why should the courts allow those who indulge in obviously wrong behaviour to escape conviction by reference to a principle which assumes that citizens take care to ascertain the law beforehand (which they usually do not), and which also assumes that the government and Parliament can be left to deal promptly with behaviour which is seen as a social problem (which they usually cannot, because of pressures on parliamentary time)? Indeed, the argument goes further. Citizens who do act in reliance on a particular view of the law could be excused by means of a defence of ignorance or mistake of law.131 As for the constitutional argument, the assumption seems to be that the principle of legislative supremacy is allpowerful. Important it may be, but there are other political and fundamental values that also have a claim to be taken into account. If one purpose of the criminal law is the deterrence of significant culpable wrongdoing and the punishment of those who engage in it, does this not supply a reason for courts to interpret criminal laws so as to achieve this end? An argument of this kind leaves a great deal to be debated—in Chapter 2 we saw how controversial the boundaries of criminalization can be132 —but its kernel is that, as with the ‘thin ice’ principle, (p. 70) it may not be unfair to penalize someone who has positioned himself on the margins of lawfulness. This may be seen as a rationalization of the appellate courts’ tendency to stretch the interpretation of statutes so as to criminalize people who, they think, have manifestly committed a serious wrong.133 As John Bell has argued, ‘if the law exists to promote collective goals, as well as to protect individual rights, it cannot be altogether unexpected that both of these aspects should come into the resolution of hard cases’.134 Two counter-arguments are often heard, in addition to the principle of legislative supremacy. One is the practical point that courts have only rarely put the legislature to the test by refusing to extend existing offences to new forms of anti-social behaviour and leaving the task to Parliament. There are some isolated examples,135 but in general the courts have not established a tradition of strict construction. If they had either brought in acquittals or quashed convictions in every case where the application of a statutory provision left some room for doubt, then the government would have been highly likely to set up a regular system for redrafting and amending criminal laws. A typical course of events was that in the case of Charles (1976): 136 the Court of Appeal favoured the acquittal of a man who, in spite of his bank's prohibition, had deliberately and substantially overdrawn on his bank account, because the Court found it difficult to bring the conduct within the definition of the offence charged. Bridge LJ recognized that social defence might be better served by a conviction, but he did not regard it as the Court's function to stretch the words of the statute. The House of Lords had no such compunction: it did stretch the statutory wording, and restored the conviction.137 Had the House of Lords adopted the same approach as the Court of Appeal, then the government and Parliament would have been left to decide on the need for an amendment to the law. In the meantime, Charles and a few others would have gone free. It is this consequence which some judges, regarding themselves as custodians of the public interest, have sought to avoid by adopting broad interpretations of statutes. Doubts have been expressed about whether appellate courts are in a proper position to assess the consequences of thus extending the law,138 and there are more recent cases in which the courts decided that it was both too difficult and inappropriate to attempt to repair defective legislation.139 Thus in Preddy 140 the House of Lords declined to adopt an interpretation of the Theft Act 1968 which would have upheld the appellant's conviction, and Parliament did move quickly to rectify the anomaly by

Page 20 of 39

Principles and Policies passing the Theft (p. 71) (Amendment) Act 1996 within five months; but, even then, Lord Goff could not resist a passing gibe at ‘the so-called principle of legality, which has a respectable theoretical foundation but can perhaps be a little unrealistic in practice’.141 A second counter-argument is that the judicial function is to uphold individual rights, leaving broader issues of social policy to Parliament. Thus Ronald Dworkin has argued that judges ought to ground their decisions in reasons which uphold individual rights and ought not to take account of policies, goals, or overall social welfare.142 Glanville Williams has advanced a similar argument specifically in relation to criminal law.143 It may be argued, however, that this adopts a particularly one-sided view of the criminal law. Principles of individual fairness are important, and some of them are absolutely fundamental, but this should not be allowed to obscure the wider sense of autonomy advocated in Chapter 2—one which emphasizes the need to provide social conditions and facilities in which a broader range of choices is available, and which may provide good reasons for pursuing a given policy, provided always that Convention rights are duly respected. (m) The presumption of innocence The principle that a person should be presumed innocent unless and until proved guilty is a fundamental principle of procedural fairness in the criminal law. Its justifications may be found in the social and legal consequences of being convicted of a crime, in which context the principle constitutes a measure of protection against error in the process,144 and a counterweight to the immense power and resources of the State compared to the position of the defendant. Article 6(2) of the Convention declares that ‘everyone charged with a criminal offence shall be presumed innocent until proved guilty according to law’, but the Strasbourg Court has not developed the presumption of innocence with any vigour. Indeed, in the leading decision of Salabiaku v France 145 the Court found no violation of Art. 6(2) in an offence (carrying a prison sentence) that placed the onus of proof on the defendant, stating merely that the reverse onus must be ‘within reasonable limits which take into account the importance of what is at stake and maintain the rights of the defence’. The English courts, on the other hand, are feeling their way towards a more robust promotion of the presumption. The famous declaration of Lord Sankey LC in Woolmington v DPP (1935)146 that ‘throughout the web of the English criminal law one golden thread is always to be seen—that it is the duty of the prosecution to prove the prisoner's guilt’, had increasingly been regarded as empty rhetoric as the numbers of statutory exceptions multiplied. However, in Lambert (2001)147 the House of Lords used the power of interpretation in s. 3 of the (p. 72) Human Rights Act to reinterpret a reverse onus provision in s. 28 of the Misuse of Drugs Act 1971 so as to impose merely an evidential burden (not the burden of proof) on the defendant. A majority of their Lordships held that the severity of the potential penalty rendered this reverse onus a disproportionate burden on D. The House of Lords adopted the same approach to an anti-terrorism offence in AttorneyGeneral's Reference No. 4 of 2002,148 although in the conjoined appeal in Sheldrake v DPP their Lordships held that it would be easier to rebut the presumption for an offence with a low maximum penalty.149 (n) The policy of ease of proof The presumption of innocence has been much neglected by the legislature: many offences are defined in such a way that the prosecution has to prove little, and then the defence bears

Page 21 of 39

Principles and Policies the burden of exculpation. Section 101 of the Magistrates’ Courts Act 1980 places on the defendant the burden of proving any excuse, exemption, proviso, or qualification in the definition of an offence tried summarily, a regime that is usually justified on grounds of expediency and economy. Neglect of the presumption is not confined to so-called regulatory offences: some 40 per cent of offences triable in the Crown Court—i.e. the most serious offences in English law—appear to violate the Woolmington principle by placing a burden of proof on the defendant.150 It seems that the presumption has been so insignificant to policymakers and legislators that often they have not even regarded it as necessary to give a reason for placing a burden on the defence. However, there is evidence of change since Lambert: in the Sexual Offences Act 2003, there were several reverse onus provisions in the Bill, which were removed after the House of Commons Home Affairs Committee pointed out the conflict with the presumption of innocence.151 Sometimes an attempt is made to justify a reverse burden by claiming that it is right to expect the defendant to prove elements relating to a defence. One difficulty here is that there is no satisfactory analytical distinction between offence and defence.152 Legislative draftsmen do not follow a single drafting rule, and it may often be a matter of chance whether a given element is expressed as a defence or is rolled up into the definition of the crime.153 A more reliable argument is that certain (p. 73) matters are much easier for one party to prove than the other: it is generally far easier for a defendant to prove that he or she had a licence or permit than for the prosecution to prove the absence of one.154 However, this should not be allowed to shade into the far less persuasive argument that D should prove any matter that ‘lies within his own peculiar knowledge’,155 a proposition that might equally apply to intention, knowledge, and many other core elements of crimes, and would thus undermine the presumption of innocence completely. The policy of ease of proof does not merely manifest itself through the imposition of burdens on the defence. Parliament has within its control the definition of offences too, and frequently inserts strict liability elements into statutory offences. This is inconsistent with the principle of mens rea and with the rule of law (see 3.6(o)), but it does not contravene the presumption of innocence enshrined in Art. 6(2) of the Convention. That presumption is procedural, not substantive, and so applies only to the burden of proof.156

3.6 Principles relating to the conditions of liability Setting the conditions for criminal liability raises further questions about ‘rule of law’ standards157 and the principle of individual autonomy (outlined in Chapter 2.1). We have already seen how these standards underlie the principle of legality in its three manifestations: the principles of non-retroactivity, maximum certainty, and strict construction. Unless a person can know what the criminal law prohibits, it is unfair to impose a conviction. Both the rule of law and the principle of autonomy emphasize respect for individuals as deliberative, choosing persons. This is often taken to suggest, as we shall see in some of the detailed principles below, that an individual should be held criminally liable only for consequences that were knowingly brought about or knowingly risked. Whatever the merits of civil liability for other consequences, an individual should not be liable to censure and punishment for them. In contrast, the principle of welfare insists that the need for social co-operation and community life may create strong arguments for extending the ambit of the criminal law and the conditions

Page 22 of 39

Principles and Policies of liability—by, for example, imposing duties to take care in certain types of situation and making the negligent liable to conviction. However, it will be argued below that this does not undermine the principle of autonomy if the appropriate conditions are fulfilled, notably that there is fair warning of the imposition of a duty of care reinforced (p. 74) by the criminal sanction, and that there is an exception for those incapable of attaining the required standard. All these points are taken further in Chapter 5: the purpose here is to express schematically the kinds of argument used. (o) The principle of mens rea In order to satisfy rule of law standards, an offence must have a (subjective) mens rea requirement in order to alert D to the fact that he is about to violate the law: some element of mens rea is needed in order to give fair warning, which would be absent if offences could be committed accidentally. The principle of autonomy may be interpreted as taking the point further, arguing that the incidence and degree of criminal liability should reflect the choices made by the individual. The principle of mens rea expresses this by stating that defendants should be held criminally liable only for events or consequences which they intended or knowingly risked. Only if they were aware (or, as it is often expressed, ‘subjectively’ aware) of the possible consequences of their conduct should they be liable.158 The principle of mens rea may also be stated so as to include the belief principle, since in some crimes it is not (or not only) the causing of consequences that is criminal but behaving in a certain way with knowledge of certain facts. Thus where the defence is one of mistaken belief, the principle of mens rea would state that a person's criminal liability should be judged on the facts as D believed them to be. All these aspects of the principle of mens rea are discussed further in Chapter 5.4 and 5.5. Although much of the principle's strength derives from the rule of law and the value of autonomy, this does not mean that negligence liability cannot be supported on the same basis: so long as there is an exception for incapacity, as argued, this may be fair. (p) The policy of objective liability In spheres of activity that are perceived to be particularly dangerous, it is often thought that there are sufficient justifications for going beyond subjective liability and imposing liability for failure to fulfil a duty of care. Perhaps the clearest example of this may be found in road traffic legislation: long-standing offences such as dangerous driving and careless driving make drivers criminally liable for the degree to which they fall below the standards expected of a competent motorist.159 Among the justifications for this is the principle of welfare, which in this respect favours the imposition of standards of behaviour on citizens because their behaviour as motorists can so easily impinge on others, with disastrous consequences. In industrial contexts there is a whole host of offences based on negligence, particularly where hazardous substances or dangerous conditions are involved. Moreover, in many commercial settings the criminal law imposes strict liability on those who sell defective products or unwholesome foodstuffs, (p. 75) convicting them in many situations where the fault is small or non-existent. The case for extending the criminal law to relatively minor harms is based on expediency, and has already been criticized in Chapter 2.4(c) . Strict liability itself is often supported by reference to considerations of welfare, ‘policy considerations’, or ‘social concern’, but it will be argued in Chapter 5.5(a) that the justifications for going beyond negligence liability to strict liability are unpersuasive. Criminal liability for negligence, however, so long as it is founded on clear and well-publicized standards and duties for people performing certain activities, may be

Page 23 of 39

Principles and Policies compatible with the rule of law and with the principle of autonomy. Thus, as will be argued in Chapter 5.5(f), there may be certain spheres in which criminal liability can properly be based on a form of negligence—taking proper account of the seriousness of the harm, the need to warn citizens of their duties, and the need to exempt those who lack the capacity to conform their conduct to the required standard. (q) The principle of correspondence Another implication of the principle of individual autonomy (with its emphasis on choice and control) and the ‘rule of law’ ideal (with its emphasis on the ability of individuals to plan) is the principle of correspondence. Not only should it be established that the defendant had the required fault, in terms of mens rea or belief; it should also be established that the defendant's intention, knowledge, or recklessness related to the proscribed harm. Thus, if the conduct element of a crime is ‘causing serious injury’, the principle of correspondence demands that the fault element should be intention or recklessness as to causing serious injury, and not intention or recklessness as to some lesser harm such as a mere assault. Another example, as we shall see,160 is the law of murder: in English law a person may be convicted of murder if he either intended to kill or intended to cause grievous bodily harm. However, the latter species of fault breaches the principle of correspondence: 161 the fault element does not correspond with the conduct element (which is, causing death), and so a person is liable to conviction for a higher crime than contemplated. In effect, murder and other crimes that breach the correspondence principle are constructive crimes. They reduce ‘rule of law’ protections and respect for autonomy by rendering D liable to conviction for a more serious offence than intended or knowingly risked; and, to that extent, the offence of conviction turns on the chance element of whether or not the more serious (unintended and unforeseen) harm results.162 (p. 76) (r) Constructive liability The argument for the extended fault element in murder, as described in the previous paragraph, favours constructive liability.163 This has a Latin tag, versari in re illicita,164 and in its widest form it argues that anyone who decides to transgress the criminal law should be held liable for all the consequences that ensue, even if they are more serious than expected. This may be termed the ‘unlawful act theory’, in so far as it holds that the commission of any crime against another supplies sufficient culpability to justify conviction in respect of whatever harm results. The decision to commit a crime is the crucial moral threshold: once D has knowingly crossed this, he should be liable for the resulting harm. This broad doctrine has now given way to what might be termed ‘moderate constructivism’, which accepts ‘the requirement of subjective mens rea introduced by the obligation to respect the rule of law’165 but argues that by intentionally attacking another D changes normative position, ‘so that certain adverse consequences and circ*mstances that would not have counted against one but for one's original assault now count against one automatically, and add to one's crime’.166 In most forms this is a more moderate doctrine than the broad ‘unlawful act theory’—it confines liability to resulting harms in the same ‘family of offences’ (usually, violence); and it does not necessarily justify the current English law of constructive manslaughter,167 since some of its supporters restrict liability to cases where there is some ‘proportionality’ or alternatively ‘no great moral distance’ between the intended attack and the resulting harm.168 However, these restrictive principles only come into play if the fundamental intuition of ‘moderate constructivism’ is

Page 24 of 39

Principles and Policies conceded. Why is a minor assault vested with such high moral importance that it is thought to justify liability for injuries much more serious than foreseen? Why should such a large slice of luck enter into the assessment of criminal liability, with the result that D is labelled as having committed a significantly worse wrong than he intended or knowingly risked? John Gardner originally argued that if the criminal law puts D on notice that this will be the consequence, the requirements of the rule of law are fulfilled.169 However, giving fair warning of an unfair rule does not turn it into a fair rule, so we still await a justification for attributing such high moral importance to the change of normative position inherent in a common assault.170 The existing law of offences against the person, stemming from an 1861 statute, is replete with examples of constructive liability, (p. 77) and the offences of murder and manslaughter are perhaps the best known instances in English law. The ‘change of normative position’ argument seems to depend on the strength of a particular intuition—that morally the most significant element in given conduct is a decision to use force on another, and that there is insufficient moral weight in the plea, ‘I only intended to punch/kick/wound slightly, not to cause injuries of that magnitude’.171 This is surely to adopt an unduly narrow view of moral responsibility. It attributes too little importance to the full context of the actor's decision, and allows a person's criminal liability to turn partly on luck.172 This argument is pursued further in the next two sections. (s) The principle of fair labelling This principle is chiefly applicable to the legislature. Its concern is to see that widely felt distinctions between kinds of offences and degrees of wrongdoing are respected and signalled by the law, and that offences are subdivided and labelled so as to represent fairly the nature and magnitude of the law-breaking.173 As James Chalmers and Fiona Leverick argue in their detailed study, labels are important chiefly to describe D's offending behaviour for the general public and to differentiate that behaviour for the purposes of those working within the criminal justice system.174 One good reason for respecting these distinctions is proportionality: one of the basic aims of the criminal law is to ensure a proportionate response to law-breaking, thereby assisting the law's educative or declaratory function in sustaining and reinforcing social standards. Fairness demands that offenders be labelled and punished in proportion to their wrongdoing; the label is important both for public communication and, within the criminal justice system, for deciding on appropriate maximum penalties, for evaluating previous convictions, for classification in prison, and so on. ‘In fairness both to offenders and to others with a relevant interest, there is a need for offence labels to convey sufficient information to criminal justice professionals to enable them to make fair and sensible decisions.’175 Similar information may also be helpful to employers and potential employers, for example.176 A deeper justification for the principle of fair labelling has a more direct connection with common patterns of thought in society. It is that where people reasonably regard two types of conduct as different, the law should try to reflect that difference. (p. 78) In principle, the criminal law should ‘track the reasonable moral convictions of the community’.177 This argument was raised against the possibility of combining the crimes of theft and obtaining by deception into a single offence: people regard stealing and swindling as distinct forms of wrongdoing, and the law should not obscure this.178 Although this proposal was not pursued, English criminal law does contain some extremely wide offences. Theft is a single offence with a maximum sentence of seven years’ imprisonment, whereas in many other jurisdictions it is subdivided into greater and lesser forms (e.g. a form of petty theft, for offences below a certain

Page 25 of 39

Principles and Policies monetary value, with a lesser maximum penalty). Criminal damage is a single offence with a maximum sentence of ten years’ imprisonment, with no subdivisions to reflect the type of property damaged or the magnitude of the damage inflicted. Perhaps the clearest example is robbery, an offence with a maximum sentence of life imprisonment which conjures up an armed raid by masked men seeking substantial money or property, and yet which in English law is fulfilled by a slight push in order to snatch a purse or handbag.179 It is regarded as axiomatic that offences of violence should be subdivided so as to distinguish different levels of injury etc.,180 and yet robbery is not subdivided to reflect the very different degrees of force used or threatened in different cases. What aspects of the offence should be reflected in the label? Four points may be raised in answer to this question. First, it was noted in sections (q) and (r) that some offences apply a much more serious label than D intended or knowingly risked—typically, the offence of manslaughter by unlawful act, for which the only fault element is in respect of the relatively minor offence of assault. The label describes the result but not the fault. This argument is examined elsewhere,181 and controversy has recently been re-kindled by the new offences of causing death by careless driving and causing death by driving while disqualified, unlicensed, or uninsured, for which the fault element is a long way below the tragic result.182 Secondly, dividing an offence into degrees may be sufficiently informative for criminal justice professionals but may be opposed by others, including defendants and (perhaps) the public. Thus the Law Commission's recommendation that the English law of homicide be divided into murder in the first degree, murder in the second degree, and manslaughter may appear to be a triumph for fair labelling, until it is noted that provoked killings are classified not as manslaughter but as murder in the second degree—which some might regard as a misuse of the ‘ultimate’ label. Defendants and victims may disagree about this, but the issue is one of labelling.183 If juries were to prove unwilling to return a verdict with the word ‘murder’ in it, this would create problems for the proposed scheme. (p. 79) Thirdly, there are good reasons for applying the principle of fair labelling to defences. Chalmers and Leverick make the point that the same reasons favour fair labelling of justificatory defences as offences, since they may guide conduct and provide a moral assessment of conduct; for excusatory defences, it may still be important to signify why a particular verdict has been reached.184 Similar points were raised by those who opposed the Law Commission's proposal for a relatively narrow definition of the partial defence of provocation, on the ground that it forced some defendants (often, women) to argue their case on a defence that reflects less well on their capacity and motivation (i.e. diminished responsibility).185 Fourthly, and separately from the above considerations, there may be a good social reason for creating a separate offence with a separate label in order to draw public attention to the wrongness of a particular course of action. Examples of this would be the creation of racially or religiously aggravated versions of certain offences, where the aggravating feature becomes part of the offence label rather than merely a matter that affects sentencing; or, as a more long-standing example, the separate offence of assault on a police officer. How far this can and should be carried is a matter of much dispute, as for example in relation to the separate offences of causing death by driving—why are they phrased in terms of ‘causing death by’ Page 26 of 39

Principles and Policies rather than ‘manslaughter by’ or ‘culpable homicide by’? Why is there no similar offence of ‘causing death by medical negligence’?186 Each of the four points above demonstrates the complex and contestable issues involved in implementing the principle of fair labelling. (t) Efficiency of administration Economic arguments would tend to favour broader drafting of offences, leaving appropriate distinctions in culpability to be made at the sentencing stage. A welcome reduction in public expenditure on the court system would follow. The labels given to offences are regarded as less important than the actual assessment of culpability, and this can be done expeditiously at the sentencing stage. A further efficiency argument stems from the limitations of juries and lay magistrates: the criminal law must be kept as simple as possible so as to avoid confusing lay people and producing erroneous verdicts, and this argues against finely graded offences which necessitate complex instructions on the law. Many of the reforms brought about by the Theft Act 1968 and the Criminal Damage Act 1971 involved broader offences with high maximum penalties, favouring efficiency of administration at the expense of the principle of fair labelling. However, as already noted, some respect was shown for fair labelling by, for example, retaining the separate offences of theft and deception, when it would have been possible to combine (p. 80) the two.187 The common law offence of conspiracy to defraud is the prime example of allowing administrative efficiency to prevail—prosecutors greatly prize its flexibility—and, as noted below, the government departed from the Law Commission's recommendation and has retained this broad ‘blunderbuss’ offence despite the enactment of new offences in the Fraud Act 2006.188 Although the Sexual Offences Act 2003 was drafted so as to include many differentiated but overlapping offences, many of those offences are drafted over-broadly and rely on prosecutorial discretion for the exclusion of cases—such as sexual familiarities between young people—that ought not to be criminalized. This creates a problem of mis labelling when prosecutorial discretion is not exercised appropriately and results in conviction for an unduly harsh offence.189 In fact, the whole ethos of ‘efficient administration’ needs to be questioned. ‘Efficiency’ and ‘practicality’ are presented as neutral concepts, when they often boil down to the convenience of prosecutors. Rarely does one hear reference to the efficiency of a rule in protecting individual rights. (u) The principle of contemporaneity Part of the basic doctrine of criminal law, as described by Hall among others,190 is that not only must the defendant cause the prohibited consequence and have the required fault, but that conduct and fault must co-exist at the same time. This is the principle of contemporaneity. We will see in Chapter 5.4(c) that the analysis of cases in the light of this principle can become rather difficult where there is a series of acts or a continuing act and where the fault element is only present for part of the time. (v) The doctrine of prior fault Even though the defendant did not have the required fault when performing the prohibited conduct, the doctrine of prior fault may be invoked to hold him liable—by fastening on to the defendant's fault at an earlier stage, which then led to an absence of fault at the time when the prohibited conduct took place. The title of Paul Robinson's seminal article, ‘Causing the Conditions of One's Own Defence’,191 explains the rationale of the doctrine. A person should not be allowed to rely on an exculpatory condition (e.g. lack of fault through automatism or Page 27 of 39

Principles and Policies intoxication) if he or she had deliberately or even negligently brought about that condition (e.g. by failing to take proper medication or by drinking alcohol to excess). Thus the doctrine operates by way of exception to—or, some would say, it conflicts with—the principle of contemporaneity. The doctrine, which shares some of the roots of constructive liability, is discussed further in Chapter 5.4(d). (p. 81) 3.7 Conclusions The purpose of this chapter has been to identify and to examine critically some of the theoretical and practical arguments for lawmaking and interpretation in the criminal law. The discussion has focused on arguments of principle, but this normative dimension has been linked to the practical issues involved in proposing legislation, interpreting statutes, and developing the common law. When referring to ‘principles’ the reference has been chiefly one of aspiration: it is not suggested that any of these principles is recognized as authoritative in English criminal law, and we have seen how often pragmatic or political arguments have held sway. Indeed, some of the normative propositions referred to here as ‘principles’ are not even regarded as sufficiently important to call upon legislators or judges who wish to depart from them to justify the departure. And even when a principle is recognized, its wording may be so indeterminate (e.g. maximum certainty, fair warning, fair labelling) as to impose only loose constraints. The Human Rights Act has not brought major changes. The Convention does not have widespread relevance to the substantive criminal law, but it has had some effect on approaches to defining offences and to judicial decision-making. Although the potential significance of Art. 7 was diluted by the Strasbourg Court in CR and SW v United Kingdom,192 there has been some willingness to apply the ‘quality of law’ standard when the government is trying to justify interference with the rights in Art. 5 and Arts. 8–11, for example.193 The judiciary has taken Convention rights seriously in some high-profile decisions, on such matters as reverse burdens of proof,194 certainty of definition in the offence of public nuisance,195 and of course the detention without trial of terrorist suspects.196 Although political forces often hold sway in lawmaking, it is vital that principled arguments continue to be pressed, and this supplies a good reason for an assessment of appropriate principles of aspiration. As we have seen, some paradoxes emerge from the different pairs of principles. For example, the advocates of a degree of constructive liability rely on the occurrence of significant harm, however unexpected, as a reason for increasing the grade of an offence; yet they may not place such importance on resulting harm when accepting criminal liability for attempts, incitement, and other inchoate offences. The main thrust of this chapter has been to argue for greater attention to the rule of law and to the principle of autonomy when determining the conditions of criminal liability. Thus the subjective principles, the non-retroactivity principle, the principle of maximum certainty, the principle of strict construction, the principle of fair labelling, and the presumption of innocence—all of them tend to emphasize the value of fair warning and predictability in the law, the importance of respecting choices (p. 82) made by autonomous individuals, and the need to control the exercise of power by state officials. They are at the heart of legality, of the rule of law, and of what has been termed ‘defensive criminal law’.197 Whereas welfare-based principles and policies of social defence are more relevant to criminalization decisions, the rule of law and

Page 28 of 39

Principles and Policies the principle of autonomy should have priority in relation to the conditions of liability, qualified only by a minimalist welfare principle. In some situations it will be justifiable to impose duties of citizenship reinforced by the criminal law, but in section 3.4(c) it was argued that this could be done in certain circ*mstances without compromising the principle of autonomy. This debate, and others connected with it, is taken forward in the context of criminal conduct in the next chapter.

Further Reading J. GARDNER, Offences and Defences: Selected Essays in the Philosophy of Criminal Law (2007), 33–56 and 246–8. J. CHALMERS and F. LEVERICK, ‘Fair Labelling in Criminal Law’ (2008) 71 MLR 217 –46. B. JURATOWITCH, Retroactivity and the Common Law (2008), 43–60, 127–38, 183–97. LAW COMMISSION Consultation Paper No. 177, A New Homicide Act for England and Wales?

(2005), 45–8. L. ZEDNER and J. V. ROBERTS (eds), Principles and Values in Criminal Law and Criminal Justice (2012), chs 1, 2, 3, and 5.

Notes: 1 As confirmed by the Home Office research by D. Brown and T, Ellis, Policing Low-Level

Disorder: Police Use of Section 5 of the Public Order Act 1986 (1994), and by T. Bucke and Z. James, Trespass and Protest: Policing under the Criminal Justice and Public Order Act 1994 (1998). 2 For a constructive critique of this use of the terms ‘principle’ and policy’, see J. Gardner,

‘Ashworth on Principles’ in L. Zedner and J. V. Roberts (eds), Principles and Values in Criminal Law and Criminal Justice (2012). 3 Law Com No. 177, para. 2.2, quoting from the submission of the Society of Public Teachers of

Law. See also A. T. H. Smith, ‘The Case for a Code’ [1986] Crim LR 285. 4 Law Com No. 177, i, para. 2.4. 5 See Law Com No. 177, i, paras. 2.5 to 2.11 for fuller discussion. 6 Law Com No. 143, Codification of the Criminal Law: a Report to the Law Commission (1985),

submitted by Professors J. C. Smith, E. Griew, and I. Dennis. 7 Law Com No. 177 in two volumes: i. report and draft Bill, ii. commentary on the draft Bill. 8 See [1990] Crim LR 141–2. This concern, even if valid at the time, seems difficult to maintain

in the light of, e.g., legislation such as the Criminal Justice Act 2003, with 339 sections and 38 Schedules. 9 Page 29 of 39

Principles and Policies 9 Law Com No. 218, Legislating the Criminal Code: Offences against the Person and General

Principles (1993). 10 Home Office, Violence: Reforming the Offences Against the Person Act 1861 (1998). 11 Law Com No. 237, Legislating the Criminal Code: Involuntary Manslaughter (1996). 12 Home Office, Reforming the Law on Involuntary Manslaughter: the Government's

Proposals (2000). 13 Law Com No. 311, Tenth Programme of Law Reform (2008); see the editional comment at

[2009] Crim LR 1. 14 Law Com No. 330, Eleventh Programme of Law Reform (2011). 15 See, e.g., Sir Henry Brooke, ‘The Law Commission and Criminal Law Reform’ [1995] Crim LR

911; Lord Bingham, ‘A Criminal Code: Must We Wait for Ever?’ [1998] Crim LR 694; Mrs Justice Arden, ‘Criminal Law at the Crossroads: the Impact of Human Rights from the Law Commission's Perspective and the Need for a Code’ [1999] Crim LR 439. 16 Chambers [2008] EWCA Crim 2467, at 24. 17 E.g. C. Wells, ‘Restatement or Reform?’ [1986] Crim LR 314; G. de Búrca, and S. Gardner,

‘The Codification of the Criminal Law’ (1990) 10 OJLS 559; S. Gardner, ‘Reiterating the Criminal Code’ (1992) 55 MLR 839. 18 A. Ashworth, ‘Is the Criminal Law a Lost Cause?’ (2000) 116 LQR 225. 19 P. W. Ferguson, ‘Codifying Criminal Law: (1) A Critique of Scots Common Law’ [2004] Crim

LR 49. 20 Section 3.4(g). 21 L. Farmer, ‘“The Genius of our Law …”: Criminal Law and the Scottish Legal Tradition’

(1992) 55 MLR 25 , 39. 22 Scottish Law Commission, A Draft Criminal Code for Scotland (2003), discussed by P. W.

Ferguson, ‘Codifying Criminal Law (2): the Scots and English Draft Code Compared’ [2004] Crim LR 105. The Scots codifying team consisted of Professors E. Clive, P. Ferguson, C. Gane, and A. McCall Smith, with Professor Sir Gerald Gordon. 23 See Chapter 1.3. 24 See further S. Peers, EU Justice and Home Affairs Law (3rd edn., 2011), ch 8. 25 For fuller discussion see B. Emmerson, A. Ashworth, and A. Macdonald (eds), Human Rights

and Criminal Justice (3rd edn., 2012). 26 Percy v DPP [2002] Crim LR 835. 27 Human Rights Act 1998, s. 19.

28 Page 30 of 39

Principles and Policies 28 E.g., the Bill that became the Sexual Offences Act 2003 (on which see Chapter 8.5 and 8.6);

and the provision that became s. 76 Criminal Justice and Immigration Act 2008 (on which see Chapter 4.6(g)). 29 Lambert [2002] 2 AC 545, and Sheldrake v DPP; Attorney General's Reference No. 4 of

2002 [2004] UKHL 43. 30 Notably in R v A [2002] 1 AC 45. 31 An example of its use was in A. v Home Secretary [2004] UKHL 56. 32 Article 15 was analysed by the House of Lords in A. v Home Secretary (n 31). 33 Now confirmed in Strasbourg by e.g. Rivas v France [2005] Crim LR 305. 34 In A. v Home Secretary [2004] UKHL 56 the House of Lords found that the provisions for the

detention without trial of suspected international terrorists were contrary to Art. 5 and were not saved by Art. 15, since, although their Lordships accepted the government's view that there was an emergency threatening the life of the nation, they held that the powers went further than strictly necessary and that they were discriminatory, by applying to non-nationals and not to British nationals. The government abandoned these powers and took different powers in the Prevention of Terrorism Act 2005. 35 Sections 12 and 13 of the Human Rights Act require British courts to ‘have particular regard

to the importance of’ the rights of freedom of expression and freedom of religion. 36 B.Goold, L. Lazarus, and G. Swiney, Public Protection, Proportionality and the Search for

Balance (Ministry of Justice Research Series 10/07). 37 The most significant appear to be H [2002] 1 Cr App R 59 on parental chastisem*nt

(Chapter 4.8); Percy v DPP [2002] Crim LR 835 (s. 5 of the Public Order Act may interfere disproportionately with the defendant's Art. 10 right); and Goldstein and Rimmington [2005] UKHL 63 (public nuisance and Art. 7). Cf. K. [2011] EWCA 1691, where it was noted that the relevant legislation actually refers to Art. 4 on servitude and forced labour. 38 See R v R [1992] 1 AC 599, and Chapter 8.5(b). 39 See e.g. Protection of Children Act 1978 and Sexual Offences Act 2003, ss. 45–50. 40 Protection from Harassment Act 1997, Chapter 8.3(g); see C. Wells, ‘Stalking: the Criminal

Law Response’ [1997] Crim LR 463, and the new offence in Protection of Freedoms Act 2012, s. 111. 41 Clingham v Royal Borough of Kensington and Chelsea [2003] 1 AC 787. 42 See Chapter 1.2, and A. Ashworth, ‘Social Control and “Anti-Social Behaviour”: the

Subversion of Human Rights?’ (2004) 120 LQR 263. 43 See further Chapter 4.4. 44 Children and Young Persons Act 1933, s. 1. 45 Page 31 of 39

Principles and Policies 45 Domestic Violence, Crime and Victims Act 2004, discussed in Chapter 7.6. 46 See A. Cadoppi, ‘Failure to Rescue and the Continental Criminal Law’ in M. A. Menlowe and

A. McCall Smith, The Duty to Rescue (1993); A. Ashworth and E. Steiner, ‘Criminal Omissions and Public Duties: the French Experience’, (1990) 10 Legal Studies 153. 47 The principle of fair warning is discussed in Chapter 3.5. 48 See Chapter 3.5(i), and also the particular problems of prosecutorial discretion under the

Sexual Offences Act 2003, discussed in Chapter 8.6. 49 M. Moore, Act and Crime (1993) 58; cf. the broader discussion by W. Wilson, Central Issues

in Criminal Theory (2002), ch 3. 50 Cf. P. Robinson and J. Darley, Justice, Liability and Blame (1995), 45–8, and particularly B.

Mitchell, ‘Public Perceptions of Homicide and Criminal Justice’ (1998) 38 BJ Crim 453 at 459. 51 A. Ashworth, ‘Criminalizing Ommissions’, in A. Ashworth, Positive Obligations in Criminal

Law (2013), ch 2. 52 A. Duff, Answering for Crime (2007), 109–10. 53 Discussed further in Chapter 4.7 and 4.8. 54 See e.g. Nachova v Bulgaria (2005) 42 EHRR 43. 55 J. Gardner, ‘Introduction’ to H. L. A. Hart, Punishment and Responsibility (2nd edn., 2008),

xxxvi. 56 B. Juratowitch, Retroactivity and the Common Law (2008), 49. 57 [1962] AC 220. 58 For the controversial nature of this approach to the relationship between law and morality

see Chapter 2.5. 59 A. T. H. Smith, ‘Judicial Lawmaking in the Criminal Law’ (1984) 100 LQR 46. 60 Knuller v DPP [1973] AC 435, a case in which, paradoxically, the court appeared to create

the offence of outraging public decency; Rimmington and Goldstein [2006] 1 AC 459, per Lord Bingham. 61 Khaliq v HM Advocate, 1983 SCCR 483. 62 Stallard v HM Advocate, 1989 SCCR 248, discussed by T. H. Jones, ‘Common Law and

Criminal Law: the Scottish Experience’ [1990] Crim LR 292 and by L. Farmer, ‘The Genius of our Law’. 63 [1992] 1 AC 599, on which see M. Giles, ‘Judicial Lawmaking in the Criminal Courts: the Case

of Marital Rape’ [1992] Crim LR 407. 64 Summarized in the first edition of this work, at 301–3. 65 Page 32 of 39

Principles and Policies 65 (1995) 21 EHRR 363; for analysis, see Juratowitch, Retroactivity and the Common Law,

127–38. 66 [1992] 1 AC 599. 67 (1995) 21 EHRR 363, at 390 cf. the even more doubtful decision in C [2005] Crim LR 238. 68 (1995) 21 EHRR 363, at 402. 69 [1962] AC 220. 70 [1973] AC 435. 71 [1983] QB 1053. 72 See nn 78–9. 73 [1996] AC 1, at 28. 74 In Attorney-General's Reference (No. 2 of 1999) [2000] 2 Cr App R 207, per Rose LJ at 218. 75 [2007] 1 AC 136, at [29]; cf. also Lord Bingham in Rimmington and Goldstein [2006] 1 AC

549 at 33 and Norris v Government of USA [2008] UKHL 16, at 55–6. 76 See P. H. Robinson, ‘Rules of Conduct and Principles of Adjudication’ (1990) 57 U Chic LR

729 , and P. Alldridge, ‘Rules for Courts and Rules for Citizens’ (1990) 10 OJLS 487. Cf. Law Com No. 177, cl. 4(4). 77 See the views of Dworkin and Williams, discussed in section 3.4(l). 78 See Willer (1987) 83 Cr App R 225 and the decisions summarized in Hasan [2005] UKHL 22,

discussed in Chapter 6.3. 79 Kingston [1995] 2 AC 355, discussed in Chapter 6.2(d). 80 [1987] AC 417, discussed in Chapter 6.3. 81 G. Fletcher, Rethinking Criminal Law (1978), 574, quoted by Smith, ‘Judicial Lawmaking’,

64–5. 82 [1995] Crim LR 163. 83 The decision in Elbekkay came a few months before Lord Lowry laid down his propositions

about judicial lawmaking in C v DPP, n 73 and accompanying text. 84 It might be argued that a well-drawn defence of reasonable mistake of law should be

introduced to protect defendants from conviction in such cases, but the argument here is that it would be unlawful for a court to reach such a decision at all. 85 See, respectively, Kokkinakis v Greece (1993) 17 EHRR 397 and Chan Chi-hung v R [1996]

AC 442. The question received a similar answer under European Community law in Cases C– 358/93 and C–416/93 Aldo Bordessa [1995] ECR I–361: see E. Baker, ‘Taking European Criminal Law Seriously’ [1998] Crim LR 361 , at 366–8, 376–7. 86 Page 33 of 39

Principles and Policies 86 [1973] AC 435. 87 For the former see Shaw v DPP [1962] AC 220; for the latter see Tan [1983] QB 1053. 88 E.g. by the Grand Chamber in Korbeley v Hungary (2010) 50 EHRR 1192, at 70; see also

the excerpt from the Kokkinakis judgment in section (i). 89 See n 65 and accompanying text; see also C. C. Murphy, ‘The Principle of Legality in

Criminal Law under the European Convention on Human Rights’ [2010] EHRLR 192 , at 200. 90 See, e.g., Kolender v Lawson (1983) 103 S Ct 1855; the Supreme Court of Canada applied

the principle in Prostitution Reference (1990) 77 CR (3d) 1. 91 Kokkinakis v Greece (1994) 17 EHRR 397, para. 52. 92 Kokkinakis v Greece (1994) 17 EHRR 397, para. 40. 93 Sunday Times v UK (1979) 2 EHRR 245, para. 49; see generally B. Emmerson, A. Ashworth,

and A. Macdonald, Human Rights and Criminal Justice, ch 16. 94 (2000) 30 EHRR 241. 95 Hughes v Holley (1988) 86 Cr App R 130. 96 Steel v UK (1999) 28 EHRR 603 thus upheld the definition of ‘breach of the peace’, even

though there remains some uncertainty in the definitions offered by the courts. 97 (2000) 30 EHRR 241, para. 39. 98 See the discussion of Gomez and Hinks in Chapter 9.2. 99 Law Commission Consultation Paper (LCCP) No. 155, Fraud and Deception (1999), Parts V

and VI; cf. the proposals in Law Com No. 276, Fraud (2002), which led to the Fraud Act 2006 (see Chapter 9.8), and which seek to avoid this problem by deploying an inchoate mode of drafting. 100 Pattni et al. [2001] Crim LR 570. 101 Cotter [2002] Crim LR 824. 102 [2005] UKHL 63. 103 [2005] UKHL 63, para. 33, quoting from Clark [2003] 2 Cr App R 363, para. 12. 104 J. Gardner, ‘Rationality and the Rule of Law in Offences Against the Person’ [1994] Camb LJ

502. 105 See the two Home Office studies of policing practice (reference at n 1). 106 A. P. Simester and A. von Hirsch, ‘Regulating Offensive Conduct through Two-Step

Prohibitions’ in A. von Hirsch and A. P. Simester (eds), Incivilities (2007), at 186 –7. 107 T. Endicott, ‘The Impossibility of the Rule of Law’ (1999) 19 OJLS 1 , 6. 108 Page 34 of 39

Principles and Policies 108 Endicott, ‘The Impossibility of the Rule of Law’, 17–18. 109 E.g. the problem of defining the conduct element in attempts: see Chapter 11.3. 110 (1926) 269 US 385, at 391. 111 See n 1. 112 Subject to the remarks of the House of Lords in Norris v Government of USA [2008] UKHL

16. 113 See the studies at n 1. 114 See further, on ‘public order’, N. Lacey, C. Wells, and O. Quick, Reconstructing Criminal

Law (2010), ch 6. 115 Discussed in Chapter 6.5. 116 A. Ashworth, ‘Ignorance of the Criminal Law, and Duties to Avoid It’ (2011) 74 MLR 1, and

Chapter 6.5. 117 See D. McBarnet and C. Whelan, ‘The Elusive Spirit of the Law: Formalism and the Struggle

for Legal Control’ (1991) 54 MLR 848. 118 See section 3.4(g). 119 See Chapter 9.9. 120 L. Hall, ‘Strict or Liberal Construction of Penal Statutes’ (1935) 48 Harv LR 748. Similar

reasoning (deriving more from the value of liberty than from capital punishment) may underlie the principle of giving the benefit to the accused when the Court of Appeal is faced with conflicting precedents: Taylor [1950] 2 KB 368. 121 See Chapter 8.3. 122 Kokkinakis v Greece (1994) 17 EHRR 397, para. 52, cited at n 91. 123 J. C. Jeffries, ‘Legality, Vagueness and the Construction of Penal Statutes’ (1985) 71 Virginia

LR 189. 124 E.g. Gomez [1993] AC 442 and Hinks [2001] 2 AC 241 in the House of Lords, and many

Court of Appeal decisions. 125 Attorney-General's Reference (No. 1 of 1988) (1989) 89 Cr App R 60, affirming the Court of

Appeal's decision at (1989) 88 Cr App R 191. 126 Cf. Black-Clawson International v Papierwerke Waldhof-Aschaffenber AG [1975] AC 591

with Pepper v Hart [1993] AC 593. 127 Jeffries, ‘Legality, Vagueness and the Construction of Penal Statutes’, and Law Com No.

177, para. 3.17. 128 For fuller discussion see A. Ashworth, ‘Interpreting Criminal Statutes: a Crisis of Legality?’

Page 35 of 39

Principles and Policies (1991) 107 LQR 419. 129 See Chapter 5.5(a). 130 Cf. Law Com No. 177, para. 3.17, criticized by Ashworth, ‘Interpreting Criminal Statutes,

425–7. 131 See Chapter 6.5. 132 Cf. Smith, ‘Judicial Lawmaking’, 58: ‘it may be doubted whether it is possible to formulate

any organising principles according to which conduct is seen to be deserving of condemnation as criminal’. 133 J. R. Spencer, ‘Criminal Law and Criminal Appeals: the Tail that Wags the Dog’ [1982] Crim

LR 260. Another example, which illustrates how some of these cases occur because the prosecutor chose the wrong charge and the Court cannot bring itself to acquit or quash the conviction, is Gomez [1993] AC 442, discussed in Chapter 9.2. 134 J. Bell, Policy Arguments in Judicial Decisions (1983), 222. 135 E.g. Oxford v Moss (1979) 68 Cr App R 183 and Gold and Shifreen [1988] AC 1063: the

Computer Misuse Act 1990 may be seen as a legislative response, at least to the latter decision. 136 (1976) 63 Cr App R 252. 137 [1977] AC 177. 138 See Smith, ‘Judicial Lawmaking’, 52–4, and also Lord Lane CJ, on credit card frauds in

Clarke (1982) 75 Cr App R 119. 139 Savage, Parmenter [1992] 1 AC 699, discussed in Chapter 8.3. 140 [1996] AC 815. 141 [1996] AC 815, at 831. 142 R. M. Dworkin, A Matter of Principle (1985), ch 1. 143 E.g. G. Williams, ‘Statute Interpretation, Prostitution and the Rule of Law’ in C. Tapper (ed),

Crime, Proof and Punishment (1981); ‘Criminal Omissions—the Conventional View’ (1991) 107 LQR 86, at 96. 144 See the judgment of Brennan J in the US Supreme Court in Re Winship (1970) 397 US 358. 145 (1988) 13 EHRR 379. 146 [1935] AC 462. 147 [2001] 3 WLR 206. 148 [2004] UKHL 43. 149 Page 36 of 39

Principles and Policies 149 For further discussion, see A. Ashworth, ‘Four Threats to the Presumption of Innocence’

(2006) 10 Evidence and Proof 241. 150 A. Ashworth and M. Blake, ‘The Presumption of Innocence in English Criminal Law’ [1996]

Crim LR 306. 151 See J. Temkin and A. Ashworth, ‘Rape, Sexual Assaults and the Problems of Consent’

[2004] Crim LR 328 , at 342–4. 152 See the discussions by G. Williams, ‘Offences and Defences’ (1982) 2 Legal Studies 233;

P. Robinson, ‘Criminal Law Defenses: A Systematic Analysis’ (1982) 82 Columbia LR 199; and K. Campbell, ‘Offence and Defence’, in I. Dennis (ed.), Criminal Law and Criminal Justice (1987). 153 As demonstrated by A. A. S. Zuckerman, ‘The Third Exception to the Woolmington Rule’

(1976) 92 LQR 402. 154 See Hunt [1987] AC 352, discussed by J. C. Smith, ‘The Presumption of Innocence’ (1987)

38 NILQ 223, and by P. Roberts and A. A. S. Zuckerman, Criminal Evidence (2nd edn., 2010), ch 6. 155 Unfortunately even Lord Bingham in Sheldrake v DPP [2004] UKHL 43 relied on this flawed

argument. 156 As decided by the House of Lords in G [2008] UKHL 37. 157 On which see the quotation from Gardner, n 55. 158 Adopted in LCCP 177, A New Homicide Act?, para. 2.101. 159 See Chapter 7.6. 160 Chapter 7.3. 161 The Law Commission accepted this, using this as a basis for its proposed reforms: LCCP

177, A New Homicide Act?, paras. 3.15–18; but it subsequently discarded the principle in order to achieve greater consensus: Law Com 304, Murder, Manslaughter and Infanticide (2006), para. 254ff, discussed in Chapter 7.3. 162 See Chapter 5.4, and J. Horder, ‘A Critique of the Correspondence Principle’ [1995] Crim LR

759 ; B. Mitchell, ‘In Defence of the Correspondence Principle’ [1999] Crim LR 195; J. Horder, ‘A Reply’ [1999] Crim LR 206; and V. Tadros, Criminal Responsibility (2005), 93–8. 163 For a fuller version of the argument in this paragraph, see A. Ashworth, ‘A Change of

Normative Position: Determining the Contours of Culpability in Criminal Law’ (2008) 11 New Crim LR 232. 164 J. Hall, General Principles of Criminal Law (2nd edn., 1960), 6. 165 J. Gardner, ‘On the General Part of the Criminal Law’ in A. Duff (ed.), Philosophy and the

Criminal Law (1998), 244. 166 Page 37 of 39

Principles and Policies 166 J. Gardner, ‘Rationality and the Rules of Law in Offences Against the Person’ (1994) 53

Camb LJ 502 , 509. 167 See Chapter 7.5(a). 168 See particularly J. Horder, ‘A Critique of the Correspondence Principle’ [1995] Crim LR 759,

763 and 766. 169 Gardner, 244 (reference at n 165). 170 See the discussion of these criticisms in J. Gardner, Offences and Defences (2007), 246–8. 171 It is not clear that the counter-argument by Gardner, ‘On the General Part of the Criminal

Law’, 236–9, deals satisfactorily with this point; cf. A. Ashworth, Positive Obligations in Criminal Law (2013), ch 5. 172 For fuller discussion and references, see Chapter 7.5 on manslaughter. 173 A. Ashworth, ‘The Elasticity of Mens Rea’ in C. Tapper (ed.), Crime, Proof and Punishment

(1981); G. Williams, ‘Convictions and Fair Labelling’ [1983] Camb LJ 85; J. Horder, ‘Rethinking Non-Fatal Offences against the Person’ (1994) 14 OJLS 335; B. Mitchell, ‘Multiple Wrongdoing and Offence Structure: a Plea for Consistency and Fair Labelling’ (2001) 64 MLR 393. 174 J. Chalmers and F. Leverick, ‘Fair Labelling in Criminal Law’ (2008) 71 MLR 217 , at 246. 175 Chalmers and Leverick, ‘Fair Labelling in Criminal Law’, 234. 176 Chalmers and Leverick, ‘Fair Labelling in Criminal Law’, 234–5. 177 V. Tadros, ‘Fair Labelling and Social Solidarity’ in L. Zedner and J. Roberts (eds), Principles

and Values in Criminal Law and Criminal Justice (2012), 79. 178 Chapter 9.2(a). 179 A. Ashworth, ‘Robbery Reassessed’ [2002] Crim LR 851. 180 Admittedly English law does not do this well: see Chapter 8.3(m). 181 In Chapter 5.4 and Chapter 7.5. 182 See the discussion in Chapter 7.7. 183 For detailed discussion, see Chapter 7.3. 184 Chalmers and Leverick, ‘Fair Labelling’, 244–6. 185 See further Chapter 7.4. 186 See A. Ashworth, ‘Manslaughter: Generic or Nominate Offences?’, drawing from several

other essays in C. Clarkson and S. Cunningham (eds), Criminal Liability for Non-Aggressive Death (2008). 187 Criminal Law Revision Committee, 8th Report, Theft and Related Offences, Cmnd 2977

Page 38 of 39

Principles and Policies (1966), para. 38. 188 See Chapter 9.8 and 9.9. 189 E.g. G. [2008] UKHL 37, discussed in Chapter 8.5. 190 J. Hall, General Principles of Criminal Law (2nd edn., 1960). 191 (1985) 71 Virginia LR 1. 192 (1996) 21 EHRR 363. 193 See Chapter 3.5(i). 194 See Chapter 3.5(m). 195 Goldstein and Rimmington [2005] UKHL 63. 196 A v Home Secretary [2004] UKHL 56. 197 N. Jareborg, ‘What Kind of Criminal Law do We Want?’, Beware of Punishment (1995),

discussed in Chapter 2.4(a).

Criminal Conduct: Actus Reus, Causation, and Permissions

Principles of Criminal Law (7th edn) Andrew Ashworth and Jeremy Horder Publisher: Oxford University Press Print ISBN-13: 9780199672684 DOI: 10.1093/he/9780199672684.001.0001

Print Publication Date: May 2013 Published online: Sep 2013

4. Criminal Conduct: Actus Reus, Causation, and Permissions Chapter: (p. 83) 4. Criminal Conduct: Actus Reus, Causation, and Permissions Author(s): Andrew Ashworth and Jeremy Horder DOI: 10.1093/he/9780199672684.003.0004

4.1 The general part of the criminal law 4.2 Involuntary conduct 4.3 Acts, states of affairs, and possession 4.4 Omissions 4.5 Causation 4.6 Self-defence and permissible force 4.7 Permissions, necessity, and the choice of evils 4.8 Conclusions Further reading

4.1 The general part of the criminal law

Criminal Conduct: Actus Reus, Causation, and Permissions This chapter and the following two chapters discuss what is usually known as the ‘general part’ of the criminal law.1 The general part is comprised of rules and principles of the criminal law whose importance and application can be analysed and debated without necessarily referring to a specific crime. For example, it has been traditional for writers on English criminal law to approach the analysis of offences by means of two concepts with Latin names, actus reus and mens rea. The actus reus consists of the prohibited behaviour or conduct, including any specified consequences arising therefrom. The mens rea is usually described as the mental element—the intention, knowledge, or recklessness of the defendant in relation to the proscribed conduct. Analysis of the two concepts is analysis of the general part of the criminal law, in so far as it involves a search for a better understanding of the two concepts, as such, or a search for a common thread of principle in the way that they are employed in the criminal law. These two concepts are often referred to by their Latin names, but will be expressed here in terms of ‘conduct’ elements and ‘fault’ elements. Questions involving these concepts relevant to the general part of the criminal law include whether it is ever right to convict someone of a crime when they were not at fault in engaging in the prohibited conduct. This is an important question, bearing (p. 84) in mind that, in English criminal law, there are thousands of strict liability offences (liability without fault).2 A question in a similar vein, but turning things around, is: is it ever right to convict someone of a crime because they possessed the fault element (say, an intention to harm another person), but did not engage in any conduct wrongful in itself to further that intention? Some commonly encountered offences all but assume that it is indeed acceptable to convict people for little more than possessing a fault element.3 Most controversially of all, we will see that there are examples of crimes where a criminal conviction has been upheld even when the person convicted was neither at fault, nor engaged in any voluntary conduct.4 What has just been said assumes that a rough-and-ready distinction between fault elements and conduct elements provides an appropriate starting point for consideration of such questions. There have also been efforts made to find more sophisticated ways of explaining the distinction.5 In particular, in so far as criminal conduct must in some sense be ‘wrongful’ conduct, some theorists and some legal systems draw a distinction between the conduct element of a crime, and the circ*mstances in which that conduct may be permissibly engaged in.6 For example, when, for adequate reason, a qualified surgeon operates on you, he or she inflicts harm on you—perhaps very serious harm—but no offence is committed. But is that because there is no ‘wrongful’ conduct in the circ*mstances, or because there is wrongful conduct but conduct that is permitted by the law given that there was adequate reason to engage in it?7 There is no space here to examine these issues more extensively, and the classification adopted may be said to be less important than an enquiry into the key question that it might obscure: ‘what the preconditions to criminal liability really are, and how far they really reflect the principles they are commonly supposed to encapsulate’.8 For convenience of exposition the conditions of criminal liability may be divided into four working groups: (i) act and causation requirements; (ii) absence of permission; (iii) capacity and fault requirements; (iv) excusatory defences. Chapter 5 deals with the requirements of criminal capacity (the doctrines of insanity, infancy,

Page 2 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions and corporate liability) as well as traditional fault requirements, i.e. the range (p. 85) of possible fault elements which the prosecution has to establish in order to construct a case for the defence to answer. Chapter 6 will discuss ‘excusatory’ defences. These are defences that are concerned with absence of fault or culpability in a broader sense than is understood by a ‘fault element’ when such an element is included in the definition of an offence. Indeed, these defences can usually be pleaded even when the defendant committed a crime requiring proof by the prosecution of a fault element. An example is ‘duress’ (threats) where that has led someone to commit a crime. In limited circ*mstances, the law gives someone a complete (excusatory) defence to a range of crimes if those crimes were committed under duress, even if the crime in question involved proof of a fault element (such as an intention to steal). When one or more such excusatory defences are pleaded the defendant has to provide some evidence of the facts that he or she claims gave rise to the defences’ relevance to the case, in order to raise them as live issues. Both positive and negative fault requirements can be analysed as aspects of the general part of the criminal law, although they do not apply invariably and to all offences. Finally, it should be noted that some defences may not fit easily into a category such as ‘permission’ or ‘excusatory’ defence.9 This chapter deals with both (i) act and causation requirements and (ii) absence of permission. It is fundamental to the characterization of certain conduct as criminal that it is not permissible. Some offences are drafted so as to exclude permissible conduct from the scope of the conduct element. An example is blackmail, where the wrongful conduct involved making an ‘unwarranted demand with menaces’.10 However, the most usual approach is to define offences without reference to the possibility that the conduct may be permissible under certain circ*mstances. This makes the permissions something best analysed as doctrines of the general part of the criminal law. Perhaps it is for this reason that permissions are often classified as ‘general’ defences. Some excuses—like duress—may be general defences as well, but it can be argued that permissions are of more fundamental importance than excuses to a sound functioning of a legal system. Conduct which is permissible gives a defendant a legal right to engage in it in appropriate circ*mstances, even if the conduct in question involves the intentional infliction of serious harm or even killing. This is quite different in theory from the operation of defences which are excuses such as duress or mistake, discussed in Chapter 6. Someone seeking excuse concedes that they had no right to do as they did and that their act was wrongful. They nonetheless claim that they should be acquitted because they lacked culpability (in the broader sense, described above) at the time.11 Where the defendant's act is regarded as permissible (say, in self-defence), the defendant claims the right to have done it (whether or not it was in some sense wrongful), even though the same act would in most situations be impermissible and wrong.12 The fundamental importance of permissions comes from the fact that they mostly (if (p. 86) not in all cases) afford some guidance to citizens on the circ*mstances in which they are permitted or right to use force, cause damage, etc. It follows that the legal limits of permissions, in particular, ought to comply with standards of fair warning.13 The chapter begins with an exploration of the doctrines of voluntariness, acts, omissions, and causation. To proceed to conviction without proof of voluntary conduct would be to fail, in the most fundamental way, to show respect for individuals as rational, choosing beings. More generally, if people were liable generally to conviction for simply having failed to stop harm being done, or (perhaps even more so) because something had been done to them, this would fail to respect their autonomy (see Chapter 2.1) and would be unlikely to give them fair warning

Page 3 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions of the incidence of the criminal sanction, unless reasonable duties had been made plain to them.14 Similarly, where it cannot be established that the defendant (D) was responsible for the conduct or consequence prohibited by the crime, there should be no conviction. Some such requirements are needed to protect individual autonomy by ensuring that both Parliament and the courts preserve fair warning and fair opportunities to choose not to offend. The chapter begins with the notion of involuntary conduct, the limits of which are examined in section 4.2. We then turn in section 4.3 to various challenges to the ‘voluntary act’ requirement—where is the act if the law criminalizes the occurrence of a state of affairs, or mere possession? Section 4.4 considers how the voluntary act requirement relates to crimes of omission. We then turn to causation, and later deal with the circ*mstances in which conduct may be recognized as justifiable.

4.2 Involuntary conduct (a) Automatism and authorship15 A claim of ‘automatism’ is a denial of authorship, a claim that the ordinary link between mind and behaviour was absent, or that the link had become distorted in some fundamental way.16 This can occur where what is prohibited by the law only occurred as a result of a set of involuntary movements of the body rather than as a result of voluntary acts. We should begin by noting that something that happens to one's body can be involuntary, without being ‘automatic’. An example would be where someone (X) is forcibly seized by another (Y) and physically made to harm the victim or to damage their property. In such a case, Y is the real aggressor, whereas X is guilty of no crime (p. 87) because X's involvement was involuntary: controlled wholly by Y. Such cases are not further discussed here, and the terms ‘involuntary’ and ‘automatic’ will be used interchangeably. The more interesting cases involve behaviour that is involuntary because it is in some sense automatic (using that term fairly loosely). The law's understanding of involuntary—automatic—conduct extends to instinctive reactions, as where the defendant's driving is affected when he or she succumbs to a panic reaction when a swarm of bees enters his or her car. It also includes cases of what might be called ‘mental disconnection’, where the defendant appears to have control over his or her behaviour, but in fact does not. Examples may include committing offences whilst sleepwalking or when affected by serious concussion. Complex behaviour—such as driving—may occur in such cases, but it does not manifest itself in the form of voluntary conduct. Automatism is often regarded as a defence to crime rather than as a denial of an essential component of criminal conduct. Certainly, the discussion that follows has more in common with the treatment of various excuses in Chapter 6 than with the rest of this chapter. However, the discussion of the issue here reflects the common understanding that automatism undermines the sense in which someone is engaging in ‘conduct’ at all, and thus amounts to a denial of the conduct element of the crime.17 As a matter of ‘general part’ thinking, the theory is that automatism prevents liability for all crimes. Since all crimes require a form of conduct, or of voluntary control over a state of affairs (as in possession cases), even if some of them do not require fault, it follows that automatism may lead to acquittal on any and every charge. Many of the early cases concerned motoring offences for which strict liability is imposed, and to which automatism is one of the few routes to acquittal. However, since a plea of automatism may apply to all, or Page 4 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions almost all, crimes, the courts have attempted to circ*mscribe its use, defining it fairly narrowly and developing three major doctrines of limitation. Where a defendant brings credible evidence to raise the possibility of involuntariness, the prosecution must establish beyond reasonable doubt that the accused was not in a state of automatism when the conduct occurred. In cases where the issue turns on mental malfunctioning (‘I don't know what happened; I just suddenly blacked out’), that may involve the prosecution in an all but impossible task if the defendant exercises his or her right not to undergo medical examination. Consequently, rare though automatism claims are, judges have said that, in cases such as that of the ‘blacking out’ example just given, the defence must be able to rely on expert medical evidence bearing on the defendant's state of mind at the relevant time. In particular, in such cases, D's own testimony as to his or her state of mind will be regarded an insufficient foundation for the judge to leave the issue to the jury, or for the magistrates to dismiss the charge.18 (p. 88) (b) The essence of automatism Examples of forms of involuntariness which might amount to automatism include convulsions, muscle spasms, acts following concussion, physically coerced movements, etc. Criminal lawyers used to express the legal position in terms of a requirement of a voluntary act, going on to say that an act is voluntary if it is willed.19 One criticism of this is that it does not explain how the act of will itself occurs, and suggests an infinite causal regress; 20 another is that it misrepresents and exaggerates our awareness of the movements involved in our behaviour.21 These criticisms led Hart to propose a ‘negative’ definition, describing involuntary actions as ‘movements of the body which occurred though the agent had no reason for moving his body in that way’.22 This switches attention to rare occasions of involuntariness, of which two types may be identified—behaviour which is uncontrollable, and behaviour which proceeds from severely impaired consciousness. Uncontrollable behaviour (it would not really be right even to describe the event as D's ‘behaviour’) may be illustrated thus: D is physically overpowered by X and is made to stab V. In these circ*mstances it is fair to say that this was not D's act but something which happened to D: the same view might be taken of a person brought to this country by ferry and then forced to leave the ferry and step onto British soil.23 Further examples may be conduct during an epileptic fit, and reflex actions. Turning to behaviour proceeding from a lack of consciousness, this can be illustrated by things done during a hypoglycaemic episode (which may be the result of taking insulin to correct diabetes). Both types of automatism should apply equally to offences of omission, excusing those who fail to fulfil a legal duty through physical incapacity arising from inability to control behaviour or through significantly reduced consciousness.24 Those final words bring us to an unresolved question. Must a court be satisfied beyond reasonable doubt that the defendant had a total lack of consciousness or of control over his behaviour, or will a lesser impairment suffice? In the first place, we should recall that the prosecution bears the burden of proof, ultimately, and all that the defence needs do is to bring credible evidence to support its case. In Broome v Perkins (1987),25 upholding a conviction for careless driving even though the defendant had been in a hypoglycaemic state, the Divisional Court held in effect that the defence must adduce credible evidence that the defendant was exercising no control over his bodily (p. 89) movements at the time. A similarly stringent test was applied in Attorney General's Reference (No. 2 of 1992),26 where

Page 5 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions there had been expert evidence about a condition known as ‘driving without awareness’, but the Court held that automatism requires a ‘total destruction of voluntary control on the defendant's part’, and that the alleged condition did not establish this. These decisions are inconsistent with earlier cases in which the defendant's consciousness was significantly reduced but not totally absent, and yet where this was held sufficient for an acquittal; 27 both Broome and the Reference case concerned road traffic offences, and it is possible that a more restrictive view is taken there because of the risk of false claims, but that is hardly a convincing reason for such a significant distinction in the application of the involuntariness requirement. What, then, should be the extent of the involuntariness doctrine? Hart's definition depends upon the absence of a reason for the movements of the body (‘the mind of a man bent on some conscious action’28 ), whereas the cases seem to have more to do with an absence of capacity. Glanville Williams, taking this point, argued that movements are involuntary if D is unable to avoid them.29 Not only does this involve a shift of emphasis to capacity, but it also strikes an unusual note in asking not only whether D did control the movements (were they uncontrolled?), but whether D could have controlled them (were they uncontrollable?). Williams's approach is preferable here, as the Law Commission's Draft Code recognizes; Hart's test dwells on cognition, whereas the essence of automatism is lack of volition. But there is no concealing the questions of judgment it leaves open. The Draft Code includes within automatism any movement which ‘(i) is a reflex, spasm or convulsion; or (ii) occurs while he is in a condition (whether of sleep, unconsciousness, impaired consciousness or otherwise) depriving him of effective control of the act’.30 The key concept here is ‘effective control’, and this, combined with ‘impaired consciousness’, shows how difficult it is to eliminate questions of degree even from such a fundamental aspect of criminal liability. The essence of automatism lies in D's inability to control the movement (or non-movement) of his body at the relevant time, but it may be thought unduly harsh to restrict the doctrine to cases of apparently total deprivation. The phrase proposed by the Law Commission, ‘depriving him of effective control’, would expressly empower the courts to evaluate and judge D's worthiness for a complete acquittal, whereas if the decisions in Broome v Perkins and Attorney General's Reference (No. 2 of 1992) represent the law (at least within the sphere of road traffic offences), the doctrine of automatism is unavailable whenever the court believes that there was a residual element of control in the defendant's behaviour at the time. The advantage of the Law Commission's formula would be to allow sensitivity to the special facts of unusual cases; its disadvantage would lie in the freedom left to courts to incorporate extraneous considerations into their judgments. (p. 90) At common law the courts have imposed at least three major limitations on the doctrine of automatism—by excluding cases involving insanity, intoxication, and prior fault— and it is to these developments that we must now turn. (c) Insane automatism Even if D's bodily movements are uncontrollable or proceed from unconsciousness, the doctrine of automatism will not be available if the cause of D's condition was a mental disorder classified as insanity.31 The courts originally developed this policy for reasons of social defence, since it ensured that those who fell within the legal definition of insanity were subject to the special verdict and (at that time) to indefinite detention, rather than being allowed to

Page 6 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions argue that their condition rendered their acts uncontrollable, and that they should therefore have an unqualified acquittal on the grounds of automatism. The social policy behind this judicial approach is expressed most clearly in Lord Denning's speech in Bratty v Attorney-General for Northern Ireland (1963).32 D based his defence to a murder charge on psychom*otor epilepsy, but the trial judge ruled that automatism was not available, holding that the true nature of the condition was a disease of the mind and that therefore insanity was the only defence. The House of Lords upheld the trial judge's approach, and Lord Denning affirmed that ‘it is not every involuntary act which leads to a complete acquittal’. D's behaviour may have been involuntary, ‘but it does not give rise to an unqualified acquittal, for that would mean that he would be left at large to do it again’. The proper verdict is one of insanity, ‘which ensures that the person who suffers from the disease is kept secure in a hospital so as not to be a danger to himself or others’. Moreover, Lord Denning was inclined to give ‘mental disease’ a broad definition for this purpose, so as to include ‘any mental disorder which has manifested itself in violence and is prone to recur’. This decision confirmed the dominance of the policy of social defence over considerations of individual responsibility. In practice, the effect of this strict approach has not been greatly to swell the numbers of people pleading insanity. Typically, if a defence is based on automatism but the judge rules that, since the origin of D's condition was a ‘disease of the mind’, the defence should be treated as one of insanity, many defendants decide to plead guilty to the charge rather than to persist with an insanity defence. The Criminal Procedure (Insanity and Unfitness to Plead) Act 1991 grants courts a discretion to choose committal to hospital, a supervision order, or an absolute discharge, if there is an insanity verdict.33 This still places considerable emphasis on social defence, and may not be an attractive option if the defendant's condition bears little relation to the common understanding of insanity. (p. 91) In recent years the courts have tended to transfer more varieties of involuntariness out of automatism and into insanity. The leading case is Quick (1973),34 where D's defence against a charge of causing actual bodily harm was that the attack occurred during a hypoglycaemic episode brought on by the use of insulin and his failure to eat an adequate lunch. The defence relied on automatism, whereas the prosecution sought and obtained a ruling that the condition amounted to insanity. The defendant then pleaded guilty and appealed. The Court of Appeal, quashing the conviction, held that a malfunctioning of the mind does not constitute a ‘disease of the mind’ within the insanity defence if it is ‘caused by the application to the body of some external factor such as violence, drugs, including anaesthetics, alcohol, and hypnotic influences’. This ‘external factor’ doctrine was accepted by the House of Lords in Sullivan (1984),35 where it was also restated that ‘diseases of the mind’ include both permanent and transitory conditions. Thus, where the malfunctioning of the mind is caused by an external factor, the legal classification is automatism rather than insanity, and the prosecution must disprove D's claim; where it arises from an internal cause, the classification is insanity, and the burden of proof lies on D. This leads to the apparently strange result that a hypoglycaemic episode (resulting from the taking of insulin to correct diabetes) falls within automatism, whereas a hyperglycaemic episode (resulting from a high blood-sugar level which has not been corrected) falls within insanity, since it is an internal condition rather than a condition caused by an external factor.36 Epilepsy falls within insanity for the same reason.

