Killing in Self-Defence (Oxford Monographs on Criminal Law and Justice) [1 ed.] 9780199283460, 019928346X

Citation preview

KILLING IN SELF-DEFENCE

This page intentionally left blank

Killing in Self-Defence Fiona Leverick

3

3

Great Clarendon Street, Oxford OX2 6DP Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide in Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With offices in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries Published in the United States by Oxford University Press Inc., New York © F. Leverick, 2006 The moral rights of the author have been asserted Database right Oxford University Press (maker) Crown copyright material is reproduced under Class Licence Number C01P0000148 with the permission of OPSI and the Queen’s Printer for Scotland First published 2006 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this book in any other binding or cover and you must impose the same condition on any acquirer British Library Cataloguing in Publication Data Data available Library of Congress Cataloging-in-Publication Data Leverick, Fiona. Killing in self-defence / Fiona Leverick. p. cm. Includes bibliographical references and index. ISBN-13: 978–0–19–928346–0 (hardback: alk. paper) 1. Justifiable homicide. 2. Self-defense (Law) I. Title. K5171.L48 2006 345⬘.04—dc22 2006032314 Typeset by Newgen Imaging Systems (P) Ltd., Chennai, India Printed in Great Britain on acid-free paper by Biddles Ltd, King’s Lynn ISBN 0–19–928346–X

978–0–19–928346–0

1 3 5 7 9 10 8 6 4 2

General Editor’s Preface Self-defence is one of the oldest ‘defences’ to homicide and has attracted a rich literature both in the common law world and beyond. In this monograph, Dr Fiona Leverick subjects the theory and practice of the law of self-defence to close scrutiny, examining whether it should be recognized as a justification or as an excuse, whether there are good arguments for some duty to retreat, whether the immediacy requirement should be maintained, to what extent killings to prevent rape or to protect property can be supported, and other key issues of principle. The author draws together statutory provisions and judicial decisions from a wide range of jurisdictions, and evaluates them in the light of a strong commitment to the right to life and to other values inherent in the European Convention on Human Rights. The boundaries of the right to self-defence are inevitably controversial, but Dr Leverick succeeds in demonstrating the important values at stake in what may sometimes be regarded as the small details of the law on self-defence. Her arguments should lead to considerable re-thinking of the law on this sensitive subject. Andrew Ashworth

This page intentionally left blank

Preface This book started off life as a doctoral thesis on the law of self-defence in Scotland and England written while the author was a student in the School of Law at Aberdeen University. The original (and, with the benefit of hindsight, naively ambitious) intention of the thesis was to undertake an analysis not just of self-defence but of all criminal defences. Self-defence was selected as the first defence to be examined on the basis that it was likely to be relatively uncontroversial and unproblematic in comparison with some other defences, such as duress, necessity or insanity, and could therefore be dealt with reasonably quickly. This assumption proved to be wrong. It soon became evident that self-defence was a complex and fascinating defence in its own right and contained more than enough substance for three years’ work. I started to focus specifically on self-defence as a defence to homicide. The more I thought about this, the more intrigued I became about why the defence is so readily accepted and why the rules governing the defence, in comparison to other related defences such as duress and necessity, are relatively relaxed. Although not usually recognized as such, self-defence is a contentious defence, permitting as it does the victim of an attack to preserve her life at the expense of another. If one holds that all human life is of equal value, explaining why this is permissible poses something of a challenge. It is particularly difficult to explain where the aggressor is, for reasons such as non-age or insanity, not responsible for her actions. I turned to philosophical material which, without a background in philosophy, was initially an uphill struggle. But I persevered and found a rich seam of material in this area, which I have drawn on in Chapter 3 of the book. But I never wanted this to be a book solely about philosophy, nor am I qualified to write one. My background before coming to law was in marketing and psychology, not the ideal preparation for delving into the world of moral philosophy. Thus it was always the intention to examine the legal rules governing the defence as well as its philosophical foundations. The book has developed and changed in emphasis from the text of the original thesis. For one thing, the thesis focused specifically on the law in England and Scotland (for no real reason other than convenience, although I did, of course, construct a convincing rationale for the selection of these two jurisdictions in my viva), whereas the book is much more comparative in nature. For another thing, the debate has moved on. The thesis was completed at the end of 2002, whereas the book manuscript was submitted in 2006. This meant that I had a lot of catching up to do and I found that while I had

viii

Preface

been away, self-defence had finally started to attract a little more attention in the legal literature. In addition, a parallel body of literature on self-defence in international law had developed at alarming speed, reflecting world events in the intervening four years. The book also contains a chapter on the implications of the European Convention on Human Rights for the law of selfdefence, which the thesis did not, and this reflects the important role that the ECHR has come to play over the last few years. For all of this, my views have changed very little since I wrote my thesis. My primary concern is with the right to life of both aggressor and victim and it is my firm conviction that the rules governing self-defence should be formulated with this in mind. As such, this is a book about the right to life, as much as it is a book about self-defence. A big thank-you is due to all of those who helped me complete this book, whether by engaging in (fascinating?) discussions about the law of selfdefence or simply by tolerating my increasingly anti-social behaviour, especially during the first few months of 2006. Some people deserve to be singled out for special attention. Chris Gane was the supervisor of my doctoral thesis and indeed the inspiration for me embarking on a PhD in criminal law in the first place. If his Vice-Principal duties ever mean that he becomes too busy to take firstyear criminal law lectures on the LLB that would be a great shame. Thankfully, given that I would be first in line to take over, this shows no sign of happening yet. Pete Duff read all the way through my thesis in the course of the week before it was submitted. Despite him having no particular interest or expertise in substantive criminal law, he came up with some very pertinent points and all of these have now been incorporated into the book. He has also continued to try and persuade me to join the Law School’s 10.30 am coffee break, despite my persistent claims of being too busy. Pete—the first thing I will do once this manuscript is submitted is re-join the coffee circuit, so I hope you’ve got some sparkling and witty conversation ready. James Chalmers read almost the entire book manuscript in the space of a weekend, despite the fact that he was trying to finish a book of his own at the same time. I am enormously grateful to him for all his helpful suggestions and for spotting all the rogue square brackets where I should have used curly ones. At an earlier stage, his ever expanding personal library meant that I never needed to venture beyond the top corridor of Taylor Building to research the law. He also kept me up-to-date on Law School gossip when I retreated into temporary solitary confinement at home to work on the book. Shona Wilson worked as a research assistant for me for most of the time I was preparing this manuscript. It is no exaggeration to say that, without her, this book would never have been completed. I came to rely on her astonishing organizational abilities and talent for spotting bold full stops perhaps

Preface

ix

more than I should have done. And yes, Shona, I have managed to sneak in a reference to the Guam Criminal Code just for you. John Finch, my husband, supported me all the way through the PhD and the book and has forcibly turned my laptop off at 10 pm every night, enabling me to forget about self-defence and listen to Book at Bedtime on Radio 4. Some final thank-yous: Victor Tadros, who was my co-supervisor before he left Aberdeen and whose outstanding work on criminal law theory has forced me to think very hard about parts of my argument; Andrew Ashworth, for being so enthusiastic and encouraging when I informally submitted the proposal to OUP; Pamela Ferguson and Tony Smith, my PhD examiners, whose helpful comments have all informed the final text of the book; John Louth, Dédé Tété-Rosenthal, Louise Kavanagh, and Gwen Booth, all of OUP (Gwen in particular always answered my questions promptly and helpfully and did not shout at me when I conformed to the academic author stereotype and asked for extra time to complete the text); Virginia Williams, my copy-editor, who managed to find errors in my text even after I had proof-read it three times; and Christine Gane, who saved my life by doing the index for me. The responsibility for errors and omissions remains, as normal, my own. Fiona Leverick 2006

This page intentionally left blank

Contents Table of Cases Table of Legislation 1. Introduction 1.1 Why Self-Defence? 1.2 Defining the Subject Matter of the Book 1.3 A Note about Terminology 1.4 Distinguishing Self-Defence from other Conceptually Similar Defences

xiii xxiii 1 1 3 5 6

2. The Classification of Defences 2.1 Introduction 2.2 What is Meant by a Criminal Defence? 2.3 The Classification of Criminal Defences 2.4 The Treatment of Mistake 2.5 The Practical Consequences of the Competing Theories

13 13 13 16 23 37

3. The Justification of Self-Defence 3.1 Introduction 3.2 Consequentialist Approaches 3.3 The Personal Partiality Approach 3.4 Appeals to the Doctrine of Double Effect 3.5 Rights Approaches

43 43 45 50 53 54

4. Retreat 4.1 Introduction 4.2 How Might the Law of Self-Defence Deal with Retreat? 4.3 The Morally Preferable Option 4.4 Retreat from the Home: A Special Case?

69 69 69 76 83

5. Imminence of Harm 5.1 Introduction 5.2 A Formal Imminence Requirement 5.3 Arguments against the Imminence Requirement: The Case of Battered Women 5.4 Alternatives to Imminence

87 87 88 89 93

Contents

xii

5.5 Should the Imminence Requirement be Replaced by a Necessity Test? 5.6 Would a Battered Woman Meet an Inevitability of Harm Test?

99 102

6. Self-Generated Self-Defence 6.1 Introduction 6.2 Categorization of Self-Generated Self-Defence 6.3 What are the Legal Options? 6.4 Deliberate Generation of a Self-Defence Situation 6.5 Starting the Trouble 6.6 Agreeing to Fight 6.7 Failing to Avoid Trouble

109 109 110 111 113 114 123 125

7. Killing to Protect Property 7.1 Introduction 7.2 Approaches to the Defence of Property 7.3 Is Killing to Protect Property ever Justified? 7.4 Is there Something Different about Defence of Property in the Home?

131 131 132 134 137

8. Killing to Prevent Rape 8.1 Introduction 8.2 Killing to Prevent Rape in the Law 8.3 The Principled Argument

143 143 143 146

9. Mistake 9.1 Introduction 9.2 Mistake about an Attack 9.3 Mistake about the Amount of Force Necessary to Repel an Attack

159 159 160 169

10. The Impact of the European Convention on Human Rights 10.1 Introduction 10.2 Article 2 of the ECHR 10.3 Defence of Property and Article 2 10.4 Mistake about the Need to Use Self-Defensive Force and Article 2 10.5 State Officials and Private Citizens

177 177 177 181

Bibliography

197

Index

211

183 190

Table of Cases E N G L A N D A N D WA L E S Albert v Lavin [1981] 2 WLR 1070 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .162 Attorney-General’s Reference (No 6 of 1980) [1981] QB 715 . . . . .111, 124 Beatty v Gilbanks (1882) 9 QBD 308 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .128 Collins v Wilcock [1984] 1 WLR 1172 . . . . . . . . . . . . . . . . . . . . . . . . . . . .164 DPP v Armstrong-Braun [1999] Crim LR 416 . . . . . . . . . . . . . . . . . . . . . .170 DPP v Bayer [2003] EWHC Admin 2567 . . . . . . . . . . . . . . . . . . . . .132, 133 DPP v Majewski [1977] AC 443 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .120 Duncan v Jones [1936] 1 KB 218 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .128 Evans v Hughes [1972] 1 WLR 1452 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 Jones and Milling v Gloucestershire Crown Prosecution Service [2004] EWCA Crim 1981 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 Mason’s case (1756) Foster 132 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 NHS Trust A v M [2001] 2 WLR 942 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .191 Percy v DPP [1995] 1 WLR 1382 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .128 R (Amin) v Secretary of State for the Home Department [2003] UKHL 51 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .178 R (Burke) v General Medical Council [2005] EWCA Civ 1003 . . . . . . . .191 R (Gezer) v Secretary of State for the Home Department [2004] EWCA Civ 1730 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .191 R (Middleton) v West Somerset Coroner and Another [2004] UKHL 10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .178 R (Rowley) v DPP [2003] EWHC 693 . . . . . . . . . . . . . . . . . . . . . . . . . . . .178 R (Sharman) v Her Majesty’s Coroner for Inner North London [2005] EWHC 857 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .169 R (Stanley) v Her Majesty’s Coroner for Inner North London [2003] EWHC 1180 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .169 R (Wright) v Secretary of State for the Home Department [2001] EWHC Admin 520 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .178 R v Abdul-Hussain [1999] Crim LR 570 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8 R v Adomako [1995] 1 AC 171 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .194 R v Ahluwalia [1992] 4 All ER 889 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .90 R v Asbury [1986] Crim LR 258 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .163 R v Beckford [1988] AC 130 . . . . . . . . . . . . . . . . . . . . . . . . . . . .133, 163, 187 R v Bird [1985] 1 WLR 816 . . . . . . . . . . . . . . . . . . . . . . . . . . . .71, 72, 73, 116 R v Cairns [2000] RTR 15 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 R v Chisam (1963) 47 Cr App R 130 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .162

xiv

Table of Cases

R v Clegg [1995] 1 AC 482 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .172 R v Clugstone The Times, 1 October 1987 . . . . . . . . . . . . . . . . . . . . . . . . .144 R v Dadson (1850) 4 Cox CC 358 . . . . . . . . . . . . . . . . . . . . . . . . . . .31, 36, 38 R v Dakin (1828) 1 Lewin 166; 168 ER 999 . . . . . . . . . . . . . . . . . . . . . .70, 72 R v DPP ex p Manning [2001] QB 330 . . . . . . . . . . . . . . . . . . . . . . . . . . . .178 R v Dudley and Stephens (1884) 14 QBD 27 . . . . . . . . . . . . . . . . . . . . .1, 7, 8 R v Edwards [1973] AC 648 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .120 R v Elkington (21 January 1993, LEXIS transcript) . . . . . . . . . . . . . . . . . .187 R v Fennell [1971] 1 QB 428 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .162 R v Field [1972] Crim LR 435 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .128 R v Fisher [1987] Crim LR 334 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .163 R v G [2003] UKHL 50 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .161 R v Hasan [2005] UKHL 22 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6, 11, 120 R v Hastings [2003] EWCA Crim 3730 . . . . . . . . . . . . . . . . . . . . . . . . . . . .138 R v Hatton [2005] EWCA Crim 2951 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .170 R v HM Coroner for East Kent ex p Spooner (1987) 88 Cr App R 10 . . . . .7 R v Howe [1987] AC 417 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 R v Hughes [1995] Crim LR 956 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .170 R v Humphreys [1995] 4 All ER 1008 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .90 R v Hussey (1924) 18 Cr App R 160 . . . . . . . . . . . . . . . . . . . . .133, 137, 138 R v Jackson [1985] Crim LR 674 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .163 R v James [2006] EWCA Crim 14 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .105 R v Johnson [1989] Crim LR 738 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .120 R v Julien [1969] 1 WLR 839 . . . . . . . . . . . . . . . . . . . . . . . . . . .70, 71, 73, 76 R v Kimber [1983] 1 WLR 1118 . . . . . . . . . . . . . . . . . . . . . . . . .162, 163, 164 R v Martin [2001] EWCA Crim 2245 . . . . . . . . .82, 133, 144, 163, 170, 172 R v Mawgridge (1707) Kel 119 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .115 R v McInnes [1971] 1 WLR 1600 . . . . . . . . . . . . . . . . . . . . . . . .116, 132, 175 R v Morgan [1976] AC 182 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .162, 166 R v Morpeth Ward Justices ex p Ward (1992) 95 Cr App R 215 . . . . . . .128 R v Morse (1910) 4 Cr App R 50 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .116 R v Nicol and Selvanayagam [1996] Crim LR 318 . . . . . . . . . . . . . . . . . . .128 R v O’Grady [1987] QB 995 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .170 R v Oatridge (1992) 94 Cr App R 367 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .163 R v Odgers (1843) 2 N & Rob 479; 174 ER 355 . . . . . . . . . . . . . . . . . .70, 72 R v Owino [1996] 2 Cr App R 128 . . . . . . . . . . . . . . . . . . . . . . . . . . .133, 170 R v P&O European Ferries (Dover) Ltd (1990) 93 Cr App R 72 . . . . . . .7, 8 R v Porritt [1961] 1 WLR 1372 . . . . . . . . . . . . . . . . . . . . .133, 137, 138, 162 R v Rashford [2005] EWCA Crim 3377 . . . . . . . . . . . . . . . . . . . . . . .116, 117 R v Rose (1884) 15 Cox CC 540 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .162 R v Safi [2003] EWCA Crim 1809 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 R v Scarlett [1993] 4 All ER 629 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .170, 172 R v Semini [1949] 1 KB 405 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75 R v Smith (1837) 8 Car & P 158 160; 173 ER 441 . . . . . . . . . . . .70, 72, 162 R v Smith [2001] 1 AC 146 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .105

Table of Cases

xv

R v Thornton (No 2) [1996] 1 WLR 1174 . . . . . . . . . . . . . . . . . . . . . . . . . . .90 R v Weston (1879) 14 Cox CC 346 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .162 R v Wheeler [1967] 1 WLR 1531 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .143 R v Whyte [1987] 3 All ER 416 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72 R v Williams [1987] 3 All ER 411 . . . . . . . . . . . . . . . . . . . . . . . . . . . .162, 163 Re: A (Children) [2001] 2 WLR 480 . . . . . . . . . . . . . . . . . . . .6, 7, 8, 9, 54, 96 Redmond-Bate v DPP [2000] HRLR 249 . . . . . . . . . . . . . . . . . . . . . . . . . .128 Reed v Wastie [1972] Crim LR 221 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .169 Semayne’s case (1604) 5 Co Rep 91a . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .85 SCOTLAND Bennett v Criminal Injuries Compensation Appeal Panel (9 December 2003, Court of Session) . . . . . . . . . . . . . . . . . . . . . . . . . . .173 Boyle v HM Advocate 1993 SLT 577 . . . . . . . . . . . . . . . . . . . . . . . . .109, 117 Brennan v HM Advocate 1977 JC 38 . . . . . . . . . . . . . . . . . . . . . . . . . .20, 120 Burns v HM Advocate 1995 JC 154 . . . . . . . . . . . . . . . . . . . . . .109, 116, 117 Cochrane v HM Advocate 2001 SCCR 655 . . . . . . . . . . . . . . . . . . . . . . . . .14 Crawford v HM Advocate 1950 JC 67 . . . . . . . . . . .109, 145, 161, 169, 173 Dawson v Dickson 1999 JC 315 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26, 38 Deakin v Milne (1882) 10 R (J) 22 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .128 Drury v HM Advocate 2001 SLT 1013 . . . . . . . . . . . .14, 161, 163, 164, 173 Elliott v HM Advocate 1987 JC 47 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .132 Fenning v HM Advocate 1985 JC 76 . . . . . . . . . . . . . . . . . . . . . . .71, 88, 173 Finegan v Heywood 2000 JC 444 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .120 Fitzpatrick v HM Advocate 1992 SLT 796 . . . . . . . . . . . . . . . . . . . . . . . . .109 Friel v HM Advocate 1998 SLT 1327 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88 Galbraith v HM Advocate (No 2) 2002 JC 1 . . . . . . . . . . . . . . . . .14, 90, 105 Garvock v HM Advocate 1991 SCCR 593 . . . . . . . . . . . . . . . . . . . . . . . . .148 Gillingham v HM Advocate The Scotsman, 24 April 2001 . . . . . . . . . . . .132 Hendry v HM Advocate 1985 JC 105 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88 Hillan v HM Advocate 1937 JC 53 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .173 HM Advocate v Doherty 1954 JC 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88 HM Advocate v Forrest (1837) 1 Swin 404 . . . . . . . . . . . . . . . . . . . . . . . . .145 HM Advocate v Kizileviczius 1938 JC 60 . . . . . . . . . . . . . . . . .109, 117, 173 HM Advocate v McGlone 1955 SLT 79 . . . . . . . . . . . . . . . . . . . . . . . . . . .109 HM Advocate v Wilson 1942 JC 75 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20 Jones v HM Advocate 1990 JC 160 . . . . . . . . . . . . . . . . . . . . . . . . . . .109, 161 Lord Advocate’s Reference (No 1 of 2001) 2002 SLT 466 . . . . . . . .145, 148 Lord Advocate’s Reference (No 1 of 2000) 2001 JC 143 . . . . . . . . . . . . . . .14 Low v HM Advocate 1994 SLT 277 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .173 McBrearty v HM Advocate 1999 SLT 1333 . . . . . . . . . . . . . . . . . . . . . . . . .71 McCluskey v HM Advocate 1959 JC 39 . . . . . . . . . . . . . . . .71, 88, 131, 132 McLean v Jessop 1989 SCCR 13 . . . . . . . . . . . . . . . . . . . . . . . . .145, 150, 161 Miller v Douglas 1988 SCCR 565 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4

xvi

Table of Cases

Moss v Howdle 1997 JC 123 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 Owens v HM Advocate 1946 JC 119 . . . . . . . . . . . . . . . . . . . . . . . . . .88, 161 Pollock v HM Advocate 1998 SLT 880 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71 Robertson and Donoghue v HM Advocate (High Court, October 1945) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .110, 117 Ross v HM Advocate 1991 JC 210 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .120 Smart v HM Advocate 1975 JC 30 . . . . . . . . . . . . . . . . . . . . . . . . . . .111, 124 Surman v HM Advocate 1988 SLT 371 . . . . . . . . . . . . . . . . . . . . . . . . . . . .109 N O RT H E R N I R E L A N D Attorney General for Northern Ireland’s Reference (No 1 of 1975) [1977] AC 105 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .95, 132 Devlin v Armstrong [1971] NI 13 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4, 132 DPP for Northern Ireland v Lynch [1975] AC 653 . . . . . . . . . . . . . . . . . . .14 Kelly v Ministry of Defence [1989] NI 341 . . . . . . . . . . . . . . . . . . . . . . . . .100 Lynch v Ministry of Defence [1983] NI 216 . . . . . . . . . . . . . . . . . . . .184, 185 R v Browne [1973] NI 96 . . . . . . . . . . . . . . . . . . . . . .114, 115, 116, 117, 121 R v Fegan [1972] NI 80 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 R v Fitzpatrick [1977] NI 20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .109, 121 R v Thain [1985] NI 457 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26, 38, 184, 185 PRIVY COUNCIL Attorney-General for Jersey v Holley [2005] UKPC 23 . . . . . . . . . . . . . . .105 Beckford v R [1988] AC 130 . . . . . . . . . . . . . . . . . . . . . . . . . . . .133, 163, 187 R v Palmer [1971] AC 814 . . . . . . . . . . . . . . . . . . . . . .96, 169, 172, 173, 175 Shaw v R [2001] UKPC 26 . . . . . . . . . . . . . . . . . . . . . . . . . .96, 131, 163, 170 CANADA Northwest v R (1980) 22 AR 522 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .74 R v Baxter (1975) 27 CCC (2d) 96 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .132 R v Charlebois [2000] 2 SCR 674 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97 R v Cinous [2002] 2 SCR 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97, 161 R v Clark (1983) 3 CCC (3d) 264 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .132 R v Faid [1983] 1 SCR 265 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .169, 172 R v Gunning [2005] 1 SCR 627 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .132 R v Kerr [2004] 2 SCR 371 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4, 14, 164 R v Latimer [2001] 1 SCR 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 R v Lavallee [1990] 1 SCR 852 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .96, 98 R v Mack [1988] 2 SCR 903 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23 R v McCraw [1991] 3 SCR 72 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .146 R v McIntosh [1995] 1 SCR 686 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .74, 117 R v Parent [2001] 1 SCR 761 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .14, 164

Table of Cases

xvii

R v Pétel [1994] 1 SCR 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .96, 97, 146, 161 R v Proulx (1998) 127 CCC (3d) 511 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .74 R v Reilly [1984] 2 SCR 396 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .161, 172 R v Ruzic [2001] 1 SCR 687 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97 R v Ward (1978) 4 CR (3d) 190 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .74 AUSTRALIA R v Howe (1958) 100 CLR 448 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .73, 173 R v Lane (1983) 8 A Crim R 182 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .146 R v Lawson and Forsythe (1985) 18 A Crim R 360 . . . . . . . . . . . . . . . . . .122 R v McKay [1957] ALR 648 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .173 R v Secretary (1996) 131 FLR 126 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98 Viro v R (1978) 141 CLR 88 . . . . . . . . . . . . . . . . . . .131, 173, 174, 175, 176 Zecevic v DPP (1987) 162 CLR 645 . . . . . . . . . . . . . . . . . . . . .73, 117, 146, 161, 169, 173, 175, 176 SOUTH AFRICA Ex p die Minister van Justisie: In re S Van Wyk 1967 (1) SA 488 (AD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .133, 134, 135 Ex p The Minister for Safety and Security and the National Commissioner of the South African Police Service: In re The State v Walters and Walters (2002) 7 BCLR 663 . . . . . . . . . . . . . .134 S v Mogohlwane 1982 (2) SA 587 (T) . . . . . . . . . . . . . . . . . . . . . . . . . . . . .136 U S S U P R E M E C O U RT Allen v United States 164 US 492 (1897) . . . . . . . . . . . . . . . . . . . . . . . . . . . .75 Beard v United States 158 US 550 (1895) . . . . . . . . . . . . . . . . . . . . . . . . . . .75 Brown v United States 256 US 335 (1921) . . . . . . . . . . . . . . . . .73, 74, 76, 80 Coker v Georgia 433 US 584 (1977) . . . . . . . . . . . . . . . . . . . . . . . . . .155, 156 ALABAMA Cooke v State 93 So 86 (1921) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78 ALASKA Castillo v State 614 Ala 2d 756 (1980) . . . . . . . . . . . . . . . . . . . . . . . . . . . .123 CALIFORNIA People v Aris 215 Cal App 3d 1178 (1989) . . . . . . . . . . . . . . . . . . . . .94, 107 People v Button 39 P 1073 (1895) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .123 People v Gleghorn 193 Cal App 3d 196 (1987) . . . . . . . . . . . . . . . . . . . . .123 People v McIlvain 130 P 2d 131 (1942) . . . . . . . . . . . . . . . . . . . . . . . . . . . .153

xviii

Table of Cases CONNECTICUT

State v Diggs 592 A 2d 949 (1991) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .123 FLORIDA Weiand v State 732 So 2d 1044 (1999) . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75 INDIANA Runyan v State 57 Ind 80 (1877) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75, 77 Whipple v State 523 NE 2d 1363 (1988) . . . . . . . . . . . . . . . . . . . . . . . . .88, 92 KANSAS State v Hundley 693 P 2d 475 (1985) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87 State v Simon 646 P 2d 119 (1982) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .167 State v Stewart 763 P 2d 572 (1988) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .95 State v Wiggins 808 P 2d 1383 (1991) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .99 M A RY L A N D Barton v State 420 A 2d 1009 (1980) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .85 MASSACHUSETTS Commonwealth v Bennett 671 NE 2d 966 (1996) . . . . . . . . . . . . . . . . . . . .83 Commonwealth v Martin 341 NE 2d 885 (1976) . . . . . . . . . . . . . . . . . . .160 MICHIGAN People v Landrum 456 NW 2d 10 (1990) . . . . . . . . . . . . . . . . . . . . . . . . . .143 People v Landrum 407 NW 2d 614 (1986) . . . . . . . . . . . . . . . . . . . . . . . . .153 People v Riddle 649 NW 2d 30 (2002) . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76 MISSOURI State v Bartlett 71 SW 148 (1902) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78 State v Martin 666 SW 2d 895 (1984) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .99 State v Muhammad 757 SW 2d 641 (1988) . . . . . . . . . . . . . . . . . . . . . . . .123 NEBRASKA State v Schroeder 261 NW 2d 759 (1978) . . . . . . . . . . . . . . . . . . . . . . . .92, 99 NEW JERSEY State v Bonano 284 A 2d 345 (1971) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 State v Garland 694 A 2d 564 (1997) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84

Table of Cases

xix

NEW YORK People v Childs 250 NYS 2d 926 (1964) . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 People v Goetz 497 NE 2d 41 (1986) . . . . . . . . . . . . . . . . . . . . . . . . . . . . .167 People v Hernandez 744 NE 2d 198 (2002) . . . . . . . . . . . . . . . . . . . . . . . . .83 People v Melendez 79 NY 2d 861 (1992) . . . . . . . . . . . . . . . . . . . . . . . . . .160 People v Tomlins 107 NE 496 (1914) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .85 People v Young 183 NE 2d 319 (1962) . . . . . . . . . . . . . . . . . . . . . . . . .27, 160 N O RT H C A R O L I N A State v Blue 565 SE 2d 133 (2002) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 State v Norman 378 SE 2d 8 (1989) . . . . . . . . . . . . . . . . . .90, 92, 94, 95, 107 State v Norman 366 SE 2d 586 (1988) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .90 OHIO Erwin v State 29 Oh S 186 (1876) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75, 77 State v Wenger 390 NE 2d 801 (1979) . . . . . . . . . . . . . . . . . . . . . . . . . . . . .160 Stoffer v State 15 Oh S 47 (1864) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .123 OKLAHOMA Bechtel v State 840 P 2d 1 (1992) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98 Fowler v State 126 P 831 (1912) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78 OREGON State v Conally 3 Or 69 (1869) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76 P E N N S Y LVA N I A Commonwealth v Daniels 451 Pa 163 (1973) . . . . . . . . . . . . . . . . . . . . . . .83 Commonwealth v Drum 58 Pa 9 (1868) . . . . . . . . . . . . . . . . . . . . . . . . . . . .76 SOUTH CAROLINA Robinson v State 417 SE 2d 88 (1992) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98 TEXAS Lane v State 957 SW 2d 584 (1997) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .92 VERMONT State v Patterson 45 Vt 308 (1873) . . . . . . . . . . . . . . . . . . . . . . . . . . .138, 140 VIRGINIA Commonwealth v Sands 553 SE 2d 733 (2001) . . . . . . . . . . . . . . . . . . . . . .92

xx

Table of Cases WA S H I N G T O N

State v Barr 39 P 1080 (1895) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .139 WYOMING Jahnke v State 682 P 2d 991 (1984) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .94 Palmer v State 59 P 793 (1900) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84 State v Bristol 84 P 2d 757 (1938) . . . . . . . . . . . . . . . . . . .111, 125, 127, 128 E U R O P E A N C O U RT O F H U M A N R I G H T S A v United Kingdom (1999) 27 EHRR 611 . . . . . . . . . . . . . . . . . . . .192, 193 Aktas v Turkey (2004) 38 EHRR 13 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .181 Andronicou and Constantinou v Cyprus (1998) 25 EHRR 491 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .180, 181, 185, 186, 188 Aytekin v Turvey (2001) 32 EHRR 22 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .189 Brady v United Kingdom (App no 55151/00) ECHR, 3 April 2001 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .189 Bubbins v United Kingdom (2005) 41 EHRR 24 . . . . . . . . . . . . . . . . . . . .190 Çacici v Turkey (2001) 31 EHRR 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .191 Calvelli and Ciglio v Italy (App no 32967/96) ECHR, 17 January 2002 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .192 Caraher v United Kingdom (App no 24520/94) ECHR, 11 January 2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .186, 187, 189 Edwards v United Kingdom (2002) 35 EHRR 19 . . . . . . . . . . . . . . . . . . .191 Finucane v United Kingdom (2003) 37 EHRR 29 . . . . . . . . . . . . . . . . . . .178 Gül v Turkey (2002) 34 EHRR 28 . . . . . . . . . . . . . . . . . . .181, 188, 190, 194 HLR v France (1998) 26 EHRR 29 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .191 Isayeva v Russia (2005) 41 EHRR 38 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .178 Jordan v United Kingdom (2003) 37 EHRR 2 . . . . . . . . . . . . . . . . . . . . . .178 Kakoulli v Turkey (App no 38595/97) ECHR, 22 November 2005 . . . . .182 Keenan v United Kingdom (2001) 33 EHRR 38 . . . . . . . . . . . . . . . . . . . . .191 Kelly v United Kingdom (2001) 11 BHRC 1 . . . . . . . . . . . . . . . . . . . . . . . .189 Kelly v United Kingdom (1993) 16 EHRR CD20 . . . . . . .100, 183, 184, 190 Kiliç v Turkey (2001) 33 EHRR 58 . . . . . . . . . . . . . . . . . . . . . . . . . . .180, 191 LCB v United Kingdom (1999) 27 EHRR 212 . . . . . . . . . . . . . . . . . . . . . .180 McCann v United Kingdom (1996) 21 EHRR 97 . . . . . . . . . . . . .178, 181, 183, 184, 185, 187, 188, 189, 190 McKerr v United Kingdom (2002) 24 EHRR 20 . . . . . . . . . . . .178, 181, 189 McShane v United Kingdom (2002) 35 EHRR 23 . . . . . . . . . . . . . . .181, 189 Makaratatzis v Greece (2005) 41 EHRR 49 . . . . . . . . . . .180, 181, 189, 190 Nachova v Bulgaria (2004) 39 EHRR 37 . . . . . . . . . . . . . . . . . .170, 181, 182 Oláh v Hungary (App no 56558/00) ECHR, 14 September 2004 . . . . . .189 Öneryildiz v Turkey (2005) 41 EHRR 20 . . . . . . . . . . . . . . . . .191, 192, 193

Table of Cases

xxi

Öneryildiz v Turkey (2004) 39 EHRR 12 . . . . . . . . . . . . . . . . . . . . . . . . . .191 Osman v United Kingdom (2000) 29 EHRR 245 . . . . . . . . . . .180, 191, 193 Ramsahai v The Netherlands (App no 52391/99) ECHR, 10 November 2005 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .182 Ribisch v Austria (1995) 21 EHRR 573 . . . . . . . . . . . . . . . . . . . . . . . . . . .178 Sabuktekin v Turkey (2003) 36 EHRR 19 . . . . . . . . . . . . . . . . . . . . . . . . . .181 Salman v Turkey (2002) 34 EHRR 17 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .191 Steel and others v United Kingdom (1999) 28 EHRR 603 . . . . . . . . . . . .129 Stewart v United Kingdom (1985) 39 DR 162 . . . . . . . . . . . . . . . . . .180, 181 Stubbings v United Kingdom (1997) 23 EHRR 213 . . . . . . . . . . . . . . . . .191 VO v France (2005) 40 EHRR 12 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .192 X v Belgium (1969) 12 Yearbook 174 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .180 X and Y v Netherlands (1986) 8 EHRR 235 . . . . . . . . . . . . . . . . . . .192, 193

This page intentionally left blank

Table of Legislation AUSTRALIA Crimes Act 1900 (New South Wales) s 420 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .133 s 421 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .173 Crimes (Amendment) Act 1986 (Victoria) s 8 . . . . . . . . . . . . . . . . . . . . . . .174 Crimes Amendment (Self-Defence) Act 2001 (New South Wales) . . . . . .173 Crimes (Homicide) Amendment Act 1982 (New South Wales) . . . . . . . .174 Criminal Law Consolidation Act 1935 (South Australia) s 15(2) . . . . . . .173 Criminal Law Consolidation (Self Defence) Amendment Act 1997 (South Australia) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .173 CANADA Criminal Code s 34(2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .74, 146 s 35 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .117, 118 s 35(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .117 s 35(c) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .74 ss 271–3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .146, 148 s 273(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .146 s 264.1(1)(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .146 E N G L A N D A N D WA L E S Criminal Law Act 1967 s 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .132 Domestic Violence, Crime and Victims Act 2004 . . . . . . . . . . . . . . . . . . . .106 Homicide Act 1957 s 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .105 Murder (Abolition of Death Penalty) Act 1965 s 1(1) . . . . . . . . . . . . . . . .174 Offences against the Person Act 1828 s 10 . . . . . . . . . . . . . . . . . . . . . . . . . .75 Sexual Offences Act 2003 s 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .148 E U R O P E A N / I N T E R N AT I O N A L C O N V E N T I O N S European Convention on Human Rights Article 2 . . . . . . . . . . . . . . . . . . . . . . . . . . .3, 57, 133, 166, 177–183, 185–195 Article 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .177, 178, 181, 192 Article 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .129 Article 10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .129 International Covenant on Civil and Political Rights Article 6(2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57

xxiv

Table of Legislation GERMANY

German Penal Code s 32 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76, 79 NEW ZEALAND Crimes Act 1961 ss 48 and 62 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .173 Crimes Amendment Act 1980 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .173 N O RT H E R N I R E L A N D Criminal Law (Northern Ireland) Act 1967 s 3 . . . . . . . . . . . . . . . . . . . . . .100 SCOTLAND Criminal Procedure (Scotland) Act 1995 s 205(1) . . . . . . . . . . . . . . . . . . .174 U N I T E D S TAT E S O F A M E R I C A Model Penal Code § 3.04 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .163 § 3.04(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34, 97 § 3.04(2)(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .77, 113, 145 § 3.04(2)(b)(ii) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72, 75 § 3.04(2)(b)(ii)(A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72, 83 § 3.06(3)(d) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .133, 141 § 3.06(3)(d)(i) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .133 § 3.09 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .163 § 3.11(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34 Alabama Ala Code § 13A-3-23(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88, 161 Ala Code § 13A-3-23(a)(2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .145 Ala Code § 13A-3-23(b)(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72, 83 Ala Code § 13A-3-23(c)(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 Ala Code § 13A-3-23(c)(2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .118 Ala Code § 13A-3-23(c)(3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .124 Alaska Alaska Stat § 11.81.330(a)(2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 Alaska Stat § 11.81.330(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .118 Alaska Stat § 11.81.335(a)(2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .145 Alaska Stat § 11.81.335(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 Alaska Stat § 11.81.350 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .137 Arizona Ariz Rev Stat § 13-404(A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97 Ariz Rev Stat § 13-404(B)(3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .118

Table of Legislation

xxv

Arkansas Ark Code § 5-2-606(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88, 161 Ark Code § 5-2-606(b)(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 Ark Code § 5-2-606(b)(2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .118 Ark Code § 5-2-606(b)(3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .124 Ark Code § 5-2-607 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 Ark Code § 5-2-607(a)(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .145 Ark Code § 5-2-607(b)(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72 Ark Code § 5-2-608(b)(2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .137 California Cal Penal Code § 198 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .161 Cal Penal Code § 776.041(2)(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .118 Colorado Colo Rev Stat § 18-1-704(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88, 161 Colo Rev Stat § 18-1-704(2)(c) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .145 Colo Rev Stat § 18-1-704(3)(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 Colo Rev Stat § 18-1-704(3)(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .118 Colo Rev Stat § 18-1-704(3)(c) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .124 Colo Rev Stat § 18-1-704.5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .137 Delaware Del Code Ann tit 11 § 464(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97, 163 Del Code Ann tit 11 § 464(c) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .145 Del Code Ann tit 11 § 464(e)(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 Del Code Ann tit 11 § 464(e)(2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72 Del Code Ann tit 11 § 464(e)(2)(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 Florida Fla Stat § 776.012 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .161 Fla Stat § 776.012(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .145 Fla Stat § 776.013(1)(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .140 Georgia Ga Code Ann § 16-3-21(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88, 145 Ga Code Ann § 16-3-21(b)(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 Ga Code Ann § 16-3-21(b)(3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 Guam 9 Guam Code Ann § 7.84 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97 9 Guam Code Ann § 7.86(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .145 9 Guam Code Ann § 7.86(b)(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 9 Guam Code Ann § 7.86(b)(2)(A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83

xxvi

Table of Legislation

Illinois 720 ICLS § 5/7-1(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .145, 161 720 ICLS § 5/7-4(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 720 ICLS § 5/7-4(c)(2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 Iowa Iowa Code § 704.6(2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 Iowa Code § 704.6(3)(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .118 Iowa Code § 704.6(3)(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 Kansas Kan Stat Ann § 21-3214(2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 Kan Stat Ann § 21-3214(3)(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .118 Kan Stat Ann § 21-3214(3)(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .124 Kentucky Ky Rev Stat Ann § 503.120(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .163 Louisiana La Rev Stat § 14:20(4)(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84, 141 Maine Me Rev Stat Ann tit 17-A § 104 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .137 Me Rev Stat Ann tit 17-A § 108(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88 Me Rev Stat Ann tit 17-A § 108(2)(A)(2) . . . . . . . . . . . . . . . . . . . . . . . . . .145 Me Rev Stat Ann tit 17-A § 108(2)(C)(1) . . . . . . . . . . . . . . . . . . . . . . . . . .113 Me Rev Stat Ann tit 17-A § 108(2)(C)(3)(a) . . . . . . . . . . . . . . . . . . . . . .72, 83 Mississippi Miss Code Ann § 97-3-15(f) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .161 Missouri Mo Rev Stat § 563.031(1)(1)(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 Mo Rev Stat § 563.031(2)(2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .145 Montana Mont Code Ann § 45-3-102 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88, 146 Mont Code Ann § 45-3-105(2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 Mont Code Ann § 45-3-105(2)(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 Nebraska Neb Rev Stat § 28-1409(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97 Neb Rev Stat § 28-1409(4) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .145 Neb Rev Stat § 28-1409(4)(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113

Table of Legislation

xxvii

Neb Rev Stat § 28-1409(4)(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72 Neb Rev Stat § 28-1409(4)(b)(i) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 New Hampshire NH Rev Stat Ann § 627.4(I)(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 NH Rev Stat Ann § 627.4(I)(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 NH Rev Stat Ann § 627.4(I)(c) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .124 NH Rev Stat Ann § 627.4(II)(c) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .145 NH Rev Stat Ann § 627.4(III)(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 New York NY UCC Law § 35.15(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88, 161 NY UCC Law § 35.15(1)(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 NY UCC Law § 35.15(1)(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 NY UCC Law § 35.15(1)(c) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .124 NY UCC Law § 35.15(2)(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72 NY UCC Law § 35.15(2)(a)(i) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 NY UCC Law § 35.15(2)(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .145 NY UCC Law § 35.20(3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .137 North Dakota ND Cent Code § 12.1-05-03(2)(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 ND Cent Code § 12.1-05-03(2)(b) . . . . . . . . . . . . . . . . . . . . . . . . . . .118, 119 ND Cent Code § 12.1-05-07(2)(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .146 ND Cent Code § 12.1-05-07(2)(b)(2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 Oregon Or Rev Stat § 161.209 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88 Or Rev Stat § 161.215(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 Or Rev Stat § 161.215(2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 Or Rev Stat § 161.215(3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .124 Or Rev Stat § 161.219(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .146 Or Rev Stat § 161.225(2)(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .137 Pennsylvania 18 Pa Cons Stat § 505(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97 18 Pa Cons Stat § 505(b)(2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .145 18 Pa Cons Stat § 505(b)(2)(i) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72, 113 18 Pa Cons Stat § 505(b)(2)(ii) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 18 Pa Cons Stat § 507(c)(4)(ii) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .137 Tennessee Tenn Code Ann § 39-11-611(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75, 161 Tenn Code Ann § 39-11-611(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .140 Tenn Code Ann § 39-11-611(d) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119

xxviii

Table of Legislation

Texas Tex Penal Code § 9.31(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97 Tex Penal Code § 9.31(b)(4)(A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 Tex Penal Code § 9.32(a)(3)(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .145 Tex Penal Code § 9.32(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 Utah Utah Code § 76-2-402(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88, 146 Utah Code § 76-2-402(2)(a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 Utah Code § 76-2-402(2)(c)(i) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 Utah Code § 76-2-402(3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 Vermont Vt Stat Ann tit 13 § 2305(2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .145 Wisconsin Wis Stat § 939.48(2)(b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 Wis Stat § 939.48(2)(c) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 Conventions American Convention on Human Rights Article 4(2) . . . . . . . . . . . . . . . . .57 Virginia Declaration of Rights s 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .56

1 Introduction 1.1 Why Self-Defence? Of all the criminal defences, self-defence is one of the most well established. In English law, which has informed the law of many other common law jurisdictions, the availability of the defence can be traced back to at least the 1200s.1 In Scotland, the availability of the plea, even in cases of homicide, can be traced back at least to the thirteenth century.2 Compared to the closely related defences of necessity and duress, selfdefence is relatively uncontroversial. It is universally accepted in common law jurisdictions as a complete defence to all crimes of violence up to and including murder3 and has not been subject to calls for its restriction or abolition in the same way as has been the case for necessity and duress.4 One might assume on this basis that self-defence is an unproblematic defence. Certainly until recently this would have been the obvious conclusion to draw from a perusal of the literature. Even now, it is rarely questioned in criminal law texts that the victim of an attack should be permitted to kill in order to save her own life at the expense of the aggressor’s life. But on closer examination, this conclusion is less obvious. A killing undertaken in self-defence is a self-interested killing and it can be an intentional killing.5 It is precisely the type of act that, on the face of it, the criminal law might wish to prohibit. Nevertheless, if one subscribes to the view that the life of an aggressor who threatens the life of another human being is worth less than that of the victim, justifying self-defence is relatively 1 Robert de Herthale (1203), noted in B Brown, ‘Self-defence in homicide from strict liability to complete exculpation’ [1958] Crim LR 583–90, at 584. For other examples, see TA Green, ‘The jury and the English law of homicide 1200–1600’ (1976) Michigan Law Review 414–99; ND Hurnard, The King’s Pardon for Homicide before AD 1307 (1969). 2 One Richard, accused of the murder of Adam the Miller, noted in TM Cooper, Select Scottish Cases of the Thirteenth Century (1944), at 58. Cf the defence of necessity, which was not even formally recognized in Scots law until Moss v Howdle 1997 JC 123. 3 Cf R v Dudley and Stephens (1884) 14 QBD 273 and R v Latimer [2001] 1 SCR 3 (both ruling out necessity as a defence to murder); R v Howe [1987] AC 417 (ruling out duress likewise). 4 RF Schopp, Justification Defences and Just Convictions (1998), at 141 (arguing that duress should never operate as a complete defence); A Norrie, Crime, Reason and History: A Critical Introduction to Criminal Law (2nd edn, 2001), at 160 (arguing that tight restrictions should be placed on necessity to prevent its abuse). 5 Not all self-defensive killings are intentional, but there are at least some circumstances in which it would be absurd to conclude that the intention of the accused was not to kill.

2

Killing in Self-Defence

unproblematic.6 If, however, one holds that all human life is of equal value, explaining why the victim of an attack is permitted to kill the aggressor is far more difficult. It is particularly difficult to explain where the aggressor is, for reasons such as non-age or insanity, not responsible for her actions. Of course, it might be concluded from this that self-defence is not a justification defence at all; that killing in self-defence is not acceptable, but merely excused, on the basis that we do not blame the accused for preserving her own life. But this is not the view taken here, nor indeed is it the view taken in criminal law texts. Self-defence is commonly regarded as the archetypal justification defence.7 If we cannot explain why self-defence is justified, we might as well abandon the category of justification defences entirely. It is against this background that the defence of self-defence is subjected to detailed consideration. The first part of the book is concerned with finding an acceptable argument to justify the use of self-defensive force. This task is less difficult than it would have been a decade ago. There has been a recent revival of interest in the theory of self-defence8 although, as yet, no consensus as to its appropriate theoretical basis. Here, the defence is examined from a rights perspective, focussing in particular on the right to life, as this is considered to be the most productive route to establishing the permissibility of self-defensive killing. As such, it builds to a certain extent on the work done by Suzanne Uniacke, in Permissible Killing.9 It will be argued here that the right to life is a fundamental right, unlike any other in that deprivation of life cannot be remedied. As such, the right should be respected as far as is reasonably possible. The permissibility of self-defence derives from the right to life in that all human beings possess this right and therefore a right to protect themselves from an aggressor10 who threatens to deprive them of it. It is acceptable to kill an aggressor because the aggressor, in becoming an unjust immediate threat to the life of another that cannot be avoided by any reasonable means, temporarily forfeits her right to life, at least as long as these conditions remain in force. The use of the language of forfeiture in this context will, I am sure, be controversial, but it will be argued that this is the most appropriate way to ground the right of self-defence and is more convincing than others, such as that of Uniacke, who attempt to avoid the term by using the language of specification.11 6 Although even if one makes this assumption, explaining the permissibility of self-defensive killing in certain situations, such as those involving multiple aggressors, is still difficult. 7 A Ashworth, Principles of Criminal Law (5th edn, 2006), at 135; AP Simester and GR Sullivan, Criminal Law: Theory and Doctrine (2nd edn, 2004), at 29. 8 For many years, the debate was confined to the philosophy journals. More recently, it has started to reach the legal arena. See R Segev, ‘Fairness, responsibility and self-defense’ (2004–5) 45 Santa Clara Law Review 383–460; S Wallerstein, ‘Justifying the right to self-defense: a theory of forced consequences’ (2005) 91 Virginia Law Review 999–1035. 9 S Uniacke, Permissible Killing: The Self-Defence Justification of Homicide (1994). 10 Or a passive threat: see Section 1.3 below. 11 See Uniacke, Permissible Killing, at 213 and the discussion in Chapter 3.

Introduction

3

The second part of the book uses this rights-based understanding to establish principles to which the law of self-defence should conform. As such, it is as much about the right to life and the limits that this places, or should place, on the substantive law, as it is about self-defence. Indeed, the most interesting debates in self-defence law all relate to the value of human life and whether this can ever be outweighed by other values. This can been seen in relation to retreat (can the value of life ever be outweighed by other values such as human dignity, such that a victim of a potentially lethal attack can benefit from the defence of self-defence if she ignores a reasonable opportunity to retreat?); mistake (is the value of life such that unreasonably mistaken belief should not be permitted to ground a claim of self-defence?); self-generated self-defence (should there be a duty to avoid trouble and, most controversially, should the value of an aggressor’s life outweigh the civil liberties of an innocent person such that the innocent person should be required to avoid situations where she knows she is likely to face a potentially life-threatening attack?); and proportionality (should the value of human life—even that of an aggressor—always outweigh the values of property and sexual autonomy?). All of these issues are examined here through the lens of a right of self-defence based on the right to life. This book is based on, although by no means identical to, the author’s doctoral thesis,12 which analysed specifically the law of self-defence in England and Wales13 and Scotland. This narrow jurisdictional focus has been extended here and the chapters that follow do not focus on the law in any particular jurisdiction, but instead traverse the case law and legislation of all major common law jurisdictions in search of the optimum formulation of the law of self-defence. Any discussion of the right to life that at least touches on the Anglo-Scottish context would not be complete, however, without reference to the impact of the European Convention on Human Rights. It will be argued that the law of self-defence in England, in particular, may violate Article 2 of the Convention in failing to rule out the use of lethal force in protecting property and in allowing an honest but unreasonable belief in the need to use self-defensive force to ground the defence.

1.2 Defining the Subject Matter of the Book The subject matter of this book is specifically self-defence, as opposed to the defence of others or private defence more generally.14 The reason for the 12 F Leverick, A Critical Analysis of the Law of Self-defence in Scotland and England (2003), unpublished PhD thesis . 13 For the sake of brevity, from this point onwards references to the law of England should be taken to mean the law of England and Wales. 14 The term private defence has been proposed as an umbrella term to cover self-defence and the defence of others (see J Chalmers, CHW Gane and CN Stoddart, Gane and

4

Killing in Self-Defence

choice specifically of self-defence is because it is the most difficult type of defensive act to justify, given the element of self-interest involved.15 In recognizing the defence of self-defence, the law permits the accused to save her own life at the expense of the life of another. It is this that distinguishes self-defence from the defence of others and makes it a particularly interesting subject to study. This is not to say, however, that the defence of others has no role to play here. Much of the case law concerning the defence of others will be drawn upon in the course of the analysis. The book also focuses primarily on self-defensive killing: conduct resulting in the death of the aggressor.16 Once again, the reason for this is that killing in self-defence is the most difficult type of self-defence to justify, involving, as it does, the death of a human being, a state of affairs that is permanent and can never be remedied.17 For this reason, although it is almost certainly less common in practice than non-lethal self-defensive force, lethal self-defence is the most interesting type of self-defensive force. If self-defensive killing can be justified, then other, lesser, types of self-defensive force can surely be justified too. That is not to say that lesser acts of self-defence play no role in this text. As before, much of the case law involving non-lethal self-defence will be drawn upon in the course of the following chapters. As a final point, the focus of the book is on the use of self-defensive force, as distinct from acts preparatory to the use of self-defensive force. There have been cases where acts performed in advance of an anticipated attack have been termed acts of self-defence.18 An example might be obtaining a weapon,19 or the preparation of petrol bombs,20 with a view to using these to ward off an expected attack. In these circumstances, ‘self-defence’ operates as a defence to an offence such as the possession of a weapon.21 The primary focus of this book, though, is on self-defence as a defence to an immediate crime of violence. It is the actual use of (especially lethal) self-defensive force that is seen here as the most challenging and interesting type of self-defensive act to justify, involving, as it does, the infliction of harm on another human being. Thus acts such as the possession of a weapon with a view to using it in self-defence at some future time will not be discussed further here.22 Stoddart: A Casebook on Scottish Criminal Law (3rd edn, 2001), at para 7.04), but it has not been widely adopted. 15 That is not to say that acts in defence of others are never self-interested. A mother who kills an aggressor to save the life of her child is likely to be, to some degree, a self-interested killer. 16 The term ‘aggressor’ is defined in Section 1.3 below. 17 At least as far as the individual who is killed is concerned. On this, see Chapter 3. 18 Miller v Douglas 1988 SCCR 565 (in Scotland); Evans v Hughes [1972] 1 WLR 1452 (in England). 19 The facts of R v Fegan [1972] NI 80. 20 The facts of Devlin v Armstrong [1971] NI 13. 21 Although one might question whether self-defence is the appropriate term for such a defence, as the Canadian Supreme Court did in R v Kerr [2004] 2 SCR 371, where a prisoner armed himself with a knife in the expectation that he would be attacked. 22 For a detailed discussion, see D Lanham, ‘Offensive weapons and self-defence’ [2005] Crim LR 85–97.

Introduction

5

1.3 A Note about Terminology At the outset, it is necessary to say something about the terms ‘aggressor’, ‘passive threat’, and ‘bystander’, as they will be used throughout the book. They will be defined here in relation to the accused; that is the person who uses self-defensive force. An aggressor is someone who directly attacks the accused, or threatens to do so, using either physical force or a weapon. An example might be the person who lunges forward with a knife, or who punches the accused in the face. An aggressor may be a ‘culpable’ aggressor or an ‘innocent’ aggressor. A culpable aggressor is an aggressor who would be held legally responsible for her act of aggression. An innocent aggressor is an aggressor who would not be held legally responsible for her act of aggression. Most commonly, this category covers the aggressor who has a lack of capacity defence,23 such as insanity or non-age. It could also include the aggressor who has an excuse defence,24 such as duress or mistaken belief in self-defence.25 A passive threat is someone who directly threatens the life or physical integrity of the accused by means other than an attack. An example is the mountaineer roped to her companion who falls and threatens to pull the companion off the mountain.26 What both aggressors and passive threats have in common is that they pose a direct threat to the life (or physical integrity) of the accused. The important point for our purposes is that both aggressors and passive threats are distinguished from bystanders, who pose no direct threat whatsoever to the life or physical integrity of the accused. In the text, the term aggressor will be used as shorthand for an aggressor or a passive threat, unless otherwise stated. It is also important to distinguish between the terms ‘necessity’27 and ‘proportionality’ as they apply to the use of self-defensive force. Necessity refers to the need to use defensive force at all. In other words, if an attack could have been avoided without resorting to force (by, for example, taking an opportunity to escape the aggressor), then any force used in self-defence was not necessary. Proportionality refers to the degree of force that is permitted, once it has been established that it is necessary to use at least some force to avoid an attack. If an individual uses lethal force to avoid being slapped in the 23

Lack of capacity defences are defined in Chapter 2. The difference between lack of capacity defences and excuse defences is discussed in Chapter 2. 25 L Alexander, ‘A unified excuse of pre-emptive self-protection’ (1999) 74 Notre Dame Law Review 1475–505, at 1483. Chapter 2 defends the claim that mistaken belief in self-defence is 26 an excuse and not a justification. This example is discussed in Section 1.4.1 below. 27 The term necessity used in this context (ie as a condition that can be applied to the use of self-defensive force) should be distinguished from the specific defence of necessity, as discussed in Section 1.4.1 below. 24

6

Killing in Self-Defence

face, the force might have been necessary to avoid the slap (perhaps it was the only way of doing so) but it was not proportional. Finally, in a book covering different common law jurisdictions, there are always going to be difficulties of terminology. Where reference is made to the law of a particular jurisdiction, every effort has been made to use the term appropriate to that jurisdiction. Where the book is referring more generally to an issue, for the sake of consistency, ‘accused’ will be used in preference to ‘defendant’; ‘duress’ will be used in preference to ‘coercion’ or ‘compulsion’; and the more commonly used ‘manslaughter’ will be used in preference to the Scottish/South African term ‘culpable homicide’.

1.4 Distinguishing Self-Defence from other Conceptually Similar Defences Self-defence is part of a family of defences, which would also include necessity and duress, that all involve the commission of a crime in order to avert a threatened harm. Necessity is used here as a term that would encompass the English defence of duress of circumstances. English law treats duress of circumstances and duress by threats as two species of the same defence governed by the same rules, although the exact nature of the relationship between duress of circumstances and necessity is not entirely clear.28 Self-defence has much in common conceptually with necessity and duress, but it can (and should)29 be distinguished from both.

1.4.1 Distinguishing self-defence from necessity Self-defence can be distinguished from necessity (or indeed the English defence of duress of circumstances) on the basis that in the case of selfdefence the act of the accused is directed towards someone who poses a direct threat to her life or physical integrity, whether as an aggressor or passive threat.30 Thus the defining feature of a claim of self-defence is that it 28 Some cases have used the terms ‘necessity’ and ‘duress of circumstances’ interchangeably (Jones and Milling v Gloucestershire Crown Prosecution Service [2004] EWCA Crim 1981, at [48]–[55]; R v Hasan [2005] UKHL 22, per Lord Bingham at [22]). Others have held that necessity is the overarching concept of which duress of circumstances and duress by threats are two possible variants (R v Cairns [2000] RTR 15, at 20; R v Safi [2003] EWCA Crim 1809, at [24]). There is even some suggestion that English law recognizes a separate defence of necessity in addition to that of duress of circumstances and duress by threats (Re: A (Children) [2001] 2 WLR 480, per Brooke LJ at 569). An examination of the relationship between necessity, duress of circumstances, and duress by threats is beyond the scope of this book, but see D Ormerod, Smith and Hogan Criminal Law (11th edn, 2005), at 324–5; J Chalmers and F Leverick, Criminal Defences and Pleas in Bar of Trial (2006), chapter 4. 29 Cf CMV Clarkson, ‘Necessary action: a new defence’ [2004] Crim LR 81–95. See Section 30 As defined in Section 1.3 above. 1.4.3 below for criticism of Clarkson’s proposal.

Introduction

7

involves defensive force being directed towards the source of the threat posed. As Kasachkoff states, ‘[t]he crucial point here is that for a killing to qualify as a killing in self-defence, it not only must have self-protection as its aim but be directed at the person but for whom we would stand in need of no protection’.31 The defence of necessity is a plea by the accused that she broke the law because it was the least harmful of two (or more) alternative courses of action.32 Like self-defence, the defence of necessity can involve the infliction of harm on a person.33 Unlike self-defence, this would involve harming a bystander who was not posing a direct threat.34 An example might be the person who kills and eats one of her companions on a shipwreck in order to avoid starving to death.35 If this were to be recognised as a defence at all,36 then it would be properly classified as an act of necessity and not one of selfdefence, on the basis that the victim posed no direct threat to the life of the accused. That is not to say that the distinction between self-defence and necessity is always a clear one. There lie difficult cases at the borderline between the two defences, especially where passive threats are concerned.37 One example is the roped mountaineer, A, who falls off a cliff edge and is left dangling on a rope tied to her companion, B.38 If B can no longer hold A’s weight without falling, and cuts the rope, causing A to fall to her death, opinion has differed on the appropriate defence for B to claim.39 It is suggested here that this is properly a case of self-defence. In terms of the distinctions above, A is a passive threat and, as such, poses a direct threat to the life of B. She is not culpable for doing so, but the culpability of the threat is not a necessary condition for an act to be classified as one of self-defence.40 Other cases are more difficult to classify. For example, evidence was given in the inquest relating to the Herald of Free Enterprise disaster41 that a man, C, stood on a ladder blocking the path to safety of a number of other 31 T Kasachkoff, ‘Killing in self-defence: an unquestionable or problematic defence?’ (1998) 17 Law and Philosophy 509–31, at 513. 32 Or, if the excuse form of the defence is recognized, that the choice she made was a reasonable one in the circumstances (Chalmers and Leverick, Criminal Defences, chapter 4). 33 Although necessity is more commonly pled as a defence to careless driving or driving with excess alcohol (ibid, chapter 4). 34 Uniacke, Permissible Killing, at 166; J Horder, ‘Self-defense, necessity and duress: understanding the relationship’ (1998) 11 Canadian Journal of Law and Jurisprudence 143–65, at 150; S Michalowski, ‘Sanctity of life: are some lives more sacred than others?’ (2002) 22 LS 35 377–97, at 395. The facts of R v Dudley and Stephens (1884) 14 QBD 273. 36 37 It was not in Dudley and Stephens. Defined in Section 1.3. 38 For a real-life case of this nature, see J Simpson, Touching the Void (1988) and the 2004 film of the same title. 39 In Re: A (Children) [2001] 2 WLR 480, Brooke LJ classifies this as a clear case of necessity. Others, such as Horder, have regarded it as a case of self-defence (‘Self-defense, necessity and 40 duress’, at 147). See eg Re: A (Children) discussed below. 41 See R v HM Coroner for East Kent ex p Spooner (1987) 88 Cr App R 10; R v P&O European Ferries (Dover) Ltd (1990) 93 Cr App R 72.

8

Killing in Self-Defence

individuals, who were in freezing water and in danger of drowning.42 One of those in the water shouted at C, for ten minutes, asking him to move, but, when C did not, he instructed one of the other passengers to push C into the water where, it was presumed, C drowned. The coroner described this as ‘a reasonable act of self-preservation’43 but did not comment on whether the appropriate defence in these circumstances would be self-defence or necessity.44 The categorization was of considerable importance, given that while self-defence was accepted in English law as a complete defence to murder, necessity was not.45 The most appropriate classification is probably one of necessity, as the man might properly be regarded not as a direct threat, but as a bystander (the direct threat being the rising water), but this is by no means an obvious conclusion.46 Another difficult case is that of the conjoined twins in Re: A (Children).47 Here, two babies, named Mary and Jodie for the purposes of reporting the case, were born with a shared heart. In any operation to separate them, only Jodie, the stronger of the two, could survive. If separation did not take place, both twins would eventually die. The Court of Appeal was asked to rule on whether the separation of the twins would be lawful. The relevant issue is the nature of any defence the doctors undertaking the separation would have to a charge of Mary’s murder, given that at least two members of the court accepted that performing the operation would involve the intentional killing of Mary.48 One member of the court, Ward LJ, identified the appropriate defence as self-defence,49 whereas another, Brooke LJ, analysed the case in terms of 42 Neither R v HM Coroner for East Kent nor R v P&O European Ferries (Dover) Ltd describes this incident in detail, but see JC Smith, Justification and Excuse in the Criminal Law (1989), at 73–4. An entirely hypothetical, but similar, example can be found in P Foot, ‘The problem of abortion and the doctrine of double effect’, in Foot, Virtues and Vices (1978), at 21–2. Foot uses the example of a fat man who is stuck in the mouth of a cave, trapping a number of pot-holers within as the tide rises. The pot-holers have at their disposal a stick of dynamite and could blow up the fat man in order to save their own lives. 43 Smith, Justification and Excuse in the Criminal Law, at 74. This raises the question of whether ‘self-preservation’ might be a third category of defence, different from either self-defence or necessity. It is suggested here that it is not and that, although it is difficult to decide whether the case is one of self-defence or necessity, it should be analysed in terms of one or other of these two 44 established defences. No criminal proceedings were ever brought. 45 R v Dudley and Stephens (1884) 14 QBD. In R v Abdul-Hussain [1999] Crim LR 570, the Court of Appeal also ruled out the availability of duress of circumstances as a defence to murder (per Rose LJ at 7 of the Lexis transcript of the case—this aspect of the case is not reported in the Criminal Law Review). Cf Re: A (Children) [2001] 2 WLR 480, per Brooke LJ at 573. 46 Both Uniacke, in Permissible Killing (at 166), and Horder, in ‘Self-defense, necessity and duress’ (at 153), regard this as a case of necessity, on the basis that the man himself is not posing 47 a direct threat. [2001] 2 WLR 480. 48 Re: A (Children) [2001] 2 WLR 480, at 531 (per Ward LJ) and 549 (per Brooke LJ). 49 Ward LJ described the case as a ‘plea of quasi self-defence, modified to meet the quite exceptional circumstances nature has inflicted on the twins’ (Re: A (Children) [2001] 2 WLR 480, at 536). The term ‘defence of others’ is perhaps more appropriate, given that the doctors were not acting in defence of their own lives, but in order to save Jodie.

Introduction

9

necessity.50 Uniacke has doubted that the case is one of self-defence, on the basis that for Mary to be a direct threat to Jodie, there needs to be some prior position of Jodie’s that Mary directly threatened.51 As the twins were born conjoined, Uniacke argues, it is difficult to think of what this prior position might be. This is not regarded here as a meaningful basis upon which to rule out the defence of self-defence. The threat that Mary posed to Jodie seems little different in nature to the passive but direct threat posed to the mountaineer by her roped companion and I would suggest that both are best classified as cases of self-defence.52 Regardless of the conclusion reached, what Re: A (Children) and the other cases discussed above demonstrate is the difficulty of distinguishing clearly between self-defence and necessity in cases that lie close to the borderline.

1.4.2 Distinguishing self-defence from duress Distinguishing self-defence from duress is comparatively straightforward. Here, the distinction lies in the nature of the act undertaken by the accused. In a case of self-defence, the accused avoids harm by warding off or blocking a threat. In a case of duress, the accused avoids harm by complying with the demands of the threatener. For example, the accused, A, is threatened with death, by B, unless she commits theft. If A complies with the threat (and commits the theft), her act is one committed under duress. If A wards off the threat of death (for example by disabling or killing B), her act is one of selfdefence.53 This relatively clear-cut distinction has not stopped some from confusing acts of self-defence with acts committed under duress. Dressler,54 for example, has suggested that a battered woman acts under duress when she kills her violent partner while he is sleeping, in order to avoid future violence. This is on the basis that she chose to engage in conduct (killing her abusive partner) because she was coerced to do so by another person’s (the abuser’s) threat of future force. As Horder points out,55 this simply does not have the gist of a claim of duress. The battered woman is not complying with a 50 ibid 573. The basis upon which the third member of the court, Walker LJ, decided the case is not clear, but he did not regard it as either a case of self-defence or necessity. 51 S Uniacke, ‘Was Mary’s death murder?’ (2001) 9 Medical Law Review 208–20, at 213. See also S Michalowski, ‘Sanctity of life: are some lives more sacred than others?’ (2002) 22 LS 52 377–97, at 395. Or, more accurately in Jodie and Mary’s case, the defence of others. 53 The distinction and the example are drawn from J Horder, ‘Killing the passive abuser: a theoretical defence’, in S Shute and AP Simester (eds), Criminal Law Theory: Doctrines of the General Part (2002) 283–97, at 296. See also P Westen and J Mangiafico, ‘The criminal defense of duress: a justification not an excuse and why it matters’ (2003) 6 Buffalo Criminal Law Review 833–950, at 843, 846. 54 J Dressler, ‘Battered women who kill their sleeping tormenters: reflections on maintaining respect for human life while killing moral monsters’, in S Shute and AP Simester (eds), Criminal Law Theory: Doctrines of the General Part (1992) 259–82. 55 Horder, ‘Killing the passive abuser’, at 296.

10

Killing in Self-Defence

demand issued by her partner, or indeed any other person. Her claim is closer to one of self-defence, although it is unlikely that she will be able to satisfy the conditions of the defence.56 It is sometimes suggested that a further distinction can be made between self-defence and duress, based on the status of the victim. That is, in cases of self-defence, the accused harms someone who was a direct threat to her life or bodily integrity, whereas in cases of duress, the accused harms an innocent victim who was not the source of the threat.57 While this is undoubtedly true in the vast majority of claims of duress involving harm to a victim, it is not a necessary feature of the defence. It is possible to envisage a scenario where the person issuing the threats demands that the accused harms not an innocent victim, but the threatener herself. The threat from A that, unless B cuts off A’s leg, A will kill B is an example of this nature.