Page 7 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions The courts have also applied the internal–external distinction to cases of somnambulism. In Burgess (1991)37 the Court of Appeal held that, since there is no external cause of sleepwalking, this condition must be regarded as arising from internal causes and therefore classified as insanity, following Quick and Sullivan. The defendant in Burgess had not changed his plea to guilty but succeeded on a plea of insanity. Now, under the Criminal Procedure (Insanity and Unfitness to Plead) Act 1991, it would be open to a judge to grant an absolute discharge in these circ*mstances. However, the ‘insanity’ label might be unwelcome to many such defendants, and English law has no satisfactory means of dealing with cases involving both danger and an absence of responsibility.38 The case for an urgent review of the ‘external factor’ doctrine is strong.39 One type of condition that has not yet been classified authoritatively in England is ‘dissociation’, which is often marked by a short period of uncharacteristic behaviour (p. 92) accompanied by some degree of memory loss. In Rabey (1978)40 the Supreme Court of Canada ruled, in the case of a defendant who attacked a woman who had rejected his admiration for her, that the dissociative state in which he acted could not be classified as automatism. Although D's rejection by the woman might be regarded as an external factor, ‘the ordinary stresses and disappointments of life which are the common lot of mankind do not constitute an external cause constituting an explanation for a malfunctioning of the mind which takes it out of the category of a “disease of the mind”’. Thus the rejection was an external factor but not the primary cause of the dissociative state: the Supreme Court thought that this lay in the defendant's ‘psychological or emotional make-up’. That approach left open the possibility that an utterly extraordinary event might suffice as an external cause, and a trial judge so ruled in T (1990).41 Here the defendant had been raped three days before she joined two others in a robbery, during which she said, ‘I'm ill, I'm ill,’ and then stabbed a bystander. Her defence was one of automatism arising from post-traumatic stress disorder caused by the rape. The judge ruled that the rape was a sufficient external cause to place the case within the doctrine of automatism rather than insanity.42 (d) Automatism through intoxication The Court of Appeal in Quick held that automatism arising from intoxication does not fall within the definition of insanity. However, this does not mean that a person who causes harm whilst in such an intoxicated state as to have significantly reduced consciousness or to be unable to control movements of the body should be brought within the doctrine of automatism. If the cause of the involuntariness is intoxication, then the courts treat the case as falling within the ambit of the intoxication doctrine. It is rare for the evidence to be strong enough to raise a reasonable doubt that D was sufficiently intoxicated as to be in a state of automatism, but this seems to have been accepted in Lipman (1970),43 where D had taken drugs and believed that he was fighting off snakes and descending to the centre of the earth, whereas he was actually suffocating his girlfriend. A defence of automatism was refused, and the case was treated as one of intoxication,44 drawing on the doctrine of prior fault discussed in (e). However, if D's condition appears to have arisen through intoxication followed by concussion resulting from a bump on the head, the court may have to establish the dominant cause of (p. 93) the condition and subsequent behaviour.45 The distinction may seem a complication too far, but consider this example. An air traffic controller goes on duty whilst heavily intoxicated. She is so intoxicated that she collapses unconscious when performing a vital part of her work, endangering many lives. If she is charged with an endangerment offence of some kind, there

Page 8 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions seems to be no pressing reason to grant the defendant an automatism-based defence. However, it should arguably be different if her intoxication led her accidentally to fall, hit her head hard, enter a mental state akin to sleepwalking in consequence, and then collapse at work as just described. In the latter case, the causal influence of the intoxication is just the background in which another cause of automatism—the concussion—governs her behaviour. (e) Prior fault The aim of the doctrine of prior fault46 is to prevent D taking advantage of a condition if it arose through D's own fault. In relation to automatism, the point was first made in Quick (1973),47 where Lawton LJ held that there could be no acquittal on this ground if the condition ‘could have been reasonably foreseen as a result of either doing or omitting to do something, as, for example, taking alcohol against medical advice after using certain prescribed drugs, or failing to take regular meals whilst taking insulin’. According to this view, the question of prior fault is resolved by applying the test of reasonable foreseeability, the test of the reasonably prudent person in D's position. But in Bailey (1983)48 the Court of Appeal held that a person should not be liable to conviction if the condition of automatism arose through a simple failure to appreciate the consequences of not taking sufficient food after a dose of insulin, even if the reasonably prudent person would have realized it. The defence of automatism should be available unless it can be shown that D knew that his acts or omissions were likely ‘to make him aggressive, unpredictable and uncontrolled with the result that he may cause some injury to others’. On this view, prior fault requires awareness of risk, sometimes called subjective recklessness.49 The conflict between the doctrine of prior fault and the principle of contemporaneity of conduct and fault is discussed elsewhere.50 The question here is whether the doctrine should apply at all in automatism cases. Consider the approach of trying to avoid the conflict with the contemporaneity principle by convicting D in respect of conduct at an earlier point in time, when there was fault. In Kay v Butterworth (1945)51 D fell asleep while driving home from night-work, and his car collided with soldiers marching down the road. It was held that he could be convicted of careless driving—not in respect of the collision (when he was asleep and therefore (p. 94) involuntarily omitting to exercise due care), but in respect of his earlier failure to stop driving when he felt drowsy. Even on its own terms, this approach is possible only where the offence is of a continuing nature, and where the charge can be appropriately worded. Having said that, one advantage of this approach is that it recognizes that the driving was at one stage involuntary, and that involuntary movements cannot be the subject of criminal liability. The application of prior fault in cases such as Quick fails to take this point, in the sense that criminal liability still depends on, or is traced through, the involuntary movements.52 Only if one maintains that the doctrine of prior fault is so fundamental to our notions of responsibility that it trumps ordinary causal principles, as well as the principle of contemporaneity, can the law's position be rationalized. (f) Reform The proposition that people should not be held liable for conduct that is involuntary is fundamental, and the common law on automatism has developed from it. However, even accepting that cases of prior fault should continue to be excluded from automatism and that cases resulting from intoxication should be classified under the intoxication rules, one major

Page 9 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions unsatisfactory feature of the law on automatism is the line drawn between this doctrine and the defence of insanity. Since the courts have flexible powers of disposal under the 1991 Act, it may be argued that judicial persistence with the internal/external distinction does not have drastic implications for defendants. Nonetheless, there can be no sense in classifying hypoglycaemic states as automatism and hyperglycaemic states as insanity, when both states are so closely associated with such a common condition as diabetes. The difference in burdens of proof (prosecution must disprove automatism, defence must prove insanity) compounds the anomaly. The proper boundaries of the defence of insanity will be examined further in Chapter 5.2(c), but it is apparent from the discussion here that the present scope of the phrase ‘disease of the mind’ is too wide. On the one hand, there are many states in which the functioning of the mind is affected but which should not sensibly be included within the concept of insanity. On the other hand, it is difficult to arrive at a clear definition of automatism: the Draft Code refers to ‘impaired consciousness … depriving him of effective control of the act’.53 This rightly recognizes that total absence of control should not be required, but it therefore leaves us with a test dependent on a judgment of degree and value (‘effective’), and does so without identifying the relevance of the defendant's capacity rather than awareness and ‘choice’.54 (p. 95) 4.3 Acts, States of affairs, and possession Accepting that a person should not be held liable for things which occur whilst he or she is in an involuntary state amounting to automatism, should there be a further requirement that liability should be based on acts? At first blush it seems wrong that people should be held liable for things that happen to them, or for a simple failure to do something. Do legal systems succeed in avoiding the creation of offences that do not require an act? Should they try to avoid such offences?55 Before sketching answers to those two questions, we must make the point that not all criminal offences are formulated so as to require proof of a particular type of act. For some offences, such as wounding and rape, the definition specifies an act and it is clearly a wrongful act in itself. For other offences, such as doing an act with intent to impede the apprehension of a person who has committed an arrestable offence,56 and all crimes of attempt, the definition requires an act, but not one that is in itself necessarily wrongful: the intention with which the act is done makes a crucial contribution to the wrongfulness of the act, but the act requirement still functions so as to exclude involuntary movements. (Whether ordinary acts should be penalized simply because of the actor's intentions is discussed elsewhere.57 ) For yet other offences, the definition refers only to a result (e.g. causing death), and the act requirement is implicit; any kind of act suffices. Those offences have a tendency to raise questions of causation (did D's act cause the death?), which draws attention to another feature of the act requirement: what is necessary is not merely an act, but an act that causes the conduct or consequence specified in the definition of the offence. This should rule out cases in which D's act is superseded by the voluntary intervening act of some third party—where it is the intervening act, and not D's original act, that is the cause. The troublesome decisions on voluntary intervening acts are reviewed in section 4.5. There are three types of offence that appear to challenge the requirement of an act. First, there are offences relating to states of affairs: is it right that a person should be liable to

Page 10 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions conviction in respect of a state of affairs that happens to him, and is not his act? Secondly, most criminal codes contain offences of possession, and it is questionable whether these require any act. Thirdly, and most obviously, there are offences of omission. The essence of these offences is that they penalize a person for doing nothing when he or she should have done something. We examine in the next section whether, and to what extent, offences of omission can be justified. In the remainder of this section, states of affairs and offences of possession are considered. (p. 96) (a) Situational liability Are there good reasons for convicting a person simply because a state of affairs exists, without the person ‘doing’ anything to create or to continue that state of affairs? The leading case is Larsonneur (1933),58 where D left England because the duration of her permitted stay had come to an end. She went to Ireland, from where she was deported back to this country. On her return, she was convicted of ‘being found in the United Kingdom’ contrary to the Aliens Order 1920. Her appeal, based on the argument that her return to England was beyond her control, was dismissed by the Court of Criminal Appeal. The case is widely criticized: her return to this country was not her own act, and was contrary to her will and desire. The Court might have held that there was no voluntary act by the defendant, since it appears that various officials compelled her return to this country. It might then have given consideration to the degree of any prior fault on her part.59 The judgment fails to discuss these points of principle, and the decision hardly shines as a beacon of common law reasoning. However, Larsonneur does not stand alone. In Winzar v Chief Constable of Kent (1983)60 the Divisional Court confirmed a conviction for being found drunk on a highway, in a case where the defendant had been taken from a hospital on to the highway by the police. Another similarly worded offence is that of being drunk in charge of a motor vehicle, and there are many other offences that impose what Peter Glazebrook has termed ‘situational liability’.61 We will see in Chapter 5.3(b) how, in certain situations, the courts have imposed ‘vicarious liability’ on shop owners and employers by construing statutory words so as to achieve convictions. In effect, these individuals and companies are being held liable simply for states of affairs—for the fact that an employee sold American ham as Scottish ham, for example, even though the shop owner had specifically warned against this.62 However, Andrew Simester has argued that in all these cases it is not the absence of a required act that is objectionable, but the absence of a fault element.63 The proper approach, he submits, is evident from two New Zealand prosecutions of visitors for staying after the expiration of a visitor's permit. In Finau v Department of Labour (1984)64 the conviction was quashed because D was pregnant and no airline would carry her. In Tifa*ga v Department of Labour (1980)65 the conviction was upheld because D was at fault in running out of money, with the result that he could not afford a ticket. The offence did not require an act (or an omission), but rather a state of affairs for which D was responsible. Thus, as argued in Chapter 5.3(b), it may be defensible (p. 97) to impose situational liability if the law is so phrased as to ensure that defendants are in control of their activities and know about their duty to avoid certain situations. This insists on a voluntariness requirement, but not an act requirement. So long as fair warning is given of the standards expected of those embarking on certain activities or enterprises, the principles of legality or ‘rule of law’ are satisfied and autonomy is respected.66 The English legislature, unfortunately, sees no objection to creating state-of-affairs offences such as ‘being found’ or ‘being drunk in charge’ without any voluntariness requirement—not

Page 11 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions even exceptions to cover the person who has been manhandled into the position in which he or she is found or the person who has been rendered drunk by the strategem of others.67 The courts have failed to develop the common law so as to provide a defence of compulsion or to insist on proof that D was responsible (i.e. voluntarily) for the conduct, result, or state of affairs proscribed. It is interesting to contrast English law in this respect with the rules developed by the Supreme Court of the United States, taking us back to the ‘prior fault’ doctrine. In Robinson v California,68 the Supreme Court held that it was unconstitutional, as a form of cruel and unusual punishment, to make someone criminally liable merely for being a drug addict. The Robinson decision has been used to strike down state laws criminalizing simple vagrancy or homelessness. However, in Powell v Texas,69 this narrow concession was not built on further. The Court held that where D, a chronic alcoholic, was charged with being found in a state of intoxication in a public place, his inability (if such it was) to stop drinking to excess did not make it cruel and unusual to punish him when he appeared in that state in public. The Powell Court distinguished the Robinson case on the grounds that in the latter case, D had been punished with imprisonment merely for being in a certain state, namely alcoholism. By contrast, the defendant Powell could have avoided public places when intoxicated even if his intoxication was involuntary (which the Court doubted that it was). In Powell, thus, D's alcoholism was regarded as nothing more than an explanation of how he came—voluntarily— to commit the crime.70 (b) Offences of possession English law contains several offences of possession, relating to such items as offensive weapons,71 any articles for use in a burglary, theft, or deception,72 and controlled drugs.73 Sometimes possession is the basic element of a crime in the inchoate mode, such as possessing drugs with intent to supply.74 In ordinary language, one might agree that it is possible to possess an item without any act on one's part. Are offences of (p. 98) this kind therefore contrary to principle? Most of the difficulties with the concept of possession have arisen in drugs cases. The leading decision is that of the House of Lords in Warner v Metropolitan Police Commissioner (1969),75 but neither the speeches of their Lordships nor subsequent cases have rendered the law clear or principled. The first proposition is that a person is not in possession of an item that has been slipped into her bag or pocket without her knowledge. The second proposition is that if a person knows that an article or container has come under her control, she is deemed to be in possession of it even if mistaken about its contents, unless the thing is of a wholly different nature from what was believed.76 The exception is extremely narrow: Warner believed that certain bags contained scent when in fact they contained cannabis, but that was held not to be a sufficiently fundamental mistake, and his knowledge that he had the bag was sufficient. In Warner Lord Pearce stated that the mistake would not be sufficiently fundamental if D thought the containers held sweets or aspirins when in fact they held heroin.77 The narrowness of this exception to the second proposition throws attention back to the first proposition, but that has also been confined tightly. In Lewis (1988)78 it was held that D was rightly convicted of possessing controlled drugs when they were found in a house of which he was tenant but which he rarely visited. His defence was that he neither knew nor suspected that drugs were on the premises. The Court of Appeal appeared to hold that, since he had the opportunity to search the house, he should be held to possess items that he did not know about but could have found. In effect, this

Page 12 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions reduces the first proposition almost to vanishing point. Surely it could equally be said, of the person into whose bag drugs are slipped by some third party, that she could have searched her bag and found them? Probably this is another example of the so-called ‘war against drugs’ resulting in the distortion of proper legal standards. The reason for enacting offences of possession is that they enable the police to intervene before a particular wrong or harm is done: in effect, these offences extend the scope of criminal liability beyond the law of attempts.79 One ground for questioning possession offences is that they may criminalize people at a point too remote from the ultimate harm, not allowing for a change of mind. Another pertinent question is whether they depart from the voluntariness requirement. Although taking possession of an article will often (but not always) involve some act of the defendant, it is surely wrong to regard the conduct as voluntary if D was substantially mistaken as to its contents. Thus the first proposition in Warner is right in suggesting that possession is not purely a physical matter but does have a mental component, although wrong in restricting that fault element to the mere realization that some item or container has arrived in one's pocket, bag, or house. The Court of Appeal has been pressed to (p. 99) broaden the fault element, notably in Deyemi and Edwards (2008),80 chiefly by reference to those House of Lords decisions such as B v DPP and K,81 which stated that the presumption of mens rea is a constitutional principle. The Court felt itself bound by previous decisions on possession of firearms, which follow the Warner approach, but certified a point of law of general public importance for the House of Lords. Until the decision in Warner is revisited, it remains objectionable that the English courts have failed to adhere to any basic voluntariness requirement, and have also ridden roughshod over normal principles of causation, which would operate so as to relieve D from liability when the voluntary act of a third party had brought about the possession.82

4.4 Omissions Omissions are controversial for two main reasons—first, there is the question whether and to what extent it is justifiable to criminalize omissions rather than acts; 83 and secondly, there is the question whether liability for omissions violates the ‘act requirement’ in criminal law. Pursuing the second point here, much has been made of the importance of requiring proof that the defendant voluntarily did something to produce the prohibited conduct or consequence. In so far as this can be termed an ‘act requirement’, are omissions a true exception to it?84 If they are, is this another argument against criminalizing them? One much-discussed preliminary question is the distinction between acts and omissions.85 Sometimes it is argued that certain verbs imply action and therefore exclude liability for omissions, and that the criminal law should respect the distinctions flowing from this. English courts have often used this linguistic or interpretive approach. It has led to a variety of decisions on different statutes,86 without much discussion of the general principles underlying omissions liability. The Law Commission's Draft Criminal Code may be said to signal the continuation of this approach, by redefining the homicide offences in terms of ‘causing death’ rather than ‘killing’, and redefining the damage offences in terms of ‘causing damage’ rather than ‘damaging’, so as ‘to leave fully open to the courts the possibility of so construing the relevant (statutory) provisions as to impose liability for omissions’.87 The Draft Code would therefore remove any linguistic awkwardness in saying, for example, that a parent killed a

Page 13 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions child (p. 100) by failing to feed it; but it does so in this specific instance, and without proclaiming a general principle that the act requirement may be fulfilled by an omission if a duty can be established. Attachment to the vagaries of the language is no proper basis for delineating the boundaries of criminal liability. In some situations the courts, following the linguistic approach, have nevertheless found themselves able to impose omissions liability. In Speck (1977)88 the defendant was charged with committing an act of gross indecency with or towards a child. The evidence was that an 8-year-old girl placed her hand on his trousers over his penis. He allowed the hand to remain there for some minutes, causing him to have an erection. The Court of Appeal held that the defendant's failure to remove the hand amounted to an invitation to the child to continue with the act, and that the offence would then be made out. In effect, the Court either held that his inactivity in those circ*mstances constituted an invitation which amounted to an act, or it created a duty in an adult to put an end to any innocent touching of this kind, with omissions liability for not fulfilling the duty. The analysis is similar to that in Miller (1983),89 where D fell asleep whilst smoking, woke up to find the mattress smouldering, but simply left the room and went to sleep elsewhere. He was convicted of causing criminal damage by fire, on the basis that a person who initiates a sequence of events innocently and then fails to do anything to stop the sequence should be regarded as having caused the whole sequence. On this view the conduct constitutes a single, continuing act; Miller caused the damage because he took no steps to extinguish the fire he had innocently started. It must be doubted whether these efforts to find an act which then coincides in point of time with the defendant's knowledge or intention are convincing.90 Surely the courts are imposing liability for an omission in these cases, by recognizing that a duty arises. Speck is a little different from Miller since the original act in Speck was that of the girl, and the duty must therefore amount to the recognition of an obligation on an adult to put an end to an indecent yet innocent touching by a child. In so far as these decisions appear to extend the statutory wording, are they objectionable on grounds of retroactivity and lack of fair warning, or defensible as applications of existing common law doctrine to new situations? In other situations it seems possible to offer plausible reasons for regarding the same event as either an act or an omission, and in some cases the courts have sought to exploit this ambiguity when dealing with problematic medical issues.91 Yet it is one thing to say that a healthcare professional who decides not to replace an empty bag for a drip-feed has omitted to do something, whereas switching a ventilator off is an act. It is another thing to maintain that the act–omission distinction should be crucial to any determination of the criminal liability in the two situations. In Airedale NHS Trust v Bland (1993)92 the House of Lords held that it would be lawful for a doctor to (p. 101) withdraw treatment from a patient in a persistent vegetative state, even though death would inevitably be hastened by that conduct. The House held that the withdrawal of treatment would constitute an omission, and thus regarded the duties of the doctor as the central issue.93 The decision was that a doctor has no duty to continue lifesupporting treatment when it is no longer in the best interests of the patient, having regard to responsible medical opinion.94 However, the Court of Appeal declined to adopt this subterfuge in Re A (Conjoined Twins: Surgical Separation),95 holding that the surgical separation of the twins would undoubtedly be an act, and subsequently deciding that carrying out an operation which would result in the death of one twin in order to save the life of the other was permissible as a necessity. This required the Court, in effect, to recognize a new defence of ‘balance of evils’ in English law—which was what the House of Lords tried to avoid in Bland, by construing

Page 14 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions the withdrawal of treatment as an omission and then focusing attention on the existence of a duty. The question thus arises again: is there any clear means of distinguishing acts from omissions? It has been argued that conduct should be classified as an omission if it merely returns the victim to his or her ‘natural’ condition, or the condition in which she would have been but for D's attempt to carry out treatment, or a rescue.96 Disconnecting a life-support machine would therefore not be classified as an act because it merely returns the patient to the condition in which he or she would have been without any treatment. This view is open to several objections, notably that of deciding what the ‘original condition’ is in relation to each actor, and the implication that a person who has saved a non-swimmer from drowning could, on discovering that the non-swimmer is an enemy, leave him in the water.97 However, one advantage of categorizing the conduct as an omission is that it then makes liability depend on the recognition of a duty, which would be straightforward in the case of the rescued nonswimmer. This approach may therefore offer comfort to those who insist that the act–omission distinction should not be used to avoid or foreclose moral arguments about the proper limits of criminal liability. But it is not a clear distinction, since it remains open to manipulation in different situations. The conclusion must therefore be that, although there are some clear cases of omission and some clear cases of act, there are many ambiguous cases in which the act–omission distinction should not be used as a cloak for avoiding the moral issues.98 (p. 102) This demonstration of the fragility of the act–omission distinction and of the vagaries of the English language indicates that it may be simplistic to oppose omissions liability in principle. There are some clear cases of omission in which it is desirable to have criminal liability, such as the parent who neglects to feed her or his child or neglects to protect it from abuse.99 Omissions can be involuntary or not, in the same way as acts; and, provided that the harm resulted because D failed to intervene, it can be argued that omissions are also causes.100 Omissions liability may therefore satisfy the principle that no one should be held liable for bodily movements that he or she did not and could not direct. It may also satisfy the principle that no person should be held liable for conduct or consequences that he or she did not cause. But one point of the act requirement is to exclude liability for mere decisions and failures to think that do not result in some kind of behaviour, and omissions fall foul of that.101 However, there are exceptions to the act requirement for a good reason—that certain positive duties to act are so important that they can rightly be made the subject of criminal liability. Of course, such a duty should also be defined with sufficient certainty, and should be adequately discoverable by those to whom it applies. So long as these formal requirements are fulfilled there can be no fairness objection to holding a person liable, provided that he or she is capable of taking some steps to carry out the duty.

4.5 Causation At the beginning of this chapter it was stated that causation can be one of the most basic requirements of criminal liability. For those offences that merely require conduct, the voluntariness requirement is crucial.102 For the many crimes which specify proof of consequences, whether or not stemming from voluntary conduct, the requirement of causation assumes a central place. Of course, as we shall see in Chapter 5, the law often goes further and insists not only that the defendant voluntarily caused the offence, but also that he did so

Page 15 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions knowingly, intentionally, and so on. Here, however, the concern is to explore the minimum conditions for criminal liability, of which causation can be one. Two kinds of consideration are of special significance in law when deciding whether a person or persons ‘caused’ something to occur. First, there are reasonable expectations about how things will or may turn out, if something is done or not done. Suppose that you leave a dog locked in a car in extremely hot weather for the whole day, and fail to feed your baby for several days, and they both die. Unless some other exceptional (p. 103) kind of explanation for the deaths is given, our reasonable expectation that the conduct you engaged in must inevitably—and therefore did—cause the deaths will prevail, when furbished (as would be normal at any trial) with scientific evidence showing the effects on the victims respectively of dehydration or lack of nutrition. The test of reasonable expectations is, though, mainly an exclusionary test. It rules out certain kinds of explanation for actions and events. Suppose D leaves a young baby in the middle of a wood, where the baby is later killed by a bird of prey or a fox. Such an outcome can be regarded as caused by D, if it is within the range of outcomes that might be expected to follow in the ordinary course of events from what D did. By contrast, if the baby left in the woods is killed by an earthquake, such an event not having occurred in that region for 1,000 years, we are unlikely to say that the baby's death was caused by D's conduct. Even though the baby would—we assume—not have been killed in the earthquake had he or she not been left in the woods, the death in the earthquake is outside the range of what might reasonably be expected to follow in the ordinary course of events from being left in the woods. In some cases, of course, there may be more than one possible cause of an event that falls within the scope of what might reasonably be expected to occur. Suppose that—at the same moment—D1 shoots V in the leg, and D2 stabs V in the stomach. V subsequently dies from loss of blood. In such a case, forensic evidence may conceivably show that either the shooting or the stabbing had no impact on the course of events. However, the likelihood is that both the stabbing and shooting will have played some part in causing V's death. As we will see, the law makes allowance for this possibility in its understanding of legal causes of outcomes, and hence both D1 and D2 may be regarded as having caused V's death. It is important to note that this ‘expectations principle’ can apply to human interventions, whether they are accidental or deliberate. Suppose, to vary the earlier example, that an armed criminal strays into the woods. Clearly, if the criminal decides to kill the baby for some reason, then this act will be the cause of the baby's death rather than D's original act in leaving the baby in the woods.103 However, suppose that the criminal's gun simply goes off unexpectedly as he is walking along, and by a tragic accident the baby is shot dead by the bullet. In such a case, we are likely to say that the sheer unexpectedness of such an outcome breaks the chain of causation leading from D's original act of abandoning the baby in the woods to the baby's death. Even though the killing was unintentional, it is thus the armed criminal who caused the death, not D. Finally, reasonable expectations must be described at the right level of specificity if they are to do the work they need to do to guide judgments in causation cases. For example, (p. 104) suppose D is chasing V with hostile intent through a dark forest, and in the dark V trips over a treasure chest, hits his head, and dies. D may be found to have caused V's death, if tripping over something, falling, and suffering a mortal wound in making reasonable efforts to escape D through a dark forest is the kind of accident we accept as being within the range of things that could well happen to V in the circ*mstances. That tripping and falling over a treasure chest was wholly unexpected and unforeseeable is quite irrelevant.104

Page 16 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions This brings us to the second kind of consideration of special relevance to the law's understanding of causation. This is the principle of autonomy, discussed in Chapter 2.1. The autonomy principle is sometimes employed by the courts to overlay or ‘trump’ the expectations-based set of causal considerations. Suppose D strikes V and leaves V unconscious in an area of town known for very frequent fatal shootings of vulnerable people so that they can be robbed. Whilst unconscious, V is shot dead by a robber. In this case, the shooting may well be regarded as something coming within the range of what might be expected to occur as a consequence of D's actions. It will, nonetheless, not be regarded as a consequence brought about by D, if it was a free, deliberate, and informed act—an autonomous act—on the part of the robber.105 We will consider this notion later in this section. Before looking further into the common law approach, a further important factor must be mentioned. There can, of course, be more than one cause of an event. It would, therefore, be possible to find that two unconnected people had a hand in bringing about that event, for the purposes of establishing the separate liability of each in criminal or civil law.106 Suppose D1 intentionally stabs V in order to cause a life-threatening injury. V is taken to hospital where, through negligence on the part of D2 (a doctor), inappropriate treatment is given to V that might have saved her life relatively easily. Let us assume that the treatment does not manifest negligence so appalling that it falls outside what might have been within the bounds of expectation, and thus does not in itself break the chain of causation from D's action to V's death (a point considered further below). In that case, both D1 and D2 may be found to have had a causal influence in killing V. It is thus perfectly possible for D1 to be found guilty of murder (causing V's death through an intention to kill or seriously injure), and for D2 to be found liable to pay damages in civil law for having caused V's death through a negligent breach of a duty of care to V.107 Notice the implicit influence of the principle of welfare in reaching the latter conclusion. We now take it for granted that there (p. 105) are emergency services under a duty to take stab victims to hospital, and that when that happens, doctors and nurses owe duties of care—shaped by exacting professional standards—to treat the victims as a high priority, and with all the skills at their disposal. These assumptions guide what falls within the scope of our expectations concerning what is likely to happen to injured victims of crime, and hence concerning who should be regarded as having had a hand in bringing about what happens to them. Analysis of who did what to whom, whether in criminal or in civil law, cannot be undertaken in isolation from broader assumptions about the rights and duties created by social and political structures.108 (a) The general principle The definitions of many crimes require that D caused a result (e.g. murder, grievous bodily harm, criminal damage) or that he caused a result by certain means (e.g. causing death by dangerous driving). In cases where it is clear that D either intended to cause the result or knowingly risked causing it, the causal enquiry is likely to be brief because no court will see much merit in the argument that the result was highly unlikely in the circ*mstances and probably a coincidence. Thus the dictum ‘intended consequences are never too remote’ is one expression of the strong effect which culpability has in hastening a finding of causation and overlooking restrictive policies which might otherwise be invoked.109 Where the culpability element does not overshadow the issue—and particularly in crimes of strict liability, where no culpability may be required—the question arises what minimum connection must be established between D's conduct and the prohibited result. Although courts have occasionally

Page 17 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions succumbed to the temptation to say that causation is a question of fact for the jury or magistrates,110 there ought to be guidance on the principles to be applied when assessing the significance of those facts. Some decisions have attempted to articulate principles, but how coherent they are is a matter of debate. The general principle is that causation is established if the result would not have occurred but for D's conduct, although support for this principle in the courts is not unwavering. In Cato (1976),111 for example, the Court of Appeal expressly stopped short of the ‘but for’ test. D had been convicted of the manslaughter of V, whom he had (p. 106) injected with a heroin compound at V's request. On the issue of whether D's injection of the heroin could be said to have caused V's death, the Court stated that: ‘as a matter of law, it was sufficient if the prosecution could establish that it was a cause, provided it was a cause outside the de minimis range, and effectively bearing upon the acceleration of the moment of the victim's death’.112 The Court later stated that the cause must be ‘a cause of substance’, although it held that the term ‘substantial cause’ would be putting the requirement too high.113 Clearly, the Court was reluctant to accept ‘but for’ causation here, fearing that the link between D's conduct and V's death might be too tenuous. Whatever one makes of the Court's reasoning on the facts of the case, it was right not to endorse the but-for test wholeheartedly. The text is both under- and over-inclusive. The Court in Cato was concerned about its underinclusiveness: that the but-for test excludes some causes of events that are highly significant even if the events could or would have occurred without them. Suppose D1 makes V drink a poison that has a 60 per cent chance of killing V. Whilst V is incapacitated by the poison, D2 (unconnected with D1) later pours a weaker version of the poison down V's throat that raises the chance of V dying from the poison to 80 per cent. V dies from the effects of the poison. In this case, V might well have died from the effects of the poison even if D2 had done nothing but stand and watch. It cannot be proved that D2's contribution to events was a but-for cause of V's death. However, D2's contribution ought almost certainly to be regarded as a cause— along with D1's conduct—of V's death. Without supplementation, the but-for test can also seem spectacularly over-inclusive, if it is not understood in a sophisticated way. Suppose that D robs V. Someone might say, ‘But-for the actions of D's grandparents in conceiving D's parents, D would never have existed to perpetrate the robbery. So, the grandparents were a but-for cause of the robbery’. That kind of reasoning takes too undiscriminating a view of causation.114 The law's starting point in its search for causes is the human conduct that led to the consequences complained of. The law's starting point is not the human conduct—or other factors—that created or shaped the person themselves whose conduct then led to those consequences.115 That principle leaves plenty of scope for the law to pay attention to, for example, the causal influence of other people's conduct on the conduct of the person that caused the consequences. That is the normal approach when the question is whether X assisted or encouraged D in some way to commit a crime against V. The principle as it has just been expressed also leaves plenty of scope for considering the causes of D's conduct for the purposes of deciding if D should, say, be excused or exempted from liability on the grounds of, for example, duress or insanity. By contrast, (p. 107) whilst the actions of D's grandparents explain how D came to exist, they do not explain how D came to commit the crime. Medical cases provide an example of another complicating factor. This is the tendency of the courts to use lack-of-causation arguments as a way of creating scope for doctors to

Page 18 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions administer treatments, such as pain-killers, that may themselves accelerate death even whilst they are having their intended pain-killing effect. At the celebrated trial of Dr Bodkin Adams (1957), charged with murdering a patient by administering excessive doses of morphine, Devlin J stated the orthodox view that to shorten life by days and weeks is to cause death no less than shortening it by years, but he added that a doctor ‘is still entitled to do all that is proper and necessary to relieve pain and suffering even if the measures he takes may incidentally shorten life’.116 This direction to the jury might be thought compatible with the principle subsequently espoused in Cato, that a de minimis contribution (i.e. a minimal cause which ‘people of common sense would overlook’117 ) is not a sufficient cause in law. However, this probably does not capture the precise point of the Adams direction, which is rather that a doctor's administration of drugs in order to relieve pain, founded upon clinical judgment, will not be regarded as causing death so long as it remains within reasonable bounds. Those bounds were transgressed in the case of Dr Cox, who administered a drug in order to stop the patient's suffering by causing her death, not simply to relieve pain.118 The Adams approach was followed in Dr Moor's case (1999),119 where the trial judge again drew a distinction between administering drugs with intent to kill the patient and administering drugs as proper treatment to relieve pain and suffering. What the courts appear to be doing here is to deny that there is causation in the latter instance, in order to avoid the need to confront the question whether a doctor can have a valid defence to an intentional killing.120 The orthodox proposition that shortening life involves causing death is neglected, and the courts apply a version of the doctrine of double effect to argue that the doctor does not cause death if the primary intention is to relieve pain, even though it is well known that this will shorten the patient's life. This is perhaps best characterized as a covert recognition, in causation doctrine, of some form of defence based on clinical medical necessity.121 To summarize, the Cato principle is that it is sufficient if D's conduct was a ‘but for’ cause which was more than minimal: it need not be a substantial cause,122 but it (p. 108) seems that a mere ‘but for’ cause will rarely be sufficient,123 and it might be best to require D's conduct to be a ‘significant cause’.124 The principle has been illustrated here in relation to ‘result-crimes’, but the same approach should be adopted to crimes that penalize conduct or possession, although for those crimes the difficulties will usually concern the exceptions in (b).125 The Draft Criminal Code re-states the general principle in terms of ‘an act which makes more than a negligible contribution to its occurrence’,126 and the Model Penal Code deals with the issue by excluding causes which are too remote to have a just bearing on responsibility.127 The requirement of ‘but for’ causation is sometimes termed ‘factual causation’, which is then contrasted with ‘legal causation’—not only to suggest that the law requires something more than ‘but for’ causation, but also to indicate that there are other aspects of the doctrine to be considered. (b) Interventions between conduct and result A natural event occurring after D's conduct may be treated as terminating D's causal responsibility, but (as suggested earlier) not if it could reasonably be expected.128 The contrast would be between D, whose assault victim catches scarlet fever in hospital and dies (which should be treated as a ‘visitation of Providence’ and as negativing any causal connection between D and the death), and E, who leaves his assault victim lying on a tidal beach, where he later drowns (this is within the risk which was reasonably foreseeable, and therefore not sufficiently unexpected to prevent causal responsibility for the death). What if

Page 19 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions D's act is followed by another human act, which intervenes before the result occurs? We say at the beginning of this section that the chain of causation can be regarded as broken in such cases if the intervention came in the form of a free, deliberate and informed act (the point is considered again shortly). But in at least three sets of situations—(i) the non-voluntary conduct of third parties; (ii) the conduct of doctors; and (iii) the conduct of the victim—this is not so, raising questions about what is the general rule and what the exception. (i) ‘Non-Voluntary’ Conduct of Third Parties: Since the general principle is said to be that the voluntary intervening act of a third party severs or supersedes the causal connection between D's act and the prohibited result, the courts have developed exceptions in cases where the third party's intervention would not be described as voluntary. If the third party is an infant or is mentally disordered, this lack of rational capacity may be sufficient to discount the third party's act in causal terms. The same (p. 109) applies if D sets out to use a responsible adult as an ‘innocent agent’, giving false information to that person in the hope that he or she will act upon it. The behaviour of the person who has been tricked is discounted as non-voluntary for these purposes. The case of Michael (1840)129 illustrates the principle. D's child was in the care of a foster-mother, and D, wishing her child dead, handed a bottle of poison to the foster-mother, saying that it was medicine for the child. The foster-mother saw no need for the medicine and placed it on the mantelpiece, from which her own 5-year-old child later removed it and administered a fatal dose to D's child. The intended result was therefore achieved through the unexpected act of an infant rather than through the mistakenly ‘innocent’ act of an adult, but neither of these intervening acts was regarded as sufficient to relieve D of causal responsibility.130 A similar approach may be taken where the intervening act is one of compulsion, necessity, or duty. If the third party brings about the prohibited harm whilst under duress from D, then D may be regarded as the legal cause of the result.131 The same analysis can be applied where D creates a situation of necessity, or where D's behaviour creates a duty to respond in the third party. Thus in Pagett (1983)132 D was being pursued by the police and took his pregnant girlfriend hostage, holding her in front of him as a shield whilst he fired shots at the police. The police fired back at D, but killed the girlfriend. The Court of Appeal upheld D's conviction for the manslaughter of his girlfriend, even though the fatal shots were fired by the police and not by him. The Court offered two reasons in support of this conclusion: first, the police officer's conduct in shooting back at D was necessary for his self-preservation and therefore was not a voluntary act; and, secondly, that the police officer was acting from a duty to prevent crime and to arrest D. Both these reasons beg important questions: did a necessity exist? Was there a duty? They contain no reference to a duty to avoid harm to the person being held hostage: should not the liberty to act in self-preservation be subject to this qualification?133 These points ought to have been explored at least. Perhaps a better rationale for this decision may be found in a doctrine of ‘alternative danger’: where D places a person in the position of having to choose between two drastic courses of action, one threatening self-danger and the other threatening danger to another, the result should be attributed causally to the creator of the emergency, and not to the unfortunate person who has to choose. This leaves open the possibility of finding that a trained police officer ought to have acted with greater circ*mspection towards the hostage on the facts of Pagett, if that is a fair judgment on the facts of that case, (p. 110) since the law might justifiably expect more of a trained official than of a hapless citizen caught up in extreme events.134

Page 20 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions Whatever one might say about the Court of Appeal's attempts to rationalize the causal responsibility of Pagett for his girlfriend's death, at least they kept some faith with the fundamental principle that a voluntary intervening act breaks the causal chain. This cannot be said of one aberrant decision of high authority, Environment Agency v Empress Car Co (Abertillery) (1999).135 In this case the company had fixed an outlet from its diesel tank which would drain towards a river, governed by a tap that was not locked. An unknown person opened the tap and the river was polluted. The company denied that it caused the polluting matter to enter controlled waters, contrary to the Water Resources Act 1991, and on normal principles one would expect the deliberate act of a third party to negative its causal responsibility. However, the House of Lords held that if the company ‘did something which produced a situation in which the polluting matter could escape but a necessary condition of the actual escape which happened was also the act of a third party or a natural event, [the court] should consider whether that act or event should be regarded as a normal fact of life or something extraordinary’.136 In this way the House of Lords discarded the general principle that a voluntary intervening act breaks the causal chain in favour of the distinction ‘of fact and degree’ between ordinary and extraordinary interventions. The conviction in this case is a clear policy decision, aimed at imposing stringent duties on companies to take steps to prevent pollution, and convicting them for omissions to fulfil those duties. When the House of Lords returned to the subject in Kennedy (No. 2), Lord Bingham held that the Empress Car decision is to be confined to its facts.137 In Kennedy No. 2 (2008)138 D handed V a syringe of heroin with which V then injected himself and died. Overruling the Court of Appeal's strained judgment in favour of a conviction for manslaughter, Lord Bingham recognized the criminal law's approach of treating individuals as autonomous beings, giving rise to the principle that ‘D is not to be treated as causing V to act in a certain way if V makes a voluntary and informed decision to act in that way’. Thus the House of Lords unanimously held that there should be no conviction for manslaughter because D did not cause V to take the heroin: it was self-administered. (ii) Conduct of Doctors: The decision in Pagett contains more than a hint that the court was far more concerned about convicting a morally culpable person than about the refinements of causation, and similar leanings may be found in cases involving (p. 111) doctors. In cases where medical attention is given to a victim, there is rarely any doubt that it may properly be described as ‘voluntary’: doctors work under pressure, occasionally having to make rapid decisions, but they are trained and trusted to exercise clinical judgment in these circ*mstances. Doctors act under a duty to treat patients, but they surely do so voluntarily. However, the courts have drawn a distinction between (a) cases where the injury inflicted by D remains a substantial and operating cause of death despite the subsequent medical treatment, in which case D remains causally responsible even if the medical treatment is negligent; and (b) those where the original wound becomes merely ‘the setting in which another cause operates’, in which case D's responsibility may be negatived by subsequent aberrant medical treatment.139 The reference to an ‘operating and substantial’ cause may be regarded as more favourable to D than the general principle of causation, unless the term ‘substantial’ is read as meaning, simply, ‘more than minimal’. This is confirmed by the statement in Cheshire (1991)140 that a significant contribution is all that is required, and that the defendant's act does not need to be the sole or even the main cause:

Page 21 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions

Even though negligence in the treatment of the victim was the immediate cause of his death, the jury should not regard it as excluding the responsibility of the accused unless the negligent treatment was so independent of his acts, and in itself so potent in causing death, that they regard the contribution made by his acts as insignificant.141

No clear reason is offered for discounting the voluntary intervening act of the doctor. If the doctor administers a drug to which the patient is known to be intolerant, or gives some other wrong treatment, surely the inappropriateness of the medical treatment should affect the causal enquiry. The courts’ reluctance to discuss the causal significance of the medical treatment probably stems from a desire to ensure the conviction of a culpable offender, and this suggests a strong attachment to a ‘wrongful act’ approach to causation, deciding the issue by reference to broader judgments of innocence and culpability. This appears to overlook the fact that D, who inflicted the original wound which gave rise to the need for medical attention, will still be liable for attempted murder or a serious wounding offence even if the medical treatment is held to negative his causal responsibility for the ensuing death. For adherents of the ‘wrongful act’ approach this would be insufficient: they want to see responsibility for the ultimate result pinned on the defendant. However, a court which declares that it is not the doctor who is on trial but the original wrongdoer142 is merely offering an unconvincing rationalization of its failure to apply the ordinary causal principle that a voluntary intervening act which accelerates death should relieve the original wrongdoer of liability for the result. If that causal principle is thought unsuitable for medical (p. 112) cases, should we not be absolutely clear about the reasons, and then look closely at a doctrine of clinical medical necessity?143 (iii) Conduct or Condition of the Victim: The general principle that the law approaches causation by considering the effect of an autonomous individual's conduct upon a ‘stage already set’ is usually taken to extend to cases where the victim has some special condition which makes him or her especially vulnerable. This is sometimes known as the ‘thin skull’ principle, or the principle that defendants must take their victims as they find them. If D commits a minor assault on V, and V, who is a haemophiliac, dies from that assault, the principle applies to render D causally responsible for the death.144 Now this principle of causation may have little practical effect on its own, since most of the serious criminal offences require proof of mens rea (proof that D intended or foresaw the risk of causing, say, serious injury), and it will usually be possible to show that the mens rea was lacking because D was unaware of V's special condition. However, where an offence imposes constructive liability (such as manslaughter in English and American law),145 the ‘thin skull’ principle reinforces the constructive element by ensuring that there is no causal barrier to convicting D of an offence involving more serious harm than was intended or foreseen. The objections to constructive manslaughter are set out in Chapter 7.5. The objection to the ‘thin skull’ principle is that such physical conditions are abnormal and that much of the standard analysis of causation turns on distinctions between normal and abnormal conditions.146 What principles should apply to the causal effect of the victim's conduct after D's original act? Should V's conduct be subject to the normal rules of voluntary intervening acts? Roberts (1972)147 was a case in which D, while driving his car, made suggestions to his passenger, Page 22 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions trying to remove her coat, at which point she opened the door and leapt from the moving car, suffering injury. The Court of Appeal upheld D's conviction for assault occasioning actual bodily harm, on the basis that a victim's ‘reasonably foreseeable’ reaction does not negative causation. Whether ‘reasonable foreseeability’ is an accurate way of expressing the point in question must be doubted; one might well say that the prospect of the woman jumping from the moving car was relatively unlikely. Surely it would be better to consider the principle of ‘alternative danger’: D's conduct had placed V in a situation of emergency in which she had to make a rapid choice about how to react. One might then say that any reaction which cannot be regarded as wholly abnormal or ‘daft’148 should remain D's causal responsibility. In this sense, V's reaction is non-voluntary. (p. 113) What if the victim refuses to accept medical treatment for the injury inflicted by D? The question presented itself starkly in Blaue (1975).149 D stabbed V four times, piercing her lung. V was advised that she would die from the wounds unless she had a blood transfusion, but, adhering to her faith as a Jehovah's Witness, she refused to undergo this treatment. She died. The Court of Appeal held D to be causally responsible for her death. Her intervening decision not to accept the ‘normal’ treatment did not negative D's causal responsibility, because, the Court argued, the situation was analogous to that covered by the ‘thin skull’ rule. Stating that ‘those who use violence on other people must take their victims as they find them’, the Court added that this ‘means the whole man [sic], not just the physical man. It does not lie in the mouth of the assailant to say that his victim's religious beliefs which inhibited him [sic] from accepting certain kinds of treatment were unreasonable.’150 Is this another example of a court stretching the principles of causation so as to ensure the conviction of a wrongdoer? The ‘thin skull’ principle applies only to pre-existing physical conditions of the victim. The principle of individual autonomy suggests that, in general, any subsequent act or omission by V should negative D's causal responsibility. Exceptions to this are where V's subsequent conduct falls within the ‘reasonable foreseeability’ notion in Roberts 151 or, perhaps, within the principle of ‘alternative danger’. D's actions in Blaue can certainly be said to have caused a situation of alternative danger and emergency, and so then the question would be whether V's reaction should be classified as wholly abnormal. In a statistical sense it surely was: it must be rare to refuse a blood transfusion knowing that death will follow that refusal. To accept this would be to make no distinction between one who refuses treatment for religious reasons and one who refuses out of spite. It could be argued that the standard of normality should be informed by social values rather than enslaved to statistical frequency, that religious beliefs are a matter of conscience which should be respected, and therefore that acts or omissions based on religious conviction should not be set aside as abnormal. Bolstering this argument is the view that matters of conscience, taken seriously, leave someone to a significant degree unfree to do other than follow their conscience. That being so, it could be said that the victim, whilst an autonomous person, was not acting freely when refusing to have a blood transfusion, preserving the link between the offender's wrongful act and her death. No doubt there was also much sympathy and respect for the victim, courageously adhering to her religious beliefs in the face of death, generating the argument that it would not be appropriate to hold her causally responsible for her own death. There can also be examples in which V has aggravated his or her condition by failure to attend to injuries or wounds, or even by deliberately re-opening them. The judicial approach is to hold that D can still be convicted if his conduct made an operative and substantial contribution to the result, even if V's own act or omission also contributed.152 Once again, the

Page 23 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions more rigorous approach of recognizing that V's own act broke the causal chain, and that D should therefore be convicted of an attempt or other offence, has been found unattractive by the courts. (p. 114) (c) Causation and omissions One of the difficulties sometimes raised about imposing criminal liability for omissions, in addition to those already discussed in section 4.4, is the problem of causation. How can an omission be said to cause harm? Or are these cases exceptions to the causal requirement?153 Starting with the most basic question, is it possible to say that, but for an omission, a harm would not have resulted? The existence of a duty justifies calling it an omission, and the nonperformance of that duty in a situation where it arises can be said to cause the result. To take an extreme example, a parent who makes no attempt to save her or his child from drowning in shallow water can be said to cause the child's death: but for the parent's inaction, the child would almost certainly have lived. It is no answer to say that the child would have drowned anyway if the parent had not been there, because in that eventuality there would have been no duty and hence no omission. On the facts as they were, the parent was present, and but for non-performance of the duty the child would not have died. When dealing with causation by acts, we have seen that the courts have used terms such as ‘significant’ and even ‘substantial’ in some cases, chiefly to rule out remote or minimal causes, but this should create no special difficulty for omissions. One counter-argument is that this approach may sometimes lead to the conclusion that many people caused a result: if, in a jurisdiction which imposes a duty of easy rescue, twenty or more people stand by without offering any help or raising the alarm, the conclusion must be that all these people caused the harm that occurred. This is true, and is hardly an argument against the causation approach. This is not to suggest that the application of causal arguments to cases of omission is without difficulty. For example, if A stabs V it is obvious that but for A's act V would not have suffered this wound; but if a parent makes no effort to save a child drowning in a pool, it is possible that the duty might have been fulfilled by summoning help (which might have caused delay, and the child's life might have been lost), or that the parent might not have been able to save the child's life anyway (if it had already been in the pool some time before the parent arrived). The point of these examples is that the ‘but for’ clause may be less concrete in some omissions cases, and may occasionally require a judgment to be made. However, at the very least there are many clear cases where ordinary causal analysis creates no more problems than it does in relation to acts. (d) Causing other persons to act Can it ever be held that one person caused another to act in a certain way? The notion would seem to be inconsistent with the general principle of individual autonomy, emphasized above by reiterating the principle that a voluntary intervening act removes or displaces the previous actor's causal responsibility. Yet we have already noted one (p. 115) case in which a person can be said to cause another to act—the case of innocent agency, where the third party lacks rationality or has been tricked. Further cases arise in the law of complicity, that branch of the criminal law which holds people liable for helping or encouraging others to commit crimes, which will be discussed at length in Chapter 10.

Page 24 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions One example of the type of case under discussion is where D goes to P and offers P money to injure or kill V: 154 the law will hold D liable for counselling and procuring P's subsequent offence, and one might say that D causes the offence, in some sense. Clearly, however, D did not cause P to act as an innocent agent: P was not, we assume, lacking in rational capacity, and so on the general principle of individual autonomy P would be regarded as causally responsible for the result. D cannot, therefore, be held to have caused that result in the usual sense, but one might follow Hart and Honoré in suggesting that D may be said to have given P a reason for committing it.155 This is a dilution of the general approach to causation, aimed specifically at rationalizing the criminal liability of certain accomplices. But it is not only those who ‘counsel or procure’ who are brought within the English law of accomplice liability. It is also persons who ‘aid and abet’ others to commit offences. Advice, information, and other acts of assistance and encouragement may be great or small, and may be readily obtainable from others if this would-be accomplice had declined. So, as an element of causal contribution to P's offence, D's ‘aiding’ may be insignificant indeed—certainly well below the ‘but for’ threshold, even in the extended sense adopted by the notion of ‘occasioning’. Many writers now acknowledge that the element of causation is absent from some cases of ‘aiding and abetting’.156 This brings us to a reconsideration of the role of causation. (e) Conclusion Causation is a complex topic, with which we have been able to deal only briefly here. Proof of causation is often said to be an essential precondition of criminal liability, but there is reason to doubt the generality of that requirement, notably in respect of accomplice liability (just discussed) and vicarious criminal liability.157 Rather than insisting on a universal requirement of causation, it may be preferable to argue that liability should be negatived, in general, by the voluntary intervening act of another. Several criticisms of the judicial approach to three exceptional categories of case have been advanced. Often the explanations given by the courts are unconvincing. Whilst traditional or standard causal theory emphasizes the significance of the last voluntary (p. 116) act, there is no reluctance to look wider or to massage the term ‘voluntary’ in certain situations, especially where D clearly started the sequence of events by doing a wrongful act. The challenge is to re-examine the intuitions that lead judges and others to their conclusions (e.g. the wrongful act theory, the approach to medical mistakes, etc.), with a view to constructing a law that ensures that the courts respect the various principles outlined in Chapter 3.

4.6 Self-defence and permissible force Many offences include a qualification such as ‘without lawful excuse’, ‘without lawful authority or reasonable excuse’, and so on. We are not concerned here with the different shades of meaning attached to such phrases,158 nor with the legislature's frequent use of the word ‘excuse’ to refer to permissions, but rather with some general doctrines which grant permissions to engage in conduct which would otherwise be criminal. Self-defence is the best known of these permissions, but there are others concerned with the prevention of crime, the arrest of suspected offenders, the protection of property, and so forth. Lawyers frequently speak of these doctrines as defences, e.g. ‘the defence of self-defence’, Page 25 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions and procedurally that is how they function. If there is evidence, usually raised by the defendant, that the conduct may have been permissible, the prosecution bears the burden of proving beyond reasonable doubt that the conduct was not permissible or otherwise lawful. ‘If the prosecution fail to do so, the accused is entitled to be acquitted because the prosecution will have failed to prove an essential element of the crime, namely that the violence used by the accused was unlawful.’159 The consequences of presenting the permissions as the element of unlawfulness required in all crimes will not be taken further here.160 Neither this, nor the procedural device of treating them as defences, should deflect attention from the fundamental significance of permissions. There are certain situations when individuals have a right or permission to do things which would generally be prohibited because they cause harm or damage. The most extreme occasions are those on which the law permits one person to kill another. It is sometimes said that defences such as self-defence involve ‘justifications’ for conduct. However, Suzanne Uniacke argues that justified conduct is conduct that one has a right to do, and defences such as self-defence cover situations in which one may—one is entitled to—choose whether to act or not (it might be morally wrong to make the choice to do the act that is covered by the defence). Such defences are thus best described as ‘permissions’.161An important point arising from this is that the rules (p. 117) governing permissions should ideally respect the various principles of legality and the rule-of-law for the same reason that offence definitions should, that is, because they may be relied upon to guide behaviour.162 (a) Self-defence and individual autonomy163 It is hardly surprising that decisions on self-defence formed an important and frequent element in the development of the English common law in days when there was no organized policing and when the carrying of deadly weapons was common. The issues here concern the basic right to life and physical safety. An individual who is either attacked or threatened with a serious physical attack must be accorded the legal liberty to repel that attack, thus preserving a basic right. A well-regulated society will provide a general protection, but it cannot guarantee protection at the very moment when an individual is subjected to sudden attack. The criminal law cannot respect the autonomy of the individual if it does not make provision for this dire situation. (b) The problem of conflicting rights In terms of individual autonomy, one difficulty with this position is that these situations involve two individuals (at least). If the law gives the person attacked the liberty to wound or kill the aggressor, what happens to the aggressor's right to life and physical safety? The answer to this question must have as its starting point the European Convention on Human Rights, Art. 2 of which declares the right to life in these terms: 1. Everyone's right to life shall be protected by law. No one shall be deprived of his life intentionally save in the execution of a sentence of a court following his conviction of a crime for which this penalty is provided by law.164 2. Deprivation of life shall not be regarded as inflicted in contravention of this Article when it results from the use of force which is no more than absolutely necessary: a. in defence of any person from unlawful violence; b. in order to effect a lawful arrest or to prevent the escape of a person lawfully

Page 26 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions detained; c. in action lawfully taken for the purpose of quelling a riot or insurrection. (p. 118) Articles 3 and 5 of the Convention protect a citizen's freedom from inhuman treatment and security of person, but, unlike Art. 2, they contain no explicit exceptions in favour of the permissible use of force, and the Court has had to imply such exceptions.165 As for the exceptions to Art. 2, two of them appear rather strange. To suggest that causing death may be absolutely necessary ‘to effect an arrest’ (Art. 2.2b) is somewhat absurd since, as Sir John Smith has pointed out, one cannot arrest a dead person.166 A killing that is permissible to prevent a riot or insurrection (Art. 2.2c) is barely conceivable. However, Art. 2 has no exception for killings in the prevention of any other non-violent crime. Thus, for example, a householder who kills a burglar ought to have no defence, if Art. 2 is applied, unless the circ*mstances can be said to have involved the defence of a person from unlawful violence. The acquittal of the householder, in a case where physical violence had not been offered by the burglar, might suggest that English law does not respect the right to life in Art. 2.167 The approaches of other legal systems differ considerably. Some maintain that an innocent person's rights are absolute and thus recognize few limitations on those rights, even when that person is repelling a minor assault or defending property.168 This suggests that the aggressor forfeits the normal rights when he embarks on an attack, and that it is his misconduct in starting the conflict which justifies the law in giving preference to the liberty of his victim. The idea of forfeiture is not objectionable in itself,169 but it should be carefully circ*mscribed lest it allows the person attacked to stand fast and use whatever force is necessary to protect his rights of ownership and liberties of passage. The forfeiture approach bears some similarity to the ‘wrongful act’ analysis in causation170 and to the theory of constructive liability,171 in that it attributes great significance to the wrongfulness of a person's initial act. However, the focus should be on the right to life, as the jurisprudence of the European Convention establishes.172 Initial wrongfulness should only be taken to permit the proportionate use of force: the innocent subject of an attack should not be free to use whatever force is necessary to vindicate his threatened rights. Such an analysis would assign no value to the rights of the attacker. If the criminal law is committed to ensuring that everyone's life is protected and that force is inflicted as rarely as possible, it cannot accept a vindicatory approach which would allow the infliction of gratuitous, or at least disproportionate, harm. Forfeiture of life to protect a person from some minor hurt, loss, or damage would promote the value of honour above respect for life and limb. The tendency of the English courts to reach for the concept of reasonableness, without setting out the relevant rights first, is an unfortunate aspect of legal culture. (p. 119) (c) The rules and the principles Self-defence is a long-standing defence in English law,173 but it must be considered in the light of two statutory provisions. Section 3 of the Criminal Law Act 1967 states that ‘a person may use such force as is reasonable in the circ*mstances in the prevention of crime …’. The section was not intended to supplant the common law rules on self-defence,174 and the courts have continued to develop those rules. It is true that in most situations of self-defence it could be said that the person was preventing crime (i.e. preventing an attack which constituted a crime), but that would still leave certain cases untouched—notably, attacks by a child under 10, by a mentally disordered person, or by a person labouring under a mistake of fact. Such aggressors would commit no offence, and so it is the law of self-defence, not the prevention of 175 Page 27 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions crime, which governs.175 Section 3 of the Criminal Law Act 1967 has now been buttressed by s. 76 of the Criminal Justice and Immigration Act 2008, which ‘is intended to clarify the operation of the existing defences’ (s. 76(9)), notably self-defence. It is rare for legislation to state on its face that it is for clarification: this curious notion must mean that the common law defence is not abolished,176 but that the new provisions supersede the common law to the extent that they apply. However, as will appear from the following paragraphs, s. 76 deals with only a few of the many issues of principle arising in the law of self-defence. Section 76(2) states that the section applies to the common law on self-defence and to s. 3 of the Criminal Law Act, which deals with force used in the prevention of crime or in effecting a lawful arrest. When the Law Commission considered the issue some years ago, it identified other possible bases for permitting the use of force (such as the prevention or termination of trespass on property), and these must not be forgotten.177 The principles should be the same as for the other permissions, as they are flexible enough to adapt to a wide variety of circ*mstances. However, there has been constant political pressure in recent years to create specially favourable rules for householders seeking to keep out or eject trespassers from their homes, and for those defending their property more generally from attack. One result of that has been the introduction of an ‘avoidance of doubt’ provision dealing with the question of whether those in—or in possession of—property can stand their ground, refuse to abandon the property, and confront a (would-be) trespasser in defence of the property. Section 148 of the Legal Aid, Sentencing and Punishment Act 2012 now amends the law to make it clear that there is no duty to retreat in the face of trespass to property (presumably, property under one's legitimate care, control, or ownership). Whether or not retreat (p. 120) or abandonment should have been the right reponse is now just a factor for the trier of fact to consider as part of the overall judgment of whether someone's response was within the bounds of reasonableness. Ironically, where the force used is not physical but consists of damage to another's property, the legal principles have long been generous to D. The permissible damaging of another's property requires only that D believed that ‘the means of protection adopted … would be reasonable having regard to all the circ*mstances’.178 This hardly embodies a legal standard at all, since it turns on D's beliefs as to what is reasonable. One feature of the Draft Criminal Code was that it would abolish the different rule for property damage. (d) The proportionality standard The law of self-defence has two elements: necessity and proportionality. The requirement that the use of force must be necessary (or, where the right to life is involved, ‘absolutely necessary’) is combined with a further requirement that the amount of force must be proportionate to the value being upheld. This shows respect for the rights of the attacker in self-defence cases, and for the rights of suspected offenders in relation to the other permissions. Even though the necessity part has subjective elements, as we shall see, the reasonableness of the force used depends not on D's beliefs but on an objective assessment.179 Thus where D misjudges the amount of force which is reasonable, e.g. to insist on passing along a path barred by another, to eject a trespasser, or to detain a poacher, this is a mistake of law rather than of fact. The Court of Appeal has confirmed that D's view of the

Page 28 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions amount of force that was reasonable is not determinative: the magistrates or jury should assess whether, in the circ*mstances existing at the time, the amount of force was reasonable.180 The standard cannot be a precise one: s. 76 of the 2008 Act states that the force must not have been ‘disproportionate’, i.e. not out of proportion to the amount of harm likely to be suffered by the defendant, or likely to result if a forcible intervention is not made. What is crucial is that it should rule out the infliction or risk of considerable physical harm merely to apprehend a fleeing non-violent offender,181 to stop minor property loss or damage, etc. As a nineteenth century Royal Commission remarked, a law whose only requirement was necessity ‘would justify every weak lad whose hair was about to be pulled by a stronger one, in shooting the bully if he could not otherwise prevent the assault’.182 On this (p. 121) view, the proper approach is to compare the relative value of the rights involved, and not to give special weight to the rights of (say) a property owner simply because the other party is in the wrong (i.e. committing a crime).183 Thus in Rashford (2006)184 the Court of Appeal rightly held that self-defence should not be ruled out simply because D was the initial aggressor. If V's response to D's aggression was out of all proportion, D would be justified in using sufficient force to protect himself. If, however, D had intended to provoke V into attacking him, in order to then use fatal force on V, it is well established that self-defence would be unavailable.185 Although Art. 2 of the Convention does not specify a proportionality requirement, the Strasbourg Court has emphasized that the use of deadly force must be both absolutely necessary and strictly proportionate if it is to come within an exception to the right to life.186 The American Model Penal Code provides that deadly force is not permitted, ‘unless the actor believes that such force is necessary to protect himself against death, serious bodily harm, kidnapping or sexual intercourse compelled by force or threat’.187 It is debatable whether this goes too far in allowing the lawful sacrifice of a life to prevent certain non-fatal assaults,188 and it should be noted that Art. 2 of the Convention is vague on this question. Deadly force may be permitted, ‘in defence of any person from unlawful violence’, but how serious a violent attack? The Strasbourg jurisprudence is no more precise than English law on this point. The Model Penal Code formulation might be a worthwhile starting point for analysis and argument, although it is arguable that the breadth and uncertainty of ‘kidnapping’ makes its inclusion in the list of threats that might warrant deadly force to resist them a controversial one.189 Should the judgment of proportionality be affected by the fact that the force was used against a law enforcement officer? Since English law renders an arrest lawful if the police officer has reasonable grounds for suspicion (even if the grounds turn out to be erroneous), this may be of importance. There are English decisions which draw a distinction between resisting a lawful —but mistaken—arrest (which is not permissible), and repelling the unlawful use of violence by police (which is permissible).190 This principle is to be found both in the Model Penal Code and the Draft Criminal Code.191 It permits individuals to defend themselves against excessive force by the police, whilst requiring them not to use force against police who are effecting an arrest (p. 122) for which the officer may believe there are reasonable grounds (even if the arrestee believes otherwise). (e) Aspects of the necessity requirement The necessity requirement forms part of most legal regimes on permissible force. The first question to be asked is: necessary for what? We have seen that force may be permitted for any one of several lawful purposes. The necessity must be judged according to the lawful

Page 29 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions purpose which the defendant was trying to pursue: for self-defence, purely reactive defensive force will often be all that is necessary; in order to apprehend a suspected offender, on the other hand, a police officer or citizen will need to behave proactively. These differences may become particularly important in cases where there is a suspicion or allegation that the force was used by way of revenge or retaliation rather than in pursuit of a lawful purpose. What was the defendant's purpose? Could the conduct be said to be necessary for that purpose? Does this reference to ‘purpose’ mean that there is a mental element in the permissions, such that a person cannot rely on a particular permission if he or she is ignorant of the basic facts needed to support that permission? In Dadson (1850)192 a constable shot a fleeing thief. Such force was permissible only against ‘felons’, and a thief was a felon if he had two previous convictions. This thief had previous convictions and so was a felon, but the constable fired at him without knowing of these convictions. It was held that the constable could not rely on the permission to use force to apprehend a felon because he was unaware of the basic fact needed to constitute the permission. The Northern Irish case of Thain (1985)193 takes this point further. D, a soldier on duty, stated from the outset that he did not fire the shot in order to apprehend V (who was running away at the time). He said that he shot in reaction to a sudden movement by V. It seems that D might have succeeded if he had maintained that his intention was to arrest, but he proffered another reason and was convicted of murder. This decision holds that D's beliefs or motives have a significant bearing on the permissibility of the use of force, and this is surely right. In many circ*mstances a greater use of force might be justifiable for law enforcement than merely for defence.194 In most cases, where no problem of the mental element arises, the main issue is necessity. The English courts have continued to develop the common law, but without always relating the issues to any general themes and without explicit reference to the (p. 123) primacy of the right to life. An attempt is made here to organize the decisions around six aspects of necessity, referring to s. 76 of the 2008 Act where relevant. (i) Imminence: Although s. 76 is silent on the matter, there is authority that the use of force can be necessary only if the attack is imminent.195 If there is time to warn the police, then that is the course which should be taken, in preference to the use of force by a private individual.196 But this apparently does not mean that it is unlawful to prepare or keep armaments for an anticipated attack. In the Attorney-General's Reference (No. 2 of 1983)197 D's shop had been looted during rioting which the police had struggled to control; D made some petrol bombs with which to repel any future attack, and the question was whether these were in his possession ‘for a lawful object’. It was held that they were, if the jury accepted that D intended to use them only against an attack on his premises which the police could not control. This is an unusually indulgent approach for the criminal courts—a conviction followed by a discharge would be more normal, since it does not signal that such conduct is permissible—but it was a response to a particular type of situation. If the police are unable to offer protection and attack is imminent, the rationale for justifiable force is made out—although objects so lethal as firebombs should rarely be approved as lawful means of defending business premises, as opposed to defending a home or human beings. This decision leaves a number of questions about the ‘imminence’ requirement unresolved. Where a woman who has been habitually subjected to physical abuse by her male partner has Page 30 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions a reasonable fear that he may kill her next time, does this satisfy the ‘imminence’ requirement if she then kills him whilst he is asleep? True it may be that ‘a “reasonable person” does not fear immediate death from a sleeping person’,198 but that reference to immediacy is surely too strict, and it may be argued that the real issue is whether the woman reasonably fears a danger to her life that she will be unable to avoid.199 Another problem concerns the lawfulness of carrying a gun or an offensive weapon in order to repel an anticipated attack: the authorities would seem to suggest that, although the use of the weapon might be lawful if an attack takes place, its possession beforehand remains an offence.200 A further unresolved question arises where a law enforcement officer shoots a fleeing suspect on the basis that the suspect is likely, (p. 124) if allowed to escape, to commit violent offences: must it be shown that those offences might or would be committed sooner rather than later?201 (ii) A Duty to Avoid Conflict?: One of the most technical but most significant elements in the common law of self-defence was the duty to retreat. Its technicality lay in its careful wording and its exceptions; its significance was that, from an early stage, the common law recognized limitations on the forfeiture principle and on the primacy of the non-aggressor's autonomy in these situations. However, the duty has now disappeared as such. In Julien (1969)202 it was rephrased as a duty to demonstrate an unwillingness to fight, ‘to temporize and disengage and perhaps to make some physical withdrawal’. In Bird (1985)203 the Court of Appeal accepted that the imposition of a ‘duty’ is too strong. The key question is whether D was acting in self-defence, or in revenge or retaliation. Evidence that D tried to retreat or to call off the fight might negative a suggestion of revenge, but it is not the only way of doing so. The modification of the law seems to derive from the suggestion in Smith and Hogan's textbook that the ‘duty’ as described in Julien is inconsistent with the liberty to make a pre-emptive strike.204 It is not. The liberty to make a pre-emptive strike can easily be cast as an exception to the general duty to avoid conflict, and, as such, it is no more inconsistent with the rule than any other exception to a rule. The difficulty with regarding the duty to avoid conflict as merely one consideration to be borne in mind here is that it says nothing about the circ*mstances which might outweigh it. If the law is to protect everyone's right to life and to pursue the minimization of physical violence, the avoidance of conflict—or what Fiona Leverick refers to as the ‘strong retreat rule’205—must be right in principle. Section 76 is silent on this. (iii) Freedom of Movement: English law also recognizes an exception to the duty to avoid conflict (if such a duty exists) in those cases where D is acting lawfully in remaining at, or going to, a place, realizing that there is a risk that someone will force a violent confrontation there. The authority for this is Field (1972),206 where D was warned that some men were coming to attack him. D stayed where he was, the men came and made their attack, and in the ensuing struggle D stabbed one of them fatally. The Court of Appeal quashed his conviction, holding that he had no duty to avoid conflict until his attackers were present and had started to threaten him. The American case of State v Bristol (1938)207 takes the point further, holding that D had no duty to avoid entering a bar where he knew his adversary (who had threatened him with attack) to be drinking. The American court declined to lay (p. 125) down a rule which might ‘encourage bullies to stalk about the land and terrorize citizens by their mere threats’. These two decisions appear to promote the value of freedom of movement above any duty to avoid conflict in advance by, 208 Page 31 of 54

209

Criminal Conduct: Actus Reus, Causation, and Permissions for example, informing the police of the threat.208 However, in Redmond-Bate v DPP209 the Divisional Court held that the defendant's right to preach should be protected, as an exercise of the right to freedom of expression under Art. 10 of the Convention, and it was only if the words spoken were likely to provoke violence in others that it would have been proper to arrest her: if her words were not provocative of violence, only those who used or threatened violence should have been arrested. This suggests a small qualification of a subject's right to freedom of expression, and there are surely strong arguments for this. Should not the minimization of physical violations (implicit in Art. 5) take precedence over freedom of expression (Art. 10) and movement? Is there not some analogy with omissions to assist in saving life, where a citizen's general liberty should also be outweighed by a specific social duty?210 These remarks concern self-defence and the defence of property only; clearly, a person who acts with the purpose of preventing crime or arresting a suspected offender cannot be expected to avoid conflict, and so the proportionality standard ought to assume primacy there. (iv) Pre-Emptive Strike: The use of force in self-defence may be justifiable as a pre-emptive strike, when an unlawful attack is imminent.211 This is a desirable rule, since the rationale for self-defence involves the protection of an innocent citizen's vital interests (life, physical security), and it would be a nonsense if the citizen were obliged to wait until the first blow was struck. The liberty to make a pre-emptive strike is not inconsistent with a duty to avoid conflict (if it were recognized), but it should be read as being subject to that duty. In other words, it would be possible and desirable to have a law which imposed a general obligation to avoid conflict but, where this was not practical, authorized a pre-emptive strike.212 A law which allows pre-emptive strikes without any general duty to avoid conflict runs the risk, as Dicey put it, of encouraging self-assertion through violence.213 (v) Necessity, Proportionality, and Law Enforcement: The point has already been made that a police officer or citizen whose purpose is to prevent a crime or to apprehend a suspected offender must behave proactively. The primary legal restriction on such conduct has been the standard of proportionality, in relation to the purpose that the actor was aiming to achieve.214 How serious an offence was being or had been (p. 126) committed? Is there a danger of serious offences in the near future? Applying Art. 2 of the Convention, not only must the permission fall within para. 2(a), (b), or (c) of the Article, but the force must be shown to have been ‘absolutely necessary’ and ‘strictly proportionate’—the adverbs emphasizing the sharper formulation of the tests under the Convention. This should be the benchmark for scrutinizing the so-called ‘shoot-to-protect’ policy adopted by the Association of Chief Police Officers in 2003 and defended after the London bombings. Although its details have not been made public, the Metropolitan Police Commissioner referred to shooting to kill ‘a deadly and determined bomber who is intent on murdering many other people’.215 Much then depends on whether reasonable grounds should be required for the belief that V is such a person. Some of the leading European decisions are not uncontroversial. In McCann and others v UK (1996)216 the European Court of Human Rights held (by a ten to nine majority) that the UK had violated the right to life of three suspected IRA terrorists who were shot dead by security

Page 32 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions forces in Gibraltar. The most important ruling was that Art. 2 requires law enforcement operations to be organized so as to ‘minimize, to the greatest extent possible, recourse to lethal force’. The Court found that the planning of the operation failed to show the required level of respect for the suspects’ right to life. It did not find that the soldiers who fired the shots violated Art. 2, although it did state that their reactions lacked ‘the degree of caution in the use of firearms to be expected from law enforcement personnel in a democratic society, even when dealing with dangerous terrorist suspects’.217 In Andronicou and Constantinou v Cyprus (1998)218 the Court (by a five to four majority) held that Art. 2 was not violated when Cypriot security forces stormed a house where a hostage was being held, firing machine guns in all directions and killing both the gunman and the hostage. This decision appeared to leave a considerable gap between the strict formulation of the tests and their application to the facts, but the Court distinguished it in Gül v Turkey (2002).219 The Court noted that in the Cyprus case the hostage-taker was known to be in possession of a gun, which he had fired twice already. In the Turkey case, there was insufficient reason to believe that Gül had a gun, and ‘the firing of at least 50–55 shots at the door was not justified by any reasonable belief of the officers that their lives were at risk’.220 It is fair to say that the Strasbourg judgments, (p. 127) particularly when applied to the facts of the cases, leave some scope for debate about what Art. 2 actually requires.221 (f) Mistaken belief as to necessity In English law the rule has become established that a person who purports to use justifiable force should be judged on the facts as he or she believed them to be.222 Section 76(3) of the 2008 Act confirms that this subjective test represents English law, and s. 76(4) goes on to state that the reasonableness of the belief may be considered when assessing whether it was genuinely held.223 However, in cases of killing under Art. 2, the Strasbourg Court has insisted on several occasions that the actions of those who take life should be judged on the basis of the facts that ‘they honestly believed, for good reason, to exist’.224 This is clearly an objective test which, in effect, places such a high value on the right to life as to require law enforcement officers to have adequate factual foundations for their beliefs before using lethal force in consequence. It is easy to argue that this may not always be possible; but a more telling point is that, where it is possible, it ought to be done so as to respect the right to life. For this reason, previous editions of this work have contended that it should be a principle of English law; the arrival of the Convention jurisprudence strengthens that case. The fact that the Strasbourg cases on Art. 2 deal only with law enforcement officers should not be crucial, since it is the State's duty to ensure that the law protects the lives of all victims, no matter who threatens them.225 However, in Bubbins v UK (2005)226 the Strasbourg Court appeared to modify its position, reiterating the requirement of an ‘honest belief, for good reason’, but then softening it considerably by emphasizing the actual belief of the police officer at the time he shot V.227 It is true to say that the Strasbourg Court has had ample opportunity to point out any incompatibility between the English law of self-defence and the Convention, and has not done so.228 Yet, as the Joint Committee on Human Rights has pointed out, adopting the same argument as this book, the preponderance (p. 128) of Strasbourg jurisprudence favours the objective test of reasonable belief, and ‘the very minimum required by human rights law’ in order to protect the right to life of ordinary citizens is that the test of belief ‘for good reason’ should be introduced ‘when force is used by state agents.’229 (g) Permissible force and the emotions

Page 33 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions The foregoing paragraphs have examined principles which might produce outcomes that consistently uphold human rights in those varied situations in which a claim of permissible force might arise. Some might regard those principles as too mechanical for the sudden and confused circ*mstances of many such cases. It is well known that a sudden threat to one's physical safety may lead to strong emotions of fear and panic, producing physiological changes which take the individual out of his or her ‘normal self’. According to this view, the most just law is the simplest: was the use of force an innocent and instinctive reaction, or was it the product of revenge or some manifest fault? This simple approach may have the great advantage of recognizing explicitly the role of the emotions in these cases. It is surely right to exclude revenge attacks from the ambit of justifiable force.230 It is also consistent with the doctrine of prior fault for the law to construe the standards of reasonableness and necessity strictly against someone whose own fault originally caused the show of violence.231 The question then is how much indulgence should be granted to the innocent victim of sudden attack who reacts instinctively with strong force. In the leading case of Palmer (1971)232 Lord Morris stated that it is ‘most potent evidence’ of reasonableness that the defendant only did what he or she ‘honestly and instinctively thought necessary’. The Strasbourg Court, despite its insistence on the requirement of ‘good reason’, deferred in Bubbins to the beliefs of ‘an officer who was required to act in the heat of the moment to avert an honestly perceived danger to his life’.233 Section 76(7) of the 2008 Act now gives legislative authority to this approach, by providing that a court should take account, when assessing reasonableness, of the considerations: • that a person acting for a legitimate purpose may not be able to weigh to a nicety the exact measure of any necessary action; and • that evidence of a person's having done only what the person honestly and instinctively thought was necessary for a legitimate purpose constitutes strong evidence that only reasonable action was taken by that person for that purpose. (p. 129) The additional flexibility of this approach suggests that it is more accurate to state the law's requirement in terms of a ‘not disproportionate’ use of force rather than a proportionate response, but even then there must be limits. It cannot be right for absolutely any reaction ‘in a moment of unexpected anguish’ to be held to be justifiable,234 particularly in the case of a trained firearms officer, even if it is right for the courts to consider ‘how the circ*mstances in which the accused had time to make his decision whether or not to use force and the shortness of the time available to him for reflection, might affect the judgment of a reasonable man’.235 To the extent that the law has moved away from objective standards towards indulgence to the emotions of innocent citizens, the rationale of permissions becomes diluted by elements of excuse.236 The fairness of this concession to what Blackstone termed ‘the passions of the human mind’237 is often supported by reference to the famous dictum of Holmes J, namely, that ‘detached reflection cannot be demanded in the presence of an uplifted knife’.238 This dictum is significant for its limited application: it concerns cases of an ‘uplifted knife’, i.e. typically, sudden and grave threats or attacks; it has no application to cases where the attack is known to be imminent and the defendant has time to consider his position. Nor should it necessarily be conclusive in relation to those who are trained to deal with extreme situations, such as the police and the army.239 As the element of sudden and unrehearsed emergency recedes, the

Page 34 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions social interest in the minimal use of force becomes a firmer precept again. In this type of situation, the law ought to give consideration to the relative importance of the sanctity of life and the physical safety of all persons, including offenders, when compared with such other interests as the free movement of citizens. The aphorism about the ‘uplifted knife’ should not be used to prevent the principled resolution of cases to which it does not apply. (h) Conclusions The law relating to self-defence and permissible force depends on resolution of a clash between two aspects of the right to life—the individual's autonomy and right to protect life by using even fatal force if necessary, and the right to life of every citizen (including offenders). English lawyers have generally been reluctant to discuss the issues in these terms, and the government has rarely acknowledged its positive obligation under Art. 2 of the Convention to have in place laws that give maximum protection to the right (p. 130) of life of all citizens. The relevant law—whether on self-defence or the other forms of permission—is mostly common law, and the enactment of s. 76 of the Criminal Justice and Immigration Act 2008 to ‘clarify’ the common law is a disappointment, a missed opportunity to legislate at the detailed level rightly recommended by the Law Commission240 and to engage with Art. 2 of the Convention and its requirements. An urgent re-assessment of the law on premissible force is called for, taking full account of the issues discussed above.

4.7 Permissions, necessity, and the choice of evils The discussion so far has focused on self-defence and the permissions relating to law enforcement and the prevention of crime. Generally speaking, the permissions relating to selfdefence may be linked directly to the principle of autonomy, in the basic sense of selfpreservation, whereas the permissions relating to law enforcement may be linked to the principle of welfare, although that principle should also be interpreted so as to insist on the minimal use of force. In some situations, however, the principle of individual autonomy is compromised because it may not be possible to protect the autonomy of all persons involved. These are the ‘choice of evils’ cases, which must now be discussed. (a) Necessity as a permission English law contains limited defences of duress and necessity, which apply when a person commits an otherwise criminal act under threat or fear of death or serious harm. The relevant law is examined in a later chapter,241 where it will become apparent that many statements about the ambit of the defences (especially in the courts) are ambivalent or even indiscriminate as to whether their basis lies in permissibility (D had a right to use this force) or excuse (the use of force was impermissible, but D did not behave unreasonably in the dire circ*mstances). One apparently clear statement came when the House of Lords, in rejecting duress as a defence to murder, held in Howe (1987)242 that, even if D's own life is threatened, it cannot be permissible to take another innocent life. What this means is that one innocent person who stands in danger of imminent death cannot be permitted to kill another innocent person. To kill an aggressor in self-defence is one thing, but to kill an uninvolved third party, even if this were the only means of preserving one's own life, could not be right—even though (p. 131) it might be excusable, as we shall see elsewhere.243 But what about the possibility of permitting the killing of an innocent non-aggressor when this will save two or more other Page 35 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions lives? One example of this emerged from the inquest into the deaths caused by the sinking of the ferry Herald of Free Enterprise in 1987.244 At one stage of the disaster several passengers were trying to gain access to the ship's deck by ascending a rope-ladder. On that ladder there was a young man, petrified, unable to move up or down, so that nobody else could pass. People were shouting at him, but he did not move. Eventually it was suggested that he should be pushed off the ladder, and this was done. He fell into the water and was never seen again, but several other passengers escaped up the ladder to safety. No English court has had to consider this kind of situation: 245 are there circ*mstances in which the strong social interest in preserving the greater number of lives might be held to override an individual's right to life? Any residual permission of this kind must be carefully circ*mscribed. It involves the sanctity of life, and therefore the highest value with which the criminal law is concerned. Although there is a provision in the Model Penal Code allowing for a defence of ‘lesser evil’,246 it fails to restrict the application of the defence to cases of imminent threat, opening up the danger of citizens trying to justify all manner of conduct by reference to overall good effects.247 The moral issues are acute: ‘not just anything is permissible on the ground that it would yield a net saving of lives’.248 Yet there may be situations in which the sacrifice of a small number of lives may be the only way of saving a much greater number of lives, as where a dam is about to burst (flooding a whole town) unless a sluice-gate is opened (flooding a less densely populated area). Could a doctrine of necessity permit the intentional killing of people in the latter area in order to save the greater number, if there were no alternative? There are strong arguments in favour of recognizing some such extreme situations as involving a permission to kill, but there are those who would oppose this and would insist that there can never be a permission for intentionally taking life—although it may be acceptable to recognize a (partial) excuse in such cases. Some situations give rise to the further moral problem of ‘choosing one's victim’, which arises when, for example, a lifeboat is in danger of sinking, necessitating the throwing overboard of some passengers,249 or when two people have to kill and eat (p. 132) another if any of the three is to survive.250 To countenance a permission in such cases would be to regard the victim's rights as less worthy than the rights of those protected by the action taken, which represents a clear violation of the principle of individual autonomy. Yet it is surely necessary to make some sacrifice if the autonomy of everyone simply cannot be protected. A dire choice has to be made, and it must be made in a way that fairly minimizes the overall harm. In an ideal world, a fair procedure for resolving the problem—perhaps the drawing of lots—would be employed. But here, as with self-defence and the ‘uplifted knife’ cases,251 one should not obscure the clearer cases where there is no need to choose a victim: in the case of the young man on the rope-ladder, blocking the escape of several others, there was no doubt about the person who must be subjected to force, probably with fatal consequences. (b) Medical necessity Is it ever justifiable for a doctor to act contrary to the letter of the law for clinical reasons? There has been little direct discussion of this by the courts or the legislature. The summing-up in Bourne (1939)252 is sometimes cited as authority that a doctor may not be convicted (there, for carrying out an abortion) if it is necessary to save the life of the patient, but that particular area of the law is now subject to express statutory provisions.253 More common in recent

Page 36 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions times has been the acceptance of ‘concealed defences’ of medical necessity, by means of stretching established concepts.254 For example, we saw how Devlin J in the Adams trial modified the general proposition that any acceleration of death satisfies the conduct element for unlawful homicide.255 And the next chapter will show how the House of Lords in Gillick v West Norfolk and Wisbech Area Health Authority (1986)256 deviated from the general proposition that intention includes foresight of virtual certainty. In these decisions the desired effect was to avoid the conviction of a doctor who acted in the ‘best interests’ of the patient, and the chosen method was to distort established concepts rather than to confront the problem openly. One way of bringing the issues into the open would be to create a special defence, which might (following Paul Robinson's suggested draft) provide a permission for reasonable treatment for the promotion of the patient's health.257 The definition would be quite elaborate, and much would turn on the criteria of reasonableness. Some would contend that ‘reasonableness’ should be determined by reference to (p. 133) practices ‘accepted at the time by a responsible body of medical opinion’,258 whereas the ultimate determination ought surely to be that of the court.259 Alternatively, the judges could be left to develop a defence at common law. One of the first English judges to confront some of the issues was Lord Goff in his speech in Re F (1990),260 where he distinguished three forms of necessity—public necessity, private necessity, and necessity in aid of another. The last category was not merely confined to medical cases (e.g. acting to preserve the life of a person who is in a condition that makes it impossible to give consent), but also extends to other cases of action to protect the safety or property of a person unable to give consent. The key element of the decision in Re F was that the necessity was determined by reference to the patient's best interests. It might be possible to reconcile the result of Bourne261 with this approach, but the reasoning in that case was that the interests of the young mother should be allowed to override those of the foetus. That kind of balancing of interests was ruled out in Dudley and Stephens, but it underlies the reasoning of Brooke LJ in Re A (Conjoined Twins: Surgical Separation),262 where he distinguished Dudley and Stephens on the ground that in Re A there was no doubt about the person whose life should be sacrificed and why (that she was incapable of separate existence, and that a failure to operate would hasten the death of both twins). This frank approach to the problem is preferable to the distorting effect of some of the earlier decisions,263 but there remains the question of how exactly a serviceable defence of necessity should be drafted. (c) Necessity and other judicial development of permissions In the past almost all permissions were developed by the judges. If the criminal law is to be codified, should an exhaustive list of permissions be included? The Law Commission thinks not. Its Draft Criminal Code includes provisions on duress and on permissible force, but clause 45(4) provides that a person does not commit an offence by doing an act that is permitted or excused by ‘any rule of common law continuing to apply by virtue of section 4(4)’.264 The intended effect is to preserve the power of the courts to develop defences, including permissions. There is an evident need for flexibility in responding to new sets of circ*mstances, but on the other hand the courts (p. 134) are not suited to the kind of wideranging review that ought to be carried out before a permission is recognized or even taken away.265 A code should go as far as it can in formulating the permissions for what would otherwise be criminal conduct, even if it must rely on terms such as ‘reasonable’ at various points. This may mean the open discussion not merely of hitherto concealed defences such as 266

267 Page 37 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions medical necessity,266 but also of broader concepts of necessity. Thus in Shayler (2001)267 (disclosure of official secrets in order to expose alleged failures by the security services to protect citizens adequately) and again in Jones, Milling et al (2006)268 (damage to an airbase in order to impede aircraft leaving for the invasion of Iraq), the appellate courts took a highly restrictive approach to the prospect of a defence of necessity, whereas decisions such as Re A (Conjoined Twins: Surgical Separation)269 demonstrate a possible category of necessity. In Shayler and in Jones, the courts were urged to apply the defence to what may be termed ‘political necessity’, to permit acts aimed at preventing a greater evil. The conditions for such a defence ought to be tightly circ*mscribed, but that is no reason to deny its existence.270 (d) Statutory recognition of purpose-based permissions There is an increasing tendency to insert into legislation some specific permissions, linked to D's purpose in acting. A long-standing example is s. 5(4) of the Misuse of Drugs Act 1971, allowing a defence to drugs charges if D's purpose in keeping the possession of the drugs was to prevent another from committing an offence or to hand them over to the authorities. Another example is s. 87 of the Road Traffic Regulation Act 1984,271 creating exemptions from speed limits for emergency vehicles. Sections 1 and 4 of the Protection from Harassment Act 1997 both include defences for persons whose course of conduct (which might otherwise amount to ‘stalking’) was ‘pursued for the purpose of preventing or detecting crime’.272 Section 73 of the Sexual Offences Act 2003 states that a person cannot be convicted of aiding, abetting, or counselling a child sex offence if he acts ‘for the purpose of (a) protecting the child from sexually transmitted infection, (b) protecting the physical safety of the child, (c) preventing the child from becoming pregnant, or (d) promoting the child's emotional well-being by (p. 135) the giving of advice’.273 The presence of the permission depends here on the purpose or motive for which D acts, a point underlined by the further requirement that D does not act for the purpose of sexual gratification or in order to encourage sexual activity. Now s. 50 of the Serious Crime Act 2007 provides a ‘defence of acting reasonably’ to a person who would otherwise be guilty of encouraging or assisting crime,274 and, again, one of the factors to be considered in assessing reasonableness is ‘any purpose for which he claims to have been acting’ (s. 50(3)(b)). Thus central to all these permissions is the purpose for which the act was done.

4.8 Conclusions This chapter has dealt with various issues relevant to criminal conduct. It began by examining the extent to which the law reflects the principle of individual autonomy through its requirements of voluntary act and of causation. Not only is the reflection imperfect, but we found that discussion of these conduct elements in criminal liability (sometimes labelled actus reus) involves mental elements and fault elements at several points, e.g. involuntariness, omissions, causation, and purpose in cases of permission. The chapter then turned to the requirement that the conduct be unlawful, in the sense of impermissible, and once again we saw that the boundaries of permissions depend on conflicting considerations which are often not openly or fully analysed. The true significance of many of these issues will not become apparent until Chapters 5 and 6, or later, but three points may be signalled at this stage. First, this chapter has provided ample evidence of the importance, in shaping the criminal law,

Page 38 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions of conflicts between the principle of individual autonomy and principles of welfare. For example, even in relation to the voluntariness requirement—the veritable sanctum of individual autonomy—there are the marks of welfare-based limitations where the rules on insanity, intoxication, and prior fault impinge. Similar conflicts appear clearly in the legislative and judicial approaches to liability for omissions. Even in the permissions for force, the strong individualism which favours the ‘innocent’ defendant has occasionally come into conflict with the underlying social goal of minimizing force in these situations. A second general point is that most of the doctrines considered yield, at crucial junctures, to malleable terminology which leaves considerable discretion to those who apply the law. This is at its plainest with the ubiquitous term ‘reasonable’ in the permissions, although there is now some evidence of a more principled approach. Discretion is also conceded by the proposition that the boundaries of omissions liability turn on (p. 136) the interpretation of particular words in statutes, by various concepts in the sphere of causation (e.g. de minimis, ‘voluntary’), and by such notions as prior fault and ‘external factor’ in automatism. The presence of these openended terms does not empty the rules of their significance, but it raises doubts about the law's commitment to the values upheld by the principle of maximum certainty outlined in Chapter 3.5(i) . It is one thing to leave the rules open-ended when persons are unlikely to rely on them as such (as with the excusatory defences discussed in Chapter 6), although even there the value of consistent judicial decisions should not be overlooked. It is another thing to leave the rules open-ended when citizens as well as courts may rely on them: thus the Law Commission's recognition that the law on self-defence can be structured more explicitly is a welcome step away from universal deference to ‘reasonableness’. In that respect, the enactment of s. 76 of the Criminal Justice and Immigration Act 2008 is almost an irrelevance, and certainly a sorely missed opportunity. Thirdly, in this chapter we have seen the first signs of the impact of the European Convention on English criminal law. More still needs to be done to bring the terms of the defence of reasonable chastisem*nt of children into line with Art. 3, and there is a strong case for going further and abolishing the defence entirely. The effects of Art. 2 on the various rules governing the permission to use force are more difficult to gauge, partly because the leading Strasbourg decisions are not as clear as some would maintain. However, before Parliament accepted the amendment that became s. 76 of the 2008 Act there should have been a proper public assessment of the positive obligations stemming from Art. 2: the government's rather late and cursory treatment of the issue suggests less than full commitment to the Human Rights Act.

Further Reading R. A. duff, Answering for Crime (2007), ch 5. R. D. Mackay, Mental Condition Defences in the Criminal Law (1995), ch 1. H. L. A. Hart and T. Honoré, Causation in the Law (2nd edn., 1985), chs XII and XIII. F. Leverick, Killing in Self-Defence (2006), passim. B. Sangero, Self-Defence in Criminal Law (2006), passim.

Page 39 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions V. Tadros, Criminal Responsibility (2005), ch 10. C. Erin and S. Ost (eds), The Criminal Justice System and Health Care (2007), chs 6–12.

Notes: 1 J. Gardner, ‘On the General Part of the Criminal Law’, in R. A. Duff (ed.), Philosophy and the

Criminal Law (1998). 2 See Chapter 5.5(a). 3 See section 4.3. 4 See section 4.3. 5 See the critical essay by P. H. Robinson, ‘Should the Criminal Law Abandon the Actus

Reus/Mens Rea Distinction?’, in S. Shute, J. Gardner, and J. Horder, (eds), Action and Value in Criminal Law (1993). 6 On German law, in this respect, see G. P. Fletcher, Rethinking Criminal Law (1978), and

Michael Bohlander, Principles of German Criminal Law (2009). 7 See John Gardner, ‘Justification under Authority (2010) 23 Canadian Journal of Law and

Jurisprudence 71; P. H. Robinson, Structure and Function in Criminal Law (1997). 8 A. T. H. Smith, ‘On Actus Reus and Mens Rea’, in P. R. Glazebrook (ed.), Reshaping the

Criminal Law (1978), 95. 9 A. Ashworth, ‘Testing Fidelity to Legal Values: Official Involvement and Criminal Justice’

(2000) 63 MLR 633. 10 Theft Act 1968, s. 21(1). 11 See, generally, J. Horder, Excusing Crime (2004), ch 1. 12 Cf. J. Gardner, Offences and Defences (2007), ch 4. 13 P. Robinson, ‘The Modern General Part: Three Illusions’, in S. Shute and A. P. Simester (eds),

Criminal Law Theory: Doctrines of the General Part (2002); because the permissions guide conduct, M. Moore (Placing Blame, (1997), ch 1) goes so far as to state that they belong to the ‘special part’. 14 This caters for criminal liability for omissions, discussed in section 4.4. 15 See generally R. D. Mackay, Mental Condition Defences in the Criminal Law (1995), ch 1;

R. F. Schopp, Automatism, Insanity, and the Psychology of Criminal Responsibility (1991). 16 See, generally, H. L. A Hart, Punishment and Responsibility (1968), at 107. 17 Thus in Attorney General's Reference (No. 4 of 2000) [2001] Crim LR 578 a driver claimed

that when he put his foot down to the brake pedal he pressed the accelerator instead and the

Page 40 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions bus shot forward out of his control; this was truly a claim of accident and not of involuntary conduct. 18 Cook v Atchison [1968] Crim LR 266. 19 The classic statement is that of J. Austin, Lectures on Jurisprudence (5th edn., 1885), 411–

24. For an excellent application of the philosophy of action to criminal responsibility see R. A. Duff, Criminal Attempts (1996), chs 9–11. 20 A. I. Melden, ‘Willing’, in A. R. White (ed.), The Philosophy of Action (1968), 77. 21 H. L. A. Hart, Punishment and Responsibility (2nd edn., 2008), 103. 22 Hart, Punishment and Responsibility, 255–6, reformulating (in response to criticism) the

passage appearing at 105. 23 The facts of Larsonneur, discussed in the text accompanying n 58. 24 See Model Penal Code, Art. 2.01(1), draft Criminal Code (Law Com No. 177) cl. 33(2), and A.

Smart, ‘Responsibility for Failing to Do the Impossible’ (1987) 103 LQR 532. 25 (1987) 85 Cr App R 321; see also Isitt (1978) 67 Cr App R 44. 26 (1993) 97 Cr App R 429. 27 E.g. Charlson [1955] 1 WLR 317; Quick [1973] QB 910. 28 Hart, Punishment and Responsibility, 106. 29 G. Williams, Textbook of Criminal Law (2nd edn., 1983), ch 29. 30 Law Com No. 177, cl. 33(1). 31 Discussed in detail in Chapter 5.2. 32 [1963] AC 386. 33 For further discussion see Chapter 5.2(c). 34 [1973] QB 910. 35 [1984] AC 156. 36 Hennessy (1989) 89 Cr App R 10; Bingham [1991] Crim LR 433. 37 [1991] 2 QB 92; cf. Bilton, The Daily Telegraph, 20 July 2005, where a person who carried

out serious sexual acts while sleepwalking was apparently acquitted entirely. 38 For an illuminating discussion, see I. Embrahim et al., ‘Violence, Sleepwalking and the

Criminal Law: the Medical Aspects’ [2005] Crim LR 614 , and W. Wilson et al., ‘Violence, Sleepwalking and the Criminal Law: the Legal Aspects’ [2005] Crim LR 624. 39 R. D. Mackay and B. J. Mitchell, ‘Sleepwalking, Automatism and Insanity’ [2006] Crim LR 901;

Page 41 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions Law Commission, Insanity and Automatism Scoping Paper (July 2012). 40 (1978) 79 DLR (3d) 414, on which see R. D. Mackay, ‘Non-Organic Automatism—Some

Recent Developments’ [1980] Crim LR 350. 41 [1990] Crim LR 256 (Snaresbrook Crown Court). 42 The cases are discussed in Jeremy Horder, ‘Pleading Involuntary Lack of Capacity’ (1993)

52 Cambridge LJ 298. 43 [1970] 1 QB 152. 44 On which see Chapter 6.2. If English law were to abandon the Majewski approach and

adopt the Antipodean solution (see Chapter 6.2(e), p. 202), cases of involuntariness through intoxication would fall to be dealt with by the general rules on automatism. 45 Stripp (1979) 69 Cr App R 318. 46 Discussed in Chapter 3.6(v), and in Chapter 5.4(d). 47 [1973] QB 910; for discussion of this development see A. J. Ashworth, ‘Reason, Logic and

Criminal Liability’ (1975) 91 LQR 102. 48 (1983) 77 Cr App R 76. 49 On which see Chapter 5.5(c). 50 See Chapter 5.4(d). 51 (1945) 173 LT 191. 52 See C. Finkelstein, ‘Involuntary Crimes, Voluntarily Committed’, in S. Shute and A. P. Simester

(eds), Criminal Law Theory: Doctrines of the General Part (2002). 53 Chapter 5.2(d). 54 Compare Law Com No. 177, paras. 11.3–11.4, with the discussion of the case of T by

Jeremy Horder, ‘Pleading Involuntary Lack of Capacity’ (1993) 52 Camb LJ 298, at 312–15. For further reflections on the need for reform, see Law Commission, Insanity and Automatism Scoping Paper (July 2012). 55 Cf. M. Moore, Act and Crime (1993), with Duff, Criminal Attempts, chs 9–11, and A. P.

Simester, ‘On the So-Called Requirement for Voluntary Action’ (1998) 1 Buffalo Crim LR 403. 56 Criminal Law Act 1967, s. 4. 57 See Chapter 11. 58 (1933) 149 LT 542. 59 Cf. the analysis by D. J. Lanham, ‘Larsonneur Revisited’ [1976] Crim LR 276, suggesting that

the decision may have been based on prior fault (see Chapter 5.4(e)). 60 Page 42 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions 60 The Times, 28 March 1983. 61 P. R. Glazebrook, ‘Situational Liability’, in Glazebrook (ed), Reshaping the Criminal Law

(1978), 108. 62 As in Coppen v Moore [1898] 2 QB 306. 63 Simester, ‘On the So-Called Requirement for Voluntary Action’, at 410–13. 64 [1984] 2 NZLR 396. 65 [1980] 2 NZLR 235. 66 See Chapter 3.5(g). 67 Cf. Kingston [1995] 2 AC 355, Chapter 6.2. 68 370 US 660 (1962). 69 392 US 514 (1968). 70 D. Husak and B. Mclaughlin, ‘Time Frames, Voluntary Acts, and Strict Liability’ (1992) 12 Law

and Philosophy 95; D. Husak, ‘Re-Thinking the Act Requirement’ (2007) 28 Cardozo LR 2437. 71 Prevention of Crime Act 1953. 72 Theft Act 1968, s. 25. 73 Misuse of Drugs Act 1971, s. 5(2). 74 Misuse of Drugs Act 1971, s. 5(3). 75 [1969] 2 AC 256. 76 These propositions were restated by the Court of Appeal in McNamara (1988) 87 Cr App R

246. 77 [1969] 2 AC, 256, at 307. 78 (1988) 87 Cr App R 270, with commentary by J. C. Smith at [1988] Crim LR 517. 79 See the discussion in Chapter 11.9(c). 80 [2008] 1 Cr App R 25. 81 [2000] 2 AC 428 and [2002] 1 AC 462 respectively, discussed in Chapter 5.5(a). 82 See A. Ashworth, ‘The Unfairness of Risk-Based Possession Offences’ (2011) 5 Criminal

Law and Philosophy 237. 83 On which see Chapter 3.4(c) and (d). 84 See Simester, ‘On the So-Called Requirement for Voluntary Action’, 427. 85 Page 43 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions 85 See Duff, Criminal Attempts, 317–20. 86 With different interpretations of words such as ‘cause’: see G. Williams, ‘What should the

Code do about Omissions?’(1987) 7 Legal Studies 92. 87 Law Com No. 177, ii, para. 7.13; see generally paras. 7.7–7.13. 88 (1977) 65 Cr App R 161. 89 [1983] 2 AC 161. 90 See the criticisms by J. C. Smith [1982] Crim LR 527 and 774, and D. Husak, Philosophy of

Criminal Law (1987), 176–8. 91 See I. M. Kennedy, Treat Me Right (1988), 169–74. 92 [1993] AC 789. 93 On this, see A. McGee, ‘Ending the life of the act/omission dispute: causation in withholding

life-sustaining measures’ (2011) 31 Legal Studies 467. 94 Cf. now the Mental Capacity Act 2005, especially s. 4, and also M. Wilks, ‘Medical Treatment

at the End of Life—a British Doctor's Perspective’, in C. Erin and S. Ost (eds), The Criminal Justice System and Health Care (2007). 95 [2000] 4 All ER 961. 96 J. Rachels, ‘Active and Passive Euthanasia’ (1975) 292 New England J of Medicine 78, as

restated by M. Moore, Act and Crime, 26. 97 Rachels, ‘Active and Passive Euthanasia’, 27. 98 Cf. N. Lacey, C. Wells, and O. Quick, Reconstructing Criminal Law (3rd edn., 2003), 680–95. 99 E.g. Emery (1993) 14 Cr App R (S) 394, and the duty imposed by the Domestic Violence,

Crime and Victims Act 2004, discussed in Chapter 7.6. 100 See R. A. Duff, Answering for Crime (2007), 111, and the discussion in section 4.5 on

Causation. 101 Duff, Answering for Crime, 112–13. 102 For the view that voluntary conduct should not be analysed in terms of causing something

to occur, see John Gardner's review of Michael Moore's Act and Crime (1994) 110 LQR 496. 103 As we will see later, though, this example is subject to a policy exception. Suppose that D1

knocks V unconscious in an area of town where killings of defenceless people are very common. V is later intentionally killed by D2 while she (V) is still unconscious. Given that this kind of outcome might reasonably have been expected, is D1 still to be regarded as a cause of V's death? No. The answer to that question would ordinarily be ‘yes’, because it is within the range of what might reasonably be expected to follow from D1's act. However, the law treats free, deliberate, and informed acts as breaking what would otherwise be a solid chain of

Page 44 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions causation: Kennedy (No. 2) [2008] 1 AC 269 (HL), and D2's act in intentionally killing V was such an act. 104 See, more generally, S. Shute, J. Gardner, and J. Horder (eds), Action and Value in

Criminal Law (1993), Introduction. 105 For an application of the autonomy principle, in preference to the expectations principle

(although these terms are not used) see Kennedy (No 2) [2005] UKHL 38. 106 E.g. Attorney-General's Reference No. 4 of 1980 (1981) 73 Cr App R 40. 107 D2 cannot be found criminally liable for manslaughter unless D2's negligence in breaching

the duty of case was ‘gross’ (see Chapter 7); but if the negligence was indeed gross, then the treatment will normally be the kind of unexpected calamity befalling V likely to lead a jury to conclude that the chain of causation leading from D1's act to V's death was broken. 108 For a critical argument along these lines, see A. Norrie, Crime, Reason and History (2nd

edn., 2001), ch 7. 109 Although, in some cases, it is in fact possible for intended consequences to be too remote

to be attributed to the person trying to bring them about. Suppose D believes he can jump across the English Channel and land in France. As he jumps, a tornado sweeps him up and carries him across to France. Even assuming that, had he not initially jumped, the tornado would not have carried him all the way to France, the result's occurrence is too remote in causal terms from the way that D envisaged the result coming about to be attributable to his action. See, generally, R. A. Duff, Intention, Agency and Criminal Liability (Oxford, 1990). 110 E.g. Alphacell Ltd v Woodward [1972] AC 824, but cf. the slightly more definite approach in

National Rivers Authority v Yorkshire Water Services [1995] 1 AC 444. See generally N. Padfield, ‘Clean Water and Muddy Causation’ [1995] Crim LR 683. 111 (1976) 62 Cr App R 41. 112 (1976) 62 Cr App R 41, 45. 113 (1976) 62 Cr App R 41, 46; cf. Cheshire [1991] 1 WLR 844, referring to a ‘significant

contribution’. 114 Although, of course, the grandparents’ role may be relevant to other kinds of judgment that

people may wish to make, such as whether ‘criminality runs in the family’. 115 See Hart and Honoré, Causation in the Law, ch 1 and passim, and the derivative

discussions by S. Kadish, Blame and Punishment (1987), ch 8, and H. Beynon, ‘Causation, Omissions and Complicity’ [1987] Crim LR 539; cf. also Williams, Textbook of Criminal Law, ch 14. 116 [1957] Crim LR 365. 117 H. L. A. Hart and T. Honoré, Causation in the Law (2nd edn., 1985), 344–5. 118 See (1992) BMLR 38. 119 Page 45 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions 119 A. Arlidge, ‘The Trial of Dr. Moor’ [2000] Crim LR 31. 120 The subsequent case of Re A (Conjoined Twins: Surgical Separation) [2000] 4 All ER 961

(n 95, and accompanying text) is one of the few to confront this question. 121 See S. Ost, ‘Euthanasia and the Defence of Necessity’, in C. Erin and S. Ost (eds), The

Criminal Justice System and Health Care (2007), and section 4.8(b). 122 There are isolated exceptions: see Corporate Manslaughter and Corporate Homicide Act

2007, s. 1(3), stating that ‘an organisation is guilty of an offence under this section only if the way in which its activities are managed or organised by its senior management is a substantial element in the breach’ (our italics). See further Chapter 7.5(c). 123 The old law on obtaining by deception (now replaced by the Fraud Act 2006, on which see

Chapter 9.8) provided a possible example of ‘but for’ causation: see p. 125 of the fifth edition of this work. 124 This was how Lord Bingham paraphrased Cato in Kennedy (No. 2) [2008] UKHL 1 AC 269,

at 274. 125 See the discussion of crimes of possession in Chapter 4.3(b). 126 Law Com No. 177, cl. 17(1)(a). 127 American Law Institute, Model Penal Code, s. 2.03. 128 Hart and Honoré, Causation in the Law, 342; the Draft Criminal Code refers to an

intervening act ‘which could not in the circ*mstances have been reasonably foreseen’, Law Com No. 177, cl. 17(2). 129 (1840) 9 C and P 356; cf. G. Williams, ‘Finis for Novus Actus’ [1989] Camb LJ 391. 130 Cf. Cogan and Leak [1976] 1 QB 217, on ‘semi-innocent agency’, discussed in Chapter

10.6. 131 Bourne (1952) 36 Cr App R 125; see also Chapter 10.6. 132 (1983) 76 Cr App R 279; see Hart and Honoré, Causation in the Law, 330–4. 133 Article 2 of the ECHR suggests so: see Chapter 4.6(b). 134 See further Chapter 6.4; cf. the arguments of P. A. J. Waddington, ‘“Overkill” or “Minimum

Force”?’ [1990] Crim LR 695, with the implications of Art. 2 of the ECHR. 135 [1999] 2 AC 22, overruling Impress (Worcester) Ltd v Rees [1971] 2 All ER 357. 136 Per Lord Hoffmann at 36. 137 [2008] 1 AC 269, at 276, adding that the House ‘would not wish to throw any doubt on the

correctness of the Empress Car case’. 138 [2008] 1 AC 269, at 276; the Court of Appeal's first decision in this case was also much

Page 46 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions criticized, see Kennedy [1999] Crim LR 65. 139 Smith [1959] 2 QB 35, distinguishing Jordan (1956) 40 Cr App R 152; cf. the critical attack

of Norrie, Crime, Reason and History, 147–8. 140 [1991] 1 WLR 844. 141 [1991] 1 WLR 844, 852. 142 Per Lord Lane CJ, in Malcherek (1981) 73 Cr App R 173. 143 Section 4.7. 144 A clear example, on these facts, is the American case of State v Frazer (1936) 98 SW (2d)

707. 145 See Chapters 5.4(a) and 7.7. 146 See the criticism of Hart and Honoré by Norrie, Crime Reason and History, 149–50. 147 (1972) 56 Cr App R 95; see also Corbett [1996] Crim LR 594. 148 (1972) 56 Cr App R 95, at 97. 149 (1975) 61 Cr App R 271. 150 Per Lawton LJ, (1975) 61 Cr App R 271, 274. 151 (1972) 56 Cr App R 95. 152 See also Dear [1996] Crim LR 595. 153 For discussions see Husak, Philosophy of Criminal Law, ch 6; A. Leavens, ‘A Causation

Approach to Omissions’ (1988) 76 Cal LR 547; H. Beynon, ‘Causation, Omissions and Complicity’ [1987] Crim LR 539. 154 Calhaem [1985] QB 808. 155 Hart and Honoré, Causation in the Law, 51; cf. G. Williams, ‘Finis for Novus Actus’ [1989]

Camb LJ 391 , at 398. 156 J. C. Smith, ‘Aid, Abet, Counsel and Procure’, in P. R. Glazebrook (ed.), Reshaping the

Criminal Law (1978); Kadish, Blame and Punishment, ch 8; cf. Part 2 of the Serious Crime Act 2007, discussed in Chapter 11.7. 157 See Chapter 5.3(b). 158 See R. Card, ‘Authority and Excuse as Defences to Crime’ [1969] Crim LR 359 , 415. 159 Per Lord Griffiths in Beckford v R [1988] AC 130, at 144. 160 See Chapter 6.4, and cf. R. H. S. Tur, ‘Subjectivism and Objectivism: Towards Synthesis’, in

S. Shute, J. Gardner, and J. Horder (eds), Action and Value in Criminal Law (1993). 161 Page 47 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions 161 S. Uniacke, Permissible Killing: the Self-Defence Justification of Homicide (1994), 26 and

ch 2 generally. 162 See section 4.1, and n 12. 163 For detailed studies of the law on self-defence and the theory underlying it, see F.

Leverick, Killing in Self-Defence (2006), and B. Sangero, Self-Defence in Criminal Law (2006). 164 Note that Protocol 6 to the Convention requires the abolition of the death penalty. Several

European States, including the UK, have agreed to this protocol, which is also brought into English law by the Human Rights Act 1998. 165 Rivas v France [2005] Crim LR 305 and RJ and M-J D v France [2005] Crim LR 307. 166 J. C. Smith, ‘The Right to Life and the Right to Kill in Law Enforcement’ (1994) NLJ 354. 167 For a similar argument, accepted by the European Court, see A v UK, discussed in section

4.8. 168 Fletcher, Rethinking Criminal Law, 862–3; cf. A. Dershowitz, Preemption (2006), 197–9. 169 Uniacke argues that there is no conceptual difficulty with the notion of forfeiture so long as

we accept that the right to life, like many other rights, is conditional on our conduct: Permissible Killing, 201 and ch 6 generally. 170 Section 4.5. 171 See Chapter 3.6(r). 172 See n 216 and accompanying text. 173 A. Ashworth, ‘Self-Defence and the Right to Life’ [1975] CLJ 282. 174 Ashworth, ‘Self-Defence and the Right to Life’, 285. 175 See the judgment of Ward LJ in Re A (Conjoined Twins: Surgical Separation) [2000] 4 All

ER 961. 176 Cf. s. 59 of the Serious Crime Act 2007 (‘the common law offence of inciting the

commission of another offence is abolished’), discussed in Chapter 11.6. 177 Law Com No. 177, cl. 44, mostly restated in Law Com No. 218, Legislating the Criminal

Code: Offences Against the Person and General Principles (1993), cl. 27. 178 Criminal Damage Act 1971, s. 5(2). 179 See e.g. Jones, Milling et al [2007] 1 AC 136. 180 Owino [1996] 2 Cr App R 128, not following Scarlett (1994) 98 Cr App R 290. See also

Tudor [1999] 1 Cr App R (S) 197. Cf. the controversy surrounding the US case of People v Goetz (1986) 68 NY 2d 96, where D had shot and wounded four youths on the New York subway after they had demanded five dollars from him. 181 Page 48 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions 181 As the European Court of Human Rights held in Nachova v Bulgaria (2006) 42 EHRR 933. 182 Report of the Royal Commission on the Law Relating to Indictable Offences (1879, C. 2345),

note B, at 44; see at 11 for an assertion of the principle that ‘the mischief done by, or which might reasonably be anticipated from, the force used is not disproportioned to the injury or mischief which it is intended to prevent’. 183 See Leverick, Killing in Self-Defence, ch 6. For a different view see the approach of the

German Supreme Court: M. Bohlander, Principles of German Criminal Law (2009), 99–114. 184 [2006] Crim LR 546. 185 For the eighteenth century authorities, see Ashworth, ‘Self-Defence and the Right to Life’,

299–301. 186 E.g. in Andronicou and Constantinou v Cyprus (1998) 25 EHRR 491, at para. 171; Gül v

Turkey (2002) 34 EHRR 719, at para. 77; Nachova v Bulgaria (2006) 42 EHRR 933. 187 Model Penal Code, s. 3.04. 188 J. C. Smith, Justification and Excuse in the Criminal Law (1989), 109 and ch 4; F. Leverick,

Killing in Self-Defence, ch 7. 189 Law Commission, Simplification of the Criminal Law: Kidnapping (Consultation Paper No.

200, 2011). 190 Fennell [1971] 1 QB 428; Ball [1989] Crim LR 579. 191 Model Penal Code, s. 3.04(2)(a)(i); Law Com No. 177, cl. 44(4). 192 (1850) 4 Cox CC 358. 193 [1985] NI 457. 194 Cf. Nachova v Bulgaria, n 181 and text thereat; see also n 198. For controversy over the

Dadson principle, see R. Christopher, ‘Unknowing Justification and the Logical Necessity of the Dadson Principle in Self-Defence’ (1995) 15 OJLS 229, and P. H. Robinson, ‘Competing Theories of Justification: Deeds v. Reasons’ with J. Gardner, ‘Justifications and Reasons’, both in A. P. Simester and A. T. H. Smith (eds), Harm and Culpability (1996). 195 E.g. Attorney-General for Northern Ireland's Reference [1977] AC 105; Chisam (1963) 47

Cr App R 130. 196 See Jones, Milling et al [2007] 1 AC 136; see also Lord Bingham's judgment on duress and

the duty to avoid using force in Hasan [2005] 2 AC 467, discussed in Chapter 6.3. See also J. Horder, Homicide and the Politics of Law Reform (2012), at 249–52. 197 [1984] QB 456; see also Cousins [1982] QB 526. 198 J. Dressler, ‘Battered Women who Kill their Sleeping Tormentors’, in S. Shute and A. P.

Simester (eds), Criminal Law Theory (2002), 269. 199 Page 49 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions 199 J. Horder, ‘Killing the Passive Abuser’, in Shute and Simester, Criminal Law Theory, at 292–

3; Homicide and the Politics of Law Reform, at 252–5; Leverick, Killing in Self-Defence, ch 5. Cf. also the proposal to create a new partial defence (in murder cases) for situations where D acted during a loss of self-control stemming from D's fear of serious violence from V, discussed in Chapter 7.4(b). 200 See Evans v Hughes [1972] 3 All ER 412; Smith, Justification and Excuse, 117–23; and

now D. J. Lanham, ‘Offensive Weapons and Self-Defence’ [2005] Crim LR 85. 201 A point left open by the European Commission in Kelly v UK (1993) 74 DR 139. 202 [1969] 1 WLR 839. 203 [1985] 1 WLR 816. 204 J. C. Smith and B. Hogan, Criminal Law (5th edn., 1983), 327, quoted by the Court of

Appeal in Bird [1985] 1 WLR 816; see now Criminal Law (12th edn., by D. Ormerod, 2011), 368. 205 Leverick, Killing in Self-Defence, 82, and ch 4 generally. For the householder's position,

see now s. 148 Legal Aid, Sentencing and Punishment Act 2012; section 4.6(c). 206 [1972] Crim LR 435. 207 (1938) 53 Wyo 304. 208 Cf. Lord Mance in Jones, Milling et al. [2007] 1 AC 136, with the arguments of F. McAuley

and J. P. McCutcheon, Criminal Liability: a Grammar (2000), 760–1. 209 [1999] Crim LR 998; cf. Beatty v Gillbanks (1882) 9 QBD 308 and Nicol and

Selvanayagam v DPP [1996] Crim LR 318. 210 See section 4.4. 211 E.g. Beckford v R [1988] AC 130, at 144. 212 See n 204 and accompanying text. 213 A. V. Dicey, Introduction to the Study of the Law of the Constitution (8th edn., 1923), 489. 214 See the discussion of Thain, n 193 and text thereat. 215 Sir Ian Blair in oral evidence to the House of Commons Home Affairs Committee, Counter-

Terrorism, 13 September 2005, Qs 59–60. 216 (1996) 21 EHRR 97; for analysis of the ECHR case law, see Leverick, Killing in Self-

Defence, ch 10, and B. Emmerson, A. Ashworth, and A. Macdonald (eds), Human Rights and Criminal Justice (3rd edn., 2012), 794–803. 217 Cf. P. A. J. Waddington, ‘“Overkill” or “Minimum Force”?’ [1990] Crim LR 695, for the

argument that if officers do not shoot to kill they risk the possibility that an injured suspect might still be able to kill or wound someone—in which case, shooting to kill might be permissible

Page 50 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions in the first place. 218 (1998) 25 EHRR 491. 219 (2002) 34 EHRR 719. 220 (2002) 34 EHRR 719, para. 82. 221 However, it is widely recognized that the State has a positive duty to ensure that the right

to life of all its citizens is protected, and to ensure that its police and military personnel conduct their operations with due respect for Art. 2. Passages in several judgments (e.g. Isaveya v Russia (2005) 41 EHRR 791, at para. 175) suggest that there may be grounds for charging senior police officers with negligent manslaughter if they fail in this duty. See also Juozaitiene v Lithuania (2008) 47 EHRR 1194. 222 By the Privy Council in Shaw v R [2002] 1 Cr App R 10; cf. Martin [2002] 1 Cr App R 27,

where the Court of Appeal nevertheless held that psychiatric evidence to assist the jury to understand D's likely perceptions was not admissible. 223 Section 76(5) states that voluntary intoxication should be left out of account. General

discussions of mistaken beliefs will be found in Chapter 6.4 and of intoxication in Chapter 6.2. 224 McCann et al. v UK (1996) 21 EHRR 97, at para. 200; Andronicou and Constantinou v

Cyprus (1998) 25 EHRR 491, at para. 192; Gül v Turkey (2002) 34 EHRR 719, at paras. 78–82. See Leverick, Killing in Self-Defence, ch 10. 225 See n 221. 226 (2005) 41 EHRR 458. 227 (2005) 41 EHRR 458, paras. 138–9. 228 Per Collins J in R (Bennett) v HM Coroner for Inner London [2006] EWHC Admin 196; see

also the government response to the Joint Committee, printed as Appendix 7 to the report. 229 Joint Committee on Human Rights, Legislative Scrutiny (15th report, session 2007–8), para.

2.35. 230 As the Law Commission has recommended their exclusion from the partial defence of

provocation: see Chapter 7.5(b). 231 See A. Ashworth (reference at n 173), and Leverick, Killing in Self-Defence, ch 6. 232 [1971] AC 814, at 832. 233 Bubbins v UK (2005) 41 EHRR 458, at para. 139. 234 A phrase from Palmer, n 232; see also the insistence of the Joint Committee on Human

Rights (n 229), para. 2.24, that allowing force which is not ‘grossly disproportionate’ would go too far and breach the State's obligations under Art. 2. 235 Per Lord Diplock, in Attorney-General for Northern Ireland's Reference [1977] AC 105. 236 Page 51 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions 236 On which see J. Horder, Excusing Crime (2004), 48–52; J. Horder, Homicide and the

Politics of Law Reform (2012), 251–2. 237 Commentaries on the Laws of England, iii, 3–4. 238 Brown v United States (1921) 256 US 335, at 343. 239 Cf., however, the use of the concession in the Andronicou case (reference at n 218), and

in Bubbins (reference at n 226). 240 In the Draft Criminal Code, Law Com No. 177, cl. 44; however, on some of the issues, this

book differs from the Commission's particular recommendations. 241 See Chapter 6.3. 242 [1987] AC 417. 243 Cf. J. J. Thomson, ‘Self-Defense’ (1991) 20 Philosophy and Public Affairs 283, with R.

Christopher, ‘Self-Defense and Objectivity’ (1998) 1 Buffalo CLR 537; on necessity as an excuse see Chapter 6.3(a). 244 Smith, Justification and Excuse, 73–9. 245 Cf. Dudley and Stephens (1884) 14 QBD 273, the case in which two men saved

themselves by killing and eating the weakest member of a threesome who had been adrift in a boat for many days; but they were rescued the following day, and some have questioned the necessity of their act. See A. W. B. Simpson, Cannibalism and the Common Law (1984), and Chapter 6.4(a). 246 Model Penal Code, s. 3.02; cf. G. P. Fletcher, Rethinking Criminal Law (1978), 788–98. 247 Cf. the remarks of the House of Lords in Jones, Milling et al [2007] 1 AC 136. 248 Thomson, ‘Self-Defense’, 309; see also J. Finnis, ‘Intention and Side-Effects’, in R. G. Frey

and C. W. Morris (eds), Liability and Responsibility (1991), for the argument that it is never morally right to choose (intend) to take another's life. 249 United States v Holmes (1842) 26 Fed Cas 360. 250 Dudley and Stephens (1884) 14 QBD 273. 251 See Brown v United States (1921) 256 US 335, and the text accompanying n 229. 252 [1939] 1 KB 687. 253 Abortion Act 1967. 254 Smith, Justification and Excuse, 64–70; A. Ashworth, ‘Criminal Liability in a Medical

Context: the Treatment of Good Intentions’, in A. P. Simester and A. T. H. Smith (eds), Harm and Culpability (1996). 255 See section 4.6(a), on the cases of Dr Adams and Dr Moor; see also the ‘medical’

Page 52 of 54

Criminal Conduct: Actus Reus, Causation, and Permissions exception to the principle that a voluntary intervening human act negatives causal responsibility in section 4.6(b)(ii). 256 [1986] AC 112, discussed in Chapter 5.5(b). 257 P. Robinson, Criminal Law Defences (1984), ii, 173. 258 Airedale NHS Trust v Bland [1993] AC 789, per Lord Browne-Wilkinson at 883. 259 As Lord Mustill argued in the Bland case. 260 [1990] 2 AC 1; see S. Gardner, ‘Necessity's Newest Inventions’ (1991) 11 OJLS 125. 261 See n 252. 262 [2000] 4 All ER 961, on which see J. Rogers, ‘Necessity, Private Defence and the Killing of

Mary’ [2001] Crim LR 515. 263 See S. Ost, ‘Euthanasia and the Defence of Necessity’, in C. Erin and S. Ost (eds), The

Criminal Justice System and Health Care (The Criminal Justice System and Health Care). 264 For discussion see Law Com No. 177, ii, para. 12.41. 265 See particularly the two articles by Rogers, on reasonable chastisem*nt (J. Rogers ‘A

Criminal Lawyer's Response to Chastisem*nt in the European Court of Human Rights’ [2002] Crim LR 98) and on necessity and private defence (reference at n 262), and the discussion of ‘the democracy problem’ by Gardner (reference at n 260). 266 See Chapter 5.3(c). 267 [2001] 1 WLR 2206; the case went to the House of Lords on other grounds: [2003] 1 AC

247. 268 [2007] 1 AC 136. 269 See n 262 and text thereat. 270 S. Gardner, ‘Direct Action and the Defence of Necessity’ [2005] Crim LR 371; cf. C.

Clarkson, ‘Necessary Action: a New Defence’ [2004] Crim LR 81. 271 As amended by s. 19 of the Road Safety Act 2006. 272 See further A. Ashworth, ‘Testing Fidelity to Legal Values: Official Involvement and Criminal

Justice’, in Shute and Simester (eds), Criminal Law Theory, at 322 –30. 273 See also the similar defence for principals in s. 14 of the Sexual Offences Act 2003

(facilitating the commission of a child sex offence). 274 For the Law Commission's approach, see Law Com No. 300, Inchoate Liability for Assisting

and Encouraging Crime (2007), 82–8.

Page 53 of 54

Criminal Capacity, Mens Rea, and Fault

Principles of Criminal Law (7th edn) Andrew Ashworth and Jeremy Horder Publisher: Oxford University Press Print ISBN-13: 9780199672684 DOI: 10.1093/he/9780199672684.001.0001

Print Publication Date: May 2013 Published online: Sep 2013

5. Criminal Capacity, Mens Rea, and Fault Chapter: (p. 137) 5. Criminal Capacity, Mens Rea, and Fault Author(s): Andrew Ashworth and Jeremy Horder DOI: 10.1093/he/9780199672684.003.0005

5.1 The issues 5.2 Agency, capacity, and mental disorder 5.3 Corporate liability 5.4 Fault and mens rea: general principles 5.5 Varieties of fault 5.6 The variety of fault terms 5.7 The referential point of fault Further reading

5.1 The issues In Chapter 4 we examined some of the fundamental requirements of a crime that are usually embraced by the notion of actus reus—voluntary act, causation, and absence of justification.

Criminal Capacity, Mens Rea, and Fault One feature of that discussion was how frequently a kind of fault element was to be found playing a role in the actus reus: for example, the doctrine of prior fault in automatism (4.2(e)), compassionate purpose in causation (4.5(a)), and elements of prior fault in the principles of permissible? force (4.6(d) and (e)). In this chapter we deal first with another fundamental requirement of a crime, criminal capacity. It is a precondition of criminal liability that the defendant is a person with sufficient capacity to be held responsible, and this leads to an examination of infancy and insanity as barriers to criminal responsibility (5.2), and then to corporate criminal liability (5.3). Having established that the defendant meets the preconditions for criminal responsibility, we then move to the fault requirements, or mens rea (as criminal lawyers often call them). An important fault element is intention, proof of which is required in major crimes such as murder, rape, and robbery. The presence of a fault element may not mean that D was necessarily culpable, in an all-things considered sense. The latter depends on whether D's conduct was justified (Chapter 4) or was excused (Chapter 6). But, those considerations apart, the fault element normally indicates culpability, and in 5.4 we explore some of the reasons for and against the criminal law requiring proof of fault in any form. We then go on, in 5.5, to give detailed consideration to the principal varieties of fault requirement in the criminal law. In broad terms there is a hierarchy of fault requirements: intention is the highest form, followed by recklessness, and then by negligence. However, the majority of crimes in English law impose strict liability, (p. 138) being offences for which neither intention, nor recklessness, nor negligence needs to be proved. Most of these are summary offences, triable only in the magistrates’ courts and carrying relatively low penalties. However, around half of the offences triable in the Crown Court have a strict liability element: that is, they do not require proof of fault in relation to one or more of the conduct, consequence, or circ*mstance elements of the crime.1 An example is assault occasioning actual bodily harm contrary to s. 47 of the Offences Against the Person Act 1861. This offence requires proof of fault (intention or recklessness) with regard to the element of assault, but does not require proof of fault in relation to the occasioning of actual bodily harm.2 It is therefore important, throughout this chapter, to distinguish between normative claims about the principles the law should observe and the realities of the law as it is. It may be argued that criminal conviction should always be founded on proof of fault, but it would require a reform affecting thousands of offences to turn that aspiration into a reality. In order to underscore this point, section 5.5 of this chapter begins with a discussion of strict liability, before turning to intention, recklessness, and negligence. Two further points must be made at this stage. One is that it is not unusual for an offence to have two or more different fault elements, relating to different aspects of the actus reus: the point may be illustrated by considering the several different elements in ‘abuse of trust’ offences in ss. 16–19 of the Sexual Offences Act 2003, which include the various requirements that D intentionally does the sexual activity, that D knows or could reasonably be expected to know of the circ*mstances by virtue of which he is in a position of trust in relation to the victim, and which requires no knowledge as to age if the victim is under 13, but requires reasonable belief that the victim is 18 if in fact he or she is aged 13–17. It is much clearer if crimes such as these are analysed in terms of their separate elements—e.g. conduct, circ*mstances, result— so as to ascertain what form of fault is required in respect of each element.3 This leads to the second point: as we go through the specific offences in Chapters 7, 8, and 9 we will see that it

Page 2 of 57

Criminal Capacity, Mens Rea, and Fault is not just the terminology of fault requirements that is diverse (many statutes contain words such as ‘maliciously’ and ‘wilfully’) but also their substance (e.g. requirements such as ‘dishonestly’ and ‘fraudulently’). The discussion in this chapter does not purport to cover all fault terms and is confined to positive fault requirements, in other words, the mental element specified in (or implied within) the offence and which the prosecution must establish. Selected for analysis below are core fault terms such as intention, recklessness, knowledge, and negligence (the core fault elements under the US Model Penal Code), together with offences that use other terminology which has been held to impose a form of ‘strict’ or no-fault liability. One difficulty in focusing on a small range of fault terms is that the existing variety of approaches to fault is not captured, and that any generalization (p. 139) on the basis of a few fault terms may lead to inaccurate conclusions.4 We will return to that difficulty at the end of the chapter, after laying some foundations for wider discussion.

5.2 Agency, capacity, and mental disorder One of the fundamental presumptions of the criminal law and criminal liability is that the defendant is ‘normal’, i.e. is able to function within the normal range of mental and physical capabilities. Many of the principles of individual fairness discussed in Chapter 3 presuppose an individual who is rational and autonomous: otherwise he does not deserve to be liable to criminal punishment. A person who is mentally disordered may fall below these assumed standards of mental capacity and rationality, and this may make it unfair to hold him responsible for his behaviour. It is for this autonomy-based reason that most systems of criminal law contain tests of ‘insanity’ which result in the exemption of some mentally disordered persons from criminal liability. A similar rationale may be given for the voluntariness requirement, discussed in Chapter 4.2. There is also the prior question of whether the defendant is fit to be tried—whether the person can participate in the trial in a sufficiently meaningful sense. It is an essential precondition of a fair trial, as Antony Duff has argued,5 that the defendant is a responsible citizen who is answerable before the court. The doctrine of ‘unfitness to plead’ embodies a procedural attempt to deal with this in relation to mentally disordered defendants.6 Once it has been decided that a person is fit to plead, there is still the question whether at the time of the alleged act D was a sufficiently responsible moral agent: the defence of insanity, discussed in paragraphs (b) and (c), addresses this issue. Before, that, however, a few words must be said about young children, where legitimate concerns about answerability to the court and moral agency have received an unsatisfactory response from English lawmakers. (a) The minimum age of criminal responsibility In England and Wales the minimum age of criminal responsibility is 10, substantially lower than the minimum age in many other European countries (although higher than in some US states), where teenage children are dealt with in civil tribunals up to (p. 140) the age of 13 (France), 14 (Germany), 15 (Scandinavia) or 16 (Spain and Portugal). At common law the presumption of doli incapax applied to children under 14, requiring the prosecution to establish that the child knew that the behaviour was seriously wrong before the case could go ahead. The presumption was much criticized,7 some arguing that children who failed to realize the 8 Page 3 of 57

Criminal Capacity, Mens Rea, and Fault wrongness of their behaviour were more in need of conviction and compulsory treatment,8 and it was abolished by s. 34 of the Crime and Disorder Act 1998.9 However, it remains important to think about fundamental issues in relation to the responsibility of young offenders. Are they fit to stand trial at the age of 10? Do they have sufficient understanding of the proceedings to participate meaningfully in them? In what sense are they responsible citizens at that age? Can it be said that, when they do criminal things with the required fault element, they are acting as moral agents, in a sufficiently full sense?10 The first two points were discussed by the European Court of Human Rights in V and T v United Kingdom (1999),11 drawing on the United Nations Convention on the Rights of the Child, which does not lay down a minimum age of criminal responsibility but does declare several other relevant standards.12 The Court held that, although the trial process to which the two 11-year-old applicants were subjected did not amount to ‘inhuman and degrading treatment’ within Art. 3 of the Convention, the trial did violate Art. 6 in its failure to ensure that the boys understood the proceedings and had the opportunity to participate, and in the failure to reduce feelings of intimidation and inhibition. A subsequent Practice Direction sets out the steps that trial judges should take in these unusual cases in order to comply with Art. 6,13 but the Strasbourg Court has held that this gives insufficient priority to the need to ensure that all young children have adequate opportunity to participate meaningfully in the criminal trial.14 The European Commissioner on Human Rights has specifically recommended that consideration be given to raising the age of criminal responsibility ‘in line with norms prevailing across Europe’, on the grounds that children of 10, 11, or 12 cannot have sufficient understanding of the nature and consequences of their actions.15 The cognitive abilities of young children may not be sufficiently developed; their self-control may not yet have developed adequately; and they may be particularly susceptible to (p. 141) peer pressure at that age.16 These are all aspects of moral development and, since childhood and adolescence are the time when moral reasoning and self-control should be learnt, it is not reasonable for the criminal law to demand as much from children as from adults.17 The case for raising the minimum age of criminal responsibility in England and Wales is overwhelming.18 (b) The special verdict of insanity19 If the defendant is thought fit to stand trial, then the issue of mental disorder may be raised as a defence; namely, that at the time of the alleged offence D was too disordered to be held liable. Medical evidence will be crucial in determining this,20 but it is for the law to lay down the appropriate test. Mental disorder is a broad concept under the Mental Health Act 1983, as amended by the Mental Health Act 2007, and few would maintain that all those who fall within one of the four classes of disorder under that Act should be exempted from criminal liability. The criminal law has settled on a much narrower conception of ‘insanity’, proof of which should lead to a verdict of ‘not guilty by reason of insanity’. In order to understand how this defence functions, however, it is important to bear in mind that until the Criminal Procedure (Insanity and Unfitness to Plead) Act 1991 came into force, the result of a successful defence of insanity was mandatory and indefinite commitment to a mental hospital. Whilst research revealed that about one-fifth of defendants thus committed were released within nine months,21 there was no certainty of, or entitlement to, early release and the potentially severe effect of the insanity verdict was enough to lead many defendants to plead guilty and to hope for a more favourable disposal at the sentencing stage.22 The 1991 Act gave the court the same discretion after an insanity verdict (except in murder cases, dealt with later) as it has 23 Page 4 of 57

Criminal Capacity, Mens Rea, and Fault after a finding of unfitness to plead: hospital order, supervision, absolute discharge.23 This still leaves the possibility that the court will order deprivation of liberty, even though the defendant has ‘succeeded’ on a ‘defence’. Insanity defences (p. 142) remain rare, with an average of fifteen per year from 1999–2001, but around half of the disposals are community-based, i.e. supervision or absolute discharge.24 The possible legal consequences of the insanity verdict show the tension between considerations of individual autonomy and policies of social protection in this sphere, and the same tension is manifest in the evidential and procedural provisions.25 Insanity is the only general defence where the burden of proof is placed on the defendant,26 a paradox when one reflects that the consequence of a successful defence may be a court order favouring social protection rather than the defendant's own interests. The prosecution may raise insanity if the defendant pleads diminished responsibility in response to a murder charge,27 and, according to one view, can do so in all cases where D puts state of mind in issue.28 The prosecution bears the burden of proving insanity here, which is much more appropriate given the consequences of the verdict of ‘not guilty by reason of insanity’. The requirements of the defence of insanity were laid down by the judges in M'Naghten's Case as long ago as 1843: 29

to establish a defence on the ground of insanity, it must be clearly proved that, at the time of committing the act, the party accused was labouring under such a defect of reason, from disease of the mind, as not to know the nature and quality of the act he was doing; or, if he did know it, that he did not know he was doing what was wrong.

A ‘defect of reason’ means the deprivation of reasoning power, and does not apply to temporary absent-mindedness or confusion.30 It is, however, limited to cognitive defects, and therefore excludes from the insanity defence those forms of mental disorder that involve significant emotional or volitional deficiencies. Although in that respect the definition of insanity is very narrow,31 in other respects it is so wide as to go well beyond even the general definition of mental disorder in the Mental Health Acts 1983–2007. Thus the phrase ‘disease of the mind’ has been construed so as to encompass any disease which affects the functioning of the mind—whether its cause be organic or functional, and whether its effect be permanent or intermittent—so long as it was operative at the time of the alleged offence.32 This means, as we saw (p. 143) in Chapter 4.2(c), that any condition which affects the functioning of the mind and which results from an ‘internal’ rather than an ‘external’ cause will be deemed to be a ‘disease of the mind’, and if D relies on it in his defence he will be held to be raising the defence of insanity. The ‘internal factor’ doctrine has resulted in epilepsy,33 sleepwalking,34 and hyperglycaemia35 being classified as insanity. This shows that the policy of social protection has gained the upper hand, and that the judiciary has been prepared to overlook the gross unfairness of labelling these people as insane in order to ensure that the court has the power to take measures of social defence against them. Even then, the policy of protection has not been carried to its logical conclusion, since the law now perpetrates the absurdity of classifying hyperglycaemia as insanity (protective measures possible under the 1991 Act) whilst, because of the external/internal distinction, classifying hypoglycaemia as automatism (resulting in an outright acquittal unless prior fault can be shown).36 More will be said about

Page 5 of 57

Criminal Capacity, Mens Rea, and Fault this later. Where it is established that there was a defect of reason due to disease of the mind, it is then necessary to show that it had one of two effects.37 First, the defence is fulfilled if D did not know the nature and quality of the act—in other words, did not realize what he was doing. In most cases this would show the absence of intention, knowledge, or recklessness; but since this mental state arises from insanity, considerations of protection are held to require the special verdict rather than an ordinary acquittal. Secondly, the defence is fulfilled if D did not know that he was doing wrong. English law appears ambivalent about the proper approach to this requirement: the Court of Appeal has confirmed that ‘wrong’ bears the narrow meaning of ‘legally wrong’,38 although there is evidence that in practice courts sometimes act upon the Australian interpretation of ‘failure to appreciate that the conduct was morally wrong’ (usually, where D believes that he must, for some distorted reason, do the act).39 The main difficulty with English law's insistence on confining this second limb to cases in which D knew that his or her act was legally wrong, is that it is using a test to judge those suffering from certain kinds of mental disorder that negate the moral relevance of the test. In terms of culpability, someone so deluded that he or she kills a boy because he or she thinks that the boy is the re-incarnation of Napoleon, is more or less on a par (p. 144) with someone so deluded that he or she kills a girl because he or she believes that the girl is at that moment trying to kill the Queen through the use of supernatural thought powers. Yet, formally, only the latter comes within the scope of the insanity defence. Only in the latter case does D's delusion mean that he or she does not appreciate that what he or she is doing is ‘legally’ wrong because he or she believes that he is acting necessarily and proportionately in defence of another. By contrast, in the former case, D may think that what he or she did was not morally wrong (preventing a full-grown Napoleon re-appearing on the world stage), but may fully appreciate that it was legally wrong. This is unsatisfactory, but it seems that legislative reform would be required to introduce the broader ‘moral wrong’ test. (c) Reform Two major issues concerning defences of mental disorder emerge from the above discussion: the question of definition and the question of protective measures. In the past, the latter has often driven the former in that the definition has been expanded to include persons against whom compulsory measures are thought to be necessary, whether or not they would be regarded by experts as suffering from a serious mental disorder. The 1991 Act altered the balance somewhat, since committal to a mental hospital is now only a possible and not an inevitable consequence of a special verdict of not guilty by reason of insanity. But the label ‘insane’ remains, and it is manifestly unsuitable for those whose behaviour stemmed from epilepsy, somnambulism, or diabetes.40 Defence lawyers will rightly challenge aspects of the insanity doctrine under the Human Rights Act.41 Article 5.1(e) of the European Convention allows that ‘persons of unsound mind’ may lawfully be deprived of their liberty, but the leading decision in Winterwerp v Netherlands42 lays down three further requirements. First, there must be a close correspondence between expert medical opinion and the relevant definition of mental disorder: that can hardly be said of a test formulated in 1843 and subsequently held to encompass epilepsy, hyperglycaemia, and sleepwalking.43 Secondly, the court's decision must be based on ‘objective medical expertise’, a requirement that could be used in conjunction with the 1991 Act to hold that

Page 6 of 57

Criminal Capacity, Mens Rea, and Fault psychiatric reports to the court should be accorded more weight than under the restrictive M'Naghten test.44 Thirdly, the court must decide that the mental disorder is ‘of a kind or degree warranting (p. 145) compulsory confinement’, and until the law was changed in 2004 the court had no opportunity to make such a determination in murder cases.45 Although arguments based on Art. 5 only have purchase at the stage where D is deprived of liberty, they are relevant in many cases and it would be best if the defence of insanity itself were reformed sensibly before piecemeal challenges are mounted under the Human Rights Act. The M'Naghten Rules are widely recognized to be outmoded. They refer only to mental disorders which affect the cognitive faculties, i.e. knowledge of what one is doing, or of its wrongness, whereas some forms of mental disorder impair practical reasoning and the power of control over actions. This is now recognized in the ‘diminished responsibility’ doctrine in manslaughter,46 which includes cases of ‘irresistible impulse’, and it should clearly be recognized as part of a reformed mental disorder defence. The Model Penal Code accomplishes this by referring to mental disorders which result in D lacking ‘substantial capacity either to appreciate the wrongfulness of his conduct or to conform his conduct to the requirements of the law’.47 The Butler Committee proposed to take this into account in a different way—by ensuring that one ground for a mental-disorder verdict is that, at the time of the alleged offence, D was suffering severe mental illness or handicap.48 In other words, if the mental disorder was severe in degree, there should be no need to establish that it affected D's cognition: so long as the court is satisfied that the conduct was attributable to that disorder, the special verdict should be returned. It therefore includes both cognitive and volitional deficiencies, and places the insanity verdict more squarely on the ground of incapacity.49 In doing so, however, it takes a somewhat static view of mental disorder, confining it more or less to the major psychoses.50 It fails to recognize the variety of mental disorders, and the fact that some of them may substantially impair the patient's practical reasoning even though the diagnosis contains some contestable elements that not all experts agree on. Psychiatry has been attacked for offering diagnoses when these inevitably contestable elements are in issue, but the proper response is to recognize and discuss the contestible elements rather than to deny that they should influence criminal liability in any circ*mstances.51 Only to a small extent is this conservative approach to mental disorder mitigated by the second limb of the Butler proposals, also to be found in a revised form in the Draft (p. 146) Criminal Code.52 This provides for evidence of mental disorder to be adduced to show that D lacked the mental element for the crime. The Law Commission, unlike the Butler Committee, would limit the type of mental disorder that may be relied upon here to ‘severe mental illness’ and ‘incomplete development of mind’. The Commission cited the danger of allowing too wide a definition, which would sweep in too many defendants.53 However, the proposed definition is framed so as to include cases of ‘pathological automatism that is liable to recur’, and again classifies diabetes and epilepsy within mental disorder for reasons of social protection.54 This is both contrary to the principle of fair labelling55 and in violation of the European Convention, and should be abandoned. A separate form of defence should be devised for this group of conditions. Of course this leads to the problem of drawing a definitional line between ‘insanity’ and ‘automatism’, and it was the difficulty of doing so that led the Law Commission to bring these cases within the mental disorder defence, believing that this would be less ‘offensive’ and ‘preposterous’ than the insanity label.56 Even if that answers the fair labelling argument, it leaves the European Convention challenge unaffected unless it is provided that no person with those conditions shall be deprived of liberty—and that, again, would require a separate

Page 7 of 57

Criminal Capacity, Mens Rea, and Fault definition.

5.3 Corporate liability (a) Natural and corporate personality Most discussion of criminal liability is concerned with individual defendants as authors of acts or omissions, raising questions of respect for the autonomy of individuals. We saw in Chapters 2.1 and 2.2 that a developed notion of autonomy is not solely about negative liberty, i.e. protecting individuals from harm, but also involves elements of positive liberty or welfare, i.e. providing facilities and social arrangements whereby individuals can exercise autonomy more fully. By providing a framework for individuals to form companies and corporations, the legal system contributes to this end. Corporate activities now play a major part in individual and commercial life—through companies as employers, as providers of goods and services, as providers of transport and of recreational facilities, and so forth. The criminal law has made increasing inroads into these spheres in recent years. The courts have developed doctrines of vicarious and corporate liability, and the legislature had introduced new offences directed specifically at corporate activities in the financial and commercial sphere (e.g. Financial Services and Markets Act 2000, and the Companies Acts 1985–9). Yet historically the criminal law has developed around the notion of individual human (p. 147) beings as the bearers of rights and duties. It is still somewhat trapped in that framework, even though the idea of companies as separate legal entities from their shareholders and their management was established in the nineteenth century. A limited liability company has long been treated as a separate legal entity from the person or persons who control it.57 The present theory, then, is that corporate personality attaches to companies just as natural personality attaches to individuals (with certain modifications). But does this theory, which has a secure grip in company law, mean that companies can be convicted of offences? The courts moved slowly in this direction in the mid-nineteenth century. Although still doubting whether companies could be said to do ‘acts’, the courts overcame any reluctance to hold companies liable for failing to act58 and for committing a public nuisance.59 The driving force behind these innovative decisions, both concerning railway companies in the early days of rail travel, was not legal theory but social protection: ‘there can be no effective means of deterring from an oppressive exercise of power, for the purpose of gain, except the remedy by an indictment against those who truly commit it, that is, the corporation acting by its majority’.60 And from there the law developed towards criminal liability for companies, acting through their controlling officers.61 (b) Towards corporate criminal liability This subject was given a pressing social importance in the late 1980s by the series of disasters connected with corporate activities and involving considerable loss of life—for example, the Piper Alpha oil rig explosion, the Clapham rail disaster, the King's Cross fire, the sinking of the Marchioness, and in 1987 the capsize of the ferry Herald of Free Enterprise. It does not make sense to present each of these, and the string of subsequent transportation disasters, as the responsibility of a few individuals. Indeed, enquiries into the disasters have tended to emphasize the role of deficiencies in the systems of corporate management and accountability. Major disasters apart, a variety of organizations offer evidence of a constant Page 8 of 57

Criminal Capacity, Mens Rea, and Fault stream of incidents of industrial pollution, unsafe working conditions, impure foods, and unfair business practices which impinge upon, or threaten to impinge upon, the lives of individual citizens. Growing recognition of the significance of corporate harm-doing has not, however, been accompanied by substantial alteration of the framework of criminal liability. The trend, as we shall see, has been to attempt to fit corporate liability into the existing structure rather than to consider its implications afresh. And, more important in social terms, there has been little change of approach at the level of enforcement. (p. 148) It is one thing to have a set of laws which penalizes corporate wrongdoing as well as individual wrongdoing. It is quite another thing to have a balanced machinery of enforcement which strives to ensure the proportionate treatment of individuals and companies according to the relative seriousness of their offences. Present arrangements seem to draw a strong line between frequent police action against individuals and the relatively infrequent action of the various inspectorates, government departments, etc. against companies.62 However, the social calculation cannot be presented simply as an imbalance in treatment between ‘crime in the streets’ and ‘crime in the suites’. We must also take into account the finding of social surveys that it is street crimes that cause real harm and fear to people, not least to those who are already among the most disadvantaged in society.63 It is therefore a question for discussion whether devoting large resources to the detection and prosecution of corporate harm-doers would be either defensible or socially acceptable. There are some straightforward applications of the doctrine that a company is a legal person, separate from the individuals involved in its operations. Thus, for example, two principal provisions of the Health and Safety at Work Act 1974 are as follows:

s 2 (1) It shall be the duty of every employer to ensure, so far as is reasonably practicable, the health, safety and welfare at work of all his employees. s 3 (1) It shall be the duty of every employer to conduct his undertaking in such a way as to ensure, so far as is reasonably practicable, that persons not in his employment who may be affected thereby are not thereby exposed to risks to their health and safety.

These provisions, in conjunction with s. 33 of the Act (which creates an offence of failing to discharge either duty), are clearly directed at companies no less than at individual employers.64 Thus in British Steel plc65 the company was convicted of failing to discharge its duty under s. 3 of the Act. During an operation to re-locate a steel platform, under the supervision of a British Steel employee, an unsafe method of working led to the collapse of the platform and a sub-contracted worker was killed. The Court of Appeal upheld the conviction, on the basis that it was the employer on whom the duty was imposed, and the duty had clearly not been discharged. Similar reasoning can be used to hold companies liable for a whole range of offences of strict liability: just as much as any individual, companies can cause pollution, sell goods, fail to submit annual returns, etc. An offence of strict liability is one which requires no fault for conviction: any person may be found guilty simply through doing or failing to do a (p. 149) certain act.66 Thus, if a company owns the business or premises concerned,

Page 9 of 57

Criminal Capacity, Mens Rea, and Fault it may be convicted of failing to control emissions of pollutants, or for causing polluting matter to enter a stream, whether or not these events come about through fault on the company's part.67 (i) The Possibility of Vicarious Liability: Outside the criminal law there have been further developments, and the law of torts has established a doctrine of vicarious liability of employers for the conduct of their employees.68 By contrast with Federal law in the USA,69 there is no such general doctrine in the criminal law, but various exceptions and quasi-exceptions are gaining a foothold. One is the ‘delegation principle’: where a statute imposes liability on the owner, licensee, or keeper of premises or other property, the courts will make that person vicariously liable for the conduct of anyone to whom management of the premises has been delegated.70 This applies whether the defendant is an individual or a company. The underlying reason for this principle seems to lie in the assumption that such offences would otherwise be unenforceable, since delegation would remove responsibility from the person in effective control.71 The second exception revolves around the interpretation of such key words in statutes as ‘sell’, ‘use’, and ‘possess’. The clearest example is where a statute prohibits the selling of goods in certain circ*mstances. Coppen v Moore (No. 2) (1898)72 held the shop owner liable as the person who sold the goods in law, even though he was away from the shop at the time and an assistant carried out the transaction—in breach of the instructions left by the owner. So long as the assistant is acting as an agent rather than as a private individual, ‘vicarious’ liability is imposed.73 In effect, a similar result flowed from the application of s. 3 of the Health and Safety at Work Act 1974 in the British Steel case, since the company was held liable for the inadequate supervision by its own employee.74 A third exception or quasi-exception arises where a statutory offence penalizes conduct that appears to require a personal act, such as ‘using’ a motor (p. 150) vehicle with defective brakes: in James & Son v Smee75 the Divisional Court held the company liable, on the basis that the use of the vehicle by an employee in the course of employment constituted use by the employer. These examples of vicarious liability in the criminal law may appear not to respect the principle of individual autonomy, in so far as they hold people (or a company) liable for something that was not their own voluntary act or omission. They can be justified only on the principle of welfare and, even then, they should be made to respect ‘rule of law’ values by ensuring fair warning of the standards expected. Fair warning is not assured if decisions are made by way of statutory interpretation in the courts, rather than clearly by the legislature. (ii) The Identification Principle: We now return to corporate liability as such. Most of the instances discussed so far concern offences of strict liability, where it is often easier to construe a statute so as to impose direct liability on a company. In 1944 the courts began to develop a new doctrine which imposes liability on companies for offences requiring a mental element. In DPP v Kent and Sussex Contractors Ltd76 the defendant company was charged with two offences—making a statement which was known to be false, and using a false document with intent to deceive. The Divisional Court held that the company could be convicted of both offences, on the basis that its officers possessed the required ‘knowledge’ and ‘intent to deceive’, and that those states of mind could therefore be imputed to the company itself. As Viscount Caldecote CJ held, ‘a

Page 10 of 57

Criminal Capacity, Mens Rea, and Fault company is incapable of acting or speaking or even thinking except in so far as its officers have acted, spoken or thought’. Thus the company is identified with those officers who are its ‘directing mind and will’ for these purposes.77 This identification principle was applied in the leading case of Tesco Supermarkets v Nattrass (1971),78 where the company had been convicted of offering to sell goods at a higher price than indicated, contrary to the Trade Descriptions Act 1968. A shop assistant at a local Tesco store had failed to follow the manager's instructions, with the result that the goods were offered at a higher price than advertised. The House of Lords quashed the conviction, holding that the manager of one of the company's supermarkets was not sufficiently high up in the organization to ‘represent the directing mind and will of the company’. The identification principle is therefore fairly narrow in its scope. It allows large companies to disassociate themselves from the conduct of their local managers, and thus to avoid criminal liability. Moreover, where a large national or multinational company is prosecuted, the identification principle requires the prosecution to establish that one of the directors or top managers had the required knowledge or culpability. Managers at such a high level tend to focus on broader policy issues, not working practices. Thus it may be considerably easier to achieve convictions in respect of the (p. 151) activities of small companies than of large corporations, because there will tend to be more ‘hands-on’ management in small companies.79 The seeds for an expansion of the Tesco v Nattrass test have been sown by Lord Hoffmann, speaking for the Privy Council in Meridian Global Funds Management Asia Ltd v Securities Commission (1996).80 Courts should be prepared to go beyond the people who represent the ‘directing mind and will’ of a company, and to enquire, in the context of the particular offence, ‘whose act (or knowledge, or state of mind) was for this purpose intended to count as the act, etc., of the company?’ The reach of this extension is unclear, since much turns on the statutory context. In this case it enabled the conviction of the company on the basis of the knowledge of two investment managers that they were making unlawful investments. When the Court of Appeal was invited to extend the identification principle along these lines in AttorneyGeneral's Reference (No. 2 of 1999),81 a case arising from the Southall train crash, it declined to do so. Rose LJ held that:

The identification theory, attributing to the company the mind and will of senior directors and managers, was developed in order to avoid injustice: it would bring the law into disrepute if every act and state of mind of an individual employee was attributed to a company which was entirely blameless.82

The terms of this statement beg a number of questions, but it was clear from the judgment that the Court of Appeal thought that any significant extension of the judge-made identification principle should be left to Parliament. (c) Individualism and corporatism The history of legal developments in this sphere suggests a somewhat slow progress towards integrating corporations into a legal framework constructed for individuals, with few gestures towards the differences between corporations and individual human beings.83 There are those who argue that social phenomena can only be interpreted through the actions and motivations 84 Page 11 of 57

Criminal Capacity, Mens Rea, and Fault of individuals, and abstractions like corporations constitute barriers to proper understanding.84 Only individuals can do things, and so the law is right to concentrate its attentions upon them. Indeed, any other view might threaten the principle of individual autonomy by holding people liable when they did no voluntary act. (p. 152) The weakness of this argument is that individual actions can often be explained fully only by reference to the social and structural context in which they were carried out. When the managing director of a company is announcing a commercial strategy, he or she is acting not merely as an individual but also as an officer of the company. Without reference to the structure and policies of the company and to that person's role within it, there can be no proper explanation of what was said and done. The argument, therefore, is that the behaviour of individuals is often shaped by their relationship to groups and collectivities—‘shaped’ in a meaningful sense, not ‘determined’ in the sense that individual autonomy is lost in the process (since individuals normally have some liberty to disengage themselves from the corporation). The thrust is that companies often have a policy structure and a dynamic of their own which to some extent transcend the actions of their individual officers. Perhaps the clearest application of this can be found in offences of omission, particularly those involving strict liability. In a case like Alphacell Ltd v Woodward (1972),85 where polluting matter escaped from the company's premises into a river, it seems both fairer and more accurate to convict the company rather than to label one individual as the offender: where the law imposes a duty, the company should be organized so as to ensure that the duty is fulfilled. None of this is meant to suggest that individuals within a corporation should not bear personal responsibility for their conduct. In appropriate cases they should do so, provided that they had fair warning of any special duties attached to the activities of the company.86 The important point is that companies should be open to both criminal and civil liability, since it is they who create the structural context for the individual's conduct qua company officer. The corporation appoints the individual and sustains him in this position—the individual is in that place, doing that thing, because of the corporation—and so it is right that the corporation should bear primary liability, or at least concurrent liability with its officer. This does not mean that legality and ‘rule of law’ principles should be neglected: companies are run by individuals, who ought to receive fair warning of their duties. All these arguments may need adjusting for small, even one-person, companies and also for non-profit organizations. Moreover, they leave open the question whether the criminal law in its traditional form is the most appropriate means of dealing with corporate harm-doing. (d) Changing the basis of corporate liability The theoretical arguments in favour of corporate criminal liability seem strong, but developments at common law have been slow. The ‘identification principle’ in Tesco Supermarkets v Nattrass87 has a relatively narrow sphere of operation, and there has (p. 153) been little judicial enthusiasm for the greater flexibility proposed in the Meridian case.88 A small number of statutes impose direct or vicarious liability on companies, the Health and Safety at Work Act 1974 being an example, but there is no such general approach. An alternative strategy of placing the emphasis on individual liability would be unlikely to work with larger companies: any particular individual might be dispensable within a corporation (e.g. the ‘Company Vice-President responsible for going to gaol’), allowing the company to continue on its course with minimal disruption. It might also be difficult to identify the individual responsible,

Page 12 of 57

Criminal Capacity, Mens Rea, and Fault not least because the lines of accountability within companies are sometimes unclear, although with smaller companies, which are far more numerous, this might be workable. A number of different approaches have been canvassed in recent years. Celia Wells, whilst emphasizing the need to re-assess the socio-political role of corporations and of individuals within them, lends support to an approach that depends partly on a version of aggregation, whereby a company's culpability should be constructed out of the knowledge and the attitudes of employees as a whole,89 and partly on the need to ensure that individual company officers are prosecuted where appropriate. Another, complementary approach would be to use company policies, or their absence, as the basis for liability. This follows the approach of Brent Fisse and John Braithwaite, and in particular their concept of ‘reactive fault’.90 In their view, rather than expending prosecutorial energy and court time trying to disentangle the often convoluted internal structures and policies of corporations, the law should require a company which has caused or threatened a proscribed harm to take its own disciplinary and rectificatory measures. A court would then assess the adequacy of the measures taken. The concept of fault would thus be a post hoc phenomenon. Rather than struggling to establish some antecedent fault within the corporation, the prosecution would invite the court to infer fault from the nature and effectiveness of the company's remedial measures after it has been established that it was the author of a harm-causing or harm-threatening act or omission. The court would not find fault if it was persuaded that the company had taken realistic measures to prevent a recurrence, had ensured compensation to any victims, and had taken the event seriously in other respects. The whole orientation of the system would be different: every death caused, whether purely accidental or not, would be treated as potentially a serious offence until the company established otherwise.91 Following a lengthy process of discussion and negotiation, a new form of corporate liability has now been introduced in the limited (but high-profile) area of homicide.92 The Corporate Manslaughter and Corporate Homicide Act 2007 introduces a new offence of corporate manslaughter, which can be committed only by ‘organizations’ and not by individuals. The legislative framework of the new offence is highly (p. 154) technical, but for present purposes we can focus on the mechanism by which liability is imposed. Section 1 of the 2007 Act provides:

An organisation…is guilty of an offence if the way in which its activities are managed or organised— • causes a person's death, and • amounts to a gross breach of a relevant duty of care owed by the organisation to the deceased.