1.4.3 A unified defence of ‘necessary action’? These conceptual distinctions have not prevented Clarkson58 from arguing that the defences of self-defence, necessity and duress should be brought together under a unified defence (termed ‘necessary action’) governed by the same rules. For Clarkson: . . . what unites these actions is more important than what separates them. In all of them a defendant is committing a prima facie wrong in order to avoid some sort of crisis. To protect herself and/or other persons the defendant is committing what would otherwise be a crime and the true issue is how the law should respond to these actions and in what circumstances it should afford a defence.59

While it is true that the defences have much in common, there is nonetheless something to be said for treating necessity, duress, and self-defence as distinct defences. Self-defence differs from necessity and duress in that it always involves action against one or more persons who pose a direct threat to life or bodily integrity. This fact alone is sufficient reason for having a set of distinct rules; distinct, that is, from those governing necessity and duress, where the victim, if indeed there was one, would not generally have posed such a threat. The most obvious example relates to whether or not the defence can be pled to a charge of murder. This is generally accepted in relation to self-defence because the murder victim was herself a direct threat to the life of the accused, but it is extremely controversial in relation to necessity and duress, where any victim was likely to have been an innocent bystander. 56

On this, see Chapter 6. Horder makes this claim in ‘Killing the passive abuser’, at 296. This is surprising because earlier (in ‘Self-defense, necessity and duress’, at 149) Horder rejected this as a distinction 58 between the two defences. In ‘Necessary action: a new defence’. 59 ibid 88. 57

Introduction

11

Likewise, the public policy issues relating to each of the defences are not identical. For example, there is an argument for placing more stringent restrictions on duress than on the other two defences because of the risk that the defence is abused by terrorist groups or other criminal organizations.60 Indeed, this was precisely the reason given by the House of Lords in R v Hasan61 for imposing stricter conditions on the defence of duress (where the defendant is judged according to an objective test) than on the defence of self-defence (where a subjective test is applied). Aside from anything else, there is an argument to be made for conceptual separation of the three defences on the grounds of fair labelling.62 The accused is surely entitled to the nature of her defensive act being labelled as accurately as possible by the courts, rather than subsumed under a general heading of ‘necessary defence’, because she may be morally judged by society on the basis of this label. ‘I did this in self-defence’ (a justification defence and thus a claim that her behaviour was acceptable) reflects more favourably on the accused than ‘I did this under duress’ (most commonly an excuse defence and thus a claim that her behaviour was unacceptable but not blameworthy).63 60 Scottish Law Commission, A Draft Criminal Code for Scotland with Commentary (2003), at 63; J Chalmers and F Leverick, Criminal Defences, chapter 5. 61 [2005] UKHL 22, per Lord Bingham at [22] and [38]. 62 Ashworth, Principles of Criminal Law, at 89–92. 63 J Gardner, ‘The mark of responsibility’ (2003) 23 OJLS 157–71. Cf Clarkson, ‘Necessary action: a new defence’, at 94, who assumes that fair labelling is important only in terms of sending a message to the public about rules of conduct.

This page intentionally left blank

2 The Classification of Defences 2.1 Introduction This chapter sets out to explore the ways in which criminal defences can be classified, why classification might be important, and where self-defence fits into such classifications.1 Much debate has centred around the nature and scope of two types of criminal defence in particular: justifications and excuses. It will be argued that self-defence is normally a justification but where the claim of selfdefence is based upon a reasonable but mistaken belief by the accused that she was facing an attack, it is more correctly thought of as an excuse. Before examining this argument in detail, it is first necessary to locate this debate in the wider context of the nature of defences.

2.2 What is Meant by a Criminal Defence? Robinson defined a criminal defence as ‘any set of identifiable conditions or circumstances which may prevent conviction for an offence’.2 But, as Husak has pointed out, this definition is too wide because it lacks a normative element, and thus potentially includes circumstances such as avoiding being caught, murdering a witness, or bribing a judge.3 Thus an improvement on Robinson’s definition might run as follows: a criminal defence is any identifiable set of conditions or circumstances that provides sufficient reason why the accused ought not to be convicted of a particular offence.4 Although it overcomes Husak’s objection, the definition 1 In doing so, it draws in part from chapter 1 of J Chalmers and F Leverick, Criminal Defences and Pleas in Bar of Trial (2006). Some passages from Criminal Defences are reproduced with the kind permission of the publisher, W Green and Son Ltd. 2 PH Robinson, Criminal Law Defenses (1984), at § 21 (emphasis added). 3 D Husak, ‘The serial view of criminal law defenses’ (1992) 3 Criminal Law Forum 369–400, at 370; D Husak, ‘On the supposed priority of justification to excuse’ (2005) 24 Law and Philosophy 557–94, at 564. 4 This definition would not encompass pleas in bar of trial, such as entrapment, time-bar and delay, and prejudicial pre-trial publicity. A wider definition of criminal defences which would include pleas in bar of trial is proposed by Chalmers and Leverick in Criminal Defences as: ‘any identifiable set of conditions or circumstances that provides sufficient reason why the accused ought not to be convicted of a particular offence or ought not to stand trial for a particular offence’ (chapter 1, emphasis in original).

14

Killing in Self-Defence

is still not perfect, including as it does a claim by the accused that ‘I didn’t do it’. It is, however, extremely difficult to define a criminal defence in such a way as to exclude a claim of this nature (other than simply stating that this situation is excluded) and the definition above may well be the best that can be achieved. At the outset, it should be said that the view taken here is that there is a valid conceptual distinction between offences and substantive defences. Offences are prima-facie wrongs; acts (or omissions) that there is reason not to do and which require explanation if criminal liability is not to be imposed. Defences are conditions or circumstances which provide a reason why this prima-facie assumption of criminal liability should be displaced.5 To take the offence of murder as an example, the wrong is the intentional killing of a human being.6 If the offence definition is satisfied, the second stage of the process allows the accused to offer a full or partial defence of her conduct. To apply this two-stage analysis to killing in self-defence, the accused has, on the face of it, committed the offence of murder.7 In order to escape criminal liability, she asserts the supervening8 defence that she was acting in self-defence. In other words, she makes an admission of wrongdoing, but offers a defence of her actions. The view that offences and defences are conceptually distinct is not one that is universally held. Williams9 has suggested that the distinction is artificial and it is merely by chance and historical accident that various elements in the criminal law have been attributed to either the offence or defence side of the equation.10 Likewise, the conceptual distinction was not recognized in Drury v HM Advocate,11 where the Scottish Appeal Court revised the definition of murder in order to exclude cases of killing under provocation from its scope.12 5 K Campbell, ‘Offence and defence’, in IH Dennis (ed), Criminal Law and Justice: Essays from the WG Hart Workshop 1986 (1987) 73–86; J Gardner, ‘In defence of defences’, in P Asp, CE Herlitz, and L Holmqvist (eds), Flores Juris et Legum: Festskrift till Nils Jareborg (1992); J Gardner, ‘Fletcher on offences and defences’ (2004) 39 Tulsa Law Review 817–27; J Horder, Excusing Crime (2004), at 101; S Shute ‘Second Law Commission Consultation Paper on Consent: (1) Something old, something new, something borrowed: three aspects of the project’ [1996] Crim LR 684–93, at 690. 6 The precise definition of murder varies from jurisdiction to jurisdiction but intentional killing would constitute murder in all major common law jurisdictions. 7 For the sake of argument. Depending on the circumstances, the appropriate offence might be either murder or manslaughter. 8 A term coined in A Simester, ‘Mistakes in defence’ (1992) 12 OJLS 295–310. 9 G Williams, ‘Offences and defences’ (1982) 2 LS 233–56. 10 Cf Campbell, ‘Offence and defence’. 11 2001 SLT 1013. But cf Lord Wilberforce in DPP for Northern Ireland v Lynch [1975] AC 653, at 679–80. The distinction was also accepted by the Canadian Supreme Court in R v Parent [2001] 1 SCR 761 (at [6]) and R v Kerr [2004] 2 SCR 371 (per Bastarache J at [28]; LeBel J at [93]). 12 This aspect of Drury is not, however, one that has been followed in later Appeal Court decisions on defences. It was not mentioned at all in two subsequent appeals on the subject of defences: Lord Advocate’s Reference (No 1 of 2000) 2001 JC 143 (necessity) or Galbraith v HM Advocate (No 2) 2002 JC 1 (diminished responsibility). It was referenced only in passing in Cochrane v HM Advocate 2001 SCCR 655 (coercion). In fact, in the Lord Advocate’s Reference

The Classification of Defences

15

Admittedly, the distinction between offence and defence elements is sometimes difficult to place, but nonetheless separating offences and defences does have some value. As Shute has argued,13 in modern Western legal systems criminal laws are intended to affect the reasoning patterns of those who live under them. Thus, the offence definition should capture the essence of the wrong in order to give fair warning of the conduct that is prohibited.14 Rape is perhaps the best example. Engaging in sexual intercourse is not wrong in itself, but engaging in sexual intercourse without consent is wrong, and thus consent is part of the offence definition and not a supervening defence.15 Of course, it is possible for people to disagree about whether it is wrong to have sexual intercourse, but if one truly believes that sexual intercourse per se is wrong, then it does not defeat the argument about the separation of offences and defences; rather it follows that if one holds this view, sexual intercourse should be part of the offence definition and consent should be a supervening defence. The distinction also has some value for a more symbolic reason. It is important to reflect that when a defence is pled, a wrong has still taken place for which the accused should be called to account. To take the example of a killing in self-defence once again, the moral significance that society attaches to the wrong of intentional killing is such that it is something for which the accused should have to answer. As Tadros puts it, the offence definition captures the wrong that has been perpetrated.16 Even though the killing took place in self-defence, a human being has still been intentionally killed.17 An event has occurred that requires explanation, not least to the family of the victim, but also to society as a whole. To consider the ‘defence’ as part of the offence definition seems to gloss over the significance of this. It may be that, all things considered, the imposition of criminal liability is not appropriate but this does not take away the fact that the accused did something that is, on the face of it, of such seriousness that it merits separate examination.18 Indeed, it may still be morally appropriate for her to display some regret for her act, even if criminal liability is not imposed.19 it was made clear that necessity does not negate mens rea but is a separate defence in its own right (at [31]). 13

Shute, ‘Second Law Commission Consultation Paper on Consent’, at 690. See also AP Simester and GR Sullivan, Criminal Law: Theory and Doctrine (2nd edn, 15 2004), at 642. Campbell, ‘Offence and defence’. 16 V Tadros, Criminal Responsibility (2005), at 114. 17 For the sake of argument. Not all self-defensive killings will be intentional, but at least some will be (unless one subscribes to the doctrine of double effect—on which, see Chapter 3). 18 Some might maintain that in this scenario, all things considered, no wrong has occurred. But as Horder has argued (in Excusing Crime, at 101), one can look at a self-defence claim and say that the accused justifiably wronged the aggressor, even if necessary and proportionate force was used. This claim becomes even stronger if an innocent aggressor, such as a child or an insane person, was involved. 19 J Gardner, ‘Justifications and reasons’, in AP Simester and ATH Smith, (eds), Harm and Culpability (1996) 103–29; Gardner, ‘In defence of defences’. 14

16

Killing in Self-Defence

2.3 The Classification of Criminal Defences There are many ways in which defences can be classified: for example according to whether they are complete or partial defences; whether they are general or specific defences; whether they are common law or statutory defences; or whether or not there are special procedural rules governing when the defence can be pled.20 The type of classification that is of interest here, however, is classification according to the rationale for admitting the defence, that is, the nature of the explanation of why the accused is not held criminally responsible for her act or omission.21

2.3.1 Robinson’s classification In one of the first attempts at classification of this nature,22 Robinson identified five categories of defence: failure of proof defences; offence modifications; justifications; excuses; and non-exculpatory public policy defences. For Robinson, failure of proof defences23 refer to those instances where the prosecution, because of the ‘defence’, are unable to prove all of the required elements of the offence. Into this category he places mistakes of fact or law that negate an element of the offence; intoxication that negates an element of the offence; mental illness that negates an element of the offence; and consent (where this is part of the offence definition). If one accepts (as it is accepted here) that there is a valid distinction between offences and substantive defences, failure of proof defences are not, strictly speaking, defences at all, referring as they do to a failure on the part of the prosecution to prove an element of the offence. Offence modifications, Robinson’s second category, operate by modifying or refining the criminalization decision embodied in the definition of the offence. An offence modification defence is a claim that ‘while the actor has apparently satisfied all elements of the offence charged, he has not in fact caused the harm or evil sought to be prevented by the statute defining the offence’.24 Robinson provides examples including the impossibility defence in attempts (in legal systems where this is recognized); the principle that a victim of an offence cannot be an accessory to that offence; and the 20 For example, in Scotland some defences are designated ‘special defences’ and require the accused to give advance notification if they are to be pled at trial. On this, see Chalmers and Leverick, Criminal Defences, chapter 2. 21 On these other grounds of classification, see Chalmers and Leverick, Criminal Defences, 22 chapter 1. Robinson, Criminal Law Defenses, at § 21. 23 Robinson sometimes refers to these as ‘absence of an element defences’ (PH Robinson, ‘Criminal law defenses: a systematic analysis’ (1982) 82 Columbia Law Review 199–291, at 204). Tadros, in Criminal Responsibility, terms them ‘evidential defences’ (at 103). 24 Criminal Law Defenses, at § 23(a).

The Classification of Defences

17

‘de minimis’ defence (in legal systems where this is recognized), whereby all the conditions of the offence are satisfied but the harm caused is so minimal as not to be worth legally recognizing. They can also be offence specific, for example the possession of a valid prescription for otherwise prohibited drugs. Offence modifications are not considered here to be a distinctive category of defence in their own right; rather they are better regarded either as failure of proof defences or as falling into one of Robinson’s other categories. Certainly there is no suggestion that self-defence could ever fall into this category and therefore offence modifications will not be considered further. Justifications and excuses are the primary concern and will be considered at length in due course. Before doing so, for the sake of completeness, Robinson’s final category is non-exculpatory public policy defences. For Robinson, these operate where the accused has satisfied all of the requirements of the offence but for reasons other than blamelessness (such as public policy) a conviction should not result. Robinson suggests that statutes of limitation (where these are recognized); diplomatic immunity; pleabargained immunity (where this is recognized); prosecution misconduct; and entrapment fall into this category. There is no suggestion that self-defence could ever operate as a non-exculpatory public policy defence, and this type of defence will not be considered further here.25

2.3.2 Justifications and excuses Justifications and excuses comprise respectively the third and fourth categories of defence in Robinson’s classification. Unlike Robinson’s failure of proof defences, both operate as substantive defences arising once the actus reus and mens rea of the offence have been made out. Thus before it is appropriate for a claim of justification or excuse to be made, there needs to be something to justify or excuse. They also have in common the fact that, unlike Robinson’s non-exculpatory public policy defences, if successful, no blame attaches to the accused. The two concepts are, however, distinct. As a working definition, it might be said that the accused who claims a justification asserts that what she did was, all things considered, an acceptable thing to do, even though it satisfied the definition of an offence. The accused claiming an excuse asserts that, although what she did was unacceptable, there is a reason why she should not be blamed for it. The distinction between justifications and excuses is a topic that has attracted an enormous amount of discussion. The starting point is often 25 For a more detailed discussion of Robinson’s classification scheme, see F Leverick, A Critical Analysis of the Law of Self-defence in Scotland and England (2003), unpublished PhD thesis, at 19–27.

18

Killing in Self-Defence

taken to be the work of George Fletcher, writing in the 1970s,26 but interest in the distinction can be traced back further than that, at least to the philosophers JL Austin27 and HLA Hart.28 According to Austin, when we are claiming a justification, ‘we accept responsibility but deny that it was bad’.29 When we are claiming an excuse, ‘we admit that it was bad but don’t accept full or even any responsibility’.30 Immediately some concern arises over Austin’s use of the term ‘responsibility’. His claim could be taken to mean that, when we claim an excuse, what we are doing is denying responsibility in the sense that we did not cause the ‘bad thing’ to happen at all. If this is what Austin meant, he would, in Hart’s terms, be referring to causal responsibility. Hart distinguishes between four types of responsibility: role responsibility (the specific duties attached to a particular role such as a parent or doctor); causal responsibility (responsibility in the sense that the accused caused or made a contribution to an outcome); legal liability responsibility (the connection between the accused and the act is sufficient that she is legally liable for it—act responsibility31 might be another term); and capacity responsibility (the ability to understand, reason, and control conduct).32 Subsequent parts of Austin’s text suggest that what he meant by responsibility was not causal responsibility, but was closest to liability responsibility (although he may have been alluding to capacity responsibility too). Hart himself refers to justified conduct as ‘something the law does not condemn or even welcomes’.33 According to Hart, an excuse is claimed when ‘what has been done is something which is deplored, but the psychological state of the agent when he did it exemplified one or more of a variety of conditions which are held to rule out public condemnation and punishment of individuals’.34 Thus the basis upon which Austin and Hart distinguish between justifications and excuses is similar. A claim of justification is a claim primarily about the act itself. The accused claims responsibility for that act but denies that, all things considered, it was an unacceptable thing to do. A claim of excuse is a claim primarily about the individual. She accepts that what she did was

26 GP Fletcher, ‘The individualisation of excusing conditions’ (1974) 47 Southern California Law Review 1269–309; GP Fletcher, ‘The right deed for the wrong reason: a reply to Mr Robinson’ (1975) 23 UCLA Law Review 293–321; GP Fletcher, Rethinking Criminal Law (1978). 27 In 1956, Austin gave a paper at the Aristotelian Society entitled ‘A plea for excuses’. This has since been reproduced in various edited collections, such as JL Austin, ‘A plea for excuses’, in AR White (ed), The Philosophy of Action (1968) 19–42. 28 In 1957, Hart addressed the New York Institute of Philosophy on the subject of ‘legal responsibility and excuses’, a paper that was later published in Hart’s own monograph: HLA Hart, Punishment and Responsibility: Essays in the Philosophy of Law (1968). 29 30 Austin, ‘A plea for excuses’, at 20. ibid, at 20. 31 V Tadros, ‘Insanity and the capacity for criminal responsibility’ (2001) 5 Edin LR 325–54. 32 33 34 Hart, Punishment and Responsibility, at 212–22. ibid, at 13. ibid, at 14.

The Classification of Defences

19

unacceptable but claims that, for some reason pertinent to her as an individual, we should not blame her for her conduct.35 The same themes are reflected in Fletcher’s work. Fletcher locates the difference between justifications and excuses in a distinction between wrongdoing and culpability: Claims of justification concede that the definition of the offence is satisfied, but challenge whether the act is wrongful; claims of excuse concede that the act is wrongful, but seek to avoid the attribution of the act to the actor. A justification speaks to the rightness of an act; an excuse, to whether the actor is accountable for a concededly wrongful act.36

It is unlikely that many of those who have written about justifications and excuses would disagree with the general gist of this distinction.37 Debate has tended to focus on issues such as which defences fall into each category (and, especially, on how mistake is to be treated)38 and on attempting to provide an overarching rationale for why certain types of conduct should be justified or excused. It has been suggested, however, that a further distinction can be made between ‘true’ excuses, such as duress, and defences such as non-age that operate as lack of capacity defences.39 Lack of capacity defences differ from excuses in that the latter relate only to those who are regarded as capable of complying with the law. The accused claiming an excuse asserts responsibility for her acts in general, but realizes that she should have acted differently on this particular occasion and puts forward a reason why she should not be blamed. As Wallace puts it, excuses ‘give us a reason to withdraw the attitudes we would ordinarily take in response to a particular action, but they do not give us a reason to

35 Although, as will be noted in due course, it is perhaps overly simplistic to conclude that justifications are solely concerned with acts and excuses with actors. 36 Fletcher, Rethinking Criminal Law, at 759. 37 J Dressler, ‘Reflecting on excusing wrongdoers: moral theory, new excuses and the moral penal code’ (1988) 19 Rutgers Law Journal 671–716, at 675–6; K Greenawalt, ‘Distinguishing justifications from excuses’ (1986) 49 Law and Contemporary Problems 89–108, at 91; DN Husak, ‘Justifications and the criminal liability of accessories’ (1989) 80 Journal of Criminal Law and Criminology 491–520, at 496; M Moore, Placing Blame: A General Theory of the Criminal Law (1987), at 483; MN Berman, ‘Justification and excuse, law and morality’ (2003) 53 Duke Law Journal 1–77, at 4. 38 A theme to which we return later: see Section 2.4 below. 39 RJ Wallace, Responsibility and the Moral Sentiments (1994); A Duff, ‘Law, language and community: some preconditions of criminal liability’ (1998) 18 OJLS 189–206; Gardner, ‘Justifications and reasons’; Horder, Excusing Crime, at 9–10; Tadros, Criminal Responsibility, at 124–9; M Baron, ‘Justifications and excuses’ (2005) 2 Ohio State Journal of Criminal Law 387–406. The term ‘exemptions’ is sometimes used (eg by Wallace and Tadros) instead of ‘lack of capacity defences’. Lack of capacity defences is, however, a more appropriate term. Exemptions could be viewed as including some non-exculpatory defences, such as diplomatic immunity. Horder terms lack of capacity defences as ‘out and out denials of responsibility’ (Excusing Crime, at 8).

20

Killing in Self-Defence

view the agent as anything other than an ordinary, accountable person in general’.40 Conversely, the accused who claims a lack of capacity defence does not need to bother making excuses as, due to an incapacity stemming from, for example, mental illness, she is regarded as incapable of making her conduct conform to the criminal law.41 The two defences most commonly described as lack of capacity defences are insanity and non-age. Describing non-age as a lack of capacity defence is relatively uncontroversial. Describing insanity as a lack of capacity defence, as some criminal law theorists have done,42 is more contentious. The leading US criminal law theorists, Dressler,43 Fletcher,44 Robinson,45 Greenawalt,46 and Kadish,47 all classify insanity as an excuse and do not generally contemplate the possibility that it might better be thought of as a different type of defence altogether.48 In practice, whether insanity is best described as an excuse or a lack of capacity defence depends largely on how it is defined in the legal system in question. Scots law, for example, might be said to recognize both forms of the defence. The substantive defence of insanity relates to the condition of the accused at the time when the offence was alleged to have been committed49 and thus is probably best classified as an excuse. An acquittal on the ground of insanity may well say nothing about whether the accused is a proper subject for the sanction of the criminal law. Scots law also recognizes a plea in bar of trial based on insanity and this might more properly be described as a lack of capacity defence in that it is concerned with the accused’s capacity to participate in the criminal process.50 40

Wallace, Responsibility and the Moral Sentiments, at 155 (emphasis in original). Colvin makes a similar distinction, using the phrase ‘defences of mental impairment’ to describe defences where the accused is not a proper subject for handling within the standard structure of criminal liability. Unlike others who have argued for the recognition of lack of capacity defences, he does not proceed to make a further distinction between justifications and excuses. Instead he places all other defences into a single category called ‘defences of contextual permission’ (E Colvin, ‘Exculpatory defences in criminal law’ (1990) 10 OJLS 381–407, at 383). 42 Tadros, ‘Insanity and the capacity for criminal responsibility’, at 341; N Lacey, State 43 Punishment (1988), at 74. Dressler, ‘Reflecting on excusing wrongdoers’, at 682. 44 Fletcher, ‘The individualisation of excusing conditions’, at 1293; GP Fletcher, ‘Rights and excuses’ (1984) 3 Criminal Justice Ethics 17–27, at 17. 45 Robinson, Criminal Law Defenses, at § 28. 46 K Greenawalt, ‘The perplexing borders of justification and excuse’ (1984) 84 Columbia Law Review 1897–927, at 1915. 47 SH Kadish, ‘Excusing crime’ (1987) 75 California Law Review 257–89, at 262. 48 Fletcher does at least consider the possibility in Rethinking Criminal Law, at 836–9, but abandons it and in his later work he is firmly of the belief that insanity is an excuse. 49 D Hume, Commentaries on the Law of Scotland Respecting Crimes (4th edn, 1844) i, 39; Brennan v HM Advocate 1977 JC 38, per the Lord Justice-General (Emslie) at 45. 50 HM Advocate v Wilson 1942 JC 75, per the Lord Justice-General (Normand) at 80. Detailed discussion of the proper basis for the defence of insanity lies beyond the scope of this book (but see Chalmers and Leverick, Criminal Defences, chapter 7). 41

The Classification of Defences

21

2.3.3 An alternative classification In the light of the preceding discussion, a refined classification scheme can be suggested as follows. (1) Justification defences A justification defence is claimed where the accused has engaged in the wrongdoing prohibited by the offence but, all things considered, her conduct is judged to be acceptable. It should be pointed out that, to some extent, the category that a particular defence falls into depends on the way in which the defence is formulated in the jurisdiction in question. There is no absolute rule, for example, that necessity must always be characterized as a justification or that duress must always be characterized as an excuse.51 That said, based on the definitions above, the category of justification defences would include most claims of self-defence,52 and some claims of necessity and duress.53 (2) Excuse defences An excuse defence is claimed where the accused has acted unacceptably, but we do not blame her for what she did. We hold an excused person morally accountable for her acts in general but do not blame her for the particular act in question. Excuse defences would include some claims of self-defence,54 the vast majority of claims of duress and necessity and claims of superior orders (if this is recognized as a defence at all). This category would also include some claims of error of law, if reasonable error of law is accepted as a defence. It would not include a claim of non-age, as this more properly falls into the third category, lack of capacity defences. The defence of insanity is more difficult to classify: as we have already seen, it can be regarded either as an excuse or a lack of capacity defence (or indeed both) depending on how it is defined. The partial defence of provocation is also difficult to classify, but it is most commonly regarded as a partial excuse (with a justificatory component).55 Claims of diminished responsibility might also fall into this category, although they might equally be regarded as partial lack of capacity defences. (3) Lack of capacity defences A lack of capacity defence is claimed where the accused is not regarded as sufficiently capable to be held accountable for her behaviour at the time of the alleged offence. This category comprises primarily non-age but might also include insanity, depending on how the defence is formulated. 51

Berman, ‘Justification and excuse, law and morality’, at 68. At least those where the accused is not mistaken about her belief in the need to use self-defensive force. On this, see Section 2.4 below. 53 Where the harm prevented far outweighed the harm threatened. There is insufficient space for this claim to be defended here, but see Chalmers and Leverick, Criminal Defences, chapters 4 (necessity) and 5 (duress). 54 Specifically those where the accused is mistaken in her belief that she is under attack. 55 J Dressler, ‘Why keep the provocation defense? Some reflections on a difficult subject’ (2002) 86 Minnesota Law Review 959–1002, at 971. 52

22

Killing in Self-Defence

(4) Non-exculpatory public policy defences This category of defences is retained from Robinson’s classification. As Robinson outlined, it covers cases where, for reasons other than blamelessness or a lack of capacity (such as reasons of public policy), a conviction should not occur. It includes defences such as diplomatic immunity, time-bar, and entrapment. While it should be acknowledged that this classification is based in part on that of Robinson,56 it differs from his in two important respects. First, a distinction is made between excuses and lack of capacity defences, which Robinson does not make. Second, two of Robinson’s categories are left out entirely. Failure of proof defences are left out because, as it was explained earlier, they are not considered to be substantive defences. Offence modifications are left out because they are not considered to be a distinctive category of defence in their own right; rather they are a collection of defences that are better regarded either as failure of proof defences, or as falling into one of the four categories above.57

2.3.4 Is there a ‘hierarchy of defences’? It has been suggested by John Gardner,58 among others,59 that the classification above has a normative structure in terms of the defence that reflects most favourably on the accused. For Gardner, justifications are the most preferable type of defence to claim, followed by excuses and then lack of capacity defences. This is because it is morally preferable to have acted acceptably (a justification) than to be excused for acting unacceptably.60 It is also morally preferable to be excused for acting unacceptably than not to have the capacity whereby one’s behaviour can be judged acceptable or unacceptable at all.61 Gardner does not make anything other than passing reference to non-exculpatory defences. It might be assumed, though, that they rank below justifications and excuses in terms of the manner in which 56

See Criminal Law Defenses, at § 21. For an alternative classification that takes into account partial defences, see Horder, Excusing Crime, at 103. In most respects, Horder’s classification does not differ significantly from the one proposed here (with the exception that he deals with partial defences separately and leaves out non-exculpatory defences). 58 J Gardner, ‘The mark of responsibility’ (2003) 23 OJLS 157–71; J Gardner, ‘The gist of excuses’ (1998) 1 Buffalo Criminal Law Review 575–98, at 590; Gardner, ‘In defence of defences’, at 262. 59 Baron, ‘Justifications and excuses’, at 389; Horder, Excusing Crime, at 99; V Tadros, ‘The structure of defences in Scots criminal law’ (2003) 7 Edin LR 60–79, at 60–1. 60 Cf Husak, ‘The serial view of criminal law defenses’; D Klimchuk, ‘Necessity, deterrence and standing’ (2002) 8 Legal Theory 339–58, at 355; Husak, ‘On the supposed priority of justification to excuse’. 61 It might be questioned whether this is indeed true of the plea of non-age. It is surely the case, though, that the self-respecting child would prefer to have her conduct deemed justified and therefore acceptable than she would to be told that she was not to be held accountable by virtue of her age. Cf Tadros, ‘The structure of defences’, at 61 n 5. 57

The Classification of Defences

23

they reflect on the accused, given that the defence is granted for reasons other than blamelessness. Some support for this proposition can be found in the Canadian Supreme Court case of R v Mack,62 which held that entrapment (which, in Canada, is regarded as a non-exculpatory defence) should only be considered after the trial has been completed and the accused found guilty. This is in order to preserve the right of the accused to the more favourable verdict of an acquittal on the basis of justification or excuse. This example notwithstanding, there is no general rule that the hierarchy dictates the order in which the defences would be considered in practice. For one thing, although morally they may be considered inferior to claims of justification, in the majority of common law jurisdictions non-exculpatory defences are matters which would be considered pre-trial and which operate to prevent a trial taking place. Likewise, if the accused had potentially a lack of capacity defence (such as non-age) and a justification defence (such as selfdefence), it would make little practical sense, given the ease of proving nonage,63 to plead the justification first and only move on to the lack of capacity defence if the justification was unsuccessful.

2.4 The Treatment of Mistake Having established this backdrop we can turn to an issue central to the law of self-defence—the treatment of mistake. There are two considerations here. The first is how to treat the accused who mistakenly believes that her act is justified. An example is Sunita, who acts in ‘self-defence’ under the mistaken belief that she is about to be attacked. The question here is whether Sunita is best described as justified or excused. The second is how to regard the ‘unknowingly justified’ accused. Here, the accused fulfils the requirements of the offence but, unknown to her, circumstances exist in which her act would have been justified. An example is Liz, who attacks her hated enemy when, unknown to her, said enemy is about to throw a bomb into a crowd that is likely to kill a large number of people. The question here is whether Liz should receive the benefit of a defence at all. It is argued that Sunita in the first scenario is excused whereas Liz in the second scenario should be held criminally liable for her act. This conclusion is based on John Gardner’s claim that the structure of at least certain types of excuse, of which mistaken self-defence is one, derives from the structure of justifications. It is useful, therefore, to start by considering Gardner’s argument in more detail. 62

[1988] 2 SCR 903. At least in the context of most major common law jurisdictions. Circumstances can be envisaged whereby proof of age might be more difficult to acquire. 63

24

Killing in Self-Defence

2.4.1 Gardner’s dual requirement theory Gardner argues that the accused is justified if and only if ‘it is true both that there was an applicable (guiding) reason for so acting . . . and that this corresponded with the (explanatory) reason why the action was performed’.64 Gardner’s terminology requires some explanation. The distinction between guiding reasons and explanatory reasons stems from Raz.65 A guiding reason is one that relates to what one ought to do. As Dancy puts it, it is a reason that makes an action ‘sensible in the circumstances, morally required, or in some other way to be recommended’.66 In any given situation there may be guiding reasons in favour of an action and guiding reasons against it but the morally appropriate action is the one supported by the balance of reasons.67 For example, when Claire comes across an accident victim who requires urgent hospital treatment and there is no means of telephoning an ambulance, there is a strong guiding reason for Claire to take that person to hospital in her own car. There may also be guiding reasons against doing so. Claire may be late for an appointment, or she may be legally disqualified from driving, but these are outweighed by the stronger guiding reason that the accident victim may die if Claire does not act. An explanatory reason relates to the reasons for which one actually does act. We use explanatory reasons to explain our behaviour (‘the reason I did X was Y’). Dancy terms such reasons ‘motivational reasons’,68 describing them as the reasons ‘in the light of which, or despite which, [the agent] acted as he did’.69 Our explanatory reasons for acting may or may not correspond with the guiding reasons for acting. If Claire chooses to drive the accident victim to hospital (despite being legally disqualified from driving) and does so in order to save that person’s life, then her explanatory reason for doing so corresponds with the applicable guiding reason for doing so and, for Gardner, she would be justified. To be justified, then, following Gardner, two conditions must be satisfied: (a) there must be an appropriate guiding reason to act in a particular way and (b) that must be the reason for which the accused did act. Neither component alone is sufficient to ground a claim of justification. For Gardner, the accused who mistakenly but reasonably believes she is justified (that is, she has appropriate explanatory reasons for acting but the appropriate guiding reasons do not exist) is excused. The structure of excuse thus derives 64

Gardner, ‘Justifications and reasons’, at 105. J Raz, Practical Reason and Norms (1975). 66 J Dancy, Practical Reality (2000), at 2. Dancy does not use the term ‘guiding reasons’, preferring ‘normative reasons’, a term that is also preferred by Tadros in Criminal Responsibility, 67 at 29. Raz, Practical Reason and Norms, at 36. 68 A term that describes them better than Gardner’s ‘explanatory reasons’ (see Tadros, in Criminal Responsibility, who prefers the term ‘motivating reasons’). To avoid confusion, given that his theory will be referred to extensively in this chapter, Gardner’s terminology will 69 Dancy, Practical Reality, at 2. nonetheless be used. 65

The Classification of Defences

25

from the structure of justification. An excused person acts for what she thinks are good reasons but what are, in fact, invalid reasons.70 In the context of self-defence, the excuse form of the defence covers the situation where the accused mistakenly but reasonably believes that she is being attacked. She does not have good reasons for acting but she has good reasons for her belief that she needs to act. At this stage it is useful to distinguish Gardner’s approach from a requirement for the accused to have knowledge of the potentially justifying circumstances. This is not what Gardner is proposing at all. Gardner’s structure relates to reasons and requires that the accused act for the justificatory reason. Simply having knowledge of the justificatory circumstances is not enough to ground a claim of justification.71 The distinction can be seen in the following scenario: Mike sees his hated enemy, Ken, attacking a woman in the street with a knife. Thinking quickly, he realizes this is an opportunity to cause harm to Ken. He rushes over the road and stabs Ken in the chest. Ken dies but the woman’s life is saved. Here, Mike has knowledge of the potentially justificatory circumstances (or, in Gardner’s terms, the guiding reason for action) but he does not act for this reason so is not, all things considered, justified. Gardner’s view of the structure of justifications has received surprisingly little support. Indeed, ‘dual requirement’72 theories are often summarily dismissed. Robinson devotes only a single page to them in a twenty-five-page article on competing theories of justification, concluding that he finds them ‘puzzling’.73 One exception is Dillof, who proposes a slight variation on Gardner’s argument whereby ‘regard’ for the justificatory circumstances is sufficient to ground a justification defence.74 To this, one might add Fletcher, whose conception of justifications is similar to Gardner’s, although it is couched in different terms. For Fletcher, justified conduct requires the existence of both justificatory circumstances and what he variously terms either ‘justificatory intent’75 or ‘good reasons’.76 He is not entirely consistent in his approach, 70 Gardner, ‘Justification and reasons’, at 119. In a later paper (‘The gist of excuses’), Gardner appears to suggest that all excuses are structured in this way. Tadros, however, has argued that this is simply one of a number of different formulations of excuses. See V Tadros, ‘The characters of excuse’ (2001) 21 OJLS 495–519. For a convincing formulation that does not depend on a ‘failed justification’, see GR Sullivan, ‘Making excuses’, in AP Simester and ATH Smith (eds), Harm and Culpability (1996) 131–52. 71 AM Dillof, ‘Unravelling unknowing justification’ (2002) 77 Notre Dame Law Review 1547–600, at 1596. 72 Robinson’s term, used in PH Robinson, ‘Competing theories of justification: deeds v reasons’, in AP Simester and ATH Smith (eds), Harm and Culpability (1996) 45–70, at 67. 73 Robinson, ‘Competing theories’, at 68. 74 By which he means ‘shorthand for the more bulky phrase “takes into account in his reasoning process when deciding what to do” or “functions as a reason (even if not the sole reason) relevant to the actor’s choice” ’ (Dillof, ‘Unravelling unknowing justification’, at 1597–8). 75 ‘The right deed for the wrong reason’, at 318. 76 GP Fletcher ‘Domination in the theory of justification and excuse’ (1996) 57 University of Pittsburgh Law Review 553–78, at 561.

26

Killing in Self-Defence

though, and has sometimes referred to the need simply for knowledge of the potentially justificatory circumstances.77 Aside from Dillof and Fletcher,78 in what might be called the mainstream of (especially) American criminal law theorists, the debate over the structure of justifications has tended to centre on whether either explanatory reasons or guiding reasons are sufficient to ground a justification. Robinson frames this debate in terms of those who support a ‘reasons’ theory of justification and those (such as himself) who support a ‘deeds’ theory.79 Others have framed the debate as one between ‘subjective’ and ‘objective’ theories of justification.80 For convenience, the terms deeds theory and reasons theory will be used here, even though they do not quite encapsulate the two opposing positions.81 Regardless of terminology, the difference between the deeds theory and the reasons theory is most apparent when mistakes are considered. The reasons theory would justify the accused on the basis of (usually reasonable) belief alone. Thus the accused who mistakenly (but reasonably) believes in the need for self-defensive force is justified regardless of whether that force was actually necessary. In Gardner’s terms, the reasons theory would justify on the basis of explanatory reasons alone. The deeds theory would justify the accused whenever justifying circumstances were present, regardless of whether she acted for this reason. Thus the accused who kills a bearded man simply because she hates men with beards will be justified on the grounds of ‘self-defence’ if the bearded ‘victim’ was about to produce a gun and shoot her. In Gardner’s terms, the deeds theory would justify on the basis of guiding reasons alone. I agree with Gardner that both an appropriate guiding reason and a matching explanatory reason are necessary for a claim of justification and will arrive at this conclusion by suggesting that both the reasons theory and the deeds theory have weaknesses.

2.4.2 The reasons theory Devinder is walking down the street when he comes across two middle-aged men apparently attacking a teenage girl. He attempts to assist the girl, but kills one of the ‘assailants’ in the process. In reality, the men were plain-clothes 77 Fletcher, Rethinking Criminal Law, at 565. There is also some indication that the dual requirement theory is accepted in Scots law: see Dawson v Dickson 1999 JC 315 (at 319) and the discussion in Chalmers and Leverick, Criminal Defences, chapters 1 and 4. See also the Northern Irish case of R v Thain [1985] NI 457, in which the defendant’s defence was rejected on the basis that he had not acted for the justificatory reason. 78 See also TL Price, ‘Faultless mistake of fact: justification or excuse?’ (1993) 12 Criminal 79 Justice Ethics 14–28. Robinson, ‘Competing theories’. 80 J Dressler, ‘New thoughts about the concept of justification in the criminal law: a critique of Fletcher’s thinking and re-thinking’ (1984) 32 UCLA Law Review 61–99. 81 A more accurate description might be ‘factual circumstances alone’ (deeds theory) and ‘accused’s reasons alone’ (reasons theory).

The Classification of Defences

27

police officers legitimately trying to arrest the girl.82 Devinder mistakenly (but reasonably) believed his action was justified. Those subscribing to the reasons theory would grant Devinder a defence based on justification rather than excuse.83 For reasons theorists, Devinder acted for what he thought (justifiably) were good reasons and this is enough to ground a claim of justification. The original supporters of the reasons theory are Dressler84 and Greenawalt,85 although other prominent criminal law theorists have since come out in support of this view.86 It will be argued here that good reasons alone should not ground a claim of justification and that a claim of mistaken but reasonable belief in self-defence87 is more appropriately regarded as an excuse. In order to reach this conclusion, it is necessary to consider Dressler’s arguments for the reasons theory.88 These collapse essentially into two points. The first is what might be termed a behaviour-guiding argument (although it also draws on rule utilitarianism).89 The second, and more important, is a moral argument.

2.4.2.1 The behaviour-guiding argument Dressler believes that justification defences should guide behaviour90 and argues that the greatest possible number of victims’ lives would be saved if we encourage people to act according to reasonable appearances. If we only grant a justification defence to those who correctly perceive the factual circumstances, a greater number of ‘victims’ are likely to die because they were too careful to check that they really were being attacked before they responded. Husak makes a similar point.91 He holds that justifications should be capable of guiding conduct. It would be wrong, for Husak, to structure justifications such that judgement of conduct is withheld until the consequences of that conduct are known. He suggests that this would equate 82 This scenario is based on the facts of People v Young 183 NE 2d 319 (1962), although in Young the police officer was not killed but merely injured. 83 In fact, in People v Young, the majority of the Court of Appeals of New York came to the surprising conclusion that the defendant, who had made a perfectly reasonable mistake, was not entitled to a defence at all. Cf the dissenting opinion of Froessel J. The case is discussed in more 84 detail in Chapter 9. In ‘New thoughts’ and ‘Reflecting on excusing wrongdoers’. 85 In ‘Perplexing borders’, and ‘Distinguishing justifications from excuses’. 86 RL Christopher, ‘Mistake of fact in the objective theory of justification: do two rights make two wrongs make two rights . . . ?’ (1994) 85 Journal of Criminal Law and Criminology 295–332; F McAuley, ‘The theory of justification and excuse: some Italian lessons’ (1987) 35 American Journal of Comparative Law 359–80; T Morawetz, ‘Reconstructing the criminal defenses: the significance of justification’ (1986) 77 Journal of Criminal Law and Criminology 277–307; Baron, ‘Justifications and excuses’; H Stewart, ‘The role of reasonableness in selfdefense’ (2003) 16 Canadian Journal of Law and Jurisprudence 317–36; Tadros, Criminal 87 Responsibility, at 280–90. Or in Devinder’s example, a claim of defence of others. 88 In ‘New thoughts’, at 92–5. These arguments appear in similar form in Greenawalt’s ‘Perplexing borders’ and ‘Distinguishing justifications from excuses’ but they are expressed with 89 more clarity by Dressler. See Chapter 3. 90 Some doubt is cast on this claim when it is examined in more detail in Section 2.5 below. 91 ‘Justifications and the criminal liability of accessories’, at 508.

28

Killing in Self-Defence

to the behaviour-guiding principle of ‘do whatever has the best consequences even though there is no way of identifying them at the time of choice’. A better conduct-guiding principle, for Husak, would be ‘act according to the best information available at the time’. This argument is flawed for many reasons. For one thing, Dressler’s concern lies exclusively with the ‘victim’ of an attack and he neglects to consider the lives of those who are mistaken for attackers. But setting this aside, if justifications are to act as an effective guide to legally permitted conduct, surely they should articulate the type of conduct that is permissible. It is difficult to dispute Husak’s maxim that people should act according to the best information available at the time but this says nothing to the general public about the type of conduct the law finds acceptable. Husak is possibly confusing rules of conduct with decision rules or rules of adjudication.92 Rules of conduct are addressed to those bound by the criminal law and relate to the types of behaviour the criminal law finds acceptable. An example of a conduct rule might be ‘do not use force against another person unless it is necessary in selfdefence’. Rules of adjudication are addressed to those who judge legal disputes. They are the rules that judges should apply when the accused has failed to follow a conduct rule. A rule of adjudication might be ‘you should be judged on what you knew or believed at the time’. If justification defences are to function as conduct rules, mistaken justification should surely be excluded from their ambit, otherwise there is the danger that people assume the type of conduct involved in mistaken justifications is approved of by the criminal law. Robinson93 gives the example of the acquittal of the police officers who administered a severe beating to Rodney King, a young African-American, in the process of trying to arrest him. The officers argued, successfully, that they were justified in doing so because they mistakenly believed, first, that Mr King had a gun and, second, that he was intoxicated with a drug said to endow its users with ‘superhuman strength’. This, suggests Robinson, sends out the signal to the general public that such conduct is acceptable and condoned. A defence based on excuse (due to mistaken belief) would be more appropriate. The conduct would be condemned but the mistake made by the officers would indicate that they were acting on what they thought were good reasons and thus they should not be blamed for what they did.94 The rule-utilitarian element of Dressler’s argument is similarly confusing. If we did not grant a complete defence of any sort to the mistaken accused, he might have a point. If the result of a mistaken but reasonable belief in the 92 M Dan-Cohen, ‘Decision rules and conduct rules: on acoustic separation in criminal law’ (1984) 97 Harvard Law Review 625–77; PH Robinson, ‘Rules of conduct and principles of adjudication’ (1990) 57 University of Chicago Law Review 729–71. 93 PH Robinson, ‘A functional analysis of criminal law’ (1994) 88 Northwestern University Law Review 857–913, at 878–82. 94 Of course, this argument does disintegrate somewhat given that the verdict system does not discriminate between acquittals through justification and acquittals through excuse. This is discussed further in Section 2.5 below.

The Classification of Defences

29

need for lethal self-defensive force was a conviction for murder, then this might well make ‘victims’ hesitate too much before warding off lifethreatening attacks. But this is not what is being suggested. What is being proposed is that the accused is still granted a complete defence, but that the defence is an excuse. The guiding principle that we are justified in using necessary and proportionate force in self-defence when faced with an attack, coupled with the reassurance that if we do make a reasonable mistake about the circumstances we will be excused, would surely not produce the increase in victims Dressler predicts. Such a rule, relating as it does to acceptable conduct, would surely not deter the accused who is mistaken. After all, she believes at the time that she is facing an attack from which she needs to defend herself and will therefore act accordingly.

2.4.2.2 The moral argument Dressler’s second argument for the reasons theory is that it is difficult to make a moral distinction between the accused who proceeds in line with a reasonably mistaken belief95 and the accused who has correctly perceived the circumstances. Both, for Dressler, are morally ‘faultless’.96 Pursuing this theme, Greenawalt suggests that the term justification relates to the reasons for which we act and if we can show that we acted for sound and proper reasons, we are entitled to a justification: ‘. . . justified action is morally proper action. Justification is used in relation to the reasons one puts forward for one’s choices; an action is justified in this sense when one has defended it with sound arguments . . . to be justified is to have sound, good reasons for what one does or believes.’97 This is also the basis upon which Tadros argues for the reasons theory. In some of his early work,98 Tadros was a supporter of Gardner’s dual requirement theory but has now, as he puts it, ‘learned the error of [his] ways’.99 Like Dressler, Tadros does not accept that there is a relevant moral distinction between having good reason to believe something and having good reason to do it:100 in both cases, the accused is justified. A similar 95 Dressler does not indicate how he would deal with the accused who holds an unreasonably mistaken belief. The issue of whether reasonable grounds for a mistaken belief should be required is discussed in Chapter 9. 96 See also Greenawalt, in ‘Perplexing borders’, at 94; McAuley, ‘The theory of justification and excuse’, at 369. 97 Greenawalt, ‘Perplexing borders’, at 1903. See also Morawetz, ‘Reconstructing the criminal defenses’, at 289. Morawetz’s reasoning leads him into something of a muddle. While he sees no reason not to justify the accused who acts according to good reasons, he also recognizes that there is a conceptual distinction between ‘mistaken belief’ justifications and ‘true’ justifications. He addresses this by creating two separate categories of justified behaviour: behaviour ‘justified tour court’ (‘true’ justifications) and ‘justified wrongs’ (mistaken justifications). He might do better to pursue the idea that his justified wrongs are actually excuses. 98 ‘The characters of excuse’ and ‘The structure of defences in Scots criminal law’. 99 100 Criminal Responsibility, at 281 n 22. ibid, at 285.

30

Killing in Self-Defence

position is taken by Baron, who argues that where the accused has ‘done everything as well as anyone could possibly have done in the circumstances’, it is wrong to say that she was not justified.101 Gardner’s response to this would most likely be that it fails to distinguish between justified beliefs and justified acts. Returning to the example used at the start of this section, we cannot fault Devinder in forming the belief that he did: he reasonably believed that the course of action he took was justified. In Gardner’s terms, he had good explanatory reasons for acting. But this is not the same as saying that his actions were justified. In reality there were not the guiding reasons for action that Devinder thought there were.102 But what of the argument that there is no moral distinction to speak of between Devinder and the objectively justified accused? This is, on the face of it, a seductive argument. Like the objectively justified accused, Devinder is entirely blameless. We cannot fault him. But while this is sufficient condition to describe him as excused, it is not sufficient condition to describe him as justified. Justification depends on more than mere belief; it relates as well to the reality of the situation. In the context of self-defence, we are talking here about a state of affairs where harm has been caused to someone who was posing no threat whatsoever to the accused. In our example, Devinder, in the mistaken belief that it was necessary to save the life of a third party, killed a plain-clothes police officer who was acting entirely legitimately. This is not a harm that we should simply ignore.103 This is because justifications speak not just to the accused but also to victims, their families, and to society. They tell us that what the accused did was acceptable. Now if Devinder was objectively correct in his belief, we could say that the harm he caused to the aggressor was justified because she threatened (with no justification) to deprive a third party of her life. But where there was no aggressor in reality, it seems odd to say that Devinder was justified and therefore what he did was entirely acceptable. After all, he harmed a third party who was posing no threat. There was, in reality, no good reason to kill this person. Devinder is excused—morally blameless for sure—but nonetheless excused, on the basis that he had good reasons for his mistaken belief. Thus, in a sense, Dressler and the other reasons theorists are quite correct to say that there is little moral difference between the objectively justified accused and the accused who has a reasonably mistaken belief: both are indeed morally blameless and entitled to a complete acquittal. This is not the 101

Baron, ‘Justifications and excuses’, at 402. See also Fletcher, ‘Domination’, at 563–4; Moore, Placing Blame, at 483 n 6; RF Schopp, Justification Defenses and Just Convictions (1998), at 38. 103 Some of the examples utilized by Tadros in defence of the reasons theory in Criminal Responsibility (at 280–90)—a mistaken belief that petrol is actually gin; visiting the doctor in the mistaken belief that one has meningitis—tend to obscure this point. 102

The Classification of Defences

31

same as saying that both are justified. Conceptually, Devinder’s behaviour has the gist of an excuse: he has done something unacceptable (killed an innocent person who was not posing a threat to the life of anyone) but we do not blame him for it. It does not have the gist of a justification. As Hurd puts it: . . . when a person kills under the reasonable but mistaken impression that the victim was an aggressor, the law ought not to gloss over the fact that she killed an innocent person by declaring her a justified self-defender. It ought instead to craft a special excuse doctrine to exonerate such a non-culpable person of the wrong she has done.104

2.4.3 The deeds theory Ashley, a police officer, has shot and wounded Danny, who was running away from a forest carrying stolen goods. Unknown to him, Danny had previously been convicted twice for theft and, under the law of England at the time, a police officer would be justified in using such force against him. Ashley had no knowledge of Danny’s previous convictions. Ashley was unknowingly justified.105 To return to our earlier example, Mike sees his hated enemy, Ken, attacking a woman in the street with a knife. Thinking quickly, he realizes that this is an opportunity to cause harm to Ken. He rushes over and stabs Ken in the chest. Ken dies but the woman’s life is saved. Unlike Ashley, Mike has knowledge of the justificatory circumstances, but he does not act for the justificatory reason. For the deeds theory, the mere existence of justificatory circumstances is sufficient to ground a claim of justification. Whether the accused has knowledge of these circumstances or acts for the appropriate reason106 is irrelevant to whether she is justified. Those subscribing to the deeds theory would describe both Ashley and Mike as justified, despite the fact that neither acted for the justificatory reason (and indeed Ashley had no knowledge of the potentially justificatory circumstances). In Gardner’s terms, the deeds theory structures justification exclusively around guiding reasons and pays no regard to explanatory reasons.

2.4.3.1 The development of the deeds theory The leading proponent of the deeds theory is Paul Robinson. The earliest statement of his position came in an article107 written while he was still a 104 H Hurd, ‘Justification and excuse, wrongdoing and culpability’ (1999) 74 Notre Dame Law Review 1551–73, at 1565. See also Fletcher, ‘Domination’, at 567; P Westen and J Mangiafico, ‘The criminal defense of duress: a justification not an excuse and why it matters’ (2003) 6 Buffalo Criminal Law Review 833–950, at 877–8. 105 Based on the facts of R v Dadson (1850) 4 Cox CC 358. 106 As has already been noted, these are not the same. 107 PH Robinson, ‘A theory of justification: social harm as a prerequisite for criminal liability’ (1975) 23 UCLA Law Review 266–92.

32

Killing in Self-Defence

student in which he argues that the unknowingly justified accused should receive a complete defence because no harm has occurred as a result of her actions. Thus in Ashley’s case, Robinson would have argued that the police officer was merely engaged in legitimate sanctions. In Mike’s case, a potentially lethal attack on a woman was prevented. If Ashley and/or Mike were punished, for Robinson (writing in 1975), this would be to punish them for their mental state alone: . . . the criminal law seeks to prevent harmful results rather than to punish evil intent that produces no harm . . . If the criminal law is extended to punish bad intent alone or the mere possibility of harmful conduct, it goes beyond its accepted role, appears unfair and overreaching, and ultimately loses its credibility and integrity.108

Robinson’s article attracted widespread criticism. Prominent among the critics was Fletcher, who pointed out that it is not the case that no harm has occurred where the accused is unknowingly justified.109 The elements of the offence have been satisfied. Ashley and Mike have both intentionally killed a human being. Unless one attaches no value whatsoever to human life, some harm has clearly occurred, albeit a harm that may be justified.110 Robinson subsequently accepted this criticism and shifted his position. By 1982, he no longer claimed that no harm occurs when the accused is unknowingly justified. Instead, he suggested, some harm is caused by the ‘justified’ behaviour but this is outweighed by the beneficial consequences.111 At this point, Robinson still appears to imply that the unknowingly justified accused should not be subject to any punishment: ‘The justified actor engages in conduct that is not culpable because its benefits outweigh the harm or evil of the offence.’112 Eventually, Robinson does concede the point that it would be unjust to allow the unknowingly justified accused to escape punishment entirely. By 1996, he developed the novel argument that the accused should be subject to a level of liability in line with that for an impossible attempt.113 The accused who engages in an impossible attempt attempts to commit a crime in circumstances where this would have been impossible. The unknowingly justified accused attempts to engage in an unjustified action in circumstances where this was impossible (due to the existence of justifying circumstances). For 108 PH Robinson, ‘A theory of justification: social harm as a prerequisite for criminal liability’ (1975) 23 UCLA Law Review 266. 109 Fletcher, ‘The right deed for the wrong reason: a reply to Mr Robinson’. See also MT Funk, ‘Justifying justifications’ (1999) 19 OJLS 631–47; CKY Lee, ‘The act-belief distinction in self-defense doctrine: a new dual requirement theory of justification’ (1998) 2 Buffalo Criminal Law Review 191–247; Simester, ‘Mistakes in defence’; Horder, Excusing Crime, at 101. 110 Dillof, ‘Unravelling unknowing justification’, at 1591. 111 Robinson, ‘A systematic analysis’, at 213. 112 Robinson, ‘A systematic analysis’, at 203, emphasis added. 113 In ‘Competing theories’.

The Classification of Defences

33

Robinson this is sufficiently conceptually similar to the impossible attempt that he feels a similar level of liability should attach. Yet in his 1996 article, Robinson makes it quite clear that he still agrees with the fundamental premise of the deeds theory: where the accused commits a crime in potentially justifying circumstances, she should be described as justified regardless of her knowledge or her reason for acting.114 Justification, for Robinson, attaches to deeds alone (or, in Gardner’s terms, to guiding reasons). In this he finds support from Byrd,115 Hurd,116 Schopp,117 and, more recently, Westen and Mangiafico.118

2.4.3.2 Robinson’s arguments for the deeds theory Robinson supports the deeds theory for three reasons.119 First, he states, pointing to his own research,120 it corresponds with public opinion. In this research, a number of hypothetical scenarios were constructed, including cases where the accused fulfilled the definition of a completed crime (with no apparent defence); fulfilled the definition of an attempted crime (again with no apparent defence); and fulfilled the definition of a completed crime but was unknowingly justified.121 Members of the public were asked to rate each scenario in terms of the liability they would attach to the accused.122 The liability attached to the unknowingly justified accused was, on average, lower than that attached to the accused who had committed the completed offence and corresponded most closely with that attached to the attempt scenario. But some doubt must be cast on Robinson and Darley’s research for two reasons. First, their sample was small and unrepresentative. Only fortyeight people were asked to rate the scenarios, of which half were students. 114

Robinson, ‘Competing theories’, at 48. SB Byrd, ‘Wrongdoing and attribution: implications beyond the justification-excuse distinction’ (1987) 33 Wayne Law Review 1289–342, at 1320. 116 Hurd, ‘Justification and excuse’, at 1551. 117 Schopp, Justification Defenses and Just Convictions, at 38. 118 Westen and Mangiafico, ‘The criminal defense of duress’, at 878. 119 ‘Competing theories’, at 48. 120 PH Robinson and JH Darley, ‘Testing competing theories of justification’ (1998) 76 North Carolina Law Review 1095–136. 121 Scenario 1 (the completed crime) ran as follows: Jake is angry with his neighbour over a dispute about the use of water from a creek that the two share. When he knows that his neighbour is away, he sets fire to the neighbour’s three acres of corn. Because the field is bounded on all sides by a dirt road, and there is no wind, it is clear that the burning will create no danger beyond that of destroying the corn. The fire destroys the entire crop. Scenario 2 (the attempt) was a variation on scenario 1 in which the neighbour returns and puts out the fire before any damage to crops occurs. Scenario 7 (the unknowing justification) was a variation on scenario 1 in which, unbeknown to Jake, lightning has started a fire in a nearby field and Jake’s burning creates a firebreak preventing the lightning fire from reaching the local town. See Robinson and Darley, ‘Testing competing theories’, at 1110. 122 The subjects could choose to assign no liability whatsoever or could assign punishment on a scale ranging from 1 (1 day’s imprisonment) to 11 (the death penalty). 115

34

Killing in Self-Defence

The claim that the research represents public opinion in any wider sense is therefore highly questionable.123 Second, there is another explanation for the findings. It may be that, when faced with a professional researcher presenting a series of different scenarios, the respondents simply did what they thought the researcher expected. The unknowing justification scenario is clearly factually different from the completed crime scenario and respondents may have taken this as a cue to attribute a lower liability rating, not because they genuinely believed that the accused was less culpable, but because they thought that this was the ‘correct’ answer. One also wonders whether the same results would have occurred if the scenario involved killing a human being, instead of damaging property. Substitute ‘Jake shot and killed his neighbour because he was angry’ for ‘Jake burnt down his neighbour’s field because he was angry’ (and make appropriate adjustments to the subsequent scenarios) and it may be that respondents would have been rather less inclined to attribute lower liability to the unknowingly justified accused. These points are speculative. More importantly, even if the research genuinely does indicate public support for the deeds theory, one has to question whether the criminal law should always follow public opinion. Robinson’s second reason for supporting the deeds theory is that it has improved theoretical clarity compared to the reasons theory. According to Robinson, the deeds theory lays bare the distinctions relevant to determining liability in cases of mistake whereas the reasons theory obscures them. By attaching justifications exclusively to acts (with no reference to the mental state of the accused), as the deeds theory does, there are clear implications for rights of resistance and assistance. The ‘victim’ in an unknowing justification scenario (Danny in the Ashley scenario or Ken in the hated enemy scenario) would have no right to resist the force directed towards him because that force is justified (even though the perpetrator is unaware of it). A third party would have a right to ‘assist’ and indeed would be justified in doing so. Robinson contrasts this with the confusion generated by the reasons theory. Here, the mistaken accused is labelled as justified, which would imply that her victim has no right to resist and third parties have rights of assistance. This would clearly be unjust124 and none of the proponents of the reasons theory would support such a proposal. Instead the reasons theory either treats the mistakenly justified accused as a ‘special case’ where the justification carries with it no rights of resistance or assistance125 or treats such an anomaly as evidence that we should not even try to make a clear 123 Robinson and Darley do not say whether the students were law students. If they were, then they may well have been exposed to Robinson’s views and the value of the results is even more questionable. 124 Think, for example, of the scenario at the start of the chapter where Devinder kills the innocent police officer attempting to make a legitimate arrest. 125 Model Penal Code §§ 3.04(1) and 3.11(1).

The Classification of Defences

35

distinction between justifications and excuses so far as rights of resistance and assistance are concerned.126 In this Robinson has a point. Granting the accused a justification defence purely on her belief that she is justified does obscure the issue of rights of resistance and assistance. In this, the deeds theory has improved conceptual clarity over the reasons theory. This in itself, however, does not make the theory convincing. Robinson’s third reason for supporting the deeds theory is that it improves the criminal law’s rule-articulation function. Attaching justifications to deeds alone better communicates to the public the conduct rules of criminal law by focussing on the types of behaviour that are permissible. In this I would agree that the deeds theory is an improvement on the reasons theory. It has already been suggested that justifying anyone who believes in the existence of justifying circumstances, regardless of whether they are mistaken, does not provide helpful guidance on the type of conduct the criminal law permits. But equally, it is not clear that the rule-articulation function of the criminal law is lost under the dual requirement theory, whereby the accused is required to act with the justificatory purpose in mind. Here the criminal law is saying to the public: in exceptional circumstances (such as when you are faced with an attacker) we permit you to kill another human being. But we do not do this lightly and it is only permissible when you do so for the primary purpose of defending your own life.

2.4.3.3 The argument against the deeds theory None of these points is a convincing argument in itself for dismissing the deeds theory. Indeed, some writers, while indicating that they do not support the deeds theory, have been unable to come up with a convincing argument against it other than it is somehow ‘wrong’.127 Others have dismissed it on the basis that it leads to the ‘logical contradiction’ that two individuals who simultaneously attack each other out of spite (or some other bad motive) would paradoxically both be justified and unjustified.128 It is suggested here that the most convincing reason why the deeds theory should be rejected is because it does not make sufficient conceptual distinction between offences and defences. The unknowingly justified accused has committed an offence, as defined by the actus reus and mens rea. If we accept that the provisions of the criminal law are (generally) intended to prevent 126

Greenawalt, in ‘Perplexing borders’. Corrado argues that the deeds theory is simply too bizarre to win acceptance (M Corrado, ‘Notes on the structure of a theory of excuses’ (1992) 82 Journal of Criminal Law and Criminology 465–97). Hogan calls the deeds theory ‘absurd’ (B Hogan, ‘The Dadson principle’ [1989] Crim LR 679–86, at 680). Greenawalt, in ‘Distinguishing justifications from excuses’, argues that the deeds theory ‘offends one’s intuitive moral sense’ (at 102). 128 RL Christopher, ‘Unknowing justification and the logical necessity of the Dadson principle in self-defence’ (1995) 15 OJLS 229–51, at 243. 127

36

Killing in Self-Defence

harm, the unknowingly justified accused has caused some harm (usually) to a third party. She has engaged in wrongdoing ‘bad’ enough for it to be prohibited by the criminal law and has done so with the required mental state for culpability. Given that this is the case, whether or not a defence is available is an entirely separate enquiry.129 There is no reason why the principles of liability governing offences must automatically govern defences. Indeed, I would suggest that we should be very careful about the circumstances in which we allow defences. Defences should only be provided to those who, given the harm they have caused, are appropriate candidates for exculpation.130 Against this, Robinson has argued that the existence of a conceptual distinction between offences and defences does not in itself explain why mistake about an element of a defence should be treated differently to mistake about an element of the offence.131 At times, even Fletcher, who has perhaps been Robinson’s strongest critic, seems to accept this: Admittedly, there might be stronger arguments for a position that seems so deeply entrenched in the world’s legal culture. Generating a convincing rationale for requiring intent in cases of justification reminds one of other practices of the criminal law that are widely shared and intuitively accepted—for example, punishing completed crimes more severely than attempts, and retributive punishment generally—but for which theoreticians have yet to generate a compelling justification.132

But Fletcher underestimates the strength of his own argument. Robinson is correct: simply stating that offences and defences are conceptually different is not in itself an argument against the deeds theory. But the argument that, where the accused has caused the harm an offence is designed to prevent and has done so with a blameworthy mental state, she should only be granted a defence when she is an appropriate candidate for exculpation is a compelling argument for rejecting the deeds theory.133 Robinson’s suggestion that the accused, although justified, should still be punished on the basis of attempt liability does not overcome this point.134 129

See Section 2.2 above. This argument would not be accepted by someone such as Williams, who sees no conceptual distinction between offences and defences. This is not the view taken here: see 131 Section 2.2 above. In ‘Competing theories’, at 57. 132 GP Fletcher, ‘The nature of justification’, in S Shute, J Gardner, and J Horder (eds), Action and Value in Criminal Law (1993) 175–86, at 179. 133 The deeds theory appears to have been rejected in R v Dadson (1850) 4 Cox CC 358, where the court held that the force used by Dadson was not justified, given that he did not know of the existence of the potentially justificatory circumstances. 134 It also fails in another context: where the accused deliberately provokes an attack, with a view to using ‘defensive’ force to kill the attacker. The deeds theory would hold that the accused is justified and here Robinson cannot justify punishment by equating the situation with an impossible attempt. In fact, Robinson’s view is that the accused should be punished—on the basis of his ‘scheming intention’—but this starts to sound very close to support for the dual requirement theory (PH Robinson, ‘Causing the conditions of one’s own defence: a study in the limits of theory in criminal law doctrine’ (1985) 71 Virginia Law Review 1–62, at 31). See Chapter 6 for further discussion. 130

The Classification of Defences

37

His argument for attributing attempt liability is based on his view that a lesser harm has occurred in a case of unknowing justification and the law should reflect this by attributing a lesser degree of punishment.135 There is no clear reason, though, why this should be the case. The unknowingly justified accused has killed a human being for a bad motive, albeit a human being who was about to cause harm. She is no different, morally, from the accused who has killed in the absence of justifying circumstances.136 That some ‘good’ came out of her unjustified conduct is not relevant, in my view, to her liability. In this respect, Fletcher has criticized Robinson for failing to distinguish between justified acts and just events.137 It is also worth questioning whether, as Robinson claims, a lesser harm has always occurred in cases of unknowing justification. Some difficulty arises in relation to innocent aggressors.138 To take a hypothetical example:139 Les kills Chesney (a 3-year-old child) out of sexual pleasure. Unknown to him, Chesney is playing with a gun and is about to shoot him. If one accepts that self-defence is justified against innocent aggressors,140 Robinson’s analysis would justify Les. Yet it is difficult to see how the harm involved is less than where no potential justification existed. Regardless of whether Chesney was about to shoot him, Les has killed an innocent child out of sexual pleasure. Adding the fact that Chesney was about to shoot Les does not change this. As Simester argues, Robinson’s approach seems to imply that the interests of the ‘victim’ are not worthy of protection.141

2.5 The Practical Consequences of the Competing Theories The view of the structure of justifications taken here is, then, that justification requires both the existence of justifying circumstances and that the accused acted for the justificatory purpose. Applying this to self-defence, the accused claims the justification form of the defence when she correctly perceives the circumstances of an attack and acts for the purpose of defending herself. When the accused is mistaken about the existence of an attack, she is claiming an excuse.142 135 It is assumed here for the sake of argument that attempt liability will attract a lesser punishment than liability for a completed offence. This is certainly the assumption Robinson 136 makes. A Ripstein, Equality, Responsibility and the Law (1999), at 192. 137 138 Fletcher, ‘The right deed for the wrong reason’, at 293. Defined in Chapter 1. 139 Based on Simester, ‘Mistakes in defence’, at 304. 140 Cf LA Alexander, ‘Justification and innocent aggressors’ (1987) 33 Wayne Law Review 1177–89. 141 Simester, ‘Mistakes in defence’, at 304. See also Horder, Excusing Crime, at 101. 142 The issue of whether her mistaken belief must be a reasonable one is dealt with in Chapter 9, where it will be argued that a reasonable belief is required.

38

Killing in Self-Defence

Does the choice between the deeds theory, reasons theory, and dual requirement theory matter in practice? One answer is that it can certainly make a difference to the unknowingly justified accused, as under the deeds theory of justification she will be granted a defence, whereas under the reasons or dual requirement theory she will not.143 Whether the choice between the three approaches has any implications for the mistaken belief scenario is less clear.144 Here, the choice of one theory over another only makes the difference between whether the accused has a justification or an excuse defence; in both instances, she would be acquitted (assuming the conditions of the defence were made out). One response might be that, as Gardner suggests, the accused would prefer to be justified rather than excused, because of the more favourable way in which this reflects on her. Indeed, Gardner professes ‘astonishment’ that anyone would doubt this: Criminal lawyers, in particular . . . tend to take it for granted that any doctrine that serves to acquit the accused, and therefore to avert the adverse normative consequences of her action, is as good as any other so far as the accused is concerned. I have always found this an astonishing assumption, which implies that nobody who is tried in the criminal courts has, or even deserves to have, any self-respect . . . The self-respecting person aspires to live up to the proper standards for success in and fitness for the life she leads, and holds herself out to be judged by those standards . . . She wants it to be the case that her actions were not truly wrongful, or if they were wrongful, that they were at any rate justified, or if they were not justified, that they were at any rate excused.145

Given that, at present, common law criminal justice systems make no pronouncement as to whether any successful defence was a justification or an excuse, the accused is, however, likely to be entirely unaware of this fact. Another way in which it has been suggested that the distinction matters is in relation to rights of resistance and assistance. It has been proposed that whereas an excused act can justifiably be resisted, there is no such right to resist a justified act. In other words, no two parties can be justified in the same conflict.146 Thus, Leanne, who has her mobile phone forcibly taken from her by Frankie in order to phone an ambulance for the dying victim of a road traffic accident, cannot permissibly resist what is, prima-facie, a robbery.147 By contrast, the victim of an excused attack or robbery (where, 143 For cases where the deeds theory was rejected and a defence was withheld from the accused because he did not act for the justificatory reason, see R v Dadson (1850) 4 Cox CC 358; Dawson v Dickson 1999 JC 315; R v Thain [1985] NI 457. 144 RA Duff, ‘Theorizing criminal law: a 25th anniversary essay’ (2005) 25 OJLS 353–67, at 361; Baron, ‘Justifications and excuses’, at 402. 145 Gardner, ‘The gist of excuses’, at 590. 146 Byrd, ‘Wrongdoing and attribution’, at 1332; Fletcher, Rethinking Criminal Law, at 760; Robinson, Criminal Law Defenses, at § 36(a). 147 It might be the case that someone resisting justified conduct could be granted an excuse defence but the suggestion here is that she would not be justified in resisting.

The Classification of Defences

39

for example, the attacker is acting under the excuse of duress) is permitted to resist that attack. Likewise, it has been suggested that, whereas it is permissible for a third party to assist someone in performing a justified act, this permission does not extent to assistance in performing an excused act.148 Thus, it is permissible for Norris to come to the aid of Emily, who is attempting to repel an attacker in legitimate self-defence, but it would not be permissible for Norris to assist Blanche, who is undertaking a bank robbery under duress. At least in relation to rights of assistance, however, whether or not these consequences follow depends in turn on the particular theory of justification preferred. If all that is required for a claim of justification is the existence of justificatory circumstances (the deeds theory), then it is true that anyone who assists the justified accused will be similarly justified, regardless of her own knowledge or motivation. Under the dual requirement theory this would not necessarily be the case, as the actor who assists would not be justified unless she acts for the justificatory reason.149 Whether justifications automatically preclude a right of resistance is a separate issue. If one actor is justified, is it the case that another actor on the ‘opposite’ side of the conflict cannot be similarly justified? This is a more convincing proposition for the dual requirement theory and it may well be that there is no right to resist justified action (although perhaps one can be excused for doing so).150 It has also been suggested that justifications function as a guide to acceptable behaviour in a way excuses do not.151 A justification defence sends out the message that the conduct in question is acceptable and that anyone who engages in it will not be convicted of a criminal offence. The accused, faced with a situation in which she must contemplate breaking the law, is entitled to rely on the existence of justification defences to guide her conduct. Thus when the court grants an acquittal on the basis of selfdefence, it is saying to the community: ‘it is acceptable to use proportionate force to save your own life where you have no alternative means of escape from an attacker—you will not be convicted of a criminal offence if you do so’. Excuses, it is said, do not function in this way. Here, the message sent out is that we disapprove of the accused’s conduct but, because we do not blame her for it, it would be unjust to burden her with the stigma of a criminal 148 Byrd, ‘Wrongdoing and attribution’, at 1332; Fletcher, Rethinking Criminal Law, at 760; Robinson, Criminal Law Defenses, at § 36(a). 149 Or, under Dillof’s version of the theory, ‘has regard’ to the justificatory circumstances before acting. 150 Although see Berman, ‘Justification and excuse, law and morality’, at 64, who contends that whether or not rights of resistance and assistance attach to particular defences is not a consequence of any distinction between justifications and excuses but is simply a policy issue. 151 A Ashworth, Principles of Criminal Law (5th edn, 2006), at 136; J Dressler, ‘Justifications and excuses: a brief review of the concepts and the literature’ (1987) 33 Wayne Law Review

40

Killing in Self-Defence

conviction. Excuses are not intended to function as a guide to acceptable behaviour and, as Gardner has pointed out,152 to attempt to benefit from an excuse by being guided by it is generally to forfeit that excuse. Thus, the accused faced with a choice between robbing a bank and suffering serious personal injury who chooses to rob the bank because she knows she will be able to rely on the defence of duress in any subsequent criminal proceedings should not, in theory at least, be granted that defence.153 There is a practical problem with the notion that justifications function as a guide to behaviour. As we have already noted, the signal sent out by the major common law legal systems is the same regardless of whether the accused is justified or excused: an acquittal. This has led Colvin to suggest that the law would be better off not even attempting to carve a distinction between justifications and excuses as the subtlety of the distinction is entirely lost on the public.154 The fact that both justifications and excuses lead to an acquittal carries the additional danger of what Simon Gardner has termed ‘quasi-justificatory effect’.155 By this he means that when the public learn of a case in which the accused was acquitted, they will assume that her conduct was judged to be acceptable, even when the defence was an excuse. To address this issue, it has been suggested that adjustments should be made to the verdict system to improve the conduct-guiding function of justification defences.156 Setting these practical difficulties aside, whether one accepts that justifications have a conduct-guiding function depends on one’s convictions about the way in which justification defences are structured. If all that is required for a justification is the existence of justificatory circumstances (the deeds theory), then justifications do play a fundamental conduct-guiding role—by setting out the exact nature of the circumstances necessary before one can claim a justification—and the public should be able to rely on this advice in making conduct decisions. If one argues, however, that to claim a justification, the accused also has to act for the justificatory reason, then the role justifications play in guiding conduct is more complex. On this understanding of justifications, to say ‘I killed this person solely because I knew it was 1155–75, at 1169; Fletcher, Rethinking Criminal Law, at 760; D Horowitz, ‘Justification and excuse in the program of the criminal law’ (1986) 49 Law and Contemporary Problems 109–26, at 112; Berman, ‘Justification and excuse, law and morality’, at 33. 152 Gardner, ‘The gist of excuses’, at 597. 153 In practice it may be difficult to distinguish, ex post, between this accused and another who, through extreme fear, and with no thought of any subsequent criminal proceedings, gives 154 in to the request to rob the bank. Colvin, ‘Exculpatory defences in criminal law’. 155 S Gardner, ‘Instrumentalism and necessity’ (1986) 6 OJLS 431–8, at 433. Gardner variously terms this phenomenon ‘quasi-justificatory drift’ and ‘quasi-justificatory effect’. 156 GP Fletcher, ‘The psychotic aggressor: a generation later’ (1993) 27 Israel Law Review 227–46, at 246; PH Robinson, PD Greene, and NR Goldstein, ‘Making criminal codes functional: a code of conduct and a code of adjudication’ (1996) 57 Journal of Criminal Law and Criminology 304–65.

The Classification of Defences

41

permitted by the law to kill anyone who attacks me’ would not be enough to claim a justification defence.157 Related to this, it is not necessarily the case that excuses can never act in a conduct-guiding manner. Certainly to say, ‘I did this because I knew that other people had received a defence in the same circumstances’ would not be an excuse at all. But it may be that excuses can guide conduct in a different way. Take the example of the excuse form of self-defence based on a reasonable but mistaken belief that one is under attack. In requiring the belief of the accused that she is facing an attack to be reasonable,158 the criminal law sends out the message that we should all take reasonable care to check the accuracy of our beliefs before acting rashly in self-defence. In this way, the excuse defence plays a guiding role as to the responsibilities of the accused to investigate the truth or falsity of her beliefs before acting. Even if one is not convinced by any of these arguments, it might be said that the distinction matters because it is important to be clear about why we are granting someone a defence. When the accused’s conduct and mental state satisfies the definition of a criminal offence, it is important to think carefully before allowing her a complete defence and to be clear about the reasons why we are doing so. The criminal law should state clearly why we are being punished and conversely why we are exempted from punishment. The justification and excuse distinction, if nothing else, can potentially assist in achieving such moral clarity in the criminal law.159 157 J Horder, ‘Killing the passive abuser: a theoretical defence’, in S Shute and AP Simester (eds), Criminal Law Theory: Doctrines of the General Part (2002) 283–97, at 285–6. 158 As most common law jurisdictions do: see Chapter 9. 159 As Austin famously wrote: ‘words are our tools, and, as a minimum, we should use clean tools: we should know what we mean and what we do not, and we must forearm ourselves against the traps that language sets us’ (Austin, ‘A plea for excuses’, at 24).

This page intentionally left blank

3 The Justification of Self-Defence 3.1 Introduction This chapter attempts to address the question of why killing in self-defence is justified.1 It is a topic that, at least in legal philosophy, has attracted little attention, with its permissibility often being presumed rather than explained. For example, in Glover’s Causing Death and Saving Lives,2 one of the most detailed examinations of the circumstances in which killing is permissible, self-defence is mentioned only twice, and even then only in passing,3 without any detailed consideration of the basis upon which it might be justified. This is in contrast to the other issues addressed in the book, such as abortion, infanticide, suicide, euthanasia, execution, and war, all of which receive substantial treatment. This is a troubling omission. Upon closer examination it is by no means obvious why self-defensive killing is justified. The killing of an aggressor in self-defence may well be an intentional killing4 and it is a self-interested killing: precisely the sort of behaviour that, prima facie, should be of concern to the criminal law. The most obvious response is that the accused is permitted to kill an aggressor in self-defence because she is entitled to protect her own life from unjust aggression. Indeed, HLA Hart explains the permissibility of self-defensive killing in something approximating these terms: ‘Killing in self-defence is an exception to a general rule making killing punishable; it is admitted because the policy or aims which in general justify the punishment of killing (eg. protection of human life) do not include cases such as this’.5 Hart sees no need for further explanation but his account is deficient. As Epps points out, if, as Hart asserts, protection of human life is a goal of the criminal law, it has failed, because a human being (the aggressor) is dead: ‘An act of violence has been committed, human life has been taken by another 1 The focus here is on acts of self-defence that result in the death of the aggressor. As Chapter 1 explained, this is because this is the most difficult type of self-defensive force to justify. 2 3 J Glover, Causing Death and Saving Lives (1977). ibid 39 and 87. 4 Not all self-defensive killings will be intentional, but there will be at least some occasions on which it would be absurd to conclude that the intention of the accused was not to kill. Those who subscribe to the doctrine of double effect as an explanation of why self-defensive killing is permissible might, however, disagree (see Section 3.4 below). 5 HLA Hart, Punishment and Responsibility: Essays in the Philosophy of Law (1968), at 13 (emphasis added).

44

Killing in Self-Defence

human being, and a legal system sincerely concerned with protecting human life should explain the exception in more rigorous terms than Hart admits’.6 Thus any convincing explanation of the permissibility of self-defensive killing must explain why it is that the victim is permitted to take the life of the aggressor in order to protect her own life. After all, in a legal system that holds the equality of lives to be a fundamental value, the aggressor has a right to life just as her victim does.7 In response, it has been argued that an aggressor who threatens the life of another human being forfeits her own right to life by virtue of her culpability.8 Whether or not one agrees with this, it is more difficult to see how it might hold in relation to an innocent aggressor9 or a passive threat,10 who is not culpable for the threat she poses. Of course, one response might be that self-defence is not a justification defence at all, but merely an excuse.11 This is precisely the view taken by Finkelstein,12 who finds it difficult to accept that such a self-interested act as self-defence can be thought of as justified.13 Alternatively, if the forfeiture account (or something similar) is accepted in relation to culpable aggressors, one might take the view that killing a culpable aggressor is justified but killing an innocent aggressor or passive threat is not.14 But instinctively this seems wrong. Killing a life-threatening aggressor, innocent or otherwise, feels intuitively like a permissible reaction. Likewise, killing an innocent bystander, who poses no threat whatsoever, feels intuitively wrong (although it may be excusable). Thus it is the aim of this chapter to put forward an account of the permissibility of self-defensive killing that explains why it is permissible to kill a culpable aggressor, an innocent aggressor, and a passive threat, but why it is not permissible to kill an innocent bystander. If a plausible theory of this nature cannot be found, however, then we may have to return to the notion that self-defence is merely an excuse (or at least to the notion that it is an excuse where the aggressor or threat is not culpable).15 6 G Epps, ‘Any which way but loose: interpretive strategies and attitudes toward violence in the evolution of the Anglo-American retreat rule’ (1992) 55 Law and Contemporary Problems 303–31, at 304. 7 C Ryan, ‘Self-defense, pacifism and the possibility of killing’ (1983) 93 Ethics 508–24, at 510. 8 Forfeiture accounts of the permissibility of killing in self-defence are considered in more 9 detail in Section 3.5.5 below. Defined in Chapter 1. 10 Defined in Chapter 1. 11 Where the accused uses self-defensive force in the mistaken but reasonable belief it is necessary, it has already been suggested that she is excused and not justified (see Chapter 2). 12 CO Finkelstein, ‘Self-defense as a rational excuse’ (1996) 57 University of Pittsburgh Law Review 621–49. 13 Cf HA Pendleton, ‘A critique of the rational excuse defence: a reply to Finkelstein’ (1996) 57 University of Pittsburgh Law Review 651–76. 14 LA Alexander, ‘Self-defense, justification and excuse’ (1993) 22 Philosophy and Public Affairs 53–66; M Otsuka, ‘Killing the innocent in self-defense’ (1994) 23 Philosophy and Public Affairs 74–94; D Wasserman, ‘Justifying self-defense’ (1987) 16 Philosophy and Public Affairs 356–78. 15 It is not questioned here that, regardless of whether it is a justification, self-defence can operate as an excuse. A person cannot be blamed for defending herself from an attack that, say,

The Justification of Self-Defence

45

It is against this background that the various approaches to explaining the permissibility of killing in self-defence are examined. It will be argued that there is a convincing justification for self-defensive killing and that it can be found in accounts based on rights. Specifically, it will be argued that selfdefensive killing is justifiable on the basis that all human beings have a right to life and therefore a right to protect themselves from an aggressor who threatens to deprive them of life. The reason why it is permissible to kill the aggressor is that the aggressor forfeits her right to life by virtue of her conduct in becoming an unjust immediate threat to the life of another. In this, the argument draws heavily, although with some modification, on the work of Suzanne Uniacke and Judith Jarvis Thomson.16

3.2 Consequentialist Approaches One of the most notable contemporary debates in legal and philosophical theory has been between those who espouse consequence-based theories and those who espouse rights-based theories.17 It is perhaps unsurprising, then, that there is a body of literature in which the permissibility of self-defensive killing is justified on the basis of appeals to consequentialism. Consequentialism can be defined as ‘the doctrine that the moral value of any action always lies in the consequences, and it is by reference to their consequences that actions, and indeed such things as institutions, laws and practices, are to be justified if they can be justified at all’.18 The term consequentialism is sometimes used interchangeably with the term utilitarianism but they are not synonymous. Consequentialism is a broad term for all approaches that justify actions in terms of their consequences. Utilitarianism is a specific type of consequentialism, concerned specifically with maximizing happiness or with minimizing unhappiness.19 Consequentialist approaches are based on the notion that the consequences of killing an aggressor in self-defence are preferable to the consequences of allowing the victim of aggression to be killed. That is, when faced with a choice between the death of the aggressor and the death of the victim, the aggressor’s death is the preferable consequence. threatens death or serious bodily injury. What we are searching for is a basis upon which it can, in addition to being excused, be described as justified. On this, see R Segev, ‘Fairness, responsibility and self-defense’ (2004–5) 45 Santa Clara Law Review 383–460, at 386. 16 S Uniacke, Permissible Killing: The Self-Defence Justification of Homicide (1994); S Uniacke, ‘In defence of Permissible Killing: a response to two critics’ (2000) 19 Law and Philosophy 627–33; S Uniacke, ‘Rights and relativistic justifications: replies to Kasachkoff and Husak’ (2000) 19 Law and Philosophy 645–7; JJ Thomson, Rights, Restitution and Risk: Essays in Moral Theory (1986); JJ Thomson, ‘Self-defense’ (1991) 20 Philosophy and Public Affairs 283–310. 17 L Lloyd and MDA Freeman, Lloyd’s Introduction to Jurisprudence (6th edn, 1994), at 380. 18 B Williams, ‘A critique of utilitarianism’, in JJC Smart and B Williams (eds), Utilitarianism: 19 For and Against (1973) 77–150, at 79. Williams, ‘A critique of utilitarianism’, at 79.

46

Killing in Self-Defence

For a consequentialist account of self-defence to work as a justification, there must be some means by which the death of the aggressor is seen as the preferable consequence. The most common way in which this is accomplished is to state that, by virtue of her moral blameworthiness, the aggressor renders her own life less valuable (the ‘discount’ approach). The discount approach can be seen in the work of Montague.20 For Montague, those who intentionally pose a lethal threat to others place themselves in a situation of ‘moral weakness’ in relation to their victim so that when a choice is forced between lives, the victim’s life is preferred, on the basis that it is worth more than that of the aggressor. A similar argument has been made by Robinson, who explains the permissibility of killing in selfdefence in consequentialist terms, with the culpability of aggressors ‘seen as discounting the value of their lives in the balance’.21 Later in his text, however, Robinson backs away from an explanation based on the discount approach alone.22 Other theorists have also expressed discomfort with the discount approach on the basis that it is wrong to devalue lives in this way. Wasserman,23 for example, points out that the idea that an aggressor’s life can be discounted through culpability sits uneasily with the premise of treating all human lives as being of equal value.24 Wasserman also makes the point that, if the aggressor’s life is discounted on the basis of culpability, then the victim’s life might be discounted in the same way if she is a ‘bad’ person or has been engaged in acts of violence in the past or is likely to do so in the future. The idea that it is only permissible to engage in self-defence if the victim has led a more blame-free life than the aggressor seems absurd, but this appears to be the implication of the notion that the culpability of the aggressor lessens the value of her life.25 Even if one agrees in principle that it is acceptable to engage in the balancing of lives, another implication of the discount approach is that selfdefensive killing might only be justified where the aggressor is, indeed, culpable. A consequentialist argument based on the culpability of the aggressor cannot justify killing innocent aggressors, such as children or the insane,26 or passive threats, such as the roped mountaineer who threatens to pull her companion 20 P Montague, ‘Self-defense and choosing between lives’ (1981) 40 Philosophical Studies 207–19; P Montague, ‘The morality of self-defense: a reply to Wasserman’ (1989) 18 Philosophy and Public Affairs 81–9. 21 PH Robinson, Criminal Law Defenses (1984), at § 131(a). See also Segev, ‘Fairness, responsibility and self-defense’, at 395, who bases his conclusion on the ‘responsibility’ of the aggressor for the conflict, although his account is a more nuanced one than that of Robinson. 22 See later passages in § 131(a) of Criminal Law Defenses (and subsequent discussion in this 23 chapter). Wasserman, ‘Justifying self-defense’, at 358. 24 See also S Wallerstein, ‘Justifying the right to self-defense: a theory of forced consequences’ (2005) 91 Virginia Law Review 999–1035, at 1003. 25 Wallerstein, ‘Justifying the right to self-defense’, at 1003–4. 26 SH Kadish, ‘Respect for life and regard for rights in criminal law’ (1976) 64 California Law Review 871–901, at 882.

The Justification of Self-Defence

47

off the mountain.27 This in itself is not necessarily a reason to reject the consequentialist approach. It may be that while consequentialism can justify killing culpable aggressors, killing innocent aggressors and passive threats is merely excused or is justified in some other way. Montague, for example, accepts that consequentialist arguments do not justify killing innocent aggressors, but suggests that this can be justified on the basis of personal partiality.28 Alexander29 has consistently argued that while killing a culpable aggressor is always justified,30 killing an innocent aggressor or a passive threat may or may not be justified, depending on whether other factors weigh in its favour. He lists such considerations as relative fault, the relative numbers of deaths, and the relative ages of the parties. Where the balance of factors does not weigh in favour of justification, the victim may be excused for killing in selfdefence.31 It has been suggested32 that a more serious difficulty with the discount approach is that, unless the aggressor’s life is devalued to the point of utter worthlessness, it is difficult to justify the killing of multiple aggressors. The problem lies in the faintly ridiculous exercise of deciding exactly how much less an aggressor’s life is worth compared to that of her victim. If it is only slightly discounted then it is difficult to see how a victim is permitted to kill, say, three culpable aggressors. Even if it is heavily discounted, there may be a point at which the combined ‘value’ of many, many aggressors adds up to more than the value of the life of the victim. The alternative is to say that the aggressor’s life is worth nothing at all, an approach with which some might be uncomfortable. Whether the multiple aggressor problem is as serious as McMahan claims, however, depends on the way in which the balancing process is undertaken. McMahan assumes that multiple aggressors’ lives are ‘jointly’ discountable. That is, their combined value is weighed up against that of the victim in a single calculation. On the other hand, if their lives are ‘severally’ discountable, the problem disappears. Under a severally discountable approach, the life of each individual aggressor would be weighed against that of the victim in a series of separate calculations. On account of the difficulties associated with the discount approach, some are more comfortable with the idea that it is justified, on a consequentialist 27 See Chapter 1, where the term passive threat is defined and where this example of a passive threat is discussed. 28 Montague, ‘Self-defense and choosing between lives’, at 211–12. The personal partiality approach is discussed in Section 3.3 below. 29 LA Alexander, ‘Justification and innocent aggressors’ (1987) 33 Wayne Law Review 1177–89; Alexander, ‘Self-defense, justification and excuse’; LA Alexander, ‘A unified excuse of pre-emptive self-protection’ (1999) 74 Notre Dame Law Review 1475–505. 30 Although it is not entirely clear whether he justifies the killing of a culpable aggressor on a consequentialist account or a rights-based account. Elements of both approaches can be found 31 in his work. Alexander, ‘Justification and innocent aggressors’, at 1187. 32 J McMahan, ‘Self-defense and the problem of the innocent attacker’ (1994) 104 Ethics 252–90, at 262.

48

Killing in Self-Defence

account, to kill an aggressor not because her life is worth less than that of her victim, but because this is likely to lead to a net saving of lives in the long term. This is because if an aggressor is allowed to go unchecked, she is likely not just to kill her present ‘victim’ but other future victims (the ‘future harm’ approach). The future harm approach could justify the killing of at least some innocent aggressors, and some cases involving multiple aggressors, if it is taken into account that the aggressor(s), if allowed to go unchecked, are likely to kill others. Davis draws on future harm considerations to justify the killing of what she terms ‘psychotic assailants’: When the attacker is a . . . psychotic assailant, I can justify killing in self-defence by pointing out that there is . . . a difference between myself and the attacker that can be seen to be morally relevant . . . [T]he psychotic is dangerous (and thus more likely to go on to harm other people than I am).33

The future harm approach is most convincing in relation to innocent aggressors who are dangerous but not responsible for their actions due to insanity. It is more difficult to see how killing a child who had accidentally come across a loaded gun would be justified by the future harm approach, as the child is less likely to go on to harm others.34 It is even more difficult to see how it could justify killing a passive threat, such as the roped mountaineer. Thus far, the approaches presented have all appealed to what can be termed ‘act’ consequentialism. That is, in weighing up the benefits of killing either the aggressor or the victim, only the direct consequences of the selfdefensive act under consideration are taken into account. It may be, however, that killing in self-defence can be justified on the basis of ‘rule’ consequentialism. That is, a rule permitting self-defensive killing would tend to deter aggression and therefore save lives. This approach can be seen in the work of Brandt, who argues that the availability of self-defence as a defence will ‘augment the public good in the long run’ because ‘awareness by a potential aggressor that force against him is legitimate is itself a deterrent factor’.35 It has been pointed out, though, that if the only consideration in justifying self-defence is one of rule consequentialism, then this would justify killing an aggressor when the killing was no longer necessary (say, when the aggressor is disarmed and lying helpless on the floor).36 This, presumably, would similarly deter aggression and produce a net saving of lives. Kadish has also 33 N Davis, ‘Abortion and self-defense’ (1984) 13 Philosophy and Public Affairs 171–207, at 189–90. Davis also draws on notions of personal partiality to explain why it is permissible to kill an innocent aggressor (see Section 3.3 below). 34 Although it might be said that the future harm approach still holds if a very short-term view of the future is taken. The child with the loaded gun could potentially kill a large number of people in a very short space of time. 35 RB Brandt, ‘Conscience (rule) utilitarianism and the criminal law’ (1995) 14 Law and Philosophy 65–89, at 88. 36 Kadish, ‘Respect for life and regard for rights in criminal law’, at 883; Wasserman, ‘Justifying self-defense’, at 360; Wallerstein, ‘Justifying the right to self-defense’, at 1004.

The Justification of Self-Defence

49

argued that a rule-consequentialism justification of self-defence means that there would be no reason to permit self-defence in a society with perfect law enforcement, for example where all aggressors are immediately caught and sentenced to death. This, he suggests, would be sufficient to deter all those aggressors who are deterrable at all, without the need for permitting selfdefensive killing by the victims of attackers.37 For these reasons, few attempt to justify self-defence on the basis of rule consequentialism alone. More commonly, rule consequentialism is utilized in conjunction with some form of act consequentialism. That is, if the benefits to society in having a rule likely to deter aggression are placed on the ‘same side of the scales’ as the life of the victim, then self-defensive killing is justified on the balance of interests. This approach can be seen in the work of Eser,38 Gur-Arye,39 and Kremnitzer.40 It is also favoured by Robinson. Having previously attempted to justify self-defensive killing in terms of act consequentialism alone, Robinson later offers combined act/rule consequentialism as a ‘better’ explanation: . . . one may properly add to the evil of physical harm to an innocent, a variety of intangible evils that arise from such aggression, evils that may well be more significant to society than the physical harm threatened. In the case of the thugs, for example, the lives of the three thugs are balanced against the lives of the three townspeople plus the compelling societal interest in preserving the right of bodily autonomy and condemning unjustified aggression.41

Consequentialist theories of self-defence have been praised for the fact that they justify conduct that results in the least harm: both in terms of the protection of human life and the minimization of the overall amount of societal violence.42 One aspect of consequentialism is, though, still troubling: to judge an action purely on the basis of its consequences is to neglect issues of individual rights and justice.43 It is possible to construct numerous hypothetical self-defence scenarios where the ‘result’ reached by a balance of interests analysis would seem unjust to an individual. One example is that of the brilliant scientist about to discover a cure for cancer who attacks her victim while deprived of mens rea due to involuntary intoxication. Consequentialist analysis would suggest that, on balance, it is better that the victim of the attack dies than the aggressor, but this seems unjust to the innocent victim, who, like the scientist, finds herself in this situation through no fault of her own. One could also conceive of situations 37

Kadish, ‘Respect for life and regard for rights in criminal law’, at 883. A Eser, ‘Justification and excuse’ (1976) 24 American Journal of Comparative Law 621–37. 39 M Gur-Arye, ‘Should the criminal law distinguish between necessity as a justification and necessity as an excuse?’ (1986) 102 LQR 71–89. 40 M Kremnitzer, ‘Proportionality and the psychotic aggressor: another view’ (1983) 18 41 Israel Law Review 178–214. Robinson, Criminal Law Defenses, at § 131(a). 42 NM Omichinski, ‘Applying the theories of justifiable homicide to conflicts in the doctrine of self-defense’ (1987) 33 Wayne Law Review 1447–69, at 1466. 43 J Rachels, The Elements of Moral Philosophy (3rd edn, 1999), at 111–13. 38

50

Killing in Self-Defence

where consequentialist analysis would justify the killing of an innocent bystander, for example where three innocent victims face an attack by an extremely dangerous aggressor in the process of a murdering rampage. The only way they can defend themselves is by shooting the aggressor but in doing so they will also kill an innocent bystander, who is being used by the aggressor as a human shield. A consequentialist could easily justify their action,44 but again, this seems unjust to the bystander, whose ‘right to life’ has not entered the equation. Consequentialists might well respond to such criticism by pointing out that, concerns of injustice apart, the balance of harms approach comes up with the ‘correct’ answer even in these hard cases. However, the lack of concern with individual rights and justice has led Thomson to dismiss consequentialist approaches as unworthy of discussion45 and they are not even mentioned by Kasachkoff46 and Uniacke47 in their surveys of the different approaches to the justification of self-defensive killing. Even some of those who espouse consequentialist explanations seem to accept that there are difficulties with the approach. Robinson, for example, describes selfdefence as a ‘troubling justification’ because of the preference displayed for one life over another.48

3.3 The Personal Partiality Approach Another approach to justifying self-defensive killing is to state that an individual is entitled, in a situation where she faces a choice between lives, to give preference to her own life. That is, faced with a choice between killing an aggressor or losing her own life, an individual is entitled to save her own life simply because she personally values it more. This approach is termed the ‘personal partiality approach’, although elsewhere it has also been described as the ‘indifference argument’.49 Accounts based on personal partiality have been used primarily to explain the permissibility of killing an innocent aggressor or a passive threat.50 The argument runs as follows. When a victim faces a threat to her life from an innocent aggressor or a passive threat, neither victim nor aggressor/passive threat is in a morally better position than the other with respect to who 44 Three lives saved compared to only two lost, even without taking into account the lives that the rampaging aggressor might go on to take if not stopped, or the benefit to society of 45 having a rule likely to deter aggression. Thomson, Rights, Restitution and Risk, at 43. 46 T Kasachkoff, ‘Killing in self-defence: an unquestionable or problematic defence?’ (1998) 47 17 Law and Philosophy 509–31. Uniacke, Permissible Killing. 48 Although he does conclude that ‘intangible societal interests break the deadlock’ (Robinson, Criminal Law Defenses, at § 131(a) ). 49 Kasachkoff, ‘Killing in self-defence’, at 520. 50 Davis, ‘Abortion and self-defense’; S Levine, ‘The moral permissibility of killing a material aggressor in self-defence’ (1984) 45 Philosophical Studies 69–78; Montague, ‘Self-defense and choosing between lives’.

The Justification of Self-Defence

51

should be killed and who should survive. This means that there is no morally relevant reason why one of them should live at the expense of the other. If there is no morally relevant reason to decide one way or the other, it is morally permissible for the victim to decide the matter in favour of herself. Both courses of action are morally permissible; neither is morally wrong.51 There is, however, another step in the argument (and this is where personal partiality becomes central). The victim values the preservation of her own life more than she values the preservation of the aggressor’s life. This is simply because it is her own life and is therefore more valuable to her. At the point when a choice between the life of the aggressor and the victim has to be made, the victim may choose her own simply because it is understandable for her to have a greater interest in her own survival than she has in someone else’s. As such, Davis argues that: When I claim that I may be justified in killing a person who poses a threat to my life, I am not claiming that my life has, or should be thought to have, greater value than that of my attacker. Nor am I claiming that it would be (more) unjust if I am the one who is killed. I am instead, merely calling for other people’s recognition of the fact that (by my lights) my life has greater value to me . . . Because of the greater value that each of us understandably attaches to the continuation of his or her own life, we are (in certain circumstances) permitted to kill another person to preserve our own life, even though we acknowledge that the other person’s claim to life is not weaker than our own.52

It might be questioned at this point whether personal partiality is indeed a separate and distinct approach. Although it is put forward as such by some,53 it might equally be conceived of as an agent-relative form of consequentialism: an individual is entitled to save her own life over that of an aggressor simply because she personally values it more. Regardless of how the personal partiality approach is classified, it is open to criticism on a number of fronts. First, if personal partiality is the sole ground for the justification of self-defence, then the aggressor may well be justified in fighting back against any self-defensive force used against her. If the victim is entitled to favour her own life, then a personal partiality account, if it is not to justify force used against self-defensive force, must find some mechanism by which the aggressor is not also entitled to favour herself. McMahan makes this point and, oddly enough, sees it as an advantage, at least in relation to innocent aggressors: ‘It also has the further advantage that it supports another intuitively plausible claim: namely that the moral reason the victim has to resist the innocent aggressor is also available to the innocent aggressor as a justification in resisting the victim’s counter-attack’.54 51 Levine, ‘The moral permissibility’, at 72; Montague, ‘Self-defense and choosing between 52 lives’, at 212. Davis, ‘Abortion and self-defence’, at 192–3 (emphasis added). 53 McMahan, ‘Self-defense and the problem of the innocent attacker’, at 268. 54 ibid, at 269.

52

Killing in Self-Defence

Unlike McMahan, I do not accept that an innocent aggressor is justified in fighting back, given that it is her conduct that has posed a direct threat to the victim’s life in the first place.55 That the personal partiality approach appears to allow this is a troubling aspect of the account and one that is difficult to defend without resort to alternative approaches. A second criticism is that if the reason why self-defence is permissible is grounded solely in the notion that an individual is entitled to prefer her own survival, there appears to be nothing to distinguish self-defence from other situations where an individual can save her life by killing another. Thus, the personal partiality approach would justify killing an innocent bystander in self-preservation: for example, the starving woman who kills and eats her companions on a shipwreck or the woman dying of kidney failure who kills an acquaintance in order to obtain her healthy kidney.56 Montague, far from feeling the need to defend the personal partiality approach from such criticism, sees this as an advantage. For Montague, a victim is justified in favouring her own life in any situation where she faces a straight choice between lives, even where to do so would cause the death of an innocent bystander.57 The view preferred here, however, is that any approach that justifies the killing of innocent bystanders should be rejected on the basis that, like consequentialism, it ignores the notion that bystanders have a right to life. Third, and perhaps most importantly by way of criticism, the personal partiality approach does not explain why the victim is permitted to override any right to life the aggressor has. In order to accept the personal partiality explanation, it is necessary to find some reason why an individual is permitted to attach more weight to her own interests than to the interests of others. Unless some way of doing this is found, the personal partiality account collapses into what Gorr terms ‘egoism’.58 As Kasachkoff argues, if a victim kills an aggressor in self-defence: . . . he violates [the aggressor’s] right to life—a violation for which he must provide a strong moral justification. But if he sacrifices himself . . . he does no moral wrong at all. It is not that sacrificing one’s life is a good thing, either sometimes or always, but it is something to which one party consents, and that is surely different from taking the life of another party against his will.59

In many ways, personal partiality accounts seem to fit more closely with the concept of an excuse than a justification. It is difficult to see why it is permissible for someone to favour her own life over that of another simply because it is her own life and therefore of value to her. It is easier to see why 55

See Section 3.5 below. Kasachkoff, ‘Killing in self-defence’, at 526; M Gorr, ‘Private defence’ (1990) 9 Law and Philosophy 241–68, at 252. 57 Montague, ‘Self-defense and choosing between lives’, at 209. 58 59 Gorr, ‘Private defence’, at 252. Kasachkoff, ‘Killing in self-defence’, at 524–5. 56

The Justification of Self-Defence

53

such conduct might be excused, especially if the choice was made under conditions of extreme fear.

3.4 Appeals to the Doctrine of Double Effect A third approach to explaining the justification of killing in self-defence has been to appeal to the doctrine of double effect. The broad thrust of the doctrine is that it is always wrong intentionally to do a bad act for the sake of good consequences that will follow, but it may be permissible to do a good act in the knowledge that bad consequences will follow. The important distinction is between intended and merely foreseen consequences. The doctrine of double effect holds that, under specific conditions, it can be morally permissible to bring about a foreseen effect, such as the death of another person, provided this is not strictly intended. An effect is strictly intended if it is the actor’s aim or purpose or is a means of achieving the actor’s aim in the circumstances. One can foresee an action that is virtually certain to occur but not intend it in this sense. Provided the intended good effect is proportionate to the foreseen bad effect, the act might be permissible.60 Applying this reasoning to self-defence, the argument is that, in using necessary and proportionate force to defend herself, the individual concerned intended merely to repel the threat to her life. What she foresaw but did not intend was that the death of the aggressor would result. The permissibility of self-defence is thus grounded in the notion that it is not an intentional killing: the death of the aggressor is merely a foreseen side-effect of a permissible act. Something approximating an appeal to the doctrine of double effect to justify self-defensive killing can be seen in the writing of Saint Thomas Aquinas and, indeed, it is from Aquinas’ writing that the term ‘double effect’ is derived. Aquinas states that: Nothing hinders one act from having two effects, only one of which is intended, while the other is beside the intention. Now, moral acts take their species according to what is intended and not according to what is beside the intention, since that is accidental . . . Accordingly, the act of self-defence may have two effects: one, the saving of one’s life; the other, the slaying of the aggressor.61

More recently, the most prominent supporter of the doctrine as an explanation of the permissibility of killing in self-defence has been Anscombe, who argues that ‘[t]he plea of self-defence . . . made by a private man who has killed someone else must in conscience—even if not in law—be a plea that 60

S Uniacke, ‘The doctrine of double effect’ (1984) 48 The Thomist 188–218. St Thomas Aquinas, Summa Theologiae (1266–73, 1975 edn with English translation by Marcus Lefébure), II-II, qu 64, art 7. 61

54

Killing in Self-Defence

the death of the other was not intended, but was a side effect of the measures taken to ward off the attack’.62 The use of the doctrine of double effect to justify self-defensive killing has been extensively criticized, primarily on the basis of whether there really is a meaningful distinction between the intended effects and merely foreseeable effects of lethal self-defensive force.63 Uniacke uses the example of repelling an aggressor by blowing her up and questions whether the defender can truly maintain that she did not intend to kill, especially given that an effect is strictly intended if it is a means of achieving the actor’s aim in the circumstances. I would agree. The moral distinction between intending to kill an aggressor in self-defence and using lethal force in self-defence but merely foreseeing (but not intending) that the death of the aggressor would result is too fine to be meaningful.64 Indeed, for this reason, Grisez,65 a supporter of the doctrine of double effect in other contexts, casts doubt on whether it can meaningfully be applied to self-defensive killing.66

3.5 Rights Approaches So far, this chapter has examined, and rejected, three different approaches to explaining the justification of self-defensive killing: consequentialist approaches, the personal partiality approach and approaches appealing to the doctrine of double effect. It now turns to explanations based on the right to life, for it is here that the most convincing explanation of the permissibility of killing in self-defence can be found. The general gist of the rights explanation is that the right to life is possessed by all human beings and a victim whose life is threatened by an aggressor is therefore permitted to use lethal force to defend her life.67 The 62 GEM Anscombe, ‘War and murder’, in Anscombe, Collected Philosophical Papers Volume 3: Ethics, Religion and Politics (1981) 51–61, at 54. 63 J Harris, Violence and Responsibility (1980), at 53; DN Husak, ‘The complete guide to self-defence’ (1996) 15 Law and Philosophy 399–406, at 405. 64 V Tadros, ‘Practical reasoning and intentional action’ (2000) 20 LS 104–23, at 109. See also Ward LJ in Re: A (Children) [2001] 2 WLR 480, at 532, who rejects the doctrine of double effect as an explanation of why the separation of two conjoined twins, where only one will survive the operation, is permissible. Although cf Walker LJ in the same case, who uses something akin to double effect to justify the separation (at 591). For a more detailed discussion of the case, see Chapter 1. 65 GG Grisez, ‘Towards a consistent natural law ethics of killing’ (1970) 15 American Journal of Jurisprudence 64–96, at 79. 66 Indeed doubt has been cast on whether even Aquinas intended double effect to be an explanation of why it is permissible to kill in self-defence (A McIntyre, ‘Doing away with double effect’ (2001) 111 Ethics 219–55, at 249). 67 Uniacke, Permissible Killing; Thomson, ‘Self-defense’. See also D Rodin, War and Self-Defense (2002), who takes a similar approach, albeit in a slightly different context, that of when it is acceptable to go to war.

The Justification of Self-Defence

55

main attraction of a rights-based approach is that it avoids the difficulties of consequentialist reasoning whereby individual rights can be overridden in the quest for maximum utility.68

3.5.1 The notion of a right to life Human rights are defined by Macfarlane as ‘those moral rights which are owed to each man or woman by every man or woman solely by reason of being human’.69 According to Macfarlane, human rights are distinguishable from other moral rights in that they possess five inherent characteristics: universality, individuality, paramountcy, practicability, and enforceability.70 Universality refers to the notion that they are possessed by all human brings simply by virtue of being human.71 Individuality refers to the idea that human rights are rights possessed by individuals, rather than groups.72 Paramountcy refers to the strength of the right: human rights can only be overridden where there is a strong justification for doing so. Practicability refers to the notion that a human right is one that the state has the power to uphold (even if it chooses not to do so). Enforceability refers to the notion that they are rights that can be asserted by the individual, most commonly against the state but in some circumstances against other individuals. The right to life has been described as the most fundamental of all human rights.73 The origin of the right to life as a specific human right can be traced back to the respect for the sanctity of life in religious doctrine. The proposition that ‘thou shalt not kill’ is fundamental to both the Jewish and Christian faiths, for example. Human life, the union of the body and the soul, is sacred, with human beings being created in the image of God, and their destiny being eternal life.74 The religious concern with the sanctity of life can, however, be distinguished from an individual ‘right to life’ in the terms defined by Macfarlane. Religious doctrine does not suggest that the individual has a right to her own life that can be asserted against God, the state, or any other individual. Rather, it suggests a duty on the part of human beings to respect the lives of others. Indeed, in Christian doctrine, life is by no means the highest value, with the Gospel teaching that the greatest good is not one’s own survival, but to lay down one’s life for others.75 68 DA Richards, ‘Rights, resistance, and the demands of self-respect’ (1983) 32 Emory Law Journal 405–35, at 415. Of course, this is only a difficulty for non-consequentialists. A supporter of consequentialism would not have a problem with overriding rights in the quest for maximum 69 utility. LJ Macfarlane, The Theory and Practice of Human Rights (1985), at 3. 70 Macfarlane, The Theory and Practice of Human Rights, at 3. 71 M Cranston, What are Human Rights? (1973), at 7. 72 The idea that human rights cannot be possessed by groups would, however, be questioned today by those who accept ‘third generation’ human rights (P Alston, People’s Rights (2001)). 73 A Ashworth, ‘Self-defence and the right to life’ (1975) 34 CLJ 282–307, at 282. See Section 3.5.3 below. 74 J Shestack, ‘The philosophical foundations of human rights’, in J Symonides (ed), Human Rights: Concept and Standards (2000) 31–66. 75 N St John-Stevas, The Right to Life (1963), at 16.

56

Killing in Self-Defence

The duty to respect the sanctity of human life recurs in the work of philosophers in the medieval period. St Thomas Aquinas,76 for example, expressed it in terms of natural law, as derived from the word of God. Aquinas referred to an inherent duty to respect the life of others, an essentially negative obligation, on the basis that human life should be regarded as sacred and God-given.77 There is no suggestion in Aquinas’ work that human beings have an individual right to life, with all the characteristics of universality, enforceability, and so on. Reference to the duty to respect life can also be found in the work of John Locke,78 writing in 1690. For Locke: The State of Nature has a Law of Nature to govern it, which obliges every one: And reason, which is that Law, teaches all mankind, who will but consult it, that being all equal and independent, no-one ought to harm another in his Life, Health, Liberty or Possessions. For Men being all the Workmanship of one Omnipotent, and infinitely wise Maker; All the Servants of one Sovereign Master, sent into the World by his order and about his business, they are his Property, whose Workmanship they are, made to last during his, not one another’s Pleasure.79

Although in some later passages Locke uses the term ‘rights’,80 like Aquinas, Locke’s concern is primarily with the duty to respect the lives of others, rather than with human rights in Macfarlane’s terms. Indeed, such was the significance attached to the sanctity of human life by philosophers such as Aquinas and Locke, that the duty extended even to one’s own life. An individual was not permitted to take her own life, which was seen as the property of God. Aquinas wrote that ‘life is a gift made to man by God, and it is subject to him who is master of death and life. Therefore a person who takes his own life sins against God.’81 It is towards the end of the eighteenth century that the first indications of the right to life as a human right in the terms expressed by Macfarlane can be found. In June 1776, a convention in Virginia adopted a declaration of rights which stated that ‘all men are by nature equally free and independent, and have certain inherent rights, of which, when they enter into a state of society, they cannot, by any compact, deprive or divest their posterity; namely the enjoyment of life and liberty’.82 In July 1776, the American Declaration of Independence commenced by stating that: ‘We hold these truths to be selfevident: That all men are created equal; that they are endowed by their Creator with certain unalienable rights; that among these are life, liberty, and the pursuit of happiness’.83 The concept of the right to life as a fundamental human right was cemented by the formation of the United Nations, created after the Second 76

Aquinas, Summa Theologiae. Summa Theologiae, II-II, qu 64, art 1; qu 64, art 2; qu 64, art 6. J Locke, Two Treatises of Government (1690, 1988 edn with introduction and notes by 79 Peter Laslett). ibid, ii, 6. 80 81 ibid, ii, 7. Aquinas, Summa Theologiae, qu 64, art 5. 82 83 Virginia Declaration of Rights s 1, emphasis added. Emphasis added. 77 78

The Justification of Self-Defence

57

World War, with one of its primary tasks being the attachment of human rights to individuals, the right to life being the most fundamental of these. In the resulting United Nations’ Universal Declaration of Human Rights, the right to life is the first substantive right to be declared. The right to life holds a similarly prominent position in subsequent conventions. In each of the European Convention on Human Rights, the American Declaration of the Rights and Duties of Man, and the International Covenant on Civil and Political Rights, the right to life is the first substantive right to be affirmed. In the American Convention on Human Rights and the African Charter on Human and People’s Rights, the right to life is the second substantive right to be declared, following on from the right to recognition as a person before the law (in the American Convention) and the right to the equal protection of the law (the African Charter).84 The importance of the right to life as an individual human right is thus well established. It is worth considering, however, first, why human beings possess it at all and, second, why the right to life should be considered of such primary importance.

3.5.2 Why do human beings have a right to life? One possible answer to the question of why human beings might have a right to life is that its existence is an ultimate belief for which no further argument can or need be made. We have already seen that the American Declaration of Independence describes the right to life as ‘self-evident’ and, as Jones states, in explaining the existence of rights, at some point ‘we must hit rock-bottom, so that demands for further justification cannot be satisfied and can be met only by the assertion of a basic principle’.85 Indeed, this seems to be the approach taken by both Uniacke and Thomson, neither of whom feels the need to defend the existence of a right to life. Both take the fact of its existence as the starting point for analysis, a stance that has led to their accounts being criticized for begging the question.86 There are various bases upon which the existence of the right to life might be grounded. One possibility is by reference to natural law. One type of natural law-based explanation would be to derive the right to life from the word of God, but an alternative, although still natural law-based, explanation can be found in the work of Finnis.87 For Finnis the right to life is a natural right derived from the existence of certain basic valuable human goods that can be 84 In none of these conventions is the right to life absolute. Most, for example, permit the death penalty when the sentence is passed by a court of law: Art 2(1) of the European Convention on Human Rights; Art 4(2) of the American Convention on Human Rights; Art 6(2) of the International Covenant on Civil and Political Rights. 85 P Jones, Rights (1994), at 97. 86 Segev, ‘Fairness, responsibility and self-defense’, at 443; McMahan, ‘Self-defense and the problem of the innocent attacker’, at 277–8. 87 JM Finnis, Natural Law and Natural Rights (1980).

58

Killing in Self-Defence

‘discovered’ simply by thinking about what it means to be human. Finnis sets out seven basic human values, of which life is one (alongside knowledge, play, aesthetic experience, sociability, practical reasonableness, and religion), all of which he claims as fundamental aspects of human well-being. From these values are derived corresponding rights owed to individuals and, for Finnis, the right to life is the most basic of these.88 Natural law is not the only basis on which the right to life has been founded. It has also been grounded in the notion that reason demands its recognition if human beings are to live together peacefully in society. As Cranston states, all societies, if they are to operate effectively, must have rules and ‘there is one rule which every society must have, and that is that no man shall use violence against the life and person of his neighbour’.89 Another approach is to ground the right to life in the basic principle of moral philosophy that it is unjust to an individual to deprive her of her life. This can be seen to an extent in the work of Glover, who examines in detail why killing is wrong.90 Glover is wary of attributing a right to life to human beings because of its connotations of absoluteness but he does accept that something akin to a right to life exists in the sense that it is wrong to kill another human being if she wants to go on living. For Glover this is grounded in the basic moral principle that it is wrong to reduce the length of a worthwhile life.91 Although not explicitly stated as such, this is the approach that seems most to inform the work of Thomson and Uniacke, both of whom discuss self-defence in terms of their intuitive feelings about what is just.92

3.5.3 Why is the right to life the most fundamental of all human rights? The right to life has been said to be the most fundamental of all human rights. Ashworth, for example, describes it as ‘the most basic claim of every human being’.93 There are perhaps two reasons for this. The first reason is to some extent self-evident, since without life no other human rights can be enjoyed. As Jones points out, being alive is a prerequisite for possessing all other rights.94 The second derives from the notion that any violation of the right to life, unlike violations of other human rights, is something that can never be remedied, at least from the perspective of the individual who has been deprived of her life. It is irremediable in two closely related senses. The first relates to the permanence of the loss involved. The deprivation of life is 88

89 Finnis, Natural Law, at 83. Cranston, What are Human Rights?, at 26. Glover, Causing Death and Saving Lives, at 39–118. 91 This is what Glover terms the ‘direct’ reason why killing is wrong. Glover also sets out various ‘indirect’ reasons why killing is wrong, which are of less relevance here. 92 For further discussion of the foundation of the right to life, see Shestack, ‘The philosophical foundations of human rights’; Jones, Rights (especially chapter 4, ‘Natural rights and human rights’ and chapter 5, ‘Justifying human rights’). 93 94 Ashworth, ‘Self-defence and the right to life’, at 282. Jones, Rights, at 83. 90

The Justification of Self-Defence

59

something from which the victim can never recover: it is irrevocable. Against this, it might be said that there are other deprivations to which this also applies. A violation of the right to liberty through unjust life imprisonment might be said to be a permanent loss if the individual is never released. Here, however, serious though the deprivation is, the individual can at least continue to develop as a human being, albeit along a different path to that originally desired. As Young has stated,95 unjust life imprisonment might prevent the realization of an individual’s original life purposes, but she may still come to engage with new life purposes. By contrast, ‘[w]hen a person is killed all such hope is irretrievably lost’.96 Ashworth has made the similar point that, with all other types of deprivation, ‘the victim retains his or her life and, therefore, the possibility of further pleasures and achievements, whereas death is final’.97 The second sense relates to the fact that the loss of life is something that can never be compensated. Given that the individual is no longer alive, no form of financial or other compensation can be paid to her in an attempt to make up for the violation.98 Again, this makes a violation of the right to life different to the violation of any other human right. As before, it might be argued that there are other types of rights to which this also applies. It could perhaps be said that no amount of money can ever recompense an individual for the unjust deprivation of her liberty or for prolonged and painful torture. In these cases, however, compensation is at least a meaningful concept, even if it is recognized that the loss suffered can never be fully compensated.99

3.5.4 The right to life and self-defensive killing For both Uniacke and Thomson, the permissibility of killing in self-defence is grounded in the notion that all human beings have a right to life. On its simplest level, the argument runs that an aggressor, by virtue of attacking her victim, threatens to violate the victim’s right to life. The victim is therefore justified in using lethal defensive force to repel the threat because, in doing so, the victim is simply protecting her right to life. It becomes clear at this point that the type of right in question has the features of a Hohfeldian ‘claim-right’.100 Hohfeld distinguished between rights (termed ‘claim-rights’ by subsequent commentators) and privileges. 95

R Young, ‘What is so wrong with killing people?’ (1979) 54 Philosophy 515–28. ibid 518. 97 A Ashworth, Principles of Criminal Law (5th edn, 2006), at 247. 98 As opposed to the family of the individual who has suffered the deprivation of life, who can at least potentially be financially compensated for the loss of their loved one. 99 There are perhaps exceptions to this in extreme cases where the deprivation involved comes close to a deprivation of life. One example might be where a victim is rendered permanently insensible by acts of torture. 100 WN Hohfeld, ‘Some fundamental legal conceptions as applied in judicial reasoning’ (1913) 23 Yale Law Journal 16–59. 96

60

Killing in Self-Defence

A claim-right invokes a corresponding duty in others to allow the right holder to enjoy her right. If X holds a claim-right to life, Y has a duty to refrain from violating X’s right to life. A privilege refers to something an individual has a liberty to do and invokes no corresponding duties in others. If the right to life was merely a privilege, it would not be correlated with anyone else’s duty to refrain from interfering with an individual’s right to life. A strength of the rights-based account is that it justifies self-defensive force against aggressors (and passive threats) but not against bystanders.101 As we have already seen, under certain types of consequentialist account, killing an innocent bystander in self-defence could be permissible (if, for example, this would lead to a net saving of lives). The rights-based account works in an entirely different way. Here, the distinctive feature of self-defensive force is that it is force used against a threat. It is precisely because an aggressor or a passive threat directly threatens the victim’s right to life that self-defence is permissible. Bystanders do not pose a direct threat to the victim’s right to life in the same way.102

3.5.5 Rights and forfeiture It has already been noted that one of the key features of a human right is that it is universal.103 It is possessed by all human beings simply by virtue of being human and cannot be lost on the basis of being a particular nationality, or colour, or holding certain views, or being a ‘bad’ person.104 The most obvious difficulty with the rights approach is, then, that there needs to be some principle that explains why the aggressor does not possess the universal right to life (or why her universal right is overridden or outweighed).105 If both aggressor and victim possess the right to life in equal measure, the implication is that when the victim uses self-defensive force, the aggressor is, in turn, permitted to fight back against the victim (who is, after all, threatening to violate her right to life). Now, if the right to life is seen as absolute, in the strict sense of the word, then maybe this conclusion simply has to be accepted. There are some parallels here with the work of Hobbes, who considers the right to selfpreservation an absolute right that ‘no law can oblige a man to abandon’.106 101

For the distinction between aggressors, passive threats and bystanders, see Chapter 1. There is, of course, another reason why the killing of a bystander would not be permissible under a rights-based approach. This is because the bystander has the same right to life as 103 the victim (see Section 3.5.5 below). See Section 3.5.1 above. 104 This is not the same as saying that the right is absolute. As Section 3.5.1 pointed out, the right might be qualified in circumstances such as a lawful execution. 105 Wallerstein, ‘Justifying the right to self-defense’, at 1016. 106 T Hobbes, Leviathan (1651), at 61. See also Wallerstein, ‘Justifying the right to selfdefense’, at 1024. Wallerstein believes that the right to life (or more specifically the right not to be killed) is an absolute and unqualified right; probably the only right that can be described as such. 102

The Justification of Self-Defence

61

Hobbes has been interpreted by some to mean that an aggressor would be justified in fighting back against self-defensive force.107 If this conclusion is to be avoided, however, the question of why the aggressor does not have a right to life has to be addressed. One way to proceed is to hold that an aggressor, by virtue of her conduct, forfeits her right to life. ‘Forfeiture’ is sometimes presented as a theory in its own right108 but it is best understood as part of a justification of self-defence that also includes appeals to rights. After all, if an aggressor does not have a right in the first place, it is difficult to see what is being ‘forfeited’. Both Thomson and Uniacke use forfeiture to explain why it is permissible for the victim to kill the aggressor, but it is not permissible for the aggressor to fight back and kill the victim.109 Their accounts of forfeiture differ in one respect: Thomson’s account is based on temporary forfeiture, while Uniacke’s is based on qualification. This difference merits some explanation. For Thomson, forfeiture is a temporary state: the right to life is forfeited only in so far as the aggressor is a threat to the life of others. The aggressor’s right to life is immediately regained once she ceases to be a threat. Thomson had, in 1986, professed to be uncomfortable with the notion of forfeiture, preferring to write of the aggressor’s right being overridden by the more stringent right of the defender.110 By 1991, however, Thomson had come round to the view that forfeiture is a necessary component of the justification of self-defensive force: ‘In short, I suggest that what makes it permissible for you to kill [in self-defence] is the fact that [aggressors] will otherwise violate your rights that they not kill you, and therefore lack rights that you do not kill them’.111 Accounts of forfeiture that rely on the temporary loss of rights have been said by some to make no sense.112 The notion of a right that comes and goes, depending on conduct, is certainly a difficult concept to grasp. It is for this reason that Uniacke tackles the issue rather differently. In explaining why it is permissible for the victim to kill the aggressor, Uniacke attempts to avoid the language of forfeiture (whether temporary or permanent) entirely by specifying the right to life in such a way that an aggressor does not possess this right in the first place, an approach that Thomson has called ‘the method of factual specification’.113 For Uniacke, the right to life is conditional and depends on conduct. Human beings possess a right to life only in so far as they do not become an 107 J Waldron, ‘Self-defense: agent-neutral and agent-relative accounts’ (2000) 88 California Law Review 711–49, at 732. 108 Omichinski, ‘Applying the theories of justifiable homicide’, at 1449. 109 As does Rodin in War and Self-Defense, at 79. 110 111 In Rights, Restitution and Risk, at 36. Thomson, ‘Self-defense’, at 302. 112 Kadish, ‘Respect for life and regard for rights in criminal law’, at 884; Wallerstein, ‘Justifying the right to self-defense’, at 1020. 113 Thomson, Rights, Restitution and Risk, at 38. This terminology has since been adopted by others, for example Wallerstein, in ‘Justifying the right to self-defense’, at 1021.

62

Killing in Self-Defence

unjust114 immediate threat to another person’s life.115 A victim is permitted to kill an aggressor who threatens her life not because the aggressor has forfeited anything, but because the aggressor does not possess the specified right to life in the first place (by virtue of being an unjust immediate threat to the life of the victim). It is suggested here, though, that Thomson’s account is the more honest. Uniacke might explicitly reject the terminology of forfeiture but forfeiture is effectively what she is describing. In an article written to deal with criticism of her account as dishonest, Uniacke re-states that she ‘explicitly reject[s] the terminology of forfeiture in the context of self-defence’116 on the basis of the term’s punitive implications. The concern to avoid any notion of penalty can also be seen in her earlier work, where she states that: Because forfeiture of rights is associated with both culpability and punishment, in setting out the theory of forfeiture necessary to a unitary account of justified homicide in self-defence I shall prefer to specify the scope of the right to life appropriately, rather than say that an unjust aggressor forfeits the right to life.117

But despite rejecting it, Uniacke slips into forfeiture terminology on numerous occasions in her text.118 It is difficult, therefore, not to interpret Uniacke as saying that the aggressor has, by virtue of her conduct, forfeited her right to life. Indeed, Uniacke herself accepts at one point that it is merely a matter of semantics and forfeiture and specification are effectively the same thing: ‘I regard an appropriate specification and theory of forfeiture of the right to life as theoretically on a par. A forfeit is a penalty and . . . the imposition of a penalty by way of a forfeited right constitutes a disadvantage which need not imply the culpability or punishment of the one who forfeits.’119 Writing about rights more generally, McConnell has also explicitly linked the idea of a conditional right with the notion of forfeiture, stating that ‘[a] right that can be forfeited is, in a sense, conditional: to retain it, the possessor must not behave improperly’.120 I would argue, then, that the rights/forfeiture accounts of the permissibility of self-defence proposed by Thomson and Uniacke are effectively the same: it 114 The term ‘unjust’ is used by Uniacke to rule out the permissibility of using self-defensive force to prevent justified killings, such as legal executions. 115 Uniacke, Permissible Killing, at 213. 116 117 Uniacke, ‘Response to two critics’, at 628. Uniacke, Permissible Killing, at 195. 118 See Permissible Killing, at 120, and the examples noted in T Kasachkoff, ‘Comment and reply to Suzanne Uniacke’s “a response to two critics” ’ (2000) 19 Law and Philosophy 635–9, at 636–7 and the accompanying footnotes. 119 Permissible Killing, at 195. In various other passages of Permissible Killing, Uniacke has described her account as ‘something akin to a theory of forfeiture’ (at 158) and has stated that ‘something like a theory of forfeiture of the unjust aggressor’s rights is necessary to the justification of homicide in self-defence’ (190). See also Uniacke, ‘Response to two critics’: ‘I regard an appropriate specification of rights as equivalent to a theory of forfeiture properly understood’ (at 628). 120 T McConnell, ‘The nature and basis of inalienable rights’ (1984) 3 Law and Philosophy 25–59, at 28.

The Justification of Self-Defence

63

is permissible to kill an aggressor because, by virtue of the unjust immediate threat she poses to the life of the victim, she forfeits her right to life. It should perhaps be said at this point that there will be some to whom a forfeiture account of self-defence will never be acceptable. If the right to life is considered an absolute right, in the strict sense, then any notion that it can be forfeited is likely to be rejected. Against this, it can only really be said that it is very difficult to justify self-defensive killing in terms of rights without appealing to the concept of forfeiture. This difficulty is illustrated by examining the account of Wallerstein, who has professed to find the notion of forfeiture problematic. Wallerstein, while accepting that the starting point for the justification of self-defence is the right to life,121 does not accept that the right is one that can be forfeited, even temporarily.122 Her reasoning, however, runs into difficulty if the notion of forfeiture is not accepted. Her account runs as follows.123 The defender has a right to life. Thus, if the aggressor makes an attack on the defender, in the absence of justifying circumstances, he wrongs him. Because the aggressor is causally responsible for the aggression, he has created a situation in which either he or the defender will have to suffer the consequences. It is this fact that means that the defender has a right to use necessary and proportionate force against the aggressor. Thus, for Wallerstein, when the defender attacks the aggressor ‘he has justifying circumstances and does not violate the aggressor’s right to life and does not wrong him’.124 But it is this final step that simply does not make sense if something akin to forfeiture is not accepted. Wallerstein starts from the basis that all human beings have a right to life. Why is it, then, that the defender does not violate the aggressor’s right to life in Wallerstein’s account? After all, the defender has deprived the aggressor of his life and may well have done so intentionally. It surely cannot simply be ‘because he [the defender] has justifying circumstances’. A further step in the argument is necessary to explain why the existence of justifying circumstances means that killing the aggressor is not a violation of his right to life. The only way of making this final step is to accept the notion of forfeiture.

3.5.6 Additional criticisms of the forfeiture account Aside from the difficulty some commentators have had in accepting the concept of temporary forfeiture, criticism of the rights/forfeiture account has been made in two further respects. First, it has been said that it does not explain why it is permissible to kill an innocent aggressor, such as a child or an insane person, or a passive threat, such as the roped mountaineer discussed in Chapter 1. At the forefront of such criticism is Kasachkoff, who 121 123

Wallerstein, ‘Justifying the right to self-defense’, at 1032. 124 ibid, at 1030–1. ibid, at 1030.

122

ibid, at 1024.

64

Killing in Self-Defence

does not accept that the right to life can be forfeited where the aggressor or threat is not culpable.125 Kasachkoff finds it hard to accept that forfeiture can be linked to anything other than fault and she criticizes Uniacke for claiming that ‘killing such blameless persons does no injustice to them. Losing their lives is, after all, not what they deserve’.126 Uniacke has responded by stressing that forfeiture is based not on fault but on the conduct of the aggressor.127 Indeed, in Permissible Killing Uniacke takes great care to stress that forfeiture is not linked to fault, as this would suggest that forfeiture is a punishment, which it is not. Rather, the right to life is forfeited simply by virtue of becoming an unjust immediate threat to the life of another. This does not imply that the aggressor ‘deserves’ to die, as Kasachkoff suggests. As Wallerstein has pointed out, it is ‘important to remember that in self-defence we are not dealing with punishment or with what the aggressor “deserves” but only with resisting or warding off an attack’.128 The second criticism made of the forfeiture account is that it permits self-defensive killing even when the death of the aggressor is not necessary. If the aggressor has forfeited her right to life, this seems to suggest that it is permissible to kill her at any time, even when she has ceased to be a threat, for example by becoming incapacitated or by desisting from her attack. Once again, Kasachkoff has been at the forefront of such criticism, stating that: ‘Once we tie the moral justifiability of killing in self-defence to the attacker’s forfeiture of the right not to be killed we seem to undercut the requirement that we use the least force necessary to defend ourselves’.129 Yet this criticism is relatively easily addressed. In neither Thomson’s nor Uniacke’s account is it suggested that an aggressor forfeits her right to life indefinitely. As we have already seen, Thomson deals with the point by using the concept of temporary forfeiture. For Thomson, an aggressor regains the right to life once she is no longer a threat to the life of the victim and thus it would not be permissible to kill an aggressor who no longer posed such a threat. For Uniacke, the point is addressed because the right to life is forfeited only if the aggressor is an unjust immediate threat to the life of the victim. The reason why it is not permissible to kill an aggressor once she ceases her attack is that the aggressor is no longer an unjust immediate threat. As Uniacke states, ‘a person’s important rights to non-interference are 125 Kasachkoff, ‘Killing in self-defence’, at 519. See also J Dressler, ‘Rethinking heat of passion: a defense in search of a rationale’ (1982) 73 Journal of Criminal Law and Criminology 421–70, at 454; BJ Smart, ‘Understanding and justifying self-defence’ (1989) 4 International Journal of Moral and Social Studies 231–44; Ryan, ‘Self-defense, pacifism and the possibility of killing’. For this reason, Smart resorts to consequentialism in order to justify killing an innocent 126 aggressor. Kasachkoff, ‘Comment and reply’, at 639, emphasis in original. 127 Uniacke, ‘Response to two critics’, at 629; Uniacke, ‘Rights and relativistic justifications’, 128 at 645. Wallerstein, ‘Justifying the right to self-defense’, at 1013. 129 Kasachkoff, ‘Killing in self-defence’, at 517. See also Dressler, ‘Rethinking heat of passion’, at 454; GP Fletcher, ‘The right to life’ (1980) 63 The Monist 135–55, at 143.

The Justification of Self-Defence

65

not forfeited or abrogated in virtue of his malice (alone) but in virtue of what he is doing’.130 Another, albeit closely related, criticism has been made by Ryan, who argues that the forfeiture account permits self-defensive killing when the victim could have saved her life by some means other than killing the aggressor.131 Ryan illustrates his point with two victims who are both threatened by the same aggressor. One of the victims has a physical disability and cannot run away, whereas the second is a champion high jumper who can jump over a wall and escape. For Ryan, it would be permissible for the disabled victim to kill, but not the high jumper, yet, he states, it does not make sense to say that the aggressor has forfeited her right to life in the first example, but not in the second, given that the aggressor’s conduct is exactly the same in both cases. Of course, at this point, it might be said that this is only a valid criticism if it is accepted that there is some sort of duty to use the least harmful method of avoiding the threat posed by an aggressor. It is suggested here that there is. Earlier it was argued that the right to life is the most fundamental of all human rights on the basis that deprivation of life is impossible to remedy.132 If there is any reasonable possibility, therefore, of doing so, the loss of the aggressor’s life should be avoided (that is, the life of both the aggressor and the victim should be saved). If it is accepted that the life of the aggressor should be preserved wherever this is reasonably possible, Ryan’s criticism is difficult for Thomson and Uniacke to rebut. Thomson says only that an aggressor temporarily forfeits her right to life during the time in which she poses a threat to the victim. Using the high jump scenario as an example, it is difficult to avoid the conclusion that the aggressor poses a threat. It is simply that the threat can be avoided by jumping over the wall. Thus, it would be permissible under Thomson’s account for the high jumper to kill in self-defence. Uniacke holds that an aggressor forfeits her right to life only in so far as she poses an unjust immediate threat to the life of another. Here too, it is difficult to avoid the conclusion that the aggressor poses just such a threat and therefore the high jumper should be permitted to kill in self-defence, even though the threat could have been avoided in a less harmful way. It is left to Smart to point out a critical difference between the high jumper and the disabled victim: the disabled victim lacks the power to save both lives whereas the high jumper can do so.133 This crucial concept adds a missing link to the explanations proposed by Thomson and Uniacke. Where an aggressor poses an unjust immediate threat, and there is no other way in which the threat can be avoided by the victim, a point is reached where a 130 131 132 133

Uniacke, ‘Response to two critics’, at 629, emphasis added. See also Segev, ‘Fairness, responsibility and self-defense’, at 440. See Section 3.5.3 above. Smart, ‘Understanding and justifying self-defence’, at 234.