Three features of the new Act's approach stand out. First, it applies to ‘organizations’, which include companies, partnerships, and various associations and government departments. This is controversial, but is not pursued here. Secondly, the offence is only committed if ‘the way in which its activities are managed and organized by its senior management is a substantial element in the breach’ of duty. This focus on ‘senior management’ is developed by s. 1(4), which provides that senior managers must be persons who ‘play significant roles’ in either

Page 13 of 57

Criminal Capacity, Mens Rea, and Fault decision-making or managing the whole or a substantial part of the organization's activities. This shows that the model of corporate liability adopted in the 2007 Act has not strayed far from the ‘identification principle’ that has developed at common law. It is, to be blunt, doubtful if the manager of a large Tesco superstore plays a significant role in managing a substantial part of Tesco's activities. If the company is to be prosecuted under the 2007 Act, it must be shown that people at a higher level in a large organization organized the relevant activities in such a way as to amount to a substantial element in the breach. Thirdly, however, s. 8 of the Act directs the jury to consider any alleged breach of health and safety legislation, and permits the jury to take account of evidence of ‘attitudes, policies, systems or accepted practices within the organization that were likely to have encouraged’ any failure to meet safety standards. This suggests that evidence of what is sometimes called ‘corporate culture’ may be determinative in some cases. Whether this new approach will be thought to achieve justice in homicide cases remains to be tested. Certainly it seems to be a rather narrow approach to serve as a model for corporate criminal liability generally, and so the quest for a fairer set of principles must go on. The new approach also serves to raise the wider question of whether it is the conviction of organizations that is the most important aspect, or whether the sentencing of organizations should be regarded as important too. A company can hardly be imprisoned, moderate fines can be swallowed up as business overheads, and swingeing fines may have such drastic sideeffects on the employment and livelihoods of innocent employees as to render them inappropriate. Fisse and Braithwaite have proposed a range of special penalties, some of which are rehabilitative (putting corporations on probation to supervise their compliance with the law), some of which are deterrent (punitive injunctions to require resources to be devoted to the development of new preventive measures), and others of which have mixed aims (e.g. community service by companies).93 In their view, the primary search should be for a regime which (p. 155) ensures maximum prevention. The 2007 Act provides for three types of sentence—publicity orders (requiring the organization to make it known that it has been convicted of this offence); remedial orders (requiring the offender to remedy the causes of the homicide); and fines (which may prove problematic for the reasons given above, and which are questionable in so far as they may have deleterious effects on the level of public service provided by organizations such as hospital trusts and the police).

5.4 Fault and mens rea: general principles (a) Choice and the subjective principles The principle of mens rea has already been outlined in Chapter 3.6(o), together with the related principles of correspondence (Chapter 3.6(q)) and of fair labelling (Chapter 3.6(s)). The essence of the principle of mens rea is that criminal liability should be imposed only on persons who are sufficiently aware of what they are doing, and of the consequences it may have, that they can fairly be said to have chosen the behaviour and its consequences. This approach is grounded in the principle of autonomy (Chapter 2.1): individuals are regarded as autonomous persons with a general capacity to choose among alternative courses of behaviour, and respect for their autonomy means holding them liable only on the basis of their choices.94 The principle of mens rea may also be claimed to enhance the constitutional values of legality and rule of law, by reassuring citizens that they will be liable to conviction,

Page 14 of 57

Criminal Capacity, Mens Rea, and Fault and to the exercise of state coercion against them, only if they intentionally or knowingly cause or risk causing a prohibited harm. If this were achieved, the criminal law would ensure that ‘each person is guaranteed a greatest liberty, capacity and opportunity of controlling and predicting the consequences of his or her actions compatible with a like liberty, capacity and opportunity for all’.95 What this liberal view rejects is an approach which holds people criminally liable solely on the ground that liability and punishment would have a general deterrent effect in preventing further harms. That approach, associated with utilitarian theories,96 looks to the probable social effects of liability and punishment, denying the individual defendant any special status in the matter: if the punishment of people in D's position would have an overall deterrent effect, then D should be punished, even though he or she cannot be said to have chosen to cause the harm. Deterrent theories therefore tend to give priority to social protection. Theories of punishment in the liberal tradition may recognize the relevance of social protection, at least at the level of justifying the criminal law itself and justifying the criminalization of certain conduct, but at the level of individual liability to conviction and censure (p. 156) they insist that respect for the principle of individual autonomy has superior value to general calculations of social utility. (This means that they must either denounce strict liability offences, or find some plausible argument in their favour.) The principle of mens rea also encompasses the belief principle, which holds that criminal liability should be based on what defendants believed they were doing or risking, not on actual facts which were not known to them at the time. Also flowing from the principle of mens rea, as we saw in Chapter 3.6(q), is the principle of correspondence, which insists that the fault element for a crime should correspond to the conduct element specified for the crime. Thus, if the conduct element is ‘causing serious injury’, then the fault element ought to be ‘intention or recklessness as to causing serious injury’; a lesser fault element, such as ‘intention or recklessness as to a mere assault’, would breach the principle of correspondence. This makes the point that the notion of choice is not an abstract phenomenon, but should in principle be linked to the circ*mstances or consequences specified in the definition of each crime. Do the notions of fault and choice that underlie the principle of mens rea have a wider application? While the principle of mens rea supports only criminal liability for intention, knowledge, and (subjective) recklessness, there are serious questions about whether gross negligence, or even negligence, can be said to involve sufficient fault and choice to justify the imposition of criminal liability. English law contains several offences of negligence, whereas the tendency of commentators has been to regard them as aberrant and calling for special justification. This discussion will be taken further in paragraphs 5(f) and 5(g). (b) Constructive liability and ‘moral luck’ We have already seen, in Chapter 3.6(q), that subjectivists tend to place high value not only on the principle of mens rea and on the belief principle (that D should be judged on the facts as he believed them to be), but also on the principle of correspondence (that in relation to each conduct element of an offence, the fault requirement should be at the same level). We also noted, in Chapter 3.6(p), that this is disputed by advocates of constructive liability, arguing that once D has crossed a significant moral threshold he should be held liable for whatever consequences follow. We will see in later chapters that among the examples of constructive liability are manslaughter by unlawful act (Chapter 7.5) and unlawful wounding (Chapter 8.3). Offences of this kind allow what is termed ‘moral luck’ to play a significant role in

Page 15 of 57

Criminal Capacity, Mens Rea, and Fault determining the level of criminal liability. As elaborated in Chapter 3.6(q) and (r), subjectivists tend to oppose the intrusion of moral luck into criminal liability. The argument is that the criminal law should blame people for what they intended or foresaw and for what lay within their control: it should draw a straight line through the vicissitudes of life and the vagaries of fortune when determining the extent of criminal liability.97 A different view is taken (p. 157) by moderate constructivists.98 Like subjectivists, moderate constructivists accept that the criminal law should respect the rule of law and should maintain ‘clear and certain offence definitions, good publicity, and conformity between announced rule and adjudicative standard’.99 However, moderate constructivists reject the principle of correspondence between the level of the conduct element and the level of the fault or mens rea, and instead argue that, so long as D intentionally commits a relevant offence, that fault is sufficient to justify conviction for a more serious offence in the same family if an unanticipated and more serious consequence results.100 This is because:

By committing an assault one changes one's normative position, so that certain adverse consequences and circ*mstances that would not have counted against one but for one's original assault now count against one automatically, and add to one's crime.101

This is John Gardner's statement of the doctrine of ‘change of normative position:’ as he accepts, to assert that such a change takes place when D intentionally commits a relevant offence does not supply a justification for imposing moderately constructive liability on D.102 More work needs to be done if the ‘change of normative position’ argument is to be a convincing rationale for moderate constructivism. In the meantime, as argued in Chapter 3.6(r), the criminal law should give precedence to the principle of correspondence. (c) The principle of contemporaneity As we saw in Chapter 3.6(u), the principle of contemporaneity states that the fault element must coincide in point of time with the conduct element in order to amount to an offence. This forms part of the ideology that the function of the criminal law is not to judge a person's general character or behaviour over a period of time; its concern is only with the distinct criminal conduct charged. According to this view, whether or not criminal conviction is deserved depends on D's conduct and mental attitude at the relevant time. But this narrow statement of the principle, if indeed it ever represented a complete statement of the law,103 has been progressively abandoned in the face of intuitions to the contrary exemplified in leading cases. In the famous case of fa*gan v (p. 158) Metropolitan Police Commissioner (1969)104 D accidentally drove his car on to a policeman's foot, and then deliberately left it there for a minute or so. The defence to a charge of assault was that the conduct element (applying force) had finished before the fault element began; the act and the intent never coincided. The Divisional Court held that D's conduct in driving the car on to the foot and leaving it there should be viewed as a continuing act, so that the crime was committed when the fault element (D's realization of what had happened and decision to leave the car there) came together with the continuing conduct. This is not the only occasion on which the courts have invoked the notion of a ‘continuing act’ to expand the timeframe of a crime and thus the application of the principle of contemporaneity.105 However, a different approach was taken by the House of Lords in Miller (1982),106 the case in which a squatter was smoking in bed,

Page 16 of 57

Criminal Capacity, Mens Rea, and Fault accidentally set the mattress on fire, but simply moved to another room without attempting to remedy the problem, with the result that the house caught fire. Lord Diplock did not reject the view that the fire was a continuing act that began accidentally but could then be connected with D's fault when he realized that the mattress was on fire. However, his Lordship expressed a preference for the ‘duty’ analysis, whereby the accidental creation of danger gave rise to a duty (a continuing duty to avert the danger caused) which in this case D knowingly failed to discharge. The continuing act approach seems to exert an influence in another area. In Thabo Meli v R (1954)107 the plan was to kill V in a hut and then throw his body over a cliff: this was what D believed he was doing, but in fact V died from the fall down the cliff and not from the beating in the hut. The argument for the appellant was based on the lack of contemporaneity (this time it was intent first, death later), but the Privy Council rejected this, holding that the beating and the disposal over the cliff formed part of a planned series of acts which should be regarded as a single course of conduct. On this analysis, the presence of the fault element at any stage during the planned sequence would suffice. That reasoning was extended in Church (1966)108 to cover a series of acts which had not been planned but which simply followed one after the other. Subsequently, in Le Brun (1991)109 the Court of Appeal had to deal with a case in which D had assaulted his wife, and then when he tried to move her unconscious body dropped her, causing her to suffer a fractured skull from which she died. The Court held that the conduct and the fault elements ‘need not coincide in point of time’ so long as they formed part of a ‘sequence of events’, particularly in a case such as this where D's later acts were attempts to conceal his initial offence. All these cases could have resulted in convictions for other offences (attempted murder in Thabo Meli, grievous or actual bodily harm in the last two cases), but the courts apparently took the view that since the consequence—death—resulted (p. 159) from D's original culpable conduct, homicide convictions ought to be registered. A similar analysis would be possible in non-homicide cases. The decisions therefore take a rather elastic view of the contemporaneity principle, and seem to be motivated by considerations akin to constructive liability.110 It is convenient to deal here with one more awkward situation relating to the link between conduct and fault. In Attorney-General's Reference (No. 4 of 1980) (1981),111 it appeared that D was arguing with his female partner at the top of a flight of stairs, that he pushed her away and she fell backwards down the stairs, that he concluded she was dead, and then dragged her back to their flat with a rope around her neck and cut up her body. The Court of Appeal held that there could be a conviction on these facts, even though it was not clear which of D's acts caused death. So long as the jury was satisfied that D had sufficient fault for manslaughter when he pushed her backwards, and sufficient fault for manslaughter when he cut up her body,112 it was immaterial which act caused death. The facts of this case are somewhat stronger than the facts of Thabo Meli, Church, and Le Brun, since in all of those cases it was clear that it was not D's initial act that caused death. Surely in the Reference case it should have been possible to convict D if the court was satisfied that there was a sequence of events and that D had the required fault element at some stage; the actual facts, however, were taken not to raise this point. (d) The doctrine of prior fault We saw in Chapter 3.6(v) that the principle of contemporaneity conflicts in certain situations

Page 17 of 57

Criminal Capacity, Mens Rea, and Fault with the doctrine of prior fault—the principle that a person should not be allowed to take advantage of any defence or partial defence to criminal liability if the relevant condition or circ*mstances were brought about by his or her own fault. Whilst the contemporaneity principle insists that the criminal law is concerned with the prohibited event itself, not with its antecedents or its sequels, the doctrine of prior fault points to circ*mstances in which the antecedents of the event ought (by way of principled exception) to affect a proper moral evaluation of D's conduct. Two examples of the doctrine's operation may be given. First, a person who deliberately drinks to excess in order to stoke up the courage to do a certain act will not be allowed to rely on that intoxication by way of defence because it arose from prior fault.113 Secondly, if D taunts another in the hope of inducing the other to attack him, D should not be able to rely on loss of self-control or self-defence as a defence to a charge of murder, because the attack on D will be regarded as self-induced.114 Examples of the doctrine of prior (p. 160) fault in operation were noted in Chapter 4, in relation to automatism and selfdefence, and will be seen in abundance in Chapter 6 (on intoxication, duress, necessity, etc.). One remaining question concerns the amount of ‘fault’ required for the doctrine to take effect. A study by Paul Robinson revealed considerable diversity of provisions in the Model Penal Code and in American laws generally,115 and a similar diversity appears in England.116 Should any causal contribution by D make a defence unavailable, or should it be a lack of proper care (for example, drinking alcohol when its possible effects are widely known,117 joining a gang which is known to use violence118 ), or should the doctrine require proof that D foresaw the possibility that certain conduct might follow? The differences between these approaches ought not to be regarded as unimportant, since the withdrawal of a defence simply on the grounds of some small amount of fault on D's part is equivalent to a principle of constructive liability for offences. One way of avoiding this difficulty would be to devise a range of offences to cover ‘faulty’ acts (e.g. excessive consumption of alcohol), and then convict D of an offence of that kind—whilst not removing any defence to the substantive crime which might otherwise be open.119 This would introduce further complexities into the law, but at least it attempts a fair solution to a difficult problem.

5.5 Varieties of fault Having introduced the subjective principles and some problems of contemporaneity of conduct and fault, we now move to the core fault elements. First to be considered is strict liability, for which there may be little or no fault at all. One reason for considering these offences first is that they are the most numerous, a fact that belies the prominence often given to intention and recklessness in the rhetoric of English criminal law. We then turn to the mens rea terms of intention, recklessness, and knowledge, before exploring the little-used concept of negligence. (a) Strict liability120 There is no clear convention about when criminal liability should be classified as ‘strict’. We will use the term here to indicate those offences of which a person may be convicted without proof of intention, knowledge, recklessness, or negligence. Some offences prescribe liability without fault but allow the defendant to avoid liability on proof of ‘due diligence’. There is dispute about whether offences with such provisos (p. 161) are properly termed ‘strict liability’ offences,121 but for our present purposes they will be included within the concept of

Page 18 of 57

Criminal Capacity, Mens Rea, and Fault strict liability. This corresponds with the Canadian approach, which separates strict liability (where a defendant can avoid liability by establishing that there was no negligence) from absolute liability (where the only defences available relate to fundamental elements of capacity or necessity).122 The term ‘absolute liability’ has its own difficulties, however, since one can argue that liability should only be described as absolute where there is no defence available at all to someone who is proved to have caused the prohibited event. What this shows, above all, is that there is no settled terminology to give simple expression to the numerous permutations of conditions for liability. If one takes account of the device of shifting the burden of proof on to the defendant, then the permutations range from requiring mens rea—with the burden of proof on the prosecution—to defining special defences or provisos with an evidential burden on D, defining special defences or provisos with a legal burden of proof on D, requiring proof of negligence by the prosecution, creating a no-negligence defence to be proved by D, imposing liability with no due diligence defence at all, and even to a dispensation from proving an element of the offence.123 (i) For and Against Strict Liability: Let us leave aside the complexities introduced by changes in the burden of proof, and formulate a central question: what are the arguments for imposing criminal liability with no due diligence defence available? The main argument is a form of protectionism or ‘social defence’. It maintains that one of the primary aims of the criminal law is the protection of fundamental social interests. Why should this function be abandoned when the violation of those interests resulted from some accident or mistake by D? Surely, Wootton argued, ‘mens rea has got into the wrong place’: it should be relevant not to the actual conviction, but to the appropriate means of dealing with the offender after conviction. ‘If the object of the criminal law is to prevent the occurrence of socially damaging actions, it would be absurd to turn a blind eye to those which were due to carelessness, negligence or even accident. The question of motivation is in the first instance irrelevant.’124 At a time when victims’ interests are receiving greater recognition, arguments of this kind may find considerable support. The infliction of the prohibited harm would become the trigger for state action, aimed at minimizing the risk of the harm being repeated. The strength of the argument lies in its concern for the welfare of citizens in general. Its weakness is to suggest that this is a justification for using the criminal law in (p. 162) this way. There are two major questions to be answered here: would it be fair? Would it be effective? The fairness issue is one which runs through this chapter and, indeed, through the whole book. The criminal law is society's most condemnatory instrument, and, as argued in Chapter 2.1, Chapter 3.6, and section 5.4(a), respect for individual autonomy requires that criminal liability be imposed only where there has been choice by D. A person should not be censured for wrongdoing without proof of choice (as distinct, perhaps, from being held civilly liable). This is a fundamental requirement of fairness to defendants. Indeed, it is not only unfair to censure people who are not culpable, but also unfair to punish them for the offence. Moreover, in so far as the criminal trial has a communicative function, strict liability impairs this by severely limiting D's ability to explain, excuse, or justify the conduct and by requiring a conviction in all but exceptional circ*mstances.125 Opponents of subjectivism may dismiss this as a mere matter of convention—and outmoded convention at that. The criminal law should simply be regarded as an efficient social resource 126 Page 19 of 57

Criminal Capacity, Mens Rea, and Fault for the prevention of harm,126 with conviction carrying no special moral connotations of ‘guilt’ or ‘blame’. Is there not something incongruous, in a world in which avoidable deaths and injuries are much too frequent and cause much grief and insecurity, for the State meticulously to observe the ‘intent’ and ‘belief’ principles, the presumption of innocence, and other fairness principles so as to facilitate the acquittal of clumsy, ignorant, but nevertheless dangerous people?127 One answer to this challenge is to reassert that the prevention of harm is neither the sole nor the overriding aim of the criminal law, and that the criminal law is not the only official means of preventing harm. Even Bentham, whose general approach was to transcend individual considerations and to weigh the social benefits against the social disadvantages of criminal liability, argued that criminal punishment is an evil which should be reserved for the worst cases, and that legislators should turn first to education, regulation, and civil liability as means of preventing harms.128 The subjective principles reflect the value of individual autonomy, but many of the harms which afflict, or threaten to afflict, citizens today are the result of the acts or omissions of corporations. Pollution, defective products, food and drugs, safety at work, transport systems —all these sources of danger are dominated by corporate undertakings. We saw in section 3 of this chapter that the traditional doctrines of the criminal law have various shortcomings when applied to corporate decision-making and responsibility. Once a secure basis for corporate liability is found, the next question would concern the appropriate conditions of liability for companies. Some corporations operate in spheres of such potential social danger, and wield such power (p. 163) (in terms of economic resources and influence), that there is no social unfairness in holding them to higher standards than individuals when it comes to criminal liability—so long as fair warning is given, since companies are run by individuals. This is particularly so when companies operate in spheres where public safety may be at risk. However, the same cannot generally be said of individuals, save in exceptional categories such as road traffic offences, where a licence to drive is required and safety is a central issue. It can therefore be argued that the conflict between social welfare and fairness to defendants should be resolved differently according to whether the defendant is a private individual or a large corporation. On the other hand, those two categories do not exhaust the range of defendants: a large proportion of British businesses have a sole proprietor, and, although their duties may well be more extensive, it can be argued that their criminal liability should follow the model for private individuals.129 That model should reject strict liability for individuals on grounds of unfairness and lack of respect for autonomy: negligence should be the minimum requirement. Moving to the second question, whether criminal liability without fault is a particularly efficacious means of preventing harm, it is important to keep in mind the differences between individual behaviour and corporate activity. At least two aspects of efficacy arise: the ease of enforcing no-fault offences, and the preventive effects of liability without fault. Ease of enforcement may be thought obvious: it must be less trouble to prepare a prosecution in which fault is not relevant than to prepare one in which proof of fault is needed. Indeed, the Court of Appeal has quashed convictions on the basis that evidence of fault is inadmissible, because not relevant, if adduced by the prosecution on a strict liability charge.130 For the more serious offences, however, evidence of fault will be needed at the sentencing stage if the courts are to pass sentence on a proper basis.131 This means that the prosecution will have to prepare some evidence on the point, which in turn diminishes any procedural benefit of strict liability. But there may still be benefits to the prosecutor in not having to prove fault for minor offences,

Page 20 of 57

Criminal Capacity, Mens Rea, and Fault and there may also be indirect benefits as a result of being able to use the threat of prosecution and conviction in order to secure compliance. Many of the regulatory agencies with the power to invoke ‘strict liability’ offences adopt what may be termed a ‘compliance strategy’ towards law enforcement—that is, aiming to secure conformity to the law without the need to process and penalize violators.132 Regulatory activities focus on obtaining compliance, and prosecution is reserved for the few cases where either the violator is recalcitrant or the violation is so large that public concern can only be assuaged by a prosecution. This may also mean that prosecutions tend to be brought only in cases where there is fault: indeed, there are regulatory agencies (p. 164) which pursue such a policy, even though they have no-fault offences at their disposal.133 There is little evidence among the regulatory agencies of a ‘deterrence strategy’, using criminal prosecution as a primary means of preventing breaches of the law. That approach to law enforcement is more typical of the police, who rarely occupy themselves with the so-called regulatory offences aimed at commercial and industrial safety, etc. It is therefore difficult to reach a firm conclusion about the preventive efficacy of strict liability. It is probably an overstatement to regard it as a ‘means of prevention’, since the no-fault offence usually forms one part of a broad regulatory scheme. Some argue that the availability of a no-fault offence strengthens the regulator's hand in ensuring compliance and, therefore, prevention. It enables regulators to use lesser measures, and then to prosecute when there is real fault.134 Others argue that no-fault offences which are followed by low penalties on conviction are almost counterproductive, resulting in the imposition of derisory fines on large organizations. Indeed, if regulation in such spheres as industrial safety had been harnessed to relatively serious offences requiring proof of fault, then those offences might now be taken much more seriously, integrated into people's thinking about offences against the person rather than being regarded as ‘merely regulatory’ and ‘not real crime’.135 This is, of course, part of a much wider issue about the conventional concepts of crime (as now embodied, for example, in the Draft Criminal Code)136 and about conventional approaches to enforcement which regard some offences as police matters and some not. Much turns on the agency through which enforcement takes place, the style of enforcement adopted, and the elements of discretion in choosing and following a style of enforcement. Is it an argument in favour of, or against, strict liability that the offence is a minor one or a grave one? The English courts have used both triviality and gravity as arguments in favour of strict liability. Many offences with low penalties are, or have been held to be, offences requiring no proof of fault.137 This reasoning derives some justification from an economic argument based on ease of prosecution: such trivial offences are not worth the public expenditure of prosecution and court time in proving fault. There is hardly any stigma in being convicted of such offences, and so it is thought to be in the public interest to dispose of them quickly (although the result may be to dilute the (p. 165) moral legitimacy of the criminal law). But none of this can apply to serious offences. Principles of individual fairness, even if overridden by economic considerations in respect of minor offences, should surely be central to the question of conviction for serious offences. One clear benchmark here is the availability of imprisonment as a punishment. The American Model Penal Code proposes that imprisonability should be a conclusive reason against strict liability.138 In Canada the Supreme Court has held that an offence of strict liability which carries the possibility of a custodial sentence is contrary to the Charter of Rights, unless there is a no-negligence defence.139 However, the jurisprudence of the European Court of Human Rights is equivocal on the 140 Page 21 of 57

Criminal Capacity, Mens Rea, and Fault matter.140 (ii) A New Constitutional Principle?: The new millennium brought an apparent change of direction. In B v DPP (2000)141 the House of Lords had to decide whether, in the offence of indecency with a child under 14 contrary to the Indecency with Children Act 1960, there was strict liability as to the age of the child or the prosecution had to establish knowledge of the child's age. The House unanimously held that ‘the common law presumes that, unless Parliament has indicated otherwise, the appropriate mental element is an unexpressed ingredient of every offence’.142 Not only does this decision apply the presumption stated in Sweet v Parsley in preference to the older view that, in sexual offences, it is morally justifiable to impose strict liability as to age,143 but Lord Steyn accepted the description of the presumption of mens rea as a ‘constitutional principle’ that is not easily displaced by a statutory text.144 The same approach was taken by the House of Lords in K (2002),145 where the charge was indecent assault on a girl under 16. The defence was that the girl had told D that she was 16. The question was whether this section of the Sexual Offences Act should continue to be regarded as imposing strict liability as to age, or whether the presumption of mens rea applied. Again, not only did the House of Lords find unanimously in favour of the presumption of mens rea, but both Lord Bingham and Lord Steyn described the presumption as a ‘constitutional principle’.146 (p. 166) (iii) Exceptions to the Constitutional Principle?: What does the term ‘constitutional principle’ mean? It is clearly intended as a principle of judicial interpretation. Whether it is a principle of which Parliament ought to take account is another matter: in the Sexual Offences Act 2003 it certainly did not, overruling the effect of both the House of Lords decisions.147 Even for the judges, it is a principle and not a rule. Thus, for example, Lord Nicholls held in B v DPP that courts may rebut the presumption of mens rea by reference to ‘the nature of the offence, the mischief sought to be prevented, and any other circ*mstances which may assist in determining what intent is properly to be attributed to Parliament when creating the offence’.148 One might comment that if the presumption can be rebutted so easily it may prove to be worth little. It may be justifiable to rebut the presumption for minor offences which may be described as ‘not criminal in any real sense’, but the failure to regard the possibility of imprisonment as a crucial distinction is a major weakness.149 A powerful example of this is Gammon v Attorney-General for Hong Kong (1985).150 Following the collapse of a building, the defendants were charged with offences against the construction regulations which carried high fines and a maximum prison sentence of three years. Lord Scarman, giving the opinion of the Privy Council, re-affirmed the presumption of mens rea laid down in Sweet v Parsley, and added that ‘the presumption is particularly strong where the offence is “truly criminal” in character’. He went on:

the only situation in which the presumption can be displaced is where the statute is concerned with an issue of social concern; public safety is such an issue…. Even where a statute is concerned with such an issue, the presumption of mens rea stands unless it can also be shown that the creation of strict liability will be effective to promote the objects of the statute by encouraging greater vigilance to prevent the commission of the prohibited act.

Page 22 of 57

Criminal Capacity, Mens Rea, and Fault

The last few words support the principle that strict liability should not be imposed where there is nothing more a defendant could reasonably be expected to do in order to avoid the harm.151 This means that liability is tethered, however loosely, to the defendant's control; liability is not completely strict in these cases. However, the earlier part of the quotation demonstrates how muddy the waters still are. In some cases the courts say that strict liability is appropriate for minor offences which are not truly criminal.152 Yet they also seem to hold, as in Gammon, that it is appropriate where offences relate to public safety or social concern—a description which (as pointed out earlier) could extend to large areas of the criminal law. That the Gammon decision was cited by the House of Lords in B v DPP without any attempt to confront this problem indicates a pessimistic outlook for the ‘constitutional principle’. (p. 167) In a search for principled exceptions, let us examine two areas of the law—firearms, and sexual offences—where the courts have taken a different view. The decision of the Court of Appeal in Deyemi and Edwards (2008)153 is of particular importance, since the defence placed strong reliance on the ‘constitutional principle’ of mens rea. The defendants were found in possession of an article that they believed to be a large torch but which was in fact a stun-gun. They were convicted of possessing a prohibited weapon, contrary to s. 5 of the Firearms Act 1968, and the judge, having examined the facts, gave them a conditional discharge. The Court of Appeal recognized the significance of the House of Lords decisions establishing the ‘constitutional principle’, but held that it was bound by a long line of authority to hold that this is a strict liability offence. The House of Lords decisions were each ‘concerned with the proper meaning of the statutory provisions in question’, held Latham LJ, a dismissal that confines their sphere of influence to statutory provisions that have not yet been the subject of an authoritative interpretation. On the substantive issue, presumably the argument is that strong reasons of public policy require the courts to impose strict liability in firearms cases, otherwise measures of control would be weakened. But the outcome here—conditional discharges for the two defendants—suggests that an acquittal in this kind of case would not weaken the law. The convictions were unfair, as the sentences indicate. The line of firearms cases which the Court applied ought to be overruled. Turning to sexual offences, the decision of the House of Lords in G. (2008)154 confirms that the offence of rape of a child under 13 in s. 5 of the Sexual Offences Act 2003 imposes strict liability as to age. The House was unwilling to accept human rights arguments to the effect that this breaches the presumption of innocence in Art. 6(2) and breaches the Art. 8 rights of the accused (who was aged 15 only).155 The majority view, expressed by Baroness Hale, was that strict liability is necessary here in order to ensure the protection of children from the sexual attentions of others. The implication is that allowing a defence of reasonable mistake (a negligence standard) would reduce that protection unacceptably; on this view, the unfairness and stigma of convicting a mistaken defendant of this serious offence is less important, even when that defendant is also below the age of consent (and could have been charged with a lesser offence). Both the empirical and the normative strands in that attempted justification call for close examination. (iv) Conclusion: Despite pronouncements of high authority on the existence of a ‘constitutional principle’ requiring fault, English law remains in an unsatisfactory state. The judgments in B v DPP and in

Page 23 of 57

Criminal Capacity, Mens Rea, and Fault K refer to the possibility of rebutting the principle or presumption; there are decisions since 2000 that show how easy it is for a court to find a reason for rebutting the presumption; 156 and it seems that courts will follow precedents on the precise statute rather than resorting to the broader authorities on (p. 168) the ‘constitutional principle’.157 So long as different statutes are promoted by different government departments without an overall grammar or standard, progress towards a consistent approach will be hampered. The first move should be for Parliament, probably prompted by the Law Commission, to establish a principled way of proceeding.158 If there are persuasive economic and social arguments in favour of strict liability for minor offences—and those arguments must be rigorously evaluated—then this may be permitted so long as imprisonment is not available. There should be recognition of the principle that no person should be liable to imprisonment without proof of sufficient fault.159 This principle should inform the distinction between minor and non-minor offences. The classification of an offence as ‘regulatory’, whatever that may mean, should be irrelevant to the imposition of strict liability: if imprisonment is available as a sanction, then fault should be required whether it is called ‘regulatory’ or not. Similarly, the ‘public safety’ test stated in the Gammon case should be discarded, not least because it points in exactly the wrong direction by arguing in favour of strict liability for more serious offences. (b) Intention The term ‘mens rea’ has conventionally been used to connote the following fault requirements: intention or recklessness as to a specified consequence, and knowledge of, or recklessness as to, a specified circ*mstance.160 In discussing offences of strict liability, we have considered the main arguments in favour of requiring fault as a condition of criminal liability, chiefly arguments of choice and fair warning. Now we move to the more detailed and specific question of drawing distinctions between the four main forms of fault which generally fall under the umbrella of mens rea. The task is important, because of the key role of intention in serious crimes. Sometimes the intent is the essence of an offence, as in doing an act with intent to impede the apprehension of an offender, all crimes of attempt, and offences defined in terms of ‘doing x with intent to do y’ (such as burglary: entering as a trespasser with intent to steal).161 Sometimes the law uses intention as the main method of grading offences: both the murder– manslaughter distinction and the dividing line between wounding under s. 18 of the Offences against the Person Act 1861 (maximum penalty of life imprison (p. 169) ment) and wounding under s. 20 (maximum penalty of five years’ imprisonment) turn on the presence or absence of intention. (i) Intention in Principle: It is quite possible—indeed, quite normal—to do things with more than one intention in mind. I can demolish a fence with the simultaneous intentions of making way for a new fence, providing wood for the fire, and so on. The approach of the criminal law, however, is generally not to ask with what intentions D committed the act, but to ask whether one particular intention was present when the act was committed. The law, generally speaking, is interested in the presence or absence of one particular intention—that specified in the definition of the offence charged—and not in conducting a general review of D's reasons for the behaviour in question. Did D intend to kill the crew of the aircraft on which he placed a bomb, as well as intending (as he admits) to claim the insurance money on the cargo? Did D intend to assist the enemy by his actions, as well as intending (as he admits) to save his family from a concentration camp?162

Page 24 of 57

Criminal Capacity, Mens Rea, and Fault The law's approach in selecting one intention, and then abstracting it from D's other reasons and beliefs at the time, calls for careful consideration. It is essential to keep in mind the particular intent required by the definition of the offence. It is quite possible to say ‘D pulled the trigger of the gun intentionally’, without implying that D intended to kill V when he pulled the trigger. The offence of murder turns (broadly)163 on the presence or absence of an intention to kill; whether the trigger was pulled intentionally or accidentally may be an important part of the case, but the legally required intention is that D intended to kill V. Loose references to whether D ‘acted intentionally’ can blur this distinction: it is unhelpful to refer to intention without relating it to a particular object or consequence, which in a legal context means the intent specified in the indictment or information.164 This approach to intention may avoid some errors, but the proper definition of intention remains the subject of theoretical debate and judicial disagreement. The core of ‘intention’ is surely aim, objective, or purpose; whatever else ‘intention’ may mean, a person surely acts with intention to kill if killing is the aim, objective, or purpose of the conduct that causes death. When drafting, however, it may be best to avoid the term ‘purpose’ (which may give rise to confusion with D's ultimate purpose in doing the act), and instead to define intention in terms of ‘acting in order to bring about’ the result.165 Similarly in Mohan (1976)166 James LJ defined intention as ‘a decision to bring about [the proscribed result], in so far as it lies within the accused's power, no matter whether the accused desired that consequence of his act or not’. This definition has the advantage of stating that desire is not essential to intention (one may act out of (p. 170) feelings of duty, for example, rather than desire); it has the disadvantage of referring to a ‘decision’, whereas in many offences of violence and other crimes the events happen so suddenly and rapidly that an action can be engaged in intentionally without there having been deliberation about the alternatives beforehand. In law, a spontaneously formed intention is as much an intention as an intention that is the product of lengthy deliberation.167 The Mohan case involved an attempted crime, and intention is thought to be crucial to attempts, because one cannot be said to attempt to produce a result unless one intends to produce it (see Chapter 11.3(a)).168 The decision in Mohan goes some way towards stating the core of the concept of intention, i.e. acting in order to bring about a result.169 It is important to note that, in intending to bring about an end, one must also intend the means adopted to achieve it, because otherwise D could always avoid liability by pointing to some ulterior motive for the action: ‘it was not my purpose to kill V, because my real purpose in shooting at V was to inherit V's money after V's death’. Such a purported detachment of the means from the end is quite unconvincing. Both are part of the intention with which D fired the shot, and the criminal law is interested only in whether the killing was intentional. Should the concept of intention be more extensive than that, in the context of criminal liability? Lawyers have long worked with a concept of intention that includes not only acting in order to bring about x, but also acting with foresight of certainty that x will result—that D can be said to have intended a result if he or she realized that the result was certain to follow from the behaviour in question. An early example of this may be found in Bentham's writings, and his distinction between direct and oblique intention is one way of expressing the point.170 One might say that a consequence is directly intended if D acts in order to produce it, and that it is obliquely intended if it is not D's aim but is known to be certain.171 To regard both these mental attitudes as forms of intention is to make a moral point. It is not necessarily being claimed that ordinary people in their everyday language use the term ‘intention’ in this way.172

Page 25 of 57

Criminal Capacity, Mens Rea, and Fault The claim is that the person who foresees a consequence as certain should be classified as having intended that result rather than as having been merely reckless towards (p. 171) it— and the claim is being made in the knowledge that some killings would thus be classified as murder rather than manslaughter, some woundings described as ‘with intent’ rather than merely as unlawful, and so on. As soon as the argument moves from the moral to the legal, such questions of classification arise. What has to be established is not that all cases of foresight of certainty are socially or morally as bad as all cases of purpose, but that it is more appropriate to classify them with ‘intention’ than with ‘recklessness’. If we pursue the moral part of the argument further, we find that the shorthand phrase ‘foresight of certainty’ is perhaps too brief in this context. Few future events in life are absolutely certain, and a reference to consequences as ‘certain to follow’ would generally mean ‘practically certain to follow’ or ‘certain, barring some unforeseen intervention’.173 A familiar example is D, who places a bomb on an aircraft with the aim of blowing it up in midflight in order to claim the insurance money on the cargo. D knows that it is practically certain that the crew of the aircraft will be killed as a result of the explosion. One might say that D's purpose is to claim the insurance money, but if the charge is murder, that is irrelevant. The key question is whether D intended to kill. Let us assume that D did not act in order to kill, i.e. that he had not intended the death of the air-crew as the means to his end. Should the law extend the definition beyond such a direct intent to cover D's awareness of the practical certainty that the crew would be killed? The argument in favour of this is that D's behaviour shows no respect for the value of human life at all: D knows that the crew will die, and yet he still pursues the aim of blowing up the aircraft. There is little social or moral difference between that and planning the explosion in order to kill the crew. It is sometimes thought that the ‘test of failure’ argues against this: 174 since D would not regard the explosion as a failure if the cargo were destroyed but the crew were not killed, this serves to differentiate him from someone whose purpose is to kill. But to establish that a philosophical distinction exists between D and the purposeful killer is not to conclude the matter: to transfer the argument from morality to law, it has to be decided whether the person who foresees death as virtually certain should be bracketed with the directly intentional killer (murder) or treated as merely reckless (manslaughter).175 Recklessness, as we shall see below, includes the taking of relatively small risks. There is a strong argument that someone who takes a risk of death that amounts to a virtual certainty comes very close, in point of culpability, to the person who chooses someone's death as the means to an end. They both show no respect at all for human life. The Law Commission accepts this, preferring a definition that includes not only the person who acts in order to bring about the prohibited consequence, but also the person who ‘thought that the result was a virtually certain consequence of his or her action’.176 (p. 172) (ii) Intention in the Courts: At present there is no legislative definition of intention. How have the courts approached the question? The leading decisions concern the crime of murder, to be discussed in a later chapter,177 but their effect can be summarized here. The first of the leading cases is Moloney (1985),178 in which the House of Lords held that judges should generally avoid defining the term ‘intention’, beyond explaining that it differs from ‘desire’ and ‘motive’. Only in exceptional cases should the judge depart from this golden rule, notably, where the essence of the defence is that D's purpose was only to frighten, not to harm, the victim. Here the jury should be instructed to decide whether D foresaw the prohibited consequence as ‘a natural

Page 26 of 57

Criminal Capacity, Mens Rea, and Fault consequence’ of the behaviour: if the answer was yes, they could infer intention from that. In the course of his speech Lord Bridge gave hints of the sort of cases he meant to include— cases where the consequence was a ‘little short of overwhelming’, or ‘virtually certain’—but unfortunately the centrepiece of his speech was the term ‘natural consequence’. When this was used by the judge to direct the jury in Hanco*ck and Shankland (1986),179 it was held to be unsatisfactory. The House of Lords overruled its own test of ‘natural consequence’, and Lord Scarman stated that juries should be told that ‘the greater the probability of a consequence the more likely it is that the consequence was foreseen, and that if that consequence was foreseen the greater the probability is that that consequence was also intended’. These decisions left unclear the precise legal meaning of intention and the proper approach to directing a jury, and Lord Lane CJ attempted to synthesize the House of Lords decisions when presiding in the Court of Appeal in Nedrick (1986): 180

Where the charge is murder and in the rare cases where the simple direction is not enough, the jury should be directed that they are not entitled to infer the necessary intention, unless they feel sure that death or serious bodily harm was a virtual certainty (barring some unforeseen intervention) as a result of the defendant's actions and that the defendant realized that such was the case.

This direction now has the authority of the House of Lords. In Woollin (1999)181 the House disapproved a direction in terms of whether D had realized that there was a ‘substantial risk’ of serious injury, and held that the Nedrick formulation should be followed—with one modification. Where Nedrick states that if D foresaw the relevant consequence as virtually certain the court is ‘entitled to infer’ intention, Woollin states that the court is ‘entitled to find’ intention.182 This change has little practical significance, and it leaves open the possibility that, if courts are ‘entitled’ but not required to find intention in these cases, then there may occasionally be cases where they may lawfully decide not to find intention despite foresight of virtual certainty.183 (p. 173) In English law, therefore, intention is not defined in terms of (a) acting in order to bring about a result or (b) acting in the knowledge that the result is virtually certain to follow. The indirect or oblique element, (b), is said to be something on the basis of which intention can be found, and not a species of intention. The reluctance of the judiciary to commit themselves to a particular definition of intention confirms that they see the need to preserve an element of flexibility so that they can continue to allow occasional divergences from the ‘standard’ (a) or (b) definition. Appellate decisions over the years reveal a variety of departures from what might be termed the ‘standard definition’ of intention. Thus in Steane (1947)184 the Court of Criminal Appeal quashed D's conviction under wartime regulations for the offence of doing acts likely to assist the enemy, with intent to assist the enemy. The Court held that if D's acts were as consistent with an innocent intent (such as saving his family from a concentration camp) as with a criminal intent, the jury should be left to decide the matter. This diverges from the standard definition, since it was never discussed whether D knew that it was virtually certain his acts would assist the enemy. The Court could probably have used the defence of duress to quash the conviction, but it evidently thought that adopting a narrow definition of intention provided a simpler route to the desired result. Similarly, in the civil case of Gillick v 185 Page 27 of 57

Criminal Capacity, Mens Rea, and Fault West Norfolk and Wisbech Area Health Authority (1986)185 the House of Lords held that a doctor who gives contraceptive advice to a girl under 16 for clinical reasons, whilst realizing that this would facilitate sexual activity, is not guilty of aiding and abetting the offence of sexual activity with a child. The decision might well have been placed on some such ground as ‘clinical necessity’,186 but instead Lord Scarman explained that ‘the bona fide exercise by a doctor of his clinical judgement must be a complete negation of the guilty mind’. With this sweeping statement it was held that the doctor did not have the intention required for aiding and abetting, even though it may be assumed that prescribing the contraceptives was foreseen as virtually certain to assist the commission of an offence. To set alongside these two decisions which favour a narrow definition of intention it is not difficult to find decisions pointing in a different direction. In Smith (1960)187 D had offered a bribe to an official, solely in order to demonstrate that the official was corrupt. The Court of Criminal Appeal upheld his conviction for corruptly offering an inducement to an official, holding that D had an intention to corrupt so long as he intended the offer to operate on the mind of the offeree. In this case D's law-abiding motivation was held to count for nothing. Similarly in Chandler v DPP (1964),188 the defendants’ convictions of acting ‘for a purpose prejudicial to the safety or interests of (p. 174) the State’ were upheld by the House of Lords. They had infiltrated a military airfield, and this was regarded as prejudicial to the State's interests. The defendants’ argument that their own purpose was to promote the safety and interests of the State (by promoting peace), rather than to prejudice them, was discounted. What these decisions demonstrate is that the courts do not adhere to a single definition of intention. Various observations may be made about this. One common reaction is to treat it as evidence for a ‘realist’ interpretation of how courts behave: they decide on the desired result, and then define the law in whatever way happens to achieve it. But the evidence is limited to a small number of appeal court decisions, and may not reflect the everyday operation of the criminal courts. Even if it were true to some degree (and few suggest that the courts have an absolute freedom in these matters), what is it that leads courts to adopt these reasons for reaching these particular results? Judges in the appellate courts are fond of referring to ‘ordinary language’ as a justification for their decisions, but this often appears to be a camouflage for moral judgments. Critical writers have made much of the tensions revealed by the varied judicial approach. Thus Nicola Lacey scrutinizes the shifting language of the appellate judges and argues that this reflects their attempt to keep the law fairly close to popular conceptions (and thereby to enhance its legitimacy) whilst trying to ensure that the interests of the powerful are not significantly challenged.189 Alan Norrie, focusing on the way in which courts sometimes regard the defendant's motive as a reason for concluding that the result was not ‘intended’ (Steane, Gillick) and sometimes do not (Smith, Chandler), argues that contemporary criminal law is trapped by a set of concepts stemming from a desire to separate ‘legal judgment from substantive moral issues’, which means that in difficult cases the courts find themselves ‘excluding and re-admitting substantive moral issues into a technically conceived set of fault categories’.190 Thus a model direction stating that, where a court is satisfied that D foresaw a result as virtually certain, it is ‘entitled to find’ that D intended the result, may operate so as to allow the courts to expand and contract the definition so as to reflect other factors, including moral judgments of a defendant's background and situation. (iii) Intention Concluded:

Page 28 of 57

Criminal Capacity, Mens Rea, and Fault In delivering the unanimous judgment of the House of Lords in Woollin, Lord Steyn observed that the appeal concerned the crime of murder and that ‘it does not follow that “intent” necessarily has the same meaning in every context in the criminal law’.191 However, the variable approaches to intention described in the previous paragraphs have not been explained by judges on an offence-specific basis, and there would surely need to be particular arguments in favour of adopting a different definition for a certain crime or class of crimes. The Woollin definition may therefore be treated as established, and yet we have seen that the House of Lords left the door ajar: the phrase ‘entitled to find’ preserves an element of ‘moral elbow-room’ (p. 175) which many judges believe to be essential to doing justice. The Law Commission accepts this view: in recommending that ‘an intention to bring about a result may be found if it is shown that the defendant thought that the result was a virtually certain consequence of his or her actions,’192 the Commission argues that this element of flexibility is ‘the price of avoiding the complexity’ needed if a comprehensive definition were attempted, and that broad terms such as ‘extreme indifference’ would create greater uncertainty. The reason judges adopted variable meanings of intention in the decisions discussed above is largely that the standard definition, in combination with the range of available defences to liability, sometimes fails to capture moral distinctions which are thought important. The term ‘intent’—sometimes the determinant of liability, sometimes a primary way of grading offences— is not one that necessarily incorporates elements of moral evaluation, unlike the other mens rea term ‘reckless’ (discussed later).193 Thus when faced with a strong moral pull towards exculpation the courts have sometimes, as in Steane and in Gillick, manipulated the concept of intention rather than developing a defence to criminal liability. However, it would surely be better to adopt a tighter definition of intention, excluding the permissive words ‘may be found’ in the Law Commission's recommended definition, and to place greater emphasis on appropriate defences. Under the Criminal Code the courts would have a power to develop new defences,194 so as to ensure that what they regard as important moral distinctions are marked appropriately. (c) Recklessness Much of the preceding discussion about the proper limits of the concept of intention in the criminal law has inevitably concerned the dividing line between recklessness and intention. The argument was that there are some cases in which D knows the risk of the prohibited consequence to be so very high (i.e. practically certain) that it is more appropriate to classify his mental attitude within the highest category of culpability (intention) rather than in the lesser category of recklessness. We may note that some would draw the dividing line lower, arguing that if D foresaw the prohibited consequence as a probable result, this should be classified as intention, leaving only the lesser degrees of risk within the category of recklessness.195 Another avenue, not explored in English law, would involve employing the US Model Penal Code term ‘knowledge’ that something will occur: to cover states of mind where D foresees a high certainty of the occurrence. We will now move away from these arguments, but they do remind us that debates about the boundaries of intention relate to the grading of culpability and so of (p. 176) offences. The same is true of the lower boundary between recklessness and negligence: when criminal lawyers refer to offences as requiring ‘mens rea’, they usually mean that either intention or recklessness will suffice for liability but that negligence will not. Thus, once again, the debate concerns not so much language as the limits of criminal liability. An abiding difficulty in discussing the legal meaning of recklessness is that the term has been

Page 29 of 57

Criminal Capacity, Mens Rea, and Fault given several different shades of meaning by the courts over the years. In the law of manslaughter, ‘reckless’ has often been regarded as the most appropriate adjective to express the degree of negligence (‘gross’) needed for a conviction: 196 in this sense, it means a high degree of carelessness. In the late 1950s the courts adopted a different meaning of recklessness in the context of mens rea, referring to D's actual awareness of the risk of the prohibited consequence occurring: 197 we shall call this ‘advertent recklessness’. Controversy was introduced into this area in the early 1980s, when the House of Lords purported to broaden the meaning of recklessness so as to include those who failed to give thought to an obvious risk that the consequence would occur: 198 as we shall see in paragraph (ii), the House of Lords has now reversed itself on this point.199 The law of manslaughter will be left for discussion later: 200 here we will focus on the other meanings of recklessness.201 (i) Advertent Recklessness: It was in Cunningham (1957) that the Court of Criminal Appeal held that, in a statute, the term ‘malicious’ denotes intention or recklessness, and that recklessness means that ‘the accused has foreseen that the particular kind of harm might be done and yet has gone on to take the risk of it’.202 There are essentially three elements in this definition, and they are the same ones found in the Model Penal Code's definition of recklessness as ‘the conscious taking of an unjustified risk’. First, it requires D's actual awareness of the risk; 203 this is why it is referred to as ‘advertent recklessness’, and it is regarded as the key element in bringing recklessness within the concept of mens rea. A person should be held to have been reckless about a particular result only if the court is satisfied that he or she was aware of the risk at the time. The second element is that a person may be held to have been reckless if he or she was aware of any degree of risk: we have seen that when the risk is so high as to be a practical certainty, D may be classed as intending the consequence, but any risk, however slight, may be sufficient as a minimum for recklessness, so long as D is aware of it and it materializes. In its recommendations for reform of the law of homicide, the Law Commission proposes a narrower definition of recklessness—that (p. 177) D must be aware of a ‘serious risk’, i.e. one that is ‘more than insignificant or remote’.204 It is not clear whether this would alter the outcome of many cases, but it is right that an offence such as murder should be more tightly defined. The third element is that the risk which D believes to be present must be an unjustified or unreasonable one to take in the circustances. This is an objective element: courts have rarely discussed it, but it exerts a significant background influence. A typical example of the objective element is the surgeon who carries out an operation knowing that death will probably result.205 In this example, assuming that there is a clinical justification for the operation, even exposing the patient to a high degree of risk from the operation itself may well be fully justified. In general, thus, ‘the responsibility line is drawn according to an evaluation of the nature of the activity and the degree of the risk’.206 However, the circ*mstances in which an activity is undertaken may be as important as the nature of the activity itself. Suppose, in the example just given, that (for a bet) the surgeon tried to conduct the same operation wearing a blindfold. Even if the surgeon was so skilled that most experts would say he or she posed no extra risk to the patient when blindfolded, the surgeon's gratuitous introduction of an extra potential source of risk would be regarded as acting recklessly. The evaluative task of determining objective risk has rarely been performed

Page 30 of 57

Criminal Capacity, Mens Rea, and Fault by the courts, but, as Alan Norrie rightly points out, this is because prosecutors have often made their own evaluations at an early stage and no prosecution (or at least no prosecution for a serious offence such as manslaughter) has been brought in most such cases.207 Thus we have scant judicial authority relating to the objective element in recklessness. The justifications for the advertent definition of recklessness are grounded in the principle of individual autonomy and the importance of respecting choice, outlined above.208 The distinction between recklessness and negligence turns on D's awareness or unawareness of the risk. In both cases there is an unreasonable risk taken, but D should only be held to have been reckless if he or she was aware of the risk. A person who is aware of the risk usually chooses to create it or to run it, and therefore chooses to place his or her interests above the well-being of those who may suffer if the risk materializes.209 Choosing to create a risk of harmful consequences is generally much worse than creating the same risk without realizing it. Moreover, holding a person reckless despite unawareness of the risk would result in a conviction in a case like Stephenson (1979).210 D, a schizophrenic, made a hollow in a haystack in order to sleep there; he felt (p. 178) cold, and so lit a small fire, causing the whole haystack to go up in flames, and resulting in damage of some £3,500. The defence relied on medical evidence that D may not have had the same ability to foresee the risk as a mentally normal person. The Court of Appeal, quashing D's conviction, held that the definition of recklessness clearly turned on what this defendant actually foresaw, and the medical evidence should have been taken into account on this point. This decision, then, strongly affirms the element of individual fairness in the advertent or subjective definition. An entirely objective test would exclude this. Does concentration on the element of awareness always produce decisions in accord with fairness? There are at least two types of awkward case for a test of liability which requires the court to be satisfied that the defendant actually saw the risk, however briefly. One is where a person acts impulsively in the heat of the moment. This is often expressed in ordinary speech by saying ‘I acted without thinking’, or ‘I just didn't think’. D denies that he or she was aware of the risk at the time of acting. In Parker (1977)211 D tried unsuccessfully to make a telephone call from a payphone; in his frustration he slammed down the receiver and broke it. The Court of Appeal upheld his conviction for causing criminal damage recklessly, despite his defence that it did not occur to him that he might damage the telephone. The Court held that he must have known that he was dealing with breakable material, even if that fact was not at the forefront of his mind when he slammed the receiver down. He had ‘closed his mind to the obvious’, or suppressed this knowledge at the time of the act.212 It is quite evident that this decision involves some stretching of the awareness element which is thought to be central to advertent recklessness. In effect, it broadens the timeframe from the moment of the act itself to an earlier and calmer time, when D would almost certainly have answered the question: ‘What might happen if you slammed down a telephone receiver?’, by saying: ‘It might break’. The reason for thus broadening the timeframe is presumably to prevent bad temper resulting in an acquittal, since this would be socially undesirable: people should control their tempers. But it does sully the subjective purity of this definition of recklessness. In that regard, an important difference between Parker and Stephenson concerns the reasons why each did not foresee the possible damage that might be done by their conduct. In Stephenson's case, the explanation was a mental disorder the effects of which he could not control. By contrast, in Parker's case, the explanation was his loss of temper, something he simply failed to control. However, as we will see, the House of Lords has rejected any attempt to finesse the definition

Page 31 of 57

Criminal Capacity, Mens Rea, and Fault of recklessness by reference to such factors. The second problem is the ‘couldn't care less’ attitude: D might not have thought about a particular consequence, because it was irrelevant to his interests. If this (p. 179) version of events is accepted, D must be acquitted on the advertent definition of recklessness. Antony Duff has argued that these cases can and should be included within the meaning of recklessness, by invoking the concept of ‘practical indifference’. This is ‘a matter, not of feeling as distinct from action, but of the practical attitude which the action itself displays’. Moreover, it may include cases in which D fails to advert to certain aspects of the situation: ‘what I notice or attend to reflects what I care about; and my very failure to notice something can display my utter indifference to it’.213 The argument is that people who are practically indifferent to certain key features of a situation may be just as much to blame as those who do advert to them. This argument was put strongly in relation to the pre-2003 law of rape, contending that judgments of practical indifference should be made on the basis that men ought to consider the victim's interests in such cases. Defensible as that approach is in that context,214 the question is whether it is subjective, since it brings within the concept of recklessness some defendants who do not actually advert to these matters. Duff's response is that requiring practical indifference is just as subjective, and just as respectful of individual autonomy, as requiring awareness of risk. The practical indifference test looks to D's attitude at the time, on the basis of his acts and words. In practice, it is likely that juries applying the test of advertent recklessness would convict such defendants on the basis that they must have realized the risk; but that merely suggests that it may be unnecessary to confront Duff's point, not that it is wrong. Thus there are at least two types of situation in which the ‘awareness’ requirement, the centrepiece of advertent recklessness, is problematic (on some views) and may fail to yield an acceptable grading of blameworthiness. One is the person who acts impulsively or in a temper, ‘without thinking’. The other is the person who fails to think about the consequences out of indifference to them. A third possibility would be where D states that he was so pre-occupied with other aspects of what he was doing as to give no thought to a particular consequence (although the courts might be reluctant to accept such a defence).215 This brings us to a discussion of two key decisions in the House of Lords. (ii) The Decisions in Caldwell (1982)216 and in G (2004):217 In Caldwell, the House of Lords introduced a new objective definition of recklessness that, incidentally, would encompass the three types of situation with which the traditional definition does not deal convincingly. It was heavily criticized, and for all practical purposes the subsequent decision in G overrules it. Nonetheless, a brief discussion is appropriate here, in order to identify some of the issues of principle raised by the Caldwell decision and 28 years of applying it (mostly in criminal damage cases, since it was never accepted (p. 180) throughout the criminal law).218 In Caldwell, Lord Diplock formulated the following model direction: a person is guilty of causing damage recklessly if:

(i) he does an act which in fact creates an obvious risk that property would be destroyed or damaged and (ii) when he does the act he either has not given any thought to the possibility of there being any such risk or has recognized that there was

Page 32 of 57

Criminal Capacity, Mens Rea, and Fault some risk involved and has nonetheless gone on to do it.

It will be noticed that this definition includes the advertent element (by referring to the person who recognizes the risk and takes it), but then goes further, extending to all those who fail to give any thought to the possibility of a risk which may be described as obvious. A major problem with Lord Diplock's test of what would have been obvious to the reasonable person was that it admitted of no exceptions. The effect was to convict young children and mentally impaired defendants by applying to them an objective standard of foreseeability that they could not meet.219 Thus, even if the Caldwell test were to be regarded as an improvement because it extended to thoughtless and inconsiderate wrongdoers, the absence of a capacity exception produced unfair convictions in some cases. The case of G220 involved two children aged 11 and 12 who set fire to some newspapers beneath a rubbish bin and then left, after which the fire spread and caused major damage to nearby shops. The House of Lords considered whether to preserve the Caldwell test and engraft a capacity exception on it so as to exempt those (such as children and the mentally disordered) who might be incapable of attaining the objective standard; but this solution was rejected on the ground that the Caldwell test was already complicated and that this would over-complicate it to the extent of risking confusion among juries and magistrates.221 The leading speech by Lord Bingham accepts the substance of the criticisms of Caldwell—the lack of legal foundation for the decision, the unfairness of its effects in some cases—and marks a reversion to the traditional, more subjective definition of recklessness based on the defendant's awareness of the risk. This closes one of the common law's less distinguished chapters,222 and more or less returns the criminal law to a single definition of recklessness. But it does not advance the debate about the types of case that strictly fall outside that traditional definition of recklessness—the indifferent D who appears not to have thought of the risk at all, and D who acts in sudden rage or temper and claims not to have realized the risk of harm.223 (p. 181) (iii) Reckless Knowledge: Both Caldwell and G were concerned with the use of recklessness as a fault term in its own right under the Criminal Damage Act 1971. Sometimes, though, recklessness as defined in G is used as the basis for inferring that D actually ‘knew’ that something was the case, where it is knowledge that is the fault term in issue. In some cases, what we can refer to as ‘G’ recklessness as to whether a circ*mstance element of a crime existed has been held to constitute ‘wilful blindness’, i.e. where D knows that there is a risk that a prohibited circ*mstance exists, but refrains from checking it. An example is Westminster City Council v Croyalgrange Ltd (1986),224 where D was charged with knowingly permitting the use of premises as a sex establishment without a licence. The House of Lords held that:

it is always open to the tribunal of fact, when knowledge on the part of a defendant is required to be proved, to base a finding of knowledge on evidence that the defendant had deliberately shut his eyes to the obvious or refrained from enquiry because he suspected the truth but did not want to have his suspicion confirmed.225

Page 33 of 57

Criminal Capacity, Mens Rea, and Fault It will be seen that Lord Bridge used the language of inference here, suggesting that a court might infer knowledge from wilful blindness in the same way as he suggested that intention might be inferred from foresight of virtual certainty.226 There is well-known authority to the effect that wilful blindness should be treated as actual knowledge.227 However, D does not know the relevant circ*mstance in such cases, since he has refrained from finding out, and it may not be easy to establish that he had an overwhelmingly strong belief (that it is virtually certain) that the prohibited circ*mstance exists. Wilful blindness should therefore be treated as a form of reckless knowledge, and relevant only when reckless knowledge is sufficient, unless it can be shown that D refrained from making inquiries because he was virtually certain that his suspicion would be confirmed.228 (d) Negligence Traditionally, books dealing with English criminal law afford an extremely brief discussion to negligence as a standard of liability. Among the common law crimes, only manslaughter rests on liability for (gross) negligence,229 and careless driving and dangerous driving are among the few common offences based on negligence. Yet there are many offences of negligence among the statutory offences regulating (p. 182) various commercial and other activities, often taking the form of an indictable offence of doing an act ‘with intent’ to contravene the regulations, supported by a summary offence of negligence in committing an act in such a way as to ‘have reason to believe’ that the regulations will be contravened.230 Moreover, other systems of law tend to have a larger group of offences of negligence, and may look askance at a set of laws which penalizes negligence where death is caused, but does not penalize it where serious injury or suffering is caused or risked. One reason for the opposition of many English text-writers to criminal liability for negligence is that it derogates from the subjective principles stated at the beginning of this chapter.231 The doctrine of mens rea, as expressed in the requirements of intention and recklessness, makes liability depend on proof that D chose the harm, in the sense of intending it or at least being aware that it might result. These elements are missing where mere negligence is sufficient: there is no need to prove that D adverted to the consequences at all, so long as the court is satisfied that a reasonable person in that situation would have done so. To have negligence as a standard of liability would therefore move away from advertence as the foundation of criminal responsibility, and in doing so might show insufficient respect for the principle of autonomy. The counter-argument to this might challenge the relevance to culpability (and to the public censure of criminal conviction) of ‘the distinction between foreseen effects and effects that were unforeseen only because the agent was not paying as much attention as he could and should have paid’.232 The proposition that human actions are sufficiently free to make blame and punishment defensible underlies most of the criminal law,233 and it might be argued that a person who negligently causes harm could have done otherwise—he could have taken the care necessary to avoid the harm. So long as the individual had the capacity to behave otherwise, it is fair to impose liability in those situations where there are sufficient signals to alert the reasonable citizen to the need to take care. Autonomy is a fundamental principle, but this does not mean that advertence should always be required so long as there is fair warning and a fair opportunity to conform to the required standard. Three features of this counter-argument should be noted. First, its focus on capacity should not be dismissed as ‘objective’, for that would be an undiscriminating use of the term. As Hart

Page 34 of 57

Criminal Capacity, Mens Rea, and Fault has shown, it is perfectly possible to allow exceptions for those who cannot be expected to attain the standard of foresight and control of the reasonable citizen. One only has to supplement the question, ‘did D fail to attain a reasonable standard of care in the circ*mstances?’, with the further question; ‘could D, given his mental and physical capacities, have taken the necessary precautions?’234 Negligence (p. 183) liability need be ‘objective’ only in so far as it holds liable those who fail to take precautions when they could reasonably have been expected to do so. Liability can be termed subjective in so far as it takes account of the limited capacities of the particular person. Taking objective and subjective aspects together, the blameworthiness may be expressed as ‘the culpability of unexercised capacity’.235 As Andrew Simester puts it:

Without external standards, judgement is impossible. Without reference to the defendant, judgement cannot lead to blame. The device of the reasonable man is, in a sense, one means by which the law seeks to reconcile the impersonal with the humane.236

In addition, empirical research suggests public support for some such individualization of negligence liability.237 Secondly, negligence liability may also derogate from any principle of contemporaneity, in the sense that the culpable failure to take precautions often pre-dates the causing of the harm: the rail worker failed to check the signals or the track, so that a crash occurred later; D misunderstood the mechanism of the gun, so that when he later pulled the trigger it killed someone. The enquiry into capacity and opportunity necessitated by negligence liability widens the timeframe of the criminal law, giving precedence to the doctrine of prior fault over the principle of contemporaneity.238 Thirdly, the argument is in favour of negligence liability, not strict liability. Existing law imposes obligations on people who engage in various activities: the obligations of those operating systems of public transport; or the obligations of driving a motor vehicle; or the obligations of owning or managing a factory; or the obligations of engaging in a particular trade or business. Strict liability was criticized in paragraph (a). Negligence liability, on the other hand, is not open to the same objections. The discussion thus far should have established that people who cause harm negligently may be culpable, in so far as they fail to take reasonable precautions when they have a duty and the capacity to do so. What it does not establish is that negligence is an appropriate standard for criminal liability, for it must be borne in mind that criminal liability is the law's most condemnatory form, and in principle it should be reserved for serious wrongs.239 How might it be argued that the English doctrinal tradition of drawing the line of criminal liability below intention and recklessness, and above negligence (at least for ‘conventional’ crimes, such as those in the Draft Criminal Code)240 is ill-founded? One approach would be to establish that some cases of negligence manifest greater culpability than some cases of subjective recklessness—the principal justification for the Caldwell decision. Thus it could be claimed that a person who knowingly takes a slight risk of harm is less culpable than another person who fails to think about or recognize a high risk of the same harm: D, a shooting champion, fires at a target, knowing that there is a slight risk that the bullet will ricochet and (p. 184) injure a spectator, which it does; E, who rarely handles guns, is invited to participate in a shooting party and fires wildly into bushes, failing to consider the possibility of others being there, and one is injured. Is D manifestly more culpable than E? A different comparison would be between

Page 35 of 57

Criminal Capacity, Mens Rea, and Fault someone who knowingly takes the risk of a small harm occurring and someone who fails to recognize the risk of a serious harm occurring: a criminal law which convicts the former and not the latter could be said to be transfixed by the notion of a ‘consistent’ general part. Why maintain that negligence is never an appropriate standard of criminal liability, even where the harm is great and the risk obvious? The argument is therefore moving towards the conclusion that negligence may be an appropriate standard for criminal liability where: (i) the (potential) harm is great; (ii) the risk of it occurring is obvious; (iii) D has a duty to try to avoid the risk; and (iv) D has the capacity to take the required precautions. This opens up further debates on various points. The thesis is that negligence may be an appropriate standard where there are well-known risks of serious harm. This argues in favour of negligence as a standard of liability for certain serious offences against the person, including some serious sexual offences,241 and also for some serious offences against the environment and property. But it must be debated whether liability for serious crime should be confined to gross negligence, not simple negligence. And it would be vital to protect ‘rule of law’ expectations, and thus to ensure that people receive fair warning of any duties that may form the basis of criminal negligence liability.242 The spread of negligence liability would not have to result in the broadening of the traditional category of mens rea: negligence could be admitted as a form of fault, whereas intention and recklessness would remain the two forms of mens rea. It would be perfectly possible for a criminal code to provide separate crimes of negligence, with lower maximum sentences, at appropriate points in the hierarchy of offences. A further issue is whether the offences of negligence should be in the inchoate mode—’failing to take reasonable precautions’—or should be tied to the occurrence of the particular harm. Careless driving is of the former type, manslaughter of the latter, and this point will be pursued further in connection with crimes of endangerment.243 Even granted this argument in favour of criminalizing certain instances of negligence, what would be the point of doing so? This takes us back to the aims of the criminal law, discussed earlier.244 It might be tempting to maintain that the general preventive aim of the criminal law cannot be served by offences of negligence: the notion of deterrence presupposes rational reflection by D at the time of offending, whereas the distinguishing feature of negligence is that D failed to think (when a reasonable person would have done). However, it can be argued that crimes of negligence may exert a general deterrent effect, by alerting people to their duties and to the need (p. 185) to take special care in certain situations. The practical prospects of deterrence here seem no less propitious than in relation to offences requiring intention or recklessness. The principal justification, however, would be that negligent harm-doers deserve criminal conviction because and in so far as they are sufficiently culpable. This is a question of degree and of judgment, on which views may differ.245 (e) Objective versus subjective Much of the discussion of the law in this section of the chapter has concerned the interplay of subjective and objective factors in the definition of the core fault terms. It has been suggested that in crimes where strict liability is imposed on individual defendants, the courts have generally placed insufficient emphasis on respect for individual autonomy and the importance of requiring fault. When dealing with recklessness and mistake, however, the tendency of some text-writers and judges has been to regard the advertent or subjective approach as axiomatic, thus excluding from conviction certain people who may be no less culpable than

Page 36 of 57

Criminal Capacity, Mens Rea, and Fault those who are convicted. The Caldwell test could be seen as a way of supplementing the narrow conception of moral fault embodied in advertent recklessness, but it was flawed in other respects (notably, the absence of an incapacity exception) and it perished.246 An alternative is Duff's test of practical indifference, which relies considerably on objective judgments as evidence of a person's attitude when behaving in a particular way. A further alternative would be to introduce more offences of negligence and, in respect of mistaken belief, more objective limitations on defences to criminal liability—a task on which the legislature embarked in the Sexual Offences Act 2003. It is evident that, in many cases examined in this chapter, an approach that focuses solely on advertence fails to capture some moral distinctions and to satisfy all social expectations.247 Subjective tests heighten the protection of individual autonomy, but they typically make no concession to the principle of welfare and the concomitant notion of duties to take care and to avoid harming the interests of fellow citizens. However, if we are to move towards greater reliance on objective standards, at least two points must be confronted. First, objective tests must be applied subject to capacitybased exceptions. This preserves the principle of individual autonomy by ensuring that no person is convicted who lacked the capacity to conform his or her behaviour to the standard required. Secondly, any improved moral ‘fit’ obtained by moving more towards objective standards must be weighed against the greater detraction from the principle of maximum certainty that is likely to result.248 Objective standards inevitably rely on terms such as reasonable, ordinary, (p. 186) and prudent. They appear much more malleable and unpredictable than subjective tests that ask whether or not a defendant was aware of a given risk, and they explicitly leave room for courts and even prosecutors to make social judgments about the limits of the criminal sanction.

5.6 The variety of fault terms Although the focus so far has been upon intention, recklessness, and knowledge, an examination of criminal legislation in force—some modern, some from the nineteenth century— reveals a diversity of fault terms. Even if the Draft Criminal Code were to be enacted, its provisions would not be restricted to the core fault terms discussed so far. Moreover, the Code would cover only some two hundred out of perhaps ten thousand criminal offences, so the diversity will inevitably remain for some years. A full survey of the different fault terms cannot be offered here, but some general remarks may be worthwhile. Nineteenth century legislation such as the Offences Against the Person Act 1861 makes considerable use of the term ‘maliciously’.249 It is now settled that this term should be interpreted to mean intention or recklessness, which simplifies the criminal lawyer's task.250 Unfortunately, certain other terms have not been interpreted consistently in line with the core terminology. Many statutory offences, both ancient and modern, rely on the term ‘wilfully’: although in Sheppard (1981)251 the House of Lords held that the term meant ‘intentionally or recklessly’ in the context of the crime of wilful neglect of a child, there are other offences in which ‘wilfully’ has been held not to require full mens rea.252 Many offences are defined in terms of ‘permitting’, a word that has usually been interpreted as requiring full knowledge but has sometimes been held to impose strict liability, even on individuals.253 More to the point, however, is the fact that many major criminal offences rely on fault terms that bear little relation to any of those discussed so far. Theft and several other Theft Act 254 Page 37 of 57

Criminal Capacity, Mens Rea, and Fault offences rely on the term ‘dishonestly’, which, as we shall see,254 encompasses a mixture of elements of subjective awareness and motivation with elements of objective moral judgment. Some fraud offences turn on whether the act or omission was done ‘fraudulently’. And a number of public order and racial hatred offences impose liability where a certain consequence is ‘likely’ to result from D's conduct, without reference to whether D is aware of this likelihood. Thus, for example, a person commits the offence of creating ‘fear or provocation of violence’ by the use of threatening, abusive, or insulting words or behaviour either with intent to cause another person (p. 187) to believe that immediate unlawful violence will be used, or ‘whereby that person is likely to believe that such violence will be used or it is likely that such violence will be provoked’.255 Similarly, the offence of publishing or distributing racially inflammatory material is committed if either D intends thereby to stir up racial hatred or ‘having regard to all the circ*mstances racial hatred is likely to be stirred up thereby’.256 Offences that rely on the court's assessment of the probable effect of certain conduct may be said to impose a form of strict liability, or at least liability for negligence, if it is assumed that the defendant ought to have known what effect was likely. However, suffice it to say that criminal offences in English law vary in their use of fault terms. The arguments for and against the core terms, examined in this chapter, should provide a framework for considering the justifications for most other fault terms that may be encountered.

5.7 The referential point of fault To say that a certain crime should require intention or recklessness is not enough. One must enquire: intention (or recklessness) as to what? It might be said loosely that ‘the crime of manslaughter requires proof of intention or recklessness’: the reason this is a loose statement is that the intent or recklessness required may be the same as that for assault or some other criminal act, whereas the liability imposed is that for homicide. Close analysis of the elements of the crime will show that the required fault and the result specified in the definition are not on the same level. This is what the principle of correspondence, outlined above, aims to eliminate.257 Whenever one is discussing intent or recklessness, its referential point should always be established. (a) Fault, conduct, and result The argument may be carried further by considering the width or narrowness of the definitions of offences. It would be far easier to establish intent for a broad offence—such as intentionally causing physical harm to another—than to establish intent in a system with a hierarchy of graded offences—such as attempted murder, causing serious injury intentionally, causing injury intentionally, and common assault—which would require proof of more specific mental states. Similarly, a law which includes a general offence of intentionally causing damage to property belonging to another makes it far easier to establish the intent than a law with a series of offences differentiated according to the type of property damaged. Do these different legislative techniques have significant implications for the subjective doctrines of fault? Surely they do: one could argue that a single broad offence of ‘intentionally causing physical (p. 188) harm to another’ would obliterate the distinction between intending a minor assault and intending a major injury, and that a single broad offence of ‘intentionally damaging property belonging to another’ obliterates the distinction between intending damage to a cheap item and intending damage to an expensive item.258 Any tendency towards broader offence 259

260 Page 38 of 57

Criminal Capacity, Mens Rea, and Fault definitions, evident in criminal damage259 but not in sexual offences,260 would give greater weight to the ‘malice principle’ of liability for the consequences of any wrongdoing (section 5.4(b)) than to the principle of correspondence (section 5.4(a)). To that extent, it would detract from the elements of choice and control which are fundamental to the subjective approach. But how should this problem be solved?261 It is hardly practical to allow each person to nominate those factors which he or she regarded as significant in any particular event: who is to say whether fidelity to individual choice and control requires two or twenty grades of criminal damage, or two or four grades of offences of violence? Nonetheless, the implications for fault principles of these labelling decisions262 should be kept firmly in mind. The argument may be taken still further, for there are cases where it is plain that D intended to cause a different result from the one which actually occurred. How ought the law to deal with such cases? Should it respect D's choice, and provide for a conviction of attempting to do X (which was what D intended to do)? Or should it regard the result as the dominant factor, ignore the difference in D's intention, and convict on the basis of ‘sufficient similarity’ between the intention and the result? English law adopts the latter, more pragmatic approach. The Law Commission, in introducing a provision into the Draft Criminal Code which follows the traditional approach, confirms the emphasis on results by stating that a conviction for attempt would be ‘inappropriate as not describing the harm done adequately for labelling or sentencing purposes’.263 The traditional English approach rests on three doctrines—unforeseen mode, mistaken object, and transferred fault. (b) Unforeseen mode When D sets out to commit an offence by one method but actually causes the prohibited consequence in a different way, the offence may be said to have been committed by an unforeseen mode. Since most crimes penalizing a result (with fault) do not specify any particular mode of commission,264 it is easy to regard the difference of mode as (p. 189) legally irrelevant. D intended to kill V; he chose to shoot him, but the shot missed; it hit a nearby heavy object, which fell on V's head and caused his death. Any moral distinction between the two modes is surely too slender to justify legal recognition. To charge D with attempting to kill V when he did kill him seems excessively fastidious. Pragmatism is surely the best approach here, and English law is generally right to ignore the unforeseen mode.265 (c) Mistaken object When D sets out to commit an offence in relation to a particular victim but makes a mistake of identity and directs his conduct at the wrong victim, the offence may be said to have been committed despite the mistaken object. The same applies if D intends to steal one item of property but mistakenly takes another. So long as the two objects fall within the same legal category, it may be said that any moral distinction between them is too slender to justify legal recognition. However, much depends on the breadth of definition of the relevant offence: there is surely some moral significance in the plea: ‘I thought the picture I damaged was just a cheap copy; I had no idea that a valuable painting would be kept in that place’.266 English law favours the pragmatic answer of reflecting shades of moral culpability at the sentencing stage, but one might argue on principle that to convict this person of intentionally or recklessly damaging a valuable painting is a gross mislabelling of the wrong. In one sphere, English law's general approach of ignoring mistake of object within the same offence is not followed. This is

Page 39 of 57

Criminal Capacity, Mens Rea, and Fault the law of complicity: where A gives assistance to D who plans to kill X, and then D decides to kill Y, there is long-standing authority to the effect that A cannot be convicted for aiding and abetting D's murder of Y.267 The complexities of the moral distinctions drawn here are discussed in Chapter 10.5, but if it is accepted that the identity of the victim is so important in this type of case, one may enquire more widely whether there really is inadequate moral significance in the plea: ‘I intended to kill my enemy, X, and never meant any harm to the poor innocent, Y’. The pragmatic approach adopted elsewhere in the criminal law (apart from complicity) may fail to mark significant moral distinctions in some cases, and many might be dissatisfied if the only conviction were for attempting to murder X. (d) Transferred fault When D sets out to commit an offence in relation to a particular person or a particular property but his conduct miscarries and the harm falls upon a different person or (p. 190) a different property, English law regards D's intent as transferred and the offence as committed against the actual victim or property. When the fault is transferred, any defence which D may have is transferred with it.268 As with unforeseen mode and mistaken object, the fault may only be transferred within the same class of offence.269 Thus, if D throws a brick at some people, intending to hurt them, and the brick misses them and breaks a window, the intent to injure cannot be transferred to the offence of damaging property.270 In this situation, the possible offences are an attempt to cause injury, and recklessly damaging property. As with the doctrine of mistaken object, the breadth of definition of the offence has some importance here. It is one thing to accept that D, who swung his belt at W and struck V, should be convicted of injuring V; 271 it is quite another thing, in moral terms, to accept that E, who threw a stone at a window, should be convicted of intentionally damaging a valuable painting which, unbeknown to him, was hanging inside. Yet English law would convict E, applying the broad wording of the Criminal Damage Act 1971 (any ‘property belonging to another’), without any need to rely on the doctrine of transferred fault.272 Thus the ambit of all three doctrines is much affected by the breadth of each offence definition. The doctrine of transferred fault and its relationship with conceptions of subjective guilt remain sources of considerable controversy.273 Rather surprisingly, in view of its long pedigree in English law, the doctrine was denounced by Lord Mustill in the House of Lords for ‘its lack of any sound intellectual basis’. In Attorney-General's Reference (No. 3 of 1994)274 D stabbed his girlfriend in the stomach, knowing that she was pregnant. Two weeks later the child was born prematurely, and because of its grossly premature birth it failed to thrive and died after four months. The House of Lords held that on these facts D could not be convicted of murder, holding that transferred malice could have no application because the foetus had no separate existence at the time the mother was attacked. The facts of this case are unusual, thankfully, but the House of Lords failed to deal convincingly with the relevance of the doctrine of unforeseen mode (should it matter that the child's death resulted from the premature birth, not from any direct wound?) and with the relevance of the extended principle of contemporaneity (if the death was part of an unbroken sequence of events following the stabbing, should not D's original intent be connected with the ultimate death?).275 It could be argued that it would go too far if three artificial doctrines (p. 191) (transferred fault, unforeseen mode, and extended contemporaneity) were combined to find someone guilty of the highest crime in the land. Indeed, it has been argued that the law should recognize a further restrictive principle, the remoteness doctrine, so as to ensure that there is no conviction of an offence (e.g.