66

Killing in Self-Defence

choice has to be made between the life of the victim and the life of the aggressor. It is at this point (and not before) that the aggressor forfeits her right to life, thus permitting the victim to kill in self-defence. I would argue, then, that the right to life is forfeited by a combination of conduct and circumstances. Two conditions must be satisfied before forfeiture takes place: the aggressor must pose an unjust immediate threat to the life of the victim (the conduct element) and there must be no other way in which threat can be avoided (the circumstances element). For Ryan’s high jumper, these two conditions have not been fulfilled, as it is still possible to save the life of both the aggressor and the victim. For Ryan’s disabled victim, however, the circumstances requirement is satisfied, as it is not possible to save both lives. A choice must be made and it is the aggressor who forfeits her right to life by virtue of her conduct (in becoming an unjust immediate threat to the life of the victim) and the circumstances of the attack (by virtue of there being no other way in which the threat she poses can be avoided).134 The difference between Ryan’s two victims still holds even if the aggressor threatens them both together at the same time and in the same place. In this situation, it would be permissible for the disabled victim to kill the aggressor on the basis that the conduct and circumstantial conditions of forfeiture are satisfied. It would also be permissible for the high jumper to kill the aggressor, but only on the basis that she was acting in defence of the disabled victim, in which case, as before, the two conditions necessary for forfeiture to occur would be satisfied. If she killed in order to save herself, the high jumper’s act would not be permissible because the circumstantial requirement would not be met.

3.5.7 The rights/forfeiture approach to the justification of self-defensive killing: a summary Under the rights/forfeiture account, killing in self-defence is permissible because all human beings possess a right to life and an aggressor threatens to violate this right. The victim therefore has a right to kill the aggressor in order to protect herself from the violation. The reason the victim is permitted to kill the aggressor, but the aggressor is not permitted to kill the victim, is that the aggressor, by virtue of her conduct in becoming an unjust immediate threat to the life of the victim that cannot be avoided by any less harmful means, forfeits her right to life. 134 In this, my account is similar to that of Wallerstein’s theory of forced consequences (see ‘Justifying the right to self-defense’). The only difference of any substance between us is that Wallerstein does not accept that the right to life can be forfeited, preferring instead the explanation that, because the aggressor has created the situation in which either she or the defender will be harmed, the defender does not wrong the aggressor in using proportionate force to defend herself. On this, see Section 3.5.5 above, where Wallerstein’s account is criticized.

The Justification of Self-Defence

67

It is recognized that an account of self-defensive killing that relies on the concept of forfeiture is likely to be controversial. This is because of the term’s association with fault, penalty, or punishment. Once again, though, it must be stressed that the term is not being used in this sense. Forfeiture is associated here with conduct alone. In no way is it being suggested that it is permissible to kill an aggressor because she deserves it or because she is being punished for her conduct. Admittedly, one of the definitions of ‘to forfeit’ in the Oxford English Dictionary is ‘to lose by misconduct’. It is also defined simply as ‘to lose; to lose the right to’ and, in another alternative, ‘to lose or give up’.135 It is in this non-punitive sense that the term is used here. Any connotations of penalty or punishment could, of course, be avoided by using Uniacke’s alternative terminology of specification whereby human beings possess the right to life only in so far as they do not become an unjust immediate threat to the life of another. For Uniacke, the right to life is not forfeited, it is just that the aggressor does not fall within its scope in the first place.136 As I have already suggested, I consider there to be very little difference between specification and forfeiture, but regard it as more honest to use the terminology of forfeiture. The term forfeiture more accurately captures the process in question. The aggressor, by virtue of her conduct, forfeits (ie loses) the right to life that she normally would have possessed if she had not engaged in that conduct. I do recognize, though, that although my own preference is for forfeiture, some would still prefer to use the language of specification. If this is the case, I will make no further attempts to persuade them otherwise because it makes no real difference to the overall argument. In summary, the rights/forfeiture account is the most convincing account because it has five key advantages over other justifications of self-defensive killing. First, unlike either consequentialist accounts or the personal partiality account, it explains why it is permissible to kill an aggressor but not a bystander. This is because the aggressor poses a direct threat to the victim and therefore forfeits her own right to life, whereas the bystander does not. Second, unlike the personal partiality account, it explains why a victim is permitted to favour her own life over that of an aggressor. This is because the aggressor, by virtue of becoming an unjust immediate threat to the life of the victim, has forfeited her right to life. Third, it provides an explanation of why it is permissible to kill an innocent aggressor and a passive threat. This is because forfeiture of the right to life is based not on fault, but on conduct (and, in my proposed modification, on circumstances). The innocent aggressor, although not responsible for her attack is, nonetheless, by virtue of her conduct, posing an unjust threat to the life of the victim. The passive threat, through no fault of her own, nonetheless does likewise. Fourth, it explains why self-defensive killing is permissible only for the duration of the threat 135

Oxford English Dictionary online. Although, as Section 3.5.5 demonstrated, Uniacke falls into the trap of using the terminology of forfeiture on occasion. 136

68

Killing in Self-Defence

posed by the aggressor, and not, for example, after her attack has ceased. This is because forfeiture is linked to conduct in that the right to life is forfeited only in so far as the aggressor poses an immediate threat to the life of the victim. Once her attack has ceased, or she has become incapacitated, she no longer poses an immediate threat and therefore it is not permissible to kill her in ‘self-defence’. Fifth, and finally, with the modification suggested here, the forfeiture approach explains why there might be a duty to use the least harmful method of avoiding a threat. This is because forfeiture is also linked to circumstances. It is only in circumstances where there is no reasonable possibility of saving both the life of the aggressor and the life of the victim that the aggressor forfeits her right to life. This means, for example, that if there is a reasonable opportunity of escaping the attack open to the victim,137 it is not permissible for her to kill in ‘self-defence’. 137

See Chapter 4 for further discussion of this.

4 Retreat 4.1 Introduction The question of whether the accused has a duty to retreat before killing in self-defence is one that has long troubled the law. On the one hand, it might be argued that the innocent victim of an attack has the right to stand her ground against an aggressor and to defend herself with force, even if this could have been avoided by running away. On the other hand, it might be argued that, if the criminal law is concerned with minimizing violence and protecting all human life, there is a duty on the victim of an attack to avoid the death of the aggressor if a reasonable opportunity to escape exists. In this chapter, it will be argued that there is a duty to retreat before killing in self-defence, whenever a reasonable opportunity to do so exists.1 The argument draws on material introduced in Chapter 3, where it was suggested that the right to life is a fundamental right possessed by all human beings and thus human life should be preserved wherever this is reasonably possible.

4.2 How might the Law of Self-Defence Deal with Retreat? There are at least four ways of dealing with the issue of retreat: an absolute retreat rule, a strong retreat rule, a weak retreat rule, or no retreat rule. In any jurisdiction, a single rule might apply to all cases of self-defence, or different rules might apply depending on the circumstances.2

4.2.1 An absolute retreat rule An absolute retreat rule treats retreat as an independent variable in the consideration of any potential defence of self-defence. It requires the accused 1 This duty only applies directly to the accused who is defending herself. A retreat rule would make little sense where the accused is defending others, as by retreating she would no longer be able to prevent the attack. The equivalent duty in this context might be a duty to facilitate the retreat of the victim, rather than prevent the attack by using force. 2 For example, some US states operate a no retreat rule if an attack takes place in the home but a strong retreat rule in all other circumstances (see Section 4.4 below).

70

Killing in Self-Defence

to make an attempt to escape before using force in self-defence regardless of the circumstances.3 If the accused did not make at least an attempt to escape from an attacker before using self-defensive force, then she would be denied the benefit of the defence. For a brief period, it seemed that English law might have had an absolute retreat rule when, in R v Julien,4 the Court of Appeal held that a defendant must demonstrate an unwillingness to fight before she could rely on the defence of self-defence. Prior to Julien, there was little English authority on retreat, other than that of the early writers such as Blackstone and Foster,5 but such authority as did exist suggested that, if anything, English law contained a strong retreat rule.6 Julien involved an altercation between the complainant and the appellant. The complainant had armed himself with ‘a chopper’,7 and threatened the appellant, who responded by throwing a milk bottle. The appellant claimed to have acted in self-defence but was convicted of assault after the trial judge instructed the jury that there was an obligation to retreat before using force in self-defence. On appeal, counsel for the appellant submitted that no such duty existed in English law and that ‘an Englishman is not bound to run away when threatened, but can stand his ground and defend himself where he is’.8 Widgery LJ rejected this submission and held that, in order to benefit from the defence of self-defence, a defendant must first demonstrate an unwillingness to fight: It is not, as we understand it, the law that a person threatened must take to his heels and run . . . but what is necessary is that he should demonstrate by his actions that he does not want to fight. He must demonstrate that he is prepared to temporise and disengage and perhaps make some physical withdrawal; and to the extent that that is necessary as a feature of the justification of self-defence, it is true, in our opinion, whether the charge is a homicide charge or something less serious.9

On one interpretation this does come close to an absolute retreat rule. Widgery LJ appears to be implying that even where a safe opportunity to escape or retreat does not present itself, the defendant must take some positive action to demonstrate that she does not wish to fight before she will be able to plead self-defence. If this ever was the case, however, English law has since rejected such a rule. English law presently contains a weak retreat rule10 and Julien was at least implicitly disapproved in what must now be regarded as the leading 3 This might initially seem an exceptionally harsh rule, but see Section 4.3.1 below and Chapter 6, which notes that it is in certain contexts, such as where the accused started the trouble, that the argument for an absolute retreat rule is most convincing. 4 [1969] 1 WLR 839. 5 For the views of Hawkins, Blackstone, Foster, and East on retreat, see Section 4.2.4 below. 6 R v Smith (1837) 8 Car & P 160, at 162; 173 ER 441, at 443; R v Odgers (1843) 2 N & Rob 479, at 483; 174 ER 355, at 356; R v Dakin (1828) 1 Lewin 166, at 167; 168 ER 999, at 1000. 7 8 9 10 ibid 842. ibid 843. ibid 844. See Section 4.2.3 below.

Retreat

71

case on retreat in self-defence, R v Bird.11 Aside from the specific circumstances of an initial aggressor’s attempt to plead self-defence,12 there is no evidence that any other common law jurisdiction operates an absolute retreat rule.

4.2.2 A strong retreat rule A strong retreat rule also treats retreat as an independent factor to be considered whenever the accused attempts to plead self-defence. The strong retreat rule differs from the absolute retreat rule, however, in that it requires the accused to make an attempt to retreat before using force in self-defence only if an opportunity to do so actually exists. The precise formulation of the rule can vary. The accused might be required to take any opportunity to escape that presents itself, however risky, or she might be required only to take any ‘clear’, ‘safe’, or ‘reasonable’ opportunity. Scotland is an example of a jurisdiction that has been relatively consistent in applying a strong retreat rule.13 The accused must have had no reasonable opportunity to escape the threat other than by using defensive force.14 If the threat could have been avoided by taking a reasonable opportunity to escape, and the accused did not take that opportunity, then the defence will fail. The courts have stressed that the opportunity must be a reasonable one if failing to take it is to bar a successful plea of self-defence. This was made clear in the leading case of McBrearty v HM Advocate.15 Here, the accused appealed against his murder conviction on the basis (among others) that the trial judge did not emphasize that any opportunity to escape must be reasonable.16 The Lord Justice-General (Rodger) commented that it would have been more precise if the trial judge had mentioned that a means of escape must be reasonable, rather than just any opportunity, however risky.17 Lord Coulsfield agreed: . . . it is no doubt true that a person who is under threat cannot be expected to use a means of escape which exposes him to equal or greater danger rather than use force to defend himself. In that sense, it is no doubt correct that, in appropriate 11 [1985] 1 WLR 816. In Bird, Lord Lane, one of the judges who was on the bench in Julien (he concurred without delivering a separate opinion) accepted that the words used in Julien were not the most suitable as, on reflection, ‘they placed too great an obligation on a defendant’ (at 12 820). For discussion of Bird, see Section 4.2.3 below. See Section 4.2.1 below. 13 Although see D Hume, Commentaries on the Law of Scotland Respecting Crimes (4th edn, 1844) i, 217, who advocates a ‘no retreat’ rule where the accused was not at fault in creating the need to use self-defensive force, reflecting the views of English writers of the time (see Section 4.2.4 below). Such a proposition would have found no support in either the Scottish case law of the time or the work of earlier Scottish writers. For discussion, see J Chalmers and F Leverick, Criminal Defences and Pleas in Bar of Trial (2006), chapter 3. 14 McCluskey v HM Advocate 1959 JC 39, per Lord Russell at 44; Fenning v HM Advocate 1985 JC 76, at 78; Burns v HM Advocate 1995 JC 154, at 157; Pollock v HM Advocate 1998 15 SLT 880, at 882. 1999 SLT 1333. 16 17 ibid 1334. ibid 1336.

72

Killing in Self-Defence

circumstances, a jury may have to consider whether any means of escape open to the person under threat were reasonable.18

It did seem for a while that English law operated a strong retreat rule,19 although this may only have been in cases of excusable homicide, where the accused was not without fault in creating the circumstances that led to the need to use self-defensive force.20 It is certainly not the case any more.21 Scotland is not alone in the common law world in having a strong retreat rule. A number of US states (following the Model Penal Code22) take the same approach, at least in relation to the use of deadly force23 although the majority of these make an exception where the defendant was in her dwelling or workplace.24

4.2.3 A weak retreat rule A weak retreat rule means that the possibility of retreat is not considered as an independent condition that must be satisfied if a plea of self-defence is to succeed. Rather, whether or not the accused made an attempt to retreat is simply one factor that is taken into account in deciding whether her actions were necessary or reasonable. Admittedly in some circumstances the strong and the weak retreat rule are very close, given that using force in self-defence when a safe opportunity to escape existed can hardly be considered reasonable, but conceptually they are distinct. The distinction lies in the fact that, under a weak retreat rule, the legal possibility is at least open that the accused could ignore an obvious and perfectly safe opportunity to escape from the attacker and still be acquitted on the basis of self-defence. Under a strong retreat rule, this legal possibility is closed.25 English self-defence law presently has a version of the weak retreat rule whereby the possibility of retreat is merely one element for the trier of fact to consider in deciding whether the conduct of the defendant was reasonable. This was made clear in R v Bird.26 Here, the appellant had been convicted of 18

1999 SLT 1337. R v Smith (1837) 8 Car & P 160, at 162; 173 ER 441, at 443; R v Odgers (1843) 2 N & Rob 479, at 483; 174 ER 355, at 356; R v Dakin (1828) 1 Lewin 166, at 167; 168 ER 999, at 20 1000. See Section 4.2.4 below. 21 English law now contains a version of the weak retreat rule. See Section 4.2.3 below. 22 § 3.04(2)(b)(ii). Some exceptions to the rule are set out in § 3.04(2)(b)(ii)(A) of the Model Penal Code, most notably where the accused is attacked in her dwelling or place of work (discussed in Section 4.4 below). 23 Ala Code § 13A-3-23(b)(1); Ark Code § 5-2-607(b)(1); Del Code Ann tit 11 § 464(e)(2); Me Rev Stat Ann tit 17-A, § 108(2)(C)(3)(a); Neb Rev Stat § 28-1409(4)(b); NY UCC Law 24 § 35.15(2)(a); 18 Pa Cons Stat § 505(b)(2)(i). Discussed in Section 4.4 below. 25 Although this would not, of course, prevent a jury from returning a not guilty verdict in a jurisdiction, such as England or Scotland, where the jury does not have to justify its decision in law. 26 [1985] 1 WLR 816. Although cf R v Whyte [1987] 3 All ER 416, where something akin to a strong retreat rule seems to have been applied. 19

Retreat

73

unlawful wounding after a fight with her ex-boyfriend at a party. He had arrived at the party with a new girlfriend and Bird poured a glass of Pernod over him causing him to slap her round the face and, according to the appellant’s version of events, pin her against a wall. She lunged at him with the hand holding the Pernod glass and hit his face, injuring him so badly that he lost an eye. The trial judge instructed the jury along the lines of Julien27 that the defendant must have demonstrated an unwillingness to fight before she could benefit from the defence of self-defence: You cannot wrap up an attack in the cloak of self-defence and it is necessary that a person claiming to exercise a right of self-defence should demonstrate by her action that she does not want to fight. At one time it was thought that in order to demonstrate that the person seeking to raise a question of self-defence had to retreat. That is not so any longer at all, but there is an obligation to see whether the person claiming to exercise the right of self-defence should have demonstrated that she does not want to fight at all.28

On appeal, Lord Lane CJ held that this was a misdirection. Instead, he held that the correct position could be found in a passage from Smith and Hogan: ‘There were formerly technical rules about the duty to retreat before using force . . . It is submitted that this is now simply a factor to be taken into account in deciding whether it was necessary to use force, and whether the force used was reasonable.’29 As such, the appellant’s conviction was quashed on the basis that the trial judge had required her to demonstrate something (an unwillingness to fight) that, on her account of events, she was unable to do, as she was pinned to the wall by the complainant. Although no authority is provided in Smith and Hogan, it is likely that the authors were influenced by R v Howe,30 where the Australian High Court held that: . . . to retreat before employing force is no longer to be treated as an independent and imperative condition if a plea of self-defence is to be made out . . . [T]here can be no doubt at this day that whether a retreat could and should have been made is an element for the jury to consider as entering into the reasonableness of the defendant’s conduct.31

The High Court relied in turn on a much earlier decision of the US Supreme Court in Brown v United States,32 where it was stated by Holmes J that: ‘Rationally the failure to retreat is a circumstance to be considered with all 27

28 R v Julien [1969] 1 WLR 839 (see Section 4.2.1 above). ibid 819. At 820. The passage stems from the second edition of JC Smith and B Hogan, Criminal Law (1969), at 230–1. This passage also appears almost word for word in what is, at the time of writing, the most up-to-date edition of Smith and Hogan (11th edn, 2005), at 336. 30 (1958) 100 CLR 448. 31 ibid 463. See also Zecevic v DPP (1987) 162 CLR 645 (High Court of Australia), per 32 Wilson, Dawson and Toohey JJ at 663. 256 US 335 (1921). 29

74

Killing in Self-Defence

the others in order to determine whether the defendant went farther than he was justified in doing; not a categorical proof of guilt.’33 Canadian law takes a similar approach. The Canadian Criminal Code contains a retreat rule in relation to provoked attacks but no such requirement in relation to unprovoked attacks,34 perhaps mirroring the historical distinction between justifiable and excusable homicide in English law.35 A weak retreat rule has, however, been read into the law, with a failure to retreat being a factor to consider in determining whether the accused believed on reasonable grounds that she could not otherwise preserve herself from death or serious bodily injury.36

4.2.4 No retreat rule Finally, it could be that retreat is of no significance whatsoever to the success or otherwise of a plea of self-defence. That is, there is no duty on the accused to take any opportunity, even a clear and safe one, to escape from an aggressor before using force in self-defence. In a jurisdiction that operates a no retreat rule, the victim of an attack has the right to stand her ground and meet force with force. For a time in English law the victim of certain types of attack (sudden and unexpected attacks for which the defendant was in no way to blame) was under no obligation to retreat. Historically, English law made a distinction between justifiable and excusable homicide.37 A justifiable homicide was one that led to a complete acquittal and no further punishment. An excusable homicide was one in which the defendant was not completely free from blame and was originally required to obtain a royal pardon before being released. The consequences of the distinction were, at one time, far reaching, with a finding of excusable homicide meaning that the defendant’s movable property was forfeited to the Crown.38 33 ibid 343. The Italian law of self-defence also contains a weak retreat rule: for discussion see S Skinner, ‘Populist politics and shooting burglars: comparative comments on the Lega Nord’s proposal to reform Italian self-defence law’ [2005] Crim LR 275–84. 34 Canadian Criminal Code s 34(2) (self-defence against unprovoked assault); s 35(c) (self-defence in case of aggression). Although see R v McIntosh [1995] 1 SCR 686, which left the distinction between these provisions in a state of some confusion. For reform proposals, see Canadian Department of Justice, Provocation, Self-Defense and Defence of Property: A Consultation Paper (1998). 35 That this was the origin of the distinction was the conclusion reached in R v McIntosh [1995] 1 SCR 686 (Supreme Court of Canada), per McLachlin J at [67]. On the distinction between justifiable and excusable homicide in English law, see Section 4.2.4 below. 36 Northwest v R (1980) 22 AR 522 (Alberta Court of Appeal), at [27]; R v Proulx (1998) 127 CCC (3d) 511 (British Colombia Court of Appeal), at [49], adopting the pronouncement of Holmes J in Brown v United States 256 US 335 (1921); R v Ward (1978) 4 CR (3d) 190 (Ontario Supreme Court, Court of Appeal), at [6]. 37 For a detailed account of the origins of this distinction, see TA Green, ‘The jury and the English law of homicide 1200–1600’ (1976) Michigan Law Review 414–99; ND Hurnard, The King’s Pardon for Homicide before AD 1307 (1969). 38 The practice of requiring a pardon in cases of excusable homicide had all but died out by the end of the sixteenth century, but forfeiture remained in operation for some time after this.

Retreat

75

From at least the eighteenth century onwards, cases of self-defence in which the defendant was subjected to a sudden and unexpected attack for which she was in no way to blame were classified as justifiable homicide.39 No retreat rule attached to cases of justifiable homicide, the defendant being permitted to stand her ground in the face of such an attack.40 Where the defendant was not wholly without fault in the incident, the case was one of excusable homicide41 and a strong retreat rule attached.42 The distinction between justifiable and excusable homicide no longer plays any part in English law,43 which now operates a weak retreat rule in relation to all cases of self-defence.44 Examples of jurisdictions in which a no retreat rule operates, however, still exist. Indeed, the no retreat rule is specifically provided for in the self-defence provisions of some US states, despite the Model Penal Code requiring retreat where a safe opportunity to do so exists.45 One example is Florida, which recently passed a ‘stand your ground’ bill,46 which permits any person who is not engaged in unlawful activity and who is attacked in any place where she has a right to be to use deadly force against an attacker without having to take a safe opportunity to escape.47 Certainly Hawkins refers to it in A Treatise of the Pleas of the Crown (3rd edn, 1739), although he does seem to be the last influential English writer to do so. 39 Hawkins, Treatise, at 69; M Foster, A Report of some Proceedings on the Commission of Oyer and Terminer and Goal (sic) Delivery for the Trial of the Rebels in the year 1746 in the County of Surry, and of other Crown Cases to which are added Discourses upon a few Branches of the Crown Law (1762), at 289; W Blackstone, Commentaries on the Laws of England: Volume 1 (1769), at 203; EH East, Pleas of the Crown (1803), at 219. For a long time, self-defence was excluded entirely from the category of justifiable homicide and it was not until the eighteenth century that cases of self-defence in which the defendant was blameless came to fall into this category (Green, ‘The jury and the English law of homicide’; F Leverick, A Critical Analysis of the Law of Self-Defence in Scotland and England (2003), chapter 4). 40 Foster, Report, at 273; Blackstone, Commentaries, at 205; East, Pleas, at 272. Cf Hawkins (who was writing at around the same time), at 73–5. 41 Hawkins, Treatise, at 75; Foster, Report, at 289; Blackstone, Commentaries, at 208; East, Pleas, at 221, 279. 42 Foster, Report, at 276; Blackstone, Commentaries, at 205; East, Pleas, at 280. By the eighteenth century the practical consequences of a finding of justifiable or excusable homicide differed very little (excusable homicide no longer required a pardon and did not, by this time, attract forfeiture). The two categories did, however, retain separate rules on retreat until the distinction itself became obsolete. 43 Offences against the Person Act 1828 s 10; R v Semini [1949] 1 KB 405, at 408–9. See also the first edition of Smith and Hogan, Criminal Law (1965), at 233 and HA Snelling, ‘Killing in self-defence’ (1960) 34 Australian Law Journal 130–8, at 135. 44 45 See Section 4.2.3 above. § 3.04(2)(b)(ii). 46 HB 0249, An act relating to the protection of persons and property. The Act amends the existing Criminal Code provisions relating to retreat (§§ 776.012 and 776.031) and adds a new section, § 776.013. See A Goodnough, ‘Florida expands right to use deadly force in self-defense’, New York Times, 27 April 2005. 47 Prior to the passing of the Bill in 2005, a ‘no retreat rule’ only operated in Florida where the defendant was attacked in her dwelling or workplace (see the Supreme Court of Florida in Weiand v State 732 So 2d 1044 (1999), at 1048). Florida is not the only US state to operate such a rule: see Tenn Code Ann § 39-11-611(a). See also the Supreme Court of Ohio in Erwin v State 29 Oh S 186 (1876), at 195; the Supreme Court of Indiana in Runyan v State 57 Ind 80 (1877), at 82; the US Supreme Court in Beard v United States 158 US 550 (1895), at 564. Cf the US Supreme Court in Allen v United States 164 US 492 (1897), at 497–8; the Supreme Court

76

Killing in Self-Defence

The German Penal Code contains a retreat rule where the attacker is an innocent aggressor, but not where the attacker is culpable,48 an approach that is discussed in more detail in due course.49

4.3 The Morally Preferable Option Of the four options identified above, the strong retreat rule is the most morally appropriate. The main reason for this is that in requiring retreat where it is possible with safety, the strong retreat rule promotes maximum respect for the right to life. As Chapter 3 argued, the right to life is a fundamental right. Maximum respect is accorded to that right if the law requires that if there is any possible way of avoiding taking the life of an aggressor, such as taking a safe opportunity to escape, then the accused should take that option. In this way, two lives are saved, rather than one life lost, whether that life is the life of the aggressor or the self-defender.50 This is also the approach taken by the US Model Penal Code,51 the commentary to which states that ‘the protection of life has such a high place in the proper scheme of social values that the law should not permit conduct that places life in jeopardy, when the necessity for doing so can be avoided by the sacrifice of the much smaller value that inheres to standing up to an aggression’.52 Indeed, it has been suggested that, in US states that have no retreat rule,53 considerable loss of life can be attributed to self-defenders who chose to stand their ground rather than take a safe opportunity to escape.54 Why, though, should a strong retreat rule be preferred over the other three options identified? To fully answer this question, it is necessary to consider why each of the three alternative options should be rejected.

4.3.1 The absolute retreat rule The absolute retreat rule is the easiest of the options to reject. Indeed, when English law seemed briefly to flirt with this possibility in Julien,55 the rule of Pennsylvania in Commonwealth v Drum 58 Pa 9 (1868), at 22; the Circuit Court of Multnomah County in State v Conally 3 Or 69 (1869), at 74. For a modern example of a no retreat rule that has developed through case law, see the Supreme Court of Michigan’s decision in People v Riddle 649 NW 2d 30 (2002). 48 German Penal Code s 32. The distinction between culpable and innocent aggressors was 49 explained in Chapter 1. See Section 4.3.2. 50 A Ashworth, ‘Self-defence and the right to life’ (1975) 34 CLJ 282–307, at 290. 51 See Section 4.2.2 above. 52 Commentary to the Model Penal Code, at 54. See also the prosecution’s argument in 53 Brown v United States 256 US 335 (1921), at 340. See Section 4.2.4 above. 54 RM Brown, No Duty to Retreat: Violence and Values in American History and Society (1994). Brown cites a study by the sociologist Henry Lundsgaarde who concluded that, of a sample of 268 homicides in Houston, a considerable number of lives would have been saved if the killer had taken a safe opportunity to retreat rather than use force. 55 R v Julien [1969] 1 WLR 839.

Retreat

77

was eventually rejected as too harsh a statement of the law.56 If a right to act in self-defence is recognized at all, then it makes little sense to require the accused always to make an attempt to escape, even where there is no safe opportunity to do so. In doing so, she may lose the opportunity to defend herself entirely, resulting in the loss of her life. An absolute retreat rule might be more appropriate where the accused has ‘started the trouble’. If the accused is the first to use physical force, and the recipient of that physical force defends herself, then it does not seem unduly harsh to require the accused who started the trouble to retreat, or to make it clear that she no longer wishes to fight, before she can benefit from the defence of self-defence.57

4.3.2 The ‘no retreat’ rule At the other end of the spectrum to the absolute retreat rule stands the no retreat rule, which would allow the accused who ignored a completely safe opportunity to escape from an aggressor, and chose instead to kill, to benefit from a complete acquittal on the basis of self-defence. In favour of the no retreat rule, it could perhaps be said that it would be dishonourable, cowardly, or would involve a lack of self-respect to require an innocent victim of an attack to retreat. Beale, in an analysis undertaken over 100 years ago, reaches the conclusion that the only argument against having a strong retreat rule is that it is dishonourable and cowardly to retreat and dishonour should not be thrust upon someone who is without fault.58 This argument can be seen in two influential nineteenth-century US cases. In Erwin v State,59 the Supreme Court of Ohio approved a passage from the English writer Matthew Hale, who wrote that no ‘true man’, who was without fault for the encounter, was obliged to fly from an assailant.60 In Runyan v State,61 the Supreme Court of Indiana seemed to think that enforcing a duty to retreat would legitimize cowardice and that this would run contrary to ‘the tendency of the American mind’.62 Beale, however, proceeds to reject the ‘dishonour’ argument as follows: A really honourable man, a man of truly refined and elevated feeling, would perhaps always regret the apparent cowardice of a retreat, but he would regret ten times more, after the excitement of the contest was past, the thought that he had the blood of a fellow-being on his hands. It is undoubtedly distasteful to retreat; but it is ten times more distasteful to kill.63 56

See Section 4.2.1 above. This is the approach taken by the US Model Penal Code (see § 3.04(2)(b)(i) and the accompanying commentary). The issue is discussed further in Chapter 6. 58 JH Beale, ‘Retreat from a murderous assault’ (1903) 16 Harvard Law Review 567–82, 59 at 581. 29 Oh S 186 (1876). 60 ibid 195. Hale actually wrote this in the context of defending against housebreakers or highway robbers, and not in the context of self-defence against any other type of violent attack 61 (see M Hale, Historia Placitorum Coronae Volume 1 (1736), at 481). 57 Ind 80 (1877). 62 63 ibid 82. Beale, ‘Retreat from a murderous assault’, at 581. 57

78

Killing in Self-Defence

The point is made even more forcefully by the Court of Appeals of Alabama in Cooke v State:64 [The] doctrine of retreat is sometimes referred to as being cowardly, but not so; it is based upon the highest consideration of civilisation, morals and our holy religion. It is better that one man should flee rather than take human life, and at a time like the present, when human life is taken upon the least provocation, and observance of the law is held so lightly, and the very fabric of our laws seem to be trembling in the balance, it is well that we should restate and emphasise the rule, that it is better and right that a man should flee rather than he take human life, if he can do so without apparently increasing his danger to life and limb. It is no cowardly doctrine. It is the law of his sovereign state, which he is bound to obey; it is the law of God, which he ignores at his peril. It is the reasonable law of God and man.65

It is certainly difficult to see how the values of personal honour, dignity, or self-respect could possibly outweigh the value of human life and thus this argument for a no retreat rule is rejected here. Arguments for a no retreat rule have also been grounded in the right to freedom of movement. In State v Bartlett,66 the Supreme Court of Missouri stated: Because the right to go where one will without let or hindrance, despite of threats made, necessarily implies the right to stay where one will without let or hindrance. These remarks are controlled by the thought of a lawful right to be in the particular locality in which he goes, or in which he stays. It is true, human life is sacred, but so is human liberty; one is as dear in the eye of the law as the other, and neither is to give way and surrender its legal status in order that the other may exclusively exist, supposing for a moment such an anomaly to be possible. In other words, the wrongful and violent act of one man shall not abolish or even temporarily suspend the lawful and constitutional right of his neighbor.67

As was the case for the values of honour and dignity, however, it is difficult to see how the value of freedom of movement could possibly outweigh that of human life. One possible approach is to argue that an aggressor, by virtue of becoming an unjust immediate threat to the life of her victim, forfeits her own right to life. This approach was considered in more detail in Chapter 3, where it was argued that the most convincing account of the right to use self-defensive force is one based on rights and an accompanying theory of forfeiture. That is, a victim is permitted to kill an aggressor because the aggressor forfeits her own right to life by virtue of becoming an unjust immediate threat to the life of the victim. It was also argued in Chapter 3, however, that the aggressor’s right to life is forfeited only when there is no reasonable possibility of saving both the life 64

65 66 93 So 86 (1921). ibid 90. 71 SW 148 (1902). ibid 151, emphasis in original. A similar argument is made by the Court of Criminal Appeals of Oklahoma in Fowler v State 126 P 831 (1912), at 833. 67

Retreat

79

of the aggressor and the life of the victim. This was on the basis that the right to life is the most fundamental of all human rights, given that a deprivation of life is impossible to remedy. Chapter 3 argued, therefore, that the right to life is forfeited only when two conditions are satisfied: a conduct condition (the aggressor poses an unjust immediate threat to the life of the victim); and a circumstances condition (there is no reasonable way in which the threat can be avoided other than by killing the aggressor). If a safe opportunity to retreat exists, then the circumstances condition is not met and the victim should not be permitted to kill the aggressor. The view is taken here that this proposition holds even in the most extreme of examples, such as the racist attacker who intends to embark upon a programme of genocide. The innocent person who is attacked by the racist attacker should still be required to retreat if a safe opportunity to do so exists. The life of even the racist attacker with genocidal intentions is always worth more than the pride, dignity, honour, or freedom of movement of another individual. To conclude otherwise is to sanction the killing of human beings simply because they are ‘bad people’, an outcome that is surely unacceptable in any civilized society that values due process.68 Even if one accepts that culpable aggressors forfeit the right to life in circumstances where a safe opportunity to retreat exists (which I do not), it is difficult to see how this could possibly apply to an innocent aggressor (such as a child or an insane person) or a passive threat.69 German self-defence law deals with this point by distinguishing between innocent and culpable aggressors: there is a strong retreat rule when an attack is made by an innocent aggressor, but no retreat rule where the attack is made by a culpable aggressor, due to her blameworthiness in instigating the attack.70 This seems an odd idea, because it will be very difficult for the victim of an attack to know whether her attacker is culpable, in which case she is entitled to stand her ground, or innocent, in which case she is not.71 It also seems grossly inappropriate to have a no retreat rule where the accused reasonably but wrongly believes that she is being attacked. If the 68 The genocidal attacker scenario was posed by one of the original reviewers of this manuscript. While I am grateful for the comment, it is misplaced. My argument is that where the accused is acting in defence purely of herself, the character of the attacker is irrelevant. My conclusion would be different if the accused was acting in the defence of others, for example if she was acting to prevent an imminent programme of genocide and there was no other way in which it could be prevented. To require retreat in this scenario would make little sense, as the accused could no longer prevent the loss of life of the other parties (although there may, in this context, be a duty to facilitate the retreat of those threatened rather than use force). 69 For an explanation of these terms, see Chapter 1. 70 German Penal Code s 32; see also the (limited) discussion in K Bernsmann, ‘Private self-defence and necessity in German penal law and in the penal law proposal: some remarks’ (1996) 30 Israel Law Review 171–87, at 178; M Kremnitzer, ‘Proportionality and the psychotic aggressor: another view’ (1983) 18 Israel Law Review 178–214, at 207. 71 The German Penal Code is silent on how to treat the accused who failed to retreat because she mistakenly believed that she was being attacked by a culpable aggressor (when in fact she was being attacked by an innocent aggressor).

80

Killing in Self-Defence

accused who mistakenly believes that she is under attack is permitted to stand her ground and ignore any safe opportunity to escape, the life of a completely innocent person who was unfortunate enough to be mistaken for an aggressor may be lost, where this could have been avoided. One further argument in favour of a no retreat rule is that it might act as a deterrent to those considering the use of violence. That is, if an aggressor knows that her victim would be protected by the criminal law if she were to ignore an opportunity to escape, this is likely to deter at least some aggressors.72 This can be described as a rule-utilitarian argument.73 Against this, it must be doubted that an aggressor would be influenced by such a rule.74 It might possibly play a role in deterring repeat or professional offenders, perhaps in relation to offences such as robbery or housebreaking. It could equally be argued, however, that a no retreat rule could actually encourage aggression, with, say, the professional housebreaker arming herself more strongly because she knows that her victim has a more extensive permission to use force.

4.3.3 The weak retreat rule If a no retreat rule is rejected, is there perhaps an argument in favour of a weak retreat rule? One argument that is sometimes made is that a strong retreat rule places too onerous a duty upon the victim of a sudden and unexpected attack. This can be seen in the US Supreme Court case of Brown v United States,75 where Holmes J explains his rationale for adopting a weak retreat rule as follows: Detached reflection cannot be demanded in the presence of an uplifted knife. Therefore in this Court, at least, it is not a condition of immunity that one in that situation should pause to consider whether a reasonable man might not think it possible to fly with safety or to disable his assailant rather than to kill him.76

But this misunderstands the nature of a strong retreat rule. A strong retreat rule can take many forms, including requiring the victim of an attack to take only a safe or reasonable opportunity to escape. It does not necessarily require the accused to pause and consider whether or not escape is possible where doing so would compromise her safety. Indeed, to require this would be to adopt something bordering on an absolute retreat rule, a policy most would agree is entirely unfair. If, for example, a strong retreat rule required that the accused did not ignore any reasonable opportunity to escape, this 72 A Harel, ‘Efficiency and fairness in criminal law: the case for a criminal law principle of comparative fault’ (1994) 82 California Law Review 1181–229; M Kremnitzer and K Ghanayim, ‘Proportionality and the aggressor’s culpability in self-defense’ (2003–4) 39 Tulsa 73 Law Review 875–99, at 886. See Chapter 3. 74 PH Robinson and JH Darley, ‘Does criminal law deter? A behavioural science investigation’ 75 (2004) 24 OJLS 173–205. 256 US 335 (1921). 76 ibid 343.

Retreat

81

would not mean that she needed to make long and drawn out checks in the face of an immediate life or death situation. Allowance can be made within the test of reasonableness for the panic that someone is likely to experience in the heat of the moment if she becomes the subject of an unexpected attack. As Horder has pointed out, writing in the context of the defence of duress, requirements of reasonableness can be sensitive to the fact that people may make mistakes under the pressure of circumstances.77 Another argument that is sometimes made in favour of the weak retreat rule is that it is easier to instruct a jury simply to decide whether, in their view, the accused’s use of force was reasonable, rather than running through a long list of independent factors for consideration. It also retains flexibility in respect of any as yet unforeseen circumstances in which enforcing a strong retreat rule would be inappropriate. This argument is not sufficiently persuasive to outweigh two arguments that can be made against a weak retreat rule: one rooted in morality, the other rooted in practicality. The most important of these is the first. If retreat is simply one factor to be taken into account in deciding whether force used in self-defence was reasonable, this implies that, in some circumstances, the law allows the accused to ignore a clear and safe opportunity to escape and still be judged to have used reasonable force. In other words, there are circumstances in which the accused could lawfully choose to stand her ground and use force rather than escape safely from an attacker. Ashworth has made this point, arguing that regarding the duty to retreat as merely one consideration to be borne in mind in an overall test of reasonableness is problematic because it says nothing about the circumstances that might outweigh it.78 If this is the correct interpretation of the weak retreat rule, it is indeed difficult to envisage what such circumstances could possibly be. Allowing the accused to make the choice to use violence on any occasion where this could have been avoided implies that the law does not respect the right to life of an aggressor and places some other value, such as honour, dignity, or vengeance, higher than that of the aggressor’s life. As the fundamental argument of this book is that all human life is of equal value, and should be preserved wherever this is reasonably possible, this position is rejected here. The second argument against the weak retreat rule is a practical argument. It has been suggested that the weak retreat rule is preferable because of its flexibility in respect of any, as yet unforeseen, circumstances, in which a strong retreat rule would be inappropriate. This flexibility presents its own dangers, though, especially in cases involving a jury, as it allows the jury too much leeway in deciding what is and what is not reasonable behaviour in the context of self-defence. 77 J Horder, ‘Occupying the moral high ground? The Law Commission on duress’ [1994] Crim LR 334–42, at 341. 78 A Ashworth, Principles of Criminal Law (5th edn, 2006), at 143.

82

Killing in Self-Defence

If a jury is not told specifically that the accused must take any safe opportunity to escape before using force in self-defence, they might acquit an accused who has ignored a safe opportunity to do so. This could be out of sympathy for a particular accused who has been the subject of an attack, or because the jury simply believe in general terms that an innocent victim should have the right to stand her ground against an attacker.79 It might equally be for more unpalatable reasons. It is not too much of a stretch of the imagination to envisage a racist jury acquitting a white accused who has stood her ground against a black attacker.80 Of course, in any jurisdiction where the jury does not have to justify its decision in law, a racist jury could do this anyway, even if a ‘strong duty to retreat’ instruction was given. The adoption of a weak retreat rule, however, increases the chances of this type of prejudiced jury decision-making. It also increases the likelihood of inconsistency of treatment.81 It might also be said that it is unfair to ask a jury to make decisions of this nature without offering precise guidance on the role that a failure to retreat should play in determining whether the defence of self-defence is made out.

4.3.4 The preferable option: a strong retreat rule The strong retreat rule is the morally preferable option. An absolute retreat rule was rejected on the basis that it places too onerous a duty on the victim of an attack always to attempt to escape before using self-defensive force. A no retreat rule was rejected on the basis that it fails to respect the right to life of the aggressor, by allowing this to be outweighed by some other value, such as the personal honour or freedom of movement of the accused. A weak retreat rule was rejected on the same basis. By failing to make retreat an independent variable in the law of self-defence, the possibility is left open that, in some circumstances, the victim of an attack could choose to use force in preference to taking a safe opportunity to escape. There is also a more minor practical argument against a weak retreat rule in that it increases the possibility of bias and 79 The letters page of the Guardian (April 21 2000) on the day after the jury verdict in the Tony Martin case suggests that this is a commonly held view, even among those who read the Guardian (a national newspaper whose editorial policy would be considered relatively left-wing). Tony Martin shot and killed a housebreaker and was denied the defence of self-defence (see R v Martin [2001] EWCA Crim 2245 and Chapter 7, where Martin is discussed in more detail). 80 Kalven and Zeisel, in their study of American juries, chart a number of cases where juries have delivered an acquittal seemingly because they sympathized with the defendant who faced ‘attack’ by a black person. In one example, the (white) defendant was acquitted on the basis of self-defence when he stabbed the (black) victim in the stomach with a knife after being forcibly ejected from a tavern. It is apparent from the facts of the case that, even if the defendant did face any threat of attack (and it was by no means clear that he did), he could have escaped without injury simply by walking away from the tavern (H Kalven Jr and H Zeisel, The American Jury (1966), at 225). 81 The ‘practical’ argument against the weak retreat rule perhaps disappears if a jury is not involved. Ashworth has suggested, though, that even judges can conceal biases and assumptions behind a test of ‘reasonableness’ (Ashworth, ‘Self-defence and the right to life’, at 307).

Retreat

83

arbitrariness in decision-making, especially where a jury is involved. A strong retreat rule is favoured over the alternatives not solely because of their weaknesses, but also because it respects the right to life of both aggressor and victim and does not allow this to be outweighed by lesser values.

4.4 Retreat from the Home: A Special Case? In some US states where retreat is required, an exception is made when an attack takes place in the accused’s dwelling82 or, in addition to this, at her place of work.83 These states follow the principle set out in the Model Penal Code, which provides that ‘[t]he actor is not obliged to retreat from his dwelling or place of work, unless he was the initial aggressor or is assailed in his place of work by another person whose place of work the actor knows it to be’.84 No explanation for the exception is provided in the Commentary to the Code beyond a brief reference to ‘sentimental factors relevant to dwellings’85 in the context of a discussion of whether the ‘dwelling’ exception should be extended to the workplace.86 The basic principle has been extended in some states. For example, in State v Blue,87 the Supreme Court of North Carolina held that a front porch could constitute part of ‘the home’ for the purposes of the no retreat rule.88 In Commonwealth v Daniels,89 the Supreme Court of Pennsylvania applied the same principle to the common hallway of an apartment block, in a case where the appellant could almost certainly have avoided killing in self-defence by retreating into his own apartment.90 Going even further, 82 Ark Code § 5-2-607; Me Rev Stat Ann tit 17-A § 108(2)(C)(3)(a); NH Rev Stat Ann § 627.4(III)(a); NY UCC Law § 35.15(2)(a)(i); Tex Penal Code § 9.32(b). The precise wording of the exception varies from state to state. In Alaska, for example, an exception is made where the defendant is on premises owned or leased by her (Alaska Stat § 11.81.335(b)). 83 Ala Code § 13A-3-23(b)(1); Del Code Ann tit 11 § 464(e)(2)(b); Neb Rev Stat § 28-1409(4)(b)(i); ND Cent Code § 12.1-05-07(2)(b)(2); 18 Pa Cons Stat § 505(b)(2)(ii). See also 9 Guam Code Ann § 7.86(b)(2)(A). In some states the exception is even wider. In Utah, there is an exception to the retreat rule whenever the defendant is in a place she has ‘lawfully entered and remained’ (Utah Code § 76-2-402(3)). These examples are all drawn from states where a retreat rule would normally operate and to which these situations provide exceptions. There are, of course, also states where there is no retreat rule in the first place: see Section 4.2.4 above. 84 85 § 3.04(2)(b)(ii)(A). Commentary to the Model Penal Code, at 56. 86 The drafters concluded that it should: although the ‘sentimental factors relevant to dwellings may not apply to one’s place of work . . . the practical considerations concerning the two locations were far too similar to sustain a distinction’ (Commentary to the Model Penal 87 Code, at 56). 565 SE 2d 133 (2002). 88 See also State v Bonano 284 A 2d 345 (1971), a decision of the Supreme Court of New Jersey, which reached the same conclusion. Cf Commonwealth v Bennett 671 NE 2d 966 (1996), where the Appeals Court of Massachusetts held that the defendant’s driveway did not constitute part of the home because it had ‘none of the closed and secure qualities of an occupied 89 dwelling’ (at 967). 451 Pa 163 (1973) (Pomeroy J dissenting on this point). 90 Cf People v Childs 250 NYS 2d 926 (1964), at 928 (Supreme Court, Appellate Division, Second Department, New York); People v Hernandez 744 NE 2d 198 (2002) (Court of Appeals of New York).

84

Killing in Self-Defence

Louisiana applies the no retreat rule to any attack taking place in a motor vehicle.91 The rationale for making such an exception would seem to lie in two interrelated themes. The first is that requiring retreat from the home would unfairly expose the accused to the very dangers that the home is supposed to protect against. The point is made by the Supreme Court of Wyoming in Palmer v State,92 which held that retreat was not required from the home because: There are many lonely ranches miles away from any help or any safe place of retreat, and they are not infrequently occupied by persons without other protection or defence than that which they can make for themselves. That any man or woman so situated must first look about for means of escape before they can defend themselves against impending danger is not the law. It would not benefit community or tend to make life safer.93

The argument has also been made in relation to battered women. To require a woman who faces physical violence from her partner to retreat, it has been said, might come at enormous personal cost, as she is unlikely to find a place of safety in which she can escape her partner94 and indeed the attempt to flee might expose her to the risk of future violence from him.95 It might even necessitate her leaving her children with her violent partner and therefore subjecting them to possible danger.96 But this argument misunderstands the nature of the retreat rule. Even under a strong retreat rule (or at least one formulated to require only a reasonable opportunity to retreat to be taken), the accused is not required to utilize an opportunity to escape that exposes her (or indeed any other occupant of the house) to greater danger. The purpose of the retreat rule is to attempt to preserve the lives of both aggressor and victim. If this would not be accomplished by requiring the accused to retreat, then there would simply be no duty to do so in those jurisdictions that have a strong retreat rule. It might also be questioned whether retreat from the home is really fraught with as much danger as supporters of this argument imply. It has been suggested that the rule stems from a time when society was less civilized and retreat from one’s dwelling might well have been associated with increased

91

92 93 La Rev Stat § 14:20(4)(a). 59 P 793 (1900). ibid 796. A Gregory, ‘To have but not to hold: can “resistance against kidnapping” justify lethal self-defense against incapacitated batterers?’ (2002) 102 Columbia Law Review 729–73. Cf State v Garland 694 A 2d 564 (1997), where the New Jersey Supreme Court reaffirmed an exception to the ‘no retreat from the home’ rule for cohabitants, with the effect that the accused who faces violence at home from her cohabitant does have a duty to retreat whereas the accused faced with violence from a stranger in her home does not. For criticism, see M Wheatcroft, ‘Duty to retreat for cohabitants in New Jersey: a battered spouse’s home is not her castle’ (1999) 95 30 Rutgers Law Journal 539–67. See Chapter 5. 96 C Wells, ‘Battered woman syndrome and defences to homicide: where now?’ (1994) 14 LS 266–76. 94

Retreat

85

peril. In a more civilized modern society this is not an assumption that can so easily be made.97 A second rationale resides in the notion that the home is a sanctuary, from which the accused should never be required to flee. Although closely related to the first rationale, there is a subtle difference between the two. This second rationale relates more to the right of the accused to remain in her home simply on the basis that it is her home and her source of shelter and sanctuary. This can be seen in People v Tomlins,98 where the Court of Appeals of New York justified the no retreat from the home exception as follows: It is not now, and never has been the law that a man assailed in his own dwelling, is bound to retreat. If assailed there, he may stand his ground, and resist the attack. He is under no duty to take to the fields and highways, a fugitive from his own home . . . Flight is for sanctuary and shelter, and shelter, if not sanctuary, is in the home.99

A similar passage can be found in Barton v State,100 where the Court of Special Appeals of Maryland held that: If the peril—the attack—occurs in his home, his dwelling place, he need not retreat from it, but may stand his ground and use whatever force is reasonable . . . in order to repel the attack and defend himself. He is not bound to flee and become a fugitive from his own home, for, if that were required, there would, theoretically, be no refuge for him anywhere in the world.101

This rationale is sometimes linked to the concept of ‘an Englishman’s home being his castle’,102 but this is almost certainly a misunderstanding of the principle. The case from which the castle doctrine derives is Semayne’s case,103 but in fact the passage does not relate to the retreat rule at all but to the permissibility of killing in defence of the home itself: ‘That the house of every one is to him as his castle and fortress, as well as his defence against injury and violence, as for his repose . . . So it is held in 21 H 7 39104 every one may assemble his friends and neighbours to defend his house against violence.’105 Notwithstanding the origins of this argument, it is rejected here. To hold that the accused has a right to remain in her home at all costs, even when she could safely retreat and spare the life of an aggressor, places the right to enjoy property above the right to life of the aggressor. This value judgement has already been rejected here. An aggressor has a right to life that is worthy of respect, even if doing so requires the accused temporarily to leave her dwelling. 97 CL Carpenter, ‘Of the enemy within, the castle doctrine, and self-defense’ (2002–3) Marquette Law Review 653–700, at 659; RV Lee, ‘A further erosion of the retreat rule in North Carolina’ (1976) 12 Wake Forest Law Review 1093–103, at 1100. 98 99 107 NE 496 (1914). Tomlins, per Cardozo J at 498. 100 101 420 A 2d 1009 (1980). ibid 1010. 102 Carpenter, in ‘Of the enemy within’, at 656; F McAuley and PJ McCutcheon, Criminal Liability (2000), at 761; J Dressler, Understanding Criminal Law (3rd edn, 2001), at 228. 103 (1604) 5 Co Rep 91a. 104 A reference to the 1532 English statute entitled ‘That a Man killing a Thief in his Defence, 105 shall not forfeit his Goods’. ibid 91b.

This page intentionally left blank

5 Imminence of Harm 5.1 Introduction In the context of self-defence, imminence refers to the time period between the harm faced by the accused and the defensive action taken to prevent that harm from materializing, with some jurisdictions only permitting the defence if the harm prevented was imminent. The term ‘imminent’ is sometimes used interchangeably with the term ‘immediate’ and, for the purposes of this chapter, it is assumed that there is no meaningful difference between the two.1 Both refer to the requirement that there be no significant time gap between the defensive response made by the accused and the time at which the harm she faced would have materialized. Straight away, it should be noted that the issue of imminence is closely related to that of retreat.2 This is because if harm is not imminent, the accused is likely to have a reasonable opportunity to retreat and avoid that harm. The issues of imminence and retreat are not, however, synonymous. Imminence raises the additional issue of probability of harm. When harm is not imminent, it is normally less likely that the harm will occur at all, because the person who poses the threat might never actually carry it out. The imminence requirement has been widely criticized,3 most commonly in the context of ‘battered women’4 who have killed their abusive partners in non-confrontational situations, such as when the deceased was sleeping. These women would find it difficult to meet an imminence requirement, yet 1 The dictionary definition of the two terms suggests that there might be a minor distinction between them. Imminent is defined in the Oxford English Dictionary as: ‘Impending threateningly, hanging over one’s head; ready to befall or overtake one; close at hand in its incidence; coming on shortly.’ Immediate is defined as: ‘Pertaining to the time current or instant.’ This could indicate a more pressing threat (in terms of a stricter time limit) than an imminent one. Indeed, this was the conclusion reached by the Supreme Court of Kansas in State v Hundley 693 P 2d 475 (1985), which allowed an appeal on the basis that there was a distinction between the two terms (at 480). For the purposes of this chapter, however, such a distinction is not significant 2 enough for the two terms to be considered separately. Discussed in Chapter 4. 3 See Section 5.3 below and the references contained therein. 4 The use of the term ‘battered woman’ is reluctantly accepted as shorthand for any person who is subjected to serious mental or physical harm by their partner over a period of time. The term is used here because of its prominence and almost universal acceptance in existing literature and not because it is necessarily felt to be an appropriate term to describe such persons. It is not intended in any way to suggest that men never experience violence in relationships or that women are never violent towards their partners. See MR Mahoney, ‘Legal images of battered women: redefining the issue of separation’ (1991) 90 Michigan Law Review 1–94, at 25–6, who reviews various terms used to describe battered women.

88

Killing in Self-Defence

it has been suggested that they should still receive the benefit of the defence of self-defence. This chapter assesses the issue of imminence primarily in this context, as this is where the vast majority of critical comment has arisen.

5.2 A Formal Imminence Requirement Some jurisdictions take a strict approach to imminence of harm and permit self-defensive force only if harm was imminent, or reasonably appeared so to the accused. Scotland is one example. Thus in Owens v HM Advocate,5 the Appeal Court held that ‘self-defence is made out when it is established to the satisfaction of the jury that the panel believed that he was in imminent danger and that he held that belief on reasonable grounds’.6 A requirement that harm be imminent can also be found in the criminal codes and case law of some US states,7 although this is not the approach taken in the US Model Penal Code.8 This is not to say that under an imminence requirement an attack must actually be under way before the accused is permitted to use defensive force. Taking Scots law as an example, even Hume, who took a relatively strict stance on imminence,9 conceded that ‘it cannot be exacted of any one, to wait till the pistol is in the very act of being fired at him; or if the enemy have drawn, and be rushing towards him, he may meet him with his fire, before the point be at his breast’.10 The purpose of the imminence rule, it has been suggested, is to deny the defence of self-defence where there were alternative courses of action open to the accused other than killing the aggressor. If a threatened harm is 5

1946 JC 119. ibid 125. The requirement for the belief to be reasonable is discussed in Chapter 9. The imminence requirement has been reiterated in HM Advocate v Doherty 1954 JC 1, at 5; McCluskey v HM Advocate 1959 JC 39, at 43; Fenning v HM Advocate 1985 JC 75, at 78; Hendry v HM Advocate 1985 JC 105, per the Lord Justice-Clerk (Wheatley) at 111; and Friel v HM Advocate 1998 SLT 1327, per the Lord Justice-Clerk (Cullen) at 1328. In practice in Scotland, where women have killed their partners against a backdrop of sustained domestic violence, the Crown Office has often been willing to accept a plea of culpable homicide in circumstances that would otherwise amount to murder (because, for example, the imminence requirement is not met). See J Chalmers, CHW Gane, and F Leverick, ‘Partial defences to homicide in the law of Scotland: a report to the Law Commission for England and Wales’, in Partial Defences to Murder (Consultation Paper No 173 (Appendices), 2003) 151–83 and the cases cited at 181–3. 7 Ala Code § 13A-3-23(a); Ark Code § 5-2-606(a); Colo Rev Stat § 18-1-704(1); Ga Code Ann § 16-3-21(a); 720 ILCS § 5/7-1(a); Me Rev Stat Ann tit 17-A § 108(1); Mont Code Ann § 45-3-102; NY UCC Law § 35.15(1); Or Rev Stat § 161.209; Utah Code § 76-2-402(1); Whipple v Indiana 523 NE 2d 1363 (1988) (Supreme Court of Indiana), at 1367. 8 On which, see Section 5.4.3 below. 9 See D Hume, Commentaries on the Law of Scotland Respecting Crimes (4th edn, 1844) i, 228, and the discussion in J Chalmers and F Leverick, Criminal Defences and Pleas in Bar of Trial (2006), Chapter 3. 10 Hume, Commentaries on the Law of Scotland (4th edn) i, 229. See also J Beale, ‘Homicide in self-defense’ (1903) 3 Columbia Law Review 526–45, at 519. 6

Imminence of Harm

89

not imminent, it is assumed that the accused could have summoned help or protection or could otherwise have avoided the threat. Thus, as Rosen states: In self-defence, the concept of imminence has no significance independent of the notion of necessity. It is, in other words, a ‘translator’ of the underlying principle of necessity, not the principle itself. Society does not require that the evil avoided be an imminent evil because it believes that an imminent evil is the only type of evil that should be avoided [or] because an imminent threatened harm is necessarily worse than a non-imminent one. Rather, imminence is required because, and only because, of the fear that without imminence there is no assurance that the defensive action is necessary to avoid harm. If the harm is not imminent then surely the actor can take steps that will alleviate the necessity for responding with fatal force.11

As the introduction to this chapter noted, however, requiring imminence of harm does not only serve to deny the defence where the threat could have been avoided by means other than killing the aggressor. It also serves to restrict the availability of self-defence to situations where the threatened harm is likely actually to materialize. If a threat of harm is not imminent, it is more difficult to know with any degree of certainty whether the harm would have occurred and whether the accused could have saved her life simply by doing nothing (and thus preserving the life of the potential aggressor). By requiring imminence of harm, the law tries to ensure that the defensive response of the accused was genuinely necessary, thus avoiding the killing of an aggressor wherever this is reasonably possible.

5.3 Arguments against the Imminence Requirement: The Case of Battered Women The primary concern of this text has been with the right to life of all human beings, including that of potential aggressors.12 With this in mind, at first glance, the imminence requirement seems to serve an important purpose. It attempts to preserve human life (of aggressor and victim alike) where this is reasonably possible, by requiring the victim to seek alternative sources of help to deal with threats whenever there is time to do so and by requiring that defensive action is taken only against those threats that have a high probability of materializing. The imminence requirement has, however, come under considerable criticism from those who feel that it places undue emphasis on the time measurement between harm and defensive response at the expense of the underlying principle of necessity.13 For the critics, whereas imminence of harm is usually an accurate indication that the defensive response was necessary, it is not always so and defensive action can 11 R Rosen, ‘On self-defense, imminence and woman who kill their batterers’ (1993) 71 12 North Carolina Law Review 371–411, at 379. See in particular Chapter 3. 13 For a rare defence of the imminence requirement, see KF Ferzan, ‘Defending imminence: from battered women to Iraq’ (2004) 46 Arizona Law Review 213–62.

90

Killing in Self-Defence

sometimes be necessary even when harm is not imminent. Nowhere has this criticism been more pressing than in relation to battered women who have killed their violent partners. The volume of literature devoted to the issue of battered women and the imminence requirement is extensive. In the USA, much of it centres around State v Norman,14 where the defendant killed her abusive partner while he was sleeping and was denied the defence of self-defence because the harm she faced was not imminent.15 In the UK, while there has been no significant case law on the availability of self-defence as a defence for battered women,16 the issue has received much critical attention. It has, for example, been the subject of a debate between Joshua Dressler17 and Jeremy Horder,18 who hold opposing views on the issue.19

5.3.1 Some preliminary observations Before examining the argument in detail, two preliminary observations are worth making. First, the assumption is often made that ‘many’ or ‘most’ battered women who kill their partners do so in non-confrontational 14 State v Norman 378 SE 2d 8 (1989) (Supreme Court of North Carolina), overturning State v Norman 366 SE 2d 586 (1988) (Court of Appeals of North Carolina). An entire issue of the University of Pittsburgh Law Review was devoted to analysis of the case (see spring issue of vol 57, published in 1996). 15 For discussion of the imminence requirement in the USA, see S Appel ‘Beyond self-defense: the use of battered woman syndrome in duress defenses’ (1994) University of Illinois Law Review 955–80; E Ayyildiz, ‘When battered woman’s syndrome does not go far enough: the battered woman as vigilante’ (1995) 4 American University Journal of Gender and the Law 141–204; S Estrich, ‘Defending women’ (1990) 88 Michigan Law Review 1430–9; CP Ewing, Battered Women who Kill: Psychological Self-Defense as Legal Justification (1987); DL Faigman, ‘Discerning justice when battered women kill’ (1987) 39 Hastings Law Journal 207–27; H Maguigan, ‘Battered women and self-defence: myths and misconceptions in current reform proposals’ (1991) 140 University of Pennsylvania Law Review 379–486; JB Murdoch, ‘Is imminence really necessary? Reconciling traditional self-defense doctrine with the battered woman syndrome’ (2000) 20 North Illinois University Law Review 191–218; CJ Rosen, ‘The excuse of self-defense: correcting a historical accident on behalf of battered women who kill’ (1986) 36 American University Law Review 11–56; S Schulhofer, ‘The gender question in criminal law’, in EF Paul, FD Miller Jr, and J Paul (eds), Crime, Culpability and Remedy (1990) 105–37; KA Shad, ‘State v Norman: self-defense unavailable to battered women who kill passive abusers’ (1990) 68 North Carolina Law Review 1159–77. 16 In both England and Scotland, the case law on battered women who have killed their partners has centred around provocation and diminished responsibility. In England, see R v Ahluwalia [1992] 4 All ER 889 (diminished responsibility); R v Thornton (No 2) [1996] 1 WLR 1174 (provocation and diminished responsibility); R v Humphreys [1995] 4 All ER 1008 (provocation). In Scotland, see Galbraith v HM Advocate 2002 JC 1 (diminished responsibility). 17 J Dressler, ‘Battered women who kill their sleeping tormenters: reflections on maintaining respect for human life while killing moral monsters’, in S Shute and AP Simester (eds), Criminal Law Theory: Doctrines of the General Part (2002) 259–82. 18 J Horder, ‘Killing the passive abuser: a theoretical defence’, in S Shute and AP Simester (eds), Criminal Law Theory: Doctrines of the General Part (2002) 283–97. 19 See also: S Edwards, Sex and Gender in the Legal Process (1996); A McColgan, ‘In defence of battered women who kill’ (1993) 14 OJLS 508–29; C Wells, ‘Battered woman syndrome and defences to homicide: where now?’ (1994) 14 LS 266–76.

Imminence of Harm

91

situations20 and therefore fail to satisfy the imminence requirement.21 This assumption is questionable. Maguigan22 studied the population of US appellate cases where battered women had killed their partners in the period between 1902 and 1991 and concluded that only 20 per cent could be described as non-confrontational. Maguigan’s figures are consistent with those reported by others.23 This suggests that the imminence requirement does not pose such a substantial problem to battered women as might initially be thought. This does not necessarily mean that the issue is not worth addressing. It has also been suggested24 that Maguigan might underestimate the proportion of non-confrontational killings by focussing on reported appellate cases, which are not necessarily representative of the proportion of non-confrontational killing that occur in real life. A second preliminary point is that the imminence requirement is not the only possible barrier to battered women’s self-defence claims. Another potential difficulty for a battered woman is the proportionality requirement. It will be suggested later that killing in self-defence is only permissible where the accused faced a threat of death (or one of near equivalent seriousness) or where such a threat might reasonably be assumed.25 The level of violence faced by the majority of battered women is unlikely to meet this threshold. Indeed, the fact that a battered woman has been assaulted on many occasions in the past but has not been killed might suggest that she is unlikely to be killed by her partner in the future. The 1996 British Crime Survey of domestic violence found that, when victims were asked to recall the most recent incident of violence they had experienced, only 6 per cent involved being hit with a weapon and 9 per cent involved being choked, strangled or suffocated. Relatively minor violence was far more common: 65 per cent of incidents involved being pushed, shoved, or grabbed and 47 per cent of incidents involved being kicked, slapped, or hit with a fist.26 It may be, of course, that the tiny minority of women who kill their violent partners are those who experience the most serious violence. Even in State v 20 Maguigan, in ‘Battered women and self-defense’, defines a non-confrontational situation as ‘a killing that occurred when either (1) the man was asleep; (2) the man was awake, but the woman was the initial aggressor on the particular occasion; or (3) the woman hired or persuaded someone else to kill the man’ (at 392). 21 DL Faigman, ‘The battered woman syndrome and self-defense: a legal and empirical dissent’ (1986) 72 Virginia Law Review 619–47, at 621; Rosen, ‘The excuse of self-defense’, at 43; Edwards, Sex and Gender in the Legal Process, at 407; C Ewing, ‘Psychological self-defense: a proposed justification for battered women who kill’ (1990) 14 Law and Human Behavior 22 579–94, at 581. Maguigan, ‘Battered women and self-defense’, at 392–408. 23 See the studies cited by Maguigan herself at ibid 398–9 and also VF Nourse, ‘Self-defense and subjectivity’ (2001) 68 University of Chicago Law Review 1235–308, at 1253. 24 CA Terrance, K Matheson, and NP Spanos, ‘Effects of judicial instructions and case characteristics in a mock jury trial of battered women who kill’ (2000) 24 Law and Human 25 Behavior 207–29, at 208. See Chapter 8. 26 C Mirrlees-Black, Domestic Violence: Findings from a New British Crime Survey SelfCompletion Questionnaire (1999), at 76. The figures do not total 100 per cent because each single incident could involve more than one type of violence.