Page 40 of 57

Criminal Capacity, Mens Rea, and Fault murder) if the way in which the death of the unanticipated victim occurred was so remote from what D intended or anticipated that to convict D of murdering the actual victim would be an unrepresentative label.276 (e) Establishing the referential point A system of criminal law which succeeded in reflecting the varying degrees of importance which people attribute to aspects of their intention (the mode of execution, the identity of the victim, the value of the property) might be a ‘law professor's dream’, but it is clearly not practical. Such an individuated or fine-grained approach to fault has to give way, at least in some respects, to claims of administrative efficiency. But that does not establish that the traditional English approach is the most appropriate. The draft Criminal Code provides for the continuation of the pragmatic approach, arguing that this is simpler for prosecutors and that an attempt conviction in the above situations would ignore the harm actually done.277 Does its pragmatism stretch too far? Would it not be better to analyse some of these cases in terms of an unfulfilled intention, combined with an accidental (or perhaps reckless) causing of harm? Some would argue that the present law of inchoate offences would not ensure a conviction in all these cases of miscarried intent and miscarried recklessness: 278 according to this view, the three doctrines are not merely effective in returning convictions and symbolically right in their emphasis on results,279 but also necessary if justice is to be done in all cases. There is, it may be argued, no serious distortion of ‘desert’ or proportionality involved in the three doctrines, since the doctrines do not misrepresent the class of harm that D set out to commit. Yet there remains the law's ambivalence about the importance of a victim's identity: if this really is significant to offenders and people's judgments of them, as the law of complicity implies, should not prosecutors make more use of the law of attempts, where it is clearly applicable?280

(p. 192) Further Reading H. L. A. HART, Punishment and Responsibility (2nd edn., 2008), chs 2 and 5. J. Gardner, ‘Introduction ‘, to H. L. A. Hart, Punishment and Responsibility (2nd edn., 2008). R. A. Duff, Answering for Crime (2007), ch 3. R. A. Duff, ‘Whose Luck is it Anyway?’ in C. Clarkson and S. Cunningham (eds), Criminal Liability for Non-Aggressive Death (2008). V. Tadros, Criminal Responsibility (2005), ch 8. A. P. Simester (ed.), Appraising Strict Liability (2005). A. Ashworth, ‘A Change of Normative Position: Determining the Contours of Culpability in Criminal Law’ (2008) 11 New Crim LR 232.

Notes: 1 Page 41 of 57

Criminal Capacity, Mens Rea, and Fault 1 A. Ashworth and M. Blake, ‘The Presumption of Innocence in English Criminal Law’ [1996]

Crim LR 306. 2 See further, Chapter 8. 3 P. H. Robinson, Structure and Function in Criminal Law (1997), ch 3. 4 For debate see J. Gardner and H. Jung, ‘Making Sense of Mens Rea: Antony Duff's Account’

(1991) 11 OJLS 559; J. A. Laing, ‘The Prospects of a Theory of Criminal Culpability: Mens Rea and Methodological Doubt’ (1994) 14 OJLS 57; J. Gardner, ‘Criminal Law and the Uses of Theory: a Reply to Laing’ (1994) 14 OJLS 217. 5 E.g. R. A. Duff, ‘Law, Language and Community: Some Preconditions of Criminal Liability’

(1998) 18 OJLS 189. 6 For further analysis, see Law Commission, Unfitness to Plead (CP No. 197, 2010); R. Mackay,

Mental Conditions Defences in Criminal Law, ch 5. 7 Notably by the House of Lords in C v DPP [1996] AC 1. 8 An argument expressed strongly by G. Williams, ‘The Criminal Responsibility of Children’

[1954] Crim LR 493, at 495–6. 9 For discussion of whether a defence of doli incapax still exists, see DPP v P [2006] 4 All ER

628 and T [2008] 2 Cr App R 17. 10 J. Horder, ‘Pleading Involuntary Lack of Capacity’ (1993) 52 Camb LJ 298, at 300–2. 11 (1999) 30 EHRR 121. 12 G. van Bueren, The International Law on the Rights of the Child (1995), ch 7. 13 Practice Direction: Crown Court (Trial of Children and Young Persons) [2000] 1 Cr App R

483. 14 SC v United Kingdom (2005) 40 EHRR 226; L. Hoyano, ‘The Cornoners and Justice Act

2009: Special Measures Directions Take Two: Entrenching Unequal Access to Justice?’ [2010] Crim LR 345. 15 Office of the Commissioner of Human Rights, Report by Mr Alvaro Gil-Robles, Commissioner

for Human Rights, on his visit to the United Kingdom (Comm DH (2005) 6), paras. 105–7. 16 F. Zimring, ‘Toward a Jurisprudence of Youth Violence’, in M. Tonry and M. Moore (eds),

Youth Violence, (1998), 447. 17 A. von Hirsch and A. Ashworth, Proportionate Sentencing (2005), ch 3. 18 See further H. Keating, ‘Reckless Children’ [2007] Crim LR 546. 19 Mackay, Mental Condition Defences in Criminal Law, ch 2; V. Tadros, Criminal

Responsibility, ch 12. 20 Page 42 of 57

Criminal Capacity, Mens Rea, and Fault 20 Section 1(2) of the Criminal Procedure (Insanity and Unfitness to Plead) Act 1991 requires

the evidence of two doctors, at least one of them approved by the Home Secretary as an experienced psychiatrist. 21 R. D. Mackay, ‘Fact and Fiction about the Insanity Defence’ [1990] Crim LR 247. 22 The presence in prison, in consequence, of many offenders with mental disorders was

criticized in Lord Bradley's Report, People with Mental Health Problems or Learning Disabilities in the Criminal Justice System (2009). 23 Hospital orders, with a restriction indicating the minimum time to be spent in hospital to

protect the public from serious harm, form roughly 38 per cent of disposals in insanity cases, with supervision orders being made in around 41 per cent of cases: R. D. Mackay, B. J. Mitchell and L. Howe, ‘Yet More Facts about the Insanity Defence’ [2006] Crim LR 399. 24 R. D. Mackay, B. J. Mitchell and L. Howe, ‘Yet More Facts about the Insanity Defence’ [2006]

Crim LR 399. 25 See A. Loughnan, ‘“Manifest Madness”: Towards a New Understanding of the Insanity

Defence’ (2007) 70 MLR 379, proposing a reinterpretation of the exceptional procedural and evidential provisions relating to insanity. 26 T. H. Jones, ‘Insanity, Automatism and the Burden of Proof on the Accused’ (1995) 111 LQR

475; a challenge to this under Art. 6.2 of the Convention now seems unlikely to succeed, following the Court of Appeal's decision in Lambert, Jordan and Ali [2001] 1 Cr App R 205 to uphold the reverse onus in diminished responsibility. 27 Criminal Procedure (Insanity) Act 1964, s. 6. 28 Per Watkins LJ, in Dickie (1984) 79 Cr App R 213, at 219. 29 (1843) 10 Cl and Fin 200; see generally N. Morris, Madness and the Criminal Law (1982),

and I. Potas, Just Deserts for the Mad (1982). 30 Clarke (1972) 56 Cr App R 225. 31 See paragraph (d). 32 Per Lord Diplock, in Sullivan [1984] AC 156. 33 Sullivan [1984] AC 156. 34 Burgess [1991] 2 QB 92, overlooked in the rape case of Bilton, Daily Telegraph, 20

December 2005. See also the decision of the Canadian Supreme Court in Parks (1990) 95 DLR (4th) 27. 35 Hennessy (1989) 89 Cr App R 10. 36 See, more fully, Chapter 4.2. 37 The fact that these limbs of M'Naghten are alternatives ought to mean that insanity may be

a defence to strict liability crimes too, since the second test is applicable there, and a

Page 43 of 57

Criminal Capacity, Mens Rea, and Fault Divisional Court ruling to the contrary is difficult to support: DPP v H [1997] 1 WLR 1406, analysed critically by T. Ward, ‘Magistrates, Insanity and the Common Law’ [1997] Crim LR 796. 38 Johnson [2008] Crim LR 132, applying Windle [1952] 2 QB 826, which had been followed by

the majority of the Supreme Court of Canada in Schwartz (1979) 29 CCC (2d) 1. 39 Stapleton v R (1952) 86 CLR 358; research evidence from R. Mackay and G. Kearns (‘More

Facts about the Insanity Defence’ [1999] Crim LR 714), and Mackay, Mitchell, and Howe (‘Yet More Facts about Insanity’, at 406–7) shows that many psychiatrists interpret ‘wrongness’ in this wider sense, and that courts seem to accept this. 40 R. D. Mackay and G. Reuber, ‘Epilepsy and the Defence of Insanity—Time for Change?’

[2007] Crim LR 782; R. D. Mackay and B. J. Mitchell, ‘Sleepwalking, Automatism and Insanity’ [2006] Crim LR 901; see also the discussion on automatism in Chapter 4.2(f). 41 A Convention challenge to the M'Naghten Rules has already met with some success in

Jersey: R. D. Mackay and C. A. Gearty, ‘On Being Insane in Jersey’ [2001] Crim LR 560. 42 2 EHRR 387 (1979). 43 P. J. Sutherland and C. A. Gearty, ‘Insanity and the European Court of Human Rights’ [1992]

Crim LR 418. 44 E. Baker, ‘Human Rights, M'Naghten and the 1991 Act’ [1994] Crim LR 84. 45 A new s. 5 of the 1991 Act, inserted by s. 24 of the Domestic Violence, Crime, and Victims

Act 2004, gives courts a choice of orders following the special verdict in a murder case. If D is suffering from a mental disorder making detention in hospital appropriate under s. 37 of the Mental Health Act 1983, the judge has the power to make a hospital detention order. If the judge finds that D poses a threat of serious harm to the public, the judge can make an additional ‘restriction’ order, imposing a minimum time period before D can be considered for discharge from hospital: Criminal Procedure (Insanity) Act 1964, s.5 (as amended). 46 Homicide Act 1957, s. 2; see Chapter 7.4(e). 47 Model Penal Code, s. 4.01. 48 Butler Report, para. 18.30. 49 Cf. Scottish Law Commission, Report on Insanity and Diminished Responsibility (2004),

paras. 2.52–63, rejecting any volitional component in the insanity defence. 50 Having said that, schizophrenia is in fact the most common basis for an insanity plea: R. D.

Mackay, B. J. Mitchell, and L. Howe, ‘Yet More Facts About the Insanity Defence’ [2006] Crim LR 399. 51 K. W. M. Fulford, ‘Value, Action, Mental Illness, and the Law’, in S. Shute, J. Gardner, and J.

Horder (eds), Action and Value in Criminal Law (1993). 52 Law Com No. 177, cll. 34–40. 53 Page 44 of 57

Criminal Capacity, Mens Rea, and Fault 53 Law Com No. 177, para. 11.27. See now the Law Commission Scoping Paper, Insanity and

Automatism (July 2012). 54 Law Com No. 177, ii, para. 11, 28. 55 See Chapter 3.6(s). 56 Law Com No. 177, para. 11.28(c). 57 Salomon v Salomon [1897] AC 22. 58 Birmingham and Gloucester Railway Co (1842) 3 QB 223. 59 Great North of England Railway Co (1846) 9 QB 315. 60 Per Denham CJ at 320. 61 Another landmark case was Mousell Bros v London and North-Western Railway Co [1917]

2 KB 836. For discussion of the history see L. H. Leigh, The Criminal Liability of Corporations in English Law (1969), ch 2, and C. Wells, Corporations and Criminal Responsibility (2nd edn., 2001), ch 5. 62 See A. Ashworth, ‘Is the Criminal Law a Lost Cause?’ (2000) 116 LQR 225, and the

discussion by D. Nelken, ‘White Collar and Corporate Crime’, in M. Maguire, R. Morgan, and R. Reiner (eds), The Oxford Handbook of Criminology (5th edn., 2012). 63 T. Jones, D. Maclean, and J. Young, The Islington Crime Survey; M. Gottfredson, Fear of

Crime (Home Office Research Study No. 84, 1986). 64 Associated Octel [1996] 1 WLR 1543, a decision of the House of Lords. See also Gateway

Foodmarkets Ltd [1997] Crim LR 512, imposing a duty under s. 2 of the Act on the employer in respect of the acts of all employees, not just those who were ‘controlling minds’. 65 [1995] 1 WLR 1356; compare the much more restrictive approach in Seaboard Offshore Ltd

v Secretary of State for Transport [1994] 1 WLR 541. 66 See Birmingham and Gloucester Railway Co (reference at n 67) and the discussion of strict

liability in section 5.5(a). 67 See the conviction under s. 85 of the Water Resources Act 1991 upheld in Environment

Agency v Empress Car Co (Abertillery) [1999] 2 AC 22, criticized in the next section. 68 P. S. Atiyah, Vicarious Liability in the Law of Torts (1967). 69 See, ‘Developments in the Criminal Law – Corporate Crime: Regulating Corporate Behaviour

Through Criminal Sanctions’ (1979) 92 Harvard Law Review 1227. 70 Cf. Allen v Whitehead [1930] 1 KB 211 with Vane v Yiannopoullos [1965] AC 486; see P. J.

Pace, ‘Delegation: A Doctrine in Search of a Definition’ [1982] Crim LR 627. 71 The doctrine was criticized by the Law Commission in, Consultation Paper No. 195 (2010),

Criminal Liability in Regulatory Contexts, paras 7.35–57. In St Regis Paper Company Ltd

Page 45 of 57

Criminal Capacity, Mens Rea, and Fault [2011] EWA Crim 2527, the delegation principle was confined to licensing cases. Modern legislation sometimes employs a different structure to deal with such situations, by (very broadly speaking) making it an offence for the owner or licensee, etc. to allow unauthorized activity to go on: see e.g. the Licensing Act 2003, s. 136. 72 [1898] 2 QB 306. 73 It could be argued that this is not an example of vicarious liability because the owner does

the actus reus himself since he is the seller. However, the physical act is that of his employee, so at least it is a form of quasi-vicarious liability. 74 See n 65 and accompanying text. 75 [1955] 1 QB 78. 76 [1944] KB 146. 77 Two other cases decided in the same year confirmed this approach: ICR Haulage Ltd

[1944] KB 551, and Moore v I Bresler Ltd [1944] 2 All ER 515. 78 [1972] AC 153. 79 Cf. JF Alford (Transport) Ltd [1997] 2 Cr App R 326 (Chapter 10.3(b)), and the conviction of

manslaughter of a small outdoor pursuit company and its managing director in respect of the deaths of young canoeists sent out in poor weather with inadequate training and supervision (OLL Ltd and Kite, The Times, 9 December 1994), with the difficulty of identification in Redfern [1993] Crim LR 43. 80 [1995] 2 AC 500. 81 [2000] 2 Cr App R 207. 82 [2000] 2 Cr App R 211. 83 See Wells, Corporations and Criminal Responsibility, ch 3, and G. R. Sullivan, ‘The

Attribution of Culpability to Limited Companies’ (1996) 55 Camb LJ 515. 84 Cf. S. Lukes, Individualism (1973), ch 17. 85 [1972] AC 824; cf. the early decision in Birmingham and Gloucester Railway Co (1842) 3

QB 223. 86 Wells, Corporations and Criminal Responsibility (2nd edn., 2001), 160–3; cf. J. Gobert and

M. Punch, Rethinking Corporate Crime (2003), ch 8, with their discussion of companies as accomplices in ch 2. 87 [1972] AC 153. 88 See n 80. 89 Wells, Corporations and Criminal Responsibility, 164–8.

90 Page 46 of 57

Criminal Capacity, Mens Rea, and Fault 90 B. Fisse and J. Braithwaite, Corporations, Crime and Accountability (1993). 91 See the analysis by G. R. Sullivan, ‘Expressing Corporate Guilt’ (1995) 15 OJLS 281. 92 See the detailed discussion in Chapter 7.5(c). 93 Fisse and Braithwaite, Corporations, Crime and Accountability; cf. Gobert and Punch,

Rethinking Corporate Crime, ch 7. 94 Cf. A. Brudner, ‘Agency and Welfare in Criminal Law’, in S. Shute, J. Gardner, and J. Horder

(eds), Action and Value in Criminal Law (1993), and R. Lippke, Rethinking Imprisonment (2007), 84–98. 95 D. A. J. Richards, ‘Rights, Utility and Crime’, in M. Tonry and N. Morris (eds), Crime and

Justice: An Annual Review (1981), iii, 274. 96 Notably those of Bentham: for extracts and discussion see von Hirsch, Ashworth, and

Roberts, Principled Sentencing (3rd edn., 2009), ch 2. 97 For further argument, compare A. Ashworth, ‘Taking the Consequences’, in Shute, Gardner,

and Horder (eds), Action and Value in Criminal Law (1993) and V. Tadros, Criminal Responsibility (2005), 90–8, with R. A. Duff, Criminal Attempts, ch 12, and R. A. Duff, ‘Whose Luck is it Anyway?’, in C. Clarkson and S. Cunningham (eds), Criminal Liability for NonAggressive Death (2008). 98 See Chapter 3.6(r); compare the different framing of the same debate in German and

French laws: J. R. Spencer and A. Pedain, ‘Strict Liability in Continental Criminal Law’, in A. P. Simester (ed.), Appraising Strict Liability (2005), 275–81. 99 J. Gardner, ‘On the General Part of the Criminal Law’, in R. A. Duff (ed.), Philosophy and the

Criminal Law (1998), 243. 100 Gardner, ‘On the General Part of the Criminal Law’, at 244; see also Chapter 3.6(r). 101 J. Gardner, ‘Rationality and the Rules of Law in Offences against the Person’ (1994) 53

Camb LJ 502, at 509. 102 J. Gardner, Offences and Defences (2007), 246–7, replying to A. Ashworth, ‘A Change of

Normative Position: Determining the Contours of Culpability in Criminal Law’ (2008) 11 New Crim LR 232. 103 An early general statement was that of Lord Kenyon CJ in Fowler v Padget (1798) 7 Term

Rep 509. 104 [1969] 1 QB 439. 105 E.g. in rape (Kaitamaki v R [1985] 1 AC 147, now confirmed by the Sexual Offences Act

2003, s. 79(2)) and in theft (on appropriation, Hale (1978) 68 Cr App R 415). Cf. the critique by M. Kelman, ‘Interpretive Construction in the Substantive Criminal Law’ (1981) 33 Stanford LR 591, and the defence by M. Moore, Act and Crime (1993), 35–7. 106 [1983] 2 AC 161, discussed in Chapter 4.4. 107 Page 47 of 57

Criminal Capacity, Mens Rea, and Fault 107 [1954] 1 WLR 228. 108 [1966] 1 QB 59. 109 [1992] QB 61. 110 Cf. the felony-murder rule and constructive manslaughter, section 5.4(b). 111 [1981] 1 WLR 705. 112 An intention merely to assault would suffice: see the discussion in 5.4(b). 113 See Attorney-General for Northern Ireland v Gallagher [1963] AC 349, and Chapter 6.2. 114 See Coroners and Justice Act 2009, s. 56(6), indicating that, for the purposes of the loss of

self-control defence, fear of serious violence or a justifiable sense of being seriously wronged will be disregarded if, ‘caused by a thing which D incited to be done or said for the purpose of providing an excuse to use violence’. For the position under the old law, see Edwards v R [1973] AC 648, and Johnson (1989) 89 Cr App 148. See also Chapter 6.7. 115 P. H. Robinson, ‘Causing the Conditions of One's Own Defence: A Study in the Limits of

Theory in Criminal Law Doctrine’ (1985) 71 Virginia LR 1. 116 Cf. the different wording in the Draft Criminal Code (Law Com No. 177) on automatism (cl.

33(1)(b) and on duress (cl. 42(5)), for example. 117 See Chapter 6.3(c). 118 Hasan [2005] 2 AC 467, and Chapter 6.5(c). 119 Robinson, ‘Causing the Conditions of One's Own Defense’. 120 See the searching exploration of this topic in A. P. Simester (ed.), Appraising Strict Liability

(2005). 121 See the study by L. H. Leigh, Strict and Vicarious Liability (1982). 122 There is ample authority that automatism is a defence to strict liability offences, but some

disagreement on whether insanity may afford a defence: cf. Hennessy (1989) 89 Cr App R 10 with DPP v H [1997] 1 WLR 1406, discussed in section 5.2(c). 123 A. Ashworth, ‘Towards a Theory of Criminal Legislation’ (1989) 1 Criminal Law Forum 41. 124 B. Wootton, Crime and the Criminal Law (2nd edn., 1981) 47. Note that a few pages later

Baroness Wootton advocates ‘a wider concept of responsibility…in which there is room for negligence as well as purposeful wrongdoing’ (50), which is less an argument for strict liability than for negligence liability. Support for negligence liability is also found in the landmark decision of Sweet v Parsley, discussed in the text at n 154. 125 See, for example, the contributions of J. Horder and R. A. Duff in Simester (ed.), Appraising

Strict Liability. 126 Page 48 of 57

Criminal Capacity, Mens Rea, and Fault 126 Cf. the nuanced argument of G. Lamond, ‘What is a Crime?’ (2007) 27 OJLS 609, at 629–31,

suggesting that strict liability may have a proper role in regulating conduct that increases the risk of violations of significant public or private interests, as in road traffic law, health and safety, and food standards. 127 J. Braithwaite, Corporate Crime in the Pharmaceutical Industry (1984), ch 9. 128 Introduction to the Principles of Morals and Legislation, ch XIII. 129 See J. Horder, ‘Strict Liability, Statutory Construction and the Spirit of Liberty’ (2002) 118

LQR 458, at 472–4. 130 Sandhu [1997] Crim LR 288. 131 Cf. Lester (1976) 63 Cr App R 144 with Hill [1997] Crim LR 459. 132 A. Reiss, ‘Selecting Strategies of Social Control over Organizational Life’, in K. Hawkins and

J. M. Thomas (eds), Enforcing Regulation (1984). 133 The leading study is K. Hawkins, Law as Last Resort (2003). For shorter reviews, see G.

Richardson, ‘Strict Liability for Regulatory Crime: The Empirical Research’ [1987] Crim LR 295, and R. Baldwin, ‘The New Punitive Regulation’ (2004) 67 MLR 351. 134 See Hawkins, Law as Last Resort; B. S. Jackson, ‘Storkwain: a Case Study in Strict Liability

and Self-Regulation’ [1991] Crim LR 892, discussing the role of the Pharmaceutical Society in regulating pharmacists. 135 Cf. decisions such as Seaboard Offshore Ltd v Secretary of State for Transport [1994] 1

WLR 541, British Steel [1995] 1 WLR 1356, Associated Octel [1996] 1 WLR 1543, and Gateway Foodmarkets Ltd [1997] Crim LR 512. 136 Cf. the justifications for confining the English codification initiative to ‘traditional’ offences

by the Code Team (Law Com No. 143, paras. 2.10–13 and Appendix A) and by the Law Commission (Law Com No. 177, paras. 3.3–6), with the critical remarks of C. Wells, ‘Restatement or Reform’ [1986] Crim LR 314 and A. Ashworth, ‘Is the Criminal Law a Lost Cause?’ (2000) 116 LQR 225. 137 Alphacell Ltd v Woodward [1972] AC 824, following the notion of ‘quasi-crimes’ outlined by

Lord Reid in Sweet v Parsley [1970] AC 132. 138 Model Penal Code, s. 6.02(4). For discussion in the context of the US Constitution see A.

Michaels, ‘Constitutional Innocence’ (1999) 122 Harv LR 829. 139 References re Section 94(2) of the Motor Vehicles Act (1986) 48 CR (3d) 289; see D. R.

Stuart, Canadian Criminal Law (4th edn., 2001). 140 Whether the presumption of innocence ought to have any implications for strict criminal

liability is a matter that has been debated extensively. See e.g. the contribution by Sullivan in Simester (ed.), Appraising Strict Liability; V. Tadros and S. Tierney, ‘The Presumption of Innocence and the Human Rights Act’ (2004) 67 MLR 402; and A. Ashworth, ‘Four Threats to the Presumption of Innocence’ (2006) 123 SALJ 62. 141 Page 49 of 57

Criminal Capacity, Mens Rea, and Fault 141 [2000] 2 AC 428. 142 Per Lord Nicholls at 460. 143 Prince (1875) LR 2 CCR 154; cf. R. Cross, ‘Centenary Reflections on Prince's Case’ (1975)

91 LQR 520 and J. Horder, ‘How Culpability Can, and Cannot, Be Denied in Under-age Sex Crimes’ [2001] Crim LR 15. 144 Per Lord Steyn at 470, borrowing the expression from Sir Rupert Cross, Statutory

Interpretation (3rd edn., 1995, by Bell and Engle), at 166. 145 [2002] 1 AC 462. 146 Lord Bingham at 17, Lord Steyn at 32. Lord Bingham made similar remarks in the

recklessness case of G [2004] 1 AC 1034, discussed in 5.5(c). 147 See the discussion of sexual offences in Chapter 8.5. 148 [2000] 2 AC at 463–4. 149 In addition to Howells and Muhamad, see also decisions such as Storkwain (n 158(),

Gammon v Attorney-General for Hong Kong [1985] AC 1 and R v Wells Street Magistrates’ Court and Martin, ex p Westminster City Council [1986] Crim LR 695. 150 [1985] AC 1 at 14. 151 See Lim Chin Aik v R [1963] AC 160. 152 For an example, see Harrow LBC v Shah [2000] 1 WLR 83 (selling lottery tickets to a

person under 16: strict liability as to age approved so as to help enforcement). 153 [2008] 1 Cr App R 25. 154 [2008] UKHL 37. 155 For fuller discussion of the decision, see Chapter 8.6(a). 156 E.g. Muhamad [2003] QB 1031 (offence of materially contributing to insolvency by

gambling); Matudi [2003] EWCA Crim 697 (offence of importing products of animal origin, not citing either B v DPP or K). 157 As in Deyemi and Edwards (reference at n 153). See also the speculation of Colin

Manchester on how the Licensing Act 2003 will be interpreted, in ‘Knowledge, Due Diligence and Strict Liability in Regulatory Offences’ [2006] Crim LR 213. 158 See now, Law Commission, Criminal Liability in Regulatory Contexts (CP No. 195, 2010), Pt

6. 159 See further R. A. Duff, Punishment, Communication and Community (2001), 149–51. 160 In that respect, ‘mens rea’ is a term narrower in scope than the term ‘fault element’,

because the latter clearly includes negligence. 161 Page 50 of 57

Criminal Capacity, Mens Rea, and Fault 161 Burglary is discussed in Chapter 9.5. On intent-based crimes see generally A. Ashworth,

‘Defining Criminal Offences without Harm’, in P. F. Smith (ed.), Criminal Law: Essays in Honour of J. C. Smith (1987), and J. Horder, ‘Crimes of Ulterior Intent’, in A. P. Simester and A. T. H. Smith (eds), Harm and Culpability (1996). 162 See the discussion of Steane in n 184 and accompanying text. 163 See Chapter 7.3(c). 164 For further study see R. A. Duff, Intention, Agency and Criminal Liability (1990), chs 3, 4,

and 6, critically discussed on this point by A. P. Simester, ‘Paradigm Intention’ (1992) 11 Law and Philosophy 235. 165 LCCP 177, A New Homicide Act?, paras. 4.36–37, adopting the argument of A. Khan,

‘Intention in Criminal Law: Time to Change?’ (2002) 23 Statute LR 235. 166 [1976] QB 1. 167 R. Cross, ‘The Mental Element in Crime’ (1967) 83 LQR 215. 168 An exception involves cases in which someone tries to do something in order to show that

it is impossible to achieve, such as seeking to jump over the roof of a very high building. Here, there may be an attempt without an intention to succeed, but such cases are rare. If such a case arose in a criminal context, it would not count as a criminal attempt, because there was no ‘intent to commit the offence’, as required by the Criminal Attempts Act 1981, s. 1(1). 169 Compare J. Finnis, ‘Intention and Side-Effects’, in R. G. Frey and C. W. Morris, Liability and

Responsibility (1991), 32, with A. P. Simester, ‘Why Distinguish Intention from Foresight?’, in Simester and Smith (eds), Harm and Culpability (1996). 170 Bentham, Introduction to the Principles of Morals and Legislation, ch VIII, on direct and

oblique intent. Bentham's definition of oblique intent was wider than that described here, a point discussed by Glanville Williams, ‘Oblique Intent’ (1987) 46 Camb LJ 417. 171 Bentham, Introduction to the Principles of Morals and Legislation, ch VIII. 172 See the discussion of ‘ordinary language’ in the next section. 173 The phrase of Lord Lane CJ, in Nedrick (1986) 83 Cr App R 267. 174 Duff, Intention, Agency and Criminal Liability, ch 3. 175 For extensive discussion, see I. Kugler, Direct and Oblique Intention in the Criminal Law

(2002). 176 Law Com No 304, Murder, Manslaughter and Infanticide (2007), para. 3.27. 177 Chapter 7.3(c). 178 [1985] AC 905. 179 [1986] AC 455. 180 Page 51 of 57

Criminal Capacity, Mens Rea, and Fault 180 (1986) 83 Cr App R 267. 181 [1999] AC 82. 182 Applied by the Court of Appeal in Matthews and Alleyne [2003] 2 Cr App R 30, although

Rix LJ commented that ‘there is very little to choose between a rule of evidence and one of substantive law’. 183 Commonly discussed examples involve situations of emergency, where a defence of

necessity is unavailable because the defence has no application, as in murder cases. Consider an example in which D and her baby are trapped by an advancing fire at the top of a high building, and D throws the baby off the edge in the vain hope that someone below may by a miracle catch the baby. In such a case, even if D foresaw the baby's death as certain to occur as a result of her action, a court might not infer from that that D intended to kill the baby by that action. 184 [1947] KB 997. 185 [1986] AC 112. 186 See the discussion in Chapter 4.8(b). 187 [1960] 2 QB 423; see also Yip Chiu-Cheung [1995] 1 AC 111, discussed in Chapter 11.5. 188 [1964] AC 763. 189 N. Lacey, ‘A Clear Concept of Intention: Elusive or Illusory?’ (1993) 56 MLR 621. 190 A. Norrie, Crime, Reason and History (2nd edn., 2001), 58. 191 [1999] AC 82, at 90; see the observations of V. Tadros, ‘The System of the Criminal Law’

(2002) 22 LS 448, at 451–5. 192 Law Com No. 304, para. 3.27; see generally paras. 3.18–26. 193 See Chapter 5.5(c)(i). 194 Expressly preserved by cll. 4(4) and 45(4) of the draft Criminal Code. See now Re A

(Conjoined Twins: Surgical Separation) [2000] 4 All ER 961, discussed in Chapter 7.2; and more generally, A. Ashworth, ‘Criminal Liability in a Medical Context: the Treatment of Good Intentions’, in Simester and Smith (eds), Harm and Culpability (1996). 195 This was one of the views expressed in Hyam v DPP [1975] AC 55, by Lord Diplock (not

dissenting on this point); see J. Buzzard, ‘Intent’ [1978] Crim LR 5, with reply by J. C. Smith at [1978] Crim LR 14. 196 See Andrews v DPP [1937] AC 576 and Adomako [1995] 1 AC 171, discussed in Chapter

7.5. 197 Cunningham [1957] 2 QB 396, adopting the definition offered by C. S. Kenny, Outlines of

Criminal Law (1st edn., 1902; 16th edn., 1952). 198 Page 52 of 57

Criminal Capacity, Mens Rea, and Fault 198 Caldwell [1982] AC 341, and Lawrence [1982] AC 510. 199 In G. [2004] 1 AC 1034. 200 Chapter 7.5(b). 201 A further meaning of recklessness was adopted in sex cases (see Kimber [1983] 1 WLR

1118, Satnam S and Kewal S (1984) 78 Cr App R 149), but the enactment of the Sexual Offences Act 2003 relegates this to a matter of historical interest only. 202 See n 197. 203 Model Penal Code, s. 2.02(s)(c). 204 Law Com No 304, paras. 3.36–40, relating to first degree murder and to reckless murder

(second degree); for further discussion, see Chapter 7.3(c). 205 Criminal Law Revision Committee, 14th Report, Offences against the Person (1980), 8. 206 D. J. Galligan, ‘Responsibility for Recklessness’ (1978) 31 CLP 55, at 70. 207 Norrie, Crime, Reason and History, 78–80. For example, prosecutions for offences of

recklessness have been unusual in respect of large-scale transportation disasters. 208 See section 5.4(a). 209 For an analysis of fault in such terms, see e.g. Alan Brudner, ‘Agency and Welfare in the

Penal Law’, in S. Shute, J. Gardner, and J. Horder (eds), Action and Value in Criminal Law (1993). 210 [1979] QB 695. 211 [1977] 1 WLR 600. 212 See the discussion by Geoffrey Lane LJ in Stephenson [1979] QB 695, and M. Wasik and M.

P. Thompson, ‘Turning a Blind Eye as Constituting Mens Rea’ (1981) 32 NILQ 328, at 339. In Booth v CPS (2006) 170 JP 305 the Divisional Court upheld a finding of recklessness on the basis that D had ‘closed his mind’ to the obvious. 213 Duff, Intention, Agency and Criminal Liability, 162–3. 214 The Sexual Offences Act 2003 alters the definition of rape in this direction: by introducing

a reasonableness test of belief in consent, it ensures that practically indifferent defendants should be convicted. See Chapter 8.5. 215 See G. Williams, ‘The Unresolved Problem of Recklessness’ (1988) 8 Legal Studies 74, at

82. 216 [1982] AC 341. The case of Lawrence (reference at n 198) was decided on the same day. 217 [2004] 1 AC 1034. 218 For a fuller discussion, see the 4th edition of this work, 183–7. 219 Page 53 of 57

Criminal Capacity, Mens Rea, and Fault 219 See e.g. Elliott v C. (1983) 77 Cr App R 103 (mentally handicapped girl of 14), Stephenson

[1979] QB 695 (man with schizophrenia). 220 [2004] 1 AC 1034. 221 Per Lord Bingham at para. 38. For substantive argument, see V. Tadros, ‘Recklessness and

the Duty to Take Care’, in S. Shute and A. P. Simester (eds), Criminal Law Theory: Doctrines for the General Part (2002), at 255–7. 222 For the detailed history of the rise and fall of Caldwell recklessness, and suggestions for

further development, see A. Halpin, Definition in the Criminal Law (2004), ch 3. 223 On this point one may compare German criminal law, which also adopts this broader form

of recklessness with a capacity exception—the question being whether D would or should have foreseen the risk, given his intellectual capacities and knowledge at the time. See J. R. Spencer and A. Pedain, ‘Strict Liability in Continental Criminal Law’, in A. P. Simester (ed.), Appraising Strict Liability (2005), 241. 224 (1986) 83 Cr App R 155; see also the draft Criminal Code, Law Com No. 177, cl. 18(a). 225 Westminster City Council v Croyalgrange Ltd (1986) 83 Cr App R 155. 226 See Moloney [1985] AC 905; cf. Woollin [1999] 1 AC 82 (reference at n 181). 227 The classic statement is that of Devlin J, in Roper v Taylor's Garages Ltd [1951] 2 TLR 284,

at 288. 228 Contrast the differing views of Shute, 196–8 (reference at n 239), and G. R. Sullivan,

‘Knowledge, Belief and Culpability’ and V. Tadros, ‘Recklessness and the Duty to Take Care’, in Shute and Simester (eds), Criminal Law Theory, at respectively 213–14 and 252–4. 229 See Chapter 7.5(c). 230 Some examples are collected at [1980] Crim LR 1. 231 See section 5.4(a). 232 J. Gardner, ‘Introduction’, summarizing the views of H. L. A. Hart, Punishment and

Responsibility (2nd edn., 2008). 233 Discussed in Chapter 4.2. 234 This is the argument of Hart, Punishment and Responsibility, chs 2 and 5. The absence of

an incapacity exception was a major argument against the Caldwell test. 235 See the detailed discussion by M. S. Moore, Placing Blame (1997), ch 9 and 588–92. 236 A. P. Simester, ‘Can Negligence be Culpable?’, in J. Horder (ed.), Oxford Essays in

Jurisprudence (4th Series) (2000), at 106. 237 Robinson and Darley, Justice, Liability and Blame, 123.

238 Page 54 of 57

Criminal Capacity, Mens Rea, and Fault 238 See section 5.4(f). 239 A. Ashworth, ‘Is the Criminal Law a Lost Cause?’ (2000) 116 LQR 225. 240 See Chapter 3.2. 241 As now adopted in the Sexual Offences Act 2003, discussed in Chapter 8.5. 242 On this and other points, see J. Horder, ‘Gross Negligence and Criminal Culpability’ (1997)

47 U Toronto LJ 495. 243 See Chapters 7.6, 7.7, and 8.3(f). 244 See Chapter 1.3. 245 Some ‘subjectivists’ might accept a case for some criminal negligence liability, while

insisting that it is categorically different from liability based on choice: see Moore (reference at n 235). 246 See the discussion in section 5.5(c). 247 Cf. Gardner's view that ‘once we go beyond the paradigm of intention…the mentalities of

crime quickly fragment and lack any intelligible ordering’: J. Gardner, ‘On the General Part of the Criminal Law’, in R. A. Duff, Philosophy and the Criminal Law (1998), 231. 248 Norrie, Crime, Reason and History, 66. 249 See the discussion of specific offences in Chapter 8.3. 250 Cunningham [1957] 2 QB 396; Savage, Parmenter [1992] AC 699; section 5.3(c). 251 [1981] AC 394. 252 See J. A. Andrews, ‘Wilfulness: a Lesson in Ambiguity’ (1981) 1 Legal Studies 303. 253 Compare, e.g., James and Son v Smee [1955] 1 QB 78 with Baugh v Crago [1976] Crim LR

72. 254 In Chapter 9.2. 255 Public Order Act 1986, s. 4(1); see Chapter 8.3. 256 Public Order Act 1986, s. 19(1). 257 See section 5.2(b). 258 Cf. the facts of G. [2004] 1 AC 1034, where the two children set fire to paper beneath a

dustbin, and the ultimate result was damage to buildings costing around £1 million. 259 See Law Com No. 29, Offences of Damage to Property (1970), and the Criminal Damage

Act 1971. 260 Sexual Offences Act 2003, discussed in Chapter 8.5. 261 Page 55 of 57

Criminal Capacity, Mens Rea, and Fault 261 See A. Ashworth, ‘The Elasticity of Mens Rea’, in C. Tapper (ed.), Crime, Proof and

Punishment (1981), and Moore, Placing Blame, ch 11. 262 See the discussion of the principle of fair labelling in Chapter 3.6(s). 263 Law Com No. 177, ii, para. 8.57 (my italics). 264 The offences of fraud form an exception: see Chapter 9.7. 265 See Ashworth, ‘The Elasticity of Mens Rea’, 46–7. 266 See Ashworth, ‘The Elasticity of Mens Rea’, 47. 267 See Law Com No. 177, ii, para. 8.31, and Chapter 10.5(a). 268 Gross (1913) 23 Cox CC 455 (partial defence of provocation transferred). 269 See A. Ashworth, ‘Transferred Malice and Punishment for Unforeseen Consequences’, in P.

Glazebrook (ed.), Reshaping the Criminal Law (1978). 270 Pembliton (1874) 12 Cox CC 607. 271 As in the leading case of Latimer (1886) 17 QBD 359. 272 See Ashworth, ‘Transferred Malice and Punishment’, 89–93. 273 Cf. e.g., D. Husak, ‘Transferred Intent’ (1996) 10 Notre Dame J Law, Ethics and Public

Policy 65, with A. M. Dillof, ‘Transferred Intent: an Inquiry into the Nature of Criminal Culpability’ (1998) 1 Buffalo Crim LR 501. 274 [1998] 1 Cr App R 91. 275 See the discussion of this principle in section 5.4(c). 276 J. Horder, ‘Transferred Malice and the Remoteness of Unexpected Outcomes from

Intentions’ [2006] Crim LR 383. 277 In effect, cl. 24 of the Draft Criminal Code is a ‘deeming’ provision: see Law Com No. 177,

ii, paras. 8.57–59. 278 G. Williams, ‘Convictions and Fair Labelling’ [1983] CLJ 85. 279 Cf. the discussion of luck and results in section 5.4(b). 280 Cf. Horder, ‘Transferred Malice and Remoteness’, with A. P. Simester and G. R. Sullivan,

Criminal Law: Theory and Doctrine (3rd edn., 2007), 156–8.

Page 56 of 57

Excusatory Defences

Principles of Criminal Law (7th edn) Andrew Ashworth and Jeremy Horder Publisher: Oxford University Press Print ISBN-13: 9780199672684 DOI: 10.1093/he/9780199672684.001.0001

Print Publication Date: May 2013 Published online: Sep 2013

6. Excusatory Defences Chapter: (p. 193) 6. Excusatory Defences Author(s): Andrew Ashworth and Jeremy Horder DOI: 10.1093/he/9780199672684.003.0006

6.1 Excuses and other defences 6.2 Intoxication 6.3 Duress and coercion 6.4 Reasonable mistake and putative defences 6.5 Ignorance or mistake of law 6.6 Entrapment 6.7 Reviewing the non-permission-based defences Further reading

6.1 Excuses and other defences Criminal lawyers sometimes speak and write as if criminal guilt turns on the presence or absence of mens rea, but observations in previous chapters have already hinted that matters

Excusatory Defences are not so simple. The notions of fault and culpability go further and deeper than mens rea and require a discussion of other doctrines broadly termed ‘defences’. It is technically incorrect to use the term ‘defence’ when referring to the ‘defence of mistake’ or the ‘defence of accident’, since these (along with intoxication) are simply ‘failure of proof’ arguments; ‘mistake’ or ‘accident’ is merely a way of explaining why the prosecution has failed to prove the required knowledge, intention, or recklessness in respect of a particular ingredient of an offence.1 Defences of various kinds have been discussed in earlier chapters. The absence of a voluntary act (Chapter 4.2) is often referred to as the defence of automatism. The various permission-based defences, such as self-defence, were analysed in Chapter 4.6, 4.7, and 4.8. Defences of lack of capacity, particularly insanity, were discussed in Chapter 5.2. In the present chapter the focus is on excuses and potential excuses for wrongful acts, the essence of which is that D ‘lived up to normative expectations’ when responding to testing circ*mstances2 —typically duress (section 6.3), but also to some extent intoxication (section 6.2), reasonable mistake and putative defences (section 6.4), and mistake of law (section 6.5). Last, there (p. 194) is a miscellany of possible defences, chiefly entrapment (section 6.6), but also some forms of mistake of law (section 6.5) that are based on elements of excuse and other public-policy arguments. We will return, at the end of the chapter, to take stock of the various rationales.

6.2 Intoxication Research confirms that many of those who commit crimes of violence and burglary (at least) have taken some kind of intoxicant beforehand.3 Alcohol is probably the most widely used of intoxicants, but narcotic or hallucinogenic drugs are involved in some cases, too, and our discussion will relate to those who have taken alcohol, drugs, or a combination of the two. The usual effects are a loosening of inhibitions and, perhaps, a feeling of well-being and confidence. It is well known that people who have taken intoxicants tend to say or do things which they would not say or do when sober, and, in that sense, intoxicants may be regarded as the cause of such behaviour. But, as we saw in Chapter 5, the criminal law's conception of fault has tended to concentrate on cognition rather than on volition. Thus the approach to intoxication has not been to examine whether D's power to choose to cause the prohibited harm was substantially reduced, but has been to focus on its relation to mens rea. However, the law has been reluctant to allow intoxication simply to negate mens rea: instead, rather like its approach to automatism (see Chapter 4.2), it has drawn on arguments of prior fault and social defence in order to prevent the simple acquittal of those who cause harm and who lack awareness at the time because of intoxication. This, as we shall see, has caused various doctrinal difficulties for English criminal law. (a) The english intoxication rules From what was said earlier about the doctrine of prior fault,4 it is not surprising to find that a person who deliberately drinks himself into an intoxicated state in order to carry out a crime will have no defence. As Lord Denning declared in Attorney-General for Northern Ireland v Gallagher (1963): 5

If a man, whilst sane and sober, forms an intention to kill and makes preparation for it …

Page 2 of 44

Excusatory Defences and then gets himself drunk so as to give himself Dutch courage to do the killing, and whilst drunk carries out his intention, he cannot rely on this self-induced drunkenness as a defence to a charge of murder.

(p. 195) Cases such as this are rare. More frequent are cases in which D has become intoxicated ‘voluntarily’, i.e. where there is no reason to regard it as ‘involuntary’,6 and has then done something which, he argues, he would not have done but for the alcohol or drugs. It could be said that there is an ‘inexorable logic’7 that if mens rea is not present, whether through intoxication or otherwise, D should not be convicted. However, considerations of social protection have led the courts to introduce an unusual distinction. The decision in DPP v Majewski (1977)8 divides crimes into ‘offences of specific intent’ and ‘offences of basic intent’, and allows intoxication as a ‘defence’ (in the form of a denial of fault) to the former but not to the latter. Murder and wounding with intent are crimes of specific intent, and there is no great loss of social defence in allowing intoxication to negative the intent required for those crimes when the amplitude of the basic intent offences of manslaughter and unlawful wounding lies beneath them—ensuring D's conviction and liability to sentence. Various theories have been advanced in an attempt to explain why those offences (together with theft, handling, and all crimes of attempt, for example) are crimes of ‘specific intent’ whereas others are not, but none is satisfactory.9 For example, to assert (as did Hughes LJ in Heard (2008))10 that crimes of specific intent require proof of purpose is unconvincing, since that is not true of, for example, handling stolen goods (a specific intent crime, after Durante11). Moreover, many crimes contain some elements for which only intent will suffice and others for which recklessness or negligence is sufficient.12 However, this rather ramshackle law has proved workable. The courts have thus restricted the operation of the ‘inexorable logic’ of mens rea to the few offences of specific intent and, since most of them are underpinned by a lesser offence of ‘basic intent’, no great loss of social defence has occurred. The policy expressed in Majewski through the idea of ‘offences of basic intent’ was expressed slightly differently in Caldwell (1982)13 in terms of ‘recklessness’. Thus, where recklessness is a sufficient fault element for the crime, evidence of intoxication is irrelevant because anyone who was intoxicated is deemed to have been reckless. This is a simpler rule to apply, although it appears not to have displaced the Majewski test.14 It is subject to an exception, as we shall see in section 6.2(d), in cases where the intoxication can be regarded as to some degree ‘involuntary’. Section 6(5) of the Public Order Act 1986, which applies only to offences in that Act, reads as follows:

a person whose awareness is impaired by intoxication shall be taken to be aware of that of which he would be aware if not intoxicated, unless he shows either that his intoxication was (p. 196) not self-induced or that it was caused solely by the taking or administration of a substance in the course of medical treatment.

This provision, though couched in the terminology of awareness instead of advertent recklessness, may be thought to express the law's general approach.15 The effect of all these rules is that voluntary intoxication rarely functions as a ground of Page 3 of 44

Excusatory Defences exculpation. Here, as with automatism and some mistakes, the ‘logic’ of the standard doctrines of actus reus and mens rea has been subordinated to considerations of social defence. Thus where it is alleged that intoxication induced a state of automatism, the case is treated as one of intoxication (the cause) rather than automatism (the effect).16 The same approach has been quite vigorously pursued in cases of intoxicated mistake, bringing them under the rules of intoxication (the cause) rather than mistake (the effect). At common law, in O'Grady (1987),17 where the defence took the form of a drunken mistaken belief in the need for self-defence, the Court of Appeal held that D could not rely on his mistake if it stemmed from intoxication. This means, in effect, that where the normal subjective rule for mistake clashes with the objective rule for intoxication, the latter takes priority. The Court of Appeal confirmed this in Hatton (2006),18 declining an invitation to depart from O'Grady and confirming a conviction for murder of someone who made an intoxicated mistake in self-defence.19 The common law approach has now been enshrined in statute, by s. 76 of the Criminal Justice and Immigration Act 2008. In relation to a mistaken belief in the need to use force in self-defence or prevention of crime, ss. 76(4) and 76(5) say:

(4) If D claims to have held a particular belief as regards the existence of any circ*mstances— (a) the reasonableness or otherwise of that belief is relevant to the question whether D genuinely held it; but (b) if it is determined that D did genuinely hold it, D is entitled to rely on it for the purposes of subsection (3), whether or not— (i) it was mistaken, or (ii) (if it was mistaken) the mistake was a reasonable one to have made. (5) But subsection (4)(b) does not enable D to rely on any mistaken belief attributable to intoxication that was voluntarily induced.

The current law may have some logic behind it. It may be appropriate to take a different approach to the relevance of intoxication depending on whether D is denying fault, or claiming a (putative) defence to a crime respecting which he or she admits have possessed the fault element. However, the approach also involves a need for fine distinction-making of a kind that becomes more troublesome the more serious the (p. 197) crime in issue. For example, consider D's potential for liability for murder, with its mandatory sentence of life imprisonment, in the following examples: 1. D, a very strong individual, was so voluntarily intoxicated that he believed at the time that the heavy wooden pole with which he repeatedly struck V until V died was made of a lightweight wood that would not cause serious harm even when repeated blows were struck. 2. D was so voluntarily intoxicated that he mistakenly thought that V, who was walking towards D, intended to attack D there and then with lethal force, and so he struck V repeatedly with a heavy wooden pole, causing V's death. In example 1, D is denying that he or she had the fault element (intention to kill or to do serious

Page 4 of 44

Excusatory Defences harm). As murder is a crime of specific intent, D's state of voluntary intoxication—and hence D's explanation of his or her actions—may be considered by the jury in determining whether D had the fault element when striking V. So, D has a route to conviction for manslaughter only, on the basis of denial of the fault element for murder. By contrast, in example 2, D admits (we may assume) having the fault element for murder, but claims that—due to voluntary intoxication—he believed that he was acting in self-defence. Applying s. 76 of the 2008 Act, D will not be able to rely on his honest but mistaken belief that V was about to attack him with lethal force, and will accordingly have no route to manslaughter because (we are assuming) he admits having killed whilst having the fault element for murder. However, stepping back from the strict application of the law, one may ask whether there is enough, morally speaking, in the distinction between examples 1 and 2 to justify providing a route to manslaughter in the first, but not in the second. Assuming that they are both telling the truth, does D clearly deserve the mandatory life sentence in example 2, in a way that D in example 1 does not? There is a case for saying that both examples should be treated as instances of reckless killing, and therefore as a highly culpable form of manslaughter. (b) The attack on the english approach The approach of the English courts has been attacked on several grounds. The distinction between ‘specific intent’ and ‘basic intent’ is ill-defined, even if it does have some moral coherence. The approach of deeming intoxicated persons to be reckless rests on a fiction, and the attempts of Lord Elwyn-Jones in DPP v Majewski to argue that intoxicated persons really are reckless because ‘getting drunk is a reckless course of conduct’20 involve a manifest confusion between a general, non-legal use of the term ‘reckless’ and the technical, legal term, which denotes (for almost all offences)21 that D was aware of the risk of the result which actually occurred. In most cases it is (p. 198) far-fetched to argue that a person in the process of getting drunk is aware of the type of conduct he or she may later indulge in. These criticisms of the courts’ attempts to stretch the established meaning of ‘intent’ and of ‘recklessness’ in order to deal with the problems of intoxication have been joined by other arguments. Some have held that the intoxication rules are inconsistent with s. 8 of the Criminal Justice Act 1967, which requires courts to take account of all the evidence when deciding whether D intended or foresaw a result: 22 but the effect of DPP v Majewski is to deny that evidence of intoxication is relevant unless the crime is one of specific intent, and s. 8 extends only to legally relevant evidence.23 Another argument is that the intoxication rules are inconsistent with the principle of contemporaneity, in that they base D's conviction (of an offence of basic intent) on the antecedent fault of voluntarily taking intoxicants: 24 but the principle of contemporaneity itself conflicts with the doctrine of prior fault, as we have noted,25 and there seems no reason why contemporaneity should be an absolute principle. The question is whether it is appropriate to apply the rival doctrine of prior fault to intoxication cases. Whatever the merit of these criticisms, it is undeniable that the intoxication rules in English law rest on fictions and apparently illogical legal devices. Is it the policy of restricting the defence of intoxication which is wrong, or merely the legal devices used to give effect to the policy? (c) Intoxication, culpability, and social policy One may concede that, in fact, a person may be so drunk as not to know what he or she is Page 5 of 44

Excusatory Defences doing when causing harm to others or damage to property, and yet maintain that there are good reasons for criminal liability. What might these reasons be? At the root of the ‘social defence’ or ‘public protection’ arguments is the proposition that one of the main functions of the criminal law is to exert a general deterrent effect so as to protect major social and individual interests, and that any legal system which allows intoxication to negative mens rea would present citizens with an easy route to impunity. Indeed, the more intoxicated they became, the less likely they would be to be held criminally liable for any harm caused. As a matter of human experience, it is far from clear that this argument is soundly based. There are several common law jurisdictions which have declined to follow the English approach and which simply regard intoxication as one way of negativing mens rea.26 Two comments may be made (p. 199) here. First, these jurisdictions can be taken to be reinforcing the important and often neglected point that it is extremely rare for a defendant to be able to raise even a reasonable doubt that he was unaware of what he was doing. All that is required for proof of intent or recklessness is a momentary realization that property is being damaged or that a person is being assaulted, etc. Thus, even if evidence of intoxication were relevant, D's condition would not usually be acute enough to prevent conviction. Secondly, and alternatively, the rarity of acquittals based on intoxication in these jurisdictions may simply be because juries and magistrates are applying a normative test rather than a purely factual test. Thus the confidence of the majority judges in the High Court of Australia that juries and magistrates will not be too readily persuaded to acquit in these cases27 may derive less from the rarity of acutely intoxicated harm-doers than from a belief that the courts will simply decline to return verdicts of acquittal where D is regarded as unworthy or culpable in some general way. This would suggest that both the English and the Australian approaches are unsatisfactory in their method—the English because it deems intoxicated harm-doers to be ‘reckless’ when they are not, the Australian because it relies on juries to make covert moral assessments and not simply the factual assessment that the law requires—even if they usually produce socially acceptable outcomes. There remains the question of individual culpability. What distinguishes evidence of intoxication from many of the other explanations for D's failure to realize what most ordinary people would have foreseen is the element of prior fault. It was D's fault for taking drink or drugs to such an extent as to lose control over his behaviour. Does this mean that, in order to support a finding of culpability, it must be established that D knew of the likely effects of the intoxicants upon behaviour? Probably not, for it would be regarded as perfectly fair to assume that all people realize the possible effects of taking alcohol or drugs (apart from the exceptional situations to be discussed in paragraph (d)). ‘It is common knowledge that those who take alcohol to excess or certain sorts of drugs may become aggressive or do dangerous or unpredictable things.’28 This is plainly an objective standard, but it is so elementary that it should not be regarded as unfair on anyone to assume such knowledge. Thus there is an element of culpability in intoxication cases which serves to distinguish them not only from insanity cases (which arise without fault) but also from many cases of simple absence of mens rea. The point was put more strongly and more directly in early modern times, when temperance was regarded as a virtue and excessive drinking as an ‘odious and loathsome sin’.29 But in what does the culpability consist? Specifically, is D to blame for becoming intoxicated or for causing the proscribed harm? It is fairly simple to establish culpability for becoming intoxicated if there is no evidence that it was ‘involuntary’. It is fairly (p. 200) difficult to

Page 6 of 44

Excusatory Defences establish culpability for causing the proscribed harm if we follow normal principles: we must assume acute intoxication at the time of the act, and if we look back to the period when D was becoming intoxicated, it is unlikely that one could establish actual foresight of the kind of harm eventually caused. Perhaps some people who regularly assault others when drunk may realize that there is a risk of this occurring, but in order to encompass the majority of cases, it would be necessary to rewrite the proposition about ‘common knowledge’ so as to maintain that people should realize that, when intoxicated, they are likely to cause damage or to assault others. The culpability, in other words, is somewhat unspecific—as in many instances where prior fault operates to bar a defence.30 Sentencing decisions suggest that intoxication may mitigate on the first occasion it is raised, if the offence can be portrayed as ‘out of character’, but it will not mitigate any subsequent offences committed in an intoxicated state.31 (d) Voluntary and non-voluntary intoxication We have already noted that non-voluntary intoxication may constitute an exception to the general intoxication rules, and we saw that s. 6(5) of the Public Order Act 1986 recognizes an exception where the intoxication was not ‘self-induced’. There is, however, no sharp distinction between the voluntary and the non-voluntary: rather, there is a continuum of states in which D has more or less knowledge about the properties of what he is consuming. The English courts, consistently with their generally restrictive approach, have been reluctant to exempt defendants from the intoxication rules. Thus in Allen (1988),32 D's argument was that he had become intoxicated because he had not realized that the wine being given to him had a high alcohol content. The Court of Appeal held that, so long as a person realizes that he is drinking alcohol, any subsequent intoxication is not rendered non-voluntary simply because he may not know the precise strength of the alcohol he is consuming. In some circ*mstances this might be quite a harsh ruling, but in broad terms it is compatible with judicial statements about the unpredictability of alcohol. A different problem arose in Hardie (1985),33 where D took a quantity of Valium tablets ‘for his nerves’ and later set fire to an apartment. The Court of Appeal quashed his conviction. The main distinguishing factor here was that Valium is widely regarded as a sedative or soporific drug, and is not thought likely ‘to render a person aggressive or incapable of appreciating risks to others’. This suggests that one basis for the distinction between voluntary and nonvoluntary intoxication is a division of intoxicants into those that are sedative and others that may have aggressive effects. The Court in Hardie added that D would nonetheless be treated as reckless if he had known, contrary to (p. 201) general beliefs, that Valium might have disinhibiting rather than sedative effects.34 It should be noted that s. 6(5) of the Public Order Act 1986, set out earlier, allows D a defence where the intoxication ‘was caused solely by the taking or administration of a substance in the course of medical treatment’. In line with the general approach, this should be confined to cases where D was not warned of the possible effects, or where those effects were not widely known. The question of non-voluntary intoxication is raised most directly by Kingston (1995).35 The evidence suggested that certain sedative drugs had been introduced into D's coffee, and that he had then carried out indecent sexual acts on a sleeping boy. The Court of Appeal quashed D's conviction, holding that if D had been placed in an altered mental state by the stratagem of another, and this led him to form an intent that he would not otherwise have formed, he should have a defence. This approach accepts that D may have had the mental element required for

Page 7 of 44

Excusatory Defences the crime, but looks to the cause of that condition: in effect, a doctrine of prior lack of fault. The House of Lords restored the conviction. If the non-voluntary intoxication is so acute as to negative mens rea, then it may lead to an acquittal of any offence requiring mens rea, whether of specific or basic intent. Where non-voluntary intoxication is not so acute as to negative mens rea, Lord Mustill held that there is no basis for an acquittal unless the courts were to create a new defence.36 This the House was unwilling to do, because their Lordships could see no significant moral difference between this case and Allen, and because the opportunity for false defences was considerable. The matter was one for the Law Commission and Parliament. What, then, is the position? If the intoxicant is in the soporific category, it seems from Hardie that D may have a defence if he can show that he lacked the mental element required for the crime. The general rule would prevent evidence of intoxication being adduced to show that he was not reckless but, if the intoxication was non-voluntary, evidence of the intoxication should be admitted. However, where the intoxicant is not so powerful as to remove D's awareness of what he is doing, it seems immaterial whether it is in the soporific or the ‘aggressive’ category. Kingston holds that there is no defence available and D is therefore convicted on the basis of his intention or recklessness. Even if D can establish that the intoxicant was administered without his awareness—the ‘laced’ or ‘spiked’ drink37 —this appears insufficient to alter the analysis, even though one might think that this presents a stronger argument than Hardie. The House of Lords in Kingston overlooked D's absence of fault in bringing about the condition, and adopted an implausibly narrow view of excuses premised on the presence or absence of mens rea. For these reasons, the decision should be reversed, but it is doubtful whether one should go further, as G. R. Sullivan (p. 202) has argued, and allow courts to look to D's character and destabilized condition in order to determine whether or not he was blameworthy, and to find a defence if D is not adjudged blameworthy.38 (e) Finding a legal solution A simple solution compatible with the ordinary logic of the liability rules is to regard evidence of intoxication as relevant on issues of mens rea, following various decisions in New Zealand and in the non-Code states of Australia.39 There will only rarely be acquittals, and these may be regarded as part of a small price for respecting the principle of individual autonomy—like occasional acquittals of clumsy and thoughtless individuals. In practice the behaviour of most defendants who allege intoxication will show some elements of intention, knowledge, or awareness.40 However, an objection to the Antipodean approach is that it seems to yield the anti-social maxim ‘more intoxication, less liability’, and public outcries at certain acquittals have led some Australian states to abandon the simple ‘logical’ approach.41 It gives no weight to the elements of choice and risk involved in getting drunk. Usually the choice is to loosen one's self-restraint rather than to commit a crime, let alone a particular kind of crime; but the retention of control over one's behaviour might fairly be regarded as a social duty, and its abandonment as a form of wrongdoing.42 This argument may be weakened where D is addicted to alcohol or drugs, since the element of choice may have been exhausted long ago.43 In this country the various proposals for reform seem to fall into one of two different camps. On the one hand there are those who argue that the essence of the wrongdoing in most cases lies in becoming intoxicated, and that it is unfair to label a defendant as a certain kind of offender

Page 8 of 44

Excusatory Defences (wounding, indecent assault, etc.) if he really was so intoxicated as not to realize what he was doing. Along these lines was a Consultation Paper issued by the Law Commission in 1993, proposing that courts be allowed to take account of evidence of intoxication on any issue of fault (following the Antipodean approach), but also introducing a new offence of causing harm whilst intoxicated—a ‘state of affairs’ offence designed to achieve a measure of social defence without unfair labelling of the offender (in line with German law).44 However, the prevailing approach (p. 203) to reform is an adaptation of the common law. A Government version is to be found in a draft Bill of 1998: 45

For the purposes of this Act a person who was voluntarily intoxicated at any material time must be treated— • as having been aware of any risk of which he would have been aware had he not been intoxicated; and • as having known or believed in any circ*mstances which he would have known or believed in had he not been intoxicated.

The first part of this means that in most cases an intoxicated actor will be deemed reckless, which is not far from the present law and its distinction between specific and basic intent.46 The 1998 Bill also deals with intoxicants taken on medical advice, and includes a definition of voluntary intoxication. It does not provide a separate defence of involuntary intoxication, and indeed creates a presumption that intoxication was voluntary. More recently, the Law Commission has proposed a way of rationalizing the current law without using the confusing terms ‘specific’ and ‘basic’ intent.47 In its place, the Law Commission draws a distinction between crimes in which the fault element is an ‘integral’ element, and crimes in which the fault element is not an ‘integral’ element. Only in the former may voluntary intoxication be used as a way of denying the fault element. The Commission lists those species of fault the inclusion of which in the definition of the offence will make them ‘integral’ to the offence, meaning that the prosecution will have to prove that D had the fault element even if D denies having it on the grounds that he or she was voluntarily intoxicated:

Recommendation 3 (1) Intention as to a consequence; (2) Knowledge as to something; (3) Belief as to something (where the belief is tantamount to knowledge); (4) Fraud; (5) Dishonesty.

This proposal is to be welcomed in so far as it requires judges, when considering if a crime is one to which voluntary intoxication may be evidence of lack of fault, only to decide (p. 204) whether the fault element falls within one of the categories just mentioned. They would no

Page 9 of 44

Excusatory Defences longer have to wrestle with the application of the more abstract and indeterminate notions of ‘specific’ and ‘basic’ intent, and hence the risk of mis-categorizations—such as that in Heard48 —might be avoided more often. In making this advance, though, the Law Commission relies on another distinction that is not without difficulty, namely the distinction between crimes where the fault element is ‘integral’ to the offence, and crimes where it is not. Clearly, the notion of fault being ‘integral’ to the offence cannot mean simply that a fault element is expressly included in the offence; otherwise, the distinction would be doing no work. In its understanding of what is ‘integral’ to the offence, the Law Commission is alluding instead to the wrongdoing that underlies the offence. The idea is that, in some crimes—murder, theft, fraud, for example—the fault element is integral to the wrongdoing itself: morally speaking, there can simply be no ‘murder’ without intention, no ‘theft’ without dishonesty, and so on. So, the argument runs, it would be unfair to deny D the possibility of denying the fault element in such crimes (whatever the reason for the absence of the fault element), given how central the fault element is to the moral wrong underlying the crime in question. By contrast, it is said, in some crimes—criminal damage is an example—the fault element may be added to the definition, as a matter of fairness to the accused, but is not integral to the underlying wrong. One may question how helpful this distinction really is. For example, is handling stolen goods ‘knowing or believing’ them to be stolen a crime in which the fault element is integral to the wrong, or not? Happily, under the Law Commission's scheme, such troublesome theoretical questions would not have to be tackled in the courts, who would look for guidance directly to Recommendation 3. All of this might be described as ‘workable’, although it ignores the moral arguments made by those who favour the modified German-Antipodean approach to which the Law Commission was temporarily attached in 1993. The case for a purely subjective approach to intoxication seems unconvincing, but the arguments about how the intoxicated wrongdoer should be labelled and sentenced remain keenly contested.49

6.3 Duress and coercion This part of the chapter deals with cases in which D's behaviour fulfils the conduct element and the positive fault requirements of an offence, but in which D acted in response to threats from another person (sometimes called ‘duress per minas’), or in order to avert dire consequences (called ‘duress of circ*mstances’), or, unusually, in (p. 205) circ*mstances of marital coercion. We have already seen, in Chapter 4.8, that some cases of necessity might give rise to a claim that what would otherwise be a criminal action was permissible, but those are likely to be rare cases where there is a net saving of lives. Having said that, in focusing here on the excusatory defences of duress, we will find that the development of the common law has been characterized by confusions over whether, when D responds to pressure by committing an offence, D acts permissibly or only excusably. (a) Requirements of the defences The courts have generally held that the requirements of duress by threats and of duress of circ*mstances are in parallel.50 Although both defences require some danger external to D,51 they arise in different factual circ*mstances, and it might be best to illustrate this by contrasting two cases. In Hudson and Taylor (1971)52 two teenagers were prosecution

Page 10 of 44

Excusatory Defences witnesses at a trial for wounding. They testified that they did not know the man charged and could not identify him as the culprit. The man was acquitted, but the young women were charged with perjury. They admitted that they gave false evidence, but said that they were under duress, having been threatened with violence by various men, one of whom was in the public gallery at the original trial. The Court of Appeal quashed their convictions because the defence of duress had been wrongly withdrawn from the jury. In Conway (1989)53 two men approached D's car, whereupon D, urged on by his passenger, drove off at great speed and in a reckless manner. D's explanation was that he knew that his passenger had recently been threatened by two men who had fired a shotgun. D feared that these two men intended harm, and his driving was in response to that emergency. The Court of Appeal quashed the conviction for reckless driving because the trial judge had failed to leave the defence of duress of circ*mstances to the jury. A difference between these two cases is that for the defence of duress there should typically be a threat intended by the threatener to coerce D into committing a particular offence, whereas for duress of circ*mstances there will typically be a situation of emergency (involving perceived danger) that leads D to do something that would otherwise be an offence. However, there are additional subtle differences between the defences (although these may not have legal consequences):

1. In duress by threats cases, the crime D is to commit to avoid the threat will be specified by the threatener; in duress of circ*mstances cases, it is D who decides that the commission of a particular crime is the only reasonable course of conduct to take if the threat is to be avoided. (p. 206) 2. In duress by threats cases, one issue for D is a character assessment of the threatener in the shape of an assessment of is or her credibility: how likely is it that the threatener can or will implement the threat if D refuses to comply with the demand? In duress of circ*mstances, the equivalent issue for D is a risk assessment: how likely is it that the threat will have an impact on D in a way that makes commission of the crime the only reasonable course of action? 3. Unlike duress by threats, duress of circ*mstances overlaps with self-defence (if that defence is given a broad understanding). For example, Conway54 could be seen as a case in which D acted in self-defence rather than under duress, if selfdefence is deemed to include avoiding action that takes a form other than the use of force. If ducking to avoid something thrown at one's head is conduct that is selfdefensive in nature, D's conduct in Conway should be analysed in a similar way.