92

Killing in Self-Defence

Norman, however, which is commonly accepted as one of the worst reported cases of sustained domestic violence, some doubt can be raised as to whether the harm the defendant suffered could be described as meeting this standard. JT Norman subjected Judy Norman to assaults that included slapping, punching and kicking her, striking her with various objects and throwing glasses, ashtrays and beer bottles at her. He also put out cigarettes on her skin, threw hot coffee at her and broke a glass on her face.27 All of these are clearly extremely distressing incidents but the only one that suggests a realistic threat of death (or of a very serious bodily harm that comes close to this standard) is the breaking of the glass. Despite this, the potential proportionality problem is rarely referred to in the literature on battered women and self-defence, with the primary focus being on the difficulties posed by the imminence requirement and by the reasonableness requirement.28

5.3.2 Battered women and the imminence requirement The imminence requirement has been criticized because it fails to deal with cases where harm was not imminent, but where it was highly likely to occur and where the accused did not have any reasonable alternative course of action open to her in order to avoid harm other than to kill the source of the threat. This, it is said, precisely describes the situation faced by battered women such as Judy Norman who have killed their violent partners in non-confrontational situations.29 While any harm posed by a sleeping person cannot be said to be imminent, nonetheless, it is argued that the battered woman has no reasonable alternative but to kill her abuser if she is to avoid future harm. This is because the pattern of previous abuse suffered by a battered woman suggests that the harm threatened by her partner is highly likely to materialize and the usual alternatives to killing—such as seeking assistance or escaping from the situation—are, for various reasons,30 not open to her. The issue is sometimes presented as one of sexism in the law. That is, in requiring imminence of harm, the law discriminates against female accused, because it confines the defence of self-defence to typically masculine situations: one-off confrontational incidents between two people of

27

State v Norman 378 SE 2d 8 (1989), at 10. For four notable exceptions, see Ewing, ‘Psychological self-defense’; JM Cohen, ‘Regimes of private tyranny: what do they mean to morality and for the criminal law?’ (1996) 57 University of Pittsburgh Law Review 757–808; Faigman, ‘Discerning justice’; and Dressler, ‘Battered women’. 29 State v Norman 378 SE 2d 8 (1989). For similar cases involving battered women, see Commonwealth v Sands 553 SE 2d 733 (2001) (Supreme Court of Virginia); Lane v State 957 SW 2d 584 (1997) (Court of Appeals of Texas). The issue does not arise solely in relation to battered women. See Whipple v State 523 NE 2d 1363 (1988) (Supreme Court of Indiana) (male defendant killed his two abusive parents while they slept); State v Schroeder 261 NW 2d 759 (1978) (Supreme Court of Nebraska) (defendant was a male prisoner who killed his abusive 30 Discussed in Section 5.6 below. cellmate while he slept). 28

Imminence of Harm

93

approximately equal strength.31 Against this, Estrich has argued that the rules of self-defence: . . . do not exist, at least in the first instance, to torment battered wives. Unlike, say, the resistance requirement in rape law, they are not born from a historic distrust of women, or a desire to keep them powerless. They exist, quite simply, to preserve human life where deadly force is not reasonably necessary.32

Nonetheless, because of the imminence requirement, battered women who have killed their passive partners, even where the harm was highly likely to occur and no reasonable alternative method of preventing harm was available, are unable to plead self-defence successfully. With this in mind, it is worth examining some alternatives to the imminence requirement.

5.4 Alternatives to Imminence There are essentially three alternatives to consider: abolishing both the imminence and the necessity requirement; interpreting imminence widely; or replacing the imminence requirement with one for necessary force.

5.4.1 Abolish the imminence requirement and the necessity requirement One option is that both the imminence requirement and any requirement for necessity of defensive response be abandoned, thus ensuring that any woman who did kill her violent partner while he was sleeping would be able to claim the defence of self-defence (assuming she met the proportionality requirement). There is no common law jurisdiction that has taken such an extreme approach. Even those jurisdictions that have no formal imminence rule have tended to retain an underlying requirement of necessity.33 This view has occasionally found currency in the academic literature. Ayyildiz, for example, has claimed that the battered woman is morally justified in killing her partner at any time, regardless of whether the killing was necessary. She comments at one point: The battered woman is by definition a victim, one who has not received justice, one who has not seen her batterer punished for the abuse he has heaped upon her. Thus, by killing her batterer, the battered woman becomes a spontaneous vigilante—she apprehends a criminal that the law has failed to bring to justice and metes out the punishment he richly deserves.34

This type of sentiment is not uncommon. Cohen comments that the death of an abusive partner ‘may be a misfortune, but it has no effective cost’.35 31 32 34 35

Wells, ‘Battered woman syndrome’, at 272. 33 Estrich, ‘Defending women’, at 1431. See Section 5.4.3 below. Ayyildiz, ‘When battered woman’s syndrome does not go far enough’, at 147–8. ‘Regimes of private tyranny’, at 790.

94

Killing in Self-Defence

Ewing states that killing to prevent domestic violence is: ‘in most cases, entirely rational, reasonable, and no less necessary and socially desirable than a killing done in traditional self-defence’.36 Sebok asks ‘whether the law should promote the life of the batterer above all interests other than the life of his abused spouse’37 and questions if ‘we must choose between the biological life of the batterer and the autonomy of an innocent woman, why should we privilege the former?’38 A similar line is taken by the dissenting judges in State v Norman39 and in Jahnke v State,40 a case where a teenage boy killed his abusive father, but was denied the defence of self-defence because he did not satisfy the imminence requirement.41 The view that the necessity requirement should be abandoned is impossible to accept, given the fundamental premise of this text: that all human beings have a right to life, regardless of whether they are ‘good’ or ‘bad’ people. Wherever it is reasonably possible, human life should be preserved, even if one of those human lives belongs to someone who has repeatedly assaulted his partner.42 This is not in any way to suggest that a perpetrator of domestic violence is not to be condemned and punished in some way. Rather, it is to suggest that his punishment should be proportionate to his conduct and should follow the requirements of due process: a criminal trial by an impartial legal tribunal. Allowing victims of domestic violence to kill their partners when there was a reasonable alternative course of action that would have preserved the life of both the victim and the abuser fails to respect the abuser’s right to life. As the Court of Appeal of California put it in People v Aris,43 a case in which a battered woman shot and killed her sleeping husband: While we recognize that applying [an imminence of harm] rule in cases such as this one is difficult because of our sympathy for the plight of a battered woman and disgust for the batterer, it is fundamental to our concept of law that there be no discrimination between sinner and saint on moral grounds. Any less exacting definition of imminence fails to protect every person’s right to live.44

5.4.2 Interpret imminence widely A second option is that imminence is interpreted in a way that brings the battered woman who has killed a passive partner within its scope. There are various ways in which this might be achieved. It might, for example, be 36

‘Psychological self-defense’, at 593, emphasis added. AJ Sebok, ‘Does an objective theory of self defense demand too much?’ (1996) 57 38 University of Pittsburgh Law Review 725–55, at 753. ibid 754. 39 State v Norman 378 SE 2d 8 (1989), per Martin J, at 21. 40 41 682 P 2d 991 (1984) (Supreme Court of Wyoming). Rose J, at 1011. 42 M Mihajlovich, ‘Does plight make right: the battered woman syndrome, expert testimony and the law of self-defense’ (1987) 62 Indiana Law Journal 1253–1282, at 1281; JC Smith, Justification and Excuse in the Criminal Law (1989), at 117. 43 44 215 Cal App 3d 1178 (1989). ibid 1189. 37

Imminence of Harm

95

presumed that the threat of harm faced by a battered women is always imminent. Although it has not found favour in the case law of any common law jurisdiction, it is the preferred approach of Diamond45 and seems to have been favoured in some dissenting opinions in US cases. Thus, in State v Stewart,46 Herd J argues that ‘the nature of the cumulative terror under which a battered woman exists and . . . a batterer’s threats and brutality can make life-threatening danger imminent to a victim of that brutality even though, at the moment, the batterer is passive’.47 Others have argued for the reinterpretation of the term imminence such that it captures a threat of future harm. Ripstein, for example, asserts that we need a ‘different understanding of imminence’48 and that while one ‘might suppose that imminence is always measured with a stopwatch . . . the requirement of imminence is itself an expression of an underlying requirement of unavoidability’49 because imminent just means ‘before any reasonable opportunity to get out of harm’s way’.50 The House of Lords seems to have taken this approach, albeit in a different context, in the Attorney General for Northern Ireland’s Reference (No 1 of 1975).51 The case stemmed from the acquittal on a charge of murder of a soldier who shot a man he thought to be a member of a terrorist organization. The incident happened during a patrol of farmland and the patrol had been briefed to expect ambush. The deceased was an innocent member of the public who had no involvement with terrorist activity. The soldier shouted ‘halt’, to which the deceased responded by running away. The soldier shot him in the back as he did so. The Reference was concerned with whether, on the facts, there was any possibility in law that the soldier could have been justified. The House of Lords answered in the affirmative.52 The leading judgment was delivered by Lord Diplock, who concluded that: In the facts that are to be assumed for the purposes of the Reference, there is material upon which a jury might take the view that the accused had reasonable grounds for apprehension of imminent danger to himself and other members of the patrol if the deceased were allowed to get away and join armed fellow-members of the Provisional IRA who might be lurking in the neighbourhood.53 45 GA Diamond, ‘To have but not to hold: can “resistance against kidnapping” justify lethal self-defense against incapacitated batterers?’ (2002) 102 Columbia Law Review 729–73. 46 763 P 2d 572 (1988) (Supreme Court of Kansas). See also Martin J (dissenting) in State v 47 Norman 378 SE 2d 8 (1989), at 19. ibid 585. 48 A Ripstein, ‘Self-defense and equal protection’ (1996) 57 University of Pittsburgh Law 49 Review 685–724, at 698. ibid 699. 50 ibid 703. See also Mahoney, ‘Legal images of battered women’, at 87; McColgan, ‘In defence of battered women’, at 518; S Wallace, ‘Beyond imminence: evolving international law and battered women’s right to self-defense’ (2004) 71 University of Chicago Law Review 51 1749–81, at 1770. [1977] AC 105. 52 The House of Lords seems to have accepted that the law of self-defence/prevention of crime contained an imminence requirement. This is not now the case in English law (and thus is probably not the case in Northern Irish law either). See Section 5.4.3 below. 53 ibid 138, emphasis added.

96

Killing in Self-Defence

With respect, it could hardly be said that the deceased threatened ‘imminent’ harm to anyone. He was running away when he was shot and there was no indication whatsoever that there were any ‘fellow-members of the Provisional IRA’ in the area. It is precisely herein that the difficulty with wide interpretation lies. It is quite simply inaccurate. In the context of battered women, when an abuser is sleeping or otherwise passive, any harm he threatens is not imminent according to the ordinary meaning of the word. An imminent harm is one that is, or gives the appearance of being, about to occur. The term does not cover a threat of harm that may or may not be realized some time in the more distant future. Against this, it might be said that imminence need only be given a legal meaning, and that legal meaning could be at odds with its ordinary usage. However, to give imminence a legal meaning so far removed from the ordinary understanding of the term is unwise as this would seriously affect the accessibility of the law in terms of its ability to communicate its provisions to the general public. It may be that those commentators who do propose such a move are simply trying to find a solution to the problem posed by battered women within the existing confines of the law, rather than through substantive change.

5.4.3 A necessity test A third option is to dispense with the imminence requirement and require instead simply that force used in self-defence was ‘necessary’. This is the approach taken in English law where imminence of harm is one factor for the jury to take into account in deciding whether or not defensive action was reasonably necessary.54 Indeed, in Re: A (Children),55 a case involving the separation of conjoined twins, it was accepted by at least one member of the Court of Appeal that the action taken by the doctors could be covered by the defence of self-defence, despite the fact that the harm threatened to the stronger twin was not imminent.56 It is also the approach taken by the Supreme Court of Canada, where, in R v Pétel,57 Lamer CJ approved the Supreme Court’s earlier decision in R v Lavallee,58 stating that: There is . . . no formal requirement that the danger be imminent. Imminence is only one of the factors which the jury should weigh in determining whether the accused had a reasonable apprehension of danger and a reasonable belief that she could not extricate herself otherwise than by killing the attacker.59 54 Shaw v R [2001] UKPC 26, per Lord Bingham at [19]; R v Palmer [1971] AC 814, per 55 Lord Morris at 831. Re: A (Children) [2001] 2 WLR 480. 56 Re: A (Children), per Ward LJ at 536. For discussion of the case, see Chapter 1. Cf DPP v Bayer [2003] EWHC Admin 2567, where Brooke LJ referred to the need for an imminent attack if the common law defence of self-defence or the defence of others is to be made out (at [21]). 57 58 59 [1994] 1 SCR 3. [1990] 1 SCR 852. ibid [24].

Imminence of Harm

97

It seemed at one time that this interpretation applied only to cases involving battered women. In R v Charlebois,60 the Supreme Court refused to relax the imminence requirement in a case where the appellant had killed another man, with whom he did not have a sexual relationship, but who had exhibited violent and controlling behaviour.61 In R v Cinous,62 however, the Supreme Court ignored Charlebois and held that the approach taken in Pétel represented the law in all cases of self-defence.63 A slight variation on requiring force to be necessary is to require it to be ‘immediately necessary’, the approach taken in some US states.64 The US Model Penal Code contains a rather puzzling variation on this whereby selfdefensive force must be ‘immediately necessary . . . on the present occasion’.65 An absence of detailed discussion in the Commentary to the Code means that it is not clear exactly how this standard differs, if at all, from a standard of imminence of harm. The ‘immediately necessary’ formulation is favoured by Robinson, who believes that the imminence requirement denies valid claims of self-defence by those whose action was necessary but was not taken in response to an imminent harm. He illustrates his argument with the following scenario: Suppose A kidnaps and confines D with the announced intention of killing him one week later. D has an opportunity to kill A and escape each morning as A brings him his daily ration. Taken literally, the imminent (sic) requirement would prevent D from using deadly force in self-defence until A is standing over him with a knife, but that outcome seems inappropriate.66

For Robinson, what matters is not the imminence of any harm, but the immediate necessity of making a response. He continues: If the concern of the limitation [of imminence] is to include threats of harm that are too remote to require a response, the problem is adequately handled by requiring simply that the response be ‘necessary’. The proper inquiry is not the immediacy of the threat but the immediacy of the response necessary in defence. If a threatened harm is such that it cannot be avoided if the intended victim waits until the last moment, the principle of self-defence must permit him to act earlier—as early as is required to defend himself effectively.67 60

[2000] 2 SCR 674. The Court distinguished Charlebois from Lavallee, where the appellant was a battered 62 woman. [2002] 2 SCR 3. 63 ibid [40]. Indeed, in R v Ruzic [2001] 1 SCR 687, the Canadian Supreme Court ruled that the imminence requirement is unconstitutional in relation to the related defence of duress. 64 Tex Penal Code § 9.31(a); Ariz Rev Stat Ann § 13-404(A). 65 § 3.04(1), emphasis added. The Model Penal Code provision has been adopted by a number of states: Del Code Ann tit 11 § 464(a); Neb Rev Stat § 28-1409(1); 18 Pa Cons Stat § 505(a). See also 9 Guam Code Ann § 7.84. 66 PH Robinson, Criminal Law Defenses (1984), at § 131(c) (emphasis in original). 67 Robinson, Criminal Law Defenses, at § 131(c). See also PDW Heberling, ‘Justification: the impact of the Model Penal Code on statutory reform’ (1975) 75 Columbia Law Review 914–62, at 932; G Williams, Textbook of Criminal Law (1978). Williams gives the example of a police officer who shoots a terrorist (who has threatened to kill his hostage) when he momentarily 61

98

Killing in Self-Defence

Robinson’s hostage example has been utilized in the context of battered women, both in case law and in academic literature. In R v Lavallee,68 explaining why an imminence requirement might not be appropriate for a battered woman who has killed in self-defence, Wilson J stated that: The situation of the battered woman . . . strikes me as somewhat analogous to that of a hostage. If the captor tells her that he will kill her in three days time, is it potentially reasonable for her to seize an opportunity presented on the first day to kill her captor or must she wait until he makes the attempt on the third day?69

Appel compares the battered woman to ‘a hostage who is told she would be killed the next day and then strangles a sleeping guard in an effort to escape’.70 The basis of these claims is that the battered woman is no more able to escape from an abusive relationship than a hostage who is held physically captive by a kidnapper. More recently, this analogy has been used by Horder, who asserts that there are: . . . important similarities between hostage situations and the relationships of domination, unpredictability, and violence in which battered women find themselves trapped. Many such women learn that if they attempt to leave, they will be followed and forced to return, to face even greater hostility and more serious violence. In any event, a concern for the welfare of children of the family will often make these women the equivalent of hostages.71

Horder is careful, however, to avoid any suggestion that he dismisses the right to life of the abuser. He continues: None of this evidence, of course, can raise even a prima facie case that the abusers in question ‘forfeit’ their right to life. What it does do, however, is give some substance to the view that, in certain exceptional cases, women whose partners are increasingly violent and possessive abusers are in a similar situation to that faced by hostages facing an unpredictable but seemingly escalating threat. They may (reasonably) take the view that there was no fair opportunity to do other than they did, in using lethal appears at a window. The police officer may not be preventing an imminent harm but, for Williams, his actions may be justified if it was reasonable to believe that it was ‘necessary to take the opportunity of killing or disabling the terrorist when it presented itself’ (at 450–1). Although he does not acknowledge it as such, Williams’s example is identical in nature to events that took place on 28 January 1978 at the Prison of Clairvaux in France. Here, in a foiled escape attempt, two prisoners, one of whom was armed, took three people hostage in a prison cell and threatened to kill them if various demands were not met within a set time period. Army snipers trained their guns on the cell. Both captors appeared at the window some considerable time before the expiry of the deadline and they were shot dead. It was accepted that the snipers acted in legitimate defence of the hostages. See MA Hay, The Criminal Law of Private Defence in England, Scotland and France (1990), at 29 n 125. 68 [1990] 1 SCR 852 (Supreme Court of Canada). 69 ibid 889. This passage is cited with approval by Mildren J of the Court of Criminal Appeal of the Supreme Court of the Northern Territory in R v Secretary (1996) 131 FLR 126, at 134; the Court of Criminal Appeals of Oklahoma in Bechtel v State 840 P 2d 1 (1992), at 12; and the Supreme Court of South Carolina in Robinson v State 417 SE 2d 88 (1992), at 91. 70 Appel, ‘Beyond self-defense’, at 975. 71 Horder, ‘Killing the passive abuser’, at 291 (Horder’s footnotes omitted).

Imminence of Harm

99

force against the abuser or hostage-taker while he was in a passive state, in order to escape the predicament.72

Two questions are raised. The first is whether any of these formulations is an appropriate standard for the law of self-defence. The second is whether, even if they were to be adopted, a battered woman who killed her partner in a non-confrontational situation could meet a test of necessity.

5.5 Should the Imminence Requirement be Replaced by a Necessity Test? A number of arguments can be made against replacing an imminence requirement with a general necessity test. Two of these can be disposed of relatively easily because they do not address the central issue. The first is that a necessity test should be rejected because not only would it bring battered women into its ambit, but it would also cover all sorts of other, less deserving, accused. Veinsreideris, for example, argues for the retention of the imminence requirement on the basis that removing it would open up the undesirable possibility of a successful self-defence claim by anyone who believes that the authorities do not offer adequate protection from violence.73 Veinsreideris suggests that claims might be made by prisoners who have killed other inmates in ‘self-protection’, pointing to a number of cases where a threat of future harm has been made by one cell-mate to another and the accused has killed in order to avoid that threat being realized.74 He suggests that it would be relatively easy for the accused to establish that he had no reasonable alternative to killing (he could not escape from the situation, prison guards do not provide adequate protection, and so on). The fear has also been raised that battered women (or indeed others) who hire contract killers to kill their abusive partners could be brought within the law of self-defence if the imminence requirement is abandoned.75 In response it might be said that this is not, in itself, an argument for retaining the imminence requirement. If it is thought appropriate to replace 72

ibid 291–2. ME Veinsreideris, ‘The prospective effects of modifying existing law to accommodate pre-emptive self-defense by battered women’ (2000) 149 University of Pennsylvania Law Review 613–44, at 625. See also Rosen, ‘The excuse of self-defense’, at 54; BC Zipursky, ‘Self-defense, domination and the social contract’ (1996) 57 University of Pittsburgh Law Review 579–613, at 588. 74 State v Schroeder 261 NW 2d 759 (1978) (Supreme Court of Nebraska); State v Wiggins 808 P 2d 1383 (1991) (Supreme Court of Kansas). In both of these cases the defence was rejected because the imminence requirement was not met. 75 Dressler, ‘Battered women’, at 261. In the USA, self-defence has been claimed by defendants who have hired contract killers to kill their abusive partners although to my knowledge these claims have all failed. See eg State v Martin 666 SW 2d 895 (1984) (Court of Appeals of Missouri). 73

100

Killing in Self-Defence

the imminence requirement with a general necessity test, then it would be absurd and discriminatory to restrict such a change to battered women. If such a change is thought appropriate, then anyone who can genuinely satisfy the necessity test should be allowed to claim the defence. The second, relatively easily dismissed, argument is that removing the imminence requirement would allow juries too much leeway in deciding what is and what is not necessary action in self-defence.76 If a jury is not instructed that only action taken against an imminent harm is permitted, they may well acquit sympathetic but undeserving defendants who killed in non-confrontational situations. Indeed, this is not solely a question relating to juries but might also be a concern where the trier of fact is a judge. This is illustrated by the Northern Irish case, Kelly v Ministry of Defence.77 Here, a 17-year-old joy-rider was shot and killed by soldiers in Northern Ireland in the mistaken belief that he and the other three occupants of the car he was driving were escaping terrorists. The deceased was running away from the soldiers at the time of the incident. Although the case was a civil action brought by the father of the deceased, the case turned on the interpretation of s 3 of the Criminal Law (Northern Ireland) Act 1967, which permitted the use of force in the prevention of crime and would have operated as a defence to both civil claims and criminal charges. The Court of Appeal in Northern Ireland held that the shooting was justified on this basis. They approved the statement of Carswell J, who heard the case at first instance, that: ‘It seems to me reasonable to regard a man who makes such a strong bid for freedom in the circumstances of the case as one who is an active and dangerous terrorist intent on evading capture, who is highly likely to commit terrorist crimes if not apprehended’.78 Rosen flags up the issue of jury decision-making,79 but concludes that ‘[t]oday’s jurors are, hopefully, in fact as well as principle, objective and disinterested fact finders’.80 This optimistic conclusion can be questioned.81 In any case, in any jurisdiction where a jury does not have to justify its decision in law, a jury could do this anyway, regardless of whether they 76

A point made in Chapter 4, in relation to retaining an independent retreat requirement. [1989] NI 341. ibid 349. For criticism of the decision, see JC Smith, ‘The right to life and the right to kill in law enforcement’ (1994) 144 NLJ 354–6, at 356; AP Simester and GR Sullivan, Criminal Law: Theory and Doctrine (2nd edn, 2004), at 627. Kelly subsequently went to the European Court of Human Rights, but the case was declared inadmissible and the imminence issue was not addressed: Kelly v United Kingdom (1993) 16 EHRR CD20. Somewhat bafflingly, the Commission proceeded on the basis that the killing was justified in effecting an arrest. 79 80 Rosen, ‘On self-defense’, at 404. ibid 405. 81 Clare Connelly’s analysis of Scottish cases in which battered women have killed abusive partners suggests that the more ‘sympathetic’ the accused appears, the more likely the jury will acquit her, regardless of whether she meets the requirements of any defence. This was the conclusion reached by Connelly in a seminar given as part of the University of Aberdeen School of Law Research Seminar Series, 7 December 2000. 77 78

Imminence of Harm

101

received an instruction on the imminence requirement.82 Concern over granting too much decision-making discretion to juries is an issue that lies beyond the scope of this book83 but in itself is not a decisive reason to retain the imminence requirement. If this is the conclusion to be drawn, it should be based primarily on the moral appropriateness of attributing criminal liability to a particular class of accused. The moral reasons why we should think very carefully before replacing an imminence requirement with a necessity test might be outlined as follows. The premise of this text is that, wherever possible, the right to life of all people should be respected. The major difficulty with any relaxation of the imminence requirement is that the further into the future a threat of harm stretches, the lower the probability that such a threat is actually going to materialize, and the higher the likelihood that killing the (potential) aggressor was not necessary at all.84 When a threat of harm is imminent, a point is reached where it is extremely unlikely that it will be possible to save the lives of both of the parties involved. A choice must be made and, because the aggressor threatens the life of the victim, the victim’s life is preferred.85 When a threat of harm is not imminent, there is a greater probability that both lives can still be saved, either because the anticipated harm will not, in fact, occur,86 or because an alternative course of action taken by the accused (other than killing) could prevent the threat from materializing. On the other hand, it is recognized here that imminence of harm is merely a signifier of the underlying requirement that any action taken in self-defence be necessary. The difficulty in abandoning the imminence requirement, though, is that imminence of harm is a very good signifier of necessity of action. If harm is not imminent, it can never be known for sure whether the aggressor would have carried out his threat of future harm or whether alternative courses of action would have prevented the attack.87 In essence, the decision that has to be made is one of probability of harm—at what 82 Kalven and Zeisel give examples of cases from the USA where defendants who did not meet an imminence requirement have been acquitted, despite the jury having received an imminence instruction. See H Kalven Jr and H Zeisel, The American Jury (1966), at 225. 83 For a summary of research into jury decision-making, see P Darbyshire, A Maughan, and A Stewart, What Can the English Legal System Learn from Jury Research Published up to 2001? (2001); P Darbyshire, ‘What can we learn from published jury research? Findings for the Criminal Courts Review 2001’ [2001] Crim LR 970–9. 84 Ferzan, ‘Defending imminence’, at 252. 85 See Chapter 3, where it was suggested that, at the point where it is no longer reasonably possible to save both lives, the aggressor, by virtue of becoming an unjust immediate threat to the victim, forfeits her own right to life. 86 In Robinson’s hostage case, for example, who is to say that the hostage-taker might not change his mind about killing his hostage over the course of the week? Against this, it might be argued that, in making a threat to kill, the hostage-taker must bear the risk of being killed in turn by his hostage. This comes perilously close, however, to arguing that the hostage-taker’s life is not worthy of consideration because of his conduct and such a view has been firmly rejected earlier. 87 Of course, even when harm appears imminent, it can never be known for sure whether it would have occurred. No future event can be predicted with certainty. The aggressor pointing

102

Killing in Self-Defence

probability level of harm do we draw the line and allow/disallow claims of self-defence?88 It is suggested here that the probability level should be set at, or very close to, inevitability: exactly the sort of level signified by an imminence requirement. Only if this ‘inevitability of harm’ probability level is met should the accused who has killed to prevent a non-imminent harm be permitted to claim the defence of self-defence.89 Meeting the inevitability of harm test would require the accused to satisfy two conditions.90 The first condition is that the harm threatened would inevitably have occurred (without her defensive action). At first glance, it might seem that this test is far too strict and that it weights the balance of concern for the two lives involved too far in favour of the aggressor’s life. As in other areas of the law of self-defence, however, allowance would be made for reasonable mistake. In other words, if there were reasonable grounds for concluding that harm was inevitable, then the defence should be allowed.91 The second condition is that there was no reasonable prospect of preventing the harm by taking an alternative course of action. This test is simply the equivalent of the strong retreat rule. In Chapter 4, it was argued that the accused should be required to take any reasonable opportunity to retreat and seek assistance before benefiting from an acquittal on the basis of selfdefence. This respects the right to life of both the aggressor and the victim and does not allow lesser values, such as personal honour or the desire for revenge, to outweigh the value of the aggressor’s life. When harm is not imminent, I can see no reason why an equivalent rule should not apply. The only difference between the imminent harm scenario and the non-imminent harm scenario is that the period of time in which the victim is able to seek an alternative course of action is longer when the harm she faces is not imminent. The same consideration of respect for the right to life of the aggressor still applies.

5.6 Would a Battered Woman Meet an Inevitability of Harm Test? It has been claimed that at least some women who kill their abusive partners at a point where harm was not imminent would, nonetheless, meet a test of the loaded gun might have a sudden change of mind; the attacker with the uplifted knife might fall and disable herself. 88 M Gorr, ‘Private defence’ (1990) 9 Law and Philosophy 241–68, at 259–60. 89 I am not alone in reaching this conclusion: inevitability of harm is the formulation proposed by the New Zealand Law Commission, in Some Criminal Defences with Particular Reference to Battered Defendants (2001), at paras 30–2. 90 Or, more accurately, it would be for the prosecution to prove beyond reasonable doubt that these two conditions were not satisfied. 91 Provided that the other requirements of the defence are made out.

Imminence of Harm

103

inevitability of harm.92 But is this a valid claim?93 This depends on (a) how likely it is that the abuser would actually carry out his threat of harm and (b) how likely it is that any reasonable alternative course of action taken would have prevented the threatened harm. On (a), the battered woman would be in a stronger position than most to meet the inevitability of harm test. The past history of the violent relationship is likely to be a rich source of evidence that the batterer has carried out his threats in the past and is likely to do so in the future.94 In this, the battered woman arguably has a stronger claim of inevitability of harm than even Robinson’s hypothetical hostage, where there is less likely to be past evidence of harm and where all sorts of other variables might mitigate against the hostage-taker carrying out her threat.95 On (b), the battered woman would be required to show that no reasonable opportunity of preventing the threatened harm existed, other than killing her abuser. Two strands of argument have been used to show that this test could be met. The first is that the battered woman wrongly (but unreasonably) believes that harm is inevitable because her psychological condition as a battered woman leaves her unable to perceive possible alternative courses of action, even though they are objectively available. This, it is suggested, is because battered women suffer from a condition termed ‘learned helplessness’. The claim stems from the work of Lenore Walker,96 a psychologist who studied a group of battered women, a small minority of whom had killed their partners. Walker claims that women who kill their violent partners may be suffering from ‘battered woman syndrome’ (BWS), a psychological condition that includes, as one of its symptoms, a degree of learned helplessness. Walker adopted the expression learned helplessness from work undertaken with dogs by Martin Seligman.97 In Seligman’s experiments, dogs in 92

New Zealand Law Commission, at para 30. It is assumed here for the sake of argument that the proportionality requirement is met. As has already been noted (see Section 5.3.1), in reality, battered women will not always meet this requirement. 94 In this, I disagree with Dressler. In ‘Battered women’, Dressler assumes that the likelihood of the batterer carrying out his threat is no different to that of any individual who threatens violence. He states that ‘even the most optimistic behavioural scientists now seem to believe that the accuracy of predictions of dangerousness whether made clinically or statistically is probably no better than one valid assessment out of two’ (at 275 n 84). This is not even the relevant type of claim in this context. Assessing the likelihood that an individual who has been violent in the past will repeat his violent behaviour in the future is not the same as assessing the likelihood that an individual will carry out a specific threat of violence. More importantly, Dressler does not take account of the history of past violence between abuser and battered woman which is likely to provide evidence that previous threats of violence had, in fact, materialized. 95 The hostage might be rescued, the hostage-taker might realize that the hostage is worth more to her alive than dead, and so on. 96 LE Walker, The Battered Woman (1979); LE Walker, The Battered Woman Syndrome (1984). 97 MEP Seligman and SF Maier, ‘Failure to escape traumatic shock’ (1967) 74 Journal of Experimental Psychology 1–9; MEP Seligman, SF Maier, and J Geer, ‘Alleviation of learned helplessness in the dog’ (1968) 73 Journal of Abnormal Psychology 256–62. 93

104

Killing in Self-Defence

cages were given electric shocks at random intervals. In the initial phase of the experiment, they were physically unable to escape. Once Seligman lifted these physical restrictions, however, he found that the dogs failed to perceive the escape routes that had opened up and chose instead to remain in the cage and experience the electric shocks. Walker suggests that sufferers of BWS remain in abusive relationships for the same reason: they become so resigned to violence that they fail to perceive escape routes, even when they are objectively available. Her argument is accepted uncritically by a number of those who argue that the situation of a battered woman who kills her partner can be likened to Robinson’s hostage situation.98 Walker’s research has, however, been subject to extensive criticism. Indeed, two of her fiercest critics, Faigman and Wright, suggest that ‘[Walker’s] 1984 book would provide an excellent case study for psychology students on how not to conduct psychological research’.99 In her study, learned helplessness was measured by scoring the women subjects on being ‘fearful’, ‘anxious’, ‘depressed’, and a number of other measures. There is no theoretical basis for these concepts being measures of learned helplessness. She also compared scores between women presently in abusive relationships with those who were no longer in them. The scores were only marginally different. No test for statistical significance was undertaken. Perhaps most fundamentally, Walker had no control group: she worked only with women who were in abusive relationships, so there is no way of telling whether other women who did not suffer from BWS would have scored highly on these measures. It might also be pointed out that the action taken by battered women who kill their partners is entirely at odds with the learned helplessness concept: on the contrary, these women have asserted their control in the most dramatic way imaginable.100 Even if it is accepted that learned helplessness exists in those battered women who kill passive partners, it does not necessarily assist claims that their action was necessary self-defence. The defence of self-defence takes two forms: that of justification (where the accused correctly believes that her action is necessary) and that of excuse (where she is mistaken in her belief). The form claimed by a sufferer of learned helplessness is the excuse form. It will be argued later that only a reasonable mistake in perception should ground an acquittal on the basis of self-defence.101 But the battered woman’s 98

For example, Ewing, in ‘Psychological self-defense’, at 270. DL Faigman and AJ Wright, ‘The battered woman syndrome in the age of science’ (1987) 39 Arizona Law Review 67–115, at 68. The following criticisms summarize those made in Faigman and Wright’s article, as well as those in Faigman, ‘The battered woman syndrome’; AM Coughlin, ‘Excusing women’ (1994) 82 California Law Review 1–93; and Schulhofer, ‘The gender question in criminal law’, at 120–2. This criticism has led some to abandon the terminology of battered woman’s syndrome entirely (New Zealand Law Commission, at para 13). 100 Indeed, Seligman himself later published a paper arguing that his work with dogs was not helpful in understanding learned helplessness in humans (LY Abramson, ME Seligman, and JD Teasdale, ‘Learned helplessness in humans: critique and reformulation’ (1978) 87 Journal of 101 Abnormal Psychology 49–77). See Chapter 9. 99

Imminence of Harm

105

mistake is not a reasonable one. On the contrary, the reasonable person would not perceive her action as necessary. The reasonable person, who was not suffering from BWS, would accurately perceive that there was some reasonable alternative to killing the abuser in order to avoid the harm. Against this, however, it might be said that the battered woman should be judged not from the perspective of the reasonable person, but from the perspective of the ‘reasonable battered woman’.102 The debate over the extent to which the hypothetical reasonable person should be endowed with the characteristics of the accused is one that has long troubled criminal law103 and it is a debate that will not be entered into in depth here. Suffice to say that the view is taken that characteristics relating to the mental condition of the accused should not enter into determinations of reasonableness. In my view, the defence claimed by the accused whose perceptions were distorted because she was suffering from BWS does not fall into the territory of reasonable self-defence but instead has the gist of a claim of diminished responsibility:104 due to an abnormality of mind, the accused’s ability to determine and control her actions was impaired.105 The second strand of argument is that some battered women genuinely do not have any reasonable alternative to killing their partner in order to escape his violence and that, for this reason, the threat of harm they face, while not imminent, is inevitable.106 Why might this be the case? One option potentially available to a battered woman is simply to leave the relationship. Yet various economic and social reasons why battered women are unable to leave abusive relationships have been advanced. Many women lack independent financial resources and would not be able to support themselves or their children if they left the relationship, especially if they are not able to draw on the support of family or friends.107 If the woman was unable to 102 In other words, she should be judged from the perspective of a reasonable woman endowed with the characteristics of a sufferer of BWS. 103 It is generally accepted that it is relevant to import some of the accused’s characteristics into the analysis—for example, her age and (more controversially) her gender. In R v Smith [2001] 1 AC 146, a case concerning the partial defence of provocation, the House of Lords held that the standard of self-control expected of the reasonable person might be adjusted downwards if the defendant was suffering from a mental illness, or indeed if she had some other (unspecified) characteristic that might reduce her capacity for self-control. In Attorney-General for Jersey v Holley [2005] UKPC 23, however, nine judges sitting as the Judicial Committee of the Privy Council held that Smith was wrongly decided. Holley was approved by the Court of Appeal in R v James [2006] EWCA Crim 14. 104 In England, see s 2 of the Homicide Act 1957; in Scotland, see Galbraith v HM Advocate 2002 JC 1. 105 See J Gardner and T Macklem, ‘Compassion without respect? Nine fallacies in R v Smith’ [2001] Crim LR 623–35, which argues against importing personal characteristics into standards of reasonableness. 106 McColgan, ‘In defence of battered women’, at 516–17; Rosen, ‘On self-defense’, at 394–5; Schulhofer, ‘The gender question in criminal law’, at 128. 107 DS Kalmuss and MA Straus, ‘Wife’s marital dependency and wife abuse’ (1982) 44 Journal of Marriage and the Family 277–86; MJ Strube and LS Barbour, ‘The decision to leave

106

Killing in Self-Defence

support her children alone, she might have to leave them in the care of her abusive partner in order to save herself from harm. Shelters run by the state or by charitable organizations might be available, but places are limited and a shelter can provide only a temporary solution. In addition, it has been suggested that battered women are at the greatest risk of harm from their partners when they attempt to leave the abusive relationship. Mahoney uses the term ‘separation assault’ to describe the experience of women who have left abusive relationships, only for their partner to track them down and severely harm them.108 She suggests that half of all women who attempt to leave an abusive partner are followed and further harassed by him. Mahoney draws on official statistics in the USA to conclude that half of all men who kill their female partners do so when they are separated. Figures for England and Wales indicate that 30 per cent of women killed by their husband were no longer living with him at the time of their death.109 The reasonableness of the battered woman’s perception that, if she attempts to leave, she will be subject to worse violence might be reinforced if she has previously attempted to leave and experienced this.110 If the battered woman does not avoid the violence by leaving her home, another alternative might be to seek the assistance of the authorities in protecting her and, if relevant, her children, from the abuser. It has been argued, however, that the ability of the police and the courts to protect battered women from their violent partners is limited as the police do not always treat domestic violence seriously.111 There is evidence to suggest that the police response to domestic violence is improving,112 but this is not unequivocal.113 It has also been suggested that the various court orders for which a woman can apply in order to prevent an abusive partner from visiting her home are ineffective.114 Indeed, a court order cannot physically prevent an abuser from harming his partner. All it can realistically do is to an abusive relationship: economic dependence and psychological commitment’ (1983) 45 Journal of Marriage and the Family 785–93. 108 Mahoney, ‘Legal images of battered women’, at 65. See also C Connelly, ‘Domestic abuse’ in J Scoular (ed), Family Dynamics: Contemporary Issues in Family Law (2001) 177–204; A Browne and KR Williams, ‘Exploring the effect of resource availability and the likelihood of female-precipitated homicides’ (1989) 23 Law and Society Review 75–94. 109 SM Edwards, Policing ‘Domestic’ Violence: Women, the Law and the State (1989), at 200. 110 See also M Wilson and M Daly, ‘Spousal homicide risk and estrangement’ (1993) 8 Violence and Victims 3–16. Wilson and Daly summarize existing literature and conclude that women face a higher risk of violence from their partner (or ex-partner) when they are separated from him than when they are cohabiting with him. 111 New Zealand Law Commission, at para 27. 112 Home Office, Domestic Violence: A National Report (2005). For the latest legislative attempt to tackle domestic violence in England and Wales, see the Domestic Violence, Crime and Victims Act 2004. 113 Connelly, ‘Domestic abuse’, at 199–202; M McLean, ‘Domestic violence and repeat victimisation’ (2005) 12 International Review of Victimology 51–74. 114 M Burton, ‘Criminalising breaches of civil orders for protection from domestic violence’ [2003] Crim LR 301–13.

Imminence of Harm

107

punish the abuser after he has broken its terms, by which time it might be too late to prevent any harm from occurring.115 Without considering individual cases, it is difficult to say whether these barriers to the success of alternative courses of action would assist the battered woman in meeting an inevitability of harm test. Those who think that this test could be satisfied by a battered woman often point to the extreme case of State v Norman, where the defendant’s attempts to seek help to protect herself from the violence of her husband were ineffective.116 Judy Norman probably would meet an inevitability of harm test,117 but such cases will not be common. Desperate as their situation is, for the vast majority of women in situations of domestic violence, reasonable alternative courses of action are available. If a battered woman kills her passive partner, rather than pursue reasonable alternative courses of action, she should not be granted the defence of self-defence. That is not to say that I have no sympathy with the position of the battered woman who feels that she has no alternative to escape domestic violence other than to kill her partner. On the contrary, sustained domestic violence is a terrible thing to experience and there is something very wrong with a society that is unable to offer effective protection against this. But the appropriate response is for society to offer better protection in order to prevent domestic violence escalating to the point where the victim feels she has no choice but to kill her partner. Achieving this is far easier said than done: effecting change in, for example, entrenched cultural attitudes towards domestic violence victims is no easy task. But this, surely, is where the emphasis for change should be placed, rather than upon the structure of the law of self-defence.118 Of course, few commentators would disagree that emphasis should be placed on improving society’s response to domestic violence. The argument has been made, though, that, until change has been effected, there is simply no point in punishing a woman who has killed her abusive partner, because the threat of punishment is unlikely to deter and she is unlikely to pose any further threat to society.119 Against this, it might be said that a criminal conviction serves an expressive function.120 It signifies that the criminal law pays equal respect to the lives of all human beings and that those who kill in 115

McColgan, ‘In defence of battered women’, at 527. Judy Norman had attempted to leave her husband on several previous occasions and on each occasion he had found her, taken her home, and beaten her. She was afraid to take out a warrant on him because he had said that if she ever had him locked up, he would kill her when she got out. He had also told her that he would see the authorities coming for him and before they got to him, he would cut her throat. See State v Norman 378 SE 2d 8 (1989), at 11. Rosen, in ‘On self-defense’, states that: ‘After reading the Norman decision, I was left with a firm belief that, had she not killed her husband, he would have inevitably killed or seriously harmed her, no matter what she tried to do to change the situation’ (at 372). 117 Although even if she did, her self-defence claim might meet difficulties in satisfying the test 118 of proportionality. People v Aris 215 Cal App 3d 1178 (1989), at 1190. 119 120 Appel, ‘Beyond self-defense’, at 978. See Chapter 9. 116

108

Killing in Self-Defence

circumstances where that course of action could have been avoided are worthy of at least some condemnation, regardless of how much sympathy we feel for their difficult circumstances. It should perhaps be reiterated that it is being advocated here that the imminence test should be replaced by one of inevitability of harm, which would bring at least some battered women within the ambit of the law of self-defence, who would previously have been denied the defence. It should also be pointed out that it is not suggested that battered women who kill, but do not meet the inevitability of harm standard, should be convicted of murder, with the automatic life sentence that would bring, at least in England and Scotland. On the contrary, there are likely to be other (admittedly partial) defences available, such as diminished responsibility and provocation, both of which would, in appropriate circumstances, allow a non-custodial sentence to be passed. It may even be that some type of complete excuse claim can be constructed for such women.121 These issues are beyond the scope of this book, which is concerned with self-defence. What I am arguing is that, in all but the most exceptional circumstances, killing a sleeping person simply does not have the gist of a claim of self-defence, because it cannot be said that the harm to be prevented was inevitable or could have been perceived as such on reasonable grounds, or that no reasonable alternative course of action would have prevented the harm. The suggestion that battered women who kill their partners should be channelled into the partial defences of provocation or diminished responsibility sometimes meets with resistance. This is because, given the choice, the rational accused would prefer to have a justification defence, such as selfdefence, as this signifies that what she did was acceptable. A (partial) excuse defence signifies that what she did was unacceptable, but that we (partially) excuse her for it. A (partial) lack of capacity defence signifies that the accused (to some extent) does not have the required mental capacity whereby her behaviour can be judged acceptable or unacceptable.122 It may be that, for all the sympathy we might feel for the battered woman, and for all the condemnation we might feel for the actions of the abuser, partial excuses or lack of capacity defences are the appropriate defences for the vast majority of battered women who kill passive abusers.123 121 Although I have some doubts about this. Dressler attempts to construct a complete defence based on duress (‘Battered women’, at 275–81), but his argument is not entirely convincing (on this, see Chapter 1). 122 J Gardner ‘In defence of defences’, in P Asp, CE Herlitz, and L Holmqvist (eds), Flores Juris et Legum: Festskrift till Nils Jareborg (2002) 1–17; J Gardner, ‘The mark of responsibility’ (2003) 23 OJLS 157–71. 123 P Westen and J Mangiafico, ‘The criminal defense of duress: a justification not an excuse and why it matters’ (2003) 6 Buffalo Criminal Law Review 833–950, at 911.

6 Self-Generated Self-Defence 6.1 Introduction Self-generated self-defence refers to the situation where the accused contributes in some way to generating the need to use self-defensive force. The issue is one that rarely receives extensive coverage.1 But this is surprising because the majority of reported self-defence cases are self-generated. Cases of ‘pure self-defence’2—where the accused uses defensive force to repel a sudden and unexpected attack for which she is in no way to blame—are rare.3 Far more common are cases in which the accused played some part, even if it was a minor one, in contributing to the situation that ultimately led to the need to use self-defensive force.4 At the outset, it is necessary to say something about terminology as it is not always obvious how best to refer to the two parties in a self-defence encounter where both are to some degree at fault. The person who wishes to plead self-generated self-defence will be referred to as ‘the accused’. The person who has been the recipient of the accused’s self-defensive force will be 1 Only two published papers written in English could be found that deal substantially with the issue in relation to self-defence: PH Robinson, ‘Causing the conditions of one’s own defence: a study in the limits of theory in criminal law doctrine’ (1985) 71 Virginia Law Review 1–62 and J Herrmann, ‘Causing the conditions of one’s own defence: the multifaceted approach of German law’ [1986] Brigham Young University Law Review 747–67. Robinson and Herrmann do not confine their analysis to self-defence, but examine the role that initial fault plays in a variety of defences, including intoxication, necessity, and duress. On this, see also: A Ashworth, ‘Reason, logic and criminal liability’ (1975) 91 LQR 103–30; AR Ward, ‘Making some sense of self-induced intoxication’ (1986) 45 CLJ 247–61; P Alldridge, ‘Self-induced provocation in the Court of Appeal’ (1991) 55 JCL 94–102. 2 F Leverick, ‘A critical analysis of the law of self-defence in Scotland and England’ (2003), unpublished PhD thesis, Chapter 4. 3 An examination of reported Scottish cases of the last 100 years did not find a single clearcut example of pure self-defence (see chapter 7 of Leverick, ‘A critical examination of the law of self-defence’, for more detail). It may be, of course, that such cases are not prosecuted, so do not find their way into the law reports. 4 In Scotland see: HM Advocate v Kizileviczius 1938 JC 60 (accused previously hit deceased on the chin); Crawford v HM Advocate 1950 JC 67 (accused previously argued with deceased and approached him with a knife); HM Advocate v McGlone 1955 SLT 79 accused and deceased previously engaged in struggle); Surman v HM Advocate 1988 SLT 371 (accused went to house of deceased, knowing he was not welcome there); Jones v HM Advocate 1990 SLT 517 (accused and deceased had earlier argument); Fitzpatrick v HM Advocate 1992 SLT 796 (accused participated in earlier pub fight); Boyle v HM Advocate 1993 SLT 577 (accused participated in pub fight); Burns v HM Advocate 1995 JC 154 (accused previously assaulted deceased’s brother); McBrearty v HM Advocate 1999 SLT 1333 (accused previously argued with friend of deceased).

110

Killing in Self-Defence

referred to as ‘the victim’. This will sometimes sound odd, for example where the contribution of the ‘victim’ is to launch a life-threatening attack and the only contribution of the accused is to fail to avoid a place where she knows she is likely to be attacked. In the interests of consistency of terminology, however, this is the approach that will be taken.

6.2 Categorization of Self-Generated Self-Defence Before proceeding, it is important to examine more closely what is meant by self-generated self-defence. The term is used to cover all occasions where the accused contributed in some way to the need to use self-defensive force. This covers four conceptually distinct categories of case as follows.

6.2.1 Deliberate generation of a self-defence situation In this first category of case, the accused deliberately provokes an attack, with a view to using ‘defensive’ force to repel that attack and kill the attacker. The initial provocation might be verbal or it might be physical force. Whichever it is, the defining feature of this type of case is that the accused right from the start intended to kill or injure the ‘victim’ under the guise of acting in ‘self-defence’.

6.2.2 Starting the trouble In this second category of case, the accused does not set out deliberately to provoke an attack but is the person who starts the trouble that leads to the need to use self-defensive force. This category can in turn be subdivided according to the response of the victim: (a) Disproportionate response: Cases where the victim responds disproportionately to the accused’s initial act. One example might be where the accused verbally insults the victim and the victim responds by physically attacking her. Another example might be where the accused partakes in a minor physical assault on the victim and the victim responds by producing a knife and attempting to stab the accused.5 (b) Proportionate response: Cases where the victim responds proportionately to the accused’s initial act. An example might be where the accused attacks the victim with a knife and the victim defends herself with a comparable weapon. 5 Gordon gives an example where a robber attacks a shopkeeper, the shopkeeper responds by using force beyond what is necessary to defend himself and the robber kills him in self-defence. These are the facts of Robertson and Donoghue v HM Advocate (High Court, October 1945), noted in GH Gordon, The Criminal Law of Scotland (2000–1), at para 24.10.

Self-Generated Self-Defence

111

6.2.3 Agreeing to fight This third category of case covers situations where one party uses selfdefensive force in a fight that was initially started with the consent of the two parties involved. Historically this scenario might have arisen in a duel, once a common way of settling disputes.6 The formal duel plays no part in modern society, but nonetheless situations where two people consent to settle their differences by a physical fight still arise in the case reports. Two examples (albeit of cases that did not involve lethal force) are the Scottish case of Smart v HM Advocate7 and the English case that led to the Attorney-General’s Reference (No 6 of 1980).8

6.2.4 Failing to avoid trouble In this fourth category of case, the accused contributes to the need to act in self-defence by failing to avoid a situation that was likely to lead to trouble. An example might be the accused who goes to a particular bar, knowing that her adversary, who has previously threatened to kill her, is likely to be drinking there.9

6.2.5 Difficulties in classification Although these distinctions have been made for analytical purposes, it will not always be easy to classify cases as one type or another or even to distinguish cases where the accused in some way, however insignificant, contributed to generating the self-defence situation from cases where she did not. It will sometimes be difficult to pinpoint exactly what the first incident was in the transaction between accused and victim, especially where they have known each other for some time. In addition, what starts as one category of case may later go on to develop features of another category. For example, the accused may be the initiator of a verbal argument in a bar (‘starting the trouble’). This may lead to a situation where the two parties agree to go outside and settle the dispute with a physical fight (‘agreeing to fight’). Notwithstanding these difficulties, there is some value in examining each of the four categories of case separately, especially given that they have, at various points in history, been treated as having different legal consequences.

6.3 What are the Legal Options? There are various ways in which the law might deal with the accused who pleads self-defence but contributed in some way to the need to do so. First, 6 7 9

DT Andrew, ‘The code of honour and its critics’ (1980) 5 Social History 409–34. 8 1975 JC 30. [1981] QB 715. The facts of State v Bristol 84 P 2d 757 (1938) (Supreme Court of Wyoming).

112

Killing in Self-Defence

the law might deny the defence entirely. Second, and at the other end of the scale, the law might find her contribution completely irrelevant and allow her to plead self-defence regardless. Third, and between these two extremes, the law might prevent the accused who contributed to the conflict from pleading self-defence as a complete defence but might allow it to act in mitigation of punishment. Fourth, the accused might be permitted to plead selfdefence, but only if she first satisfies certain conditions, such as making an attempt to withdraw from the conflict. It is rare for someone to argue for the first of these options whenever the accused is responsible for creating the self-defence situation, however minor or blameless her contribution. Harel10 is one example. He argues that denying the defence would minimize the incidence of violence in society by providing an incentive to individuals to moderate their behaviour. If individuals know that they will not be permitted to use self-defensive force if they contribute in some way to the need to do so, then they will act more carefully, thus creating a society with a lower incidence of crime and violence. Harel’s argument must be seen in the context of his overall thesis, which might best be described as an economic analysis of law, with the primary purpose of the criminal law being to promote certain types of socially beneficial behaviour.11 He is operating under the assumption that it is possible for the criminal law to influence behaviour in this way.12 It is doubtful that, in reality, Harel’s proposal would influence behaviour in the way that he hopes.13 But even if it did, it would seem unfair always to deny the defence of self-defence in these circumstances, especially where the initial provoking act was a minor one to which the victim reacted disproportionately. In addition, Harel’s central thesis (that the primary purpose of the criminal law is to influence behaviour) can be questioned. The criminal law might have, for example, a protective function, in detaining those who are incapable of being deterred by its provisions but who, nonetheless, are a danger to society. On the other hand, it would seem inappropriate to go to the other extreme and always allow an initial aggressor to plead self-defence. It would be particularly unfair to the innocent victim of an initial attack if the attacker was permitted to respond defensively to the victim’s use of defensive force. The appropriate response of the criminal law would seem to depend on the circumstances. With this in mind, each category of self-generated selfdefence identified above will be examined in turn. 10 A Harel, ‘Efficiency and fairness in criminal law: the case for a criminal law principle of comparative fault’ (1994) 82 California Law Review 1181–229, at 1185. See also Chapter 4, where Harel’s views on retreat were discussed. 11 Harel, ‘Efficiency and fairness’, at 1185. 12 Cf PH Robinson and JH Darley, ‘Does criminal law deter? A behavioural science investigation’ (2004) 24 OJLS 173–205. 13 See Sections 6.5.2.2 and 6.6 below, where it is suggested that Harel’s proposal might well have the opposite effect to the one he hopes for.

Self-Generated Self-Defence

113

6.4 Deliberate Generation of a Self-Defence Situation The first situation is likely to be the least common in practice and seems, on the face of it, the easiest to settle. It is not particularly controversial to suggest that the accused who deliberately provokes the victim with a view to killing or injuring her with ‘self-defensive’ force should not be acquitted on the basis of self-defence. After all, taking an overview of the situation, it would seem that the accused does not act primarily in self-defence, but acts for an entirely different motive. On the very rare occasions where the issue has arisen, this is the approach that has been taken. In Mason’s case,14 the defendant verbally provoked the deceased and taunted him with a cudgel,15 before dropping the cudgel on the floor. The deceased picked it up and hit the defendant twice, at which point the defendant produced a sword he had been concealing inside his coat and stabbed the deceased. He attempted to plead self-defence but was convicted of murder. The judge noted that ‘there are in this Case so many circumstances of deliberate Malice and deep revenge on the Defendant’s part, that his Offence cannot be less than Wilful Murder’.16 Foster, who reported the case, commented that ‘all Circumstances considered, [the defendant] appeareth to have returned with a deliberate Resolution to take a deadly Revenge for what had past: and the Blows were plainly a Provocation Sought on his part, that he might execute the wicked purpose of his Heart, with some colour of Excuse’.17 The defence was ruled out by the historical English writers Hale,18 Foster,19 and East.20 This is also the position taken by the US Model Penal Code, which provides that the use of deadly force in self-defence is not justifiable if ‘the actor, with the purpose of causing death or serious injury, provoked the use of force against himself in the same encounter’.21 The Model Penal Code provision has been adopted by a number of US states.22 None of the influential Scottish writers of criminal law texts even mention the issue. 14

15 16 (1756) Foster 132. A short thick stick. ibid 134. ibid 134–5, emphasis in original. 18 M Hale, Historia Placitorum Coronae Volume 1 (1736), at 480, 457. 19 M Foster, A Report of some Proceedings on the Commission of Oyer and Terminer and Goal (sic) Delivery for the Trial of the Rebels in the Year 1746 in the County of Surry, and of other Crown Cases to which are added Discourses upon a few Branches of the Crown Law 20 (1762), at 277. EH East, Pleas of the Crown (1803), at 278. 21 § 3.04(2)(b)(i). 22 Ala Code § 13A-3-23(c)(1); Alaska Stat § 11.81.330(a)(2); Ark Code § 5-2-606(b)(1); Colo Rev Stat § 18-1-704(3)(a); Del Code Ann tit 11 § 464(e)(1); Ga Code Ann § 16-3-21(b)(1); 720 ICLS § 5/7-4(b); Iowa Code § 704.6(2); Kan Stat Ann § 21-3214(2); Me Rev Stat Ann tit 17-A § 108(2)(C)(1); Mont Code Ann § 45-3-105(2); Neb Rev Stat § 28-1409(4)(a); NH Rev Stat Ann § 627:4(I)(a); NY UCC Law § 35.15(1)(a); ND Cent Code § 12.1-05-03(2)(a); Or Rev Stat § 161.215(1); 18 Pa Cons Stat § 505(b)(2)(i); Utah Code § 76-2-402(2)(a); Wis Stat § 939.48(2)(c). See also 9 Guam Code Ann § 7.86(b)(1). 17

114

Killing in Self-Defence

The English Law Commission has considered the matter and reached the conclusion that the defence should not be available.23 It is perhaps not quite as simple as that. If one subscribes to the deeds theory of justification,24 as Robinson does,25 it would seem to follow that the accused who deliberately generates the need to use self-defensive force, even though she does not act exclusively for motive of self-defence, should be allowed the defence. To recap, under the deeds theory of justification, the accused who commits an offence in potentially justifying circumstances (here, to repel a life-threatening attack, albeit a self-generated one) should be granted a justification defence regardless of her motivation for doing so. On this basis, it would seem that the accused who deliberately generates the need to use self-defensive force would be justified under Robinson’s analysis. As Chapter 2 demonstrated, however, Robinson’s views on this issue have changed over time and while his early work suggests that the accused who deliberately generates a self-defence situation should escape entirely from criminal liability,26 his later work makes it clear that she should not.27 Robinson eventually accepted that the culpable state of mind of the accused (her ‘scheming intention’28) does justify punishment, even though her conduct is objectively justified.

6.5 Starting the Trouble 6.5.1 How has the law dealt with the accused who starts the trouble? Historically, in determining claims of self-defence English law attached considerable significance to the contribution of the defendant. As we have already seen,29 the law recognized two clear categories of self-defence, one being regarded as justifiable homicide and the other as excusable homicide. The blameworthiness of the defendant was central to this distinction. Where the defendant was not at fault in any way, the case was classified as justifiable homicide and resulted in an acquittal. Where the defendant was not wholly without fault, the case was at best one of excusable homicide, resulting in the defendant being remitted to jail to obtain a royal pardon and, even when this was obtained, attracting the forfeiture of her movable property. Indeed, influential English writers of the eighteenth century are unanimous in concluding that the defendant who made a premeditated murderous 23 Law Commission, Codification of the Criminal Law: A Report to the Law Commission (Law Com No 143, 1985), at para 13.35. Although the issue did not arise directly, the defence 24 also appears to be ruled out in R v Browne [1973] NI 96, at 106–7. See Chapter 2. 25 PH Robinson, ‘A theory of justification: social harm as a prerequisite for criminal liability’ (1975) 23 UCLA Law Review 266–92; PH Robinson, ‘Criminal law defenses: a systematic analysis’ (1982) 82 Columbia Law Review 199–291; PH Robinson, ‘Competing theories of justification: deeds v reasons’, in AP Simester and ATH Smith (eds), Harm and Culpability 26 (1996) 45–70. Robinson, ‘A theory of justification’, at 266. 27 28 29 Robinson, ‘Causing the conditions’, at 31. ibid, at 31. See Chapter 4.

Self-Generated Self-Defence

115

attack on the victim could not claim the defence of self-defence at all, even if she made an attempt to withdraw from the conflict.30 This was on the basis that the victim of the attack could not trust the initial aggressor not to renew her attack even if she does appear to have retreated. This is illustrated in R v Mawgridge,31 where Holt CJ commented as follows: Now it hath been held, that if A of his malice prepensed assaults B to kill him, and B draws his sword and attacks A and pursues him, then A for his own safety gives back, and retreats to a wall, B still pursuing him with his drawn sword, A in his defence kills B. This is murder in A. For A having malice against B and in pursuance thereof endeavouring to kill him is answerable for all the consequences, of which he was the original cause. It is not reasonable for any man that is dangerously assaulted, and when he perceives his life in danger from his adversary, but to have liberty for the security of his own life, to pursue him that maliciously assaulted him; for he that hath manifested that he hath malice against another is not fit to be trusted with a dangerous weapon in his hand.32

Where the English writers differ is in the extent to which a defendant who resorts to self-defence in the course of a sudden quarrel is permitted to plead self-defence. All agree that the defence would be available in some circumstances and that the result would be a finding of excusable homicide. After all, the whole basis of the category of excusable homicide was that it excused those who were partially at fault in generating the need to use self-defensive force. The differences lie in exactly what would be included in the category of excusable homicide. Hawkins takes a restrictive approach, allowing selfdefence to be pled only if the defendant did not strike the first physical blow.33 Foster and East are less restrictive. Both see the issue of who struck the first blow as irrelevant to the plea of excusable self-defence. As long as the fight developed from a sudden and unplanned quarrel between the parties and was not a premeditated attack on the part of the defendant, then the defendant will be able to plead self-defence, provided that she first makes an attempt to withdraw from the conflict.34 Blackstone would seem closest to the position of Foster and East, although he acknowledges that there are differences of opinion on the matter and does not clearly express a view one way or the other.35 In modern case law, the issue has rarely arisen directly. One exception is the Northern Irish case, R v Browne.36 Here, the defendant hijacked a car and was stopped by police. He refused to get out of the car when ordered to do so and one of the police officers hit him on the head with a gun. The 30 East, Pleas, at 278; Foster, Report, at 277; Hawkins, A Treatise of the Pleas of the Crown 31 (3rd edn, 1739), at 75. (1707) Kel 119. 32 ibid 128–9. The main issue in the case was not one of self-defence, but whether Mawgridge should be found guilty of murder or manslaughter. The case is discussed in detail by Jeremy Horder in the context of provocation: see J Horder, ‘The duel and the English law of homicide’ (1992) 12 OJLS 419–30, at 429 and 425; J Horder, Provocation and Responsibility (1992), at 52–3. 33 34 Hawkins, Treatise, at 75. Foster, Report, at 277; East, Pleas, at 281. 35 W Blackstone, Commentaries on the Laws of England: Volume 1 (1769), at 206. 36 [1973] NI 96.

116

Killing in Self-Defence

defendant responded by firing his own gun, killing one police officer and wounding another. He was convicted of murder after the trial judge withdrew self-defence from the jury. On appeal, Lowry LCJ held that in order for the defence to be available ‘[t]he need to act must not have been created by conduct of the accused in the immediate context of the incident which was likely or was intended to give rise to that need’.37 This seems, on the face of it, a more restrictive view even than that of the historical English writers. Browne appears to suggest that any conduct of the defendant that starts trouble will rule out the defence of self-defence. Unfortunately Lowry LCJ does not explain further, focussing instead on the issue of whether the action of the police officer in hitting the defendant with his gun was lawful. It is clear, however, that Browne does not represent the position of modern English law, although it is only relatively recently that the issue has been addressed directly, in R v Rashford.38 Prior to Rashford, there were a number of self-defence cases where the defendant obviously did start the trouble and the defence was not denied. In R v Morse,39 the conflict was initiated because the defendant had sexual intercourse with the victim’s wife. In R v McInnes,40 the defendant started an altercation between himself and the victim. In R v Bird,41 the defendant poured a glass of Pernod over her ex-boyfriend. In none of these cases was there any suggestion that the defendant’s initial conduct would bar a plea of self-defence. In Rashford, the appellant and the deceased had initially fallen out in a minor argument that did not involve violence. The prosecution case was that, later that day, the appellant had armed himself with a knife and sought out the deceased in order to ‘teach [him] a lesson’.42 Rashford was convicted of murder after he stabbed the deceased in the heart. At trial, he did not claim to have acted in self-defence, but the trial judge, nonetheless, instructed the jury on selfdefence, stating that if it was the defendant who ‘was the aggressor . . . who was out for trouble . . . well self-defence doesn’t come into it’.43 The Court of Appeal held that this was a misdirection, commenting that: ‘the mere fact that a defendant goes somewhere in order to exact revenge from the victim does not of itself rule out the possibility that in any violence that ensues selfdefence is necessarily not available as a defence’.44 The Court noted the absence of direct authority in English law, but placed considerable reliance on the Scottish case of Burns v HM Advocate,45 which was described as an ‘important decision [that] should be more widely known than it appears to be’.46 37

[1973] NI 106–7. [2005] EWCA Crim 3377. Browne is not referred to in Rashford, but the judgment of the Court of Appeal makes it quite clear that a defendant who starts the trouble will not automati39 cally be denied the defence of self-defence. (1910) 4 Cr App R 50. 40 41 42 [1971] 1 WLR 1600. [1985] 1 WLR 816. ibid [4]. 43 44 45 46 ibid [15]. ibid [19]. 1995 JC 154. ibid [19]. 38

Self-Generated Self-Defence

117

Burns is the leading authority on claims of self-defence by an initial aggressor in Scotland. In Burns, the Lord Justice-General (Hope) stated that: It is not accurate to say that a person who kills someone in a quarrel which he himself started, by provoking it or entering into it willingly, cannot plead self-defence if his victim then retaliates. The question whether the plea of self-defence is available depends, in a case of that kind, on whether retaliation is such that the accused is entitled then to defend himself. That depends on whether the violence offered by the victim was so out of proportion to the accused’s own actings as to give rise to the reasonable apprehension that he was in immediate danger from which he had no other means of escape, and whether the violence which he then used was no more than was necessary to preserve his own life or protect himself from serious injury.47

Thus, in Scots law, even where the accused was the first to use physical violence, she will be able to plead self-defence if the victim responded disproportionately and the accused had no other means of saving her life.48 In Rashford, the Court of Appeal held that this passage in Burns ‘should be regarded as accurately representing English law as well’.49 Likewise, there is no support for Browne in the law of other common law jurisdictions. In Zecevic v DPP,50 the High Court of Australia held that, when the accused is the initial aggressor: . . . it will be for the jury to consider whether the original aggression had ceased so as to have enabled the accused to form a belief, upon reasonable grounds, that his actions were necessary in self-defence. For this purpose, it will be relevant to consider the extent to which the accused declined further conflict and quit the use of force or retreated from it.51

The Canadian Criminal Code specifically provides for the possibility of an initial aggressor pleading self-defence, with a separate provision— s 35—relating to the accused who was the first to use aggression or who otherwise provoked an attack upon himself. Under s 35, the defence will be available provided that, among other things, the accused did not at any point before the need to preserve himself from death or serious bodily harm ‘endeavour to cause serious bodily harm’52 and he ‘declined further conflict and quitted or retreated from it as far as it was feasible to do before the necessity of preserving himself from death or grievous bodily harm arose’.53 These provisions have been criticized as unduly rigid and potentially unjust.54 In R v McIntosh,55 however, the Canadian Supreme Court held that 47 ibid 159. The Lord Justice-General drew on the previous authorities of Boyle v HM Advocate 1993 SLT 577; Robertson and Donoghue v HM Advocate (High Court, October 1945) (see Gordon, at para 24.10); and HM Advocate v Kizileviczius 1938 JC 60. 48 This leaves open the question of what happens when the victim responds proportionately. 49 On this, see Section 6.5.2.2 below. ibid [19]. 50 (1987) 162 CLR 645. 51 Zecevic v DPP (1987) 162 CLR 645, per Wilson, Dawson and Toohey JJ at 663. 52 53 s 35(b). s 35(c). 54 D Stuart, Canadian Criminal Law: A Treatise (4th edn, 2001), at 484. 55 [1995] 1 SCR 686.

118

Killing in Self-Defence

an initial aggressor need not rely on s 35, but instead can utilize the general self-defence provisions in s 34 of the Code, which do not contain strict rules on retreat. This has left the law in a state of some confusion, as s 34 applies only to an accused who has caused death or grievous bodily harm. This means that the accused who started the trouble but caused only minor injuries (or indeed no injury at all) in defending himself is subject to the strict rules of s 35, whereas the accused who started the trouble and killed in his own defence is not, a situation that has led to calls for reform.56 The approach taken in s 35 of the Canadian Criminal Code of requiring the accused who has started the trouble to decline further combat before being permitted to plead self-defence is an approach that is common in the USA where most state codes contain specific rules about the conditions under which an initial aggressor can plead self-defence. Some state codes contain a provision that where the defendant started the trouble but was met with a disproportionate response from her victim, the defence of self-defence should be available to her.57 Others, either instead of or in addition to this, contain a provision that, where the defendant was the initial aggressor, she must have withdrawn from the attack and communicated this to the victim before being permitted to plead self-defence.58 We will return to these possibilities later.

6.5.2 How should the law respond where the accused has started the trouble? It is suggested that where the victim responds disproportionately to the accused’s initial act, the accused should not be prevented from pleading selfdefence (assuming, of course, that all of the other conditions of the defence are satisfied). Where the victim responds proportionately to the accused’s initial act, the defence should be denied unless the accused withdraws from the conflict and took all reasonable steps to communicate this to the victim (again assuming the other conditions of the defence are satisfied). In both of these cases, where the initial conduct of the accused is serious enough to constitute a criminal offence, she should be held liable for that offence. The reasons for these conclusions are outlined below.