What, then, are the general requirements of the two defences? They appear to be restricted to cases where D acted as he did out of fear of death or serious injury.55 Threats to property or to reputation have been held to be insufficient,56 but there was a dictum in Steane (1947)57 that a threat of false imprisonment would suffice. In a sense it seems strange that the degree of threat or danger should be fixed in this way, since the seriousness of the crimes in respect of which duress is raised may vary considerably. A dire threat should be necessary to excuse a person who caused a grave harm, but it does not follow that some lesser threat should not be sufficient to excuse a lesser offence. However, the courts have continued to insist on threats of death or serious harm as the standard requirement and, additionally, that the threats must 58 Page 11 of 44

Excusatory Defences be such that ‘a sober person of reasonable firmness’ would not have resisted them.58 This objective condition has been tested in cases in which there have been attempts to introduce evidence to the effect that D's personality rendered him or her particularly susceptible to threats. In Bowen (1996)59 the Court of Appeal held that the question is whether D responded ‘as a sober person of reasonable firmness sharing the characteristics of the defendant would have done’. In applying this test a court should not admit evidence that D was more pliable, vulnerable, timid, or susceptible to threats than a normal person, and characteristics due to self-abuse (alcohol, drugs) should also be left out of account. But the Court did suggest that it would be proper to take account of age, sex, pregnancy, serious physical disability, or a ‘recognized psychiatric condition’.60 In view of the re-affirmation of the objective standard of self-control in loss of self-control (p. 207) cases,61 this broadening of the ambit of duress may appear anomalous. However, the comparison with the operation of a partial defence to murder (i.e. provocation) is not an apt one, since duress operates as a complete defence to offences other than murder. Whereas the objective standard in provocation can be maintained in the expectation that mentally disordered defendants will have resort to the partial defence of diminished responsibility, the possibility of relaxing the requirements of duress for mentally disturbed defendants can only be realized through a complete defence or conviction followed by mitigation of sentence.62 In some cases the question of mixed threats and mixed motives has arisen: in ValderramaVega (1985),63 a case in which D was both under severe financial pressure and subject to blackmail threats, the Court of Appeal held that duress would be available if the jury found that D would not have acted as he did but for the death threats he had received. The other pressures may have exerted an influence, but so long as the threats were causally significant, this was sufficient. The threats need not be addressed to D personally: 64 the defence is available if the threats are against D's family or friends, but it now seems that there must be some connection with D.65 In the unusual case of Shayler (2001),66 the defence was that D revealed official secrets because he believed that (unidentified) people were placed in danger by MI5's activities. The Court of Appeal held that, for the defence to be available, the threat or danger must be to D himself or ‘towards somebody for whom he reasonably regarded himself as being responsible’.67 This arguably introduces an unnecessary complication to the defence. Suppose that a bank robber threatens to shoot a bank customer or employee, unless an employee or customer volunteers to help with the robbery; D, a customer, agrees to help the robber. In this example, D is not in any way responsible for the welfare of any other customer saved by his or her decision to help the robber; yet, it would seem harsh to deny the defence to D. Would it really be held by the courts that D, being a customer, could not invoke the defence whereas, had a bank employee volunteered instead of a customer, the defence would have been available, because bank employees have a degree of responsibility for their customers? Such a distinction seems arbitrary. The threat must be ‘present’ and not a remote threat of future harm, but how long an interval may elapse? In Hudson and Taylor,68 the facts of which were outlined earlier, (p. 208) the Court of Appeal held that it is not necessary that the threat would be carried out immediately, so long as its implementation was imminent. The same approach was taken in Abdul-Hussain et al. (1999),69 where a group of Shiite Muslims from Iraq had hijacked an aircraft to Stansted airport. When they surrendered, they claimed that they had acted out of fear of persecution

Page 12 of 44

Excusatory Defences and death at the hands of the Iraqi authorities. The trial judge withdrew the defence from the jury on the ground that the threat was not sufficiently close and immediate, but the Court of Appeal held that imminence is sufficient and that the execution of the threat need not be immediately in prospect. However, in Hasan (2005)70 Lord Bingham opposed the drift towards the looser concept of ‘imminence’ and held that older authorities in favour of a requirement of immediacy should be restored: there is a duty to take evasive action where possible, particularly where the threat ‘is not such as [D] reasonably expects to follow immediately or almost immediately on his failure to comply with the threat’.71 Lord Bingham regarded Hudson and Taylor as wrongly decided, commenting that he could not accept ‘that a witness testifying in the Crown Court at Manchester has no opportunity to avoid complying with a threat incapable of execution there and then’. This strong line may, though, leave intact the concession to the defendants’ youth in Hudson and Taylor—‘having regard to his age and circ*mstances, and to any risks to him which may be involved in the course of action relied upon’.72 Having said that, more broadly, Lord Bingham's conception of duress evidently finds no place for those who cannot measure up to reasonable expectations. Another objective element is that the defendant is not entitled to be judged on the facts as he believed them to be. Contrary to the generally subjective approach to mistaken beliefs,73 the Court of Appeal in Graham (1982)74 held that the test for duress is whether, as a result of what D reasonably believed that the duressor had said or done, he had good cause to fear death or serious injury. Lord Lane offered no convincing reasons for departing from the subjective orthodoxy of the time, and in Safi (2003)75 the Court of Appeal appeared to favour a subjective approach, although the point was not argued to a clear conclusion.76 Both duress by threats and duress of circ*mstances are subject to the doctrine of prior fault.77 In Sharp (1987)78 D joined a gang of robbers participating in crimes where guns were carried, but when he tried to withdraw he was himself threatened (p. 209) with violence. The Court of Appeal held that the defence of duress is unavailable to anyone who voluntarily joins a gang ‘which he knows might bring pressure on him to commit an offence and was an active member when he was put under such pressure’. In Shepherd (1987)79 the Court added that: ‘there are certain kinds of criminal enterprises the joining of which, in the absence of any knowledge of propensity to violence on the part of one member, would not lead another to suspect that a decision to think better of the whole affair might lead him into serious trouble’. The doctrine of prior fault does not only operate in the context of joining criminal enterprises: it also applies where drug users become indebted to drug dealers who have a reputation for violence. The leading decision now is Hasan,80 where D associated with a man who was known to use violence, and who allegedly forced D (by threats) to carry out two burglaries. Lord Bingham held that there should be an objective test, based on the foreseeability of violence being threatened by the people with whom D was associating, and not requiring foresight of coercion to commit crimes of a particular kind: 81

The policy of the law must be to discourage association with known criminals, and it should be slow to excuse the criminal conduct of those who do so. If a person voluntarily becomes or remains associated with others engaged in criminal activity in a situation where he knows or ought reasonably to know that he may be the subject of compulsion by them or their associates, he cannot rely on the defence of duress to excuse any act which he is thereafter compelled to do by them.82

Page 13 of 44

Excusatory Defences

Thus the subjective element in Sharp, ‘which he knows …’, is superseded by the strongly objective approach running through Lord Bingham's speech in Hasan. However, Baroness Hale's speech favours the subjective approach in Sharp, and also argues that the foreseen threat must have been a threat to commit crimes rather than a general threat of violence.83 The latter point is surely right: the likelihood of being subjected to violence or threats thereof is different from the foreseeability of threats being used to force D to commit crimes, and the latter should be required. (b) Theoretical foundations for the defences Why should defences of duress be allowed? One argument is that acts under duress are permissible in the sense that they constitute a lesser evil than the carrying-out of the threat: the credentials of this rather narrow argument were discussed in Chapter 4.8. In general the courts have tended to mix arguments about permissions with those of excuse, without noticing the distinction. How strong are the arguments for excusing D rather than saying that the act is permissible? It is fairly clear that duress does not negative intent, knowledge, or recklessness: D will know only too well the nature and consequences of the conduct. It also seems unlikely that duress negatives the voluntary (p. 210) nature of D's conduct: the elements of unconsciousness and uncontrollability of bodily movements which are regarded as the hallmark of involuntary behaviour84 are not typically to be found in duress cases. Two separate rationales, with somewhat different implications, warrant further discussion—the first seeing duress as characterized by moral involuntariness, the second regarding it more as a reasonable response to extreme pressure. Although conduct in response to duress or necessity is not involuntary, it may be described as non-voluntary. The argument is that there is a much lower degree of choice and free will in these cases than in the normal run of actions. George Fletcher has termed this ‘moral or normative involuntariness’, arguing that the degree of compulsion in these cases is not significantly less than in cases of physical involuntariness.85 The phrases used by the Court of Appeal in Hudson and Taylor—‘effective to neutralize the will of the accused’, and ‘driven to act by immediate and unavoidable pressure’—have been repeated in many subsequent decisions. Even though ‘neutralizing the will’ puts it rather too strongly, the idea of moral involuntariness seems to encapsulate the approach of English judges, who also draw on a supposed analogy with loss of self-control. Full acceptance of the ‘moral involuntariness’ rationale might lead to an entirely subjective version of duress, in which the degree of pressure experienced by D would be the main issue. In fact English law imposes a standard of reasonable steadfastness, but of course that could be explained as a means of avoiding false defences (as courts and reform committees often state) rather than a rejection of the basic rationale.86 The weakness of the moral involuntariness account of duress is that it misdescribes most actions taken under duress, which are coerced but are neither involuntary nor non-voluntary. It is possible to imagine circ*mstances in which a threat could break D's will to resist. An example might be where D is subjected to prolonged and agonizing torture, and then threatened with a continuation of the torture unless he or she complies with a threatener's demand. In such a case, D may conceivably lack the will to resist, but such cases will be rare. An alternative rationale is to regard the successful duress defence as recognition that D responded in a reasonable way to the pressure of circ*mstances which involved extreme Page 14 of 44

Excusatory Defences danger. It is important not to read too much into the element of reasonableness here. It is not being claimed that D had a right to respond as he did, save perhaps in the small group of cases where a net saving of lives is in prospect.87 Duress usually operates as an excuse, recognizing the dire situation with which D was faced and limiting the defence to cases where D responded in a way that did not fall below the standard (p. 211) to be expected of the reasonable citizen in such circ*mstances. On this rationale the person of reasonable firmness assumes a central role, not so much in announcing a standard that should be followed, or reducing the risk of false defences, but rather in recognizing that D was not lacking in responsibility for what was done.88 D is excused for giving way to the threat or danger when resistance could not reasonably be expected in the circ*mstances—which means that selfsacrifice is required in certain (lesser) situations. There is thus a pale reflection of doctrines of self-defence, which requires a proportionate response to the threatened harm if D is to be acquitted.89 That leaves the issue of citizens who, for one reason or another, cannot attain the standard of reasonable firmness in these situations. We saw that in Bowen90 the Court of Appeal recognized a small group of conditions which might be allowed to modify the standard of reasonable firmness, whilst maintaining that the standard should be upheld for those falling outside that short list. Arguments of this kind have been encountered in other contexts.91 The ‘moral involuntariness’ rationale argues in favour of including these people in the defence of duress, on the basis of the severe reduction of their free will. But they fall outside the ‘reasonable response’ rationale, to which the standard of reasonable steadfastness is central, and so the most appropriate form of defence would ideally be one that rests on diminished capacity or extreme mental or emotional disturbance.92 A more radical approach would be to argue that, since there are so many questions of degree in duress cases (degree of threat, degree of immediacy, seriousness of crime), they are much more appropriate for the sentencing stage than the liability stage.93 On that view, the duress defences should be abolished altogether. At present English law takes the view (except in murder cases) that there is a point at which threats or an emergency may place so much pressure on an individual that it is unfair to register a conviction at all, so long as the individual does not fall below the standard of reasonable firmness, but that in lesser situations claims of duress sound only at the sentencing stage. Mitigation may be right if ‘desert’ is the basis for sentence, but supporters of deterrent sentencing have a particular problem. Their general approach is to maintain that the stronger the temptation or pressure to commit a crime, the stronger the law's threat should be in order to counterbalance it.94 The law and its penalties should be used to strengthen the resolve of those under pressure. Yet Bentham also accepted that criminal liability and punishment are inefficacious where a person is subject to such acute threats (e.g. death, serious injury) that the law's own threat cannot be expected (p. 212) to counterbalance it: in these cases, he said, there should be a complete defence.95 The difficulty with this analysis is that it suggests heavy deterrent sentences for all cases except the most egregious, where it prescribes no penalty at all—a distinction with momentous effects but no clear reference point. There is surely a sliding scale of intensity of duress and necessity. If, in the dire circ*mstances that confront D, he or she responds in such a way that one could not reasonably expect more of a citizen, then surely neither conviction nor punishment is deserved. Mitigation of sentence should be available for less extreme cases, to reflect strong elements of pressure that did not amount to the full defence.

Page 15 of 44

Excusatory Defences (c) Duress and the taking of life Although most of the elements of these defences seem to be based on a rationale of excusing a person's understandable submission to the threat, the troubled issue of whether the defences should be available to murder has led the courts to draw on permission-based rationales. The tone was set in the late nineteenth century with Dudley and Stephens (1884),96 where two shipwrecked mariners killed and ate a cabin-boy after seventeen days adrift at sea. Lord Coleridge CJ held that no defence of necessity (now called duress of circ*mstances) was available in a case of taking another person's life. In the first place, he argued, there is no necessity for preserving one's own life, and there are circ*mstances in which it may be one's duty to sacrifice it. Then, secondly, if there were ever to be a similar case, who would judge which person is to die? (This point might be overcome by drawing lots.) So he concluded that, terrible as the temptation might be in this kind of case, the law should ‘keep the judgment straight and the conduct pure’. The sentence of death was later commuted to six months’ imprisonment, thus emphasizing the obvious conflict between the desire to reaffirm the sanctity of life and the widely felt compassion for people placed in an extreme situation. In DPP v Lynch (1975)97 the House of Lords accepted, by a majority of three to two, that duress by threats should be available as a defence to an accomplice to murder, reflecting the law's compassion towards a person placed under such extreme pressure. But then the Privy Council in Abbott v R (1977)98 held that duress was unavailable as a defence to the principal in murder, and in Howe (1987)99 the House of Lords had to decide whether to perpetuate this distinction between principals and accomplices. Their Lordships decided not to do so, unanimously favouring a rule which renders duress and necessity unavailable as defences in all prosecutions for murder.100 The primary reason for their decision was that the law should not recognize that any individual (p. 213) has the liberty to choose that one innocent citizen should die rather than another. All duress cases involve a choice between innocents, D and the intended victim, and the law should not remove its protection from the victim. Thus D is required to make a heroic sacrifice. A secondary argument, similar to that employed a century earlier in Dudley and Stephens, was that executive discretion could take care of deserving cases—either by releasing D on parole at an early stage or even by refraining from prosecution.101 Both these arguments are open to criticism. The argument based on protection for the innocent victim seems to assume that duress is being advanced as a permission for killing: this enables the judges to assume that, because the killing of an innocent person is impermissible, duress should not be a defence. It was argued earlier that a killing under duress might be justifiable if there were a net saving of lives,102 but that is not the issue here. Where it is a question of liability for taking one innocent life to save another, the rationale must be one of excuse, not permission. It can therefore be put alongside other situations in which a killing may be excused in whole or in part (e.g. mistaken self-defence, intoxication, loss of self-control), without being permitted.103 Utilitarians might argue that a rule denying duress as a defence to murder is preferable because over the years it might achieve a net saving of lives: 104 this not only fails to take the defendant's interests into account, but also assumes that persons under duress will know of the law's approach and will be influenced by it, an assumption which will rarely be true. The second argument, in favour of convicting the person under duress and then invoking executive clemency to reduce the punishment, also smacks of an unrealistic

Page 16 of 44

Excusatory Defences utilitarian solution. For one thing, there can be no certainty that the Parole Board will view these cases more favourably than others. For another, if we are satisfied that D was placed under extreme pressure, we ought to declare that publicly either by allowing a defence or, if not, by allowing a partial defence to murder on an analogy with loss of self-control. The argument in favour of merely a partial defence should not be understated: as Chapter 7.4(c) will show, it is possible both to recognize the sanctity of life as a fundamental value and to demonstrate compassion. The case for a partial defence in duress cases has arguably been strengthened by the extension of the partial defence to murder of loss of self-control to cover cases in which D lost self-control and killed V when in fear of serious violence from V.105 The requirement in duress cases that the threatener have insisted on immediate or almost immediate compliance106 is strongly analogous to a requirement for loss of self-control. Further, whilst it is true that in loss of self-control cases it must be the person threatening the violence who is killed, rather than an innocent third party (as in duress cases), this is arguably no more than a factor to take into account when deciding (p. 214) whether the person of reasonable steadfastness might have done as D did. In its 2005 Consultation Paper, the Law Commission proposed that duress should be available as a partial defence to first degree murder, reducing it to second degree murder, and that the Bowen test of relevant characteristics should be tightened so as to run in parallel to the partial defence of loss of self-control.107 There were other, complicated proposals about how this approach should be adapted to defences of duress to second degree murder and to manslaughter, but the consultation process persuaded the Commission to abandon this whole approach. Although recognizing that consultees were ‘more divided on duress’ than on any other aspect, the Commission has now reverted to its earlier view that in principle it would be morally wrong to convict of any crime a defendant who satisfies the stringent requirements of the defence of duress, having reacted as a person of reasonable fortitude might have done.108 The Commission recognizes that recommending duress as a partial defence might have been a compromise acceptable to many, but it states that the argument against a complete defence based on the sanctity of life is not conclusive because of cases of ‘ten year olds and peripheral secondary parties becoming involved in killing under duress’.109 The reference to age is sharpened by the subsequent decision in Wilson (2007),110 where a boy of 13 was pressed by his father into helping with the killing of a neighbour and no defence of duress was available on the charge of murder, despite considerable evidence that he was so frightened that he could not disobey his father. The Law Commission's principal argument is that as a matter of moral principle a person who is found by a jury to have reacted to extreme circ*mstances as a reasonable person might have done ‘should be completely exonerated despite having intentionally killed’, adding that youth is a relevant factor in determining reasonableness.111 Thus the Commission insists that the threat must be believed to be lifethreatening, and that D's belief that the threat has been made is based on reasonable grounds. The argument for adopting the Graham approach is that, compared with provocation and selfdefence (which have no such requirement of reasonable belief), there is a less immediate temporal or physical nexus between the threat and the killing in duress cases.112 This also becomes the primary argument in favour of the Commission's recommendation that the burden of proof be reversed where duress is raised as a defence to homicide113 —that the separation of the threat from the killing creates extra difficulties for the prosecution. However, the Commission supports the tightening of the law by the House of Lords in Hasan, one aspect of

Page 17 of 44

Excusatory Defences which was the replacement of the former ‘imminence’ requirement with (p. 215) one of immediacy; so the temporal separation cannot be great, and reference to ‘time to reflect’ takes insufficient account of the great emotional turmoil brought about by threats of this kind.114

6.4 Reasonable mistake and putative defences For the first three-quarters of the twentieth century, the approach of the common law to mistake was that if the defendant wished to rely on this defence it must be shown that he had reasonable grounds for his mistaken belief. The leading case was Tolson (1889),115 where the Court for Crown Cases Reserved held that a mistake of fact on reasonable grounds would be a defence to any criminal charge. Despite being cited as the leading case, the ambit and status of Tolson were never clear, since Stephen J devoted much of his judgment to the proposition that if the mental element of the crime is proved to have been absent, the crime so defined is not committed.116 Certainly it is authority for the proposition that reasonable mistake is a defence to crimes of strict liability.117 It is also authority on the crime of bigamy, and was expressly preserved by the House of Lords in DPP v Morgan (1976)118 when it introduced (or, in the light of Stephen J's judgment, reintroduced) the proposition that if the mental element is missing in respect of one of the conduct elements specified in the definition of the crime, then as a matter of inexorable logic D should be acquitted even if the mistake was wholly unreasonable. The ‘inexorable logic’ argument may be accepted as a starting point, but the question is whether considerations of moral fault indicate that in certain types of case it should be abandoned. We have already seen that the ‘inexorable logic’ has not been followed in respect of intoxication (where special restrictive rules have been created). When the House of Lords in Morgan opted for the ‘inexorable logic’ approach, treating the claim of mistake as a mere denial of the required mental element, it expressly left undisturbed two different rules—the Tolson principle, as applied to bigamy, and the requirement that mistakes relating to a defence should be reasonable. This second requirement relates to ‘defences’ resting on permission: if there is a mistake about the circ*mstances giving rise to the permission, this makes it a putative defence (i.e. an excuse rather than a (p. 216) permission, because the circ*mstances for permission were absent and D merely believed they were present). The persistence of the objective approach to mistake in these cases owed more to assumption and repetition than to principled argument. Its chief application was in self-defence, where courts had tended to require that any mistake about the circ*mstances should be based on reasonable grounds.119 But this reasonable mistake doctrine, left intact in Morgan itself, was swept away by decisions of the Court of Appeal and Privy Council in the 1980s.120 Thus a putative defence will succeed wherever the prosecution fails to prove that D knew the relevant facts (i.e. that D did not hold the mistaken belief claimed), no matter how outlandish that belief may have been. Thus in Williams (1984)121 V saw a man, X, snatch a bag from a woman in the street; V ran after X and forcibly detained him; D then came upon the scene and asked V why he was punching X; V said, untruthfully, that he was a police officer; D asked V for his warrant card, and when V failed to produce the card, D struck V. D was charged with assaulting V, and his defence was that he mistakenly believed that his actions were permissible in the prevention of crime. It is plain that his actions were not in fact permissible, since V was acting lawfully in trying to detain X.122 The law requires the prosecution to satisfy the court that D was aware of the facts which made his action unlawful, and he was not. He

Page 18 of 44

Excusatory Defences was mistaken. The Court of Appeal held that his conviction should be quashed: ‘The mental element necessary to constitute guilt is the intent to apply unlawful force to the victim. We do not believe that the mental element can be substantiated by simply showing an intent to apply force and no more.’ The courts in Williams and Beckford123 presented this as an application of the ‘inexorable logic’ approach in Morgan (overlooking the fact that Morgan left this aspect of the law unchanged), reasoning as follows:

• unlawfulness is an element in all crimes of violence; • intention, knowledge, or recklessness must be proved as to that element; and therefore; • a person who mistakenly believes in the existence of circ*mstances which would make the conduct lawful should not be criminally liable.

The crucial step is the first: how do we know that unlawfulness is a definitional element in all crimes?124 Not all crimes are defined explicitly in this way. So it is, rather, a doctrinal question. Andrew Simester has argued that unlawfulness cannot be an (p. 217) ingredient of the actus reus, since only when there is actus reus with mens rea can we conclude that conduct was unlawful.125 Might this not be a question of terminology? Some would argue, as we saw in Chapter 4.6, that there is no actus reus where the conduct is permissible. If ‘absence of permission’ is substituted for ‘unlawfulness’ in the above reasoning, does not the difficulty claimed by Simester disappear? A stronger argument is that, irrespective of the definitional boundaries of the actus reus, there is a need to confront the moral issue whether there should be some grounds for doing so before using force against another. Using force is prima facie wrongful and so citizens should not use force without grounds for doing so, bearing in mind that the circ*mstances in which D has to act will affect what is regarded as adequate grounds. This distinguishes cases of putative defence from other cases of mistake in which D does not think what he is doing is wrongful or dangerous.126 Rather than relying on the logic of steps (i), (ii), and (iii), the law should adopt this more context-sensitive approach, taking some account of the circ*mstances of the act, of D's responsibilities, and of what may reasonably be expected in such situations.127 The consequence may be not to require knowledge of a certain circ*mstance in the definition of the offence, but to require reasonable grounds for a belief. In rape cases these considerations militate in favour of a requirement of reasonable grounds for any mistake, as the Sexual Offences Act 2003 now provides.128 Reasonable grounds should also be required in respect of age requirements for consensual sexual conduct, although in this respect the 2003 Act goes further and imposes strict liability in some circ*mstances.129 In principle, it is also right to require reasonable grounds before allowing the acquittal of a police officer with firearms training, as in Beckford v R (1987).130 Of course, any such infusion of objective principles must recognize the exigencies of the moment, and must not demand more of D than society ought to expect in that particular situation.131 That is a necessary safeguard of individual autonomy. The general point, however, is that there may be good reasons for society to require a certain standard of conduct if the conditions were not such as to preclude it, Page 19 of 44

Excusatory Defences particularly where the potential harm involved is serious. These arguments may be no less strong in many cases of putative defences of duress, where a reasonableness requirement has been imposed.132 Any move in the direction of requiring reasonableness may have (p. 218) the effect of raising the question whether cases of mistaken belief in permission are necessarily cases of excuse, or whether they may be treated as forms of permission.133 In fact they have elements of both: Antony Duff proposes that this is best expressed by describing D's conduct as wrong but warranted—wrong because there is no objective reason for it, but warranted because D (reasonably) believed in the existence of circ*mstances that would have made it the right thing to do.134 English law currently takes variable approaches to these questions. In recent years the judges have often seemed to be firmly in the embrace of the ‘inexorable logic’ approach to mistake,135 but there have been some deviations which perhaps suggest recognition of the complexity of the issues. As noted in the discussion of duress in section 6.3(a), the poorly reasoned decision in Graham,136 holding that a mistake about the nature of the threat must be a reasonable one if the defence of duress is to be available, has now been championed on strong protectionist grounds by Lord Bingham in Hasan and adopted by the Law Commission.137 However, although the objectivist approach in the Sexual Offences Act suggested a more context-sensitive treatment of mistake and putative defences, such considerations were neglected in the drafting of the self-defence provisions in s. 76(3) of the Criminal Justice and Immigration Act 2008, which confirms a wholly subjective test of belief with no variations between trained police or military personnel and ordinary citizens caught up in a sudden incident.138

6.5 Ignorance or mistake of law (a) The english rules English criminal law appears to pursue a relatively strict policy against those who act in ignorance of the true legal position, but the maxim ignorantia juris neminem excusat (ignorance of the law excuses no one) is too strong as a description. Ignorance or mistake as to civil law, rather than criminal law, is capable of forming the basis of a defence; indeed, the crimes of theft and criminal damage explicitly provide for defences where D believes that he has a legal right to take or to damage property.139 But it would be unsafe to state the rule by reference to a distinction between matters of civil law and criminal law, because offences are often defined in such a way as to blur the two. Whether goods are classified as ‘stolen’ for the purposes of the offence of handling stolen property seems to be a question of criminal law, so if D knows all (p. 219) the facts but misunderstands their legal effect, this is irrelevant. Whether an auditor is disqualified from acting for a certain company seems to be a question of civil law, so where D was unaware of the relevant law, his conviction for acting as an auditor knowing that he was disqualified was quashed.140 One difference between these two offences is that the latter contains the word ‘knowingly’, whereas the crime of handling includes the words ‘knowing or believing’; it is certainly true that a number of English decisions have allowed mistake or ignorance of the law to negative ‘knowingly’,141 but this cannot explain all the decisions.142 English law does recognize that the obligations are not all on one side. The state has duties to declare and to publicize laws and regulations: non-publication of a Statutory Instrument will usually afford a defence to any crime under that Instrument to a 143 Page 20 of 44

Excusatory Defences person unaware of its existence,143 and failure to publish a government order in respect of a particular person will also afford a defence to that person if he or she is unaware of the order.144 (b) Individual fairness and public policy It could be argued that individual fairness demands the recognition of ignorance or mistake of law as an excuse: a person who acts in the belief that conduct is non-criminal, or without knowing that it is criminal, should not be convicted of an offence. Although ignorance of the law may not negative the fault requirements of a particular offence, respect for individual autonomy supports the excuse in its own right: a person who chooses to engage in conduct without knowing that it is criminal makes a choice which is so ill-informed as to lack a proper basis. The counter-arguments are based on conceptions of intrinsic wrong and of social welfare. One is that it can fairly be assumed that people know that certain morally wrong conduct is criminalized, even if they are unaware of the precise terms of the law.145 The utilitarian argument that it is desirable to encourage knowledge of the law rather than ignorance, and any rule which allowed ignorance as a defence would therefore tend to undermine law enforcement.146 This does not establish that ignorance of the law is always wrong, merely that it may be socially harmful. Another is the argument that, if we judge defendants on their particular view of the law rather than on the law as it is, we are contradicting the essential objectivity of the legal system.147 This is, to say the least, an exaggeration: so long as the court states what the law is, the law's objectivity remains unimpaired. It would also seem to suggest (p. 220) that for a court to allow any excuse amounts to a denial of the offence. This not only confuses the element of excuse with the element of wrongdoing,148 but also overlooks the value of a publicized trial, where reasonable mistake of law is allowed, as a means of public education. Is it generally wrong to be ignorant or mistaken about the law? It may be argued that it is a duty of citizenship to know the law. Thus, to convict a person despite ignorance of the law is not to attack the principles of choice and individual autonomy which were identified earlier as fundamental to the principles of fairness.149 Rather it is to forsake an atomistic view of individuals in favour of a recognition of persons as social beings, with both rights and responsibilities within the society in which they live.150 It has already been argued that in many situations it is fair to expect citizens to take care to enquire into the surrounding circ*mstances before they act, and the case for requiring some mistakes to be reasonable has been put.151 A similar line of argument might support a duty on each citizen to take reasonable steps to become acquainted with the criminal law.152 There are few problems in making the duty known, since ‘ignorance of the law is no excuse’ is a widely known principle even now.153 The duty should not be an absolute one, however. First, there is often uncertainty in the ambit of the law. Sometimes the legislature acknowledges the difficulty of stating the law by allowing D's own standards as a benchmark of lawfulness, as in the crime of blackmail.154 Sometimes it resorts to a broad standard such as ‘reasonable’ or ‘dishonest’, leaving the courts to concretize the norm after each event, which goes against the principles of maximum certainty and fair warning.155 This is not to suggest that every case in which the courts change the law should inevitably give the defendant a defence of ignorance of the law; indeed, the European Court of Human Rights has held not only that judicial extensions of the law conform to Art. 7 if they are ‘reasonably foreseeable’, but also that the application of that test varies according to the subject-matter of the law and that ‘a law might still satisfy the

Page 21 of 44

Excusatory Defences requirement of foreseeability even if the person concerned had to take legal advice’ to determine its practical scope.156 A second reason for not making the policy absolute is the possibility that the State has not fulfilled its duties in respect of making a new offence known and knowable.157 The State clearly (p. 221) has this duty when it seeks to impose criminal liability for an omission,158 and the duty applies generally to the publication of laws. This, indeed, is an aspect of the principle of legality, requiring both certainty of definition and fair warning.159 One way of maintaining the general duty to know the law, whilst allowing exceptions based on respect for individual autonomy, would be to provide that a mistake of law may excuse if it is reasonable. This, in combination with the argument in section 6.4, would have the advantage of narrowing the present gulf, wide and difficult to defend, between the effects of ignorance of law (no general defence) and ignorance of fact (frequently negativing liability).160 Ignorance of the law would clearly be reasonable if fair warning of a prohibition had not been given: this would accommodate the second point above. Mistake or ignorance of law might also be reasonable if D had no cause to suspect that certain conduct was criminal, or if D had been misinformed or wrongly advised about the law (see (c)), or perhaps in other circ*mstances.161 Ignorance and mistake would be unlikely to be held reasonable if D was engaging in a business or an activity (such as driving a car) that is known to have changing rules; 162 but the merit of a reasonableness requirement is that it would not absolutely rule the defence out. A defendant would be able to argue that there were special circ*mstances warranting exculpation. To rebut the claim that such an excuse might be raised so often as to impede the administration of the criminal law, one has only to refer to the lengthy experience of Scandinavian countries in providing for defences of this kind.163 The Draft Criminal Code states that ‘ignorance or mistake as to a matter of law does not affect liability to conviction for an offence except (a) where so provided, or (b) where it negatives the fault element of the offence’.164 This is traditional, inflexible, and unsatisfactory: it would prevent the courts from developing a wider defence, and would relegate most of these matters to mitigation of sentence.165 Moreover, exception (b) hardly corresponds to any general moral distinction. The legislature has not pursued a consistent policy in deciding whether or not ‘knowingly’ should form part of the definitions of offences, and it certainly cannot be assumed that Parliament had considered whether particular offences justify an exception in favour of ignorance (p. 222) or mistakes of criminal law (including unreasonable ones). The courts have veered between allowing ignorance of law to negative ‘knowingly’ and declaring that this approach would be ‘wholly unacceptable’.166 There is a need to adopt a clear principle (a duty with circ*mscribed exceptions) and then to interpret statutory offences in the light of it. The same approach should be adopted where the offence includes a phrase such as ‘without lawful excuse’ or ‘without reasonable excuse’.167 (c) The reliance cases Another benefit of moving away from the relatively strict English policy against defences based on mistake or ignorance of criminal law towards a ‘reasonable grounds’ defence would be to deal more fairly with the ‘reliance’ cases. In Cooper v Simmons (1862)168 an apprentice absented himself from his apprenticeship after the death of his master, having sought the advice of an attorney and having been counselled that he was no longer bound. The Court nevertheless convicted him of unlawfully absenting himself from his apprenticeship, and

Page 22 of 44

Excusatory Defences Pollock CB stated that ‘it would be dangerous if we were to substitute the opinion of the person charged … for the law itself’. In Arrowsmith (1975)169 D had on occasions distributed leaflets urging British soldiers not to serve in Northern Ireland. In the past the Director of Public Prosecutions had declined to prosecute her under the Incitement to Disaffection Act 1934, but now she was charged with an offence under that Act. One line of defence was that she reasonably believed, as a result of a letter from the Director, that her conduct did not contravene the Act. The Court of Appeal upheld her conviction, stating that ‘a mistake as to the law would not avail the appellant except perhaps in mitigation of sentence’. Both these cases would surely be better analysed in terms of reasonable reliance. If it is established that D relied on advice from officials with regard to the lawfulness of the proposed conduct, that ought to be sufficient to support reasonable grounds for the mistake of law. Confusion may arise about the entitlement of a particular agency or official to advise a member of the public about the law, as one English case vividly demonstrates,170 but since reasonable mistake of law would be an excuse, the key question is whether D reasonably assumed that the person giving the advice was duly authorized. In the element of reliance, these cases can call upon a kind of estoppel reasoning—the State and the courts should not convict a person whom they or their officers have advised otherwise.171 Thus the Control of Pollution Act 1974, s. 3(4)(a), specifically creates a (p. 223) defence to the crime of unlicensed waste-disposal where D ‘took care to inform himself from persons who were in a position to provide information’, recognizing both individual fairness and an estoppel on officials. Thus, even if one were persuaded by the argument that allowing mistake of law as a general defence would encourage ignorance of the law, the reverse of that argument applies here: to recognize officially induced error as a defence would signal the value of citizens checking on the lawfulness of their proposed activities. Indeed, all the values that support the principle of fair warning militate in favour of recognizing officially induced error, since a citizen who seeks advice is showing respect for the law.172 One reason for the rarity of appellate cases on mistake of law may be that it is often accommodated in other ways. An appeal is unlikely if a person receives substantial mitigation of sentence, perhaps an absolute or conditional discharge. On some occasions a person who acted on a mistaken view of the law might not be prosecuted at all, or the prosecution might be discontinued.173 In one case a company was advised by members of the local council's planning department that the erection of advertising boards would not require planning consent. The company erected the boards, and the council then brought a prosecution. The Divisional Court held that the prosecution should have been stayed as an abuse of process,174 Schiemann LJ stating that it is ‘important that the citizen should be able to rely on the statements of public officials’. The council had argued that these were junior officials and that the company was wrong to rely on their opinion, but the Divisional Court replied that ‘it was not as though they had requested planning advice from one of the council's gardeners’. This is a significant decision, employing the powerful procedural approach of staying the prosecution where a mistaken view of the law has been implanted by an official. The courts might well decline to recognize a substantive defence of officially induced error of law,175 but it can be argued that staying the prosecution is a more appropriate remedy inasmuch as D might not have brought himself within the offence definition at all if the official advice had not been given. Should the doctrine extend to acting on the advice of a lawyer? Glanville Williams, although a

Page 23 of 44

Excusatory Defences strong supporter of a defence of reasonable reliance on official statements, pointed out the danger that allowing reliance on a lawyer's advice (rather than official advice) might open up a broad route to exculpation for corporate defendants in particular.176 On the other hand, for an individual to take legal advice might be even (p. 224) more reasonable, in terms of citizenship duties, than to rely on the advice of a junior official.177

6.6 Entrapment There are cases in which the police arrange either for one of their own officers or for some other person to approach D and tempt him to commit an offence. If D commits the offence after the officer or agent provocateur has over-stepped the boundary of permissible conduct, should there be a defence of entrapment? Some jurisdictions admit such a defence. Until 2001 English law relied merely on the exclusion of evidence or mitigation of sentence in such cases. Now, following the decision of the House of Lords in Looseley; Attorney-General's Reference No. 3 of 1999,178 proof of entrapment leads to a stay of the prosecution. What amounts to entrapment? The House of Lords held that where ‘the police conduct preceding the commission of the offence was no more than might be expected from others in the circ*mstances’ it is acceptable.179 This is the ‘unexceptional opportunity test’: if all that the official does is to offer D an unexceptional opportunity, this is permissible conduct. If the official goes further than that—as by inciting, instigating, persuading, or pressurizing—it would be a case of entrapment. Where there are reasonable grounds for suspecting a particular individual, or individuals frequenting a certain place, of involvement in a type of offence (e.g. drug dealing), it seems that it is permissible to test the person(s) by approaching them and making an enquiry.180 The rationale of this approach to entrapment seems to have two strands. First, the courts are concerned to prevent abuse of executive power: it would be a misuse of power for the State's agents to lure citizens into breaking the law and then to prosecute them for doing so. Secondly, entrapment must be prevented in order to protect the integrity of the criminal justice system—which would be undermined if the courts allowed the prosecution of crimes created by state officials.181 These rationales led the House of Lords in Looseley to adopt the procedural remedy of staying the prosecution for abuse of process, rather than allowing the trial to proceed and according D a defence to criminal liability. This approach is consistent with that of the Strasbourg Court in Teixeira de Castro v Portugal,182 which held that the entrapped applicant had been ‘deprived of a fair trial from the outset’, and therefore that Art. 6 had been violated. The Supreme Court of the United States still upholds an entrapment defence,183 and the Model Penal Code includes one.184 The Supreme Court's version focuses on whether (p. 225) D would have committed the offence otherwise, which then becomes a question of whether he was ‘pre-disposed’ to commit such offences. This notion is also to be found in the Strasbourg decision in Teixeira de Castro, but it shifts the enquiry back towards the character and previous record of the person incited—and into dangerous waters. The House of Lords was wise to reject the notion of predisposition in its Looseley judgment, but it remains to be seen whether its requirement of ‘reasonable suspicion’ that the person targeted was involved in that type of offending will be any more robust. The rationale and remedies for entrapment just described are dependent on the involvement of the State and its officials in instigating crime. No such rationale would apply if it were a private Page 24 of 44

Excusatory Defences individual who, on his or her own initiative, incited D to commit the offence: the fact that one person incites another does not relieve the other of criminal liability, since the law regards each of them as autonomous individuals who are able to choose what to do. However, there is an argument that few people would wish to live in a society where they were liable to have their virtue tested unexpectedly (by, for example, journalists in search of a story), and that therefore the exclusion of evidence ought to be available in egregious cases of private entrapment.185 The courts seem to accept this to some extent, in that they seem to have made little of the distinction between official entrapment and private entrapment (typically engineered by journalists), although there has been no case in which a stay of prosecution on grounds of private entrapment has been ordered and upheld.186

6.7 Reviewing the non-permission-based defences In Chapter 4 we dealt with permissions for the use of force, often regarded as defences, and also with involuntary conduct. Permissions are clearly separate from the exculpatory doctrines in this chapter, but the reason for placing involuntariness (automatism) in Chapter 4 is that it relates to the basic requirement of a voluntary act. From the functional point of view, however, automatism tends to operate as a defence, and its rationale belongs properly with the capacity requirements (expressed in terms of infancy and insanity) discussed in Chapter 5.2. Reference will be made below to these incapacity ‘defences’, as we consider some general questions about the rationales, functions, and appropriate responses to the various conditions discussed in this chapter. First, we shall examine the implications of the threshold question: should a suggested excuse be recognized as a defence, or merely as a mitigating factor in sentencing, or even marked in a different way? Secondly, we consider the roots of fault and the excuses (p. 226) in conceptions of individual responsibility. Thirdly, we go on to examine the arguments in favour of policies of social defence and social responsibility. Whether it is possible to travel beyond a demonstration of the conflicting policies and principles and to achieve a unifying theory is then the question which remains. (a) The recognition of exculpatory doctrines In moral and social terms there is probably a scale of exculpation, running from the most acute forms that affect agency itself by denying responsibility (such as insanity and automatism) to mere matters of difficulty and extra pressure at the other extreme. Most forms of exculpation can be manifested to a different degree (strong or weak circ*mstances of duress, mild or acute mental disorder). During the course of the chapter it has often been remarked that the courts strive to keep the ambit of a particular ‘defence’ as narrow as possible, so as to capture only the full or extreme cases. This approach leaves other cases which have exculpatory elements to be dealt with in some other way. In some spheres of criminal law it is not simply a question of whether there is a defence or not. Loss of self-control and diminished responsibility187 are available as partial defences to murder, reducing the crime to manslaughter, and there is no procedural reason why they and other partial defences should not be granted a wider application—wherever there is a ladder of offences, the partial defence might serve to reduce the higher to the lower.188 There are obvious counter-arguments, grounded in the increased complexity and length of trials of cases where the unique stigma of ‘murder’ is not present,189 but these concede rather than weaken the moral/social arguments for allowing the reduced culpability in, say, loss of self-control cases to be signified by a

Page 25 of 44

Excusatory Defences reduction in the offence of conviction. This may be regarded as an example of fair labelling: 190 just as there is a ‘scale of excuse, running downwards from excusing conditions, through partial excuses to mitigating excuses’,191 so the law should reflect these gradations through complete defences, partial defences, and then mitigation of sentence. In some spheres, English courts have faltered and have refused to recognize a defence at all, leaving all degrees of exculpation to be reflected by procedural means, chiefly at the sentencing stage. This has been the predominant approach to entrapment,192 and for many years it was the courts’ approach to excuses based on necessity.193 Indeed, the House of Lords has gone further by proposing executive discretion as a desirable way of mitigating the effective punishment of those who kill under duress.194 (p. 227) It would be procedurally possible to deal with all excuses, and, indeed, with all fault requirements, by excluding them from the criminal trial and dealing with them at the sentencing stage. As we saw in Chapter 5.4(a), one could create a strict liability system in which proof of conduct and causation was sufficient for conviction, and fault would then be considered at the sentencing stage as a pointer to the most appropriate means of state intervention to prevent any repetition. The objection to this is that a criminal conviction is rightly regarded as condemnatory: it is unfair to apply this official censure when the absence of fault is so high on the ‘scale of excuse’ that there should be no formal blame. Supporters of strict liability, such as Baroness Wootton, would reply that on their system a conviction would not carry such a stigma, since it would not imply culpability.195 Such an approach would sacrifice the underlying deterrent and censuring elements of the criminal law, as well as reducing the protection of individual autonomy by reducing the individual citizen's ability to plan and to predict the law's interventions. So long as the criminal law is the principal censuring institution, conviction should carry the moral connotation of culpable wrongdoing, and so there ought to be the possibility of recognizing compelling excuses by means of acquittal. Even if defences to criminal liability are recognized for ‘strong’ exculpatory factors, it will remain necessary to deal appropriately with ‘weak’ or imperfect cases of exculpation. This is where mitigation of sentence should be the principal tool. Unless the penalty is mandatory (as, in English law, for murder), courts will be able to reflect the strength of the excuse in the sentence they pass. However, there are two difficulties in treating this as an ideal way of reflecting the defendant's desert. First, there is the question of establishing the factual basis for mitigation. Sometimes this will have emerged during a trial, if a trial has taken place,196 but more often it will be necessary to lay a foundation after conviction and before sentence. Procedures to ensure proper fact-finding are still developing, but there has been insufficient recognition of the importance of ensuring that defendants have the same evidential safeguards as they would have had in a criminal trial.197 Secondly, there is no clear recognition that mitigation of sentence is a right. It is often presented as discretionary, suggesting that courts may withhold a reduction in sentence if they wish to do so.198 This is unsatisfactory, and reflects the general lack of structure of English sentencing law in respect of mitigating factors. If defences to liability and mitigation of sentence should be the two principal responses to excuses, what should be the role of procedural remedies? The most powerful procedural approach is not to prosecute at all. Thus prosecutors are expected to take account of the likely line of defence in a particular case, and they might therefore bring no prosecution if

Page 26 of 44

Excusatory Defences convinced that a certain defence will probably succeed. Where they are not so convinced, they may still decide that a prosecution would not (p. 228) be ‘in the public interest’. The Code for Crown Prosecutors mentions cases where ‘the offence was committed as a result of a genuine mistake or misunderstanding’, and cases where ‘the defendant is elderly or is, or was at the time of the offence, suffering from significant mental or physical ill health’.199 Both of these factors are to be weighed against the seriousness of the offence. In practice, nonprosecution and discontinuance of prosecution are responses to many cases involving mentally disordered persons, who may then be admitted to hospital or a treatment programme informally. However, a deeper issue is whether a prosecution should be stayed once it has been commenced. This powerful remedy has been held appropriate in cases of entrapment200 and in one case of officially induced mistake of law.201 The reason it is particularly appropriate in these types of case is that the involvement of officials in ‘creating’ the offence makes it wrong for the prosecution to be heard by the courts at all. Providing a defence to liability would not be enough: it is so fundamentally wrong for the state to prosecute that D should not be put to the trouble of defending himself. (b) Individual responsibility It was shown in Chapter 5.4(a) that the roots of the conception of individual responsibility which underlies the principle of mens rea lie in respect for the autonomy of the individual. Thus defendants who did not have the capacity to choose—who were not responsible moral agents —should be dealt with by means of the denial-of-responsibility ‘defences’ of infancy, insanity, and automatism. Those who were responsible moral agents should then be judged, if they raise defences such as duress, necessity, or reasonable mistake, according to the standard of what we ought reasonably to expect of a person in that situation. These defendants are, as John Gardner puts it, ‘asserting their responsibility’ (in the sense that they are claiming to have the capacities of a normal citizen of full age and sound mind) but claiming an excuse on the ground that their response to a testing situation ‘lived up to expectations in a normative sense’.202 Thus, as we saw in section 6.3, the standard of the person of reasonable firmness is central to the defences of duress and necessity. Now in one sense this might be thought to be indulgent to D—there is no requirement that he should have felt totally deprived of his freedom of action, merely that a reasonably steadfast citizen would have found the pressure intolerable203 —but that may be explained on the ground that this is not a denial of responsibility but an excuse, where a high but reasonably achievable standard is more appropriate than perfectionism.204 (p. 229) In another sense, however, the standard of ‘normative expectations’ may be thought insufficiently indulgent to D: it precludes actual enquiry into the pressures experienced by this defendant. Even if D felt totally overwhelmed by the pressures, the law would not allow a defence of duress unless a reasonably steadfast person would also have been seriously affected. In other words, there is no scope for a plea of diminished personal capacity, based on D's inability to meet the normative expectations. In the past the courts have been reluctant to lower the standard, because of a fear that false defences may succeed if the law were totally subjective, or a fear of a significant loss in the deterrent effect of the law, or perhaps for other protectionist reasons. But in more recent years, as we have seen, courts have occasionally been willing to lower the standard in order to take account of certain individual susceptibilities and conditions.205 Lowering the standard blurs the rationale, however. Strictly speaking, if D lacks the capacity to attain the standard normatively expected,

Page 27 of 44

Excusatory Defences then the essence of plea is not that D behaved as a responsible citizen might be expected to, but rather that D is to some extent denying responsibility for what was done.206 This would reduce the grade of liability for defendants who to a significant extent felt coerced, compelled, or ‘pressured’ to do what they did and where there is evidence of an underlying condition to explain this. It might, for example, be open to those who narrowly fail to satisfy the requirements of a defence of insanity, and could include such conditions as pre-menstrual syndrome.207 It would also deal with those unable to attain the standard of reasonable steadfastness in duress and necessity.208 In order to preserve the distinct grounds for different complete excuses (such as duress), and to respect fair labelling, it would be preferable to articulate as many discrete defences as possible, and to have any defence or partial defence of diminished capacity in a kind of residual or ‘sweeper’ role. Whether its availability would unduly complicate trials is for careful enquiry and debate. There is no need for citizens to have fair warning of its existence,209 but it is important to ensure that the courts exercise their power fairly and consistently as between similarly or equivalently situated defendants. (c) Social responsibilities and social defence In practice, the objective standard of the person of reasonable firmness in excuses such as duress and necessity may be sustained less by the doctrine of ‘normative expectations’ (p. 230) or Hart's ‘fair opportunity’ rationale than by judicial fear of false defences.210 The latter may also be a prominent reason for the presence of restrictive conditions in intoxication (the limitation to crimes of ‘specific intent’) and in ignorance or mistake of law (the virtual denial of such a defence). There are, however, stronger protectionist arguments for restrictions. One is the importance of taking compulsory measures against persons shown to be capable of causing harm in their condition. This is a major plank of the ‘special defence’ in insanity cases, where absence of capacity leads to a special verdict which, in turn, may give rise to compulsory measures of social protection. Yet we saw in Chapter 5.2 that the terms of the defence are not designed to demonstrate that D is a dangerous person, likely to cause further serious harm if given a simple acquittal. The same might be said of the restrictions placed on intoxication as a defence (section 6.2), where beliefs about future dangerousness may play some part, but probably the chief reason for restricting the defence is the belief that people who do harm whilst intoxicated are blameworthy.211 One argument often mobilized against the infiltration of objective requirements into excusing defences is ‘logic’. We have noted this in relation to intoxication (section 6.2(b)) and particularly mistake of fact (section 6.4). Consistency of approach to excusing conditions would certainly seem to be an element in fairness, but it does not follow that the excuses should be consistently and utterly subjective in their requirements. The subjective principles have their foundation in the principle of individual autonomy, and its emphasis on choice, control, and fair warning. But we have seen that modern liberal philosophy has begun to emphasize that individuals should be viewed as members of society with mutual obligations rather than as abstracted and isolated individuals. The subjective principles and the contemporaneity principle,212 ingrained as they are in much academic writing in the common law world, in some judicial pronouncements, and in many Law Commission proposals, seem premised on an atomistic view of individual behaviour.213

Page 28 of 44

Excusatory Defences An alternative approach would spell out certain duties of citizenship which should form part of membership of a legal community and which might have some bearing on issues of criminal responsibility. One such duty might be to show reasonable steadfastness in the face of pressure, and to avoid uncontrolled behaviour that might lead to harm to others. This might be applied to cases of intoxication, based on the general social proposition that persons who take large amounts of alcohol or certain drugs constitute a greater and well-known risk of causing harm. A similar argument might be used to justify the refusal to admit loss of self-control as a general defence, rather than as a partial defence to murder. In principle, no exceptions should be admitted (p. 231) to the principle that citizens should control their tempers. However, certain loss of self-control cases contain strongly exculpating elements, in terms of justified anger or fear combined with a disturbed emotional state, and these make a convincing case for loss of self-control as a qualified defence.214 How might the ‘duties of citizenship’ approach apply to mistake cases? Citizens may surely be expected to make reasonable efforts to acquaint themselves with the contours of the criminal law, but this does not support the refusal of the English courts and legislature to recognize a general excuse based on ignorance or mistake of law. On the contrary, the citizen's duty is fulfilled by making reasonable enquiries, and this would support a defence of reasonable ignorance or mistake of law. Indeed, where there is reasonable reliance on official advice the prosecution should be stayed, since D has acted as a good citizen should.215 Strangely, the English courts erred in the opposite direction in cases of mistake of fact, seduced by the allure of what Lord Hailsham described as ‘inexorable logic’. The courts have failed to show proper sensitivity to the rights of others in particular situations which ought to alert the citizen, but Parliament has now intervened in relation to sex cases. Thus, rather than regarding the defendant in a rape case as abstracted from the situation of close proximity to the victim and subject only to the momentary and ‘inexorable’ logic of the question: ‘did he at that time realize that there was a risk that she was not consenting?’, the law now requires that D reasonably believed that the other party consented.216 Similarly, rather than applying broad subjective principles to a defendant who alleges mistake as to the age of a young person with whom he had (consensual) sexual relations, the law now requires a reasonable belief that the child is 16 or over where the actual age is 13–15 (inclusive),217 although it goes further and (controversially) imposes strict liability as to age where the child is under 13.218 Would it not also be proper to require higher standards of those trained for special roles? Thus, rather than regarding a police officer as abstracted from his or her training and knowledge of alternative means of resolving a situation and subject only to the momentary and ‘inexorable’ logic of the question: ‘did he at that time believe that his life was in danger from V?’, the law should ask whether he took care (so far as possible) to ensure that V was armed, before injuring V or taking V's life.219 The drift of this argument is towards the idea of duties of citizenship which relate in part to control of one's own passions or ‘vices’220 and in part to one's respect for the rights of others in situations which obviously concern those rights (e.g. sexual intercourse, the use of deadly force). The doctrine of prior fault should prevail over (p. 232) the principle of contemporaneity, as the various duties tug the enquiry away from the momentary conduct towards a broader consideration of the situation and its antecedents. Those wedded to traditional theory will doubtless regard this as the spread of negligence liability, and so it is. In this sense, it is compatible with much of what was said by Lord Diplock in Caldwell,221 in that failure to give thought to those matters which the reasonable citizen might regard as obvious

Page 29 of 44

Excusatory Defences may be just as culpable as momentary advertence to such matters. But the idea of duties of citizenship does not require full adherence to the Caldwell doctrine. Two modifications are particularly important. First, the general notion that citizens with ordinary powers of perception and self-control should exercise those powers must be subject to an exception in favour of persons incapable of attaining that general standard.222 But it was argued above that it would be clearer to deal with such persons separately under a (partial) defence of diminished capacity, rather than to distort the general ‘normative expectations’ of citizens. Secondly, a full-blown notion of individual responsibility should be responsive to the relative magnitude of the wrongs or harms. The paradox of Caldwell is that it applied chiefly to criminal damage, an offence which is, in most instances,223 well down the scale of seriousness. A socially sensitive doctrine would impose greater duties of care on citizens in situations where serious harm is widely known to be possible (e.g. use of firearms, fire-raising, irregular driving), where great harm is a possibility (e.g. the operation of transport systems, sports stadiums), and particularly where the means of avoiding the wrong or harm are relatively simple (as in sexual intercourse, enquiring about the other's willingness). It will be evident that these arguments do not promise a simplified system of fault and excuses, but Chapters 5 and 6 should have demonstrated that the present system is far from simple. Conflicts between ‘pure’ individual responsibility and questions of social responsibility are endemic in this sphere. The allure and ‘logic’ of orthodox subjectivism need re-appraisal in the light of considerations of welfare and social responsibility, and a proper adjustment of the different claims debated. Lastly, discussion of citizens’ responsibilities should not lead one to neglect the positive duties of the State in these matters. The obligation to publicize new criminal laws is one such duty, particularly strong in respect of duties to act. It is also time to recognize more fully the wrongness of entrapping citizens into committing offences (section 6.6) and the wrongness of convicting those who rely on official advice (section 6.5). And then there is the more general issue of the State's responsibility for social conditions which foster crime. This is not an outrageous notion: the preamble to the European Convention on Compensation for Victims of Crimes of Violence refers to the idea that the State's duty to provide compensation arises from its failure to prevent crimes,224 (p. 233) and this suggests at least an obligation to take reasonably determined measures to reduce crime. One such measure is to relieve those criminogenic social conditions of poverty, bad housing, unemployment, lack of social facilities, and so forth which have an established link with law-breaking.225 Even if we are not prepared to go so far as to accept social deprivation as an excuse for crime,226 it may be regarded as significantly reducing an offender's ‘desert’, and also as an example of state neglect of a duty towards its citizens.227 (d) Exculpation and ‘desert’ Modern writings on the criminal law have made substantial advances in uncovering and criticizing the reasons for admitting, rejecting, and shaping the various fault requirements in criminal liability. Some ‘defences’ are essentially denials of capacity and responsibility (notably infancy, insanity, and automatism); others are denials of the positive fault requirements of offences (usually, of intention, recklessness, or knowledge). Another important conceptual distinction is that between permission and excuse, which improves the clarity of analysis and might avert confusions in the courts.228 However, once the conceptual distinction is made, it must be recognized that some defences (or partial defences) contain elements of both,229 and that some others rest on neither rationale—for example, principles of integrity and coherence

Page 30 of 44

Excusatory Defences support a decision to stay the prosecution if there is a finding of entrapment or reliance on official advice. It is probably true that defendants would prefer to be acquitted on grounds of permissibility (recognizing that the conduct was acceptable in the circ*mstances) rather than on grounds of excuse (conduct unacceptable, but D insufficiently culpable), and indeed that many defendants would prefer to be acquitted on grounds of excuse than on grounds of denial of responsibility (D lacking capacity at the time, not acting as a responsible moral agent).230 This is one reason women defendants may be unwilling to accept a diminished responsibility defence when they claim loss of self-control.231 The search for a unifying theory of excuses has been less productive, partly because different authors set out to rationalize different groups of defences (some including (p. 234) denials of capacity, others excluding them).232 Hart's influential doctrine, that a person should be held criminally liable only if he or she had the capacity and a fair opportunity to act in conformity with the law,233 captures the essence of individual autonomy in the importance of having fair warning and being able to plan and predict. However, it leaves much work to be done on appropriate criteria of the ‘fairness’ of opportunities. Gardner's theory of ‘normative expectations’ is clear about its rationale,234 but of course requires interpretation in practice. However, both Hart and Gardner recognize that the enquiry should not be entirely subjective, and that there are good grounds for expecting people to attain certain standards of behaviour. The idea of duties of citizenship, aired in the previous section, might thus be developed to broaden out the concept of desert. Although some of the objective requirements mentioned in this chapter are based on principles of protection, it should not be thought that all of them are derogations from a properly social theory of individual autonomy.235 Desert theory—maintaining that individuals should be liable to punishment only when they deserve it, and to the extent that they deserve it236 —may find its application in one of three forms in modern writings.237 One is the character theory, which argues that D's ‘desert’ is ‘gauged by his character’ and therefore that ‘a judgment about character is essential to the just distribution of punishment’.238 Behaviour should be excused when it does not reflect D's true character, but D should be held responsible whenever the behaviour can be regarded as genuinely expressive of his dispositions.239 The Court of Appeal came close to espousing this theory in Kingston,240 when it held that D should not be held liable for acts done whilst involuntarily (but not totally) intoxicated. Full espousal of the theory would have excused D if he had no general disposition to paedophilia, but would have convicted him if paedophilia was part of his general character. There are several difficulties with this approach, one of which is the breadth of the conception of character it employs (although that, in turn, raises the question of one's responsibility for one's character), and another is its lack of sharpness in distinguishing between acceptable and unacceptable excuses.241 Fletcher's attempt to limit the theory to the particular act charged, by reference to the principle of legality and the value of privacy, is unconvincing.242 A second approach is choice theory, emphasizing respect for D's autonomy and for the choices he or she made and not imposing liability for conduct which cannot be said to be chosen. Gardner's ‘normative (p. 235) expectations’ theory falls into this category, as does much of Hart's famous doctrine of fair opportunity. A third strand may be found in capacity theory, which focuses on D's capacity to conform conduct to the law's requirements. Although some regard this as a general rationale, it can be argued that its proper place is as a supplement to choice theory, not denying or altering ‘normative expectations’ theory, but adding to it a further ground of (partial) defence for those unable to attain the objective standards inherent in the ‘normative expectations’ approach.243

Page 31 of 44

Excusatory Defences Acknowledging that there are some who cannot attain the general normative standards requires an assessment of other principles, facts, and rationales, as we have seen (for example) when examining the case for a defence based on social deprivation. But this is not to reject a framework based on desert, however, since that would be to reject the foundations for many of the safeguards and protections for individuals that are constructed out of respect for autonomy, and that is not the road we should go down.

Further Reading H. L. A. HART, Punishment and Responsibility (2nd edn., 2008), chs 2 and 7, and Introduction by J. Gardner. S. Kadish, Blame and Punishment (1987), ch 5. J. Gardner, Offences and Defences (2007), chs 4, 6, and 7. J. Horder, Excusing Crime (2004), ch 3. V. Tadros, Criminal Responsibility (2005), chs 10 and 11. R. A. Duff, Answering for Crime (2007), ch 11. P. Westen, ‘An Attitudinal Theory of Excuse’, (2006) 25 Law and Philosophy 289.

Notes: 1 See P. H. Robinson, ‘Criminal Law Defenses: A Systematic Analysis’ (1982) 82 Columbia LR

199, Structure and Function in Criminal Law (1997), ch 5, and Criminal Law Defences (1984), for a five-fold classification of defences: (i) failure of proof defences; (ii) offence modifications (e.g. withdrawal in complicity); (iii) justifications; (iv) excuses; and (v) non-exculpatory publicpolicy defences (e.g. time limitations). This chapter is concerned with (iv) and with some forms of (i). 2 J. Gardner, Offences and Defences (2007), ch 6. 3 G. Dingwall, Alcohol and Crime (2006); the British Crime Survey reports that 45 per cent of

victims of violent incidents believed the offender(s) to be influenced by alcohol and 19 per cent believed their offender(s) to be influenced by drugs: C. Kershaw et al., Crime in England and Wales 2007/08 (2008), 76–7. 4 See Chapter 5.4(d). 5 [1963] AC 349, at 382. 6 See the discussion in section 6.2(d). 7 The phrase of Lord Hailsham in DPP v Morgan [1976] AC 182, at 214, criticized in Chapter

5.5(d) and section 6.5. 8 Page 32 of 44

Excusatory Defences 8 [1977] AC 443. 9 See the discussion in A. Ward, ‘Making Some Sense of Self-induced Intoxication’ [1986] CLJ

247. 10 [2008] QB 43, on which see D. Ormerod, [2007] Crim LR 654. 11 [1972] 3 All ER 962. 12 S. White, ‘Offences of Basic and Specific Intent’ [1989] Crim LR 271. 13 [1982] AC 341. 14 The discussion of ‘drunken accidents’ in Heard pays little regard to the possibility of

recklessness in certain situations (reference at n 10). 15 Aitken (1992) 95 Cr App R 304, Richardson and Irwin [1999] 1 Cr App R 392. 16 Lipman [1970] 1 QB 152; see Chapter 4.2. 17 (1987) 85 Cr App R 315; there is debate about whether this ruling was merely obiter dictum,

but it has been applied in O'Connor [1991] Crim LR 135 and in Hatton. 18 [2006] 1 Cr App R 16. 19 See also Fotheringham (1989) 88 Cr App R 206, on rape. 20 [1977] AC 443, at 475. 21 See Chapter 5.5(c) for discussion. 22 See J. C. Smith, ‘Intoxication and the Mental Element in Crime’, in P. Wallington and R. Merkin

(eds), Essays in Honour of F. H. Lawson (1987). 23 [1977] AC 443, at 475; Woods (1981) 74 Cr App Rep 312. See C. Wells, ‘Swatting the

Subjectivist Bug’ [1982] Crim LR 209. 24 Voiced by majority judges in the High Court of Australia, in O'Connor (1980) 146 CLR 64. 25 See Chapter 5.4(d) and (e). 26 See Keogh [1964] VR 400, and O'Connor (1980) 146 CLR 64 in Australia, and Kamipeli

[1975] 2 NZLR 610 in New Zealand: compare G. Orchard, ‘Surviving without Majewski—a View from Down Under’ [1993] Crim LR 426 with S. Gough, ‘Surviving without Majewski?’ [2000] Crim LR 719. 27 Cf. S. Gough, ‘Intoxication and Criminal Liability: the Law Commission's Proposed Reforms’

(1996) 112 LQR 335, at 337. 28 Bailey (1983) 77 Cr App R 76, per Griffiths LJ at 80. 29 J. Horder, ‘Pleading Involuntary Lack of Capacity’(1993) 52 Camb LJ 298, at 308–9.

30 Page 33 of 44

Excusatory Defences 30 See P. H. Robinson, ‘Causing the Conditions of One's Own Defence’ (1985) 71 Virginia LR 1,

at 50–1, discussed in Chapter 5.4(e). 31 For the uncompromising judicial response to repeated offences of drunken violence, see

Sheehan and O'Mahoney [2007] 1 Cr App R (S) 149 and McDermott [2007] 1 Cr App R (S) 145. 32 [1988] Crim LR 698. 33 (1985) 80 Cr App R 157. 34 This follows the reasoning in Bailey (1983) 77 Cr App R 76 on diabetes and automatism:

see Chapter 4.2. 35 [1995] 2 AC 355. 36 His Lordship concluded that the few distant authorities in favour of the defence were

unpersuasive, and so the House of Lords (rightly) considered the issue afresh. 37 For an example see Blakely and Sutton [1991] Crim LR 763. 38 G. R. Sullivan, ‘Making Excuses’, in A. P. Simester and A. T. H. Smith (eds), Harm and

Culpability (1996). For discussion of character-based theories of excuse, see section 6.7(b). 39 See n 26. 40 R. Shiner, ‘Intoxication and Responsibility’ (1990) 13 Int J Law and Psychiatry 9; C. N.

Mitchell, ‘The Intoxicated Offender—Refuting the Legal and Medical Myths’ (1988) 11 Int J Law and Psychiatry 77. 41 Gough, ‘Surviving without Majewski?’, also discussing the Canadian decision in Daviault

(1995) 118 DLR (4th) 469 and its consequences. 42 Gough, ‘Intoxication and Criminal Liability’. 43 Cf. H. Fingarette, ‘Addiction and Criminal Responsibility’ (1975) 84 Yale LJ 413 with J. Tolmie,

‘Alcoholism and Criminal Liability’ (2001) 64 MLR 688. 44 Law Commission Consultation Paper No. 127, Intoxication and Criminal Liability (1993).

German law adopts this approach, allowing intoxication to negative intention (applying the ‘inexorable logic’) but then applying an offence of ‘dangerous intoxication’ that consists of committing the conduct element of another offence while culpably intoxicated: see J. R. Spencer and A. Pedain, ‘Strict Liability in Continental Criminal Law’, in A. P. Simester (ed), Appraising Strict Liability (2005), 244–5. 45 Home Office, Violence: Reforming the Offences Against the Person Act 1861 (1998), draft

Bill, cl. 19, based on the criminal code proposals in Law Com No. 177, draft Bill, cl. 22. An intervening report from the Law Commission, Law Com No. 229, Legislating the Criminal Code: Intoxication and Criminal Liability (1995), was not adopted in its central recommendations. 46 Sir John Smith rightly questioned (b), which might have unexpected consequences in

attributing to people beliefs they did not hold: ‘Offences Against the Person: the Home Office Page 34 of 44

Excusatory Defences Consultation Paper’ [1998] Crim LR 317, at 321. 47 Law Commission, Intoxication and Criminal Liability (No. 314, 2009). 48 [2008] QB 43 (CA). 49 Cf. A. Ashworth, ‘Intoxication and the General Defences’ [1980] Crim LR 556 with Gough,

‘Intoxication and Criminal Liability’. See now R. Williams, ‘Voluntary Intoxication – A Lost Cause?’ (2012) Law Quarterly Review (forthcoming). 50 See Willer (1986) 83 Cr App R 225, Conway (1988) 88 Cr App R 159, Martin (1989) 88 Cr

App R 343, discussed by D. W. Elliott, ‘Necessity, Duress and Self-Defence’ [1989] Crim LR 611. 51 Rodger and Rose [1998] 1 Cr App R 143 (D's own suicidal tendencies cannot found either

defence). 52 [1971] 2 QB 202. 53 [1989] 3 All ER 1025. 54 [1989] 3 All ER 1025, discussed earlier. 55 DPP for Northern Ireland v Lynch [1975] AC 653; Bowen [1996] 2 Cr App R 157. 56 In, respectively, DPP v Lynch [1975] AC 653 at 687, and Valderrama-Vega [1985] Crim LR

220. 57 [1947] KB 997. 58 Graham (1982) 74 Cr App R 235, confirmed by the House of Lords in Howe [1987] AC 417. 59 [1996] 2 Cr App R 157. 60 For the suggestion that this phrase has wider implications than the Court realized, see A.

Buchanan and G. Virgo, ‘Duress and Mental Abnormality’ [1999] Crim LR 517. 61 Coroners and Justice Act 2009, discussed in Chapter 7.4(b). 62 See further J. Horder, Excusing Crime (2004), 183–5. 63 [1985] Crim LR 220; P. Alldridge, ‘Developing the Defence of Duress’ [1986] Crim LR 433. 64 Valderrama-Vega (reference at n 63); Gill (1963) 47 Cr App R 166; and Law Com No. 83,

Defences of General Application (1977), 2–3. 65 However, it has been held that the defence is unavailable where D himself is the source of

the danger, through his (conditional) determination to commit suicide: Rodger and Rose [1998] 1 Cr App R 143. 66 [2001] 1 WLR 2206. 67 Lord Bingham in Hasan [2005] 2 AC 467 approved this formulation as ‘consistent with the

Page 35 of 44

Excusatory Defences rationale’ of duress (para. 21(3)). 68 See n 52 and accompanying text. 69 [1999] Crim LR 570. 70 [2005] 2 AC 467. The decision is also known as Z. 71 [2005] 2 AC 467, para. 28. 72 [1971] 2 QB 202, at 207. 73 See Chapter 5.5(d) and section 6.4; the Divisional Court erroneously applied this general

approach to duress in DPP v Rogers [1998] Crim LR 202. 74 (1982) 74 Cr App R 235; much of Lord Lane's judgment proceeds on an analogy with

provocation, even though the preponderance of authority favours a subjective test for belief in provocation cases—see section 6.4, and W. Wilson, ‘The Structure of Criminal Defences’ [2005] Crim LR 108. 75 [2003] Crim LR 721. 76 It is tolerably clear from the strong objectivism of Lord Bingham's speech in Hasan [2005] 2

AC 467, notably at para. 38, that he would support the Graham test. 77 Discussed in Chapter 5.4(d). 78 [1987] QB 853. 79 (1988) 86 Cr App R 47. 80 [2005] 2 AC 467. 81 Overruling Baker and Ward [1999] 2 Cr App R 335 on this point. 82 [2005] 2 AC 467, at para. 38; see also Ali [2008] EWCA 716. 83 [2005] 2 AC 467, para. 77. 84 See Chapter 4.2. 85 G. P. Fletcher, Rethinking Criminal Law (1978), 803, adopted by Dickson J in the Supreme

Court of Canada in Perka v R (1984) 13 DLR (4th) 1. For discussion see C. Wells, ‘Necessity and the Common Law’ (1985) 5 OJLS 471. 86 Acceptance of the ‘moral involuntariness’ rationale might also raise questions about the

law's rejection of social and financial pressures as grounds of defence: see section 6.8(b). 87 Discussed in Chapter 4.8. 88 Gardner, Offences and Defences, ch 6; Horder, Excusing Crime, 99–109. 89 See the argument of C. Clarkson, ‘Necessary Action: a New Defence’ [2004] Crim LR 81. 90 Page 36 of 44

Excusatory Defences 90 [1996] 2 Cr App R 157, n 59 and accompanying text. 91 E.g. where negligence is the fault element for crimes (see Chapter 5.5(d)), and to a small

extent in self-defence (see Chapter 4.6(g)). 92 Cf. Horder, Excusing Crime, 183–5, and the discussion in section 6.7. 93 See section 6.8(a), and M. Wasik, ‘Duress and Criminal Responsibility’ [1977] Crim LR 453. 94 J. Bentham, Introduction to the Principles of Morals and Legislation (1789), ch XIV, para. 9. 95 Bentham, Introduction to the Principles of Morals and Legislation, para. 11. 96 (1884) 14 QBD 273. 97 [1977] AC 653. 98 [1977] AC 755; cf. I. Dennis, ‘Duress, Murder and Criminal Responsibility’ (1980) 96 LQR

208. 99 [1987] AC 417. 100 The House of Lords held in Gotts [1992] 2 AC 412 that, by logical extension, duress should

not be available as a defence to attempted murder. 101 Per Lords Griffiths and Mackay, at 446 and 457. 102 Chapter 4.8. 103 P. Alldridge, ‘The Coherence of Defences’ [1983] Crim LR 665. 104 A. Kenny, Freewill and Responsibility (1978), 38. 105 Coroners and Justice Act 2009, s. 55(3). See further, Chapter 7. 106 Hasan [2005] 2 AC 467 (HL), para. 28. 107 LCCP 177, A New Homicide Act? (2005), Part 7. 108 Law Com No. 304, Murder, Manslaughter and Infanticide (2006), Part 6; the earlier report

adopting the same approach was Law Com No. 218, Legislating the Criminal Code: Offences against the Person and General Principles (1993). 109 Law Com No. 304, Murder, Manslaughter and Infanticide (2006), Part 6, para. 6.46. 110 [2007] QB 960. 111 Law Com No. 304, paras. 6.53 and 6.142–3. 112 Law Com No. 304, para. 6.79. 113 For criticism of this recommendation, see A. Ashworth, ‘Principles, Pragmatism, and the Law

Commission's Recommendations on Homicide Law Reform’ [2007] Crim LR 333, at 340–2. 114 Page 37 of 44

Excusatory Defences 114 The Ministry of Justice's Consultation Paper 19 on Murder, Manslaughter and Infanticide:

Proposals for Reform of the Law (2008), much discussed in Chapter 7, does not cover the duress recommendations. 115 (1889) 23 QBD 168. 116 Compare E. Griew, ‘States of Mind, Presumptions and Inferences’, in P. F. Smith (ed.),

Criminal Law: Essays in Honour of J. C. Smith (1987), with A. P. Simester, ‘Mistakes in Defence’ (1992) 12 OJLS 295, and R. H. S. Tur, ‘Subjectivism and Objectivism: Towards Synthesis’, in Shute, Gardner, and Horder (eds), Action and Value in Criminal Law (1993). 117 Confirmed by the House of Lords in Sweet v Parsley [1970] AC 132. 118 [1976] AC 182, discussed in Chapter 5.5(d). 119 The leading cases were probably Rose (1884) 15 Cox CC 540 and Chisam (1963) 47 Cr

App R 130. The only careful analysis was that of Hodgson J in the Divisional Court in Albert v Lavin (1981) 72 Cr App R 178. Cf. however the subjective approach to mistake in provocation cases: Letenock (1917) 12 Cr App R 221, Wardrope [1960] Crim LR 770. 120 Kimber (1983) 77 Cr App R 225, followed by Gladstone Williams (1984) 78 Cr App R 276

and by Beckford [1988] 1 AC 130. 121 (1984) 78 Cr App R 276. 122 S. Uniacke, Permissible Killing (1994), discussed in Chapter 4.6, would say that D's

conduct was agent-perspectivally permitted to act but not objectively permitted. 123 See n 120. 124 Cf. the discussion by S. Yeo, Compulsion in the Criminal Law (1991), 198–208. 125 Simester, ‘Mistakes in Defence’; see also Gardner, Offences and Defences, ch 5. 126 McCann v UK, discussed in Chapter 4.6(g); see also Simester, 307 (reference at n 125). 127 Cf. A. Brudner, ‘Agency and Welfare in the Penal Law’, in Shute, Gardner, and Horder

(eds), Action and Value in Criminal Law, 35 and 43. 128 See Chapter 5.5(d) and Chapter 8.5(c). 129 See Chapter 8.6(d), discussing G. [2008] UKHL 37. 130 Compare Beckford [1988] AC 130 with McCann v UK, Chapter 4.7(f)(vi); cf. J. Horder,

‘Cognition, Emotion and Criminal Culpability’ (1990) 106 LQR 469. The High Court of Australia has required ‘reasonable grounds’ in all cases of mistaken self-defence: Zecevic v R (1987) 162 CLR 645. 131 See, e.g. the provision in s. 76(7) of the Criminal Justice and Immigration Act 2008, that in

determining whether force was reasonable in self-defence a court should take account of ‘what the person honestly and instinctively thought necessary’, discussed in Chapter 4.6(g). 132 Page 38 of 44

Excusatory Defences 132 See section 6.3(c), and the Law Commission Consultation Paper No. 139, Consent in the

Criminal Law (1995), ch 7. 133 See J. Horder, ‘Killing the Passive Abuser’, in S. Shute and A. P. Simester (eds), Criminal

Law Theory: Doctrines of the General Part (2002). 134 R. A. Duff, Answering for Crime (2007), 270–6. 135 For recent affirmations see B v DPP [2000] 2 AC 428 and K [2002] 2 AC 462 discussed in

Chapter 5.5(a). 136 (1982) 74 Cr App R 235. 137 [2005] 2 AC 467, discussed in part 3(a) of this chapter. 138 See the discussion in Chapter 4.6(f). 139 Theft Act 1968, s. 2(1)(a); Criminal Damage Act 1971, s. 2.5(2)(a). 140 Secretary of State for Trade and Industry v Hart [1982] 1 WLR 481. 141 Williams, Textbook of Criminal Law, ch 20. 142 E.g. Grant v Borg [1982] 1 WLR 638, Jones, The Times, 19 August 1994. 143 Statutory Instruments Act 1946, s. 3(2). 144 Lim Chin Aik v R [1963] AC 160; see also Toulson LJ in Chambers [2008] EWCA Crim 2467. 145 Christian v R. [2006] 2 AC 400 (defendants from Pitcairn Island knew rape and sexual

abuse were seriously wrong and criminal, though unaware of terms of English law); see H. Power, ‘Pitcairn Island: Sexual Offending, Cultural Difference and Ignorance of the Law’ [2007] Crim LR 609. 146 O. W. Holmes, The Common Law (1881), 48. 147 J. Hall, General Principles of Criminal Law (1960), 388, and Chapter 4.1. 148 Fletcher, Rethinking Criminal Law, 734, and Chapter 4.1. 149 See Chapter 5.4(a). 150 J. Raz, The Morality of Freedom (1987), 206–7, and Chapter 4.2 and 4.3. 151 In Chapter 5.5(d) and in section 6.4. 152 A. Ashworth, ‘Ignorance of the Criminal Law and Duties to Avoid it’ (2011) 74 MLR 1, and R.

Goodin, ‘An Epistemic Case for Legal Moralism’ (2010) 30 OJLS 615. 153 Cf. D. Husak, ‘Ignorance of Law and Duties of Citizenship’ (1994) 14 Legal Studies 105,

110: ‘the problem arises from the fact that few persons are likely to be aware of the existence of the alleged duty to know the law’. 154 Page 39 of 44

Excusatory Defences 154 Theft Act 1968, s. 21(1), discussed in Chapter 9.4. 155 See Chapter 3.5(h). 156 Cantoni v France (1997) VIII HRCD 130, on the French offence of selling prohibited

pharmaceutical products; cf. Chapter 3.5(g) on Art. 7. 157 Husak, ‘Ignorance of Law’, at 115, rightly emphasizes that the state has duties as well as

citizens. 158 See the American case of Lambert v California (1957) 355 US 225 on omissions (and

Chapter 4.4(c)). 159 See Lord Bingham in Rimmington and Goldstein [2006] 1 AC 459 at [30]; and Chapter

3.5(i). 160 Cf. D. Husak and A. von Hirsch, ‘Culpability and Mistake of Law’, in Shute, Gardner, and

Horder (eds), Action and Value. 161 See Husak and von Hirsch, ‘Culpability and Mistake of Law’, proposing that the only way of

avoiding unfairness is to allow courts to assess the moral legitimacy of D's beliefs. Cf. the remark of Brooke LJ in R (on application of W) v DPP [2005] EWHC Admin 1333, that a boy of 14 ‘might well not know what was a criminal offence and what was not’. 162 Thus the distinction drawn by Brudner, ‘Agency and Welfare’, 36, between ignorance of

‘true crimes (as distinct from welfare offences)’ is not convincing, since there may be strong duties in the latter category too. 163 J. Andanaes, ‘Error Juris In Scandinavian Law’, in G. Mueller (ed.), Essays in Criminal

Science (1961); cf. generally P. Brett, ‘Mistake of Law as a Criminal Defence’ (1966) 5 Melb U LR 179. 164 Law Com No. 177, cl. 21. 165 As in Thomas [2006] Crim LR 71, where D was unaware that the Sexual Offences Act 2003

had changed the law by criminalizing sexual acts by foster parents with former foster children under 18 (not 16). 166 Cf. Secretary of State for Trade and Industry v Hart [1982] 1 WLR 481, with Grant v Bord

[1982] 1 WLR 638, two decisions of the House of Lords in the same year; see generally A. T. H. Smith, ‘Error and Mistake of Law in Anglo-American Criminal Law’ (1984) 14 Anglo-American LR 3. Cf. Attorney-General's Reference (No. 1 of 1995) [1996] 2 Cr App R 320, where the absence of ‘knowingly’ was one factor in the Court's decision to hold that ignorance of the law was no excuse. 167 See R. Card, ‘Authority and Excuse as Defences to Crime’ [1969] Crim LR 359 and 415. 168 (1862) 7 H and N 707, discussed by Brett, ‘Mistake of Law as a Criminal Defence’. 169 [1975] QB 678. 170 Cambridgeshire and Isle of Ely CC v Rust [1972] 1 QB 426. 171 Page 40 of 44

Excusatory Defences 171 A. Ashworth, ‘Excusable Mistake of Law’ [1974] Crim LJ 652. 172 A. Ashworth, ‘Testing Fidelity to Legal Values’, in Shute and Simester (eds), Criminal Law

Theory, and the refinements proposed by Horder, Excusing Crime, 270–6. 173 Code for Crown Prosecutors, para. 4.17(d) (‘genuine mistake or misunderstanding’). 174 Postermobile v Brent LBC, The Times, 8 December 1997, discussed at [1998] Crim LR

435, and by Ashworth, ‘Testing Fidelity to Legal Values’, at 303. 175 Cf. Kingston [1995] 2 AC 355, discussed in section 6.3(d), where the House of Lords held

that it must be for Parliament to decide whether or not to introduce a new defence. 176 G. Williams, ‘The Draft Code and Reliance on Official Statements’ (1989) 9 Legal Studies

177, at 186–7; the Model Penal Code, s. 2.04(3), also allows reliance on official advice, but not a lawyer's advice, as a defence. 177 See further Ashworth, ‘Testing Fidelity to Legal Values’, at 306–7. 178 [2002] 1 Cr App R 29. 179 Per Lord Nicholls at [23]. 180 See the discussion by A. Ashworth, ‘Re-Drawing the Boundaries of Entrapment’ [2002]

Crim LR 161, and in ‘Testing Fidelity to Legal Values’, at 310–22. 181 See, e.g., Lord Nicholls at 1 and Lord Hoffmann at 39–40. 182 (1999) 28 EHRR 101. 183 Jacobson v US (1992) 112 S Ct 1535. 184 Section 2.13 (official inducement of offence; not available if offence involves bodily injury). 185 K. Hofmeyr, ‘The Problem of Private Entrapment’ [2006] Crim LR 319. 186 Cf. the decision of the Court of Appeal in Shannon [2000] 1 Cr App R 168 with that of the

Strasbourg Court in Shannon v UK [2005] Crim LR 133; see also Ashworth, ‘Re-Drawing the Boundaries of Entrapment’, at 175–6. 187 See Chapter 7.4(b) and (e). 188 As proposed by the Criminal Law Revision Committee in its 1976 Working Paper, ‘Offences

against the Person’. 189 See M. Wasik, ‘Partial Excuses in the Criminal Law’ (1982) 45 MLR 515, and Horder,

Excusing Crime, 143–52. 190 See Chapter 3.5(l). 191 Wasik, ‘Partial Excuses’, 524. 192 See section 6.6. 193 Page 41 of 44

Excusatory Defences 193 See section 6.3(c). 194 See Howe [1987] AC 417, but cf. the discussion in section 6.3(c). 195 The views of Baroness Wootton were discussed in Chapter 5.5(a). 196 See Chapter 1.4 on the prevalence of guilty pleas. 197 See A. Ashworth, Sentencing and Criminal Justice (5th edn., 2010), 372–6. 198 Ashworth, Sentencing and Criminal Justice, 372–6, Ch 5.7. 199 Crown Prosecution Service, Code for Crown Prosecutors, discussed in Chapter 1.4. 200 Looseley; Attorney General's Reference No. 3 of 1999 [2002] 1 Cr App R 29, discussed in

section 6.6. 201 Postermobile plc v Brent LBC, The Times, 8 December 1997, discussed in section 6.5(c). 202 Gardner, Offences and Defences, 124. 203 See Hasan [2005] 2 AC 467, and the theoretical discussion by A. Brudner, ‘A Theory of

Necessity’ (1987) 7 OJLS 338. 204 E. Colvin, ‘Exculpatory Defences in Criminal Law’ (1990) 10 OJLS 381, 395. 205 In duress, see the loosening in Bowen [1996] 2 Cr App R 157 and the tightening in Hasan

[2005] 2 AC 467, section 6.3(a); in provocation, see the loosening in Smith (Morgan) [2001] 1 AC 146 and the tightening in Attorney General for Jersey v Holley [2005] UKPC 23, Chapter 7.4(b). 206 For an argument that there should be an intermediate category of cases where the

essence of D's response is diminished capacity, see Horder, Excusing Crime, ch 3; R. Lippke, Rethinking Imprisonment (2007), 88–101. 207 P. Taylor and G. Dalton, ‘Pre-Menstrual Syndrome: a New Criminal Defense?’ (1983) 19 Cal

WLR 269; Sandie Smith [1982] Crim LR 531; and J. Dressler, ‘Reflections on Excusing Wrongdoers: Moral Theory, New Excuses and the Model Penal Code’ (1988) 19 Rutgers LJ 671, 707. 208 See section 6.3(b). 209 See Chapter 3.5(i). 210 In Hasan [2005] 2 AC 467, Lord Bingham's primary reason for taking a restrictive,

objectivist approach to the duress defence was one of ‘public policy’, including (para. 22) fear of false defences. 211 For other arguments in favour of taking coercive measures against those acquitted on

certain grounds, see Colvin, ‘Exculpatory Defences’, 392. 212 See Chapter 5.4(a) and (d). 213 Page 42 of 44

Excusatory Defences 213 M. Kelman, ‘Interpretive Construction in the Substantive Criminal Law’ (1981) 33 Stanford

LR 591. 214 Horder, Excusing Crime, Chapter 4, and Chapter 7.4(b). 215 Section 6.5(c). 216 Sexual Offences Act 2003, s. 1(1)(c), and Chapter 8.5(i). 217 Cf. Sexual Offences Act 2003, s. 9(1)(c) with the decisions of the House of Lords in B v

DPP [2000] 2 AC 428 and K. [2002] 2 AC 462, discussed in Chapter 5.5(a). 218 E.g. Sexual Offences Act 2003, ss. 5–8, interpreted in G. [2008] UKHL 37 and criticized in

Chapter 8.6. 219 See McCann v UK, discussed in Chapter 4.6(f)(vi), and Gardner, Offences and Defences,

128–30. 220 Fletcher, Rethinking Criminal Law, 514; V. Tadros, Criminal Responsibility (2005),

Chapter 3.3. 221 [1982] AC 341, discussed in Chapter 5.3(c). 222 See the discussion in Lippke, Rethinking Imprisonment, 88–101. 223 Cf. criminal damage by fire, which is often serious. 224 Council of Europe, European Convention on Compensation for the Victims of Crimes of

Violence (1984). 225 For reviews, see D. Farrington, ‘Childhood Risk Factors and Risk-Focused Prevention’, and

D. Smith, ‘Crime and the Life Course’, in M. Maguire, R. Morgan, and R. Reiner (eds), Oxford Handbook of Criminology (4th edn., 2007). 226 See text at nn 209–212. 227 N. Lacey, State Punishment (1988), ch 3 and 140–1. 228 See, e.g., the discussion of Howe, Chapter 6.3(d). 229 For an introduction to the literature, see Fletcher, Rethinking Criminal Law, ch 10; K.

Greenawalt, ‘The Perplexing Borders of Justification and Excuse’ (1984) 84 Columbia LR 1897; J. Dressler, ‘Justifications and Excuses: a Brief Review of the Concept and the Literature’ (1987) 33 Wayne LR 1155; G. Williams, ‘The Theory of Excuses’ [1982] Crim LR 732; W. Wilson, Central Issues in Criminal Theory (2002), chs 10 and 11; R. A. Duff, Answering for Crime (2007), ch 11. 230 See D. Husak, ‘The Serial View of Criminal Law Defences’ (1992) 3 Crim L Forum 369,

developed by Horder, Excusing Crime, ch 3. 231 See further Chapter 7.5.

232 Page 43 of 44

Excusatory Defences 232 P. Westen, ‘An Attitudinal Theory of Excuse’ (2006) 25 Law and Philosophy 289, 330. 233 H. L. A. Hart, Punishment and Responsibility (2nd edn., 2008), chs 2 and 7, and the re-

assessment in J. Gardner's ‘Introduction’, xxxiv–liii. 234 Gardner, Offences and Defences, ch 5. 235 See Chapter 4.1 for discussion. 236 See Chapter 1.5. 237 See Westen, ‘Attitudinal Theory of Excuse’. 238 Fletcher, Rethinking Criminal Law, 800. 239 For analysis and discussion, see Lacey, State Punishment, 65–78; Horder, Excusing

Crime, ch 1; Tadros, Criminal Responsibility, ch 1. 240 Discussed in section 6.2(d); see further Sullivan, ‘Making Excuses’. 241 Cf. Dressler, ‘Reflections on Excusing Wrongdoers’, 692–701, with Tadros, Criminal

Responsibility, ch 11. 242 Fletcher, Rethinking Criminal Law, 800, criticized by Brudner, ‘A Theory of Necessity’,

344–7. 243 Horder, Excusing Crime, ch 3.

Homicide

Principles of Criminal Law (7th edn) Andrew Ashworth and Jeremy Horder Publisher: Oxford University Press Print ISBN-13: 9780199672684 DOI: 10.1093/he/9780199672684.001.0001

Print Publication Date: May 2013 Published online: Sep 2013

7. Homicide Chapter: (p. 236) 7. Homicide Author(s): Andrew Ashworth and Jeremy Horder DOI: 10.1093/he/9780199672684.003.0007

7.1 Death and finality 7.2 The conduct element: causing death 7.3 Defining murder: the inclusionary question 7.4 Defining murder: the exclusionary question 7.5 ‘Involuntary manslaughter’ 7.6 Causing or allowing the death of a child or vulnerable adult 7.7 Causing death by driving 7.8 Reviewing the structure of the law of homicide Further reading There are a surprising number and variety of homicide and homicide-related offences in English law.1 This chapter deals with the most important instances in which the criminal law prohibits and punishes behaviour that causes death or risks causing death. Murder, manslaughter, and several other homicide offences are discussed, and one recurrent issue

Homicide here is fair labelling: does English law respond proportionately to the different degrees of culpability manifested in cases where death is caused?

7.1 Death and finality The culpable causing of another person's death may fairly be regarded as the most serious offence in the criminal calendar. It is sometimes argued that treason or terrorism are more serious offences, since they may strike at the very foundations of the State and its social organizations, but this line of thinking has little substance. Both treason and terrorism-related offences cover a large range of conduct, much of which is very remote from any harm done.2 Treason-related offences, for example, include the common law offence of ‘compounding treason’. This involves agreeing, in exchange for some benefit, not to prosecute someone who has committed treason. This is wrongdoing that in (p. 237) all probability would now be treated rather less dramatically either as bribery or as perverting the course of justice. The harm caused by homicide is absolutely irremediable, whereas the harm caused by many other crimes is remediable to a degree. Even in crimes of violence which leave some permanent physical disfigurement or psychological effects, the victim retains his or her life and, therefore, the possibility of further pleasures and achievements, whereas death is final. This finality makes it proper to regard death as the most serious harm that may be inflicted on another person, and to regard the culpable causing of death without justification or excuse as the highest wrong. Although many deaths arise from natural causes, and many others from illnesses and diseases, each year sees a large number of deaths caused by ‘accidents’, and also a number caused by acts or omissions which amount to some form of homicide in English law. In 2000, for example, the statistics showed that there were some 13,000 accidental deaths, of which some 3,200 occurred on the roads and the remainder either at work or in the home.3 By comparison, the number of deaths recorded as criminal homicide is much smaller: it rose from around 600 per year at the start of the 1990s to around 700 per year in the early years of this century, since when it has fallen back to 638 in 2010–11 and to 550 in 2011–12.4 This includes all the murders and manslaughters, but it leaves further questions to be confronted. For example, are we satisfied that the 700 deaths recorded as homicide are in fact more culpable than all, or even most, of the deaths recorded in other categories? In other words, does English criminal law pick out the most heinous forms of killing as murders and manslaughters, or are the boundaries frozen by tradition? For example, the number of offences of causing death by dangerous driving, causing death by careless driving whilst intoxicated, and causing death by careless driving now stands at around 400 per year5: some of these offences result in sentences more severe than those handed down for some forms of manslaughter,6 which prompts the question whether these offences should be brought into manslaughter or other offences should be removed from that category. We will see below that Parliament has created four new homicide offences in the last few years. For one of them (corporate manslaughter) it has used the term manslaughter. For the others it has either used the terminology of ‘causing death by’ (the two offences introduced by the Road Safety Act 2006)7 or provided no label at all (the offence under s. 5 of the Domestic Violence, Crime and Victims Act 2004).8 Various questions may be raised: are these labelling decisions acceptable? Are these extensions of homicide law defensible, or is the distance Page 2 of 69

Homicide between the defendant's fault and the (p. 238) consequent death too remote? What implications, if any, should the different labels have for sentence levels? These and other problems in the reform of homicide law will be examined, after the contours of the present law have been discussed.

7.2 The conduct element: causing death English law distinguishes between the offences of murder and manslaughter, as we shall see, but the two crimes do have a common conduct element. It must be proved that the defendant's act or omission caused the death of a human being. The requirements of causation in the criminal law were discussed in Chapter 4.5, and already some problems came to light. Thus, the standard doctrine is that to shorten life by days is to cause death no less than shortening it by years, and this raises questions about the liability of doctors who administer drugs which they know will have the effect of shortening life, even though their primary purpose is to relieve pain. We noted that, in the rare trials of doctors for murder, the approach has been to direct the jury (in effect) to determine whether the doctor's primary motive was to relieve pain or to accelerate the patient's death9 —an approach that conflicts with the orthodox approach to intention. It would be more consistent with prevailing doctrine for the courts to accept that a doctor may be clinically justified in administering a form of treatment he or she foresees as almost certain to hasten death (so long as death is not directly intended), but may have a suitably refined defence of clinical necessity in this situation, in spite of fulfilling both the conduct and fault elements for murder.10 At what points does an organism start and cease to be a person within the protection of the law of homicide? The current view, both in English law and in that of many other European countries,11 is that a foetus is not yet a person and therefore cannot be the victim of homicide. Thus, in the difficult case of Attorney-General's Reference (No. 3 of 1994),12 where D had stabbed his pregnant girlfriend, also injuring the foetus, and the child was born prematurely and died some four months later from the wound, the House of Lords held that the doctrine of transferred intention could not be applied because it could only operate to transfer intention from one person to another, and not from a person to (what was at the relevant time) a foetus. Only when the child is born alive and has an existence independent of its mother does it come within the protection of the law of homicide, although there are other serious offences capable (p. 239) of commission before birth, notably child destruction (which carries a maximum of life imprisonment).13 The point at which the protection of the law of homicide begins was a crucial factor in the case of Re A (Conjoined Twins: Surgical Separation).14 The twins were conjoined and both would have died within months if left conjoined, but the stronger twin had good prospects of survival if surgical separation was performed. The Court of Appeal held that the weaker twin was sufficiently capable of independent breathing to be classed as a human being: she was independent of her mother, even though she was dependent on the vital organs of her twin for survival. Once it was decided that she was a person within the protection of the law of homicide, it followed that the operation to separate the twins would constitute the conduct element of murder in relation to the weaker twin (who would inevitably die shortly afterwards) unless there was some legal justification for the homicide, which the Court, invoking a version of necessity, held that there was. It seems that a person will be treated as dead if he or she has become irreversibly ‘brain 15 Page 3 of 69

Homicide dead’, the definition of brain death being largely left to medical practice.15 Thus switching off the life support machine of someone who already fulfils the criteria of brain death would not amount to the conduct element of murder. What if the patient does not fulfil those criteria, but is in a persistent vegetative state? This was the situation in the case of Tony Bland, who was being kept alive by food from a naso-gastric tube and by occasional administrations of antibiotics.16 The House of Lords held that it would be lawful to discontinue treatment, thus allowing the patient to die. The elements of criminal homicide would not be present, they held, because discontinuing treatment was not causing death: it was allowing the patient to die of his pre-existing condition. Discontinuing treatment was properly regarded as an omission, not as an act. Further, it was not a criminal omission because there was no duty to treat the patient, given that there was no hope of recovery and it was no longer in his best interests to be kept alive. The controversial aspects of this decision cannot be pursued here.17 At common law there was also a rule that a person could only be convicted of a homicide offence if the death occurred within a year and a day of the accused's act or omission. Advances in medical science now make it possible for some victims to be kept alive for years after being injured or wounded, and the argument that the passage of years should not prevent a homicide prosecution was accepted by Parliament in the Law Reform (Year and a Day Rule) Act 1996, which abolished the old rule. The Act relies on prosecutorial discretion to prevent oppressive or unfair prosecutions: s. 2 provides that, where more than three years have elapsed since the injury and the defendant has already been convicted of a non-fatal offence, a prosecution for homicide may only be (p. 240) instituted with the Attorney General's consent. It would be helpful to see the publication of some principles or guidelines on which this discretion should be exercised.

7.3 Defining murder: the inclusionary question (a) The procedural context If causing death is to be regarded as the most serious harm, it would seem to follow that the most blameworthy form of homicide (the greatest wrong) should result in the most severe sentences imposed by the courts. Indeed, many systems of criminal law impose a mandatory sentence for murder (or whatever the highest form of homicide is called in that system). In some jurisdictions this is a mandatory sentence of death.18 In the United Kingdom the penalty for murder is the mandatory sentence of life imprisonment.19 The existence of the mandatory sentence has a significant impact on the shape and content of the remainder of the law of homicide: as we shall see, the dividing line between murder and manslaughter may be affected by the inability of courts to give different sentences for murder, and there are those who believe that the strongest reason for retaining provocation or ‘loss of control’ as a partial defence to murder is that otherwise a judge could not reflect degrees of culpability in the sentence (for murder). The mandatory sentence of life imprisonment is divided into three portions: the first is now known as the minimum term (formerly, the tariff period), and is intended to reflect the relative gravity of the particular offence. It is a term that is served in full, and the early release provisions applicable to all determinate custodial sentences do not apply here. Once the minimum term expires, the second part consists of imprisonment based on considerations of public protection, and a murderer who is thought still to present a danger may be detained until Page 4 of 69

Homicide the Parole Board decides that it is safe to order release. The third portion is after release from prison: the offender remains on licence for the rest of his life. Although until 2003 the Home Secretary had the final say on the minimum term and ultimate release, those decisions have now passed to the courts and the Parole Board respectively.20 However, the government wished to fetter the judges so as to ensure that minimum terms were not set too low. Thus s. 269 of the (p. 241) Criminal Justice Act 2003 requires a court, when setting the minimum term to be served by a person convicted of murder, to have regard to the principles set out in Sch 21 of the Act. The structure of that Schedule is to indicate four starting points: • (for those over 21 at the time of the offence) a whole life minimum term for exceptionally serious cases, such as premeditated killings of two people, sexual or sad*stic child murders, or politically, religiously, or ideologically motivated murders. • (for those over 18 at the time of the offence) 30 years for particularly serious cases such as murders of police or prison officers on duty, murders involving firearms or explosives, sexual or sad*stic killings, or murders aggravated by racial or sexual orientation, or cases that would have attracted a ‘whole life’ term starting point, had the offender been 21 at the time of the offence. • (for those over 18 at the time of the offence) 25 years for murder committed with a knife or other weapon intentionally taken to the scene in order to commit the crime, or to have the weapon available for use, and it was used. • (for those over 18 at the time of the offence) 15 years for other murders not falling within either of the higher categories. It should be borne in mind that to compare the minimum term with a determinate sentence one should double it: in other words, a minimum term of 15 years is the equivalent of a determinate sentence of about 30 years.21 However, the language of Sch 21 leaves considerable latitude to the sentencing judge. Although criteria are enumerated for the whole life and 30-year starting points, they are expressed as factors that would ‘normally’ indicate such a sentence. There is then provision for the court to take account of any further relevant factors, and an explicit statement that ‘detailed consideration of aggravating and mitigating factors may result in a minimum term of any length (whatever the starting point)’. The Lord Chief Justice amended the previous guidance to reflect the 2003 provisions when he issued a Practice Direction in May 2004.22 He has subsequently emphasized that s. 269(3) merely states that the judge must specify the minimum term that ‘the court considers appropriate’, and indeed went on to say that so long as the judge bore in mind the principles set out in Sch 21, ‘he is not bound to follow them’—although an explanation for departing from them should be given.23 Nonetheless, anomalies can arise. Suppose that a farmer's wife chooses to accede to her terminally ill husband's request by shooting him with a shotgun. On the face of it, in such a case, 30 years is the indicated minimum and premeditation may take the minimum term even higher, although the sentencing judge might have to balance these factors against the mitigating factors mentioned in para. 11(d) and 11(f) (p. 242) of Sch 21, namely provocation brought about by prolonged stress, and the fact that the killing was intended as an act of mercy.24 The system introduced by the 2003 Act means that judges can vary the minimum term to reflect degrees of culpability in murder, but the overall framework of the mandatory sentence means that judges cannot set the maximum term to be served, as they do for other serious

Page 5 of 69

Homicide offences. On expiry of the minimum term, release is determined on public protection grounds by the Parole Board. This constraining effect of the mandatory life sentence means that the justifications for retaining it must be scrutinized afresh. One argument in favour of the mandatory life sentence is that it amounts to a symbolic indication of the unique heinousness of murder. It places the offender under the State's control, as it were, for the remainder of his or her life. This is often linked with a supposed denunciatory effect—the idea that the mandatory life sentence denounces murder as emphatically as possible—and with the supposed general deterrent effect of declaring that there is no way of avoiding the life-long effect of this sentence. It might also be argued that the mandatory life sentence makes a substantial contribution to public safety. None of these arguments is notably strong, let alone conclusive. The mandatory penalty does indeed serve to mark out murder from other crimes, but whether the definition of murder is sufficiently refined to capture the worst killings, and only the worst killings, remains to be discussed. As we shall see in section 7.3(c), it is sufficient for murder if D killed without intent to kill but with intent to cause serious harm, and the lesser intent is merely a mitigating factor from the various starting points in Sch 21 of the 2003 Act. Whether the life sentence is regarded as a sufficient denunciation depends on the public's perception of what life imprisonment means: if it is widely believed that it results in an average of about ten years’ imprisonment, the effect will be somewhat blunted, even if the belief is untrue. The same applies to the general deterrent argument: its effectiveness depends on whether the penalty for murder affects the calculations of potential killers at all, and, if it does, whether the prospect of life imprisonment influences them more than the alternative of a long, fixed-term sentence. As for public protection, this depends on executive decisions with regard to release; it raises the question whether it is necessary for public protection to keep most ‘lifers’ in for so long.25 It is sometimes claimed that murderers should be treated differently because they are particularly dangerous: anyone who chooses to kill once can choose to kill again. But this is an over-generalization that takes little account of the situational variations of murder cases. Moreover, the argument will seem less persuasive when we have discussed cases of manslaughter by reason of diminished responsibility: where a murder is reduced to manslaughter, the judge has a wide sentencing discretion and (p. 243) may, according to the facts of the case, select a determinate prison sentence, a hospital order, or life imprisonment. There is no evidence that those who kill and are convicted of manslaughter by reason of diminished responsibility are less dangerous than those convicted of murder, and yet the judge has sentencing discretion in one case and not in the other. Considerations of this kind led the House of Lords Committee on Murder and Life Imprisonment to recommend the abolition of the mandatory sentence for murder.26 A committee chaired by the former Lord Chief Justice, Lord Lane, reached the same conclusion in 1993.27 Both committees favoured judicial sentencing discretion to mark the relative heinousness of the murder, subject to review on appeal.28 A discretionary sentence of life imprisonment would still be available for those cases in which it was thought appropriate. Such a reform could bring improvements in natural justice without loss of public protection, but successive governments have been reluctant to contemplate the abrogation of the mandatory sentence for murder, and the latest reform proposals are premised on the retention of the mandatory penalty.29 (b) The structure of homicide law 30 Page 6 of 69

Homicide The structure of the law of homicide varies across jurisdictions,30 and recent proposals for reforming English law will be discussed below. It must be said that the current structure of the English law of homicide is rather strange. Although formally there are two offences—murder and manslaughter—the latter includes two distinct varieties: ‘voluntary’ manslaughter (killings which would be murder but for the existence of defined extenuating circ*mstances); and ‘involuntary’ manslaughter (killings that are in fact the product of voluntary conduct, but for which there is no need to prove any awareness of the risk of death being caused, but for which there is nonetheless thought to be sufficient fault to justify liability for a killing). The arguments therefore tend to focus on three borderline questions: What is the minimum fault required for conviction of murder? In what circ*mstances should murder be reduced to manslaughter? What is the minimum fault required for a conviction of manslaughter? (c) Requirements for murder In English criminal law, satisfying the fault requirement for murder involves the prosecution in proof that D possessed one of two states of mind: either an intent to kill, or (p. 244) an intent to cause grievous bodily harm. What do these requirements mean? Do they extend the definition of murder too far, or are they too narrow? An intent to kill may be regarded as the most obvious and least controversial form of fault element for murder. In part, though, that judgment hinges on the meaning given to ‘intent’ in the criminal law. The meaning of intent has been the subject of a number of House of Lords decisions,31 and yet the definition is still not absolutely clear. A broad definition would be that a person intends to kill if it is his or her aim to kill by the act or omission charged, in the sense that he or she would regard it as a failure if V was not killed by the act or omission (although the courts do not themselves use this so-called ‘test of failure’, which is merely a helpful explanatory tool). In practice, the ‘golden rule’ is the first to be applied—that intention should be left without description or definition in most cases, and the full definition should be reserved for cases where D claims that his purpose was something other than to cause injury. This is because the full definition includes a further possibility, in addition to proof of aim or purpose of killing on D's part. This is proof that D foresaw that V's death was virtually certain to follow from his or her act or omission, whether or not V's death was aimed at (i.e. whether or not D would have regarded V's survival of the incidence as a ‘failure’ on D's part to achieve his or her goal). If the jury is sure that D acted with this state of mind, they are entitled to infer that D intended to kill (we will look further at this issue). To this extension of the meaning of intent must also be added that, as was mentioned above, it is sufficient for the prosecution to show that, in killing V, D intended to cause grievous (serious) harm to V, and hence it will be enough for the prosecution to show that D foresaw grievous bodily harm as virtually certain to occur. So, in a case where D has killed V, in the absence of any justification or excuse: 1. If D intended to kill: murder; 2. If D intended to cause grievous (serious) bodily harm: murder; 3. If D foresaw death or grievous bodily harm as virtually certain to occur: murder, if the jury infers that D intended to kill or cause grievous bodily harm. A fairly typical set of facts is provided by Nedrick (1986),32 where D had a grudge against a woman and had threatened to ‘burn her out’. One night he went to her house, poured paraffin

Page 7 of 69

Homicide through the letter-box and onto the front door, and set it alight. One of the woman's children died in the ensuing fire. When asked why he did it, D replied: ‘Just to wake her up and frighten her’. A defence of this kind, a claim that the purpose was only to frighten and not to cause harm, requires the full definition of intention (i.e. including the reference to foresight of death or grievous bodily harm as a virtual certainty) to be put to the jury. The question is: granted that D's aim was to frighten, did he nonetheless realize that it was virtually certain that his act would cause death (p. 245) or grievous bodily harm to someone? The jury should answer this, as in all criminal cases, by drawing inferences from the evidence in the case and from the surrounding circ*mstances. As was pointed out in Chapter 5.5(b), the decision of the House of Lords in Woollin33 leaves some leeway in the application of the test by holding that, where the jury concludes that D foresaw that death or grievous bodily harm was virtually certain to ensue, it is ‘entitled to find’ that D had the intention necessary for murder. The test remains a permissive principle of evidence rather than a rule of substantive law, although the Court of Appeal has accepted that, once there is an appreciation of virtual certainty of death, ‘there is very little to choose between a rule of evidence and one of substantive law’.34 However, the test is so formulated in order to leave a degree of indeterminacy,35 and this could allow juries to make broader moral or social judgments when deciding whether the fault element for murder is fulfilled in a case where death (or grievous bodily harm) was known to be virtually certain.36 What about the alternative element in the definition, an intent to cause grievous bodily harm? This has considerable practical importance, since this is all that the prosecution has to prove in order to obtain a verdict of guilty of murder. It must be shown that the defendant intended (which, again, includes both acting in order to cause the result and knowledge of practical certainty) to cause ‘really serious injury’ to someone, although the use of the word ‘really’ is not required in all cases.37 The House of Lords confirmed this rule in Cunningham (1981): 38 D struck his victim on the head a number of times with a chair, causing injuries from which the victim died a week later. D maintained throughout that he had not intended to kill, but there was evidence from which the jury could infer—and did infer—that he intended to cause grievous bodily harm. The House of Lords upheld D's conviction for murder: an intent to cause really serious injury is sufficient for murder, without any proof that the defendant even contemplated the possibility that death would result. Does the ‘grievous bodily harm’ rule extend the definition of murder too far? If the point of distinguishing murder from manslaughter is to mark out the most heinous group of killings for the extra stigma of a murder conviction, it can be argued that the ‘grievous bodily harm’ rule draws the line too low. The rule departs from the principle of correspondence (see Chapter 5.4(a)), namely that the fault element in a crime should relate to the consequences prohibited by that crime. By allowing an (p. 246) intent to cause grievous bodily harm to suffice for a murder conviction, the law is turning its most serious offence into a constructive crime. Is there any justification for ‘constructing’ a murder conviction out of this lesser intent? One argument is that there is no significant moral difference between someone who chooses to cause really serious injury and someone who sets out to kill. No one can predict whether a serious injury will result in death—that may depend on the victim's physique, on the speed of an ambulance, on the distance from the hospital, and on a range of other medical and individual matters unrelated to D's culpability. If one person chooses to cause serious injury to another, he or she has already crossed one of the ultimate moral thresholds and has shown a sufficiently

Page 8 of 69

Homicide wanton disregard for life as to warrant the label ‘murder’ if death results. The counterarguments, which would uphold the principle of correspondence, are that breach of that principle is unnecessary when the amplitude of the crime of manslaughter lies beneath murder, and also that the definition of grievous bodily harm includes a number of injuries which are most unlikely to put the victim's life at risk. In the leading case of Cunningham Lord EdmundDavies (dissenting) gave the example of breaking someone's arm: that is a really serious injury, but one which is unlikely to endanger the victim's life.39 So in practice the ‘grievous bodily harm’ rule goes beyond the point at which the arguments of its supporters still carry weight. In its charging standards, the Crown Prosecution Service gives the following account of when it is appropriate to charge D with the offence of wounding with intent to do grievous (serious) bodily harm, an account that will also serve as a guide to cases in which, if V dies, it will be appropriate to charge murder: • injury resulting in permanent disability, loss of sensory function or visible disfigurement; • broken or displaced limbs or bones, including fractured skull, compound fractures, broken cheek bone, jaw, ribs, etc; injuries which cause substantial loss of blood, usually necessitating a transfusion or result in lengthy treatment or incapacity; • serious psychiatric injury. As with assault occasioning actual bodily harm, appropriate expert evidence is essential to prove the injury.40 It must be recognized that many other legal systems also have a definition of murder that goes beyond an intent to kill.41 What other approaches might be taken? The fault element for many serious offences is intent or recklessness: why should this not suffice for murder? One question is whether all killings in which the defendant is aware of a risk of death are sufficiently serious to warrant the term ‘murder’. An answer sometimes given is that they are not, because a driver who overtakes on a bend, knowingly (p. 247) taking the risk that there is no vehicle travelling in the opposite direction, should not be labelled a murderer if a collision and death happen to ensue.42 This example assumes that sympathy for motorists will overwhelm any tendency to logical analysis: the question is whether motorists are ever justified in knowingly taking risks with other people's lives. Yet if the example is modified a little, so that the overtaking is on a country road at night and the risk is known to be slight, it becomes questionable whether the causing of death in these circ*mstances should be labelled in the same way as, say, an intentional killing by a hired assassin. This is not to suggest that motorists, in particular, should be treated differently. The point is rather that, even though knowingly taking risks with other people's lives is usually unjustifiable, taking a slight risk is less serious than intentionally causing death. In discussing the boundaries of murder, we are concerned with classification, not exculpation. To classify all reckless killings as murder might be too broad, but the point remains that some reckless killings may be thought no less heinous than intentional killings. Can a satisfactory line be drawn here? One approach would be to draw the line by reference to the degree of probability. Murder would be committed in those situations where D caused death by an act or omission which he knew had death as the probable or highly probable result. A version of this test of foresight of high probability is used in several other European countries; 43 it was introduced into English law by the decision in Hyam v DPP (1975),44 but abandoned in Moloney (1985)45 on grounds of uncertainty. A related approach, applicable to certain terrorist situations, would be to maintain that someone who intends to create a risk of death or serious

Page 9 of 69

Homicide injury endorses those consequences to the extent that, if they occur, they can fairly be said to be intended.46 A second approach is to frame the law in such a way as to make it clear that the court should make a moral judgment of the gravity of the defendant's conduct. Section 210.2 of the Model Penal Code includes within murder those reckless killings which manifest ‘extreme indifference to the value of human life’.47 Scots law treats as murder killings with ‘wicked recklessness’, a phrase which directs courts to evaluate the circ*mstances of the killing.48 Both the Model Penal Code test and the Scots test may be reduced to circularity, however, for when one asks how extreme or how wicked the recklessness should be, the only possible answer is: ‘wicked or extreme enough to justify the stigma of a murder conviction’. Admittedly, the Model Penal Code does contain a list (p. 248) of circ*mstances which may amount to extreme indifference, which assists the courts and increases the predictability of verdicts in a way that Scots law does not. Having said that, under both approaches there is no precise way of describing those non-intentional killings which are as heinous as intentional killings. The advocates of this approach argue that the law of murder has such significance that the principle of maximum certainty should yield to the ability of courts to apply the label in ways more sensitive to moral/social evaluations of conduct. Opponents argue that the principle of maximum certainty is needed here specifically to reduce the risk of verdicts based on discriminatory or irrelevant factors, such as distaste for the defendant's background, allegiance, or other activities, especially if the mandatory life sentence is at issue.49 A third, more precise formulation now favoured by the Law Commission is that a killing should be classified as murder in those situations where there is an intention to cause serious injury coupled with awareness of the risk of death.50 Neither an intention to cause serious injury nor recklessness as to death should be sufficient on its own, but together they could operate so as to restrict one another and perhaps to produce a test which both satisfies the criterion of certainty and marks out some heinous but non-intended killings. A fourth approach, adopted by English law until 1957 and still in force in many American jurisdictions, is some form of felony-murder rule: anyone who kills during the course of an inherently dangerous felony should be convicted of murder.51 Thus stated, there is no reference to the defendant's intention or awareness of the risks: the fact that D has chosen to commit rape, robbery, or another serious offence, and has caused death thereby, is held to constitute sufficient moral grounds for placing the killing in the highest category. Plainly, this is a form of constructive criminal liability: the murder conviction is constructed out of the ingredients of a lesser offence. Presumably the justification is that D has already crossed a high moral/social threshold in choosing to commit such a serious offence, and should therefore be held liable for whatever consequences ensue, however accidental they may be.52 The objections would be reduced if awareness of the risk of death was also required: in other words, if the test were the commission of a serious offence of violence plus recklessness as to death. The effect of that test would be to pick out those reckless killings which occurred when D (p. 249) had already manifested substantial moral and legal culpability, and to classify them as murder. Four alternative approaches have been described, and others could be added. The point is that the traditional concept of intention does not, of itself, appear to be sufficiently well focused to mark out those killings which are the most heinous. The law must resort to some kind of

Page 10 of 69

Homicide moral and social evaluation of conduct if it is to identify and separate out the gravest killings. Some would defend the GBH rule on this basis as a form of ‘rough justice,’ and that argument could be extended to some of the cases in the fourth category above, as William Wilson has proposed.53 At the other end of the spectrum were the Law Commission's provisional proposals, which proposed to deal with the issue by means of a distinction between first degree murder and second degree murder. The mandatory life sentence would remain for first degree murder, the definition of which would be refined by confining it to cases where there is an intent to kill.54 Second degree murder would then include cases where D is proved to have killed while intending to do serious harm, defined more tightly than in existing law,55 and also cases where D is proved to have killed with reckless indifference as to causing death. Second degree murder would carry a maximum sentence of life imprisonment, together with the label ‘murder’, and is an attempt to allow some ‘moral elbow-room’ in the definition of murder outside the mandatory penalty.56 However, in its final report the Law Commission sought a compromise that enlarges first degree murder beyond intention to kill and yet does not encounter the objections made against the GBH rule. In effect, the Commission adopts the third approach above, arguing that first degree murder should extend beyond an intent to kill to those cases where there is an intent to cause serious injury coupled with an awareness of a serious risk of causing death.57 The Law Commission's view is that cases involving both these elements are morally equivalent to cases of intent to kill, or at least are closer to those cases than to cases placed in the other offence of murder in the second degree. (p. 250) What cases would fall within murder in the second degree? The Law Commission identified two types of case—where D kills with an intention to do serious injury (those not accompanied by an awareness of the risk of death and therefore not within murder in the first degree), and where D ‘intended to cause injury or fear or risk of injury aware that his or her conduct involved a serious risk of causing death’. The latter category is designed to capture bad cases of reckless killing and to sweep them into an offence with the label murder (in the second degree). One issue with this provision is whether the breadth of the concepts of ‘injury’ and ‘serious risk’ enables the proposal to distinguish fairly between these cases and others that fall into manslaughter. Another issue is whether the introduction of the provision would place too many choices between closely related mental states that the jury must juggle, in deciding whether D is guilty of first degree, second degree murder, or manslaughter. There is a considerable likelihood that, were the provision introduced, there would be more disagreements between individual jurors as to which offence category (if any) it fell into, and hence more cases in which the jury failed to agree on a verdict. This would mean that a fresh trial would have to be ordered, with all the witnesses having to given their evidence again, with no guarantee that a fresh jury would be more likely to agree on a verdict.58 In any event, the Government indicated that it was not minded to pursue the first degree/second degree distinction, and would be focusing on reform of the partial defences to murder.59

7.4 Defining murder: the exclusionary question Even in a legal system which had the narrowest of definitions of murder—say, premeditated intention to kill—there would still be an argument that some cases which fulfil that criterion should have their labels reduced from murder to manslaughter because of extenuating

Page 11 of 69

Homicide circ*mstances. Just as the discussion of the inclusionary aspect of the definition of murder travelled beyond the concepts of intent and recklessness, so the discussion of the exclusionary aspect (i.e. which killings fulfilling the definition should be classified as manslaughter rather than murder?) must consider the circ*mstances in which the killing took place and other matters bearing on the culpability of the killer. (a) The mandatory penalty The existence of the mandatory penalty has significant effects on the shape of the substantive law of homicide. One argument is that the main reason for allowing such matters (p. 251) as loss of self-control to reduce murder to manslaughter is to avoid the mandatory penalty for murder: if the mandatory penalty were abolished, it would be sufficient to take account of loss of self-control when sentencing for murder. However, this argument neglects the symbolic function of the labels applied by the law and by courts to criminal conduct. Surely it is possible that a jury might decline to convict of murder a person who intentionally killed following a loss of self-control, even though they knew that the judge could give a lenient sentence, because they wished to signify the reduction in the defendant's culpability by using the less stigmatic label of manslaughter. Since there are two offences—and particularly in jurisdictions where there are three or more grades of homicide—surely it is right and proper to use the lesser offence to mark significant differences in culpability. This may be seen as an application of the principle of fair labelling. When a jury takes the decision between the two grades of homicide, this may also assist the judge in sentencing, and help the public to understand the sentence imposed.60 A second major effect of the mandatory penalty for murder derives from the long minimum terms (plus the detention for public protection and then the licence for life) imposed for murder, as compared with the considerably shorter sentences for manslaughter upon loss of self-control or by reason of diminished responsibility. Thus the sentencing guidelines for manslaughter upon loss of self-control indicate a starting point of twelve years’ imprisonment where minimal provocation led to a loss of self-control is low, and starting points of eight and three years for more serious provocation leading to a loss of self-control.61 The difference between the highest starting point for manslaughter following a loss of self-control—twelve years (release on licence after six years)—and the lowest starting point for murder—fifteen years (which means fifteen years at least before release on licence)—is so considerable that, in practice, ‘there is the greatest of pressure to distort the concepts of provocation and diminished responsibility to accommodate deserving or hard cases. This pressure will continue as long as each case of murder carries the mandatory life sentence’.62 (b) Manslaughter following a loss of self-control Killings are generally thought to be less heinous when they are the result of grave provocation, or of a fear of violence stemming from words or conduct on V's part. Before important legislation in 2009, only provocation (not a fear of violence) had historically been accepted as a ground for reducing to manslaughter a killing which would otherwise be murder.63 From time to time, there continue to be cases where the provocation (p. 252) is so gross and so strong that a court imposes a very short prison sentence or even a suspended sentence for the manslaughter—typically, cases where a wife, son, or daughter kills a persistently bullying husband or father. In cases of this type (amongst others), though, the motivation may not be provocation so much as a fear of continuing violence, but where—for one reason or another— the complete defence of self-defence or action in prevention of crime cannot succeed. The

Page 12 of 69

Homicide Coroners and Justice Act 2009 sought to address this gap in the law. The 2009 Act put a fear of serious violence on a par with provocation, as a basis for reducing the offence from murder to manslaughter, so long as—putting on one side a number of restrictions and qualifications— in either case (or where both are pleaded together) D acted during a loss of self-control. The issues here are whether either or both of these ‘partial’ defences (defences to murder that reduce the crime to manslaughter) have any place in the modern law, or contrariwise, whether there is a case for extending them.64 The 2009 Act made highly significant changes to the law governing the partial defences. However, it is helpful to start with a brief explanation of the old structure, built from a mixture of common law and legislation, that was replaced by the 2009 Act: to a considerable extent, the 2009 Act still relies on that old structure. Before the coming into force of the 2009 Act, the doctrine of provocation (as it was commonly known) had two main elements. These emerge in s. 3 of the Homicide Act 1957, which the new law replaces:

Where on a charge of murder there is evidence on which the jury can find that the person charged was provoked (whether by things done or by things said or by both together) to lose his self-control, the question whether the provocation was enough to make a reasonable man do as he did shall be left to be determined by the jury; and in determining that question the jury shall take into account everything both done and said according to the effect which, in their opinion, it would have on a reasonable man.

This was not intended to be a complete statement of the law on provocation, but it settled the form of the main requirements. First, there had to be evidence that D was provoked to lose self-control and kill. Then the jury had to decide whether the provocation in question was enough to make a ‘reasonable man’ who had lost self-control do as D did. Importantly, D did not bear the burden of proving any of this him or herself. In a murder trial, when evidence of provocation was given or emerged in the course of the trial, the burden of showing beyond reasonable doubt that the requirements of s. 3 were not satisfied lay upon the prosecution.65 If the prosecution failed to discharge that burden to that standard, D was entitled to an acquittal on the murder charge on the grounds of provocation, and would be convicted of manslaughter instead. In practice, then, in any case where evidence of provocation became relevant, the prosecution would focus (p. 253) on one, or both, of two things. First, the prosecution could seek to convince the jury that D never lost control, however grave the provocation might have been. Secondly, the prosecution could seek to convince the jury that, even if D did lose selfcontrol, the provocation was not of such a grave kind that it might have moved even a reasonable person to lose self-control and kill with the fault element for murder. (i) The Subjective Requirement and its Replacement: The first requirement of the pre-2009 qualified defence of provocation was predominantly subjective—evidence that D was provoked to lose self-control and kill. Without this, there would be no way of excluding planned revenge killings as a response to provocation, and the argument is that they should be excluded from the defence. A person who coolly plans a murder as a response to an affront or a wrong is defying the law in so doing, and (barring the applicability of some other defence) there can be no excuse for that. By contrast, the killer

Page 13 of 69

Homicide provoked to lose self-control is, for the duration of the loss of self-control, not fully master of his or her mind, and so—in theory—is not deliberately defying the law in the same way. As Richard Holton and Stephen Shute argue, the paradigm case is where D normally had sufficient self-control to suppress violent inclinations, but the provocation aroused those inclinations and undermined D's controls.66 In the old case of Duffy,67 Devlin J (before he became Lord Devlin) expressed this idea in the following famous passage:

[C]irc*mstances which induce a desire for revenge are are inconsistent with provocation, since the conscious formulation of a desire for revenge means that a person has had time to think, to reflect, and that would negative a sudden temporary loss of self-control, which is of the essence of provocation.

In what ways has this requirement for a loss of self-control been altered by the reforms effected by the 2009 Act? The wording of s. 3 of the Homicide Act 1957 (the residual elements of the common law having been abolished) has now been replaced by ss. 54 and 55 of the 2009 Act. These sections introduce a partial defence renamed ‘loss of control’. The defence includes a new version of the old ‘provocation’ defence (the second limb of the defence), but also includes a new defence focused on a loss of self-control stemming from a fear of serious violence at V's hands (the first limb of the defence). In relation to the subjective requirement, the new law retains the requirement that D's acts or omissions in doing or being party to a killing resulted from D's loss of self-control (s. 54(1) of 2009 Act).68 The 2009 Act also re-states in a new form two ancient common law doctrines that served to restrict the defence in important ways. The first, just mentioned, is that the defence of loss of self-control will not apply where D acts on, ‘a considered desire for (p. 254) revenge’ (s. 54(4)). The second is that the defence will not apply if D's loss of self-control stemmed from something that D him or herself incited to be done or said in order to provide an excuse to use violence (s. 55(6)(a) and (b)).69

More significantly, the 2009 Act makes the following changes to the old law: 1. Section 54(6) says that the defence must only be left for the jury to consider where, ‘sufficient evidence is adduced to raise an issue with respect to the defence … [namely where] … evidence is adduced on which, in the opinion of the trial judge, a jury, properly directed, could reasonably conclude that the defence might apply’. 2. Section 54(2) says that the loss of self-control need no longer be ‘sudden’.

Let us turn first to consideration of this second change. In the passage cited earlier, Devlin J referred to the need at common law for a ‘sudden temporary’ loss of self-control. This supposed requirement at common law was ever-afterwards controversial. To begin with, it is unclear what is added by stipulating that a loss of control must be ‘temporary’. A loss of selfcontrol cannot indefinitely sustain itself or be sustained for very long, even if, having subsided, it then periodically overtakes D as he or she reflects from time to time on some provocation or Page 14 of 69

Homicide on some threat of violence from V. Secondly, it was never clearly the case at common law in any event that a loss of self-control had to be ‘sudden’ (i.e. immediately following a provocation, rather than happening at a slightly later point as D reflected on something said or done earlier).70 In many cases prior to 2009, this supposed requirement was ignored by both judges and juries.71 Critics of the suddenness requirement also claimed that it slanted the defence in favour of male defendants and prejudiced it against female defendants, in so far as the latter, unlike the former, tend not to lose self-control instantly but to react in a more ‘slow burning’ way.72 Whatever one's view about this, though, there was always a certain amount of tension at common law, and now in the legislation, between the removal of the supposed requirement of suddenness, and the exclusion by s. 54(4) from the scope of the defence of killings prompted by a ‘considered desire for revenge’. To begin with, although this is not acknowledged by the legislation, a provoked killing, even following a loss of self-control, normally has a vengeful motivation: the ‘desire for retaliatory suffering’ as ancient Greek philosopher Aristotle called it. So, on the one hand, it is perfectly consistent with pleading the defence that D did not (p. 255) react instantaneously and brooded on what was said or done, perhaps even forearming him or herself in readiness to confront V.73 That is the effect of s. 54(2) even though, ironically, evidence of delay (such as the detour to the garage to fill up the car in Baillie74 )—and even more so, evidence of preplanning—can be powerful evidence that D's supposed loss self-control at the time of the killing was not genuine or never occurred. On the other hand, by virtue of s. 54(4), in cases in which D acted from a ‘considered’ desire for revenge the defence is simply inapplicable. The combined effect of these provisions seems to be that, even if D does take time to brood on vengeful thoughts and even if D to some extent has prepared for a retaliatory attack on V, so long as D (a) had lost self-control at the time of the killing and (b) was not at that time motivated by any preconceived decision to exact vengeance that he or she may have come to previously, the defence may apply. That position is intelligible, legally and morally, but is based on fine distinctions that it may be very difficult to draw, not least in circ*mstances where it is likely that D's account of his or her own thoughts, feelings, and actions will be the only or main source of evidence for those thoughts, feelings, and actions. The Law Commission had originally argued strongly against the retention of any loss of self-control requirement.75 In the Commission's view, there should instead simply be a negative test of whether D acted on a considered desire for revenge; if he or she did not, then the defence would be available in principle. However, it is not entirely clear that this approach would have resolved the difficulties in distinguishing deserving from undeserving cases in this area.76 It is one thing to exclude cases like Ibrams77 from the defence: not merely was there a gap of some five days between provocation and killing, but there was evidence of planning and premeditation. It is another thing to exclude defendants with slow-burning temperaments, who do not react straight away to an insult or wrong, but go away and then react after minutes or even hours of festering anger. However, the problem then is that allowing lapse of time between the provocation and the retaliation—as suggested in Ahluwahlia78 —not only helps women defendants but also broadens the timeframe for men, and may thus weaken the excusatory force that derives from acting in uncontrolled anger. (ii) The subjective requirement and its replacement: Turning to the first change, the opening words of the old s. 3 of the Homicide Act 1957 (‘Where on a charge of murder there is evidence on which the jury can find that the person charged

Page 15 of 69

Homicide was provoked.’) were construed by the courts in such a way that almost any evidence of a provoked loss of self-control would make provocation an issue in the trial, and the prosecution would come under an obligation to show that the elements of the defence were not made out, as described above. What was required, stated Lord Steyn in Acott (1997),79 is ‘some evidence of a specific act or words of provocation resulting in a loss of self-control’, whereas ‘a loss of self-control caused by fear, panic, sheer bad temper or circ*mstances (p. 256) (e.g. a slow down of traffic due to snow) would not be enough’. However, what may properly be defined as ‘provocation’ in this context proved controversial. In Doughty80 the crying of a 17-day-old child was held to be sufficient to satisfy the requirement for a ‘provoked’ loss of self-control, even though a child's crying is scarcely even voluntary conduct, let alone conduct that is intended to or known to be likely to provoke. Doughty did not, of course, decide that a baby's crying was to be regarded as such that a reasonable person might have done as D did in that case. The point the Court of Appeal made was that even a baby's persistent crying was ‘evidence … that the person charged was provoked’ such that the defence became an issue at trial and the prosecution had to convince the jury that its elements were not satisfied. It was a waste of the court's time, and of counsels’ time, to have to address defences when, as in Doughty, they had such little chance of success. What is more, albeit exceptionally, some juries were acquitting Ds of murder, following a direction from the judge on the provocation issue, in cases where the provocation was so trivial that, as a matter of justice, it ought never to have been put before the jury. An example is Naylor.81 D picked up a prostitute (V) in his car, and then refused to pay as agreed for the services he had received. When V remonstrated with him he strangled her with such force that he broke bones in her neck. As there was ‘evidence … that the person charged was provoked’, the defence of provocation was put to the jury, which (surprisingly) acquitted him of murder and convicted of manslaughter only. So far as the second the second limb of the defence is concerned,82 it may be argued that s. 54(6) seeks to snuff out the possibility that cases such as Doughty and Naylor might end in manslaughter verdicts on the grounds of loss of self-control. It does this by providing that the loss of self-control defence is not to be put to the jury unless the judge is of the opinion that a jury, having been given proper directions, might reasonably conclude that the defence applied. This gives the trial judge an important and broad responsibility and discretion to exercise, in the role of ‘gatekeeper’, to the loss of self-control defence. Were cases with facts similar to those in Naylor and Doughty to recur, on any view it is hard to see how they would pass the judicial ‘gatekeeper’, given the height of the new hurdle (to be considered shortly) that D must surmount if he or she is to make loss of self-control an issue in the case in relation to the second limb of the defence. So far as the first limb of the defence (addressed below) is concerned—a fear of serious violence from V—s. 54(6) is less dramatic in its effect in this respect. It is still meant to deny the defence to Ds who raise wholly implausible claims to have lost self-control due to a fear of serious violence at V's hands, but as we will now see, s. 54(6) is a less powerful tool in that regard in relation to the first limb of the defence than in relation to the second limb. (p. 257) In relation to the first limb of the defence, s. 55(4)(b) is at the heart of the 2009 Act's replacement for the subjective requirement in what were formerly provocation cases. It requires that a loss of self-control as a result of something said or done (or both together) cause D ‘to have a justifiable sense of being seriously wronged’. These are the words designed to replace the old notion of a ‘provoked’ loss of self-control under the 1957 Act, and

Page 16 of 69

Homicide are referred to in the legislation as being part of a ‘qualfying trigger’ bringing the defence into play (s. 55(2)).83 It seems immediately apparent how the wording of s. 55(4)(b) will be likely to exclude cases such as Doughty from the scope of the loss of control defence, at the point when the judge has to decide whether a properly directed jury might reasonably conclude that the defence may apply. This is because s. 55(4)(b) avoids the language of ‘provocation’, by employing instead the notion that D must have had a justifiable sense of being seriously wronged. In this way, the law introduces the idea that, for the purposes of the second limb of the defence, D must have had good cause to feel that he or she was the victim of some kind of serious injustice, insult, or other glaring instance of denigration or derogatory conduct at the hands of another: nothing short of that will, assuming it sparks a loss of self-control, be sufficient to bring the defence into play (other things being equal). A crying baby is, quite simply, not capable of behaving in such a way towards another person, however frustrated or enraged that person may be by the crying. The way that the law achieves this, though, is by introducing an objective (judgmental) element into what was formerly an almost purely factual, subjective question concerning whether D was provoked to lose self-control. The objective, judgmental element is represented by the requirement that D's sense of being seriously wronged by some piece of conduct must be a ‘justifiable’ sense of being seriously wronged.84 It is worth noting the contrast, in this respect, with the subjective requirement for the first limb of the defence (the other ‘qualifying trigger’ bringing the defence into play), which does not involve this heavily judgmental element. Section 55(3) requires no more than evidence of a ‘fear of serious violence from V against D or another identified person’. Such evidence will bring the loss of self-control defence into play, when this limb of the defence is relied on, whether or not the fear of serious violence was ‘justifiable’.85 This is likely significantly to complicate the judge's task in discharging the gatekeeper function, and in directing the jury, when D pleads (as he or she is entitled to do) both limbs of the defence together. (p. 258) (i) The objective requirement, and its modification: It is helpful to begin by focusing on what the 2009 Act puts in place of the old provocation defence (the second limb of the defence, under s. 55(4)), and in that regard starting with a brief discussion of the pre-2009 law. Under s. 3 of the Homicide Act 1957, the jury had to decide not only whether D had lost self-control at the time of the killing, but whether the provocation was such as might have made a reasonable man do as D did. The latter question is the objective requirement, and it is still part of the post-2009 law (in a more sophisticated form, as we will see). Its function is to ensure that not every homicidal loss of self-control reduces the offence from murder to manslaughter. It would, for example, be unacceptable if the defence served to reduce from murder to manslaughter an intentional killing involving the act of a possessive spouse angered by no more than his or her partner's decision to go out alone for an evening, or the act of someone who flies into rage when lawfully arrested for a crime he or she has committed. Only those provocative acts serious enough to unbalance the reactions of a person with reasonable self-control should suffice.86 However, the 1957 Act did not elaborate on the kinds of provocation that might form the basis of a successful plea.87 The authors of the 1957 legislation were content to leave that question to be decided by each jury on a case by case basis.88 Regrettably, though, the Appeal Courts could not resist the temptation to introduce legal complexity to the relatively simple provisions of the 1957 Act. They sought to spell out, case by case, the characteristics of the ‘reasonable person’ as a matter of law, and hence dictate

Page 17 of 69

Homicide to the jury what should, and should not, be taken into account in deciding whether the reasonable person might have done as D did in the circ*mstances. In so far as it affected his or her reaction, could the reasonable person still be reasonable when jealous, depressed, temperamental, or touchy on the subject of a disability or a previous criminal record? Could the ‘reasonable’ person possibly be someone who had a mental disorder that affected their capacity to maintain self-control? Judges needlessly set themselves the task of answering all these questions—and more of a similar kind. Naturally, when considering these questions, judges disagreed amongst themselves over the right answers, over the kinds of characteristics that could, and could not, be regarded as features of the reasonable person. This served only to push the law even further into obscurity.89 We will return to this issue shortly. (p. 259) A very significant effect of the 2009 Act is to bring to the fore, and make explicit, an objective requirement that was only implicit in the pre-2009 law: the requirement for the conduct that sparked D's loss of self-control to be an instance of very grave provocation. Section 55(4) says that, in addition to being something that gave D a justifiable sense of being seriously wronged (a requirement that we have already considered), the conduct that sparked D's loss of self-control will not amount to a qualifying trigger unless (s. 55(4)(a)) it, ‘constitutes circ*mstances of an extremely grave character’. It will still be, as it was under the law prior to the 2009 Act, a matter for the jury whether what was done or said that led D to lose self-control and kill constituted circ*mstances of an extremely grave character. However, it is arguable that s. 55(4)(a) was intended to raise the bar that D must surmount in order to bring the defence into play: the provocation must not merely be serious, but ‘extremely grave’. In the Parliamentary debate on the issue in 2008–9, Baroness Scotland, speaking for the government, gave as an example a hypothetical case in which a refugee living in the UK encountered a man responsible for rounding up members of his old village, locking them in a church, and then setting the church alight. The man laughs at the incident and describes in some detail what happened to the refugee's family killed in the fire. The refugee thereupon lost self-control and killed the man. Baroness Scotland went on the say:

We consider that the words and conduct limb of the partial defence needs to be included in this kind of extremely grave example, where the defendant would have a justifiable cause to feel seriously wronged. We remain of the view that the partial defence should succeed only in the gravest of circ*mstances.90

Baroness Scotland's example, coupled with this explanation, suggests two things. First, what is to count as an ‘exceptionally grave’ provocation must be judged in context. There can never be a list of exceptionally grave provocations detached from consideration of circ*mstantial and contextual issues such as who the provoker and the provoked person were, the relationship between them, and the manner in which the provocation was given. For example, if the hypothetical war criminal in the example given by Baroness Scotland had given his account of the events at the church to a young British person who was born long after the incident took place, and knew nothing of it, the gravity of the provocation would be reduced. It might still be a grave provocation, but arguably no longer an exceptionally grave provocation. So, as the Law Commission put it (considering its own version of the test), ‘The jury should be

Page 18 of 69

Homicide trusted to evaluate the relative grossness of provocation, in whatever form it comes, according to their own sense of justice in an individual case’.91 Under the 1957 Act, this was also the position, following the overruling in DPP v Camplin (1978)92 of the old case of Bedder v DPP (1954).93 In Bedder v DPP, D, who was sexually impotent, was taunted (p. 260) about his impotence and kicked in the groin by a prostitute with whom he had been attempting to have sexual intercourse, whereupon he lost self-control and killed her. The House of Lords held that the jury should consider the effect of these acts on a ‘reasonable’ man, without regard to the sexual impotence. That was a difficult rule to apply, because it is hard to see how the gravity of the provocation constituted by a taunt about impotence can be properly understood without reference to whether the person taunted is indeed impotent. The House of Lords in Bedder v DPP failed to consider whether it was possible to take into account D's impotence, in assessing the gravity of the provocation, whilst at the same time insisting that the provocation must be so grave that it might lead even a person with reasonable powers of self-control to lose control and kill.94 The effect of the decision in DPP v Camplin was to adopt the latter approach. Lord Diplock held that a court should consider the effect of the provocation on ‘a person having the power of self-control to be expected of an ordinary person of the sex and age of the accused, but in other respects sharing such of the accused's characteristics as they think would affect the gravity of the provocation to him’. We will consider the first part of this ruling in due course. So far as the second part of the the ruling is concerned—characteristics affecting the gravity of the provocation—following the decision in Camplin, increasingly few, if any, limits were set to the kinds of characteristics that may affect the gravity of the provocation. In Morhall (1996),95 D was a glue-sniffing addict who had been taunted about his glue-sniffing by the victim, whom D subsequently stabbed. The Court of Appeal held that in applying the objective test the jury should be directed not to take account of discreditable characteristics such as glue-sniffing (or paedophilia). The House of Lords disagreed, and held that a jury should be directed to take account of any matter relevant to an assessment of the strength of the provocation. As we will see, in broad terms, this is now the approach under the new law. Does this mean that there are no boundaries at all to what personal attributes may be taken into account in assessing the gravity of the provocation? What about the case of a racist who believes that it is gravely insulting for a non-white person to speak to a white man unless spoken to first?96 Lord Taylor in Morhall put the case of ‘a paedophile upbraided for molesting children’,97 which raises similar issues. The implication of the House of Lords decision in Morhall is that the judgment of such matters must be left to the jury without much guidance. It could be argued strongly that is unsatisfactory: there ought to be a normative element that excludes attitudes and reactions inconsistent with the law or inconsistent with the notion of a tolerant, pluralist society that upholds the right to respect for private life without discrimination (Arts. 8 and 14 of the Convention). However, the significance of that argument has sharply diminished in the light of the reforms effected by s. 55(4)(a) of the 2009 Act, and its requirement that (p. 261) provocation be of an ‘extremely grave’ character. Further, Baroness Scotland's second point in the passage cited earlier from her speech indicates that s. 55(4)(a) is meant to be a restatement of what was originally envisaged by the legislature when debating the 1957 Act, namely that commonly encountered provocations—even very annoying or wounding ones, such as persistently inconsiderate behaviour by a neighbour or the discovery that a spouse is in another relationship—should not be capable of forming the basis of a successful plea.98 In the years following the reform of the defence in 1957, it would

Page 19 of 69

Homicide be fair to say that courts had—whilst making the law ever more complex—also allowed juries to consider a provocation plea in a far wider range of cases, involving commonly encountered provocations, than the legislature had envisaged in 1957. To that extent, s. 55(4)(a) is a welcome development. It is against that background that we should consider s. 55(6)(c), which provides that, ‘In determining whether a loss of self-control had a qualifying trigger … the fact that a thing done or said constituted sexual infidelity is to be disregarded’. As these words indicate, s. 55(6)(c) concerns a (dis)qualifying trigger, but it is illuminating to discuss it more generally as part of the post-2009 modification of the objective requirements of the loss of self-control defence. In policy terms, the provision was intended to stop possessive men, in particular, pleading a partner's infidelity as a basis for reducing murder to manslaughter: something that it was open to them to do (with no guarantee of success, of course) under the old law.99 The provision is, though, so clearly and obviously incapable of doing that (except in rare cases), that it is probably better to regard its importance as lying not so much in the extent to which it narrows the scope of the law, as in its symbolism as a commitment to taking domestic violence more seriously. The case for regarding it as more important symbolically than normatively is expressed in Baroness Scotland's explanation of the section, worth citing at length:

Even accepting that a great deal has been done in recent years to address this problem, and that pleas of provocation on the basis of sexual infidelity generally do not succeed, it is still true that, under the current law, the defence can be raised and could technically succeed. We want to make it clear in the Bill that this can no longer be the case, and that it is unacceptable for a defendant who has killed an unfaithful partner to seek to blame the victim for what occurred. It is important to correct a misconception here. By doing this, we are not saying that people are not entitled to feel upset and angry at a partner's unfaithfulness: we are concerned here with a partial defence to murder and the circ*mstances in which it is appropriate to reduce liability for murder to that of the less serious offence of manslaughter. We are saying that killing in response to sexual infidelity is not a circ*mstance in which such a reduction can be justified.100

(p. 262) The case for saying that s. 55(6)(c) is of little normative significance was to an extent conceded by the government itself, when it stated that it is only sexual infidelity in itself that must be disregarded in a provocation case.101 If there is other evidence that constitutes a qualifying trigger, then the case may be considered by the jury notwithstanding the part played in D's reaction by sexual infidelity. This point was hammered home by the Lord Chief Justice in the Court of Appeal's decision in Clinton.102 Indeed the Court of Appeal went on to say that even evidence of sexual infidelity itself could be given, ‘where sexual infidelity is integral to and forms an essential part of the context in which to make a just evaluation whether a qualifying trigger properly falls within the ambit of subsections 55(3) and (4)’.103 The case for saying that s. 55(6)(c) will have only a negligible impact in cases where possessive men haved used violence against women (and against their sexual partners in particular) has been much commented on, not least in Clinton itself.104 To adapt an earlier example, if D loses control and kills V simply because V wishes to have a night out without D present, that evidence will not be ruled out as a basis for a plea of loss of self-control by s.

Page 20 of 69

Homicide 55(6)(c). Similarly, a plea of loss of self-control will not be ruled out by that section in any of the following examples: (a) D ‘stalks’ a celebrity with whom he is obsessed (although they have never met), losing control and killing her when he sees her having dinner with a man; (b) D loses control and kills V, his 16-year-old daughter, when V says that she intends to start dating; (c) D loses self-control and kills V, his partner, when V says that she will never have sexual intercourse with him again. In these examples, it is highly unlikely that the defence of loss of self-control would be put to the jury, or if put to the jury it is highly unlikely that the defence would succeed; but that is not because of the existence of s. 55(6)(c). It is because of the combined effect of s. 54(6)—that requires the judge to withdraw a case from the jury if a properly directed jury could not reasonably conclude that the defence might apply—and s. 55(4)(a), that requires the evidence to constitute circ*mstances of an extremely grave character. Were it not for the perceived great importance of the symbolic function of (p. 263) s. 55(6)(c), it is strongly arguable that sexual infidelity cases should have been left to be weeded out by the same combination of ss. 54(6) and 55(4)(a). (ii) The objective requirement and its modification: Section 55(3) creates a new basis for reducing murder to manslaughter, under the heading of loss of self-control. As we have seen, a qualifying trigger for the loss of self-control defence is that D's loss of self-control ‘was attributable to D's fear of serious violence from V against D or against another identified person’. This short statement of the first limb of the defence captures both the subjective element, ‘fear of violence’ (already considered), and an objective requirement that the violence feared must have been ‘serious’; s. 55(3) does not say as much, but it would be wholly inconsistent with the tenor of the provisions as a whole were it to be a matter solely for D whether the violence he feared was serious. It should not, for example, be possible for a gang member to say that he responded with lethal violence to a punch from a member of a rival gang, because he regarded violence offered by rival gang members to be much more serious than other kinds of violence. ‘Serious’ should probably be understood to mean, ‘constituting at least serious bodily harm’. However, the application of the first limb may not be all that straightforward in some circ*mstances. Consider the facts of the case of Uddin, that arose under the old pre-1957 common law.105 D was a Moslem who killed another Moslem (V) when V threw a pigskin shoe at him. At the trial, expert evidence of the religious significance of shoe throwing (and in particular, no doubt, of pigskin shoe throwing) was given to assist the jury in understanding the gravity of the provocation. That would almost certainly also happen were D in a similar case now to plead loss of self-control under the second limb, on the basis that this was exceptionally grave provocation that gave him a justifiable sense of having been seriously wronged. However, were D to claim that he intentionally killed V solely on the grounds that he faced ‘serious’ violence from V, how should that claim be treated? It is submitted that the claim should fail. The incident was serious only as a form of provocation, not as a form of violence (a far more objective question). However, as has already been pointed out, D would be entitled to plead the two limbs of the loss of self-control defence together, entailing a tricky task for the judge in directing the jury. That direction would be further complicated if, as indicated earlier, D was Page 21 of 69

Homicide mistaken about what V threw at him, thinking it was a pigskin shoe when in fact it was a plastic container of some kind. The underlying explanation for s. 55(3) is principally the difficulty that may arise, highlighted in the Law Commission's analysis of the provocation defence,106 when a woman kills an abusive husband or partner (V) without reacting suddenly in the face of a final provocation or threat from V. In such cases, it will often make more sense, in terms of the moral narrative, to describe her ultimate reaction in killing the husband as attributable to fear for her own or her children's safety rather than to anger (although (p. 264) motives may be understandably mixed).107 It will also make sense, in those terms, to explain her reaction as delayed because of (amongst other things) an inequality between her size and strength and his—that necessitates waiting until he is off guard to defend herself—rather than because of sheer malevolence or a purely vengeful motivation. Neither of these points was capable of accommodation under a defence of provocation tied, as it was under the old law, to the notion of a sudden and angry loss of self-control. Moreover, although solid evidence to prove this one way or another has been hard to come by, juries have been considered reluctant to apply the complete defence of self-defence in such circ*mstances, even though that defence is sensitive to considerations such as a disparity in size between those involved, such that the weaker person may (more justifiably) use a weapon or wait until the stronger person is off his guard.108 As we have seen, so far as the loss of self-control defence is concerned, the new law has dispensed with the suddenness requirement (s. 54(2)), but crucially, it retained the requirement for a loss of self-control (s. 54(1)). The government departed from the Law Commission's view: that the latter should be abolished, as it was doing as much as the former to make things difficult for a battered woman who had killed her abusive partner, having waited until he was off his guard. Although there is certainly truth in this, that does not mean that the government's retention of the loss of selfcontrol requirement was necessarily wrong. This first limb of the loss of self-control defence has an application in a number of contexts when what would otherwise be murder is the result of a fear of serious violence: such as when someone lashes out with lethal intent when attacked in a public house brawl, or when an armed police officer shoots a man dead without warning because the officer believes that the man may be threatening him. In both such cases, a jury may be reluctant to acquit completely on the grounds of self-defence, but has the option of reducing murder to manslaughter under the first limb of the defence. In such cases, it is much more understandable that the law would be reluctant even to see murder reduced to manslaughter, unless the killer could say, as part of the excusatory explanation for his or her reaction, that he or she had lost self-control.109 (iii) The objective requirement and its modification: Should D by some miracle surmount the hurdles to pleading loss of self-control just discussed, there is still one hurdle left to jump. That is the test set out in s. 54(1)(c), reflecting a requirement of the old common law, alluded to in the previous version of s. 3 of the 1957 Act. The test—formally, part of the qualifying trigger for the defence—is whether someone of D's sex and age, with a normal degree of tolerance and self-restraint, and in D's circ*mstances, (p. 265) might have reacted in the same or in a similar way. Added to this, in s. 54(3), is an elaboration of the meaning of D's ‘circ*mstances’:

Page 22 of 69

Homicide

[T]he reference to ‘the circ*mstances of D’ is a reference to all of D's circ*mstances other than those whose only relevance to D's conduct is that they bear on D's general capacity for tolerance or self-restraint.

It is unclear how much is really added by these two provisions, given that the loss of selfcontrol defence cannot come into play, in any event, unless one of the limbs of the defence is satisfied, and the disqualifying triggers (a ‘considered revenge’ motive; self-induced loss of self-control; reliance solely on sexual infidelity, and so forth) do not apply. If, for example, under the second limb of the defence, someone is found to have intentionally killed having lost control in the face of an exceptionally grave provocation that gave them a justifiable sense of being seriously wronged, is that not just another way of saying precisely that, in the circ*mstances, a normal person in D's position might indeed have acted in the same or in a similar way?110 In fact there may be some room for these provisions to have an impact. An example is where, at the time of the killing, D occupied a role that demanded of him or her greater self-restraint than an ordinary member of the public might be expected to show. A case in point might be where D, a police officer from an ethnic minority seeking to control a violent crowd, is subjected to continued pushing and shoving, coupled with continuous racist abuse. After an hour of this treatment, the officer loses self-control, knocks over a demonstrator and stamps on his head, killing him. In such a case, the first limb of the defence might be satisfied (exceptionally grave provocation giving D a justifiable sense of having been seriously wronged). However, a jury might nonetheless find that a person with a normal degree of tolerance and self-restraint who was a police officer on crowd-control duty (these being ‘the circ*mstances of D’) would not have reacted in the same or in a similar way.111 As indicated above, under the pre-2009 law, the higher courts took upon themselves the task of explaining to the jury the characteristics of the ‘reasonable man’ (as it was then termed), but continually disagreed amongst themselves over which characteristics were and which were not attributable to the ‘reasonable man’. Indeed, following the enactment of the Homicide Act 1957 the House of Lords or Privy Council considered some aspect of the provocation defence (and this is not a complete list) in 1968, 1973, 1978, 1996 (twice), 1997, 2001, and 2005, with the Court of Appeal considering it in many more cases. In few areas (with the exception of complicity, dealt with in Chapter 10) can Dickens’ characterization of Chancery lawyers have been more (p. 266) appropriate in the criminal law: ‘some score of members of the High Court … mistily engaged in one of the ten thousand stages of an endless cause, tripping one another up on slippery precedents, groping knee-deep in technicalities, running their goat-hair and horse-hair warded heads against walls of words…’.112 It would not be appropriate to go through those ‘slippery precedents’ now that s. 54(1)(c) replaces them, as (in broad terms) that provision aimed to set the law as described in Attorney-General for Jersey v Holley113 on a statutory footing. What s. 54(1)(c) seeks to ensure is that the jury does not take into account, in D's favour, some feature of his or her psychological make-up that makes him or her prone to explode into violent rage in circ*mstances where ordinary people would have kept their tempers in check, or would have responded in a less violent (or non-violent) way. For example, D may be someone for whom jealousy leads to extreme anger,114 who may suffer from Intermittent Explosive Disorder, or who may be suffering from a tumour affecting the brain such that his or her reactions to stress have become unpredictable or uncontrollable (or both).115 Whether or Page 23 of 69

Homicide not any of these factors are D's ‘fault’, they are not to be taken into account by the jury. This is because they affect the level of restraint that can be expected of D in general; they do not necessarily relate to the gravity of the provocation or to the seriousness of the violence feared on the occasion in question. Where D simply struggles to—or cannot—control him or herself in the way that a person of ordinary tolerance and self-restraint should do, and it is that—rather than the gravity of the provocation or a fear of serious violence—which explains his or her loss of self-control and lethal use of violence, the right plea is diminished responsibility, and not loss of self-control. This explanation of s. 54(1)(c) is something of an over-simplification. For example, even the person of ordinary tolerance and self-restraint may be affected by irritability or tiredness in such a way that they are more likely to ‘fly off the handle’. In such a case, D's irritability or tiredness may well form part of the ‘circ*mstances of D’ that are relevant to judging the possible reaction of a person of ordinary tolerance and self-restraint, for the purposes of s. 54(3). However, it is hard to see how that will help D much, if his or her violent reaction was not in response to an exceptionally grave provocation or a fear of serious violence (although it might help to explain, in an appropriate case, why D made a mistake about the provocation or threat offered). Section 54(1)(c) also creates its own exception to the general rule, by stipulating that D's ‘sex and age’ can affect the level of self-restraint and tolerance to be expected of D in the circ*mstances. Quite why, for example, D's sex is thought especially likely to affect D's level of tolerance (or, for that matter, her capacity for self-restraint) is something of a mystery. Even if men are, say, generally less racially or religiously tolerant than women,116 should a jury (p. 267) really be making allowances for this? Again, the issue is unlikely to arise, because the need to satisfy the requirements of the first and second limbs (a fear of—objectively—serious violence, or an extremely grave provocation giving S a justifiable sense of being seriously wronged) will tend to mean that reactions attributable to racial stereotyping or religious intolerance will fall at the first hurdle. (c) Manslaughter by reason of diminished responsibility117 General considerations: Diminished responsibility was formerly one of the most frequently used qualified defences to murder, but in recent years the numbers have fallen from eighty per year in the early 1990s to around twenty per year (22 in 2004, for example). Diminished responsibility was introduced into English law only in 1957, in response to long-standing dissatisfaction with the insanity defence. Insanity was, and still is, a complete defence to crime, as we saw in Chapter 5.2, but its confines are narrow, and on a murder charge a verdict of not guilty by reason of insanity requires the court to make a hospital order with restrictions.118 Diminished responsibility has a wider ambit, but its effect is merely to reduce murder to manslaughter. Moreover, by way of contrast with the defence of loss of self-control, the burden is on the accused to show (on the balance of probabilities) that he or she is suffering from diminished responsibility.119 The judge then has a discretion in sentencing, and in recent years about half of the cases have resulted in hospital orders without limit of time.120 The existence of the diminished responsibility defence is one of the reasons for insisting that the loss of self-control defence is insensitive to mental disorders affecting D's levels of tolerance and powers of self-restraint: D must be ‘normal’ in this respect (s. 54(1)(c)). For, if D is clinically abnormal in this respect (and it is that which explains D's reaction), he or she is free to plead diminished responsibility which, if successful, has the same effect as a plea of loss of self-control in reducing murder to manslaughter.

Page 24 of 69

Homicide (p. 268) The post-2009 law: The wording with which s. 2 of the Homicide Act 1957 introduced diminished responsibility was generally regarded as unsatisfactory and has now been replaced (by s. 52 of the 2009 Act) with the following provisions:

(1) A person (D) who kills or is a party to the killing of another is not to be convicted of murder if D was suffering from an abnormality of mental functioning which— (a) Arose from a recognized medical condition, (b) Substantially impaired D's ability to do one or more of the things mentioned in subsection (1A), and (c) Provides an explanation for D's acts or omissions in doing or being a party to the killing. (1A) Those things are— (a) To understand the nature of D's conduct; (b) To form a rational judgment; (c) To exercise self-control. (1B) For the purposes of subsection (1C), an abnormality of mental functioning provides an explanation for D's conduct if it causes, or is a significant contributory factor in causing, D to carry out that conduct.