6.5.2.1 The disproportionate response situation At the outset, it is worth considering why the law might wish to withhold the defence of self-defence from the accused who started the trouble. One reason 56 Canadian Department of Justice, Provocation, Self-Defense and Defence of Property: A Consultation Paper (1998). 57 Iowa Code § 704.6(3)(a); Kan Stat Ann § 21-3214(3)(a); ND Cent Code § 12.1-0503(2)(b). 58 Ala Code § 13A-3-23(c)(2); Alaska Stat § 11.81.330(b); Ariz Rev Stat § 13-404(B)(3); Ark Code § 5-2-606(b)(2); Cal Penal Code § 776.041(2)(b); Colo Rev Stat § 18-1-704(3)(b);

Self-Generated Self-Defence

119

is that she is not free from fault but has brought her predicament on herself. After all, earlier in this text,59 it was proposed that defences should be granted only to those who are appropriate candidates for them. It might be said that the accused who starts trouble is morally blameworthy and therefore does not deserve a defence. It might also be said that the action she takes in self-defence is not necessary as she could have avoided it by not starting the trouble in the first place. Against this, it might be argued that denying the defence in the disproportionate response situation is unduly harsh. Take the example of Karen, who starts the trouble with provocative words, such as a confession of infidelity. Steve responds by attacking her with a knife and Karen kills him in the process of trying to defend herself. While Karen may have behaved in a way that would solicit disapproval from some, Steve responded in a disproportionate manner. To punish Karen for acting in self-defence would seem to be punishing her solely for her initial conduct and while some in society would disapprove of infidelity, few would argue that it merits a criminal conviction. This argument still has some force where the accused starts the trouble with physical force. Take the example of Charlie, who slaps Shelley in the course of an argument. Shelley responds by attacking Charlie with a weapon, but Charlie then kills Shelley in the process of defending himself. Once again, while Charlie may have behaved in a way that occasions disapproval, it would seem unnecessarily harsh to convict him of murder or even manslaughter. If his initial slap was serious enough to constitute an offence such as assault, then he should be convicted of that offence. To convict him of a more serious offence would seem unfair. A second argument that might be made in favour of withholding the defence is that the response of the victim is an entirely foreseeable consequence of the accused’s actions. Take the example of Kevin, who has, in the past, told Sally that he would kill her if she ever had an affair. Sally confesses her infidelity to him and he reacts disproportionately, but arguably foreseeably, by attempting to strangle her. Should Sally be denied the defence of selfdefence if she uses lethal force to defend herself? After all, it could be said that she was reckless in making the confession of infidelity, and should be held responsible for the consequences of her recklessness. This does, though, seem a harsh conclusion to reach. After all, it might be said that it is precisely because we do not view a lethal attack as a natural and foreseeable consequence of a confession of infidelity that provocation is not a complete defence to murder. On the contrary, we still hold the accused Ga Code Ann § 16-3-21(b)(3); 720 ICLS § 5/7-4(c)(2); Iowa Code § 704.6(3)(b); Mo Rev Stat § 563.031(1)(1)(a); Mont Code Ann § 45-3-105(2)(b); NH Rev Stat Ann § 627:4(I)(b); NY UCC Law § 35.15(1)(b); ND Cent Code § 12.1-05-03(2)(b); Or Rev Stat § 161.215(2); Tenn Code Ann § 39-11-611(d); Tex Penal Code § 9.31(b)(4)(A); Utah Code § 76-2-402(2)(c)(i); Wis Stat § 939.48(2)(b). 59

See Chapter 2.

120

Killing in Self-Defence

criminally liable, albeit to the reduced charge of manslaughter, if she kills in response to provocation. Therefore, even where the victim’s response was foreseeable, such as in Kevin’s case, this should not prevent the accused from pleading self-defence. Kevin’s response may have been predictable, but it was still disproportionate. It seems unduly harsh to require Sally simply to submit to his life-threatening attack on the basis that she ought to have known that it would take place.60 Before rejecting this argument entirely, it is worth noting that this is precisely the basis on which the law has dealt with the issue in relation to other defences. Take, for example, the defence of duress. In R v Hasan,61 it was held by the House of Lords that duress is not available to a defendant who joins a criminal organization where there was a foreseeable risk that she would be put under pressure to commit an offence by members of that organization.62 In Scotland, voluntary intoxication is ruled out as a defence on a similar basis. In Brennan v HMA,63 it was held that self-induced intoxication is itself evidence, at least in part, of the criminal recklessness that can constitute the mens rea of crimes of violence such as murder.64 The Lord Justice-General (Emslie) stated in Brennan that ‘[t]here is nothing unethical or unfair or contrary to the general principle of our law that self-induced intoxication is not by itself a defence to any criminal charge including in particular the charge of murder’.65 In Finegan v Heywood,66 it was suggested that the defence of automatism would also be ruled out if there was any prior fault on the part of the accused that resulted in her being in that state.67 If the defences of duress, intoxication, and automatism are withheld from the accused who voluntarily exposes herself to the risk that she will engage in criminal activity, why should it be any different in the case of self-defence? There are two possible explanations for this apparent inconsistency. First, self-defence, unlike any other complete defence, always involves causing harm to someone who was a direct threat. The availability of the defence derives from the conduct of the victim.68 Unlike self-defence, the availability of defences such as automatism, duress, and intoxication does not normally 60 This situation should be distinguished from the Mason-type situation discussed in Section 6.4, where the accused deliberately provokes the victim into attacking her, with the intention all 61 along of using ‘self-defensive’ force to kill. [2005] UKHL 22. 62 63 R v Hasan, per Lord Bingham at [37]; Baroness Hale at [77]. 1977 JC 38. 64 Later cases have suggested that evidence of intoxication may result in mens rea being presumed even in relation to offences such as assault that cannot be committed recklessly (see Ross v HM Advocate 1991 JC 210). The position on voluntary intoxication is slightly more complicated in English law—see DPP v Majewski [1977] AC 443. But English law also accepts that where a crime can be committed recklessly, voluntary intoxication is no defence. 65 66 Brennan, at 51. 2000 JC 444. 67 Finegan v Heywood, per the Lord Justice-General (Rodger) at 451–2. 68 The only other defence where this can be said always to be the case is the partial defence of provocation and here, consistent with self-defence, the Privy Council has held that the fact that provocation was self-induced does not necessarily rule out the defence (R v Edwards [1973] AC 648; R v Johnson [1989] Crim LR 738).

Self-Generated Self-Defence

121

depend on the conduct of the victim. If a victim is harmed under duress or under the influence of intoxication, she is likely to be an innocent bystander, who posed no threat to the accused and made no contribution to the situation in which the accused finds herself. In these circumstances, it would seem reasonable to require that the accused, in order to benefit from a complete defence, is herself free from fault in creating the situation.69 The second is that, unlike the justification defence of self-defence, and except in some limited circumstances,70 defences such as duress and intoxication operate as excuse defences. The basis of an excuse is that we accept that the accused has done something wrong but we do not blame her for doing so. Thus there is an argument for withholding an excuse defence from someone who is partially blameworthy. The accused who voluntarily associates with a criminal gang or voluntarily becomes extremely intoxicated is not free from fault and it may be quite consistent that she cannot claim an excuse.71 To sum up, it is suggested that where the accused starts the trouble with provocative words or with a minor assault and the victim responds disproportionately, the accused should not be denied the defence of self-defence. To do so would be to punish her too harshly for her initial act. As we have already seen, this is, by and large, the approach taken in common law jurisdictions, none of which would deny the defence in these circumstances.72 Two final points should be made. First, it is worth reiterating that the accused who starts the trouble should be acquitted on the basis of selfdefence only if all of the other conditions of the defence are made out. Elsewhere in this text, it is argued that self-defence should be governed by strict rules on proportionality and necessity and that, in particular, any reasonable opportunity to retreat, rather than use lethal force, should be taken.73 These conditions should apply to self-generated self-defence just as they should apply to any other self-defence situation. Second, it is important to stress that the accused who starts the trouble remains criminally liable for her original act, if this is serious enough to constitute a criminal offence. It is not necessarily the case, therefore, that the accused who starts the trouble will remain entirely free from penal consequences. If she starts the trouble with minor physical violence, this may constitute assault. Even provocative words or gestures may constitute a public order offence, such as breach of the peace. But the argument made here is that it is this initial act that should determine the severity of her punishment and not the act of self-defence that follows. 69 Lowry LCJ makes something akin to this point in the Northern Irish case of R v Fitzpatrick 70 [1977] NI 20 (at 30), which concerned the availability of duress. See Chapter 2. 71 This is perhaps a less convincing argument. Even if it is accepted, one might question whether it applies to the excuse form of self-defence (where the accused is mistaken about the need to use self-defensive force). 72 73 See Section 6.5.1 above, although cf R v Browne [1973] NI 96. See Chapter 4.

Killing in Self-Defence

122

6.5.2.2 The proportionate response situation At the outset, it was suggested that the defence of self-defence should be denied in the proportionate response situation unless the accused withdraws from the conflict and takes reasonable steps to communicate this to the victim. It is helpful, in explaining why this is the case, to use the example of Violet, who attacks Jason with a knife. Jason responds to Violet’s attack by picking up an iron bar and attempting to defend himself. It has been argued that the defence of self-defence is permitted because, by virtue of her conduct in launching a disproportionate, and therefore unjust, life-threatening attack, the aggressor’s right to life is forfeited. With this in mind, it follows that because Violet has launched an unjust life-threatening attack, to which Jason responded proportionately, she cannot automatically be permitted to plead self-defence in turn. Her initial life-threatening attack means that she forfeits her right to life and Jason is permitted to respond proportionately in self-defence. Jason’s right to life is not similarly forfeited because he has acted entirely justly in using proportionate self-defensive force. If Violet withdraws from the conflict that she has created, however, she ceases to be a threat to Jason. We have already seen that the permissibility of self-defensive force relies on the existence of an immediate threat to life. If this is no longer the case, and Jason continues to hit Violet with the iron bar, Violet should be permitted to use defensive force. After all, Jason is no longer acting justly out of necessity, but has himself become an unjust immediate threat to the life of Violet. One very important proviso should be made, however, and that is that the withdrawal made by Violet must be a genuine one and she must take all reasonable steps to communicate it to Jason. This would mean that, by continuing his attack on Violet, there is no longer any possibility that Jason is acting for the purposes of self-defence but is instead acting for some other purpose, such as revenge. The requirement for communication is an important one. As McGarvie J stated, in R v Lawson and Forsythe,74 ‘it would never be enough that the original aggressor had a secret change of intention after the initial confrontation. A secret change of his own intention would not of itself have led him reasonably to believe that the original victim was no longer acting in selfdefence but as an aggressor’.75 There is a question of whether actual communication is required or whether it is sufficient that the initial aggressor took all reasonable steps to communicate her withdrawal from the attack. The latter approach would guard against the possibility of the entirely unreasonable victim who, in the face of all evidence to the contrary, continues to believe that she is under 74

(1985) 18 A Crim R 360 (Supreme Court of Victoria).

75

ibid 383.

Self-Generated Self-Defence

123

attack, although it might be argued that, on a ‘take your victim as you find her’ basis, the accused should still not be allowed to plead self-defence. The courts have occasionally gone as far as suggesting that there must be actual communication of the intention to withdraw before self-defence can be successfully pled. In People v Button,76 the Supreme Court of California refused to allow the defence where the aggressor’s initial attack dazed the victim and he did not realize that the aggressor had withdrawn. This does not seem entirely unreasonable on the facts, because the reason why the victim did not perceive the withdrawal was because the initial aggressor’s attack left him dazed. The more common approach in jurisdictions where the issue has arisen seems to be that reasonable steps will suffice.77 That said, it might still be argued that there are public policy reasons for disallowing the defence entirely. If we deny the defence to the accused who starts the trouble by launching a life-threatening attack, even where the accused communicates her withdrawal from the conflict, this might act as a disincentive to violence and aggression in society.78 But even if one accepts that the criminal law is capable of acting as an incentive in this way,79 this might equally lead to the opposite conclusion. That is, if an initial aggressor is permitted to plead self-defence provided that she ceases her attack and communicates this to her victim, this provides a very clear incentive for her to do so. If that incentive is not in place, she has every reason for continuing to use violence rather than withdrawing. In other words, there are equally compelling public policy reasons for allowing the defence as there are for disallowing it. It is proposed here, then, that where the victim’s response to the accused’s initial act is proportionate, the accused should be permitted to plead selfdefence only where she withdraws from the conflict and takes all reasonable steps to communicate this to her victim. As before, two final points are worth noting. First, the other conditions of the defence, including the retreat rule, must be made out. Second, the accused remains criminally liable for her original attack. Where this is a life-threatening one, the appropriate offence might well be attempted murder, despite the accused being acquitted of the subsequent killing on the basis of self-defence.

6.6 Agreeing to Fight Gail and Eileen become involved in an argument in a bar. They agree to settle their differences by stepping outside and trading blows to the death. It soon 76

39 P 1073 (1895). Stoffer v State 15 Oh S 47 (1864) (Supreme Court of Ohio), at 53; Castillo v State 614 Ala 2d 756 (1980) (Supreme Court of Alaska), at 766; People v Gleghorn 193 Cal App 3d 196 (1987) (Court of Appeal of California), at 85; State v Diggs 592 A 2d 949 (1991) (Supreme Court of Connecticut), at 951; State v Muhammad 757 SW 2d 641 (1988) (Missouri Court of 78 Appeals), at 643. Harel, ‘Efficiency and fairness’, at 1185. 79 Cf Robinson and Darley, ‘Does criminal law deter?’. 77

124

Killing in Self-Defence

becomes apparent that Eileen is the physically stronger of the two and that she will kill Gail if she is permitted to continue. Gail, realizing this, fortuitously manages to put her hands round Eileen’s neck and strangle her. Should she be permitted to plead self-defence to a charge of murder? The most productive way of examining the situation is to do so using the same principles as were applied in the previous section. By agreeing to fight to the death, Gail and Eileen both simultaneously become an unjust immediate threat to the life of the other. Their right to life is therefore forfeited. Thus, in order to plead self-defence, Gail should first make an attempt to withdraw from the conflict and only if Eileen continues her attack after Gail has taken all reasonable steps to communicate her withdrawal should Gail be permitted to plead self-defence. By attempting to withdraw, Gail signals that she is no longer an unjust immediate threat to the life of Eileen and therefore by continuing to attack, Eileen violates Gail’s (now restored) right to life. If one of the parties in the fight unexpectedly uses disproportionate force, the situation can also be analysed using the same principles as were applied in the previous section. For example, if Gail and Eileen had consented (expressly or impliedly) to a fist-fight and Gail produced a gun, it could be said that she responded disproportionately and it would be unfair not to allow Eileen to use defensive force to save her life. That said, there might still be public policy reasons for denying the defence entirely.80 It could be argued that, in the interests of maintaining public order in society, the law should discourage consensual fighting. One way to achieve this could be to deny the defence of self-defence in all circumstances where the fight was entered into on a consensual basis. Indeed, this is precisely the approach taken in some US states. One example is Alabama, where the use of force in defence of a person is not permitted where ‘[t]he physical force involved was the product of a combat by agreement not specifically authorized by law’.81 Two arguments might be marshalled against this. First, it would seem unfair to the accused who consents to one thing (say, a fist-fight) only for her opponent to react in a completely disproportionate manner (say, by attacking with a knife). To deny the accused a defence if she uses defensive force against her knife-wielding opponent would seem harsh. This type of fight was not, after all, what she consented to and she will not escape the consequences of the criminal law entirely. She will still be liable for her original act of agreeing to and commencing the fight, which could constitute an assault82 80

Harel, ‘Efficiency and fairness’, at 1185. Ala Code § 13A-3-23(c)(3). For similar provisions, see Ark Code § 5-2-606(b)(3); Colo Rev Stat § 18-1-704(3)(c); Kan Stat Ann § 21-3214(3)(b); NH Rev Stat Ann § 627:4(I)(c); NY UCC Law § 35.15(1)(c); Or Rev Stat § 161.215(3). 82 Assuming that consent is not a defence to assault where two parties have agreed to a fight. This is the position taken by Scots law (see Smart v HM Advocate 1975 JC 30). The position taken in English law is slightly more complicated, but in Attorney-General’s Reference (No 6 of 81

Self-Generated Self-Defence

125

or form the basis for conviction of a public order offence such as breach of the peace. Second, if it is accepted that it is possible for the law to influence human behaviour in this way (and this is an assumption made in the public policy argument)83 then surely it is preferable to provide an incentive for one of the parties in a consensual fight to attempt to end the contest? If the accused knows that, once she has consented to a fight, she will be denied the defence of self-defence regardless of whether or not she makes an attempt to withdraw from that fight, there is absolutely no incentive for her to do so. The central concern of this text has been with the preservation of human life and the suppression of violence wherever this is reasonably possible and to deny self-defence even to the party who attempts to withdraw from a consensual fight would seem incompatible with this end.

6.7 Failing to Avoid Trouble Janice works in a bar. Cilla, who has a history of violent conduct, buys a drink and accuses Janice of giving her too little change from her £20 note. Janice is adamant that she has given her the correct amount. An argument develops and Cilla storms out of the bar, threatening as she does so to kill Janice if she ever sees her again. When her shift ends later that night, Janice goes to another bar where she has arranged to meet friends. As she walks in the door, she sees Cilla sitting with another group of women. Cilla is shouting and obviously drunk. Nonetheless, Janice enters the bar and joins her friends. Cilla sees her and attacks her with a broken glass. Janice takes a knife from her bag and kills Cilla in order to save her life. Should she be permitted to plead self-defence?84 It seems, on the face of it, that the proper response of the law is obvious. Janice has done nothing that is sufficiently blameworthy to merit criminal sanctions. She has simply gone to a public place, albeit one where she might reasonably expect trouble to result. To punish Janice would be to punish her for doing something that might merit, at the most, mild disapproval. Is there any possible argument in favour of denying a defence to Janice? Ashworth certainly seems to think so.85 For Ashworth, there are two competing values at stake here: freedom of movement and the minimization of violent conflict. He argues that to allow the accused to plead self-defence 1980) [1981] QB 715 it was held that an act undertaken in a consensual fight can constitute assault, regardless of any consent to fight, if actual bodily harm was caused or there was an intention to cause actual bodily harm. This effectively rules out consent as a ‘defence’ to assault in anything other than a ‘minor struggle’ (at 719). 83 Cf Robinson and Darley, ‘Does criminal law deter?’ 84 This scenario is based on the facts of State v Bristol 84 P 2d 757 (1938) (Supreme Court of 85 Wyoming). A Ashworth, Principles of Criminal Law (5th edn, 2006), at 145.

126

Killing in Self-Defence

is to promote freedom of movement above any duty to avoid violent conflict that might result in the loss of life. Where it does not conflict with any other value, the accused should have the freedom to go to any public place she chooses, but here the value of freedom of movement is outweighed by the weightier value of minimizing violence.86 At this point, it is worth noting that the competing values would not always be so obviously different in magnitude as those of life and freedom of movement. Consider the example of Martin, the violent hospital porter, and Katy, who needs to visit the hospital where Martin works in order to receive a life-sustaining blood transfusion. Martin has previously threatened to kill Katy if he ever saw her again.87 Here, the two competing values at stake are identical: the life of Katy weighed against the life of Martin. It seems absurd in this situation to require Katy to risk her life by missing her blood transfusion in order to avoid a potentially violent conflict initiated by Martin. Where the two values in question are freedom of movement and life, though, the question is a more difficult one. Support for Ashworth comes from Dicey.88 Dicey uses the example of A, who knows that if he walks down a particular footpath he is in danger of an assault by X because X believes (wrongly) that he owns the footpath and A has no right to walk there. A could avoid the footpath by taking a slightly longer route to his destination. For Dicey, if A walks down the footpath and is assaulted by X, there are two reasons why he would not be justified in using self-defensive force. First, A’s use of force would be disproportionate because the interest he is trying to protect is his right to use the footpath and this is outweighed by the interest of X’s life. In response, it might be said that, at the time of the attack, this is simply not correct. If X launches a life-threatening attack on A and A repels this with defensive force, A is surely protecting his life, not his right to walk on the footpath. Second, Dicey argues that the force used by A is not necessary because he could easily have avoided all danger by choosing the alternative route. This may be correct, but brings us back to Ashworth’s question of whether or not A should be required to sacrifice his freedom of movement in order to avoid a situation that could lead to his own death or to that of his opponent. This argument is an important one to consider because the central concern of this text, in assessing the law of self-defence, has been to attempt to pay maximum respect to the right to life of all human beings. This would seem to suggest that we should be required to refrain from acting where it is 86

A Ashworth, Principles of Criminal Law (5th edn, 2006), at 145. Thanks are due to Professor Chris Gane for this imaginative example. The choice of names is my own. 88 AV Dicey, Introduction to the Study of the Law of the Constitution (4th edn, 1893), at 426. 87

Self-Generated Self-Defence

127

reasonably foreseeable that our course of action would result in another party responding violently. This would, after all, be likely to lead to a net saving of lives. But this surely goes too far. If a duty was imposed on individuals to avoid any place where there was a chance, however remote, that they would be attacked, it is difficult to see where the limits of such a duty would lie. For example, would it prevent an employee from going to her place of work if she had, on a previous occasion, been threatened by another employee? Would it prevent a woman who had previously been threatened by her violent partner from returning to her home? If so, this neglects the fact that the person who issues the threats also has a choice in whether or not to engage in violent behaviour. If such a rule existed, it would seem to condone aggressive and threatening behaviour by placing the onus on the victim of threats to avoid engaging in entirely lawful acts rather than on the issuer of the threats to curb her violent behaviour.89 A slightly different approach to arguing for the Ashworth/Dicey position is taken by the prosecution in State v Bristol,90 the case from which the Janice/Cilla scenario was drawn. Here, an analogy is drawn between the duty to avoid trouble and the duty to take any safe opportunity to retreat. In both cases, two competing values must be weighed. When deciding whether there should be a retreat rule, the competing values are the honour of the victim and the right to life of the attacker. In deciding whether there is a duty to avoid trouble, the competing values are freedom of movement and the right to life of the attacker. It is suggested by the prosecution in State v Bristol that if we require the victim of an attack to sacrifice her honour and retreat, we should also require the victim of threats to sacrifice her freedom of movement and avoid trouble. Given that a strong retreat rule has been proposed here on exactly this basis—that the life of the attacker should always be preferred over the honour of the victim91—the argument is initially attractive. There is, however, an important distinguishing feature of the avoiding trouble situation compared to the retreat situation: the magnitude of the burden placed on the victim. Requiring the victim of an attack to retreat, if a safe opportunity to do so exists, requires her to do this at a single point in time as part of a single incident of violence and, once she has reached a place of safety and informed the authorities of the attack, has no continuing impact on the manner in which she is required to lead her life. Requiring the victim of a threat to avoid her adversary potentially requires her to take evasive action for a much longer period of time and covering a much wider geographical scope. Exactly how long this should be is not clear. If Janice meets Cilla unexpectedly in a bar a year later, is she still required to curtail her freedom 89 91

State v Bristol 84 P 2d 757 (1938), at 762. See Chapter 4.

90

84 P 2d 757 (1938).

128

Killing in Self-Defence

of movement, just in case Cilla remembers her threat and acts upon it? Does the duty to avoid trouble extend to any place Janice visits in her entire lifetime? There is no clear answer to these questions if one accepts that the duty to avoid trouble exists. In addition, unlike in the retreat situation, it is difficult for Janice to know what she must do to avoid Cilla’s threat as Cilla’s own movements cannot easily be predicted. It is argued here, then, that the duty to avoid trouble can be distinguished from the duty to retreat. This was the approach taken by the court in State v Bristol,92 where the prosecution argument was rejected. Another case in which the issue arose directly was R v Field,93 a case decided by the English Court of Appeal. Here, there was a history of bad feeling between Field and the victim. On the day in question, Field was warned by a third party that the victim was looking for him. Field stayed outside, even though the third party warned him that the victim was approaching. When the victim arrived, Field declared that he did not want to fight, but the victim attacked him and Field stabbed the victim with a knife in selfdefence.94 Field was initially convicted of manslaughter, but the Court of Appeal made it clear that it was not the law that a man could be driven off the streets and compelled to avoid a place he could lawfully be, just because he had reason to believe he might be attacked.95 Field is supported by a series of cases in which defendants had initially been convicted of breach of the peace (or a related public order offence) because their behaviour was likely to provoke others to violence but where their convictions were overturned on appeal. In Redmond-Bate v DPP,96 for example, the appellant was one of three women who had been preaching on the steps of Wakefield Cathedral. A crowd had gathered and started behaving in a hostile way towards the preachers. A police officer, fearing trouble, asked the women to stop preaching but they refused and he arrested them. The appellant was convicted of obstructing a police officer in the execution of his duty. In overturning the conviction, the Divisional Court held that the appellant had been behaving in an entirely lawful way and that she could not be held responsible for the crowd’s response. The court held that there is an obligation to avoid conduct likely to cause trouble only where such conduct ‘gives rise to a reasonable apprehension that it will, by interfering with the rights and liberties of others, provoke violence’.97 The appellant was not interfering 92

93 84 P 2d 757 (1938), at 762. [1972] Crim LR 435. One might question on these facts whether this bordered on the deliberate generation of a self-defence situation on Field’s part, but this was not considered by the court. 95 96 ibid 435. [2000] HRLR 249. 97 ibid 258 (emphasis added). See also Beatty v Gilbanks (1882) 9 QBD 308; R v Morpeth Ward Justices ex p Ward (1992) 95 Cr App R 215; Percy v DPP [1995] 1 WLR 1382; R v Nicol and Selvanayagam [1996] Crim LR 318. For a contrary view expressed in English law, but criticized in Redmond-Bate, see Duncan v Jones [1936] 1 KB 218. For a Scottish case that reached the opposite conclusion, see Deakin v Milne (1882) 10 R (J) 22. 94

Self-Generated Self-Defence

129

with the rights or liberties of others and thus had no duty to refrain from acting. Indeed, the decision of the European Court of Human Rights in Steel and others v United Kingdom,98 suggests that to penalize an individual solely for being in a place that was likely to provoke a reaction from others might be a breach of the European Convention on Human Rights.99 98

(1999) 28 EHRR 603. It would be a breach of Art 5 (the right to liberty and security of person) and Art 10 (the right to freedom of expression). Steel concerned three individuals who had been convicted of breach of the peace by attending a military conference and handing out leaflets protesting against the sale of fighter helicopters. The European Court held that their rights under Arts 5 and 10 had been violated. 99

This page intentionally left blank

7 Killing to Protect Property 7.1 Introduction Thus far, the focus has been on issues relating to the necessity of using defensive force. This chapter turns to the issue of proportionality.1 At the outset, it can be said that the requirement for proportionality between force used in defence and the magnitude of the threat faced is a cornerstone of the law in common law jurisdictions.2 Thus this chapter and the one that follows will focus not on the general principle of proportionality, which is relatively uncontroversial, but on two particular examples: killing to protect property and killing to prevent rape. The question this chapter attempts to address is whether there are any circumstances in which the accused should be permitted a complete defence where she has killed to protect property.3 It might be thought that the answer to this question is obvious, given the line of argument already developed. The premise of this text is that all human beings, even those who commit or attempt to commit serious crimes, have a right to life. The reason we are permitted to kill someone who threatens our life is that the right to life of the aggressor is temporarily forfeited by virtue of becoming an unjust immediate threat to the life of another. Starting from this premise, it is difficult to see how lethal force could possibly be permissible against lesser threats than a threat to life. This is not a universally accepted position, though, and this chapter assesses some of the arguments that have been made in support of the principle that lethal force is permissible in defence of property. None of the arguments is convincing and the conclusion is reached that the right to life cannot be forfeited by becoming a threat to property because no item of property is of such value that it outweighs the value of human life. 1

For an explanation of the distinction between necessity and proportionality, see Chapter 1. Shaw v R [2001] UKPC 26, at [19] (Privy Council); McCluskey v HM Advocate 1959 JC 39, at 43 (Scotland); Viro v R (1978) 141 CLR 88, per Mason J at 147 (High Court of Australia). 3 See also M Jefferson, ‘Householders and the use of force against intruders’ (2005) 69 JCL 405–13; B Hogan, ‘Defence of property’ (1994) 144 NLJ 466–8 and 471; P Lanham, ‘Defence of property in the criminal law’ [1966] Crim LR 368–79 (part one) and 426–35 (part two); RJ Muth and AL Blumstein, ‘The use of deadly force in the protection of property under the Model Penal Code’ (1959) 59 Columbia Law Review 1212–32; DR Stuart, ‘Killing in defence of property’ (1967) 84 SALJ 123–31. 2

132

Killing in Self-Defence

7.2 Approaches to the defence of property In some common law jurisdictions, statute or case law specifically provides that the accused who kills to protect property does not have a defence to homicide. The Canadian Supreme Court, for example, confirmed in R v Gunning4 that lethal force can be used only in defence of persons and never in defence of property.5 In Scots law, no modern reported case has directly addressed the issue. The leading authorities on self-defence suggest, however, that it would not be sanctioned. In McCluskey v HM Advocate,6 the Lord Justice-General (Clyde) refused to extend the plea of self-defence to the prevention of a sexual assault by one man on another because of the ‘principle of the sanctity of human life’.7 When the same issue was considered in Elliott v HM Advocate,8 the Lord Justice-General (Emslie) approved the trial judge’s charge to the jury that ‘homicide will not be justified by self-defence unless it is committed of necessity in the just apprehension on the part of the killer that he cannot otherwise save his own life’.9 Given these comments, and the strict rules relating to other aspects of self-defence,10 it can probably be concluded that the acquittal of an accused who had killed in defence of property could not be legally justified.11 The situation is slightly less clear in English law. Here, defence of property can fall to be considered either under the common law or statute. Whichever approach is taken, the test applied is the same: whether or not the force used was reasonable.12 The relevant statutory provision is the Criminal Law Act 1967, which provides that ‘a person may use such force as is reasonable in the circumstances in the prevention of crime’.13 No further guidance is given in the Act as to the level of force that might be permissible in preventing a crime against property and thus what constitutes reasonable force is a question for the jury.14 If the defence of property falls to be considered under the common law, the same principle applies. In DPP v Bayer,15 Brooke LJ stated that ‘[i]t is a 4

[2005] 1 SCR 627. ibid [26]. The Supreme Court merely confirmed what lower courts had previously held: R v Baxter (1975) 27 CCC (2d) 96 (Ontario Court of Appeal); R v Clark (1983) 3 CCC (3d) 264 6 (Alberta Court of Appeal). 1959 JC 39. 7 8 9 ibid 43. 1987 JC 47. ibid 50. 10 J Chalmers and F Leverick, Criminal Defences and Pleas in Bar of Trial (2006), chapter 3. 11 This seems to have been borne out in Gillingham v HM Advocate The Scotsman, 24 April 2001. Here, a man who set an automated trap in a toolshed on his land in order to try and catch vandals was convicted of attempted murder when the trap injured an intruder. Although cf Hume, i, 220, 221, and the discussion in chapter 3 of Chalmers and Leverick, Criminal Defences. 12 Devlin v Armstrong [1971] NI 13, per MacDermott LCJ at 33; R v McInnes [1971] 1 13 WLR 1600, per Edmund Davies LJ at 1610. Criminal Law Act 1967 s 3. 14 Attorney-General for Northern Ireland’s Reference (No 1 of 1975) [1977] AC 105, per 15 Lord Diplock at 137. [2003] EWHC Admin 2567. 5

Killing to Protect Property

133

principle of the common law that a person may use a proportionate degree of force to defend himself, or others, from attack or the threat of imminent attack, or to defend his property or the property of others in the same circumstances’.16 If force has been used in defence of property, an English court should ‘consider the facts as the defendants honestly believed them to be, and should then determine objectively whether the force used was no more than was reasonable in all the circumstances’.17 As it is under the statutory provisions, this question is for the jury and is not a point of law.18 The issue of whether killing to protect property is permissible has never directly arisen in any reported English case.19 There is certainly no indication in existing case law that a judge could withdraw the defence on the basis that, as a matter of law, the use of lethal force could never be reasonable to protect property. This leaves open the possibility that a jury might acquit a defendant who had used lethal force in this context, although this is an unlikely outcome.20 The US Model Penal Code explicitly rules out the use of deadly force to protect property21 and US states appear, for the most part, to have followed this.22 Likewise, the use of deadly force to protect property is explicitly ruled out in at least one Australian state.23 By contrast, a jurisdiction that does permit killing to protect property is South Africa. This is clear from Ex p die Minister van Justisie: In re S Van Wyk.24 Van Wyk was a shopkeeper who had experienced a number of nighttime burglaries. He eventually resorted to setting up a shotgun that would be triggered by anyone coming into the shop outside of opening hours and put up a notice stating what he had done. The shotgun was triggered by a burglar, who was hit in the chest and killed. Van Wyk was charged with murder but acquitted on the basis of private defence. This resulted in a reference to the Appellate Court on the issue of whether or not a person can rely on the 16

17 ibid [21]. See also R v Martin [2001] EWCA Crim 2245, at [4]. Bayer, at [30]. R v Martin [2001] EWCA Crim 2245, at [7], following R v Beckford [1988] AC 130 and R v Owino [1996] 2 Cr App R 128. 19 It did not in Bayer. Here, the respondents were initially acquitted after using force to prevent the planting of genetically modified crops and thus, they argued, to protect the surrounding environment and property from damage. The Queen’s Bench Division of the High Court overturned the acquittal on the basis that they were not defending the property in question from an unlawful act. See, however, the obiter comments in R v Hussey (1924) 18 Cr App R 160 and R v Porritt [1961] 1 WLR 1372 discussed in Section 7.4 below. 20 And indeed, if this is a legal possibility, it may well violate Art 2 of the European Convention on Human Rights (see Chapter 10). 21 § 3.06(3)(d), but with an exception where the defendant faces an attempt to dispossess her of her dwelling (§3.06(3)(d)(i)) (discussed in Section 7.4). 22 Although, following the Model Penal Code, some state codes make an exception to this general principle where the property in question is the defendant’s home. 23 Crimes Act 1900 s 420 (New South Wales). 24 1967 (1) SA 488 (AD), extracted in J Burchell and J Milton, Cases and Materials on Criminal Law (2nd edn, 1997), at 151, and discussed in Stuart, ‘Killing in defence of property’. 18

134

Killing in Self-Defence

doctrine of private defence when he kills in order to protect his property. The court concluded in the affirmative and the majority of the judgment was concerned with the narrower issue of what would constitute adequate warning to a burglar in circumstances akin to the Van Wyk case. It may be, however, that if a case of this nature was now to arise for decision, the outcome would be different. In Ex p The Minister for Safety and Security and the National Commissioner of the South African Police Service: In re The State v Walters and Walters,25 it was hinted that allowing killing in defence of property might be unconstitutional.26 The case concerned the use of lethal force in effecting an arrest and the Constitutional Court concluded that killing a fleeing suspect who did not pose a threat of violence or was suspected on reasonable grounds of having committed a crime involving the infliction of serious bodily harm violated her right to life under the South African Constitution.27

7.3 Is Killing to Protect Property ever Justified? As a starting point, it should be said that it is accepted here that it is permissible to use some force in the defence of property. The issue in question is the degree of force that should be permitted and, in particular, whether there are any circumstances in which lethal force is acceptable. It should be said immediately that the view taken here is that there are not. Although killing in defence of property is not regarded as justifiable here, however, that is not to say it cannot be justified on an alternative theory of self-defence. If one accepts that the right to life is forfeited in circumstances other than when a person becomes an unjust threat to the life of another person, then killing to protect property is potentially justifiable. This is the way in which the permissibility of killing in defence of property was justified in South Africa by the Appellate Court. In Ex p die Minister van Justisie: In re S Van Wyk,28 the Lord Chief Justice took the approach that the person who attacks property infringes the rights of its owner and in doing so should accept that there is an accompanying risk to his own life: One who invades another’s rights, who defiantly ignores the prohibition, warning and resistance of the defender so that he can only be prevented by the most extreme measures can with good reasons be seen as the author of his own misfortune. It is he who is the outlaw, and if he is prepared to risk death by violating another’s rights, 25

(2002) 7 BCLR 663 (Constitutional Court of South Africa). ibid [53]. This was certainly the conclusion reached about the effect of Walters in J Burchell, Principles of Criminal Law (3rd edn, 2005), at 254. 27 For further discussion of the case, see D Bruce, ‘Killing and the constitution: arrest and the use of lethal force’ (2003) 3 South African Journal on Human Rights 430–54. 28 1967 (1) SA 488 (AD). 26

Killing to Protect Property

135

why should the defender, who is unquestionably entitled to protect his rights, be viewed as the one acting unlawfully if he uses deadly force rather than sacrifice his rights?29

This is effectively a version of the rights/forfeiture argument,30 where the right to life is forfeited by becoming a threat to property. Such an argument is comprehensively rejected here on the basis that human life is always worth more than property.31 Equally, if the use of defensive force is justified via a consequentialist approach,32 a different outcome in respect of the defence of property might result. This can be seen in German law, where there is only a very limited requirement for proportionality between the defended interest and the means of defence. Under the consequentialist approach taken by German law, on one side of the equation lies the value of the life of the attacker or thief. On the other side of the equation lie two distinct interests. The first is the value inherent in the life, bodily integrity, or property of the individual threatened. The second is the public interest in defending the legal order. Permitting defensive force signals to an attacker that she cannot violate the legal order without consequence. It is thought under German law that the value of these two interests together (the individual’s life/property and the protection of the legal order) outweighs the value of the life of the person who threatens these interests.33 In this way, killing in defence of property, at least in certain circumstances, is justified in German law. Earlier in this text,34 however, consequentialist justifications of the permissibility of selfdefensive force were rejected because they fail to pay adequate attention to issues of rights. For the same reason the German approach to the justification of lethal force in defence of property is not accepted here, on the basis that it ignores any notion that the taker of the property might have a right to life. The permissibility of killing in self-defence is grounded here in the notion that the right to life is possessed by all human beings and is forfeited only by virtue of becoming an unjust immediate threat to the life of another person. The person who threatens merely to deprive us of or destroy our property does not pose a threat to our life or any equivalent interest. If killing in defence of property were to be permitted, it would involve a violation of that person’s right to life and a value judgement that property (or at least some items of property) is worth more than human life. That value judgement is rejected here.35 29 Van Wyk, per Steyn CJ (translated) at 151 of Burchell and Milton, Cases and Materials on 30 Criminal Law. See Chapter 3. 31 And it may be that a South African court would now reject this argument too: see Section 32 7.2 above. See Chapter 3. 33 This explanation is drawn from TM Funk, ‘Justifying justifications’ (1999) 19 OJLS 631–47; and K Bernsmann, ‘Private self-defence and necessity in German penal law and in the penal law proposal: some remarks’ (1996) 30 Israel Law Review 171–87. 34 35 See Chapter 3. See also Hogan, ‘Defence of property’, at 466.

136

Killing in Self-Defence

This line of argument rests, however, on the fundamental assumption that human life is always worth more than property. Is it credible that there might be some types of property that are worth more than human life? Simester and Sullivan put forward this possibility: It is often said that it is never justified to kill in order to protect property. Yet it is not obviously a categoric truth that any individual’s life is of more intrinsic worth than any item of property. Is it necessarily unreasonable to kill a terrorist if such is the only means of preventing the destruction of priceless historic artefacts?36

Muth and Blumstein also see this as an argument that is ‘theoretically defensible’, on the basis that ‘in situations involving the irretrievable loss of property of great value, the property may be as important as life to the owner’.37 There are both practical and conceptual difficulties with this line of argument. In practical terms, it is difficult to envisage how a rule permitting lethal force based on the value of the property would operate. Deciding where to draw the line between property deemed worthy of protection and property not deemed so could be nothing other than entirely arbitrary. A rule based on financial value seems absurd. It could lead to a criminal conviction for someone killing to protect a painting worth £100 million but not for someone killing to prevent the destruction of a painting worth £101 million. Difficulties would also arise in obtaining agreement on valuation, especially given the social, cultural, and historical factors upon which the value of some items of property are contingent. A rule based on something other than objective financial value, such as the subjective value of the property to the owner, seems even more difficult to apply and potentially subject to abuse.38 More importantly, the ‘valuable property’ argument is rejected here on the basis that the value of human life outweighs that of any mere property, even the value of a terrorist’s life over Simester and Sullivan’s valuable and irreplaceable historic artefacts. As Chapter 3 argued, the deprivation of life is a loss that cannot be remedied or compensated for in any meaningful way. This cannot be said of items of property. It may be that the loss of a particularly valuable or treasured item of property can never be fully compensated in financial terms, but compensation is at least a meaningful concept in this context. There is little more that can be said in defence of this position other than arguing for it on the basis of the core values that one holds,39 however, and it is recognized here that if one genuinely believes in the value of certain 36

AP Simester and GR Sullivan, Criminal Law: Theory and Doctrine (2nd edn, 2004), at 626. Muth and Blumstein, ‘The use of deadly force’, at 1225. Cf M Kremnitzer and K Ghanayim, ‘Proportionality and the aggressor’s culpability in self-defense’ (2003–4) 39 Tulsa Law Review 875–99, at 894. 38 Although in South African law, which permits lethal force in defence of property, account is taken of the value the accused placed on the property defended and not merely its objective value (S v Mogohlwane 1982 (2) SA 587 (T)). 39 Simester and Sullivan, Criminal Law, at 626; Muth and Blumstein, ‘The use of deadly force’, at 1222–3. 37

Killing to Protect Property

137

types of property over human life, then it follows that killing in defence of such property might be permissible. This is not the view held here and even if it was, the formidable practical difficulties of formulating a rule to this effect remain.

7.4 Is there Something Different about Defence of Property in the Home? Those who reject the permissibility of killing to protect property as a general principle have sometimes suggested that it is different if the property in question is the home.40 Indeed, in a number of US states, an exception to the general rule that deadly force cannot be used to protect property is made where the defendant is threatened with the dispossession of her dwelling41 or in order to prevent burglary42 or arson.43 There is some suggestion that a similar principle once operated in English law, based on comments made in R v Hussey44 and R v Porritt.45 In Hussey, the appellant had rented a room but had been asked to leave by the landlady. When he refused to do so, the two complainants attempted to force their way into Hussey’s room with a view to evicting him. Hussey fired a gun and injured both of them. He was convicted of unlawful wounding. In quashing the conviction on appeal, the Lord Chief Justice quoted with approval the following passage from Archbold’s Criminal Pleading, Evidence and Practice: In defence of a man’s house, the owner or his family may kill a trespasser46 who would forcibly dispossess him of it, in the same manner as he might, by law, kill in self-defence a man who attacks him personally; with this distinction, however, that in defending his home he need not retreat, as in other cases of self-defence, for that would be giving up his house to his adversary.47

A similar line was taken in Porritt where, although it was not the central issue, it was assumed by the court that killing in defence of the home was 40 J Dressler, Understanding Criminal Law (3rd edn, 2001), at § 20.03; SP Green, ‘Castles and carjackers: proportionality and the use of deadly force in defence of dwellings and vehicles’ 41 (1999) University of Illinois Law Review 1–41. 18 Pa Cons Stat § 507(c)(4)(ii). 42 Alaska Stat § 11.81.350; Ark Code § 5-2-608(b)(2); NY UCC Law § 35.20(3). In Maine, deadly force is permitted where the defendant reasonably believes such force is necessary ‘to prevent or terminate a criminal trespass by such other person’ who she reasonably believes ‘has entered or is attempting to enter the dwelling place’ and ‘is committing or is likely to commit some other crime within the dwelling place’, although the defendant has a duty first to ask the intruder to terminate her trespass, unless the making of such a request would be dangerous to herself or another person (Me Rev Stat Ann tit 17-A § 104). 43 Alaska Stat § 11.81.350; Ark Code § 5-2-608(b)(2); Colo Rev Stat § 18-1-704.5; Me Rev 44 Stat Ann tit 17-A § 104; Or Rev Stat § 161.225(2)(b). (1924) 18 Cr App R 160. 45 [1961] 1 WLR 1372. 46 There is some doubt as to whether the victims in Hussey actually were trespassers. Nonetheless, this seems to be the basis upon which the case was decided. 47 Hussey, at 161.

138

Killing in Self-Defence

permissible. The passage in question comes from the judgment of Ashworth J, who stated that: At the trial it was conceded on behalf of the Crown that if the jury took the view that the firing was done in the honest belief that it was necessary for the protection of his stepfather, then the proper verdict was one of not guilty, and a similar concession was made in regard to the possibility of an honest belief that it was reasonably necessary to protect the house by shooting.48

Porritt and Hussey seem to suggest that English law did permit killing in defence of property where an attempt was made to dispossess the defendant of her home. Hussey was, however, decided over 75 years ago and commentators have doubted that it would apply today. Lanham, for example, comments that: ‘in an era when the sanctity of life takes precedence over the sanctity of possession, Hussey’s case makes strange reading’.49 Smith and Hogan suggest that ‘[e]ven if this were the law at the time, it would seem difficult now to contend that such conduct would be reasonable’.50 Getzler has pointed out that the Lord Chief-Justice was a ‘notoriously tough judge’ and that in later editions of Archbold the passage in question was removed.51 Is there any merit, then, in the argument that lethal force to protect the home should be permissible? At the outset, it is necessary to define precisely the issue that is being addressed. In particular, defence of the home should be distinguished from retreat from the home.52 The issue of defence of the home is one of proportionality: should a defendant be permitted to kill someone who, say, threatens permanently to dispossess her of her home? The issue of retreat is one of necessity: should a defendant be required to take a safe opportunity to escape rather than use force in self-defence if escape would require her to leave her home? In addition to this, defence of property is sometimes confused with whether or not it is permissible to use lethal force against an intruder, an issue that periodically attracts political and media attention,53 with a recent example being the campaign of a British newspaper to lobby for a change in the law in England in the belief that this was necessary to prevent householders from being prosecuted for protecting themselves or their families from violent 48

49 Porritt, at 1375, emphasis added. Lanham, ‘Defence of property’, at 372. D Ormerod, Smith and Hogan Criminal Law (11th edn, 2005), at 340. This conclusion does seem to have been borne out in R v Hastings [2003] EWCA Crim 3730. Here, the appellant was convicted of manslaughter after disturbing a housebreaker. It was dark and the appellant thought that the deceased was armed so he used a knife and stabbed him twelve times. 51 J Getzler, ‘Use of force in protecting property’ (2005) 7 Theoretical Inquiries in Law (on-line edition). 52 Considered in Chapter 4. For an example of confusion between the two concepts see the passage in State v Patterson 45 Vt 308 (1873) commencing ‘The idea that is embodied in the expression that, a man’s house is his castle . . .’ (at 9), where the Supreme Court of Vermont begins with a statement about defence of the home but, by the end of the passage, is referring to retreat from the home. 53 For comment on the way the media has dealt with the issue, see Jefferson, ‘Householders and the use of force against intruders’. 50

Killing to Protect Property

139

intruders.54 This led the Crown Prosecution Service in England to issue guidelines reassuring householders that they are extremely unlikely to be prosecuted for using even lethal force against an intruder, unless they act in malice or revenge.55 The campaign was always misplaced, as it confused the issue of the use of lethal force where the householder believed that the safety of herself or her family was threatened and where the householder was acting solely to protect her property. The former has always been permitted in English law.56 It is the latter that is of concern here. Finally, it is important to define exactly what is meant by the property in question being the home. A distinction can be made between someone who forcibly enters the accused’s home in order to steal the property contained within the home and someone who threatens to dispossess the accused of her home itself. The first of these is the more common scenario. When someone breaks into a dwelling, it is most likely because she wishes to steal the property contained within, not because she wishes to take possession of the dwelling itself. There are, perhaps, two possible lines of argument for permitting the defence where lethal force has been used in these circumstances.57 The first is that there is something about a threat of theft that takes place in the home that is intrinsically more dangerous than a threat of theft outside the home. This argument can be seen in the Commentary to the US Model Penal Code, which justifies the defence of the home exception to the rule prohibiting lethal force in defence of property by stating that ‘any persistent effort to break into a home is likely to arouse in the householder a reasonable fear for his own safety, unless he knows that the assailant is breaking in under a claim of right’.58 A similar argument can be found in State v Barr:59 ‘It is common knowledge that burglaries under such circumstances often result in the death 54 The Daily Telegraph, ‘The right to fight back’, 12 December 2004. See also the private members bills that were drafted in the wake of the Tony Martin case: the Criminal Law (Amendment) (Householder Protection) Bill 2004; the Criminal Justice (Justifiable Conduct) Bill 2004. Neither was successful. On the same theme, see the proposals in Italy to create an irrebuttable presumption that force used against an intruder in the home is legitimate selfdefence, discussed by S Skinner, ‘Populist politics and shooting burglars: comparative comments on the Lega Nord’s proposal to reform Italian self-defence law’ [2005] Crim LR 275–84. 55 Householders and the Use of Force against Intruders, Joint Public Statement from the Crown Prosecution Service and the Association of Chief Police Officers, issued 1 February 2005. The guidance is available at ⬍www.cps.gov.uk/publications/prosecution/householders.html⬎ (accessed on 12 April 2006). For critical comment on the CPS guidance, see Jefferson, ‘Householders and the use of force against intruders’, at 405–6. 56 As the Law Commission suggested in its Report on Partial Defences to Murder (Law Com No 290, 2004), public fear about defending against intruders is almost certainly based on a misunderstanding of the law. In fact, as the law in England permits the defendant to act on the basis of an honest but unreasonable belief that force was reasonably necessary, the leeway allowed to householders in England is greater than in Scotland, where only a reasonable belief in an imminent attack will suffice (see Chapter 9). 57 These are drawn, with some modification, from Green, ‘Castles and carjackers’. 58 Commentary to the Model Penal Code, at 93. 59 39 P 1080 (1895) (Supreme Court of Washington).

140

Killing in Self-Defence

of some of the inmates of the dwelling upon which the burglary is committed and for that reason it might well be held that a burglary of that kind could rightfully be prevented by such means as might result in death.’60 But it is important to unpack exactly what is being argued here. The argument seems to be that because break-ins to the home are sometimes accompanied by a threat of death or serious injury, this threat should be presumed whenever an intruder unlawfully enters the home. Indeed, this is precisely the presumption made in the criminal codes of some US states.61 This presumption, however, demonstrates conceptual confusion between the threat to property and any additional and accompanying threat to life. Simply because some housebreakers threaten the life of the inhabitants, it does not follow that all housebreakers should be assumed to carry such a threat.62 This is not even an obviously reasonable presumption. The primary concern of the majority of housebreakers might equally be to escape with the householder’s property without having any contact with the occupants of the house.63 Where there are reasonable grounds for a belief that a housebreaker poses a threat to the life of the occupants,64 lethal force should be permitted on the basis of that threat to life with, of course, some allowance being made for reasonable mistake on the part of the householder in the heat of the moment. As Green states, ‘having one’s home invaded by an unknown intruder, particularly at night, is likely to be a terrifying experience for most homeowners. One cannot reasonably expect occupants to interview intruders about their intentions before using defensive force.’65 But to permit lethal force in defence of property contained within the home where there are no reasonable grounds for a belief that the housebreaker threatens the life of the occupants is to show a lack of respect for the life of the housebreaker and should be rejected on this basis.66 A second possible line of argument is that the thief who attempts to steal property from the home does more than simply threaten to deprive us of that property. In addition, by entering our home without permission, she violates our right to privacy. It is this additional violation that grounds the permissibility of killing a housebreaker. This line of argument is used by Dressler: The home is a . . . source of privacy where the most intimate activities in life are conducted, and from which people seek to exclude the prying eyes and ears of 60

39 P 1082 (1895) (Supreme Court of Washington). eg Tenn Code Ann § 39-11-611(b). Florida goes further and presumes a threat of death or great bodily injury where an intruder attempts to enter an occupied vehicle: Fla Stat § 776.013(1)(a). 62 This conceptual confusion can be seen in State v Patterson 45 Vt 308 (1873) (Supreme Court of Vermont), at 316. 63 In England and Wales in 2004/5, 321,459 domestic burglaries were recorded by the police. In only 13 per cent of these was violence used or threatened. See S Nicholas, D Povey, A Walker, and C Kershaw, Crime in England and Wales 2004–05 (2005), at section 4.2 and supplementary 64 table 1.08. The need for such a belief to be reasonable is defended in Chapter 9. 65 Green, ‘Castles and carjackers’, at 29. 66 Cf David Burnside, the former Ulster Unionist MP, quoted in the Sunday Telegraph 24 October 2004: ‘Once a burglar enters someone’s home with the intent of robbing them or 61

Killing to Protect Property

141

strangers and of the government . . . When a wrongdoer seeks to enter a person’s dwelling, therefore, more than property is invaded. In common law terms, the fortress has been attacked; a person’s primary source of safe and private habitation has been jeopardised.67

This line of argument is comprehensively rejected here. To accept it would be to accept the proposition that the right to life is forfeited by an invasion of another person’s privacy. This conclusion would contradict the central principle of this text: that the right to life is possessed by all human beings and is forfeited only by becoming an unjust immediate threat to the life (or equivalent interest) of another. It might also open up the permissibility of killing in other situations in which privacy is seriously invaded, such as exceptionally intrusive photography or journalism.68 Let us turn, then, to whether there are any arguments that support the killing of someone who threatens to dispossess us of our home itself (as opposed to the property within it). Certainly the drafters of the US Model Penal Code seem to think that there is something special about killing in defence of the home, as the Code contains an exception to the general rule that lethal force may not be used in defence of property ‘where the person against whom the force is used is attempting to dispossess [the accused] of his dwelling otherwise than under a claim of right to its possession’.69 There are two arguments that might be put forward for this position.70 The first is that one’s home is such an expensive and significant possession that its value outweighs the value of human life and lethal force should therefore be permitted in its defence. The ‘valuable property’ argument has already been considered and rejected71 and the fact that the valuable property in question is one’s home does not seem a good reason in itself to change the conclusion reached that human life is always of more intrinsic value than mere property.72

hurting them, they should lose their rights to protection from the law’ (the quote is taken from Jefferson, ‘Householders and the use of force against intruders’, at 412 n 11). 67

Dressler, Understanding Criminal Law, at 238–9. Although, against this, it might be argued that no intrusion of privacy other than one that breaches the sanctity of the home is serious enough to ground the permissibility of killing the invader. 69 Model Penal Code § 3.06(3)(d). The state of Louisiana has extended this concept to cover attempts to dispossess a person of her motor vehicle (La Rev Stat § 14:20(4)(a)). Here, only the arguments relating to the home will be considered, because if killing in defence of the home is rejected, it is assumed that killing in defence of a motor vehicle is likewise. For criticism of the Louisiana provisions, see SM Gerling, ‘Louisiana’s new “kill the carjacker” statute: self-defense or instant injustice?’ (1999) 55 Washington University Journal of Urban and Contemporary Law 109–34. 70 Once again, these are drawn (with some modification) from Green, ‘Castles and carjackers’. 71 See Section 7.3 above. 72 If this line of argument were to be accepted (which it is not here), it would seem anyway to rule out killing to protect homes that have only a very low financial value or where the occupant has no property rights in the home. 68

142

Killing in Self-Defence

A second possible argument is that there are certain types of property that are essential in sustaining life and the home is one of these. It is conceded that there are certain types of property that are essential to the preservation of life. One might take the example of insulin to a diabetic or a life-support machine to a person in a coma. If an individual threatens unjustly to remove genuinely life-sustaining property, this is not just a threat to property, but is also a threat to life and it follows that killing in self-defence should be permitted, providing the necessity requirements of the defence are made out. It is unlikely, however, that this would ever be the case where the property in question is the home. If an individual’s home is taken away, in all but the most exceptional circumstances this would not constitute a threat to life. That individual would generally be able to find somewhere else to shelter and even if she could not (say her property was uninsured and destroyed by fire, she had no money and no one else would take her in), she would be unlikely to die simply because she does not have a roof over her head. Extreme hypothetical scenarios can be constructed where the loss of one’s home would genuinely threaten one’s life (say, the isolated cabin in the wilderness in the middle of a blizzard), but otherwise, the ‘home as essential to the preservation of life’ argument does not hold as a justification of killing in defence of the home. What is perhaps true is that a home is necessary for a certain quality of life. Without doubt, if an individual’s property is burnt down by arsonists, leaving her temporarily homeless, that person’s life, at least for a time, will be disrupted, uncomfortable, and stressful. But accepting that it is permissible to kill in defence of a certain quality of life is a slippery slope. It could, for example, open the way to permitting the accused to kill her employer if this was the only way to prevent the employer from unjustifiably dismissing her from her job. In conclusion, then, the argument that it is permissible to kill in defence of property is rejected here. The value of human life is thought always to outweigh the value of property. This is the case even where the property in question is one’s own home. Granted, some threats to property will carry with them an accompanying threat to life. One might think here of the hostage-taker who deprives her diabetic hostage of insulin. Other threats to property might lead the accused reasonably to believe that her life is threatened. One might think here of the night-time housebreaker who carries a torch that could reasonably be mistaken for a gun. The use of lethal force should be permissible to defend against both of these types of threat but this is on the basis of the actual or reasonably perceived threat to life, not on the basis of the threat to property.

8 Killing to Prevent Rape 8.1 Introduction This chapter considers the question of whether killing to prevent rape is permissible. On the face of it, the issue is uncontroversial. Those colleagues to whom I put the question in the early stages of writing generally answered without hesitation in the affirmative. Indeed, this view can be seen in the leading US case, People v Landrum,1 where Riley CJ began the majority judgement by stating: ‘Only an archaic system of justice would suggest that a woman cannot use deadly force to defend herself against common law rape’.2 It will be demonstrated here, however, that this conclusion is not as obvious as it initially seems, especially if the threat of rape is separated out from any accompanying threat to life. This chapter will propose that the criminal law should provide a defence to the accused who has killed to prevent rape. The reason for this is not because of the physical or psychological effects of rape. Rather it is because rape is a denial of humanity and thus approaches the same level of seriousness as a threat to the accused’s life.

8.2 Killing to Prevent Rape in the Law It is generally accepted that the accused will be able to plead self-defence if she kills to prevent rape. The early English writers were clear on this point, with, for example, both Foster3 and Hawkins4 viewing killing to prevent rape as a case of justifiable self-defence. In modern times, the issue has rarely arisen but there seems to be no reason to doubt that this position is still correct. In R v Wheeler5 it seems to have been assumed, without any discussion, that killing to prevent rape is permissible. Here, the defendant was charged with murder but claimed that he was acting to prevent the violent rape of his 1

456 NW 2d 10 (1990) (Supreme Court of Michigan). ibid 22. The central issue in Landrum was not whether killing to prevent rape is permissible but whether the defendant was entitled to a specific jury instruction to this effect. The majority held that she was not. 3 M Foster, A Report of some Proceedings on the Commission of Oyer and Terminer and Goal (sic) Delivery for the Trial of the Rebels in the Year 1746 in the County of Surry, and of other Crown Cases to which are added Discourses upon a few Branches of the Crown Law (1762), at 274. 4 Hawkins, A Treatise of the Pleas of the Crown (3rd edn, 1739), at 71. 5 [1967] 1 WLR 1531. Note that the All England Law Reports version of the case report ([1967] 3 All ER 829) leaves out the passages of the judgment relevant to self-defence and rape. 2

144

Killing in Self-Defence

girlfriend. He was eventually convicted of manslaughter, his claim not being accepted, but in the course of his charge to the jury, the trial judge stated that: ‘If you are satisfied that he had no such intention at all but nevertheless used unlawful violence which resulted in death, then he is guilty of manslaughter. If the attack on [the deceased] was justified and was no more than was reasonably necessary to prevent rape then it would not be an unlawful attack.’6 This passage was approved on appeal as ‘a clear, accurate, precise and helpful definition of the issues’.7 There is also the case of R v Clugstone,8 where the defendant was charged with murder after killing a man who was attempting to rape her. The trial judge stated that there was no evidence to contradict the defendant’s assertion that she was acting to prevent the rape and directed the jury to return a verdict of not guilty. Recent English case law has tended not to specify the nature of the threat(s) which justify the use of lethal defensive force, but to leave the issue to the jury as part of a general test of whether any force used was reasonable. In R v Martin,9 for example, the Court of Appeal sets out the law relating to self-defence as follows: A defendant is entitled to use reasonable force to protect himself, others for whom he is responsible and his property . . . In judging whether the defendant had only used reasonable force, the jury has to take into account all the circumstances, including the situation as the defendant honestly believes it to be at the time, when he was defending himself.10

This leaves open the question of whether lethal force used to prevent rape is permissible. It will depend on whether the jury in question view killing as a reasonable response to a threat of rape, potentially leaving defendants subject to inconsistency of treatment. In Scots law, the position is more clear-cut. The proportionality requirement is satisfied where a female person uses lethal force to prevent a threat of rape.11 This is the conclusion of Hume, who states that: ‘In like manner as a man may kill in resistance of an attempt on his life, so may a woman in resistance of an attempt to commit a rape on her person, an attempt at which she is entitled to feel the highest indignation and resentment’.12 Hume’s account is gender specific: the person threatened with rape must be female, although the prevention of the attack can be undertaken by either the victim or another person.13 As Hume refers specifically to rape, which in 6

7 [1967] 1 WLR 1534, emphasis added. ibid 1534. The Times, 1 October 1987, and also noted by JC Smith, Justification and Excuse in the 9 Criminal Law (1989), at 109. [2001] EWCA Crim 2245. 10 ibid [4]–[5]. 11 Even where she faces no accompanying threat of death or serious bodily injury. 12 D Hume, Commentaries on the Law of Scotland Respecting Crimes (4th edn, 1844) i, 218. 13 In the first edition of Hume’s Commentaries, published in 1797, only the victim herself or the husband, father, or brother of the victim were permitted to kill to prevent a rape. By the time 8

Killing to Prevent Rape

145

Scotland is a gender-specific crime,14 this is not surprising. But it seems that it would not be permissible to kill in order to prevent an equivalent sexual assault on a male person under Hume’s account. This is because of the very serious injury rape is assumed to cause to the honour of (specifically) a woman.15 Presumably Hume does not consider non-consensual sexual intercourse to carry the same degree of harm to a male person.16 The leading (relatively) modern authority is McCluskey v HM Advocate.17 McCluskey concerned a claim by the (male) accused that he killed in order to prevent a sexual assault by another man. He was convicted of culpable homicide after the trial judge directed the jury that homicide was justified by self-defence only ‘if the homicidal acts are done to save the man’s own life’.18 This was confirmed on appeal.19 In the course of his judgment, the Lord Justice-General (Clyde) contrasted this with the situation of a woman faced with a threat of rape, who would be permitted to kill in her defence. The US Model Penal Code specifically provides for a defence of killing to prevent rape, stating that the use of deadly force is not justifiable ‘unless the actor believes that such force is necessary to protect himself against death, serious bodily injury, kidnapping or sexual intercourse compelled by force or threat’.20 The penal codes of a number of US States also specifically provide for a defence where deadly force has been used to prevent rape,21 a forcible sex offence,22 or sexual assault.23 of the fourth edition in 1844, this permission had been widened to include ‘anyone who is with her’. 14 At the time of writing, rape can be committed only by a male perpetrator on a female victim. The definition of rape was considered in the Lord Advocate’s Reference (No 1 of 2001) 2002 SLT 466. While the case was noteworthy in that it redefined the actus reus of rape as sexual intercourse with a female person without her consent (a change from the previously accepted definition of sexual intercourse with a female person by overcoming her will), the gender-specific nature of the crime was not addressed, leaving Scots law out of line with almost every other common law jurisdiction. The Scottish Law Commission’s Discussion Paper on Rape and other Sexual Offences (DP 131, 2006) recommends the adoption of a gender neutral definition, at 15 See Section 8.3.2 below. para 4.18. 16 On this, see Burnett: ‘[A] woman may kill with impunity in hindrance of a rape; but the same will not hold in other injuries, though of the grossest kind; a man will not be justified in killing another to prevent an indignity to his person, as if one should wantonly invade him, to horsewhip, or to pull his nose; though his killing the invader in such circumstances would not infer the guilt of murder. To extenuate, however, in any great degree, the injury must be real, and of the highest class’ (at 53, emphasis in original). 17 1959 JC 39. Although see also HM Advocate v Forrest (1837) 1 Swin 404 and Crawford v 18 HM Advocate 1950 JC 67. ibid 41. 19 It will be interesting to see whether this remains the case if the Scottish Law Commission’s recommendation in its Discussion Paper on Rape and other Sexual Offences that the definition 20 of rape is widened to include male victims becomes law (para 4.18). § 3.04(2)(b). 21 Ala Code § 13A-3-23(a)(2); Del Code Ann tit 11 § 464(c); Mo Rev Stat § 563.031(2)(2); Neb Rev Stat § 28-1409(4); NY UCC Law § 35.15(2)(b); 18 Pa Cons Stat § 505(b)(2). See also 22 9 Guam Code Ann § 7.86(b). NH Rev Stat Ann § 627.4(II)(c). 23 Alaska Stat § 11.81.335(a)(2); Colo Rev Stat § 18-1-704(2)(c); Me Rev Stat Ann tit 17-A § 108(2)(A)(2); Tex Penal Code § 9.32(a)(3)(b); Vt Stat Ann tit 13 § 2305(2). Others permit lethal force to prevent a felony involving force or violence, which would include rape: Ark Code § 5-2-607(a)(1); Fla Stat § 776.012(1); Ga Code Ann § 16-3-21(a); 720 ICLS

146

Killing in Self-Defence

The position in Canada is complicated by the fact that no specific crime of rape is recognized. Instead, there is a single offence termed ‘sexual assault’ that is intended to cover all assaults committed in circumstances of a sexual nature.24 It is not entirely clear whether the accused is permitted to use lethal force to prevent sexual assault. Under s 34(2)(a) of the Criminal Code, a reasonable apprehension of death or grievous bodily harm is required if lethal force is used in self-defence, a standard reiterated in the leading case on self-defence from the Canadian Supreme Court, R v Pétel.25 Given that s 272(1) of the Criminal Code makes specific provision for a sexual assault that ‘causes bodily harm to the complainant’ and s 273(1) makes specific provision for aggravated sexual assault (sexual assault that ‘wounds, maims, disfigures or endangers the life of the complainant’), it might be concluded that lethal force cannot be used to prevent a sexual assault that does neither of these. Alternatively, it may be that a sexual assault involving penetration without consent (what in other jurisdictions would be termed rape) would always be regarded as causing grievous bodily harm, but this has not been explicitly stated in the case law. Some assistance might be gained from R v McCraw,26 where the Canadian Supreme Court equated a threat to commit rape to a threat of serious bodily harm.27 This decision was made in a different context,28 though, so its value as an authority is limited. There is no clear authority in Australia either, although it was implied by the High Court in Zecevic v DPP29 that the use of deadly force to prevent rape might be permissible.30 One might also point to an obiter remark in R v Lane,31 where Lush J stated that ‘[s]elf-defence against rape must, I think, be capable in appropriate circumstances of justifying homicide’.32 As this was not the central issue in the case, however, its value as an authority is limited.

8.3 The Principled Argument 8.3.1 Some Preliminary Points The permissibility of killing to prevent rape has received very little attention, perhaps because it is simply assumed that it is easily justifiable. On the rare occasion where the issue is mentioned at all, it tends to be touched § 5/7-1(a); Mont Code Ann § 45-3-102; ND Cent Code § 12.1-05-07(2)(b); Or Rev Stat § 161.219(1); Utah Code § 76-2-402(1) (forcible felony specifically defined in (4) as including rape). 24 Canadian Criminal Code ss 271–3. The change took place in 1983, prior to which the 25 specific offence of rape was recognized. [1994] 1 SCR 3. 26 27 [1991] 3 SCR 72. ibid [34]. 28 The accused had threatened to rape the complainant and the question was whether this constituted a threat to cause serious bodily harm contrary to s 264.1(1)(a) of the Canadian 29 Criminal Code. (1987) 162 CLR 645. 30 31 32 Brennan J at 667. (1983) 8 A Crim R 182. ibid 183.

Killing to Prevent Rape

147

upon without further analysis or elaboration.33 One exception is a paper by Kates and Engberg,34 in which the authors argue for the permissibility of killing to prevent rape and their arguments will be considered subsequently. In the leading UK criminal law texts, there is a notable absence of discussion of the issue. Among the English texts, neither Ashworth35 nor Simester and Sullivan36 make any specific reference to rape at all when they discuss the proportionality requirement in self-defence. Smith and Hogan state only that ‘[t]he whole question is somewhat speculative’ and then frame the question ‘[i]s it reasonable to kill or cause serious bodily harm in order to prevent rape?’ without attempting to address it.37 Among the Scottish texts, Gordon simply states that rape is a ‘unique exception’38 to the rule that lethal force can be used only where there is danger to life, without considering why this might be so. Jones and Christie note that ‘the one exception to the danger to life requirement for killing in self-defence is in the case of rape’39 and question the ‘appropriateness of [such] an exemption from the normal rules of self-defence’40 but do not consider the issue further. Among the leading US texts, the issue is similarly neglected. Robinson does not consider the question at all.41 LaFave mentions the issue only in a footnote, noting briefly that some US penal codes permit the use of deadly force to prevent kidnapping and rape.42 Should, then, the criminal law permit killing to prevent rape? At the outset, some preliminary points need to be addressed. The first of these is what is meant here by rape. The definition of the crime of rape varies 33 GP Fletcher, ‘Defensive force as an act of rescue’, in EF Paul, FD Miller Jr, and J Paul (eds), Crime, Culpability and Remedy (1990) 170–9, at 172; DN Husak, ‘The complete guide to self-defense’ (1996) 15 Law and Philosophy 399–406, at 400; T Kasachkoff, ‘Killing in self-defence: an unquestionable or problematic defence?’ (1998) 17 Law and Philosophy 509–31, at 511–12; SH Kadish, ‘Respect for life and regard for rights in criminal law’ (1976) 64 California Law Review 871–901, at 888; M Kremnitzer and K Ghanayim, ‘Proportionality and the aggressor’s culpability in self-defense’ (2003–4) 39 Tulsa Law Review 875–99, at 893. 34 DB Kates Jr and NJ Engberg, ‘Deadly force self-defense against rape’ (1982) 15 University of California Davis Law Review 873–906. Subsequently Kates and Engberg. 35 A Ashworth, Principles of Criminal Law (5th edn, 2006). 36 AP Simester and GR Sullivan, Criminal Law: Theory and Doctrine (2nd edn, 2004). 37 D Ormerod, Smith and Hogan Criminal Law (11th edn, 2005), at 334. 38 GH Gordon, The Criminal Law of Scotland (3rd edn, 2000–1), at para 24.18. 39 TH Jones and MGA Christie, Criminal Law (3rd edn, 2003), at para 8–35. 40 ibid, para 8–35. For a more detailed discussion in the Scots law context, see J Chalmers and F Leverick, Criminal Defences and Pleas in Bar of Trial (2006), chapter 3. 41 PH Robinson, Criminal Law Defenses (1984). On the issue of proportionality in self-defence, Robinson states only that ‘[w]here a person’s life or health is threatened, proportionality is less often a limitation [than it is in cases of defence of property]’ (at § 131(d)). 42 WR LaFave, Criminal Law (3rd edn, 2000). The text of the footnote states: ‘Some of these provisions extend to the most extreme intrusions on freedom of the person (eg kidnapping and rape). Perhaps such provisions may be properly classified as primarily serving defence of the person, even though they apply when the crime does not threaten death or serious bodily harm’ (at 493 n 15). LaFave does not explain either why killing might be justified in prevention of intrusions on freedom or why rape is seen as an example of such an intrusion.

148

Killing in Self-Defence

between jurisdictions.43 Indeed, as we have already seen, Canadian criminal law does not have a crime of rape at all, but instead classifies it as part of a broader offence of sexual assault.44 In this chapter, the term rape will be used as shorthand for sexual intercourse without consent with either a male or female person45 or a sexual assault of equivalent seriousness on either a male or female person. It is not the intention to engage in debate over the way in which rape, or indeed any other sexual offence, should be defined. Rather, it is to question the assumption that killing to prevent ‘rape’ or an equivalently serious sexual assault is permissible. It is recognized here that it is extremely difficult to draw a clear distinction between ‘rape’ and other sexual assaults,46 or indeed between ‘more’ and ‘less’ serious types of sexual assault. At best, all that can be done is to place sexual offences on a scale ranging from the most serious to the least serious.47 What is clear, however, is that if the assumption can be questioned in relation to rape and the types of sexual assault placed at the top of that scale, then it can also be questioned in relation to less serious sexual assaults. Second, it should be stressed that what is under consideration is the issue of proportionality in self-defence. It will be assumed that the necessity requirement of the defence is satisfied: that is, there is no reasonable means of escape from the rape and the attack could not be prevented other than by killing the attacker. Thus the question that we are really attempting to consider is not whether the criminal law should permit killing to prevent rape, but whether the criminal law should permit killing to prevent rape, where this is the only way in which the rape can be prevented. One’s instinctive reaction to the question when it is phrased in this way is to say yes. But that brings us on to a third preliminary point: the separation of the threat of rape from any accompanying threat of death. It may be that one reason why people instinctively think that killing is permissible in this context is that the threat of rape is confused with an accompanying threat to life. Admittedly in practice the two things might be difficult to separate. When faced with a threat of rape, it would often be reasonable for the victim to believe that the attacker also poses a threat to life. If this is the case, it is not in any way disputed that she should be permitted to kill to repel that 43 In Scots law, the actus reus of rape is sexual intercourse with a woman without her consent (Lord Advocate’s Reference (No 1 of 2001) 2002 SLT 466, emphasis added). In English law, ‘[a] person (A) commits [rape] if (a) he intentionally penetrates the vagina, anus or mouth of another person (B) with his penis, (b) B does not consent to the penetration, and (c) A does not reasonably believe that B consents.’ (Sexual Offences Act 2003 s 1, emphasis added). 44 Canadian Criminal Code, ss 271–3. 45 It is not accepted here that the wrong of rape is any different according to whether the victim is male or female. The only exception to this is the risk of pregnancy for a female person. This will be considered in more detail in due course. 46 Cf Garvock v HM Advocate 1991 SCCR 593, where the court held that the fact that the Crown could not prove that the victim had been penetrated by a penis as opposed to a blunt instrument meant that her attacker’s sentence should be reduced from ten to eight years. 47 And even here, there is likely to be disagreement.

Killing to Prevent Rape

149

perceived threat. The permissibility of killing in defence of life (or where the accused has a reasonable belief48 that she is killing in defence of her life) is not disputed here.49 Likewise it is accepted that, when assessing whether or not the victim’s belief is a reasonable one, allowance should be made for mistakes made in the heat of the moment.50 The concern of this chapter is whether the victim should be permitted to kill to prevent a threat of rape alone, where there are no reasonable grounds for a belief that the attacker poses a threat to life. At this point, the merit of this line of enquiry might be questioned because such a situation is unlikely to arise in reality. In the circumstances of many attempted rapes, it is likely to be entirely reasonable for the victim to fear for her life. But scenarios can be envisaged where it would not be reasonable to fear death, perhaps where the attacker is well known to the victim, carries no weapon and has never demonstrated any tendency towards violence in the past. In a survey of female rape victims in the USA, just over half of the respondents reported that they thought they might be killed or seriously injured during the rape.51 While this constitutes the majority of respondents, it does mean that just under half of the respondents did not fear death or serious injury during the attack. Aside from the frequency with which a situation of this nature is likely to arise, it is worth considering the issue of killing to prevent rape (in the absence of a threat to life), if only to clarify the criminal law. As we have seen, the criminal law in the majority of common law jurisdictions either assumes or states explicitly that killing to prevent rape is permissible. If, in fact, there is no convincing argument in support of this proposition and what is actually permissible is killing where there are reasonable grounds for perceiving a threat to life, for the sake of conceptual clarity in the law, the ‘killing to prevent rape’ provision should be removed. Fourth, and finally, it should be stressed that if it is concluded that it is not permissible to kill to prevent rape, this conclusion would not necessarily leave the accused facing a murder conviction. Killing to prevent rape would generally be treated as manslaughter rather than murder and in appropriate circumstances the accused might receive no punishment at all, other than the fact of the conviction. What is being considered here is whether the accused should receive the benefit of the complete acquittal that the defence of self-defence would provide. 48 It will be argued later that only a reasonable mistaken belief should ground an acquittal on 49 the basis of self-defence (see Chapter 9). See Chapter 3. 50 See Chapter 9. 51 Of a sample of 464 women who had been raped, 54 per cent reported that they thought they might be killed or seriously injured during the incident. See R Acierno, M Gray, C Best, H Resnick, D Kilpatrick, B Saunders, and K Brady, ‘Rape and physical violence: comparison of assault characteristics in older and younger adults in the National Women’s Study’ (2001) 14 Journal of Traumatic Stress 685–95, at 690.

150

Killing in Self-Defence

8.3.2 Should the law provide a defence to the accused who has killed to prevent rape? Having disposed of these preliminary points, we can move on to consider the question of whether the criminal law should provide a defence where the accused has killed to prevent rape, where there was no accompanying threat to life (or reasonable grounds for perceiving one) and where there was no other way in which the rape could have been prevented. Historically the justification for such a defence was grounded in the particularly serious injury that rape caused to (specifically) a woman’s personal honour. For example, in setting out the defence in Scots law, Hume describes rape as a ‘cruel and irreparable injury’52 and ‘a robbery of that in which a woman’s honour, her place in society, and her estimation in her own eyes depend’.53 Foster, writing at around the same time54 in the English context, uses similar language: The Injury intended can never be Repaired or Forgotten. And Nature to render the Sex amiable hath implanted in the Female Heart a quick Sense of Honour, the Pride of Virtue, which kindleth and enflameth at every such Instance of Brutal Lust. Here the Law of Self-Defence plainly coincideth with the dictates of Nature.55

But the modern law surely cannot sanction killing to prevent an affront to personal honour or dignity?56 An alternative rationale must be found if the defence is to be justified in modern times. A helpful starting point is to consider the arguments made by Kates and Engberg,57 who conclude that killing in such circumstances is permissible. They draw upon two arguments in support of their conclusion. The first is that rapists so often kill their victims that it is entirely reasonable to assume that every threat of rape carries with it the additional threat of death.58 This is not a convincing argument for two reasons. First, the assertion made by Kates and Engberg that rapists ‘often kill their victims’ is questionable. It is questionable even on the figures provided by Kates and Engberg themselves. When they examine US empirical studies they find that between 1 and 3 per cent of rapists kill their victims. This is hardly a basis upon which it can be assumed that every threat of rape carries with it the threat of death.59 52

53 Hume, Commentaries on the Law of Scotland (4th edn) i, 218. ibid 301. Foster was writing in 1762. The first edition of Hume was published in 1797, although the same explanation can still be found in the fourth edition, published in 1844. 55 Foster, Report, at 274, emphasis in original. 56 Indeed, in the leading Scottish case, McCluskey v HM Advocate 1959 JC 39, the Appeal Court made it quite clear that this would not be permitted (at 43). The Court did not, however, provide a convincing alternative justification for the defence: see Chalmers and Leverick, 57 Criminal Defences, chapter 3. Kates and Engberg. 58 ibid 885. See also RJ Muth and AL Blumstein, ‘The use of deadly force in the protection of property under the Model Penal Code’ (1959) 59 Columbia Law Review 1212–32, at 1222. 59 And as this chapter has already noted, there is evidence to suggest that a significant proportion of rape victims do not perceive their life to be under threat during the rape. See Acierno et al, ‘Rape and physical violence’, at 690. 54

Killing to Prevent Rape

151

Perhaps more importantly, the second reason that this is not a convincing argument is that even if any or most (or whatever the relevant proportion is) of rapes do carry with them a threat to life, this is not in itself a good reason to permit killing to prevent all rapes. Granted, it will in some circumstances be difficult to distinguish between a threat of rape alone and an accompanying threat to life and where there is some reasonable basis for the victim to think that her life is in danger from the attacker, she should be permitted to use lethal force. However, the fact that some rapes carry with them an accompanying threat of death simply does not work as a justification of killing to prevent rape in cases where there is no threat of death or where there are no reasonable grounds for perceiving such a threat.60 This brings us on to Kates and Engberg’s second argument: that the threat of rape is equivalent in seriousness to a threat of serious bodily harm and because we are generally permitted to kill to prevent serious bodily harm, we should also be permitted to kill to prevent rape. Kates and Engberg support their argument by setting out the various harmful effects of rape reported by rape victims, all of which are, undoubtedly, serious harms, and we will return to these later.61 A major assumption is made by Kates and Engberg though: that it is always permissible to kill to prevent serious bodily harm. That the criminal law should provide a defence to those who have killed to prevent ‘serious bodily harm’ is a conclusion (like that of killing to prevent rape) that seems to have been accepted without question. But why should this be the case? One answer would be that, as in the case of rape, it is often difficult to distinguish between a threat of serious bodily harm and a threat of death. This is undoubtedly true, especially where an attack involves a weapon. But if this is the reason for including a provision that permits killing to prevent serious bodily harm, it is redundant. This possibility could be covered simply by stating that killing is permitted in defence against a threat to life or where there are reasonable grounds for perceiving such a threat, which leaves open the question of killing to prevent serious bodily harm where there is no reasonable basis to think that there is a threat to life. When might such a threat arise? Thinking up hypothetical possibilities is a somewhat speculative exercise but it is not impossible to imagine, say, the situation where a persistent car thief is faced with the threat of ‘kneecapping’ to dissuade her from further criminal activity, or where a kidnapper threatens to cut off a hostage’s finger to send to relatives with a ransom demand. Both of these examples would be considered by most people to be serious bodily harms (without any accompanying threat of death or reasonable grounds for 60

See Chapter 7, where the same point was made in relation to the defence of property. The same line of argument is made in GA Diamond, ‘To have but not to hold: can “resistance against kidnapping” justify lethal self-defense against incapacitated batterers?’ (2002) 102 Columbia Law Review 729–73, at 733. His argument is subject to the same objections as will shortly be made to that of Kates and Engberg. 61

152

Killing in Self-Defence

fearing death).62 But should it be permissible under the criminal law to kill to prevent these harms? The premise of this book suggests that the answer may well be no. The argument throughout this text has been that we are permitted to kill in self-defence because we have a right to life and therefore a right to defend ourselves from unjust threats to our life. That we are permitted to kill an aggressor who is threatening to kill us is because the aggressor, by virtue of becoming an unjust immediate threat to our life, forfeits her own right to life. The kneecapper and kidnapper above do not threaten the life of their victims. On the basis of my justification for the use of self-defensive force, they do not forfeit their own right to life and we should not, therefore, be permitted to kill them. It is considered here that the life of even a terrorist or kidnapper is worth more than the bodily integrity of another person. Attacks on bodily integrity, although painful, are generally events from which the victim can recover and go on to lead a fulfilling life. Deprivations of life are not. As Ashworth states: 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.63

But maybe this is not the case. Maybe there are some attacks on bodily integrity that are so harmful that the victim can never recover and go on to lead a fulfilling life. Such attacks might constitute a harm as serious as the harm of deprivation of life itself. Prolonged and painful torture is perhaps one example,64 although it is difficult to think of a realistic scenario that would not also be reasonable to consider a threat to life. Where does this leave us in our consideration of killing to prevent rape? It has been suggested so far that the assumption that it is always permissible to kill to avoid serious bodily harm is questionable. Kates and Engberg’s argument that killing to prevent rape is permissible because the harm of rape is equivalent to serious bodily harm is therefore rejected. Rather, it is suggested that if killing to prevent rape is permissible, an argument must be constructed that the harm of rape is worse than that; that the harm of rape is so serious that it is equivalent to, or at least approaches, the harm of deprivation of life itself.65 62 It might be argued that the kneecapping is potentially life threatening, but with the wide availability of medical treatment, this is extremely unlikely. 63 Ashworth, Principles of Criminal Law, at 249. His view is echoed in R Young, ‘What is so wrong with killing people?’ (1979) 54 Philosophy 515–28, at 518. 64 In discussion, colleagues suggested the example of having one’s limbs cut off or having one’s spinal cord severed. If this were ever to arise in reality, however, it would surely be reasonable to regard such a serious assault as life threatening and killing would be permissible on this basis. 65 In reaching this conclusion, I seem to be in agreement with both Kadish and Kasachkoff. Kadish comments that killing to prevent rape is permissible because ‘one may plausibly argue that the interests protected are comparable to that of the victim’s life’ (Kadish, ‘Respect for life’,

Killing to Prevent Rape

153

At this point, it is useful to return to Kates and Engberg, who list four types of ‘harm’ associated with rape. They do so in order to establish that rape is equivalent to serious bodily harm. These will now be examined in order to consider whether the harm of rape can possibly be equivalent to, or at least approach, the harm of deprivation of life. The first three types of harm all relate to the physical impact of rape and these can be disposed of most easily. First, state Kates and Engberg, rape potentially causes pain and physical injury such as the tearing of genital tissue or injury to internal organs. But the authors themselves recognize that such injuries are seldom permanent and therefore it is difficult to argue that these alone justify killing. Second, there is the possibility of what Kates and Engberg term ‘venereal disease’.66 They were writing in 1982 before there was wide awareness of the possibility of transmission of the HIV virus through unprotected sexual intercourse but today this could be added to the list of possible harms. But here too, it is difficult to argue that the possibility of catching a sexually transmitted disease is a good reason for permitting the victim to kill a rapist. Admittedly the transmission of the HIV virus is potentially an extremely serious harm. The possibility of contracting the virus (or one of equivalent seriousness) from a rapist is remote,67 however, and must be weighed up against the certain loss of life of the rapist. Third, Kates and Engberg refer to the possibility of unwanted pregnancy that can result from rape.68 One might counter, however, that while pregnancy is an extremely unpleasant potential consequence of a rape, the harm caused is not permanent because the rape victim will generally be able to have an abortion. Kates and Engberg’s response to this is that abortion itself is traumatic and might run contrary to the victim’s principles and that for this reason pregnancy constitutes a great bodily harm.69 I would have to disagree. First, it is worth remembering that the standard we are looking to meet is not Kates and Engberg’s standard of great bodily harm, but the higher standard of a harm so serious that it at least approaches the harm of at 888). Kasachkoff suggests that killing to prevent rape is permissible ‘because of the seriousness of [the attack] on one’s person and because . . . [it is] closest to the case of a threat to one’s life, because the danger with which the person is threatened is, or may be, similarly final and irreversible’ (Kasachkoff, ‘Killing in self-defence’, at 511). Unfortunately Kadish and Kasachkoff make these comments only in passing and do not consider the issue further. 66 Kates and Engberg, at 896. See also People v Landrum 407 NW 2d 614 (1986) (Court of Appeals of Michigan), at 617. 67 See table 2 of M Fisher, P Benn, B Evans, A Pozniak, M Jones, S MacLean, O Davidson, J Summerside, and D Hawkins, ‘UK guideline for the use of post-exposure prophylaxis for HIV following sexual exposure’ (2006) 17 International Journal of STD and AIDS 81–92. Even if the HIV virus were to be transmitted, the victim may not go on to develop AIDS, and even if she did, developments in treatment are such that this would not necessarily constitute a threat to life. 68 Kates and Engberg, at 897. See also People v Landrum 407 NW 2d 614 (1986), at 617; and the dissenting opinion of Schauer J in People v McIlvain 130 P 2d 131 (1942) (District 69 Court of Appeal of California), at 334. Kates and Engberg, at 898.

154

Killing in Self-Defence

the taking of life. It surely cannot be argued that an unwanted pregnancy, or even an unwanted birth (to a woman who has principled objections to abortion), constitutes a harm on this scale. But more importantly, the consequence of unwanted pregnancy simply does not capture the essential harm of rape.70 If it did, rape would be a less serious crime when the victim was pre-pubertal, menopausal, infertile, sterilized, using contraception or, for that matter, male.71 This is not a tenable argument.72 So far, none of the physical effects of rape can be convincingly argued to constitute harm equivalent to a deprivation of life. The essential harm of rape is simply not captured by a focus on its physical effects. This brings us on to Kates and Engberg’s fourth category of harm, the psychological impact of rape.73 If a justification for killing lies anywhere, it may lie here. The proposal that the criminal law should permit killing to prevent serious psychological harm has been made elsewhere, albeit in a different context. Charles Ewing74 and others75 have suggested that ‘battered women’ who kill their abusive partners should be able to plead ‘psychological self-defence’, in recognition of the difficulty that some might have in meeting the established proportionality requirement in self-defence. Specifically, Ewing suggests that the law should permit deadly force to prevent serious psychological injury, defined as ‘gross and enduring impairment of one’s psychological functioning that significantly limits the meaning and value of one’s physical existence’.76 Ewing’s proposal has met with extensive criticism in the context of the debate over defences for battered women who kill their abusive partners.77 Does it have any merit in the context of killing to prevent rape? There is little doubt that the psychological impact of a rape can be considerable. Resick78 has conducted an extensive review of studies of the 70

Exactly what does constitute the essential harm of rape is considered shortly. A point made by Jennifer Temkin in Rape and the Legal Process (2nd edn, 2002), at 61. The possibility of pregnancy is also the only real argument for confining the defence to female persons. As we have seen, this argument does not stand up to examination and therefore it is concluded here that if it can be established that killing to prevent rape is permissible, this should apply equally to male and female persons. 73 See also CR Essique, ‘The use of deadly force by women against rape in Michigan: justifiable homicide?’ (1991) 37 Wayne Law Review 1969–87, at 1986. 74 CP Ewing, Battered Women who Kill: Psychological Self-Defense as Legal Justification (1987); CP Ewing, ‘Psychological self-defense: a proposed justification for battered women who kill’ (1990) 14 Law and Human Behavior 579–94. 75 AJ Sebok, ‘Does an objective theory of self-defense demand too much?’ (1996) 57 University of Pittsburgh Law Review 725–55, at 754; JM Cohen, ‘Regimes of private tyranny: what do they mean to morality and for the criminal law?’ (1996) 57 University of Pittsburgh Law Review 757–808, at 790–1; C Wells, ‘Battered woman syndrome and defences to homicide: where now?’ (1994) 14 LS 266–76, at 273. 76 Ewing, ‘Psychological self-defense’, at 587. 77 DL Faigman, ‘Discerning justice when battered women kill’ (1987) 39 Hastings Law Journal 207–27; SJ Morse, ‘The misbegotten marriage of soft psychology and bad law: psychological self-defense as a justification for homicide’ (1990) 14 Law and Human Behavior 595–618. 78 PA Resick, ‘The psychological impact of rape’ (1993) 8 Journal of Interpersonal Violence 223–55. Resick’s study is referred to here because it summarizes the results of numerous other studies of the issue. 71 72

Killing to Prevent Rape

155

psychological impact of rape. She found that rape victims were more likely to suffer higher levels of fear and anxiety than either those who had been the victim of no crime at all or those who were victims of offences other than rape. This relationship held for at least three years after the rape. Rape victims were also likely to suffer from Post Traumatic Stress Disorder (PTSD), the major symptoms of which include re-living the experience through nightmares and flashbacks, social withdrawal, avoidance behaviour, and increased physiological arousal, leading to sleep disorders and difficulty in concentrating. Rape victims were more likely than non-rape victims to suffer from depression, low self-esteem, social adjustment problems, and difficulties in sexual functioning. Perhaps most tellingly, rape victims were more likely than non-rape victims to have attempted suicide.79 Resick concludes that ‘rape is a life event that causes considerable upheaval in a victim’s psychological functioning for a considerable period of time, perhaps for the rest of her life’.80 The question is whether this constitutes harm so great that it justifies killing to prevent rape.81 It is doubted here that it does. Traumatic as the harmful psychological effects of a rape may be, they are psychological effects that are almost certainly remediable over time. Granted, the rape victim’s life may never be quite the same again, but the psychological harm of rape does at least leave open the possibility of recovery or readjustment. As has already been pointed out, the harm of killing does not leave open this possibility. Precisely this point was made by the Supreme Court of the United States in declaring the death penalty a disproportionate and unconstitutional penalty for the crime of rape. In Coker v Georgia,82 the majority commented that: Rape is without doubt deserving of serious punishment; but in terms of moral depravity and of the injury to the person . . . it does not compare with murder, which does involve the unjustified taking of human life. Although it may be accompanied by another crime, rape by definition does not include the death of or even serious injury to another person. The murderer kills; the rapist, if no more than that, does not. Life is over for the victim of the murderer; for the rape victim, life may not be nearly so happy as it was, but it is not over and normally is not beyond repair.83

In addition, it might be said that opening up the legal possibility of psychological self-defence is something that should be approached with extreme 79 This finding is mirrored in another study: D Kilpatrick, C Edmunds and A Seymour, Rape in America: A Report to the Nation (1992). The authors found that 44 per cent of rape victims had considered suicide (compared to 8 per cent of non-rape victims) and 19 per cent had attempted suicide (compared to 1 per cent of non-rape victims). 80 Resick, ‘The psychological impact of rape’, at 234–5. 81 There is an associated question of whether killing one’s attacker would actually be effective in preventing the psychological consequences from occurring, given that a victim might suffer from these types of psychological symptoms as a result of the trauma of killing her attacker. 82 Thanks are due to Professor Peter Duff for this point. 433 US 584 (1977). 83 ibid 598. See also Young, in ‘What is so wrong with killing people?’, who rejects the permissibility of killing to prevent rape on the basis that rape is a wrong that can be remedied,

156

Killing in Self-Defence

caution, given that great psychological harm can be caused by experiences other than rape. Some thefts of property, for instance, might cause enormous psychological harm. An example might be the person who is threatened with the theft of a much-treasured last remaining photo of their dead child.84 If killing to prevent rape is permissible on the basis that rape causes psychological harm, then it would be difficult to argue against allowing killing in defence of property where the loss of that property would cause psychological harm of a similar magnitude.85 But the ‘psychological harm’ argument also raises a difficulty of a more conceptual nature. Constructing a case that killing is permissible because of the particular psychological harm rape causes to individual victims seems to run into trouble because not all rape victims will necessarily suffer psychological harm of the magnitude described above. For this reason, if one is attempting to justify killing to prevent rape, it is perhaps a more productive line of argument to focus not on the psychological impact of being raped, but on the wrong that rape does to a victim, regardless of whether she suffers from certain specific psychological effects. At this point, it is worth considering wider arguments about the reason why rape is wrong. Of particular relevance is Hampton’s account.86 The reason why rape is wrong, according to Hampton, is that it causes ‘moral injury’ (defined as an injury to a person’s value)87 that does not necessarily involve material or psychological damage. For Hampton, this injury is effected by the expressive content of the act of rape. The rapist expresses the idea that the rape victim is a ‘mere object who [is] there to be used whenever the rapist feels the need to do so’88 and Hampton concludes that it is the expressive content of the action ‘representing the rapist as master and the victim as inferior object . . . that accounts for why it is wrongful’.89 A similar account is provided by Gardner and Shute.90 Like Hampton, Gardner and Shute conclude that the wrongness of rape lies not in the harm that it causes, but in the fact that the rapist uses the rape victim as an object and in doing so dehumanizes her. They reach this conclusion by considering ‘pure rape’, where the rape victim is unconscious, uninjured, and never becomes aware of what has happened to her. Because there is no harm caused to the victim and yet the act of the rapist is still wrong, Gardner and whereas killing is not (at 520). Cf the minority judgment of Powell J in Coker v Georgia: ‘Some [rape] victims are so grievously injured psychologically that life is beyond repair’ (at 603, emphasis in original). 84 Thanks are due to Professor Chris Gane for this example. 85 This is only a problem, of course, if it is held, as it is here, that killing in defence of property is never justified (see Chapter 7). 86 J Hampton, ‘Defining wrong and defining rape’, in K Burgess-Jackson (ed), A Most Detestable Crime: New Philosophical Essays on Rape (1999) 118–56. 87 88 89 Hampton, ‘Defining wrong’, at 131. ibid 135. ibid 135. 90 J Gardner and S Shute, ‘The wrongness of rape’, in J Horder (ed), Oxford Essays in Jurisprudence, Fourth Series (2000) 193–217.

Killing to Prevent Rape

157

Shute conclude that the primary reason why rape is wrong cannot lie in the harm that it causes but must lie in the dehumanization and objectification of the rape victim by the rapist.91 As Gardner and Shute state, the element of objectification and dehumanization is present even when the rape is unaccompanied by any physical violence or injury: Rape, in the pure case, is the sheer use of a person. In less pure, but statistically more typical, cases this use is accompanied by violence, terror, humiliation, etc. Our only point is that when someone feels humiliated by rape itself, this feeling is justified. Rape is humiliating even when unaccompanied by further affronts, because the sheer use of a person, and in that sense the objectification of a person, is a denial of their personhood. It is literally dehumanising.92

At this point, it might be said that other offences also contain this element of ‘sheer use’. It is possible to envisage a non-sexual assault in which the victim is used by the perpetrator as an object upon which to inflict physical injury. But the crime of rape contains an additional element, that of sexual penetration, and it is here that Gardner and Shute locate what is particularly wrong about rape over and above other physical assaults. It is because of the significance with which society has imbued the act of sexual penetration: The most fundamental element of wrongdoing in rape, which differentiates rape from (most) assaults and gives rape a separate theme from the family of assault crimes, is the sheer use of the person raped, whether that is how the rapist saw what he was doing or otherwise. To understand how rape counts as sheer use, the social meaning of sexual penetration has to be kept in focus.93

Hampton makes a similar point, stating that: Our sexuality is deeply important to each of us and in certain ways central to our sense of self. To express the subordination of another through the sexual act will therefore likely be wounding to that person in a profound way—not only psychologically but objectively, given the way in which the act will be powerfully expressive of the idea of inferiority.94

It is argued here that the criminal law should provide a complete defence to the accused who has killed to prevent rape. The reason for this is not to be found in any of the justifications advanced by Kates and Engberg. Rather it is because rape approaches the standard of a wrong equivalent to a deprivation of life itself. That is, as Hampton and as Gardner and Shute argue, the rapist uses the victim as an object through the act of sexual penetration, an act that 91

92 Gardner and Shute, ‘The wrongness of rape’, at 204. ibid 205. ibid 212. 94 ibid 151. See also the Criminal Law Revision Committee’s Fifteenth Report on Sexual Offences (Cmnd 9213, 1984), at para 2.2. If the wrong of rape is socially constructed, this does raise the question of whether rape would be equally wrong in a society where sexual penetration was not imbued with such significance. 93

158

Killing in Self-Defence

has been given a particular significance by society, and, in doing so, denies the victim’s humanity.95 Arriving at this conclusion was not straightforward, as so many of the arguments that have been made in support of a defence of this nature do not stand up to critical examination. Oddly enough, Kates and Engberg seem to accept that their arguments are not particularly persuasive. On the one hand, they claim that it is ‘intuitively obvious’ that killing to prevent rape is permissible.96 On the other hand, they confess to finding their own arguments in support of the proposition unconvincing and conclude by stating that: . . . some additional element inherent in the crime of rape—something in the nature of the violation which the law and common sense recognise—justifies the privilege to resist rape by any and all means. Although impossible to verify or quantify, this subtle understanding of the crime of rape may be the true justification for the self-defence privilege.97

This mystery ‘additional element’ sought by Kates and Engberg may well lie in the Hampton/Gardner and Shute account of the wrongness of rape as a denial of the victim’s humanity. 95 For another example of someone who argues that a ‘denial of personhood’ should ground a claim of lethal self-defence, see KF Ferzan, ‘Defending imminence: from battered women to Iraq’ (2004) 46 Arizona Law Review 213–62, at 253. Ferzan is writing in the context of killing to prevent kidnapping and (more dubiously) killing to prevent non-life-threatening domestic abuse. 96 They also follow the rather dubious logic that because they have raised the issue in classroom discussions with students and have met with universal agreement that killing to prevent rape is permissible, it must, therefore, be permissible. 97 Kates and Engberg, at 906.

9 Mistake 9.1 Introduction The final substantive issue for consideration is mistake in self-defence.1 Thus far, it has been assumed that the accused’s perceptions—about the existence of an imminent attack or about the amount of force necessary to repel that attack—are correct, but this will not always be so. Mistakes can be made and the law must decide how best to treat the accused who makes such a mistake. Of course, self-defence is not the only defence that can be grounded on the basis of a mistaken belief. The issue could arise in relation to, for example, necessity, duress, or provocation.2 While this chapter is concerned specifically with self-defence, this is not to say that it does not have relevance for the issue of mistakes in defence more generally.3 It should also be said that the debate about mistake in self-defence sits beside a parallel debate about mistake in relation to an offence element. Here, most attention has been devoted to mistake about consent in sexual offences,4 and many of the arguments made in this context also resonate in the context of mistake in self-defence. There are essentially two types of mistake that might be made by the accused claiming self-defence: a mistake relating to the perception of an attack5 and a mistake relating to the level of force necessary to repel that attack.6 While they have much in common, there is some merit in discussing each separately, given that in some jurisdictions, they have been treated differently. 1 This chapter draws in part from F Leverick, ‘Mistake in self-defence after Drury’, which was first published in the Juridical Review (2002 JR 35–48). Some passages from this article are reproduced with the kind permission of the publisher, W Green and Son Ltd. 2 Although it does seem less likely that a mistake would be made about whether a third party is issuing a threat of death unless one complies with her demands than it does about whether one is about to face an attack. 3 On which, see J Chalmers and F Leverick, Criminal Defences and Pleas in Bar of Trial (2006), chapter 12; AP Simester, ‘Mistakes in defence’ (1992) 12 OJLS 295–310. 4 EM Curley, ‘Excusing rape’ (1976) 5 Philosophy and Public Affairs 325–60; F McAuley, ‘The grammar of mistake in criminal law’ (1996) 31 Irish Jurist 56–82; T Pickard, ‘Culpable mistakes and rape: relating mens rea to the crime’ (1980) 30 University of Toronto Law Journal 75–98; DP Bryden, ‘Redefining rape’ (2000) 3 Buffalo Criminal Law Review 317–479; VJ Dettmar, ‘Culpable mistakes in rape: eliminating the defense of unreasonable mistake of fact as to victim consent’ (1984–5) 89 Dickinson Law Review 473–99; H Power, ‘Towards a re-definition of the mens rea of rape’ (2003) 23 OJLS 379–404; C Gane, Sexual Offences (1992), at 44–6. 5 Where the accused believes that she is being attacked but, in reality, she is not. 6 Where the accused correctly perceives that she is being attacked, but wrongly believes that a certain (excessive) degree of force is necessary to repel that attack.

Killing in Self-Defence

160

9.2 Mistake about an Attack 9.2.1 The legal options There are a number of ways in which the law might respond to a mistaken belief in the existence of an attack. One possibility, which can be swiftly rejected, is that any mistake serves to defeat the defence of self-defence. This seems unduly harsh to the accused, especially where she had good reasons for her belief. It would result in a criminal conviction for someone who, in the heat of the moment and faced with circumstances in which it was reasonable to believe that she was about to be attacked,7 used force to defend herself. There is no basis upon which a criminal conviction is merited. We cannot blame her for her conduct and she did not demonstrate any failing with which the criminal law should be concerned. This was, however, precisely the approach taken by the majority of the New York Court of Appeals in People v Young,8 albeit in the context of the defence of others. Here, the defendant came across two men in civilian dress who appeared to be assaulting a teenager. The boy was crying and his trousers had almost been removed in the struggle. The defendant stepped in to prevent the ‘attack’, breaking the leg of one of the ‘attackers’ in the process. It turned out that the attackers were plain-clothes police officers attempting to arrest the boy. Although it was accepted that the mistake was a reasonable one, the majority upheld the defendant’s conviction for assault on the basis that allowing a defence would ‘not be conducive to an orderly society’.9 As Frossel J pointed out in his dissenting judgment, it is hardly conducive to an orderly society to promote ‘a principle which would deter one from coming to the aid of a fellow citizen who he has reasonable ground to apprehend is in imminent danger of personal injury at the hands of assailants’.10 It is generally accepted that Young has now been overruled by statute11 and, to my knowledge, there is no common law jurisdiction that has ever taken this approach to reasonable mistake in self-defence, as opposed to the defence of others. What is more debatable is whether the belief of the accused needs to be reasonable in order to ground the defence, or whether 7 Take the accused who is out on a dark night and is suddenly confronted by a hooded figure who knocks her to the ground. It may be entirely reasonable for her to believe that she is about to be attacked, even though the hooded figure turns out to be a runner who pushed her over by 8 accident. 183 NE 2d 319 (1962). 9 10 ibid 319. ibid 322. 11 The 1965 revised version of the New York Penal Code provides that ‘[a] person may . . . use physical force upon another person when and to the extent he reasonably believes such to be necessary to defend . . . a third person from what he reasonably believes to be the use or imminent use of unlawful physical force by such other person’ (§ 35.15(1)(b)). For the effect of this provision on Young, see People v Melendez 79 NY 2d 861 (1992) (Court of Appeals of New York); Commonwealth v Martin 341 NE 2d 885 (1976) (Supreme Court of Massachusetts); State v Wenger 390 NE 2d 801 (1979) (Supreme Court of Ohio).

Mistake

161

any honestly held belief, reasonable or not, will suffice. The majority of common law jurisdictions take the approach that only a reasonable mistake about the existence of an attack should be permitted to ground an acquittal. Scots law is a typical example. Thus Hume stated: It is not sufficient that the pannel have killed out of an apprehension, though ever so serious on his part, of danger to his life, if it was not also a reasonable apprehension, and well grounded in the circumstances of the situation . . . In a matter of this kind, the law cannot entirely pardon any vain terror or sudden panic; for against this, so far at least as concerns his conduct towards others, it is the duty of every man to be on his guard.12

The reasonableness requirement was reiterated in the leading case of Owens v HM Advocate,13 where it was held that ‘self-defence is made out when it is established to the satisfaction of the jury that the panel believed that he was in imminent danger and that he held that belief on reasonable grounds. Grounds for such belief may exist though they are founded on a genuine mistake of fact.’14 Owens has been followed in a line of cases including Crawford v HM Advocate;15 McCluskey v HM Advocate;16 McLean v Jessop;17 Jones v HM Advocate;18 and Burns v HM Advocate.19 The same approach can be found in Canadian law where, in R v Reilly,20 the Canadian Supreme Court concluded that: . . . the accused may still be found to have acted in self-defence even if he was mistaken in his perception. Reasonable and probable grounds must still exist for this mistaken perception in the sense that the mistake must have been one which an ordinary man using ordinary care could have made in the same circumstances.21

It was also the approach taken by the Australian High Court in Zecevic v DPP22 and represents the law in a number of US states.23 By contrast, in English law, where a defendant acts in ‘self-defence’ under the genuine but mistaken belief that she is being attacked, she is entitled to an acquittal (assuming the other conditions of the defence are made out), even if her mistake is an unreasonable one.24 The manner in which English law arrived at this result is worthy of some attention. 12

D Hume, Commentaries on the Law of Scotland (4th edn, 1844) i, 224. 14 1946 JC 119. Owens, per the Lord Justice-General (Normand) at 125. 15 16 17 1950 JC 67, at 72. 1959 JC 39, at 43. 1989 SCCR 13, at 17. 18 1990 JC 160, at 172. 19 1995 JC 154, at 159. Although cf Drury v HM Advocate 2001 SLT 1013, discussed in 20 Section 9.2.2 below. [1984] 2 SCR 396. 21 ibid [19]. Reilly has since been confirmed by the Supreme Court in R v Pétel [1994] 1 SCR 3, per Lamer CJ, Sopinka, Cory, McLachlin, and Iacobucci JJ at [12] and R v Cinous [2002] 2 SCR 3, per McLachlin CJ, L’Heureux-Dubé, Bastarache, and LeBel JJ at [41]. 22 (1987) 162 CLR 645. 23 Ala Code § 13A-3-23(a); Ark Code § 5-2-606(a); Cal Penal Code § 198; Colo Rev Stat § 18-1-704(1); Fla Stat § 776.012; 720 ICLS § 5/7-1(a); Miss Code Ann § 97-3-15(f); NY UCC Law § 35.15(1); Tenn Code Ann § 39-11-611(a). 24 Reflecting a more general trend towards the subjective approach in English criminal law, a trend noted by the House of Lords in R v G [2003] UKHL 50, per Lord Steyn at [55]. 13

162

Killing in Self-Defence

Until R v Williams,25 a long line of English decisions had suggested that while an unreasonable mistake in relation to an offence element would result in an acquittal, an unreasonable mistake in relation to a defence element would not do likewise.26 This distinction seemed to be endorsed in R v Morgan,27 where a mistaken belief in consent to sexual intercourse was held to warrant acquitting a defendant of rape, even where that mistake was not reasonable. It did not appear to be the aim of the House of Lords to extend this to mistakes in defence. Lord Hailsham, for example, made a clear distinction between cases like Morgan, where ‘the issue relates to the primary intention which the prosecution has to prove’ and ‘cases where there is a special defence, like self-defence’28 and assumed there was a requirement of reasonableness in relation to the latter. Lord Hailsham’s distinction was followed in Albert v Lavin,29 a case where the appellant, mistakenly believing he was under attack, had used force to repel a plain clothes police officer who was in the process of trying to arrest him. The appellant had tried to argue that the mens rea of assault is an intention to use or threaten force unlawfully and that, following Morgan, an unreasonable belief that one is acting in self-defence negates mens rea as there is no such unlawful intention. This argument was rejected by Hodgson J: ‘It does not seem to me that the element of unlawfulness can properly be regarded as part of the definitional elements of the offence. In defining a criminal offence the word unlawful is surely tautologous (sic) and can add nothing to its essential ingredients.’30 The Divisional Court went on to confirm that a belief in the existence of an attack must be reasonable in order to ground an acquittal on the basis of self-defence.31 Given the rejection of the ‘honest belief’ principle in Albert v Lavin, how did the Court of Appeal in Williams arrive at the reverse decision? Their approach can be attributed at least in part to comments made in R v Kimber,32 where the ‘honest belief’ principle was extended to indecent assault. In Kimber, Lawton LJ considered Albert v Lavin and did not accept Hodgson LJ’s argument that, in defining the offence of assault, the word ‘unlawful’ is tautological. For Lawton LJ, the word imported an ‘essential element’ into the offence because without it, he claimed, ‘social life would be unbearable because every touching would amount to a battery unless there was an evidential basis for a defence’.33 It is not surprising, then, that in Williams, the Court of Appeal held that a defendant charged with assault could successfully plead self-defence on the 25

[1987] 3 All ER 411. R v Smith (1837) 8 Car & P 158 160; 173 ER 441; R v Weston (1879) 14 Cox CC 346; R v Rose (1884) 15 Cox CC 540; R v Chisam (1963) 47 Cr App R 130; R v Fennell [1971] 1 QB 27 428. Cf R v Porritt [1961] 1 WLR 1372, at 1376. [1976] AC 182. 28 29 30 ibid 214. [1981] 2 WLR 1070. ibid 1083. 31 The case reached the House of Lords but this point was not considered (Albert v Lavin 32 [1982] AC 546). [1983] 1 WLR 1118. 33 ibid 1123. On this claim, see Section 9.2.2 below. 26

Mistake

163

basis of an honest but unreasonable belief that she was being attacked. In Williams, the defendant intervened in the course of a lawful arrest, thinking that the police officer was an assailant. The trial judge directed the jury that only a reasonable belief that the police officer was an attacker would entitle Williams to an acquittal. On appeal, this was held to be a misdirection. Lord Lane CJ agreed with Lawton LJ’s comments in Kimber and defined assault as an act where the defendant applies unlawful force to the complainant, the mental element being an intention to apply unlawful force. If a defendant believes, however unreasonably, that she is under attack, or that she is protecting another from an attack (as Williams did), then she does not have an intention to apply unlawful force and must be acquitted. Williams was swiftly followed in a number of English Court of Appeal cases34 and the approach was cemented by Beckford v R,35 a Privy Council decision and R v Oatridge.36 Both involved charges of murder, indicating that the approach taken in Williams was not confined to assault.37 A similar approach is taken in the US Model Penal Code, where § 3.04 contains no requirement that the belief of the defendant in the need to use defensive force be reasonable. This must be read in conjunction with § 3.09, however, which provides that where the defendant’s belief was formed recklessly or negligently, self-defence will be unavailable as a defence to any offence where recklessness or negligence would establish culpability.38 This proposal has largely been ignored by US states, the more common approach to mistake about an attack simply being to require reasonableness of belief in relation to all offences.39

9.2.2 Should mistake about the existence of an attack be required to be reasonable? The main argument put forward in the English case law for allowing unreasonable mistake to ground an acquittal is, as we have already seen, one of logic. The mens rea of assault is an intention to use or threaten force unlawfully, this supposedly being essential in order to prevent any sort of touching from falling into the scope of the offence.40 A similar approach was taken by the Appeal Court in the Scottish case of Drury v HM Advocate,41 where the mens rea of murder was defined not 34 R v Jackson [1985] Crim LR 674; R v Asbury [1986] Crim LR 258; R v Fisher [1987] Crim 35 LR 334. [1988] AC 130. 36 (1992) 94 Cr App R 367. More recently, see the Court of Appeal in R v Martin [2001] EWCA Crim 2245 and the Privy Council in Shaw v R [2001] UKPC 26. 37 On which, see Lord Griffiths in Beckford: ‘it is an essential element of all crimes of violence that the violence or the threat of violence should be unlawful’ (at 144, emphasis added). 38 § 3.09 relates to justification defences generally, of which self-defence is one. 39 Two states have adopted the Model Penal Code provision: Delaware (Del Code Ann § 464(a)) and Kentucky (Ky Rev Stat Ann § 503.120(1)). 40 41 R v Kimber [1983] 1 WLR 1118, per Lawton LJ at 1123–4. 2001 SLT 1013.

164

Killing in Self-Defence

simply as an intention to kill but as a wicked intention to do so,42 the term wickedness being necessary to prevent any applicable justification or excuse defence, such as self-defence, from attracting criminal liability.43 Indeed, if the approach taken in Drury becomes accepted in Scots criminal law,44 this has the potential to change the way in which Scots law deals with mistake in self-defence. If murder requires a wicked intention (that is, one which excludes action taken in self-defence), then it would seem to follow that the accused who honestly believes she is acting in self-defence, whether or not her mistake is reasonable, does not act with wicked intent and will therefore not have the requisite mens rea for murder.45 This would leave her facing conviction for, at most, culpable homicide. But defining offences in this way is not the only way to avoid the outcomes that were feared in Kimber and Drury. Minor touches can be excluded from the definition of assault by the sensible use of prosecutorial discretion or by recognizing the concept of implied consent as a defence.46 Indeed, as Simester has argued, if an unlawful intention really was a requirement of the offence of assault, the consequences would be enormous. For example, any mistake of law, even an unreasonable one, would operate as a defence, since the defendant would have no unlawful intention.47 Likewise, the self-defender can be excluded from criminal liability simply by recognizing self-defence as a separate defence operating once the offence definition has been satisfied.48 If these alternative approaches are taken, then the question of whether or not mistakes in self-defence require some reasonable basis remains up for consideration. It is argued here that only a reasonable mistake about the perception of an attack should be permitted to ground the defence of self-defence. To understand why this is so, it is instructive to return to the material presented in Chapters 2 and 3 and examine the basis of the defence of self-defence: why we are permitted to kill someone who is attacking us. As Chapter 2 outlined, the defence of self-defence can take two forms: that of justification or excuse. In both instances, the accused accepts that she has done something wrong (taken the life of another human being) but offers a reason for her conduct. The justification form of the defence applies where 42

Or alternatively as wicked recklessness. Drury, per the Lord Justice-General (Rodger) at [11]. See also Lord MacKay at [9]; Lord Johnston at [18] and Lord Nimmo-Smith at [3]. Cf the Canadian Supreme Court in R v Kerr [2004] 2 SCR 371, per Bastarache J at [28] and per LeBel J at [93]; R v Parent [2001] 1 SCR 761, at [6]. 44 Drury was a case about provocation and has not yet proved influential outside of the narrow confines of the law relating to this defence. On this, see Chapter 2. 45 Leverick, ‘Mistake in self-defence’, at 40–2 and J Chalmers, ‘Collapsing the structure of criminal law’ 2001 SLT (News) 241–5; Chalmers and Leverick, Criminal Defences, chapters 3 and 12. 46 As was recognized by Goff LJ in Collins v Wilcock [1984] 1 WLR 1172, at 1177. 47 Simester, ‘Mistakes in defence’, at 301. 48 Leverick, ‘Mistake in self-defence’, at 41–3. 43

Mistake

165

the accused correctly perceives that she is being attacked. The reason she offers for killing the aggressor is the fact that the aggressor was threatening to kill her. We all, by virtue of being human, possess the right to life. The only way the accused could save her life was by using force to repel the attack. The excuse form of self-defence applies where the accused is mistaken in her perception that she is being attacked. Here, the accused offers a reason for thinking that she had a reason to do what she did. That is, she accepts that she has done something wrong (killed another human being) but asserts that she had good reasons for thinking she was under attack. Given these good reasons, she thought that the only way she could save her own life was by using proportionate force to repel the aggressor.49 Why does she need ‘good reasons’ for her belief? Why will just any mistake, reasonable or otherwise, not do? The answer lies in the role of the perceived aggressor. As Chapter 3 argued, the reason why we are permitted to kill in self-defence is because we, as human beings, possess a right to life. An aggressor, by virtue of unjustly attacking her victim, violates or threatens to violate the victim’s right to life. The victim is therefore justified in using defensive force to repel the threat, even where this leads to the death of the aggressor. As Chapter 3 noted, however, this explanation is not complete because it does not explain why the aggressor is not permitted to fight back against the self-defensive force. After all, if all human beings possess a right to life, then presumably the aggressor also possesses this right. If the aggressor does possess a right to life, then it is not immediately obvious why the victim is permitted to kill her, even in self-defence. To explain why the aggressor is not permitted to fight back against the self-defensive force, something akin to forfeiture of the right to life must enter the equation. As Chapter 3 argued, the right to life is forfeited by virtue of becoming an unjust immediate threat to the life of another. This explains why we are permitted to kill an aggressor, but the aggressor is not permitted to kill us: the aggressor, in becoming an unjust immediate threat to the life of another, forfeits her own right to life. This, then, is the reason why we are permitted to kill someone who is actually attacking us: the justification form of self-defence. But when we kill a human being in the mistaken belief that she is attacking us (the excuse form of self-defence), the situation is different. There is no aggressor in reality. The perceived ‘aggressor’, like all human beings, possesses a right to life and she has not forfeited her right to life by becoming an unjust immediate threat to the life of another. Instead of killing someone who was threatening to violate our right to life, we have killed a human being who was posing no threat but was unfortunate enough to do something to suggest that she was about to 49 Not everyone would agree with this conception of justification and excuse but in Chapter 2, following Gardner, I argue that it is the preferable one.

166

Killing in Self-Defence

attack us.50 This is why, in order to benefit from a complete acquittal, we need to offer a good reason for what we have done. If it were otherwise this would ignore the fact that the perceived aggressor had a right to life and a right to the protection of the criminal law against having her life taken for no good reason. As Lord Simon put it in R v Morgan:51 The policy of the law in this regard could well derive from its concern to hold a fair balance between victim and accused. It would hardly seem just to fob off a victim of a savage assault with such comfort as he could derive from knowing that his injury was caused by a belief, however, absurd, that he was about to attack the accused.52

It is argued here, then, that allowing an unreasonably mistaken belief that one is acting in self-defence to ground an acquittal fails to provide adequate protection of the right to life of the person who was posing no threat but who has been mistaken for an attacker.53 But are there any arguments (other than ‘logic’) that might be made in favour of allowing an unreasonable mistake about the existence of an attack to ground an acquittal? One possibility is that the accused who makes an honest but unreasonable mistake is not sufficiently blameworthy to merit the imposition of criminal liability. As Singer puts it, ‘the engine of the criminal law is to be ignited primarily for imposing blame, and an actor who honestly believes the facts allow her to act legally, is not blameworthy, however badly mistaken’.54 In response, it is suggested that there are two compelling reasons why punishment is appropriate. First, punishment serves an expressive function, expressing the indignation of society that the accused failed to show adequate consideration to the interests of others.55 She has, after all, killed a human being who was posing no threat in the mistaken belief that this person was about to attack her, and she had no acceptable reason for this belief. As Fletcher puts it, to allow unreasonable mistaken belief in self-defence is to ‘sanction thoughtless, negligent over-reaction’ when instead, ‘the lack of restraint, the indulgence, the failure to discipline one’s reactions . . . are all grounds for blaming the person who claims his wrongdoing is excused’.56 50 CJ Rosen, ‘The excuse of self-defense: correcting a historical accident on behalf of battered women who kill’ (1986) 36 American University Law Review 11–56, at 47–8. 51 [1976] AC 182. 52 At 221. See also Simester, ‘Mistakes in defence’, at 309; OE Woodruff Jnr, ‘Mistake of fact as a defense’ (1958–9) 63 Dickinson Law Review 319–33, at 325–6. 53 In this, I am in agreement with Simester, ‘Mistakes in defence’. For an argument of this nature based on Art 2 of the European Convention on Human Rights, see Chapter 10. 54 R Singer, ‘The resurgence of mens rea: II—honest but unreasonable mistake of fact in self defense’ (1986–7) 28 Boston College Law Review 459–519, at 512. 55 V Tadros, ‘The characters of excuse’ (2001) 21 OJLS 495–519, at 517. Tadros would almost certainly not agree with my classification of mistaken belief in self-defence as an excuse. See V Tadros, Criminal Responsibility (2005), at 280–90 and the discussion in Chapter 2. 56 GP Fletcher, ‘The psychotic aggressor: a generation later’ (1993) 27 Israel Law Review 227–46, at 241.

Mistake

167

It is helpful at this point to think about the type of belief that might be deemed unreasonable in this context. One example could be the paranoid racist, who honestly believes that all black people are uncontrollably violent and likely to attack her at the slightest opportunity. If she kills a black person carrying an umbrella whom she meets on the stairs of a block of flats solely because that person is black and therefore perceived to be dangerous, surely we should consider this person blameworthy to the extent of imposing criminal liability? If we acquit the paranoid racist on the basis of her honest but entirely unreasonable belief, the message expressed is that society does not blame or condemn its members for acting on the basis of such abhorrent beliefs.57 The hypothetical example of the paranoid racist is not as far-fetched as might initially be thought. Lee documents a number of cases in the USA where racist defendants have been acquitted after killing black or Asian people in ‘self-defence’ because they thought that such a person was more likely to attack them than a white person.58 Second, punishment serves a conduct-guiding function.59 When we say that we will punish people who are too quick to use ‘defensive’ force, this sends the message that we should take reasonable care to check the accuracy of our beliefs before acting. If an individual is unwilling or unable to conform to this standard of behaviour, they surely represent something of a danger to society.60 The point is made by the Court of Appeals of New York in People v Goetz,61 a notorious case that might be added to the examples of unreasonable belief noted above. The case concerned the ‘self-defensive’ shooting of a group of black men on a subway train by a white defendant. One of the men had approached Goetz, stating ‘give me five dollars’. Neither he nor any of the other four men were armed. Goetz responded by pulling a handgun out of his coat and shooting all four of the men. He claimed that he was justified in shooting because he feared, based on his prior experiences, that the men would ‘maim’ him. He admitted that he knew that none of them had a gun. Goetz was ultimately acquitted of attempted murder by a jury, but in overturning a decision to dismiss the indictment at an earlier stage of the case, Wachtler CJ commented: To completely exonerate . . . an individual, no matter how aberrational or bizarre his thought patterns, would allow citizens to set their own standards for the permissible 57 Canadian Department of Justice, Provocation, Self-Defense and Defence of Property: A Consultation Paper (1998), section 2. 58 CKY Lee, ‘Race and self-defense: towards a normative conception of reasonableness’ (1996) 81 Minnesota Law Review 367–500. One example is State v Simon 646 P 2d 119 (1982) (Supreme Court of Kansas), where an elderly man killed his Chinese neighbour in the course of an argument. Simon believed that all Chinese people were experts in martial arts and he shot the neighbour in what appears to have been the entirely unreasonable belief that he was about to launch a lethal attack. 59 See Chapter 2, where it was suggested that excuse defences can serve a conduct-guiding 60 function. Cf Singer, ‘The resurgence of mens rea’, at 511. 61 497 NE 2d 41 (1986).

168

Killing in Self-Defence

use of force. It would also allow a legally competent defendant suffering from delusions to kill or perform acts of violence with impunity, contrary to fundamental principles of justice and criminal law.62

This is not to suggest that the law should require long and drawn-out checks to be made in the face of an immediate life or death situation. Allowance should be made for the panic that someone is likely to experience in the heat of the moment if she becomes the subject of an unexpected attack. As Horder has pointed out, writing in the context of the defence of duress, requirements of reasonableness can be sensitive to the fact that people may make mistakes under pressure.63 It has been said that this is not a convincing argument for the reasonableness requirement, based as it is on the assumption that the law can guide our conduct in this way. Singer has argued that the potential for deterrence is low because the honestly mistaken accused believes that she is acting reasonably, even when she is not. The chance that she will (a) have sufficient time to reflect that she might be wrong about something she believes to be correct and (b) reconsider something about which she is already convinced is too slim a basis upon which to require reasonableness of belief.64 In response, it must be conceded that the potential for deterrence is indeed low. That is not to say, however, that it is non-existent.65 A reasonableness requirement does send out the message that we should all should take reasonable care to check the accuracy of our beliefs before acting rashly in self-defence. Where the consequences of a mistake are as serious as the death of a human being who was posing no threat, it does not seem unreasonable to conclude that we are all under an obligation, as responsible citizens in society, to control our behaviour in this way. Even if one does not accept the deterrence argument in relation to private citizens, it gains greater force in the case of public officials such as police officers. Unlike the private citizen, the public official is trained to assess the situation before using force in self-defence or the defence of others. Incidents where innocent individuals have been shot and killed by police officers who mistakenly believed that they were acting in self-defence or the defence of others suggest that the need for the law to serve such a conduct-guiding function is pressing.66 62 497 NE 2d 41 (1986) at 50. For detailed discussion of the case, see GP Fletcher, A Crime of Self-Defense: Bernard Goetz and the Law on Trial (1988); Singer, ‘The resurgence of mens rea’, at 493–7. 63 J Horder, ‘Occupying the moral high ground? The Law Commission on duress’ [1994] 64 Crim LR 334–42, at 341. Singer, ‘The resurgence of mens rea’, at 510. 65 One may take the view that the criminal law has no deterrent effect at all. See PH Robinson and JH Darley, ‘Does criminal law deter? A behavioural science investigation’ (2004) 24 OJLS 173–205. 66 The case of Harry Stanley is one of the most troubling. Mr Stanley was shot dead by police marksmen as he walked home from a pub carrying a wooden table leg. The police had received a telephone call from a member of the public who had seen Mr Stanley leaving the pub with the table leg and thought he had an Irish accent (he was actually Scottish). On this basis, armed

Mistake

169

9.3 Mistake about the Amount of Force Necessary to Repel an Attack 9.3.1 The options The other type of mistake that the accused might make is a mistake in relation to the amount of force needed to repel an attack. At the outset, it can be said that one option would be simply to deny the defence to the accused who honestly but mistakenly believed that a certain (excessive) degree of force was necessary to repel an attack. As has already been suggested in relation to mistakes about the existence of an attack, this would seem unnecessarily harsh. In the heat of the moment, especially where the attack is sudden and unexpected, it is going to be difficult to judge accurately the precise level of force that is necessary to repel an attack.67 The more relevant question is whether such a mistake needs to be reasonable. While at least some common law jurisdictions permit an unreasonable mistake about the perception of an attack to ground an acquittal on the basis of self-defence, the position taken about a mistake in the amount of force necessary to repel an attack is more uniform. In none of the major common law jurisdictions would an unreasonable mistake about the amount of force necessary to repel an attack ground a complete acquittal for the accused,68 although in a very small minority of jurisdictions, a partial defence of excessive force in self-defence is recognized, which would negate liability for murder, but result in a conviction for a lesser offence, such as manslaughter.69 The rejection of a partial defence of excessive force in self-defence is a feature even of English law, although in one case it was suggested that police officers approached him from behind, shouted at him to stop and, when he turned round in curiosity, shot and killed him. They claimed to have believed that he was an Irish terrorist about to shoot them. The police officers involved were not prosecuted. The decision not to prosecute has been reviewed on more than one occasion, most recently in October 2005, but the decision has not been changed. In a press release, one of the reasons the Crown Prosecution Service gave for this was that ‘it would be very difficult for prosecutors to disprove the officers honestly believed that they were facing a sawn-off shotgun’ (CPS Press Release, 13 December 2001). The Harry Stanley case is considered in more detail in F Leverick, ‘Is English self-defence law incompatible with the ECHR?’ [2002] Crim LR 347–62 and F Leverick, ‘What has the ECHR done for victims: a United Kingdom perspective’ (2004) 11 International Review of Victimology 177–200. See also R (Stanley) v Her Majesty’s Coroner for Inner North London [2003] EWHC 1180 (where Mr Stanley’s widow successfully compelled the authorities to hold a fresh inquest into her husband’s death) and R (Sharman) v Her Majesty’s Coroner for Inner North London [2005] EWHC 857 (where the verdict of unlawful killing returned at this second inquest was quashed by the Administrative Court). 67 As it was famously stated in Reed v Wastie [1972] Crim LR 221, ‘one does not use jewellers’ scales to measure reasonable force’ (at 221). 68 R v Palmer [1971] AC 814 (England and Wales); Crawford v HM Advocate 1950 JC 67 (Scotland); Zecevic v DPP (1987) 162 CLR 645 (Australia); R v Faid [1983] 1 SCR 265 69 (Canada). See Section 9.3.4 below.

170

Killing in Self-Defence

a mistaken belief that excessive force was necessary did not have to be based on reasonable grounds in order to ground an acquittal. The case in question was R v Scarlett,70 where the landlord of a public house used excessive force in the honest but unreasonable belief it was necessary to eject a drunken and abusive customer. The trial judge directed the jury that excessive force in the removal of an intruder could not ground a defence, regardless of the belief of the defendant. The conviction was overturned on appeal, with Beldam LJ stating that: If the mental element necessary to prove an assault is an intention to apply unlawful force to the victim, and the accused is to be judged according to his mistaken view of the facts whether that mistake was on an objective view reasonable or not, we can see no logical basis for distinguishing between a person who objectively is not justified in using force at all but mistakenly believes he is and another who is in fact justified in using force but mistakenly believes that the circumstances call for a degree of force objectively regarded as unnecessary.71

It soon became apparent, though, that English law was not going to go down this path after all, as Collins J made clear in R v Owino:72 The jury have to decide whether a defendant honestly believed that the circumstances were such as required him to use force to defend himself from an attack or a threatened attack. In this respect a defendant must be judged in accordance with his honest belief, even though that belief may have been mistaken. But the jury must then decide whether the force used was reasonable in the circumstances as he believed them to be.73

This does mean that the approach taken by English law is something of a logically confused one. A mistake about the existence of an attack does not need to be reasonable but a mistake in relation to the amount of force required to repel that attack does require a reasonable basis.74 In R v Hughes,75 the same judge who delivered the Scarlett judgment appears to have changed his mind and outlines the law exactly as stated in Owino, and Owino is also followed in DPP v Armstrong-Braun.76 The Owino test has subsequently been affirmed by the Privy Council in R v Shaw77 and by the Court of Appeal in R v Martin.78

70

71 72 [1993] 4 All ER 629. ibid 636. [1996] 2 Cr App R 128. ibid 132–3. 74 English law is also inconsistent in that it requires all mistakes in defence induced by self-induced intoxication to be reasonable, including mistake in the perception of an attack. See R v O’Grady [1987] QB 995, where it seems that the Court of Appeal realized that the logical consequence of Williams would be to allow unreasonable intoxicated mistake and for public policy reasons drew back from this outcome. The approach taken in O’Grady, where the defendant was convicted of manslaughter, was followed in R v Hatton [2005] EWCA Crim 2951, where the Court of Appeal confirmed that O’Grady also applied to murder. 75 76 77 [1995] Crim LR 956. [1999] Crim LR 416. [2001] UKPC 26. 78 [2001] EWCA Crim 2245. 73

Mistake

171

9.3.2 Should mistakes about excessive force be required to be reasonable? It is suggested that mistakes in relation to excessive force should be required to be reasonable if they are to ground an acquittal. The overriding concern of this book is with protecting the right to life of all persons, even those who commit crimes, wherever this is possible. In order to achieve this, the law should require that an aggressor be repelled with the minimum amount of force necessary to prevent the attack and should allow only reasonable mistake in relation to this. If the aggressor does not threaten the accused’s life, her own right to life is not forfeited, and the accused who unreasonably uses lethal force to ward off that aggressor violates the aggressor’s right to life for no good reason and should be potentially subject to at least some punishment. The same objection can be made to this conclusion as was made in relation to mistakes about the existence of an attack: that punishment serves no purpose, as the accused who honestly but mistakenly believes that a certain (excessive) level of force is necessary will not be deterred from doing so by the criminal law. The same response can also be made. Punishment would serve both an expressive and a conduct-guiding function. Strangely enough, it seems to be accepted by some who argue in favour of allowing unreasonable mistake in relation to the existence of an attack that an unreasonable belief should not be permitted in relation to the level of force necessary to repel an attack. The English Law Commission,79 for example, has argued that unreasonable mistaken belief should be permitted in relation to the existence of an attack, stating that: In such circumstances it would be unjust . . . if A, providing he did no more than would have been reasonably required to avoid an unexpected attack on himself, and not in a spirit of aggression or revenge, were to be exposed to criminal liability simply because of his mistaken or even negligent belief.80

The Law Commission go on, however, to conclude that unreasonable mistake should not be permitted in relation to the degree of force necessary to repel an attack, explaining this as follows: It is not for the defendant himself to adjudicate upon the reasonableness of the steps that he takes to prevent [an attack], because that would unfairly and dangerously exculpate defendants who had an irresponsible, irrational or anti-social notion of the extent to which it is acceptable to react when threatened with attack.81

I think the Law Commission is correct in this, but it does seem logically to contradict the earlier statement that unreasonable mistaken belief should be permitted in relation to the existence of an attack. This is, of course, precisely 79 Law Commission, Legislating the Criminal Code: Offences Against the Person and 80 General Principles (Law Com No 218, 1993). ibid 66. 81 ibid 66–7.

172

Killing in Self-Defence

the approach taken by English law at present: while an unreasonable mistake about the existence of an attack can serve to ground an acquittal, only a reasonable mistake about the amount of force necessary to repel that attack can do likewise. But surely the defendant who reacts rashly to a perceived ‘attack’ in circumstances that any reasonable person would have realized did not constitute a threat is equally irresponsible, irrational, and anti-social? After all, if the mistaken defendant kills her attacker, both situations result in the death of a human being who, if the defendant had taken more care to evaluate the situation, would still be alive.

9.3.3 A partial defence? It is argued here, then, that unreasonable mistake in relation to both the existence of an attack and the degree of force permitted to repel that attack should not ground an acquittal on the basis of self-defence. That is not to say, though, that a conviction for murder should necessarily follow. Another option is to permit excessive force in self-defence to operate as a partial defence, reducing murder to manslaughter.82 While this is an option that has on more than one occasion been recommended by law reform bodies,83 it is rare to find a common law jurisdiction that operates it in practice. In Canada, the Supreme Court ruled out its availability in R v Faid,84 a decision confirmed in R v Reilly.85 In England, it has been clear since R v Palmer86 that excessive force in self-defence is not a partial defence, this being reaffirmed by the House of Lords in R v Clegg87 and the Court of Appeal in R v Martin.88 The English Law Commission’s 2004 review of partial defences to murder recommended that this remain the case89 but that where there has been an excessive response to a fear of violence, this should be covered by the defence of 82 Or an equivalent offence, such as culpable homicide in Scotland. This discussion is predicated on the assumption that the original charge is one of murder. Where the original charge is one of assault, the relevant question is whether the fact that the accused acted in self-defence (albeit using excessive force) should operate as a mitigating factor in punishment. 83 Criminal Law Revision Committee 14th Report, Offences against the Person (Cmnd 7844, 1980), at para 288; Report of the House of Lords Select Committee on Murder and Life Imprisonment, House of Lords Session 1988–9, HL Paper 78–1, at para 89; Law Commission, Criminal Law: A Code for England and Wales (Law Com No 177, 1989), at clause 59. Cf Law Reform Commission of Canada, Homicide (Working Paper 33, 1984). 84 [1983] 1 SCR 265, per Dickson J at [9]–[19]. 85 86 [1984] 2 SCR 396, per Ritchie J at [5]–[10]. [1971] AC 814. 87 [1995] 1 AC 482. For criticism, see M Kaye, ‘Excessive force in self-defence after R v Clegg’ (1997) 61 JCL 448–57. 88 [2001] EWCA Crim 2245. Cf R v Scarlett [1993] 4 All ER 629, per Beldam LJ at 636 (discussed in Section 9.3.1 above). 89 Report on Partial Defences to Murder (Law Com No 290, 2004), at para 4.31. For comment, see SSM Edwards, ‘Abolishing provocation and reframing self-defence: the Law Commission’s options for reform’ [2004] Crim LR 181–97; C Elliott, ‘What future for voluntary manslaughter?’ (2004) 68 JCL 253–63.