The aim of the reforms was to bring greater clarity of definition to the terms employed to describe diminished responsibility, and to ensure that those terms were capable of adaptation to developing clinical diagnostic practices. Hence, the abnormality of mental functioning from which D suffers must arise from a ‘recognized medical condition’, where it is understood that, over time, what count as recognized medical conditions may change as knowledge about mental functioning advances.121 In practice, under the old law, where a diminished responsibility plea was backed by medical evidence, it was simply accepted by the prosecution—meaning that D would be convicted of manslaughter only, without the need for a full trial to go ahead—in 77 per cent of cases. Whether or not there is reason to think that this statistic will alter, in the light of the change of definition of diminished responsibility, and in the light of the new relationship it forms with the loss of self-control defence, is something to be considered when the ingredients of the defence have been analysed. The nature of the conditions: The notion of an ‘abnormality of mental functioning’ arising from a ‘recognized medical condition’ suggests that whether or not D suffers from such a condition is essentially an expert question. It would be inconsistent with the tenor of the legislation, and unfair to D, to leave members of the jury to decide whether or not D's condition meets the criteria if that matter is not in dispute between the experts. The emphasis in the new law is different to the path adopted by the old law, under which an abnormality of mind was said to be ‘a state or mind so different from (p. 269) that of ordinary human beings that the reasonable man would term it abnormal’.122 This was unsatisfactory, not least because it suggests that the state of mind in question must be something not experienced in any form by mentally normal people (schizophrenia, for example); whereas, many abnormalities of mental functioning are

Page 25 of 69

Homicide states of mind very familiar—at least in a moderate form—to ‘ordinary human beings’, such as pathological jealousy or intermittent explosive disorder.123 However, in cases where experts disagree over whether or not there is an abnormality of medical functioning, or over whether the abnormality stems from a recognized medical condition, the matter will have to go to the jury. Even if this issue is not in dispute, the jury should be told that they are free to decide whether or not D's condition, established on the basis of expert opinion, ‘substantially impaired D's ability to do one or more of the things mentioned in subsection (1A)’. It will also be for the jury to decide whether the abnormality of mental functioning was the cause, or a significant contributory factor in causing, D's condition, although they may be assisted on this issue by expert evidence.124 Again, these are matters on which D bears the burden of proof on the balance of probabilities.125 What kinds of conditions will meet the criteria? In Byrne,126 D strangled and then mutilated a woman after her death. Evidence was given that from an early age Byrne had suffered from extremely strong perverse desires that he found it all-but impossible to control, and it was such a desire that had overwhelmed him when he killed the woman. Such a case would fall within the scope of the defence, because it involves an abnormality of mental functioning substantially impairing D's ability to exercise control over violent impulses, an abnormality recognized as a medical condition. The well-respected Diagnostic and Statistical Manual of Mental Disorders (DSM-4) lists some 16 different kinds of recognized mental disorder, including: • Disorders Usually First Diagnosed in Infancy, Childhood, or Adolescence • Delirium, Dementia, and Amnestic and Other Cognitive Disorders • Mental Disorders Due to a General Medical Condition • Substance-Related Disorders • Schizophrenia and Other Psychotic Disorders • Mood Disorders • Anxiety Disorders (p. 270) • Somatoform Disorders • Dissociative Disorders • Sexual and Gender Identity Disorders • Eating Disorders • Sleep Disorders • Impulse-Control Disorders Not Elsewhere Classified • Adjustment Disorders • Personality Disorders • Other Conditions That May Be a Focus of Clinical Attention This is not a list of disorders drawn up for legal purposes, and so not all of these disorders, even if they affected D's conduct, would be capable of substantially impairing D's ability to (s. 52(1)(a)) understand the nature of his or her conduct, to form rational judgment, or to exercise self-control. Even so, the DSM gives an indication of the breadth of expert opinion concerning abnormalities of mental functioning. Certain to be included are the effects of alcohol 127 Page 26 of 69

Homicide dependency,127 of depressive illnesses resulting from, for example, long-term abuse at the hands of a violent partner,128 or from the stress of long-term care for a terminally ill relative.129 Picking up on the final example, some critics of the new law complained that the 2009 reforms would end the practice of ‘benign conspiracy’ (allegedly sometimes entered into by the prosecution and the defence, in agreeing not to contest the case) to allow those who had taken a premeditated and rational decision to kill a terminally ill relative (at the latter's request) to plead guilty to manslaughter only.130 That is a curious objection, since the very fact that the practice could only be sustained by a benign conspiracy shows that the practice was, in fact, inconsistent with the legal requirement for any abnormality of mental functioning arising from the stress of long-term care for the terminally ill person to have a medically recognized origin (to use the modern language of the 2009 law). Excuses for rationally perpetrated euthanasia, whether or not they are partial excuses, should be introduced after open democratic debate on their merits, and not introduced through a back door route created by lawyers manipulating defences that were intended as a humane way of treating only those with medically recognized abnormalities of mental functioning. The irony is that, as the Law Commission was at pains to point out in an almost wholly neglected part of its (p. 271) Consultation Paper in the subject,131 many of those who kill terminally ill relatives following years of stressful longterm care are indeed suffering from such an abnormality of mental functioning that both substantially impairs their judgment and control, and makes a significant contribution to their conduct in killing V; and that can be true whether or not V consented to be killed. We consider this issue further below. Where a mental disorder has been aggravated by the effects of voluntary intoxication, the pre-2009 law was that the judge should instruct the jury to answer the question, ‘Has the defendant satisfied you that, despite the drink, his mental abnormality substantially impaired his mental responsibility for his fatal acts?’132 This focuses on the ‘substantial impairment’ element of the partial defence, and the Court of Appeal took the same approach in Wood,133 where the underlying clinical condition was alcohol dependency syndrome and D had also drunk much alcohol. The jury should decide whether the clinical condition ‘substantially impaired’ D's responsibility, discounting any effects of alcohol consumed voluntarily. In effect, the jury are left to determine how much of D's drinking derived from his alcohol dependency and how much was ‘voluntary’. This inevitably involves a good deal of speculation by the jury on which the assistance that expert evidence can provide may be limited. In such cases, it is arguable that evidence of voluntary intoxication should not simply rule out a plea of diminished responsibility, even if the voluntary intoxication made some causal contribution to D's conduct in killing V. This approach is warranted, because the issue of voluntary intoxication is more complicated than when alcohol affects a mentally normal person. Research shows both that alcohol dependency or heavy drinking may generate psychiatric disorders and that, vice versa, those with psychiatric disorders often become alcohol dependent or heavy drinkers: in an effort, for example, to offset the unwanted and unpleasant effects of a disorder.134 The focus should be on whether or not the abnormality of mental functioning (which may include the lasting effects of excessive drinking over a long period) substantially impaired D's understanding, judgment, or control, and whether that made a significant contribution to D's conduct in killing V. That brings us to the first decision on diminished responsibility under the new law, in Dowds.135 In that case, D had, whilst heavily intoxicated, killed his partner V with a knife, inflicting some

Page 27 of 69

Homicide sixty stab wounds in the process. The evidence suggested that D periodically drank very heavily, but retained control over when he started drinking. In support of his plea of diminished responsibility, D nonetheless claimed that his acute intoxication at the time of the offence was a ‘recognized medical condition’ for the purposes of the new law, even though the intoxication was voluntary. D relied on a World Health Organization classification of acute intoxication as a medical condition (p. 272) (WHO ICD-10). D was convicted and his appeal was dismissed by the Court of Appeal. Hughes LJ said (para. 40) that, even when D is suffering from a recognized medical condition, that is only a necessary element to be satisfied if D is to raise diminished responsibility. The presence of a medically recognized condition will not in and of itself always be sufficient for that purpose. In support of that conclusion. Hughes LJ pointed to the fact that the Dictionary of Scientific Medicine (DSM) itself warns that there is what it refers to as an ‘imperfect fit’ between clinical diagnosis and legal concepts.136 By way of example, Hughes LJ highlights the fact that the DSM includes as possible basis for a clinical diagnosis ‘unhappiness’, ‘irritiability and anger’, and ‘paedophilia’, none of which could ever come to be regarded as ‘recognized medical conditions’ for the purposes of the defence of diminished responsibility (a point made earlier). In Hughes LJ's view, voluntary intoxication, however acute, should be placed in the same category. That is to say, even if it is a recognized medical condition, it is not a medical condition appropriate for recognition in law as capable of giving rise to an abnormality of mental functioning substantially impairing D's ability in the relevant respects. The Court of Appeal's approach, then, is to admit that acute voluntary intoxication can be a recognized medical condition, but to carve out some space for judicial discretion to rule that not all recognized medical conditions will suffice to bring the diminished responsibility defence into play. As we have just indicated, that approach is consistent with what the DSM itself says about the ‘imperfect fit’ between clinical and legal analysis. However, even if one puts aside the case of voluntary intoxication as a special case, the approach has the broader potential both to give rise to significant difficulties for trial judges in deciding whether or not to put the defence to the jury in contested cases, and to give rise to tensions between legal and clinical analysis that it was one purpose of the s. 52 reform to reduce. A more radical view of the reform would be that (a) any recognized medical condition is in principle capable of founding a defence of diminished responsibility, (b) that it is D's task to prove (on the balance of probabilities) that the condition gave rise to an abnormality of mental functioning that had the effects s. 52 requires it to have, and (c) that it is the prosecution's task to show the contrary beyond a reasonable doubt. On this radical view, if such an approach means more contested cases in which the prosecution seeks to rebut defence evidence that a recognized medical condition met the conditions of s. 52, then so be it. However, not the least of the problems with the radical view is that it places in jeopardy the prospect of establishing a satisfactory relationship between the defences of diminished responsibility and of loss of self-control.137 Pleading loss of self-control and diminished responsibility together: There is nothing to stop D pleading both loss of self-control and diminished responsibility together. One cynical reason for D to take this course is the hope that if, say, the loss of self-control (p. 273) that led to the killing was not triggered by very grave provocation, and D was suffering only from a mild form of mental disorder, the jury will nonetheless put these two pieces of evidence together and bring in a manslaughter verdict even though, strictly speaking, that would be a case of adding 2 and 2 together in order to reach the required ‘5’. This is because if neither the evidence relating to the loss of self-control defence itself, nor the evidence for the diminished

Page 28 of 69

Homicide responsibility defence itself, is independently sufficient to satisfy the criteria respectively for each of those defences, the verdict should be murder. There should be no mixing and matching of half-fulfilled criteria on each side to make a ‘whole’ defence to murder. To that end, a judge should instruct a jury considering both defences to ignore evidence relevant only to one of them, when considering the other.138 So, suppose that D claims that he reacted with murderous rage to a mildly offensive remark made by V not only because he (D) lost selfcontrol, but also due to the influence on him of a medically recognized abnormality of mental functioning. D calls evidence from a psychiatrist who has examined him to say that D suffers from a mental abnormality that means that he sometimes finds it impossible to control his temper. The judge should tell the jury that this evidence is not relevant to the plea of loss of self-control (other than to show, if need be, that D had in fact lost self-control at the time of the killing). In particular, the psychiatrist's evidence cannot affect (a) whether D ‘justifiably’ has a sense of being seriously wronged, (b) whether the loss of self-control was attributable to circ*mstances of an extremely grave character, or (c) whether a person of D's sex and age with a normal degree of tolerance and self-restraint might have reacted as D did. Consequently, we may expect to see a shift in popularity towards diminished responsibility and away from old-style provocation in its newly restricted form as the loss of self-control defence. This is because—financial resources permitting—it is likely to prove easier to find a medical practitioner somewhere in the UK (or, if need be, in the world) willing to give evidence that D suffers from an abnormality of mind stemming from a ‘medically recognized condition’ that influenced his or her conduct, than it is to fulfil the requirements of the narrow loss of selfcontrol defence.139 To that end, a ‘recognized’ medical condition is not necessarily one that a substantial body of medical practitioners would accept as such: it can include, for example, a condition discovered by an individual practitioner who has published the results of his or her own medical research in a peer-reviewed Journal.140 All this suggests that, in the future, we may see more contested trials than we have been used to on the issue of diminished responsibility. At a superficial level, the two defences of loss of self-control and diminished responsibility can seem as if they are two sides of a single coin, the former excusing normal (p. 274) people, and the latter abnormal people, who—for the reasons specified in each defence—could not be expected to contain an urge to kill. Moreover, they are often raised in similar circ*mstances: where, for example, a man has killed his partner following an argument or alleged infidelity.141 However, the appearance is something of an illusion. In each case, the reasons for excusing differ so greatly that, ethically speaking, these defences have little or nothing in common. In cases of loss of self-control, the basis for excuse is that no more could reasonably have been expected of D—any ordinary person might have reacted in that way—although, given the nature of the motive for which D acted (retaliation; revenge-taking), there is not enough in this to warrant a full excuse and hence an acquittal. By contrast, ‘reasonableness’ plays no part at all in the assessment of D's homicidal conduct in diminished responsibility cases. In effect, D is saying, ‘I was only to an extent morally responsible for my actions, and to the extent that I was not morally responsible, I should not be judged by the standards of ordinary people at all’. Those acting under the influence of provocation or a fear of violence are ‘morally active’ (in evaluative control), and hence full morally accountable—if not fully to blame—for their conduct, whereas those acting under diminished responsibility are morally more ‘passive’ (less capable of evaluative control), and less justly held fully to account for their conduct.142 Accordingly, the defence of diminished responsibility may have just as clear an application to

Page 29 of 69

Homicide cases of, for example, premeditated, sexually motivated, or mass killing as it does to killing in the heat of the moment. Even a homicidal war crime could be reduced to manslaughter if there was evidence that the perpetrator's responsibility was diminished, whereas no amount of provocation could ever excuse—let alone justify—such an act. The absence of theoretical connection between the two partial defences to murder is an illustration of a broader problem in the law of homicide. Currently, there is little more than a jumble of instances in which murder can be reduced to manslaughter, if one also adds in the curious case of part-completed suicide pact (s. 4 of the Homicide Act 1957), all with their own rationales for existence, none of which is wholly convincing in its own terms.143 Were the mandatory sentence for murder to be abolished, it would be possible—with the assistance of statutory guidelines if need be—to rid the law of these anomalous excuses and regard the issues (provocation, fear, mental disorder, etc.) as matters of sentence mitigation. Not the least of the benefits of such a scheme would be that it would cease to matter which precise defence pigeonhole D's actions fitted into, with all the current implications that has for the admissibility of evidence (ir)relevant to the defence in question. It would be possible to take account of both diminished responsibility and fear of violence or provocation, depending on the degree to which they had a just bearing on the appropriate sentence. Ironically, under current sentencing guidelines, when D is convicted of murder because the (p. 275) specific defences based on these factors have been rejected by the jury, it is precisely this approach that influences what sentence D receives.144 (d) Killing in pursuance of a suicide pact Section 4 of the Homicide Act 1957 provides that a person who kills another in pursuance of a suicide pact is guilty of manslaughter, not murder. A suicide pact exists where two or more people, each having a settled intention of dying, reach an agreement which has as its object the death of both or all.145 Some suicide pacts may be regarded as the highest expression of individual autonomy, by means of a mutual exercise of the individuals’ rights of selfdetermination, but the Law Commission reported concern that the majority of cases involve men taking decisions to end the lives of the spouses or partners for whom they are caring.146 The Criminal Law Revision Committee recommended that killings in pursuance of a suicide pact should be a separate offence, on the ground that the stigma and maximum penalty for manslaughter are inappropriate in these cases.147 This is a more contentious stance than might at first glance be supposed. In principle, for example, s. 4 would cover the following cases: (A) a cult leader secures the agreement of all his 300 followers that they will die together with him in a barn that he will set alight. He sets the fire with that intention in mind but, finding it hot, changes his mind and makes his escape while the 300 followers die. (B) Two terrorists are on the run from the police. They return to their apartment and agree that one will shoot the other dead before shooting himself dead, to avoid the capture and questioning of either. One shoots the other dead, but then changes his mind about killing himself and decides to fight to the death with the pursuing police instead. When the police arrive later he gives up on that intention as well, and allows himself to be arrested without resistance. An offence dealing with suicide pacts must be up the task of fair labelling in such cases, and it 148 Page 30 of 69

Homicide is strongly arguable that the right label is (at least) manslaughter.148 Consequently, the Law Commission recommended no change to the law until there is a wider review of ‘consensual’ and ‘mercy’ killings. However, it abandoned its earlier proposal that s. 4 should be abolished and all cases dealt with under the partial defence of diminished responsibility if they were to fall outside the scope of murder. This was because there might be cases in which a suicide pact was the product of a rational (p. 276) decision by mentally normal people, and such cases required consideration alongside similar ‘ending of life’ decisions, rather than being dealt with in isolation: 149

Y has terminal cancer and is determined to bring about her own death one way or another. X, Y's husband, does not wish to live on if Y is dead. So, X and Y decide to end both their lives by jumping off a high cliff on to the rocks below. They hold hands at the top of the cliff, count to three in unison, shout, ‘Go!’, and then jump off. X survives the fall but Y dies.

Were it not for s. 4, X would in theory be guilty of the murder of Y because, with the intention that they should both die, his acts of assistance and encouragement play a causal role in bringing about Y's death, although the law almost always treats such a role in another's death as the specific offence of doing an act capable of encouraging or assisting suicide rather than as murder.150 Had Y survived instead of X, it would be Y who was guilty of murder or of encouraging/assisting suicide, for identical reasons. Yet, it is perhaps not all that easy to see why X's decision to end both their lives with Y's agreement should make such a fundamental difference to the legal outcome. Suppose Y had asked X to push her off the cliff, and X had agreed; but secretly, X always harboured the intention to jump off the cliff immediately afterwards, a decision he did not communicate to Y for fear of causing her even greater distress. In such circ*mstances, if X jumps off the cliff immediately afterwards and survives, he is nonetheless guilty of the murder of Y and has no defence. Indeed, given that he pushed her off the cliff, there is no case for treating what X did as merely encouraging or assisting suicide. This kind of example illustrates that it would make more sense for the law to consider all forms of ‘ending of life’ decisions together, when considering how far to extend the scope of defences to murder, rather than permitting only an exception for suicide pact cases that is morally so arbitrary in its range and application. For similar reasons, the relationship between murder and s. 4 manslaughter needs to be considered in the light of the specific offence of encouraging or assisting suicide. For example, why is the latter not manslaughter, if V dies having been influenced by D's encouragement or assistance, with killing in the course of a suicide pact being treated as a separate specific offence instead? (e) Doing an act capable of assisting or encouraging suicide or attempted suicide Suicide and attempted suicide ceased to be criminal when the Suicide Act 1961 became law, in recognition of the right to self-determination. However, as indicated at the end of the last section, it is an offence contrary to s. 2(1) of the Suicide Act 1961 if :

(a) Does an act capable of encouraging or assisting the suicide or attempted

Page 31 of 69

Homicide suicide of another person, and (p. 277) (b) D's act was intended to encourage or assist suicide or an attempt at suicide. (1A) The person referred to in subsection (1)(a) need not be a specific person (or class of persons) known to, or identified by, D. (1B) D may commit an offence under this section whether or not a suicide, or an attempt at suicide, occurs. (1C) An offence under this section is triable on indictment and a person convicted of such an offence is liable to imprisonment for a term not exceeding 14 years.151

This version of the offence is a version modified by s. 59 of the Coroners and Justice Act 2009, although the changes were intended simply to modernize the language of s. 2(1),152 and clarify the law.153 It may appear paradoxical to legalize an activity (suicide), but at the same time make it a serious offence to encourage or assist that lawful act. Further, many of the cases governed by s. 2 involve compassionate assistance, where many may think there is little case for prosecuting. However, the rationale for the offence is illustrated by McShane (1977),154 where a woman was convicted of an attempt to counsel her mother's suicide by encouraging her repeatedly to take an overdose, and it was shown that the mother's death would greatly alleviate the defendant's financial problems. There remains a need to protect the vulnerable from persuasion on such a crucial matter as the ending of life. Just because suicide is not a criminal offence does not mean that it has ceased to involve an unjustified harm through the elimination of a human life. That being so, there can be a legitimate case for criminalizing the encouragement or assistance of suicide, and McShane is an illustration of the kind of case where it seems justified to employ the deterrent and retributive powers of the criminal law.155 The consent of the Director of Public Prosecutions (DPP) is required before a prosecution under s. 2 is commenced, and that has brought under scrutiny the prosecution policy adopted by the DPP in relation to such prosecutions. A number of difficulties have arisen in relation to scope of the offence. For example, suppose someone (V) attempts to kill themselves through an overdose. The emergency services arrive whilst V is still conscious and in need of urgent treatment, but V refuses treatment under any circ*mstances. It might seem as if to attempt forcibly to treat V would be to assault V, and the courts have confirmed that this is indeed so.156 However, doctors and the emergency services are under a duty to take positive action to help those in their care, and those needing immediate treatment for whom they have assumed (p. 278) responsibility, meaning that an omission can count as an ‘act’ for legal purposes. So, do doctors and the emergency services do ‘an act capable of … assisting’ V to commit suicide by allowing V to die unaided? It would seem that they do not.157 So, in seeking provide an escape route from the dilemma in which care workers may find themselves, in such situations, the law errs on the side of the right to self-determination and avoids paternalism (seeking to improve someone's prospects even when they are themselves fully capable of deciding where their own best interests lie). It is probably right to do so, in so far as to permit the forcible administering of treatment on a sane and mature adult against their will is likely to be an inhuman and degrading process for that person, and will hence involve a breach of Art. 3 of the European Convention on Human Rights. In some countries, the law has gone further, and does not look unfavourably even on doctors who take positive steps to assist suicide.158

Page 32 of 69

Homicide That still leaves unclear what policy should be adopted towards the ordinary citizen who seeks to assist,159 in a variety of circ*mstances, someone who wishes to die. All prosecutions must pass tests not only of evidential sufficiency, but also of public interest, and it is in relation to the public interest factor in prosecuting s. 2(1) cases that devising prosecution policy is most difficult and controversial. On the one hand, taking it for granted that the criminalization of assisting suicide is the right legal policy in general,160 there would be a strong public interest in prosecuting someone who set up a commercial operation in the UK to help people end their lives.161 On the other hand, if a wife simply agrees with her terminally ill husband's request not to give him any more of the medicine he needs to stay alive for the few more days that the medicine can realistically give him, or buys tickets for them both to visit a country where he can be assisted to die, there may—depending on the precise facts—be little or no public interest in a prosecution. In R (on the Application of Purdy) v DPP,162 the applicant sought to compel the DPP to reveal or devise a published policy for prosecuting under s. 2, so that she could make a properly informed decision on whether to ask her husband to assist her to travel abroad to die. The House of Lords agreed with the argument that Art. 8—the right to respect for private and family life—was engaged by the prohibition on s. 2. Even so, Art. 8(2) provides that public authorities may legitimately interfere with respect for private and family life, but only—amongst other things—if the interference is ‘in (p. 279) accordance with the law’.163 The prohibition in s. 2 itself is, of course, in accordance with the law. However, the DPP is himself also a ‘public authority’ seeking to interfere or impinge on a matter of private and family life. In that regard, the House of Lords said that the absence of a published prosecution policy in relation to s. 2, meant that there was a risk that prosecutions would not be ‘in accordance with the law’, if it was not possible for individuals like Purdy to make highly important personal decisions against a sufficiently clear legal policy background. So, the DPP was obliged to publish a code setting out the factors to be consided in any prosecution decision in relation to s. 2.164 This argument involves a sleight of hand, because there was never any suggestion that a prosecution would be undertaken other than ‘in accordance with the law’.165 It is true that forcing the DPP to draw up a comprehensive prosecution policy in relation to s. 2 would assist citizens the better to assess the risk that they might be prosecuted if they took certain steps. However, that argument might equally apply to cases in which someone is contemplating committing, assisting or encouraging, or conspiring to commit euthanasia (murder, in English law).166 Moreover, the issuing of official guidelines in relation to the prosecution of a particular offence (s. 2, in this instance) opens up the prospect of secondary litigation testing whether, in any allegedly borderline case, the guidance had been correctly followed. That could lead to prosecutions being stayed for lengthy periods whilst the secondary litigation is conducted to answer the point, or to convictions being quashed long after the event, on procedural fairness grounds that would have no application to crimes closely analogous to assisting suicide (where there was no official guidance), such as euthanasia. Perhaps the answer is that all serious crimes should come with comprehensive official guidance on prosecution policy, whose implementation in any individual case can be challenged by the individual prosecuted or convicted (or perhaps also by a third party, such as Dignity in Dying). That would take the law into a new era in which public law principles of judicial review were potentially as important as criminal law principles to the outcome of cases, something that English law has witnessed in private litigation against the State.167 That might not be such a bad thing, although it would draw the courts into making authoritative decisions not only on the scope of the substantive

Page 33 of 69

Homicide law (as they have always done), but now also on the boundaries within which a decision based on a pre-announced policy will be regarded as reasonable, and as having taken into account relevant—and disregarded irrelevant—considerations. Inevitably, there (p. 280) would be tensions between judicial interventions of this kind, and the broader position so far as the accountability of the DPP on policy matters is concerned: the DPP is answerable politically, on policy questions, to the Attorney General (a member of Parliament). Amongst the factors listed as being relevant to a prosecution decision are: In favour of prosecution: 1. The would-be suicide is under 18 years of age; 2. The suspect not being wholly motivated by compassion; 3. The suspect being unknown personally to the would-be suicide, and encouraging or assisting through, for example, the provision of specific information through website material; 4. The suspect being paid by the would-be suicide, or providing assistance to more than one person; 5. The suspect acting in his or her capacity as a health professional; Against prosecution: 1. The suspect was wholly motivated by compassion; 2. The assistance or encouragement was only minor; 3. The suspect had sought to dissuade the would-be suicide, and participated only reluctantly; 4. The suspect reported the would-be suicide's death and co-operated fully with the police. These factors span across excusatory, justificatory, and after-the-fact mitigating features of a case, but in a distinctive way. In D's favour, the factors are sensitive to excusatory features such as a compassionate motive or only reluctant participation, and to ‘good citizenship’ features such as reporting the death to, and co-operating with, the authorities. On the other hand, against D, the factors are sensitive to whether someone acts under a false cloak of justification (in his or her capacity as a professional, as a commercial provider, or as a selfappointed actor ‘in the public interest’). The logic of this approach is that it is precisely in those cases where someone sets themselves up as giving help or guidance in opposition to the law that a prosecution is desirable to underline the authority of the law itself.168 By contrast, where someone acts from compassion, or only reluctantly (given the law's demands), whilst there is still a case for prosecution—after all, even reluctant murderers or thieves are still murderers or thieves—the public interest in the prosecution may be less compelling. (p. 281) (f) ‘Mercy’ Killing This concept has no special legal significance in English criminal law. Where there is a clear case of mercy killing by a doctor, he or she is likely to avoid prosecution or to benefit from Devlin J's concession to good motive in the Adams case.169 Under the old law, the usual response to a ‘genuine’ case involving a non-professional defendant is that ‘legal and medical consciences are stretched to bring about a verdict of manslaughter by diminished 170 Page 34 of 69

Homicide responsibility’.170 In Mackay's study of 157 cases in which diminished responsibility was raised, it seems that six were probably cases of mercy killing.171 Practitioners seemed to accept that worthy cases of mercy killing should be eased into diminished responsibility, but this informal approach provides the defendant with no legal basis for a defence—he or she is truly at the mercy of the psychiatrists, the prosecutor, and the judge.172 The decision of the House of Lords in the Purdy case (discussed in section (d)) to force the DPP to issue guidelines on prosecution policy in relation to the offence of assisting suicide seems equally significant in relation to all these ‘manoeuvres’ to escape the full force of the law of murder. Time will tell whether the judges see sufficient force in the analogies to prompt the issuing of yet more official guidance on prosecution policy in this difficult area. The Criminal Law Revision Committee (CLRC) regarded the bending of the law as unsatisfactory, and we have seen—in relation to diminished responsibility—that it may no longer be possible to engage in it in quite the same way under the post-2009 law.173 The CLRC tentatively proposed a new offence of mercy killing where a person, out of compassion, unlawfully kills another who is, or is believed by him to be, permanently helpless or in great pain. The proposal attracted strong opposition, some arguing that it might withdraw legal protection from the weak and vulnerable, others arguing that the fundamental ethical problems could not be satisfactorily resolved by legal definition. In respect of doctors, some flexibility is achieved through such distinctions as that between bringing about a patient's death through omission (which may be lawful) and (p. 282) bringing it about by a positive act (which is not),174 and between intending to cause death and intending to relieve pain while knowingly accelerating death.175 A ‘blind eye’ may be turned to the practices of some doctors. But doctors cannot be assured that a ‘blind eye’ will be turned, and relatives and friends may be exposed to the strict law. In terms of protection for the vulnerable,176 the chief difference between the present system and the CLRC's proposed offence is that the latter had a maximum penalty of two years’ imprisonment, whereas life imprisonment is now available even where there is a conviction for manslaughter on grounds of diminished responsibility. The Law Commission concluded that a separate review and consultation on ‘consensual’ and ‘mercy’ killings would be necessary before well-founded proposals could be made. (g) Conclusion: the murder—manslaughter boundary In this section we have been examining the partial defences which mark out cases where, despite the presence of the mental element for murder, culpability is thought to be sufficiently reduced to warrant a reduction in the class of offence. Our discussion has taken a broad view of partial defences, commenting also on some possible defences which are not (yet) accepted in English law. Various reasons have been advanced for recognizing partial defences to murder. Some regard the mandatory penalty for murder as the chief, even the sole, reason for these doctrines. However, whilst highly significant, the mandatory penalty is not the only argument for partial defences. A key issue is the proper legal classification of an offence which contains strong exculpatory features: should killings influenced by diminished responsibility really be treated as cases of murder, whether or not the judge has flexibility on sentence? The Law Commission found strong support for the murder–manslaughter distinction, and for the view that partial defences should have the effect of reducing the crime from the most heinous to something lesser.177 Worldwide, the label ‘murder’ (or its equivalent), and the stigma thought to accompany it, are very widely reserved for the most heinous group of killings,178 with lesser forms of homicide classified differently where the culpability is significantly 179 Page 35 of 69

Homicide lower.179 The Law Commission recommended a three-tier structure for the law of homicide which includes two degrees of murder (first degree murder, second degree murder) and (p. 283) manslaughter.180 Assuming—until we examine the matter in detail—that the scope of manslaughter corresponds roughly with the existing law of involuntary manslaughter, the Law Commission's recommendations would have three grades of conviction and therefore three thresholds to consider. To distinguish between first and second degree murder on the basis of the existence of an intent to kill or an intent to cause serious injury with an awareness of a serious risk of causing death may be acceptable, but it would focus much argument on the boundaries of ‘serious injury’ and of ‘serious risk’. Similarly, the proposal that cases of intention to cause injury or fear or risk of injury with an awareness of a serious risk of causing death should qualify for conviction of second degree murder may be acceptable, but it will lead to much argument over the boundaries of ‘injury’ and of ‘serious risk’.181 Under the Law Commission's revised structure, the partial defences would reduce first degree murder to second degree murder, rather than from murder to manslaughter as under the current structure. The government chose not to explore further this three-tier structure for the law of homicide. Arguments in favour of some of the partial defences have been set out above. Since it is possible that more than one defence might be raised in each murder case, sometimes in combination with a defence of lack of intent, or self-defence, a system of criminal law which offers a number of partial defences to murder risks undue complication and confusion in contested cases.182 There is a risk—albeit small—that a jury may be divided over which partial defence applies, if any, on particular facts, leading to a situation in which no clear verdict can be reached. On the other hand, one merit of separate partial defences is that they focus the evidence and the legal argument, giving the jury (in contested cases) an opportunity to assess the particular arguments for partial exculpation. The challenges posed by having separate but closely allied defences, such as loss of control and diminished responsibility, could be overcome—as they are in some US states—by merging the two into a partial defence of ‘extreme emotional disturbance’, following the lead of the Model Penal Code.183 This might encompass all those partial defences with an element of excuse in them. A provoked loss of self-control could fall within this new doctrine, as could diminished responsibility, although a general defence of mental disorder may be a better way of labelling and dealing with cases of clinical mental disorder. Cases now treated as infanticide often involve extreme emotional disturbance, as do mercy killings (although not those performed by professionals or commercial providers), suicide pacts, and cases of duress. One advantage of this (p. 284) amalgamation might be that there would be less potential for the jury to become confused, and yet the jury would still be empowered to reduce murder to manslaughter in appropriate cases. One disadvantage of the change might be that the more precise moral distinctions currently incorporated within the law would become submerged within the sentencing discretion, where the signposts are less clear and the arguments less structured. A variant on this approach, a version of which is to be found under Art. 345 of the French Penal Code, is to give the jury in any murder case the right to find ‘extenuating circ*mstances’, which would mean—were it to be introduced in English law—that the judge would be free to pass an appropriate sentence rather than having to impose the mandatory life penalty. A sophisticated version of this was put forward as an amendment in the House of

Page 36 of 69

Homicide Lords during the debate that preceded the 2009 Act, but it was not adopted. It read:

56: Before Clause 46, insert the following new Clause—“Murder: extenuating circ*mstances (1) In a trial for murder the trial judge may in the course of his summing up direct the jury that if they are satisfied that the defendant is guilty of murder, but are of the opinion that there were extenuating circ*mstances, they may on returning their verdict add a rider to that effect. (2) The judge may not give such a direction unless there is evidence on which a reasonable jury might so find. (3) Where the jury has so found, the judge shall not be obliged to pass a sentence of life imprisonment but may pass such other sentence as he considers appropriate having regard to any extenuating circ*mstances found by the jury. (4) If the judge passes a sentence other than a sentence of life imprisonment, he shall be obliged to state his reasons. (5) If it appears to the Attorney General that the sentence so passed is unduly lenient he may refer it to the Court of Appeal under section 36 of the Criminal Justice Act 1988 (c. 33) (reviews of sentencing).184

This would be a highly sophisticated means of dealing with excusatory claims in homicide cases, although it is less clear that it would operate in a fair way when it is mental disorder that is the basis of the plea of extenuating circ*mstances. Juries are prone to find that horrific killings deserve to be treated as murder irrespective of the severity of any mental disorder that led the killer to do the acts in question. That is not fair to the killer whatever we may think of his or her acts, and does not ensure that he is treated more ‘severely’ in any event: prisoners with mental disorders may simply be transferred to hospitals for treatment. (p. 285) It is arguable that both partial defences and ‘extenuating circ*mstances’ provisions only exist because legislatures in many countries have always wished to underpin the supposed uniqueness of murder, as a crime, with a unique (meaning uniquely severe) sentencing system to match, requiring a complex system for ensuring that at least some cases can fall down into a lesser (and less severely treated) category of homicide. Another alternative, then, would be to preserve the distinction between murder and manslaughter—but making it depend solely on the nature of the offender's fault element—and underpin this with separate but overlapping sentencing regimes for each offence. So, murder might involve a starting point of (say) fifteen years’ imprisonment, going up to life imprisonment, whereas manslaughter might involve complete discretion over sentences up to a maximum of twenty years’ imprisonment.185

7.5 ‘Involuntary manslaughter’ The category of killings which has come to be known as involuntary manslaughter has nothing to do with involuntariness, properly so-called. These are not cases where the accused has caused death while in an involuntary state.186 They are cases where death has been caused

Page 37 of 69

Homicide with insufficient fault to justify labelling it as murder, but with sufficient fault for a manslaughter verdict. The word ‘involuntary’ is therefore used merely to distinguish these killings from ones which have the necessary intent for murder but which are reduced to manslaughter by one of the doctrines just considered, such as loss of self-control or diminished responsibility. The legal debate in involuntary manslaughter is over the lower threshold of homicide liability— where to draw the line between manslaughter and cases of death by misfortune which are not serious enough to deserve a manslaughter conviction. There are now three forms of involuntary manslaughter—two at common law (manslaughter by unlawful and dangerous act, manslaughter by gross negligence) and a statutory addition, corporate manslaughter. These three offences raise some deep issues of general principle. For example, manslaughter by unlawful and dangerous act is a species of constructive liability, which was criticized in Chapters 3.6(r) and 5.4(b). In none of the cases of involuntary manslaughter is death or grievous bodily harm intended. Can constructive liability be justified by reference to the magnitude of the harm resulting, i.e. death? Or would it be fairer to convict the wrongdoer of a lesser offence, thus ignoring the chance result of death occurring? Consider this example:

D is arguing with V over whether V took D's place in the queue. V insults D, and D punches V in the stomach. V falls over, hits his head, and dies.

(p. 286) In law, so long as D is shown to have caused V's death, D commits manslaughter by unlawful and dangerous act in this example. The risk that V might die from D's punch is not relevant to the question of whether D is guilty of the crime. It is sufficient, as we shall see, that D commits the offence of assault (an act causing or posing a danger of ‘some harm’), and that the assault causes death.187 The other two forms of involuntary manslaughter are based on liability for negligence, albeit gross negligence: as we saw in Chapter 5.5(f), this is regarded as insufficient for liability for most serious offences.188 Is it right that liability for the second most heinous crime in English law, which carries a maximum penalty of life imprisonment, should be satisfied by this relatively low grade of fault?189 These questions will be discussed in more detail once the elements of the offence have been outlined. (a) Manslaughter by unlawful and dangerous act This species of involuntary manslaughter is based upon constructive liability. In broad terms, the law constructs liability out of the lesser crime which D was committing, and which happened to cause death. In fact, the courts have progressively narrowed this form of manslaughter over the last century or so: 190 there was a time when the mere commission of a tort or civil wrong sufficed as the ‘unlawful act’, and when there was no additional requirement of ‘dangerousness’ to be satisfied. What the prosecution must now prove is that D was committing a crime (not being a crime of negligence or a crime of omission), that in committing this crime he caused V's death, and that what he did when committing this crime was objectively dangerous. Let us examine each of these requirements in turn. First, D must have been committing a crime. In many cases the crime which constitutes the

Page 38 of 69

Homicide ‘unlawful act’ will be a battery or an assault occasioning actual bodily harm, arising from a push, a punch, or a kick. The prosecution must establish that all the elements of the crime relied upon as the unlawful act were present: to this extent a mental element is required for this form of manslaughter, and so in assault or battery this would be the mental elements of intent or recklessness. This point had been overlooked by the trial judge in Lamb (1967),191 in assuming that an assault had taken place when two young men were joking with a revolver, without noting that fear was neither caused nor intended to be caused.192 The point appears to have been overlooked by all the courts, including the House of Lords, in Newbury and Jones (1977): 193 two boys (p. 287) caused the death of a railwayman by pushing a paving stone off a railway bridge on to a train below, but none of the judges identified the precise crime which constituted the ‘unlawful act’. No doubt the boys’ act was a crime (a form of criminal damage, or a specific railway offence), and so this case does not call into question the proposition that all the elements of the crime relied upon must be established. In Dhaliwal (2006)194 D had subjected V to a long course of abuse, including physical assaults. One evening he abused her again, striking her once, and she subsequently committed suicide. The Court of Appeal quashed the manslaughter conviction, on the ground that V's severe emotional trauma caused by D's long course of abuse was not a recognized psychiatric condition and therefore not ‘bodily harm’. Thus the abuse was not an unlawful act. The ‘unlawful act’ requirement also means that D must not have any defence to the crime relied upon. Intoxication would supply a defence to a crime of specific intent in this context,195 but in most cases the prosecution will rely on a crime of basic intent or recklessness and therefore intoxication would be no defence.196 In a case where the prosecution relies on assault or battery as the ‘unlawful act’ and D claims that it was a justifiable use of force, the court must be satisfied beyond reasonable doubt that the force was not justified if it is to proceed to a manslaughter conviction.197 There appear to be two types of crime which will not suffice as the unlawful act—crimes of negligence and crimes of omission. The reasons for excluding crimes of negligence were stated in Andrews v DPP (1937),198 where a driver had killed a pedestrian whilst overtaking another car. There was little dispute that D had committed the offence of dangerous driving, but did that automatically make him guilty of manslaughter when death resulted? The House of Lords held that it did not: since the essence of dangerous driving was negligence, a driver should only be convicted of manslaughter if his driving was so bad as to amount to the gross negligence required under the second head of involuntary manslaughter (see below). Whether or not the decision was motivated by tenderness towards motorists is hard to tell, but there is certainly some logic in keeping offences of negligence out of the ‘unlawful act’ doctrine when a separate head of manslaughter by gross negligence exists. The logic of the second exception is less evident, and cases of omission have not always been treated differently. In Senior (1899)199 a man who belonged to a religious sect called the Peculiar People refused to call a doctor to his child, who subsequently died; he was held guilty of manslaughter on the ground that he had committed an unlawful act (wilful neglect of the child) which caused death. However, this very reasoning was abjured in Lowe (1973),200 where D failed to ensure that medical help was summoned to his child, and (p. 288) it died. The Court of Appeal held that a manslaughter verdict would not necessarily follow from a conviction for wilful neglect:

if I strike a child in a manner likely to cause harm it is right that if the child dies I may be

Page 39 of 69

Homicide charged with manslaughter. If, however, I omit to do something with the result that it suffers injury to health which results in death, we think that a charge of manslaughter should not be an inevitable consequence, even if the omission is deliberate.

This passage suggests that the law should, and does, draw a distinction between the blameworthiness of acts and omissions, even where the omission is deliberate. Yet the connection between withholding medical aid and subsequent death is surely closer than that between striking a child once and subsequent death. The father's duty in Senior and in Lowe is manifest and incontrovertible. If the ‘unlawful act’ doctrine is thought sound, these cases should fall squarely within it. If the doctrine is thought unsound, it should be abolished. This manifestation of the distinction between acts and omissions is morally untenable.201 Once it has been established that D was committing a criminal offence, the second step is to establish that this caused the death. In most cases of battery or actual bodily harm the causal connection will be plain, but cases involving drugs have presented difficulties. In Kennedy (No 2)202 D passed a syringe containing heroin to V, who injected himself and later died. The Court of Appeal upheld D's conviction for manslaughter, on the basis that the unlawful act was causing a noxious substance to be taken by V,203 and that D was acting in concert with V and therefore bore joint responsibility for the offence. However, the House of Lords overruled this decision,204 and re-affirmed the principle that a voluntary act (i.e. V's self-administration of the drugs) breaks the causal chain and prevents D from bearing responsibility for the death. The third requirement is that the defendant's conduct in committing the crime must have been objectively dangerous. This was seen as a slight restriction of the doctrine when it was imposed in Church (1966),205 where the Court held that ‘the unlawful act must be such as all sober and reasonable people would inevitably recognize must subject the other person to, at least, the risk of some harm resulting therefrom, albeit not serious harm’. The House of Lords has declined to narrow this test by requiring that D recognized the risk.206 The test remains largely objective, but not entirely so. The dangers inherent in the situation should be judged on the basis of a reasonable person in that position, endowed with D's knowledge of the surrounding circ*mstances. Thus an ordinary person who burgled the house of an elderly resident would realize the possible dangers as (p. 289) soon as the age and frailty of the householder became apparent,207 whereas the ordinary person would not know (if D did not know) that an apparently healthy girl of 15 had a weak heart.208 However, the reasonable person does not make unreasonable mistakes, and so the mistake of D who carelessly loaded a gun with a live cartridge thinking that it was blank was not taken into account.209 One element of the Church test—‘some harm … albeit not serious harm’—has been construed restrictively. In Dawson (1985) D, wearing a mask and carrying a pickaxe handle, approached a petrol-station attendant and demanded money; D fled when the attendant pressed the alarm bell, but the attendant then suffered a heart attack and died. The Court of Appeal held that the unlawful act would be regarded as ‘dangerous’ only if it was likely to cause physical harm, not if mere emotional shock (unaccompanied by physical harm) was foreseeable. The manslaughter conviction was quashed, partly because the trial judge had given the impression that conduct likely to produce emotional disturbance would be sufficient.210 Both the elements of ‘unlawfulness’ and ‘dangerousness’ in this form of manslaughter can be criticized. One important criticism is that these elements together fail to identify sufficiently

Page 40 of 69

Homicide clearly conduct that is sufficiently blameworthy to justify conviction for manslaughter, but is not in essence conduct where negligence as to unforeseen consequences plays the crucial role in determining blamworthiness. For, as we saw just now in the discussion of Andrews v DPP, where negligence is playing that role it should ordinarily not be sufficient to justify liability for manslaughter unless it amounts to gross negligence in breach of a duty of care not to cause death (a basis for manslaughter discussed in the next section). By contrast, in French law, the focus in the equivalent offence is on neither the unlawfulness of conduct, as such, nor on the danger of minor harm that it poses. Instead, Art. 222-7 says simply that, ‘acts of violence causing an unintended death’ may be punished by up to fifteen years’ imprisonment.211 (b) Manslaughter by gross negligence This second variety of ‘involuntary’ manslaughter has suffered no fewer changes of direction than the first. Gross negligence became well established as a head of manslaughter in the nineteenth century, and then all but disappeared from the law in the 1980s. Thus in Finney (1874),212 where an attendant at a mental hospital caused the death of a patient by releasing a flow of boiling water into a bath, the test was whether he had been grossly negligent. In Bateman (1925),213 where a doctor had attended the confinement of a woman who died whilst giving birth, the Court of Criminal Appeal held that there must be negligence over and above that which is sufficient to establish (p. 290) civil liability, and which shows ‘such disregard for the life and safety of others’ as to deserve punishment. This test was approved by the House of Lords in Andrews v DPP (1937).214 In Lamb (1967)215 two young men were joking with a gun; D pointed it at V and pulled the trigger, believing that it would not fire because neither bullet was opposite the barrel. The gun was a revolver, however, and it did fire, killing V. The Court of Appeal held that D might properly be convicted if his belief that there was no danger of the gun firing had been formed in a criminally negligent way. The beginnings of a change of direction appeared in Stone and Dobinson (1977),216 where two people were convicted of manslaughter for allowing a sick relative, whom they had permitted to live in their house, to die without medical attention. The Court of Appeal's grounds for finding a duty of care in this case are scrutinized elsewhere.217 The fault element required was expressed as recklessness, and defined thus: ‘a reckless disregard of danger to the health and welfare of the infirm person. Mere inadvertence is not enough. The defendant must be proved to have been indifferent to an obvious risk of injury to health, or actually to have foreseen the risk but to have determined nevertheless to run it.’ This passage contrasted ‘mere inadvertence’ with ‘indifference to an obvious risk’, perhaps foreshadowing the change that was about to take place. In the 1980s it was the concept of recklessness, in the Caldwell sense,218 that came to dominate this variety of manslaughter. Both the House of Lords in Seymour219 and the Privy Council in Kong Cheuk Kwan220 propounded this as the proper test, and it was widely assumed that manslaughter by gross negligence had been absorbed into and replaced by reckless manslaughter. In Adomako (1995)221 the House of Lords re-established manslaughter by gross negligence, and jettisoned manslaughter by Caldwell recklessness. Lord Mackay held that manslaughter by gross negligence requires the prosecution to prove (i) that D was in breach of a duty of care towards the victim, (ii) that the breach of duty caused the victim's death, and (iii) that the breach of duty amounted to gross negligence.

Page 41 of 69

Homicide What determines the existence of a duty? Lord Mackay took the view that this was simply a matter of consulting the law of tort, and some decisions can be thus explained. Certain duty situations are well established, such as parent–child and doctor–patient. Others have been recognized in previous decisions: there are the omissions cases where D has a contractual duty to ensure safety,222 and where D was initially responsible for creating a hazardous situation.223 New duty-situations may be recognized, as in R v West London Coroner, ex p Gray (1988),224 where the Divisional Court recognized that police officers have a duty of care towards persons they arrest, particularly persons who are intoxicated. In Prentice (1994)225 the Court of Appeal recognized the duty of an (p. 291) electrician to leave the house safe for the householder who employed him. A company has a duty to ensure the health and safety of its employees and of others affected by its activities,226 for example. More recently, interest has focused on the criminal law's recognition of duties beyond those of the law of torts. This is not a new phenomenon, because the reasons adduced to support a duty to care for a sick relative in Stone and Dobinson (e.g. blood relationship, guest in house) remain controversial. In Wacker (2003)227 D was involved in a plan to bring sixty illegal immigrants into the UK in a container on his lorry, and fifty-eight of them died from suffocation. The point was taken that, since the sixty would-be immigrants had concurred in the plan, D did not have an enforceable duty of care towards them because of their voluntary involvement in the illegality. The Court of Appeal responded that, while this may be the situation in the law of tort, the criminal law has the wider function of protecting the public and it is therefore not subject to the same restrictions as the law of tort. In Willoughby (2005)228 D asked V to come to a disused public house that D owned and to help him set fire to it with petrol. The ensuing fire killed V and injured D. The Court of Appeal held that D was rightly convicted of the manslaughter of V, as well as the offence of arson endangering life, because D owed a duty to V. The Court held that the trial judge had been wrong to hold that the duty arose simply by virtue of D's ownership of the pub, but held that the duty stemmed from D's recruiting V to help and assigning him the dangerous task of spreading the petrol. As in Wacker, the effect is to go beyond dutysituations recognized by the law of negligence and to impose on D a duty towards people who are willing participants in the same enterprise. The decision in Willoughby recognizes that the existence of a duty is a question of law. But, in the absence of criteria for determining dutysituations,229 this appears to be common law decision-making at its retrospective worst. It is in a way understandable that the courts should wish to avoid tying the concept of duty in the law of criminal negligence to the concept as it is currently understood in the law of tortious negligence. The duty concept has evolved in tort to reflect considerations concerned with monetary compensation of one private citizen for damage done by another (or by the state or one of its agents) that have no direct application in the substantive criminal law. So, for example, it can be a relevant consideration limiting the duty of care in tort that D might—if a duty were imposed—be exposed in a disproportionate way to an unlimited liability to compensate to an unlimited class of persons; but that concern is not relevant in the criminal law.230 Moreover, in an ideal world, criminal trials would not be delayed (or followed) by lengthy secondary (p. 292) litigation over whether a duty of care arose in law, in the way that this heavily litigated question has pre-occupied civil courts for over a century. Accordingly, the criminal courts have settled for a simpler position, involving a two-step process: (a) the judge rules whether, as a matter of law, D's relationship with V was capable of giving rise to a duty, and if the ruling is that the relationship may involve in law a duty, then (b) it is for the jury to decide whether that duty arose on the facts of the case before them.

Page 42 of 69

Homicide Much simpler though this approach may be, it sacrifices a great deal in point of certainty, in that it will be difficult for an ordinary person to know in advance when a duty of care will arise, and therefore whether they must do or not do certain things to avoid another's death if they are to escape criminal liability for manslaughter. A good example of the kind of retrospective lawmaking to which this approach can lead is to be found in Evans.231 D had given her 16year-old half-sister heroin at home, even though her sister was a recovering addict. V selfinjected, and then began to show clear symptoms of having overdosed. D and her mother appreciated that there was a danger to V's health, but were afraid of getting into trouble if they called the emergency services. Consequently, they simply put V to bed and checked on her periodically, sleeping in the same room as her. By the morning, V had died from heroin poisoning. D was convicted of manslaughter and appealed, but the Court of Appeal upheld her conviction. It did so, however, not in virtue of D having supplied V with the heroin in the first place, but on the basis that, having contributed (by that act of supply) to the creation of a situation in which there was a risk of V's death, she had a duty to seek to rectify it by taking reasonable steps, and in the circ*mstances breached that duty in a grossly negligent way through failing to summon help before V died. This finding involved an extension to the Miller232 doctrine that the unjustified creation of a dangerous situation may lead to a duty to take reasonable steps to eliminate the danger (in that case, the danger posed by letting a cigarette fall and set fire to the room). The extension comes about because it was V who herself created the danger of death by self-injecting with heroin; D merely assisted in the creation of that danger by V. Small extension to the Miller doctrine though it might seem, it does involve judicial extension of the ambit of criminal liability for manslaughter through case by case development. This is unsatisfactory, in that in the controversial area of dangerous drug-taking, the issues ought to be addressed through a proper consultative process that it is beyond the courts to undertake before they extend the law.233 In French law, a combination of Art. 221-6 and Art. 121-3 provides for just the kind of situation encountered in Evans. By Art. 221-6, the causing of death by negligence is an offence punishable by up to three years’ imprisonment, and by Art. 121-3, that offence can be committed in the following circ*mstances:

natural persons who have not directly contributed to causing the damage, but who have created or contributed to create the situation which allowed the damage to happen who (p. 293) failed to take steps enabling it to be avoided, are criminally liable where it is shown that they have … committed a specified piece of misconduct which exposed another person to a particularly serious risk of which they must have been aware.

As a matter of doctrine, a less convoluted route to conviction in Evans might have been to say that D's supply of the heroin to V was a grossly negligent act that was itself still an operating cause of V's death. It was still an operating cause, in that—being only 16—V's decision to selfinject a controlled drug was not adequately free and informed, and hence did not break the chain of causation from D's supply; only self-injection by someone 18 years old or older (in full knowledge of the facts) would adequately meet the ‘free, deliberate, and informed’ criterion and hence break the chain of causation.234 Even taking that course would, though, involve an extension of the scope of the law. Once it is established that there was a duty, that it was breached, and that this caused the

Page 43 of 69

Homicide death, there is the question of the terms in which the test of gross negligence is to be put to the jury. Lord Mackay LC in Adomako held that gross negligence depends:

on the seriousness of the breach of duty committed by the defendant in all the circ*mstances in which he was placed when it occurred and whether, having regard to the risk of death involved, the conduct of the defendant was so bad in all the circ*mstances as to amount in the jury's judgment to a criminal act or omission.235

Lord Taylor CJ in the Court of Appeal had earlier stated that there were other types of case that might justify a finding of gross negligence, notably cases where there was actual awareness of a risk combined with indifference to it or a grossly negligent attempt to avoid it.236 It does not seem difficult to encompass these other cases within the Adomako test, so long as the focus remains on ‘the risk of death’. If such a risk was reasonably foreseeable, then the jury must decide whether D's conduct fell so far below the expected standard as to justify conviction for manslaughter. It has often been observed that this test is circular: if members of the jury ask how negligent D must have been if they are to convict him of manslaughter, the answer is ‘so negligent as to deserve conviction for manslaughter’. Significant as the circularity point may be, more powerful are the arguments that a) it fails to meet the test of certainty properly required of a criminal law by Art. 7 of the Convention, and b) its breadth leads to unfair inconsistencies in prosecution policy. The Adomako test was challenged on the former basis in Misra (2005),237 and drew the unconvincing response that the question ‘is not whether the defendant's negligence was gross and whether, additionally, it was a crime, but whether his behaviour was grossly negligent and consequently criminal’. This is a distinction without a difference and, despite the Court's discussion of some of the (p. 294) Strasbourg authorities, it should not be the last word on the subject. Indeed, Judge LJ went on to state that ‘this is not a question of law, but one of fact’. Lord Mackay's words in Adomako make it clear that the jury is, in effect, deciding a question of law when it decides whether the conduct was bad enough to be classed as manslaughter. The second criticism of the Adomako test emerges from research by Oliver Quick into the decision-making of prosecutors in cases of fatal errors by medical staff,238 finding a number of unexplained variations in prosecution decisions that the broad terminology of the offence permits. Finally, we should note that, closely allied to manslaughter by gross negligence is the offence of manslaughter by (subjective) recklessness. There is some confusion over the relationship, in that in Adomako Lord Mackay indicated that it would not be wrong for a judge in directing the jury on the fault element in gross negligence manslaughter cases to use the term ‘reckless’ to describe D's conduct. However, it seems clear that causing death by recklessness is a form of manslaughter in its own right. In Lidar,239 V died under the wheels of D's car when he lost his grip on the car as it was being driven at speed by D following a fight involving them both. D was convicted of manslaughter by recklessness. The Court of Appeal upheld Lidar's conviction on the basis that he had foreseen a risk that V might suffer death or serious injury if he (D) continued to drive at speed, and yet he continued to do so, and caused V's death thereby. (c) Corporate manslaughter The Corporate Manslaughter and Corporate Homicide Act 2007 introduced a new form of

Page 44 of 69

Homicide manslaughter, corporate manslaughter, to English law. It was formerly possible to convict a company of manslaughter by gross negligence at common law, but the ‘identification doctrine’ (discussed in Chapter 5.3) proved so restrictive in practice that only a few convictions of smaller companies were obtained.240 Public concern at the considerable loss of lives resulting from companies’ operations, and belief in the fairness of imposing the censure of homicide convictions (rather than merely of convictions under the Health and Safety at Work Act 1974) in bad cases, led to the lengthy and politically controversial process of bringing forward legislation.241 The new offence can be committed only by an ‘organization’, and organizations can no longer be convicted of manslaughter by gross negligence (although individuals can). An individual cannot be held liable for the offence of corporate manslaughter, or for complicity in it. (p. 295) The provisions of the Act are beset with considerable technicality,242 but five key elements of the definition may be identified—(i) an ‘organization’ must (ii) owe a relevant duty of care and (iii) the way in which the activities were managed or organized must amount to a gross breach of that duty, (iv) a substantial element in that gross breach being the way that the organization's activities were organized by senior management, and (v) death must be caused by the way in which the activities were managed or organized. First, what qualifies as an ‘organization’ for the purpose of the Act? The definition goes well beyond companies and includes partnerships, some unincorporated associations, and most public bodies (such as hospital trusts, the police, and government departments).243 The practical implications of these broad categories are considerably narrowed down by the second requirement, that the organization must owe a relevant duty of care to the deceased person. The concept of duty of care, elaborated in s. 2, is confined to duties recognized by the law of negligence.244 Especially controversial in this respect was the prospect that liability could arise for deaths in custody, particularly where these occurred through a suicide that it is alleged could have been prevented. Intially, the Government delayed making the 2007 Act applicable to such deaths, but since 2011 it has been applicable in that situation as specified in s. 2(2). There are also several exclusions in ss. 3 to 7 from duties that would otherwise apply—exclusions dealing with public bodies’ decisions on resource allocation and public policy, military activities, the operations of emergency services, and duties under the Children Act 1989.245 Whether there was a relevant duty of care is a question of law for the judge; the technicality of the tests and the exclusions may give rise to considerable legal argument. Once the prosecution has satisfied the court that the defendant is an organization to which the Act applies, and that there was a relevant duty towards the deceased, the third element is that the way in which its activities were managed or organized amounted to a gross breach of the organization's duty. How can this be established? A breach is gross if the conduct allegedly amounting to the breach ‘falls far below’ what could reasonably be expected of the organization in the circ*mstances. This is a question of degree for the jury, similar to that which has to be decided when determining whether negligence is ‘gross’ for the purposes of manslaughter by gross negligence (7.5(b)). In this connection s. 8(2) requires the jury to consider whether there was a breach of health and safety legislation. If there was, then the jury should take account of how serious the breach was, how much of a risk of death it posed, and (by s. 8(3)) (p. 296) any evidence of what is often termed corporate culture, i.e. evidence of ‘attitudes, policies, systems or accepted practices within the organization that were likely to have encouraged’ any breach of safety legislation. This broadens the timeframe of the new offence, by reducing the possibility that the grossness of the breach is assessed

Page 45 of 69

Homicide simply on a ‘snapshot’ taken at the time of the fatal incident. The fourth element in the definition is that the way in which the organization's activities ‘are managed and organized by its senior management is a substantial element in the breach’. The term ‘senior management’ refers (s. 1(4)) to persons who play ‘significant roles’ in either decision-making about or actual managing of the whole or a substantial part of the organization's activities. This may prove to be a fairly restrictive definition, especially in relation to large organizations, since scrutiny falls only on those who play a significant role in relation to a substantial part of all the organization's activities. The test is probably a factual one (who played a significant role?), and courts are unlikely to be deflected by nomenclature. But then, once the persons who are senior managers are identified for the purposes of the Act, the jury must also be satisfied that those persons’ role in the activities was a ‘substantial element in the breach’, a restrictive phrase that invites argument about the role of a particular employee's fault, and thereby revisits some of the problems of the ‘identification doctrine’ in corporate liability generally.246 The fifth requirement is that the way in which the organization's activities were managed caused the death. The issue of causation is not straightforward: presumably a ‘more than minimal’ cause is sufficient,247 and, since there is no provision in the Act to prevent the application of normal principles, a voluntary intervening act (such as the conduct of an employee) would break the causal chain,248 clearly not the intention behind the Act. The new offence of corporate manslaughter is an important step towards recognition that corporate liability in this sphere is fair and that it has to be constructed differently from individual liability. There are, however, various respects in which the Act could have been improved. The Act combines great technical complexity in some respects with considerable open texture in key terms, such as ‘gross’, ‘significant’, and ‘substantial’. Moreover, it remains to be seen whether its application is easier and more extensive than the identification doctrine that applied previously: the notion of ‘senior management’ still requires the court to identify people within an organization who had a certain amount of influence and whose failure was a substantial element in the breach. On the other hand, if the test proves to be applicable without difficulty to larger organizations, the kernel of this approach could and should be adapted to other forms of corporate liability. The requirement of the consent of the DPP for any prosecution is regrettable, since the possibility of private prosecution could have (p. 297) operated to prevent any official ‘cover-ups’. The relationship between the new offence and offences under the Health and Safety at Work Act 1974 remains to be worked out: 249 a reconsideration of that legislation, with a view to reformulating its offences, would be a sensible next step. In the meantime, the Sentencing Guidelines Council has issued sentencing guidelines on both corporate manslaughter (for which only three forms of sentence—fines, remedial orders, and publicity orders—are possible) and offences under the Health and Safety at Work Act resulting in death.250 (d) The contours of involuntary manslaughter The English law of manslaughter exhibits a tension between the significance of the harm caused and various principles of fairness such as the principles of correspondence and fair labelling. It is the resulting harm (death) which still dominates, and the enormous moral distance between D's conduct and the fatal result is evident from the fact that in many

Page 46 of 69

Homicide situations there may be nothing more than a conviction for common assault if death does not result (manslaughter by unlawful act if death results) or even no criminal offence at all (manslaughter by gross negligence if death results). Much is made of the unique significance of human life and the need to mark out, and to prevent, conduct which causes its loss. But does this really justify the present contours of the law of involuntary manslaughter? It is important not to neglect the fact that manslaughter currently covers a wide range of culpability. The focus thus far has been on the lower borderline, but there are some forms of manslaughter that fall little short of murder, such as manslaughter by recklessness.251 There is no doubting the substantial culpability of the person who embarks on a course of conduct knowing that there is a risk of death or serious injury to another (e.g. the man who administered carbon tetrachloride to the woman in Pike,252 knowing the danger of physical harm to her). We noted earlier that the Law Commission has recommended that some forms of reckless killing (where D intended to cause injury or fear or risk of injury, knowing that the conduct involved a serious risk of causing death) should be included in the new offence of murder in the second degree,253 with lesser varieties falling within the offence of manslaughter.254 The main focus of our discussion above was on the lower threshold of manslaughter, where its minimum requirements form the boundary with accidental (non-criminal) homicide. It can be strongly argued that to apply the label ‘manslaughter’ to the conduct of a person who envisaged no more than a battery, e.g. by a single punch, is both disproportionate and unfair. It is only luck that makes the difference between (p. 298) the summary offence of common assault (maximum, six months’ imprisonment) and the grave offence of manslaughter (maximum, life imprisonment). In such cases, the manslaughter label exaggerates the amount of culpability, producing an extreme form of constructive liability.255 The Law Commission originally accepted this reasoning and recommended the abolition of unlawful act manslaughter,256 but this has now been replaced—without detailed justification—by a recommendation that adopts the Government's own formulation of a possible offence of manslaughter based on death caused by an act that D intended to cause injury or was aware carried a serious risk of injury.257 This reversal will be welcome to those who argue that D's responsibility for the death cannot be avoided: it is something D did (he killed by acting dangerously), and he bears some moral responsibility for it.258 It will also be welcome to those who maintain that a person who changes his or her normative position by attacking another ought to be held liable for manslaughter if death results. Any ‘bad luck’ in holding the attacker guilty is clearly traceable to the offender's fault in making the attack.259 However, it can be strongly counter-argued that the language of ‘attack’ and ‘change of moral position’ fails to address the enormous gulf between what was intended and what (by mischance) resulted.260 Statistically speaking, the risk of death from a single punch is far too remote to enter into reasonable contemplation, and it is not clear how significant the proposed restriction to causing ‘injury’ would be. Death would not be an intrinsic risk when injuring another, rather than injuring another seriously. It seems wrong to attribute too much weight to chance: ‘the offender's fault falls too far short of the unlucky result. So serious an offence as manslaughter should not be a lottery’.261 If D's conduct was not serious enough to constitute reckless manslaughter (as described in the previous paragraph), and does not amount to manslaughter by gross negligence, the proper course is simply to convict D of whatever other offence he has committed and to pass sentence for that.262

Page 47 of 69

Homicide Where does this leave the crime of manslaughter by gross negligence? Negligence is not usually a sufficient fault element for serious offences. The Law Commission, after a review of the arguments, concluded that it is justifiable in homicide cases to criminalize gross negligence where D unreasonably takes a risk of causing death, where the failure to advert to the risk is culpable because the risk is obviously foreseeable and D (p. 299) has the capacity to advert to the risk.263 The Commission recommends an offence of killing by gross negligence, with the following elements:

(1) a person by his or her conduct causes the death of another; (2) a risk that his or her conduct will cause death would be obvious to a reasonable person in his or her position; (3) he or she is capable of appreciating that risk at the material time; and (4) either his or her conduct falls far below what can reasonably be expected of him or her in the circ*mstances; or he or she intends by his or her conduct to cause some injury, or is aware of, or unreasonably takes, the risk that it may do so, and the conduct causing (or intended to cause) the injury constitutes an offence.264

This formulation has a number of good features. It incorporates a capacity requirement (3), and now insists that the objective risk must be one of death (2).265 There is no special provision for omissions, but such cases should fall within (4) and the criteria for deciding whether or not there was a duty to act will continue to be left for development at common law. The formulation of condition (4)(a) does not go far in the direction of maximum certainty, but the Commission argues that at least the new test would not be circular (a criticism levelled at the Adomako test), and that the only alternative to leaving ‘a large degree of judgment to the jury’ would be ‘to define the offence in such rigid and detailed terms that it would be unworkable’.266 As for the second limb of (4), this is advanced as a test that may be simpler for juries to apply than the test in (4)(a), and one that is likely to be co-extensive in practice with (4)(a). It does not, of course, make any explicit reference to gross negligence, and its form is similar to the existing ‘unlawful act’ doctrine—a doctrine whose abolition the Law Commission originally recommended but whose retention, in modified form, the Commission now supports. The difference in this context is that condition (2) must be satisfied before condition (4)(b) would be applied. The Law Commission envisages that the maximum penalty for killing by gross negligence would be a determinate sentence, not life imprisonment, but no conclusions are reached on its precise grading. However, before we leave the question of how the law of involuntary manslaughter should be structured, it is important to assess various other homicide offences and to consider their proper place. (p. 300) 7.6 Causing or allowing the death of a child or vulnerable adult There has long been concern about the difficulty of achieving a homicide conviction when the death of a young child has been caused by one of the child's parents or carers but it cannot be proved which. Parliament has responded by introducing the offence of causing or allowing Page 48 of 69

Homicide the death of a child or vulnerable adult, contrary to s. 5 of the Domestic Violence, Crime and Victims Act 2004. Whether the creation of a new offence was necessary, as distinct from special procedural means of bringing such cases within mainstream homicide offences, warrants further examination.267

7.7 Causing death by driving English law now contains some five offences of causing death by driving: (1) causing death by dangerous driving, contrary to s. 1 of the 1988 Act; (2) causing death by careless driving when under the influence of drink or drugs, contrary to s. 3A of the 1988 Act; (3) causing death by careless driving, contrary to s. 2B of the 1988 Act; (4) causing death by driving when unlicensed, disqualified, or uninsured, contrary to s. 3ZB of the 1988 Act; (5) causing death by aggravated vehicle-taking, contrary to s. 1 of the Aggravated Vehicle-Taking Act 1992. It is not possible to go into the detail of all these offences, but as they take the law of involuntary homicide well beyond the scope of gross negligence or unlawful and dangerous act manslaughter, it is worth considering the issues of principle that some of them raise. The offence of causing death by dangerous driving replaced the former offence of causing death by reckless driving and, unlike that offence, is defined in the legislation. This is a considerable step towards greater certainty in the criminal law. In outline, s. 2A(1) provides that a person drives dangerously if ‘(a) the way he drives falls far below what would be expected of a competent and careful driver; and (b) it would be obvious to a competent and careful driver that driving in that way would be dangerous’. Section 2A(2) adds that a person also drives dangerously if it would be obvious to a competent and careful driver that driving the vehicle in its current state would be dangerous—for example, driving an obviously defective vehicle or driving with an (p. 301) unsteady load.268 Section 2(3) defines ‘dangerous’ in terms of danger either of injury to any person or serious damage to property, and provides that any special knowledge possessed by the driver should be taken into account. This is an objective standard, but its extension to cases where only serious damage to property (and not death or injury) is foreseeable may be considered too wide. Since it applies to conduct on the road that falls ‘far below what would be expected’, the standard may therefore be higher than that of negligence in the law of tort, and is approaching (or equivalent to) a standard of gross negligence. The maximum for this offence is fourteen years, following the Criminal Justice Act 2003, compared with a maximum of five years for the offence of dangerous driving. It remains possible to convict drivers of murder where the required fault element can be proved,269 and likewise of manslaughter where the prosecution can establish, following Adomako,270 that D was grossly negligent as to the risk of death, compared with a high degree of negligence as to injury or damage, as required for causing death by dangerous driving. The Road Safety Act 2006 has introduced two further offences. The first is causing death by careless driving, with a maximum sentence of five years’ imprisonment. The Act states that careless or inconsiderate driving means driving that ‘falls below what would be expected of a 271 Page 49 of 69

Homicide competent and careful driver’,271 and this contrasts with the dangerous driver whose driving must fall ‘far below’ that standard. However, the offence of careless driving is chiefly intended to penalize small errors of judgment, and it can hardly be said that the offence is intended to protect people's lives (unlike dangerous driving). Thus it can be said that the moral distance between the underlying offence of careless driving—for which Parliament has provided only a fine as the penalty—and causing death by careless driving, with its maximum of five years, is too great, and that this is an improper use of a homicide offence. This is not to downplay the concern and grief of the families of victims of these offences; but that must be responded to in a different way, rather than by excessive punishment of someone whose error may have been both slight and momentary. The second new offence is causing death when driving while unlicensed, disqualified, or uninsured. The essence of this offence is simply a minimal element of causation: if the driver causes death when he is committing one of these three other offences (no valid licence, disqualified from driving, no insurance), he is guilty of this homicide offence without proof of any fault in the driving.272 Indeed, if there were fault in the driving, one would expect a prosecution for one of the three offences with higher maxima. The reason for creating a homicide offence for deaths caused in these circ*mstances is that the driver should not have been on the road at all: his or her decision to drive when not permitted to do so was a sine qua non of the incident that caused death. Parliament regarded this as the least serious of the offences, (p. 302) assigning it a maximum penalty of two years’ imprisonment. That may be taken as an indication of the absence of a fault requirement for the actual driving. However, research among members of the public shows unequivocally that this view is widely rejected: where a disqualified driver takes to the road, and happens to cause death through no fault in the manner of driving, most people regard this as more serious than causing death by careless driving273 —whereas Parliament evidently viewed it as less serious by a factor of 2 to 5. This research was commissioned in order to assist with the drafting of guidelines for sentencers, but in fact it made that task more difficult, once it became apparent that there was widespread public rejection of the hierarchy of maximum penalties that had been created. Most people, it seems, would wish to see a reversal of the maxima for causing death by careless driving and causing death when driving while unlicensed, disqualified, or uninsured; but the sentencing guidelines must reflect the law as it stands.274 Is there a need for these separate homicide offences? The first offence of causing death by dangerous driving was introduced in 1956, largely because juries were unwilling to convict culpable motorists of such a serious-sounding offence as manslaughter. Ever since its introduction there have been those who have pointed to its ‘illogicality’.275 The difference in practice between an offence of dangerous driving (maximum penalty of five years) and one of causing death by dangerous driving (maximum penalty of fourteen years) may simply be one of chance. Bad driving may or may not lead to an accident, depending on the chance conjunction of other factors and other people's behaviour. And an accident may lead to death (in which case the more serious offence is committed) or merely to serious injuries or to minor damage. The response to this ‘illogicality’—which is, of course, the very problem with the law of involuntary manslaughter too—has varied. Both the James Committee in 1976276 and the Criminal Law Revision Committee in 1980277 recommended the abolition of the offence of causing death by dangerous driving, thereby accepting the ‘illogicality’ argument. This accords with the CLRC's proposal that ‘unlawful act’ manslaughter should be abolished.278 However, the North Report on road traffic law reversed this trend. The report accepted the principle that, in general, persons should be judged according to the intrinsic quality of their

Page 50 of 69

Homicide driving rather than its consequences, but argued that the law should depart from this in cases where death is caused and the driver's culpability is already high.279 There is a well-known risk in motoring that certain kinds of driving may cause accidents, and that accidents may cause death. The rules of the road are designed not only to produce the orderly and unhampered movement of traffic, but also to protect property, safety, (p. 303) and lives. One who deviates so manifestly from these rules as to drive dangerously ought to realize— because the driving test requires a driver to realize—that there is a considerable risk of an accident. If an accident happens as a result of driving which falls well below the proper standard, then that may well be a case of culpable negligence even if the driver had never thought of the risk in that particular case, because the driver is presumed to know the Highway Code. Sound as this reasoning may be where it is dangerous driving that results in death, it is significantly less convincing in other cases.280 Perhaps the next most serious cases would be those where D drives after taking considerable alcohol or drugs, and cases where D drives after being disqualified from driving, and with both of these it is debatable whether the fault is sufficient to justify conviction and sentence for a homicide offence. One view is that it is sufficient—a decision to drive while intoxicated flies in the face of widely advertised safety campaigns, and a decision to drive while disqualified ignores the road safety reasons that led to the disqualification. Another view is that, in these cases and more generally, the current trend places far too much emphasis on the occurrence of death. These are not cases in which death is intended or knowingly risked: many of them are cases of negligence, to a greater (dangerous) or lesser (careless) degree, albeit that the risks to safety involved in motoring are well known. The great significance attributed to the accident of death is more appropriate to a compensation scheme than to a system of criminal law. Yet if the criminal law in motoring cases were to focus on ‘intrinsic’ fault rather than the consequences of the bad driving, it would come down much harder on many people who by good fortune did not cause any or much harm even though their driving fell appallingly below the required standard. Those who think it wrong that the courts should respond so readily to the ‘accident of death’ would equally have to harden themselves to reject the pleas of drivers who say ‘at least I did no harm’. If the courts were to focus on the intrinsic fault in a defendant's driving, sentencing would become more difficult in itself and more controversial in the view of the mass media.

7.8 Reviewing the structure of the law of homicide In this chapter we have discussed a wide array of different homicide offences. Towards the end of the chapter, questions about the proper contours and boundaries of the law of homicide have become more and more pressing. Although some suggestions and criticisms have been ventured at appropriate points, we may conclude the chapter with some broader reflections on the structure of the law of homicide. (p. 304) Three main interlinked issues present themselves—questions about appropriate fault requirements, questions about appropriate labels, and questions about appropriate sentence levels.281 Let us begin with fault: the principle should be that of equal treatment of offences of equal seriousness. There should therefore be an alignment of the minimum culpability requirements for homicide offences, unless there are strong reasons to the contrary. For example, manslaughter by gross negligence requires the risk of death to have been obvious,

Page 51 of 69

Homicide whereas causing death by dangerous driving may be committed if there was danger of injury or serious damage to property—requirements which are surely far too low for a homicide offence. The offence of corporate manslaughter is rather reticent on the whole issue, although s. 8 of the Corporate Manslaughter and Corporate Homicide Act 2007 does state that, where the death arose from a breach of health and safety regulations, the jury should consider how serious a risk of death existed.282 However, we find major departures from this principle when it comes to the two new homicide offences of causing death by careless driving and causing death when driving while unlicensed, disqualified, or uninsured. Causing death by careless driving falls significantly below the threshold of fault required for manslaughter by gross negligence, and for causing death by dangerous driving. Moreover, the rationale for having an offence of careless driving is not so much to save lives (one purpose of the offence of dangerous driving) as to protect from injury and damage to property, so the claims for the new offence to be admitted as a form of criminal homicide are low. The arguments relating to causing death when driving while unlicensed, disqualified, or uninsured are different, because it is the initial criminal act of driving when not permitted to do so that colours the consequences. The parallel is therefore with manslaughter by unlawful act, and the question is whether the moral distance between the originating criminal act and the tragic (accidental) result is too great to justify its inclusion as a homicide offence. This does not mean that the defendant is not convicted, since the underlying criminal offence is still there, and the legislature could create another (non-homicide) offence if it were thought morally and socially appropriate. The chief argument in favour of including these as homicide offences is that the deliberate commission of a criminal offence changes D's normative position such that it is fair to hold him liable for the fatal (if unanticipated) consequences—a view which, as we have seen, begs enormous questions, such that the original progenitor of the ‘change of normative position’ reasoning no longer supports it.283 This leads us into the second general issue, that of labelling. In principle the label applied to an offence should be a fair represe