Mistake

173

provocation.90 The partial defence is also ruled out by statute in New Zealand.91 In Scotland, despite some historical confusion,92 it seems settled that there is no partial defence of excessive force in self-defence,93 although some doubt might be cast on this conclusion by the Appeal Court’s definition of murder in Drury v HM Advocate.94 If, as it was held in Drury, murder requires a wicked intention (that is, one which excludes action taken in self-defence), then it is possible that the accused who honestly believes that she is acting in self-defence, albeit using excessive force, is not acting with wicked intent and will therefore not have the requisite mens rea for murder.95 A rare example of a common law jurisdiction where excessive force in self-defence does operate as a partial defence to murder is India, where it has been in the Indian Penal Code since its inception.96 At various times, a partial defence of excessive force in self-defence has also operated in Australian common law. The partial defence was recognized in R v McKay97 and R v Howe,98 but some doubt over its existence was raised by the Privy Council case of R v Palmer,99 which reached the opposite conclusion. In Viro v R,100 however, the High Court of Australia, in a unanimous decision, held that a defendant who had acted in self-defence, but used excessive force in doing so, should be convicted not of murder but of manslaughter. This recognition of the partial defence was only temporary. Viro was overturned nine years later by a five–two majority decision of the High Court in Zecevic v DPP,101 although this was not primarily for reasons of principle, but because the defence was thought to be too complex for juries to understand.102 Some Australian states have since reintroduced the partial defence of excessive force through legislation.103

9.3.4 Should there be a partial defence of excessive force in self-defence? The main argument for permitting excessive force in self-defence to operate as a partial defence is that the moral culpability of someone who kills 90

Report on Partial Defences to Murder, part 3. The Crimes Act 1961 ss 48 and 62, as amended by the Crimes Amendment Act 1980. 92 Hillan v HM Advocate 1937 JC 53, at 58; HM Advocate v Kizileviczius 1938 JC 60, at 63. 93 Crawford v HM Advocate 1950 JC 67, at 69; Fenning v HM Advocate 1985 JC 76, at 80; 94 Low v HM Advocate 1994 SLT 277, at 286. 2001 SLT 1013. 95 Although cf Bennett v Criminal Injuries Compensation Appeal Panel (9 December 2003, Court of Session), at [18], where it is suggested that wicked intent might be inferred in such circumstances. For discussion, see Chalmers and Leverick, Criminal Defences, chapter 3. 96 97 98 See exception 2 to s 300. [1957] ALR 648. (1958) 100 CLR 448. 99 100 101 [1971] AC 814. (1978) 141 CLR 88. (1987) 162 CLR 645. 102 See Section 9.3.4 below. For criticism of the decision, see S Yeo, ‘Applying excuse theory to excessive self-defence’, in S Yeo (ed) Partial Excuses to Murder (1990) 158–77; D Lanham, ‘Death of a qualified defence?’ (1988) 104 LQR 239–49. 103 In South Australia, see s 15(2) of the Criminal Law Consolidation Act 1935 (SA), as amended by the Criminal Law Consolidation (Self Defence) Amendment Act 1997 (SA). In New South Wales, see s 421 of the Crimes Act 1900 (NSW), as amended by the Crimes Amendment 91

174

Killing in Self-Defence

as a result of making an honest mistake in relation to either the existence of an attack or the amount of force necessary to repel that attack is less than that of the accused who kills in a cold-blooded and premeditated fashion.104 The mistaken accused is, after all, acting in the heat of the moment and under the pressure of urgency. One way in which the criminal law could take account of this is to convict the accused of murder, but take account of the circumstances of the killing as a mitigating factor in the sentencing process. In jurisdictions where there is a mandatory life sentence for murder,105 the opportunity for doing this is limited at present, but if the mandatory life sentence were to be abolished,106 this would not be the case.107 Even if it was possible for sentencing to reflect the moral blameworthiness of the accused who has used excessive force in self-defence, there is still an argument for allowing excessive force in self-defence to operate as a partial defence. There is a labelling issue at stake as well as the issue of the appropriate sentence to pass. It may not be appropriate to label the accused who makes a mistake in self-defence as a murderer, except perhaps where the force used was so grossly excessive as to be incompatible with a claim that the accused was acting in self-defence at all. The label of murder should be reserved for the most serious and morally blameworthy examples of killing.108 This was precisely the argument used by the High Court of Australia in Viro v R,109 the case which briefly brought the partial defence of excessive force in self-defence into Australian law. As Aickin J stated: . . . there is a real distinction in the degree of culpability of an accused who has killed having formed the requisite intention without any mitigating circumstance, and an

(Self-Defence) Act 2001 (NSW). For a detailed survey of the approach taken to the partial defence by common law jurisdictions, see the Appendices to the Law Commission’s Consultation Paper, Partial Defences to Murder (Law Com CP No 173, 2003). 104 Kaye, ‘Excessive force in self-defence’, at 452–3; P Smith, ‘Excessive defence: a rejection of Australian initiative?’ [1972] Crim LR 524–34, at 533–4; AP Simester and GR Sullivan, Criminal Law: Theory and Doctrine (2nd edn, 2004), at 626; LH Leigh, ‘Manslaughter and the limits of self-defence’ (1971) 34 MLR 685–90, at 689; S Yeo, ‘Revisiting excessive self-defence’ (2000–1) 12 Current Issues in Criminal Justice 39–57, at 41. 105 Criminal Procedure (Scotland) Act 1995, s 205(1); Murder (Abolition of Death Penalty) Act 1965 s 1(1) (England and Wales). 106 As it has been in, for example, some Australian jurisdictions, including New South Wales (see the Crimes (Homicide) Amendment Act 1982) and Victoria (see s 8 of the Crimes (Amendment) Act 1986). 107 The abolition of the mandatory life sentence in England was recommended by the House of Lords Select Committee on Murder and Life Imprisonment (HL Paper, 78–1, 1989) but not implemented. The Law Commission’s review of the definition of the law on murder was required to assume the continuing existence of the mandatory life sentence for murder (see the resulting consultation paper A New Homicide Act for England and Wales? (Law Com CP No 177, 2005), at para 1.1). 108 Yeo, ‘Revisiting excessive self-defence’, at 41–2; G Woods, ‘The sanctity of murder: reforming the homicide penalty in New South Wales’ (1983) 57 Australian Law Journal 161–6. 109 (1978) 141 CLR 88.

Mistake

175

accused who, in response to a real or a reasonably apprehended attack, strikes a blow in order to defend himself, but uses force beyond that required by the occasion and thereby kills the attacker. Such a killing is undoubtedly unlawful but it appears to me to differ significantly from murder.110

In addition to the culpability argument, the Law Commission has pointed out that the absence of a partial defence may not only result in harsh convictions, but in inappropriate acquittals because juries, faced with the all or nothing choice of a murder conviction or an acquittal, might opt for the latter in cases where excessive force has been used.111 On the other hand, as the Law Commission recognize, this argument can be turned on its head. If a partial defence did operate, in borderline cases juries might be inclined to opt for the middle ground of a manslaughter conviction instead of a complete acquittal.112 The main arguments put forward against a partial defence of excessive force in self-defence are twofold. First, it is sometimes said that it is unnecessary, given that the use of force in self-defence is not judged on too fine a scale and that evidence that the defendant only did what he honestly believed was necessary in the heat of the moment is likely to be evidence that the level of force used was, in fact, reasonable. As the Privy Council put it in R v Palmer:113 If there had been an attack so that defence is reasonably necessary it will be recognised that a person defending himself cannot weigh to a nicety the exact measure of his necessary defensive action. If a jury thought that in a moment of unexpected anguish a person attacked had only done what he honestly and instinctively thought was necessary that would be the most potent evidence that only reasonable defensive action had been taken.114

In addition, the accused who has used excessive force in self-defence may be able to take advantage of the partial defence of provocation if there is evidence of a loss of self-control.115 It has also been said that the recognition of a partial defence of excessive force in self-defence would make the law too complex and the task of charging the jury too difficult. This, and not any moral objection to the partial defence, was the main reason why 110 ibid 180. See also Stephen J at 133 and Mason J at 146 in Viro and the dissenting opinions of Deane J and Gaudron J in Zecevic v DPP (1987) 162 CLR 645, the Australian High Court decision that overturned Viro. 111 Consultation Paper on Partial Defences to Murder, para 12.84. 112 113 ibid, para 12.85. [1971] AC 814. 114 Palmer, per Lord Morris of Borth-y-Gest at 832. 115 R v McInnes [1971] 1 WLR 1600, per Lord Edmund Davies at 1608. See also the Law Commission’s Consultation Paper on Partial Defences to Murder, at para 9.9. In their eventual report, the Law Commission recommended reform of the provocation defence in English law so that it is available if the defendant acted in response to ‘fear of serious violence towards the defendant or another’ if a person of the defendant’s age and of ordinary temperament in the circumstances of the defendant might have reacted in the same or a similar way (at para 3.168). This, they state, would make the need for a partial defence of excessive force in self-defence less pressing (at para 4.29).

176

Killing in Self-Defence

the High Court of Australia overruled Viro in Zecevic and returned to the position where self-defence could operate only as a complete acquittal.116 Neither of these arguments is particularly convincing. Not every situation where excessive force has been used will be covered by either the wide allowance made for the heat of the moment or, failing this, by the provocation doctrine. Likewise, the difficulties of formulating a jury direction could equally apply to many of the other doctrines that are recognized in the criminal law and could surely be overcome.117 It is argued here that the accused who uses excessive force in self-defence has displayed morally blameworthy behaviour and should not receive the benefit of a complete acquittal. A conviction for murder, though, is unnecessarily harsh. As Lacey has stated, someone who reacts excessively in self-defence admittedly departs from the normal standards of practical reasonableness but this departure is morally distinguishable from that of the accused who is not under pressure of urgency and immediacy. Certainly there seems to be no less compelling reason for accepting this as a partial defence compared to others that are already accepted, such as provocation and diminished responsibility.118 116 Zecevic v DPP (1987) 162 CLR 645, per Mason CJ at 653. This was not surprising, given that the rules on excessive force in self-defence set out in Viro were unnecessarily confusing and involved several double negatives. On this, see Lanham, ‘Death of a qualified defence?’, at 240; Yeo, ‘Revisiting excessive self-defence’, at 45. 117 For some suggestions as to how this might be achieved, see Yeo, ‘Revisiting excessive self-defence’, at 45–50. 118 N Lacey, ‘Partial defences to homicide: questions of power and principle in imperfect and less imperfect worlds’, in A Ashworth and B Mitchell (eds), Rethinking English Homicide Law (2000) 107–32, at 124–5; J Horder, Excusing Crime (2004), at 57.

10 The Impact of the European Convention on Human Rights 10.1 Introduction This final chapter examines the implications that the European Convention on Human Rights1 might have for the law of self-defence.2 Given that our concern throughout has been specifically with self-defensive killing, the focus is on Article 2 of the ECHR, which protects the right to life, but that is not to say that Article 3, the right to freedom from inhuman and degrading treatment, may not have similar implications in relation to non-lethal force. Thus far, this text has taken a comparative approach, drawing on the law of all the major common law jurisdictions (and occasionally civil law jurisdictions). In assessing the impact of the ECHR, however, it is the UK that is placed in particular focus, as other common law jurisdictions are not bound by its provisions. That is not to say that this chapter has no relevance beyond the UK context. The right to life is given constitutional protection in, for example, the Canadian Charter of Rights and Freedoms, the South African Constitution and the US Bill of Rights. While importing concepts from one jurisdiction’s constitutional jurisprudence to another is something that must always be done with care, there are clearly parallels that can be drawn between the jurisprudence of the European Court of Human Rights3 and other jurisdictions where the right to life is similarly protected.

10.2 Article 2 of the ECHR 10.2.1 Article 2 generally Article 2(1) of the ECHR provides that: ‘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 1

Subsequently ‘the ECHR’. This chapter draws in part on F Leverick, ‘Is English self-defence law incompatible with the ECHR?’, which was first published in the Criminal Law Review ([2002] Crim LR 347–62). Some passages from this article are reproduced with the kind permission of the publisher, 3 Subsequently ‘the European Court’. Sweet & Maxwell Ltd. 2

178

Killing in Self-Defence

which this penalty is provided by law.’ At the outset, it can be said that Article 2 of the ECHR may be relevant to cases of killing in self-defence in a number of ways, both procedurally and regarding the substantive criminal law. In relation to the former, it is well established that there is a procedural obligation associated with Article 2 that places a duty upon the state to undertake an effective investigation where there has been a deprivation of life as a result of the use of force. This was set out in McCann v United Kingdom4 and developed in Jordan v United Kingdom.5 Any investigation must satisfy certain requirements if it is to comply with Article 2. These include a requirement for independence in terms of the persons responsible for undertaking the investigation to be independent from those implicated in the events;6 a requirement that the investigation is capable of leading to some form of determination of whether or not the force used was justified and ‘to the identification and punishment of those responsible’;7 a requirement that the authorities must have taken all reasonable steps to secure evidence relating to the events in question;8 a requirement that the investigation is undertaken in good time;9 and a requirement for public scrutiny and accountability and for victim involvement,10 whereby ‘the next-of-kin of the victim must be involved in the procedure to the extent necessary to safeguard his or her legitimate interests’.11 As well as the general obligation to investigate, the European Court has read into Article 2 a duty for the state to give reasons for decisions not to prosecute, albeit in limited circumstances. This stemmed from Jordan but was set out in more detail in Finucane v United Kingdom,12 where it was held that a failure to give reasons for a decision not to prosecute in a case where a death had occurred could constitute a violation of Article 2. The duty to give reasons was limited to ‘controversial cases’.13 The domestic courts arrived at the same conclusion in R v DPP ex p Manning,14 although here the duty to give reasons was limited to cases where an inquest jury, following a proper direction, has returned a verdict of unlawful killing implicating an 4 (1996) 21 EHRR 97. Ribisch v Austria (1995) 21 EHRR 573 sets out a similar obligation 5 in relation to Art 3. (2003) 37 EHRR 2. 6 7 8 9 Jordan, at [106]. ibid, at [107]. ibid, at [107]. ibid, at [108]. 10 ibid, at [109]. 11 ibid, at [109]. The Jordan requirements were repeated in McKerr v United Kingdom (2002) 24 EHRR 20, at [111]–[115]; Finucane v United Kingdom (2003) 37 EHRR 29, at [67]–[71]; Nachova v Bulgaria (2004) 39 EHRR 37, at [116]–[119]; and Isayeva v Russia (2005) 41 EHRR 38, at [209]–[214]. They were approved by the House of Lords in R (Amin) v Secretary of State for the Home Department [2003] UKHL 51 (at [20]) and have been applied in a number of English cases. See R (Wright) v Secretary of State for the Home Department [2001] EWHC Admin 520; R (Rowley) v DPP [2003] EWHC 693; R (Middleton) v West Somerset Coroner [2004] UKHL 10. For a more detailed discussion of the procedural obligations derived from Art 2, see F Leverick, ‘What has the ECHR done for victims? A United Kingdom perspective’ (2004) 11 International Review of Victimology 177–200, at 182–8. 12 13 14 (2003) 37 EHRR 29. At [82]. [2001] QB 330.

The Impact of the ECHR

179

identifiable person whose whereabouts are known and a prosecution does not follow.15 Our primary interest, however, is not in the procedural implications of Article 2, but is in its possible impact on the substantive law of self-defence. Here we must turn to the exceptions to Article 2(1) contained in Article 2(2).

10.2.2 The exceptions in Article 2(2) Article 2(2) provides that: 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 detained; (c) in action lawfully taken for the purpose of quelling a riot or insurrection.

Two issues arise. The first is that there is no reference to the defence of property in Article 2(2) as a purpose for which lethal force is permissible and thus, on the face of it, any jurisdiction within the United Kingdom that does not contain criminal sanctions for those who kill to protect property risks contravening Article 2. The second is the issue of mistaken belief. Here, clause (a) is relevant, which permits the use of lethal force in self-defence or the defence of others, but only when ‘absolutely necessary’. As Chapter 9 described, English selfdefence law allows an unreasonable belief in the existence of an attack to ground an acquittal on the basis of self-defence.16 The defendant who has no good reason for her belief that she is facing an attack is hardly undertaking a deprivation of life that is ‘absolutely necessary’. Thus, in failing to attach criminal sanctions to the defendant who kills in the entirely unreasonable belief that she is being attacked, it might be said that English criminal law fails to protect the right to life of the person who is mistaken for an attacker. Before either of these points can be addressed, however, some preliminary comment on whether killing in self-defence would even be governed by Article 2 is necessary.

10.2.3 Does self-defensive killing fall under the scope of Article 2? It has occasionally been suggested that English self-defence law would not be caught by Article 2 because action taken in self-defence does not constitute an intentional killing. After all, Article 2 provides that no one shall be 15 At [33]. For a more detailed discussion of the Art 2 derived requirement to give reasons, see Leverick, ‘What has the ECHR done for victims?’, at 186–8; M Burton, ‘Reviewing Crown Prosecution Service decisions not to prosecute’ [2001] Crim LR 374–84. 16 This approach is not shared by other common law jurisdictions such as Scotland, Canada, and Australia, where any mistaken belief must be reasonable (see Chapter 9).

180

Killing in Self-Defence

deprived of her life ‘intentionally’ and, as Smith put it, ‘[t]he person acting in good faith in private defence does not have the purpose of killing or causing injury. His sole purpose is to preserve his own or another’s life or safety. He may intend to kill for the purposes of English law, but not, it seems, for the purpose of the Convention.’17 Smith cited no authority for his view and in response at least two points might be made. First, unless one subscribes to the doctrine of double effect,18 it is surely incorrect to say that the person acting in self-defence does not have the purpose of killing or causing injury. This may be true of some selfdefenders, but examples can easily be constructed where it seems absurd to say that the accused did not intend to kill.19 Second, even if it is accepted that killings in self-defence are not intentional, they are still covered by Article 2. Aside from the specific provision that ‘no-one shall be deprived of his life intentionally’, the first sentence of Article 2 imposes a general obligation on the state to protect the right to life.20 This would seem to suggest that Article 2 is not concerned solely with intentional killing but places an obligation on the state to protect the individual from any unjust deprivation of life, intentional or not. As the Court put it in LCB v United Kingdom,21 ‘[t]he first sentence of Article 2(1) enjoins the State not only to refrain from the intentional and unlawful taking of life, but also to take appropriate steps to safeguard the lives of those within its jurisdiction’.22 This interpretation is supported by an examination of the case law on Article 2. Admittedly, in X v Belgium,23 a Commission decision of 1969 which involved the shooting of a demonstrator by a police officer, the case was declared inadmissible on the basis that it was not an intentional killing. All subsequent decisions and judgments, however, have conclusively rejected this interpretation. The first case to do so explicitly was Stewart v United Kingdom,24 where the Commission concluded that the exceptions in Article 2(2) are: . . . not concerned exclusively with intentional killing. Any other interpretation would hardly be consistent with the object and purpose of the Convention or with a strict interpretation of the general obligation to protect the right to life. In the Commission’s opinion, the text of Article 2, read as a whole, indicates that paragraph 2 does not primarily define situations where it is permitted intentionally to kill an 17 J Smith, ‘Surgical Separation: Case Comment on Re: A (Children)’ [2001] Crim LR 400–5, at 403. See also J Smith, ‘The use of force in public or private defence and Article 2’ [2002] Crim 18 LR 958–62. See Chapter 3. 19 For example, the case of Andronicou and Constantinou v Cyprus (1998) 25 EHRR 491 (discussed in detail in Section 10.4.3 below) where the authorities fired multiple machine gun rounds towards the deceased at close range in the belief that he was about to shoot them. 20 Art 2 commences with the sentence ‘everyone’s right to life shall be protected by law’ before continuing ‘no-one shall be deprived of his life intentionally’. 21 (1999) 27 EHRR 212. 22 At [36], emphasis added. This passage is repeated in Osman v United Kingdom (2000) 29 EHRR 245, at [115]; Kiliç v Turkey (2001) 33 EHRR 58, at [62]; Makaratatzis v Greece (2005) 23 41 EHRR 49, at [57]. (1969) 12 Yearbook 174. 24 (1985) 39 DR 162.

The Impact of the ECHR

181

individual, but defines the situation where it is permissible to ‘use force’ which may result, as the unintended outcome of the use of force, in the deprivation of life.25

In all subsequent cases, the Court has followed the Commission and held that killings undertaken in self-defence and the defence of others are covered by Article 2, regardless of the precise nature of the intention of the perpetrator.26 Indeed, in Makaratatzis v Greece,27 the actions of police officers who opened fire on the applicant after he drove through a roadblock fell to be considered under Article 2 even though it was accepted by the Court that they did not intend to kill him (and in fact they did not kill him—he survived).28 With this issue disposed of, we can turn to an examination of two substantive issues: killing in defence of property and the mistaken belief in the need to kill in self-defence.

10.3 Defence of Property and Article 2 As we have seen in Chapter 7, there are no specific rules in English law on killing to defend property. There is certainly no guarantee that a particular jury or trier of fact might not find it reasonable to have used lethal force to protect property and to acquit a defendant on this basis. Under the ECHR, however, the position seems clear: the use of lethal force by state agents29 to protect property would constitute a violation of Article 2. Article 2 permits the use of lethal force for the purposes set out in Article 2(2).30 There is no reference to the defence of property in Article 2(2). The European Court has made it clear that Article 2 is a fundamental provision that is to be strictly construed. For example, in McCann v United Kingdom,31 the Court stated that: It must also be borne in mind that, as a provision which not only safeguards the right to life but sets out circumstances when the deprivation of life may be justified, Article 2 ranks as one of the most fundamental provisions in the Convention—indeed one which, in peacetime, admits of no derogation under Article 15. Together with Article 3 of the Convention, it also enshrines one of the basic values of democratic societies . . . As such, its provisions must be strictly construed.32 25

ibid, at [15]. The passage from Stewart is repeated in McCann v United Kingdom (1996) 21 EHRR 97, at [148]; Andronicou and Constantinou v Cyprus (1998) 25 EHRR 491, at [160]; Gül v Turkey (2002) 34 EHRR 28, at [77]; McKerr v United Kingdom (2002) 34 EHRR 20, at [110]; McShane v United Kingdom (2002) 35 EHRR 23, at [101]; Nachova v Bulgaria (2004) 39 27 EHRR 37, at [93]. (2005) 41 EHRR 49. 28 29 At [55]. For the position of private citizens, see Section 10.5 below. 30 31 See Section 10.2.2 above. (1996) 21 EHRR 97. 32 At [147]. This passage has consistently been repeated in subsequent cases: see Andronicou and Constantinou v Cyprus (1998) 25 EHRR 491, at [171]; Sabuktekin v Turkey (2003) 36 EHRR 19, at [91]; Aktas v Turkey (2004) 38 EHRR 13, at [289]. 26

182

Killing in Self-Defence

This would suggest that, because it is not specifically provided for in Article 2(2), the use by state agents of lethal force to protect property would constitute a violation of the ECHR. The issue has never been directly addressed by the Court. An opportunity to do so arose in Ramsahai v The Netherlands,33 where the applicant claimed that her son was shot dead by a police officer after he stole a scooter. The Court did not accept this interpretation of the facts, however, and held that the deceased was shot only after he raised a loaded pistol towards the police.34 Thus the shooting was justified on the basis of preventing a perceived threat to life.35 The case law on the degree of force permitted to effect an arrest would seem to suggest that lethal force is permissible only in defence of life or to prevent bodily injury. In Nachova v Bulgaria,36 for example, it was held that the use of lethal force by state agents to prevent the escape of individuals who did not pose a threat of violence constituted a violation of Article 2. Here, the two deceased were serving sentences for being absent without leave from military service. They escaped (without resorting to violence; they were working outside the prison and simply walked away). They were not armed and had no record of violent offending. They were both shot and killed by military officers to prevent them escaping. This was held to be a violation of Article 2, on the basis that: . . . it can in no circumstances be ‘absolutely necessary’ within the meaning of Article 2(2) of the Convention to use such firearms to arrest a person suspected of a nonviolent offence who is known not to pose a threat to life or limb, even where a failure to do so may result in the opportunity to arrest the fugitive being lost.37

After a request for referral by the Bulgarian government,38 the violation of Article 2 was confirmed by the Grand Chamber of the European Court, which concluded that: . . . with reference to [Article 2(2)(b)] the legitimate aim of effecting a lawful arrest can only justify putting human life at risk in circumstances of absolute necessity. The Court considers that in principle there can be no such necessity where it is known that the person to be arrested poses no threat to life or limb and is not suspected of having committed a violent offence, even if a failure to use lethal force may result in the opportunity to arrest the fugitive being lost.39

Likewise, in Kakoulli v Turkey,40 the Court held that lethal force is not permissible under Article 2 where ‘it is known that the person arrested poses no threat to life or limb and is not suspected of having committed a violent 33

34 (App no 52391/99) ECHR, 10 November 2005. At [378]–[380]. 36 37 At [383]. (2004) 39 EHRR 37. At [105]. 38 A procedure permitted under Art 43 of the ECHR. In fact, the Bulgarian government had accepted the European Court’s findings in relation to Art 2 (but were contesting an issue relating to Art 14). Under Art 43, however, if a case is referred to a Grand Chamber, all of the substantive 39 issues are considered afresh. Nachova v Bulgaria [2005] ECHR 465, at [95]. 40 (App no 38595/97) ECHR, 22 November 2005. 35

The Impact of the ECHR

183

offence’.41 In this case, the Turkish authorities shot and killed the deceased, who was in the border area between the Turkish and Greek controlled areas of Cyprus, after he failed to stop when asked to do so. This does seem a very clear indication that killing is not permissible under Article 2 where no ‘threat to life or limb’ is involved.42 Thus the use of lethal force by state agents to protect property would constitute a violation of Article 2.

10.4 Mistake about the Need to Use Self-Defensive Force and Article 2 English self-defence law permits an unreasonable belief in the need to use self-defensive force to ground an acquittal, assuming all the other conditions of the defence are made out. It is argued that this runs contrary to Article 2. As we have already seen, lethal force is permitted under Article 2(2) of the Convention when it is ‘absolutely necessary’ for one of the purposes set out in Article 2(2). This does not mean that reasonable mistake about the need to use lethal force is not permitted under Article 2. In order to understand the approach taken by the European Court on mistaken belief, and why English self-defence law might violate Article 2, it is worth examining the case law in some detail.

10.4.1 The decision of the Commission in Kelly The first case to arise under Article 2(2) in which a killing was actually based on a mistaken belief was Kelly v United Kingdom.43 In Kelly, a 17-year-old joy-rider was shot and killed by soldiers in Northern Ireland in the mistaken belief that he and the other three occupants of the car he was driving were escaping terrorists. In its decision, the Commission began by emphasizing the ‘strict and compelling’ test of ‘absolute necessity’ contained in Article 2(2).44 On this basis, it might be thought that any case involving a mistaken belief cannot possibly satisfy the test of ‘absolute necessity’. After all, the force used was not necessary at all: the individual simply thought that it was.45 The Commission rejected this interpretation and noted the finding of the High Court in Northern Ireland that the soldiers reasonably believed the occupants of the 41

At [108]. M Jefferson, ‘Householders and the use of force against intruders’ (2005) 69 JCL 405–13, at 412; A Ashworth, Principles of Criminal Law (5th edn, 2006), at 139; D Ormerod, Smith and Hogan Criminal Law (11th edn, 2005), at 332; AP Simester and GR Sullivan, Criminal Law: 43 Theory and Doctrine (2nd edn, 2004), at 631. (1993) 16 EHRR CD20. 44 At CD21. 45 This was in fact the view of Mr Loucaides, one of the dissenting Commissioners in the Commission’s admissibility decision in McCann v United Kingdom (1996) 21 EHRR 97. 42

184

Killing in Self-Defence

car to be terrorists.46 The Commission went on to point out that this finding was supported by the facts of the case: the car was stolen; it had been seen in the vicinity of a house and car belonging to a member of the security forces; and the driver of the car made ‘determined and even desperate’47 efforts to evade the checkpoint. Taking these factors into account, the Commission concluded that the force used by the soldiers, despite it being based on a mistaken belief, met the absolute necessity test.48 Nowhere in Kelly does the Commission explicitly state that a mistaken belief must be based on reasonable grounds, although the Commission’s emphasis of the facts supporting the belief of the soldiers suggests that this was their view. In any event, the issue was soon addressed directly in McCann v United Kingdom, the first judgment of the Court under Article 2(2).

10.4.2 The judgment in McCann Like Kelly, McCann v United Kingdom49 involved a killing based on a mistaken belief. The case stemmed from the shooting of three members of a terrorist organization in Gibraltar by soldiers. The soldiers had been told that the three deceased had planted a car bomb (this was not true) and were able to detonate it by remote control, by pressing a single button concealed on their person (even if there was a bomb, it is extremely unlikely that this would have been possible). An inquest in Gibraltar returned a verdict of lawful killing and no criminal proceedings against the soldiers were brought. The European Court considered the question of whether the Gibraltar Constitution was compatible with Article 2. The Gibraltar Constitution then provided that force used to defend a person from violence is permitted if it is ‘reasonably justifiable’. The Court interpreted this phrase according to what they termed ‘UK law’, drawing on the three cases of Lynch,50 Thain,51 and Williams,52 the first two of which originate from Northern Ireland and the last of which is an English case. On the basis of these cases, the Court decided that the domestic law to be assessed was as follows: . . . the reasonableness of the use of force has to be decided on the basis of the facts which the user of force honestly believed to exist: this involves the subjective test as to what the user believed and an objective test as to whether he had reasonable grounds for that belief. Given that honest and reasonable belief, it must then be questioned whether it was reasonable to use the force in question in the prevention of crime or to effect an arrest.53

It is difficult to tell from this paragraph whether the Court thought that domestic law required a mistaken belief to be reasonable or whether they thought that an honest but unreasonable belief would suffice. In the first 46 50 52

47 48 49 At CD22. At CD22. At CD22. (1996) 21 EHRR 97. 51 Lynch v Ministry of Defence [1983] NI 216. R v Thain [1985] NI 457. 53 R v Williams [1987] 3 All ER 411. At [134].

The Impact of the ECHR

185

sentence they seem to conclude that the test of what the accused believed is subjective but in the next sentence, they refer to an ‘honest and reasonable belief’. It may be that their confusion stemmed from the three cases cited. Of these, Lynch requires a mistaken belief to be reasonable. Williams, however, was the first case in English law to allow an honest but unreasonable mistake of fact to ground a claim of self-defence. The reasonableness or otherwise of a mistaken belief was not the central issue in Thain, but the court followed Williams in its statement of the law on the matter. From this basis, the European Court stated that domestic law did not have to be formulated identically to the test in Article 2, as long as ‘the substance of the Convention right was protected by domestic law’54 and, in this case, they concluded that it was. On the basis of this confused assessment, it is difficult to reach any firm conclusions on whether English self-defence law might be incompatible with Article 2. It is impossible to tell whether the Court correctly understood ‘domestic law’ and whether they were approving the honest belief test or not. Later parts of the McCann judgment suggest that they were not. The Court went on to consider the actions of the individual soldiers and concluded that ‘the use of force by agents of the State in pursuit of one of the aims [in Article 2(2)] may be justified under this provision where it is based on an honest belief which is perceived, for good reasons, to be valid at the time but which subsequently turns out to be mistaken’.55 This suggests that English law’s lower standard of mere ‘honest belief’ is contrary to Article 2. The confused discussion of domestic law by the Court in McCann does not, though, allow us to conclude this with any great confidence.56

10.4.3 The judgment in Andronicou The emphasis in McCann on the need for a mistaken belief to be based on good reasons was repeated by the European Court in Andronicou and Constantinou v Cyprus.57 This case stemmed from the actions of the Cypriot authorities in using officers of the MMAD58 to attempt to bring to an end a domestic dispute between a man (Lefteris Andronicou) and his fiancée (Elsie Constantinou). Andronicou had been holding Constantinou hostage in the flat they shared and it was known that he was armed with a double-barrelled shotgun, a weapon capable of firing two bullets before it needed to be reloaded. On the basis of a number of ambiguous statements made by Andronicou, the 54

55 At [152]. At [201], emphasis added. The claim that the individual soldiers acted contrary to Art 2 was eventually rejected, the Court concluding that their belief that the suspects were about to detonate a bomb was indeed based on good reasons. The United Kingdom was, though, found to have violated Art 2 in 57 respect of its planning and control of the operation. (1998) 25 EHRR 491. 58 Mihanokiniti Monada Amesis Derasis, a Cypriot government unit trained specifically to shoot to kill. 56

186

Killing in Self-Defence

authorities formed the view that he would kill Constantinou and himself at midnight and, as a result, the MMAD, armed with machine guns, stormed the flat. They found Andronicou in a small room, using Constantinou as a human shield. Andronicou fired two shots, one of which hit an MMAD officer. The remaining MMAD officers returned fire with their machine guns, killing both Andronicou and Constantinou. At least some of the shots were fired when Andronicou was on the ground and no longer holding the shotgun (which by this time was no longer a threat anyway as its two bullets had been discharged). He had no other weapons, although in explaining their actions the MMAD officers claimed that they thought he was still armed. The Court concluded that no violation of Article 2 had occurred,59 stating that: . . . the Court cannot with detached reflection substitute its own assessment of the situation for that of the officers who were required to react in the heat of the moment in what was for them a unique and unprecedented operation to save life. The officers were entitled to open fire for this purpose and to take all measures which they honestly and reasonably believed were necessary to eliminate any risk either to the young woman’s life or to their own lives.60

No consideration was given specifically to domestic criminal law but once again the Court stressed the need for a mistaken belief to have some reasonable basis.

10.4.4 The missed opportunity in Caraher Andronicou was fairly quickly followed by Caraher v United Kingdom,61 an admissibility decision in which a direct opportunity to consider English law’s honest belief standard was frustratingly missed by the European Court. The application stemmed from the shooting of two men by soldiers in Northern Ireland. The soldiers had opened fire on a car that was driving away from a car park, killing one man and injuring the other. The soldiers who fired the shots claimed that the car had driven off with another soldier on the bonnet and they were firing in order to prevent injury to him. There was no suggestion that any of the men in the car were armed or terrorists. 59 A surprising decision, given that the operation to end the siege suffered from numerous flaws. To list but a few: despite Andronicou’s well-known distrust of the police, negotiations were undertaken by an untrained police officer, who was heard to threaten Andronicou on a number of occasions; the area around the flat was not sealed off during negotiations, nor were the telephone calls to and from the flat limited (this meant Andronicou could make and receive telephone calls during the negotiations as he chose); a plan to administer soporific drugs in food sent into the flat failed because the drug used took two hours to take effect (an alternative drug would have worked within ten minutes). It was a narrow decision (five votes to four) and it differed from that of the Commission, who concluded that there had been a violation of Art 2. 60 At [192], emphasis added. 61 (App no 24520/94) ECHR, 11 January 2000. See the case comment in (2000) 3 EHRLR 326–9.

The Impact of the ECHR

187

The two soldiers who fired the shots were charged with murder but acquitted at trial. The application required the European Court to consider whether the standard governing the use of lethal force in domestic law was weaker than the Convention standard under Article 2(2). Unfortunately, the Court in doing so entirely misunderstood the nature of English law. The trial took place in Northern Ireland and it is clear from the report of the case62 that the trial judge, Hutton LCJ, was proceeding on the basis that an honest but unreasonable belief in imminent danger would suffice to ground a plea of self-defence or the defence of others. In his judgment, he surveyed the authorities and concluded that ‘in deciding whether the Crown has proved that the force used was reasonable beyond a reasonable doubt, the tribunal of fact must have regard not to what was actually happening but to what the accused at the time honestly thought was happening’. He continued, stating quite clearly that, in the light of Williams and Beckford:63 ‘the issue is to be determined in the light of the facts as the accused honestly believes them to be, whether his belief was reasonable or unreasonable’. The European Court, however, proceeded on an entirely different understanding of English law.64 They concluded that English law was not substantially divergent from the ‘good reasons’ test in McCann, stating that ‘the approach taken by the domestic judge in the trial, in having regard to the honest and reasonable belief of the two soldiers that one of their colleagues was at risk . . . is compatible with the principles established in McCann’.65 With respect, this was not the standard that the trial judge applied. Hutton LCJ was working on the basis of an honest belief alone, not the ‘honest and reasonable belief’ standard assumed by the European Court. The misunderstanding seems to have arisen because the Court focused on one part of the judgment to the exclusion of all other parts and indeed took this passage completely out of context. In the passage thought by the Court to be a pronouncement on English self-defence law, Hutton LCJ stated that: . . . there was a reasonable possibility that Marine B was carried away on the bonnet of the [car] and that, in the emergency of the moment, there was a reasonable possibility that the two accused fired at the driver because they honestly believed it was necessary to do so to save Marine B from death or serious injury and that in the circumstances as the accused honestly believed them to be there was a reasonable possibility that this constituted reasonable force.

Taken out of context, it is possible to see why the Court took this to mean that English law requires a mistake of fact to be reasonable. Read in the context of 62

R v Elkington (21 January 1993, LEXIS transcript). For discussion of these cases, see Chapter 9. 64 Thus when David Ormerod states that ‘the European Court has not condemned English law which it has had the opportunity to do’ (Ormerod, Smith and Hogan Criminal Law, at 332), he tells only half the story. The opportunity arose in both McCann and Caraher, but because the Court misunderstood the nature of the law, we cannot conclude that the Court has 65 Emphasis added. in any way given its seal of approval to English law as it stands. 63

188

Killing in Self-Defence

the rest of the judgment, however, it is clear that, when the trial judge uses the term ‘reasonable’ in this passage, he is actually referring to the standard of proof to be applied in criminal cases: beyond reasonable doubt. As has already been demonstrated, he is absolutely clear when outlining the relevant law on self-defence that an honest mistake of fact does not need to be reasonable.

10.4.5 The first case to find a violation: Gül The next significant case is that of Gül v Turkey,66 the first (and, to date, only) case in which the European Court has found Article 2 to be violated on the basis that the state authorities used force in self-defence with no good reason for their belief in its necessity. This case stemmed from the actions of a group of Turkish police officers who shot Mehmet Gül on the suspicion that he was an armed terrorist. The police had received information from an informant that the address at which Gül lived housed a terrorist. There was no evidence prior to this that Gül had any involvement in terrorist activity. The police officers knocked on his door in the middle of the night. While Gül was in the process of opening the door, the officers fired at him and killed him without warning. There was evidence that at least 50 shots were fired. The police officers claimed that Gül had fired the first shot as he was opening the door, but this claim was rejected by the Court. For the Court, the only possible basis for the actions of the police officers was that the click of the key turning, as Gül opened the door, sounded like a gun being cocked. The Court concluded unanimously that the actions of the police officers constituted a violation of Article 2. They began their judgment by repeating the ‘good reasons’ test from McCann and went on to state that: . . . 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 from the occupants of the flat. Nor could the firing be justified by any consideration of the need to secure entry to the flat as it placed in danger the lives of anyone in close proximity to the door.67

In Gül, as in Andronicou, domestic criminal law was not up for consideration but once again the Court can be seen stressing the need for a mistaken belief to be based on good reasons. In addition, unlike in Andronicou, Gül shows that the Court is willing to find a violation of Article 2 where the actions of individuals (albeit officers of the state) are based on a belief for which there is no reasonable basis.

10.4.6 Completing the account of the Article 2(2) case law At the time of writing, aside from Gül, there have been no further cases in which the European Court has held that Article 2(2) was actually violated on 66

(2002) 34 EHRR 28.

67

At [82], emphasis added.

The Impact of the ECHR

189

the basis of a belief in the need to use lethal force that was not supported by good reasons. Likewise, aside from the missed opportunities in Caraher and McCann, the substance of English self-defence law, or indeed the law of any other jurisdiction that allows an unreasonable mistake to ground the defence of self-defence, has not arisen for consideration. There have been a number of cases where the facts would seem to fall squarely into the territory of a mistaken belief in the need to use lethal force, but where, somewhat frustratingly for our purposes, the European Court declined to consider the substance of the application because domestic remedies had not been exhausted.68 In addition, there have been a number of cases in which the McCann ‘good reasons’ test has been reiterated but where the actions of the authorities were found to meet this test. In Brady v United Kingdom,69 the application was declared inadmissible by the Court on this basis. It stemmed from a police operation to arrest the perpetrators of a robbery, where a police officer shot one of the suspects who was carrying a torch, mistaking it for a gun. The European Court assessed the actions of the police officer in the light of the McCann test, stressing the need to have good reasons for an honest belief, and concluded that this test was met. Likewise, in Oláh v Hungary,70 an application involving a mistaken belief in the need to use force was declared inadmissible by the European Court because the state authorities did have ‘good reasons’ for their mistaken belief. Here, police officers had shot and killed the applicant’s husband, in the course of a chase. The deceased had been ordered to stop and was shot when he made a sudden move towards one of the police officers with a shiny object in his hand (in the event he turned out to be unarmed). In Makaratzis v Greece71 the applicant was shot after he drove through a red light and a series of roadblocks in the centre of Athens near to the United States embassy.72 He was shot by police officers both during the course of the chase and when he did eventually come to a halt at a petrol station and refused to get out of his car as requested. In the event, he turned out to be unarmed and the belief of the police officers that he was a terrorist proved to be incorrect. The European Court considered the case under the McCann ‘good reasons’ test and concluded that ‘even though it was subsequently discovered that the applicant was unarmed and that he was not a terrorist . . . the use of force against him was based on an honest belief which was perceived for good reasons to be valid at the time’.73 68 McShane v United Kingdom (2002) 35 EHRR 23; Aytekin v Turkey (2001) 32 EHRR 22; McKerr v United Kingdom (2002) 34 EHRR 20; Jordan v United Kingdom (2003) 37 EHRR 2; Kelly v United Kingdom (2001) 11 BHRC 1. 69 (App no 55151/00) ECHR, 3 April 2001. See the case comment in (2001) 65 JCL 417–19. 70 71 (App no 56558/00) ECHR, 14 September 2004. (2005) 41 EHRR 49. 72 The application was considered under Art 2 even though the applicant survived the shooting. 73 At [66]. The Court did conclude, however, that Art 2 was violated on the basis of the planning and control of the operation.

190

Killing in Self-Defence

Finally, in Bubbins v United Kingdom,74 the applicant’s sister claimed that a violation of Article 2 had occurred after police officers shot dead her brother, Michael Fitzgerald, in his flat. An intruder had been reported entering Mr Fitzgerald’s flat. In fact, it was Mr Fitzgerald himself, but a series of mistakes led police officers to believe that the man in the flat was not Mr Fitzgerald. After a lengthy siege, Mr Fitzgerald appeared at the window holding what appeared to be a gun. One of the police officers asked him to drop it, and when he failed to do so, he was shot dead. It turned out that the object was a replica gun. The European Court applied the McCann test and concluded that, while the use of force was ‘highly regrettable’, it did not constitute a violation of Article 2.75

10.4.7 What can we conclude about the ECHR and mistaken belief in self-defence? As we have seen, the European Court has consistently held that where force is used under one of the three exceptions contained in Article 2(2), any mistaken belief of fact must be held ‘for good reasons’. This was implied in the Commission decision of Kelly and has subsequently been directly stated in a line of cases stemming from McCann and including Gül, Makaratzis, and Bubbins. All of these cases, however, involved state officials. Where the party involved in using force is a state official, the analysis is a straightforward one. The state is directly implicated when one of its own officials uses lethal force in contravention of Convention principles. Indeed, this was exactly what happened in Gül, where the European Court concluded that the conduct of the Turkish authorities in using lethal force in the belief that they were acting in self-defence was a violation of Article 2 because the officials had no good reason for this belief. Thus, where a state official is concerned, a violation of Article 2 occurs when lethal force is used in the unreasonably mistaken belief that it is necessary in self-defence or the prevention of crime.76

10.5 State Officials and Private Citizens It has already been concluded that it is a violation of Article 2 where a state official has used lethal force in order to protect property. Likewise, where a state official has used lethal force in the mistaken belief that it was necessary in self-defence where no good reasons existed for her belief, then this constitutes a violation of Article 2. 74

(2005) 41 EHRR 24. At [140]. For comment on Bubbins, see N Martin, ‘Bubbins v United Kingdom: civil remedies and the right to life’ (2006) 69 MLR 242–9. 76 A conclusion that is also reached by Ashworth, in Principles of Criminal Law, at 147. 75

The Impact of the ECHR

191

The position of private citizens is less straightforward. It should be said immediately that a case could not be brought under Article 2 against any private individual in either the domestic courts or the European Court. The Convention creates obligations for the state, not for private individuals. This does not necessarily mean, though, that there would be no remedy available. The ECHR imposes a positive obligation on the state to take appropriate steps to safeguard the Convention rights of those within its jurisdiction.77 This positive obligation extends in some circumstances to protecting citizens from breaches of their Convention rights caused by the acts of other private citizens.78 In Osman v United Kingdom,79 it was accepted that this obligation required the state to put in place ‘effective criminal law provisions to deter the commission of offences against the person backed up by law enforcement machinery for the prevention, suppression and [punishment] of breaches of such provisions’.80 In Öneryildiz v Turkey,81 the Grand Chamber of the European Court specifically commented on the positive obligation associated with Article 2, concluding that this ‘entails above all a primary duty on the state to put in place a legislative and administrative framework designed to provide effective deterrence against threats to the right to life’.82 The case concerned an explosion at a landfill site which resulted in 39 deaths, which could have been prevented if the authorities had not acted negligently.83 The two local mayors were found guilty of negligence in the course of their duties but the punishment imposed was minor.84 A violation of Article 2 was found on the basis that the mayors were prosecuted for ‘negligence in the performance of their duties’,85 rather than a more serious offence which made some reference to their life-endangering acts.86

77 Çacici v Turkey (2001) 31 EHRR 5; Salman v Turkey (2002) 34 EHRR 17. See also NHS Trust A v M [2001] 2 WLR 942; R (Gezer) v Secretary of State for the Home Department [2004] EWCA Civ 1730, at [27]; R (Burke) v General Medical Council [2005] EWCA Civ 1003, at [39]. For further discussion of the positive obligations doctrine, see A Mowbray, The Development of Positive Obligations under the European Convention on Human Rights by the European Court of Human Rights (2004); J Rogers, ‘Applying the doctrine of positive obligations in the European Convention on Human Rights to domestic substantive criminal law in domestic proceedings’ [2003] Crim LR 690–708. 78 Stubbings v United Kingdom (1997) 23 EHRR 213; HLR v France (1998) 26 EHRR 29. 79 (2000) 29 EHRR 245. 80 At [115]. This passage was repeated in Keenan v United Kingdom (2001) 33 EHRR 38, at [88]; Kiliç v Turkey (2001) 33 EHRR 58, at [62]; Edwards v United Kingdom (2002) 35 EHRR 81 19, at [54]. (2005) 41 EHRR 20. 82 At [89]. 83 In this case more serious criminal law provisions were available; it was just that the state chose not to use them. 84 A fine of 610,000 Turkish lire was imposed but this was suspended by the Court on the basis that it was satisfied the defendants would not re-offend. See Öneryildiz, at [34]. 85 Öneryildiz, at [116]. 86 See also the earlier decision of the Chamber, who reached the same conclusion: Öneryildiz v Turkey (2004) 39 EHRR 12.

192

Killing in Self-Defence

In A v United Kingdom,87 the failure of English criminal law to provide adequate protection to a child who was beaten by his stepfather was found to constitute a violation of Article 3. English criminal law at the time allowed the defence of ‘reasonable chastisement’ to a beating administered by a parent that would otherwise have constituted an assault occasioning actual bodily harm. Likewise, in X and Y v Netherlands,88 the failure of the law of the Netherlands to provide for the criminal prosecution of someone who was suspected of having indecently assaulted a 16-year-old woman with learning difficulties was found to constitute a violation of Article 8 of the Convention, the right to respect for one’s private and family life. Any prosecution required the alleged victim to make a legal complaint, something the woman was not capable of doing. There was no mechanism in place to allow someone to make a complaint on her behalf unless she was aged under 16 or had been placed in guardianship, a procedure available only to those aged over 21. Thus if it is not to be in breach of its obligations under the Convention, the state must, via the substantive criminal law, provide adequate protection of the lives of its citizens. On the basis of the authorities on positive obligations, it could certainly be argued that any system of criminal law which does not contain sanctions for the private individual who uses lethal force to protect property leaves the state open to a finding that Article 2 has been violated. As a counter-argument, it has been said that where an infringement of the right to life was not caused intentionally, the positive obligation imposed by Article 2 does not necessarily require the provision of a criminal law remedy in every case.89 Two points can be made here. First, it might be said that many defensive killings are intentional killings, in which case this point simply does not apply.90 Second, the case in which this principle was established (Calvelli and Ciglio v Italy) was a medical negligence case that involved the death of a baby due to the negligent provision of medical care. This seems substantially different to a case of killing in self-defence or the defence of property, which involves a positive act of force applied to a person. Indeed, this was conclusion reached by the Grand Chamber of the European Court in Öneryildiz v Turkey,91 where the Court contrasted Calvelli with cases involving ‘the use of lethal force’.92 Thus, the conclusion reached above still stands: in order to comply with Article 2, the criminal law must contain sanctions for the private individual who kills in defence of property. The position in relation to mistaken belief is more complicated and there are at least two hurdles to be overcome before it can be concluded that 87

88 (1999) 27 EHRR 611. (1986) 8 EHRR 235. Calvelli and Ciglio v Italy (App no 32967/96) ECHR, 17 January 2002, at [51]; VO v France (2005) 40 EHRR 12, at [90]. See the case comment on Calvelli in (2002) 4 Human 90 Rights Law Review 531–4. On this, see Section 10.2.3 above. 91 92 (2005) 41 EHRR 20. At [93]. 89

The Impact of the ECHR

193

English criminal law, in allowing an unreasonable mistaken belief in the existence of an attack to ground the defence of self-defence, violates Article 2. The first is that it is not entirely obvious that a change in the law would provide any increase in the level of protection offered to citizens. The obligation of the state is to put in place ‘effective criminal law provisions to deter the commission of offences against the person’.93 But it might be said that the substantive provisions of the law would never have a deterrent effect on those who make honest mistakes because such individuals believe that they are acting reasonably, even when they are not. This is unlike the situation in, for example, A v United Kingdom, where a change in the law on reasonable chastisement has more obvious potential to influence the behaviour of parents, who, assuming that the change in the law was publicized, would know as a result what they are and are not permitted to do by way of physical punishment of their children. In response, it could be said that requiring a ‘good reasons’ standard might have some, admittedly limited, deterrent effect as it would send out the message that we should all take reasonable care to check the accuracy of our beliefs before acting rashly in selfdefence.94 On a related note, Buxton has argued that English law would not be found contrary to Article 2 under the positive obligations doctrine because it would be impossible to establish that the criminal law caused the defendant in question to act in the way she did.95 With respect, this objection can be dismissed far more easily, given that establishing causation in this way was not a requirement in either A v United Kingdom or X and Y v Netherlands. The second hurdle to be overcome before reaching a conclusion that English self-defence law is contrary to Article 2 is that Article 2 does not require that the unreasonably mistaken self-defender be convicted specifically of murder. It may be that a conviction for a lesser offence would suffice, so long as that offence is one that reflects the fact that a death has occurred.96 It has been suggested that, in English law, if the self-defender’s mistake was so unreasonable as to amount to gross negligence, she would be liable to be convicted of manslaughter.97 Thus it is not correct to say that English criminal law provides inadequate protection of the lives of its citizens. Two points might be made in response. First, this is all somewhat speculative as there is no direct authority in English law that a conviction for gross negligence manslaughter can be sustained on the basis of an unreasonable belief in the need to use force. Second, the test for ‘gross negligence’ as it 93

94 Osman, at [151], emphasis added. See Chapter 9. R Buxton, ‘The Human Rights Act and the substantive criminal law’ [2000] Crim LR 331–40, at 338. 96 On this requirement, which has been read into Art 2, see Öneryildiz v Turkey (2005) 41 EHRR 20, at [116]. 97 Smith, ‘The use of force in public or private defence’; Ormerod, Smith and Hogan Criminal Law, at 332; Simester and Sullivan, Criminal Law, at 631. 95

194

Killing in Self-Defence

stands in English law at the time of writing is a vague one and there is the risk that instances of unreasonably mistaken belief will slip through the net. In R v Adomako,98 Lord Mackay held that the question of whether gross negligence has occurred is ‘supremely a jury question’99 and involves consideration of whether ‘having regard to the risk of death involved, the conduct of the defendant was so bad in all the circumstances as to amount in [the jury’s] judgment to a criminal act or omission?’100 This test has been criticized as vague and circular,101 involving as it does simply a decision by the jury about whether or not the defendant’s conduct was sufficiently serious to be deemed criminal. Essentially the jury is left to decide the scope of the crime without any clear guidance,102 leaving ‘considerable scope for different verdicts on similar facts’.103 This, it is submitted, can hardly be said to amount to adequate protection of the lives of English citizens. The possibility that English self-defence law might violate Article 2 of the ECHR is an issue I first wrote about in 2002. It was clear even then from the case of Gül that state officials killing in the unreasonable belief that it was necessary in self-defence will render the state liable to a finding that Article 2 has been violated. The issue of whether the state might also be liable where private citizens have acted in this way is more debatable and has still not arisen directly for consideration by the European Court. Despite some critics expressing dissenting views since 2002, I have not changed my view that (a) English law is morally objectionable and (b) it violates Article 2 of the ECHR.104 This does not mean, however, that the appropriate offence is necessarily murder. It was argued earlier in this text that there are compelling reasons for criminalizing the act of the unreasonably mistaken self-defender, as opposed simply to imposing liability in tort.105 It is accepted, though, that murder should be reserved for the most serious acts of killing and that a manslaughter conviction is probably more appropriate. It may be that a conviction for manslaughter could be achieved through the existing law of manslaughter by gross negligence, but this is by no means a foregone conclusion, as noted above. An alternative approach would be for English law to recognize unreasonable self-defensive mistake as a partial defence to a charge of murder, in the same way as it already recognizes a 98

99 100 [1995] 1 AC 171. At 187. At 187. Including by those who argue that the possibility of a conviction for gross negligence manslaughter means that self-defence law does not contravene Art 2. See Ormerod, Smith and Hogan Criminal Law, at 486. That the test ‘involves a element of circularity’ was even accepted by Lord Mackay of Clashfern, who delivered the judgment of the House of Lords in Adomako ([1995] 1 AC 171, at 187). 102 A state of affairs that Ormerod describes as ‘objectionable in principle’ (in Smith and 103 Hogan Criminal Law, at 486). Simester and Sullivan, Criminal Law, at 371. 104 A view that is shared by Ashworth, in Principles of Criminal Law, at 147. 105 See Chapter 9. Cf Ormerod, Smith and Hogan Criminal Law, at 332. 101

The Impact of the ECHR

195

number of other partial defences such as provocation or diminished responsibility.106 Although beyond the scope of this book, which is primarily concerned with the use of lethal force, this would have an advantage over the manslaughter by gross negligence route in another context: where a nonfatal injury has been caused by an unreasonable self-defensive mistake. The gross negligence approach would rule out conviction in such circumstances because, as John Smith put it, in cases of non-fatal injury ‘there is no fallback offence corresponding to manslaughter’.107 The partial defence approach would allow conviction for assault, with the unreasonable mistake acting in mitigation of punishment. Regardless of the procedural route by which it is achieved, the appropriate label for the self-defender who has killed under an unreasonable belief that she is being attacked is almost certainly manslaughter and not murder. This does not take away from the conclusion reached here that if this outcome cannot reliably be achieved in English law at present, the law is incompatible with Article 2 of the ECHR. 106 As has been argued in relation to the use of excessive force in self-defence. On this, see Chapter 9 and N Lacey, ‘Partial defences to homicide: questions of power and principle in imperfect and less imperfect worlds’, in A Ashworth and B Mitchell (eds), Rethinking English Homicide Law (2000) 107–32. 107 Smith, ‘The use of force in public or private defence’, at 961.

This page intentionally left blank

Bibliography Abramson, LY, Seligman, ME, and Teasdale, JD, ‘Learned helplessness in humans: critique and reformulation’ (1978) 87 Journal of Abnormal Psychology 49–77. Acierno, R, Gray, M, Best, C, Resnick, H, Kilpatrick, D, Saunders, B, and Brady, K, ‘Rape and physical violence: comparison of assault characteristics in older and younger adults in the National Women’s Study’ (2001) 14 Journal of Traumatic Stress 685–95. Alexander, LA, ‘A unified excuse of pre-emptive self-protection’ (1999) 74 Notre Dame Law Review 1475–505. —— ‘Self-defense, justification and excuse’ (1993) 22 Philosophy and Public Affairs 53–66. —— ‘Justification and innocent aggressors’ (1987) 33 Wayne Law Review 1177–89. Alison, A, Principles of the Criminal Law of Scotland (Edinburgh: W Blackwood, 1832). Alldridge, P, ‘Self-induced provocation in the Court of Appeal’ (1991) 55 JCL 94–102. Allen, FA, The Habits of Legality: Criminal Justice and the Rule of Law (Oxford: Oxford University Press, 1996). Alston, P, People’s Rights (Oxford: Oxford University Press, 2001). Andrew, DT, ‘The code of honour and its critics’ (1980) 5 Social History 409–34. Anscombe, GEM, ‘War and murder’, in Anscombe, Collected Philosophical Papers Volume 3: Ethics, Religion and Politics (Oxford: Blackwell, 1981) 51–61. Appel, S, ‘Beyond self-defense: the use of battered woman syndrome in duress defences’ (1994) University of Illinois Law Review 955–80. Aquinas, St T, Summa Theologiae (1266–73, 1975 edn with English translation by Marcus Lefébure) (London: Blackfriars, 1975). Ashworth, A, Principles of Criminal Law, 5th edn (Oxford: Oxford University Press, 2006). —— ‘Reason, logic and criminal liability’ (1975) 91 LQR 103–30. —— ‘Self-defence and the right to life’ (1975) 34 CLJ 282–307. Austin, JL, ‘A plea for excuses’, in AR White (ed), The Philosophy of Action (London: Oxford University Press, 1968) 19–42. Ayyildiz, E, ‘When battered woman’s syndrome does not go far enough: the battered woman as vigilante’ (1995) 4 American University Journal of Gender and the Law 141–204. Baron, M, ‘Justifications and excuses’ (2005) 2 Ohio State Journal of Criminal Law 387–406. Beale, JH, ‘Homicide in self-defence’ (1903) 3 Columbia Law Review 526–45. —— ‘Retreat from a murderous assault’ (1903) 16 Harvard Law Review 567–82. Berman, MN, ‘Justification and excuse, law and morality’ (2003) 53 Duke Law Journal 1–77.

198

Bibliography

Bernsmann, K, ‘Private self-defence and necessity in German penal law and in the penal law proposal: some remarks’ (1996) 30 Israel Law Review 171–87. Blackstone, W, Commentaries on the Laws of England: Volume 1 (1769, 1979 reprint) (Chicago: University of Chicago Press, 1979). Brandt, RB, ‘Conscience (rule) utilitarianism and the criminal law’ (1995) 14 Law and Philosophy 65–89. Brown, B, ‘Self-defence in homicide from strict liability to complete exculpation’ [1958] Crim LR 583–90. Brown, RM, No Duty to Retreat: Violence and Values in American History and Society (New York: Oxford University Press, 1994). Browne, A and Williams, KR, ‘Exploring the effect of resource availability and the likelihood of female-precipitated homicides’ (1989) 23 Law and Society Review 75–94. Bruce, D, ‘Killing and the constitution: arrest and the use of lethal force’ (2003) 3 South African Journal on Human Rights 430–54. Bryden, DP, ‘Redefining rape’ (2000) 3 Buffalo Criminal Law Review 317–479. Burchell, J, Principles of Criminal Law, 3rd edn (Cape Town: Juta and Co Ltd, 2005). —— and Milton, J, Cases and Materials on Criminal Law, 2nd edn (Cape Town: Juta and Co Ltd, 1997). Burnett, J, A Treatise on Various Branches of the Criminal Law of Scotland (Edinburgh: George Ramsay and Company, 1811). Burton, M, ‘Criminalising breaches of civil orders for protection from domestic violence’ [2003] Crim LR 301–13. —— ‘Reviewing Crown Prosecution Service decisions not to prosecute’ [2001] Crim LR 374–84. Buxton, R, ‘The Human Rights Act and the substantive criminal law’ [2000] Crim LR 331– 40. Byrd, SB, ‘Wrongdoing and attribution: implications beyond the justification-excuse distinction’ (1987) 33 Wayne Law Review 1289–342. Campbell, K, ‘Offence and defence’, in IH Dennis (ed), Criminal Law and Justice: Essays from the WG Hart Workshop 1986 (London: Sweet & Maxwell, 1987) 73–86. Canadian Department of Justice, Provocation, Self-Defense and Defence of Property: A Consultation Paper (Ottawa: Canadian Department of Justice, 1998). Carpenter, CL, ‘Of the enemy within, the castle doctrine, and self-defense’ (2002–3) Marquette Law Review 653–700. Chalmers, J, ‘Collapsing the structure of criminal law’ 2001 SLT (News) 241–45. —— Gane, CHW, and Leverick, F, ‘Partial defences to homicide in the law of Scotland: a report to the Law Commission for England and Wales’, in Partial Defences to Murder (Consultation Paper No 173 (Appendices)) (London: Law Commission for England and Wales, 2003). —— —— and Stoddart, CN, Gane and Stoddart: A Casebook on Scottish Criminal Law, 3rd edn (Edinburgh: W Green, 2001). —— and Leverick, F, Criminal Defences and Pleas in Bar of Trial (Edinburgh: W Green, 2006). Christopher, RL, ‘Unknowing justification and the logical necessity of the Dadson principle in self-defence’ (1995) 15 OJLS 229–51.

Bibliography

199

—— ‘Mistake of fact in the objective theory of justification: do two rights make two wrongs make two rights . . . ?’ (1994) 85 Journal of Criminal Law and Criminology 295–332. Clarkson, CMV, ‘Necessary action: a new defence’ [2004] Crim LR 81–95. —— and Keating, HM, Criminal Law: Text and Materials, 5th edn (London: Sweet & Maxwell, 2003). Cohen, JM, ‘Regimes of private tyranny: what do they mean to morality and for the criminal law?’ (1996) 57 University of Pittsburgh Law Review 757–808. Colvin, E, ‘Exculpatory defences in criminal law’ (1990) 10 OJLS 381– 407. Connelly, C, ‘Domestic abuse’ in J Scoular (ed), Family Dynamics: Contemporary Issues in Family Law (Edinburgh: Butterworths, 2001) 177–204. Cooper, TM, Select Scottish Cases of the Thirteenth Century (Edinburgh: W Hodge and Co, 1944). Corrado, M, ‘Notes on the structure of a theory of excuses’ (1992) 82 Journal of Criminal Law and Criminology 465–97. Coughlin, AM, ‘Excusing women’ (1994) 82 California Law Review 1–93. Cranston, M, What are Human Rights? (London: Bodley Head, 1973). Criminal Law Revision Committee, Sexual Offences (Cmnd 9213) (London: Home Office, 1984). Criminal Law Revision Committee, Offences against the Person (Cmnd 7844) (London: Home Office, 1980). Curley, EM, ‘Excusing rape’ (1976) 5 Philosophy and Public Affairs 325–60. Daily Telegraph, ‘The right to fight back’, 12 December 2004. Dan-Cohen, M, ‘Decision rules and conduct rules: on acoustic separation in criminal law’ (1984) 97 Harvard Law Review 625–77. Dancy, J, Practical Reality (Oxford: Oxford University Press, 2000). Darbyshire, P, ‘What can we learn from published jury research? Findings for the Criminal Courts Review 2001’ [2001] Crim LR 970–9. —— Maughan, A, and Stewart, A, What Can the English Legal System Learn from Jury Research Published up to 2001? Research Papers in Law, Occasional Paper No 49 (London: Kingston Business School, 2001). Davis, N, ‘Abortion and self-defense’ (1984) 13 Philosophy and Public Affairs 171–207. Dettmar, VJ, ‘Culpable mistakes in rape: eliminating the defense of unreasonable mistake of fact as to victim consent’ (1984–5) 89 Dickinson Law Review 473–99. Diamond, GA, ‘To have but not to hold: can “resistance against kidnapping” justify lethal self-defense against incapacitated batterers?’ (2002) 102 Columbia Law Review 729–73. Dicey, AV, Introduction to the Study of the Law of the Constitution, 4th edn (London: Macmillan, 1893). Dillof, AM, ‘Unravelling unknowing justification’ (2002) 77 Notre Dame Law Review 1547–600. Dressler, J, ‘Why keep the provocation defense? Some reflections on a difficult subject’ (2002) 86 Minnesota Law Review 959–1002. —— ‘Battered women who kill their sleeping tormenters: reflections on maintaining respect for human life while killing moral monsters’, in S Shute and AP Simester (eds), Criminal Law Theory: Doctrines of the General Part (Oxford: Oxford University Press, 2002) 259–82.

200

Bibliography

Dressler, J, Understanding Criminal Law, 3rd edn (New York: Lexis, 2001). —— ‘Reflecting on excusing wrongdoers: moral theory, new excuses and the moral penal code’ (1988) 19 Rutgers Law Journal 671–716. —— ‘Justifications and excuses: a brief review of the concepts and the literature’ (1987) 33 Wayne Law Review 1155–75. —— ‘New thoughts about the concept of justification in the criminal law: a critique of Fletcher’s thinking and re-thinking’ (1984) 32 UCLA Law Review 61–99. —— ‘Rethinking heat of passion: a defense in search of a rationale’ (1982) 73 Journal of Criminal Law and Criminology 421–70. Duff, RA, ‘Theorizing criminal law: a 25th anniversary essay’ (2005) 25 OJLS 353–67. —— ‘Law, language and community: some preconditions of criminal liability’ (1998) 18 OJLS 189–206. East, EH, Pleas of the Crown (1803, 1987 reprint) (Oxford: Professional Books Ltd, 1987). Edwards, S, ‘Abolishing provocation and reframing self-defence: the Law Commission’s options for reform’ [2004] Crim LR 181–97. ——Sex and Gender in the Legal Process (London: Blackstone, 1996). ——Policing ‘Domestic’ Violence: Women, the Law and the State (London: Sage, 1989). Elliott, C, ‘What future for voluntary manslaughter?’ (2004) 68 JCL 253–63. Epps, E, ‘Any which way but loose: interpretive strategies and attitudes toward violence in the evolution of the Anglo-American retreat rule’ (1992) 55 Law and Contemporary Problems 303–31. Eser, A, ‘Justification and excuse’ (1976) 24 American Journal of Comparative Law 621–637. Essique, CR, ‘The use of deadly force by women against rape in Michigan: justifiable homicide?’ (1991) 37 Wayne Law Review 1969–87. Estrich, S, ‘Defending women’ (1990) 88 Michigan Law Review 1430–9. Ewing, CP, ‘Psychological self-defense: a proposed justification for battered women who kill’ (1990) 14 Law and Human Behavior 579–94. ——Battered Women who Kill: Psychological Self-Defense as Legal Justification (Lexington, Mass: Lexington Books, 1987). Faigman, DL, ‘Discerning justice when battered women kill’ (1987) 39 Hastings Law Journal 207–27. —— ‘The battered woman syndrome and self-defense: a legal and empirical dissent’ (1986) 72 Virginia Law Review 619– 47. —— and Wright, AJ, ‘The battered woman syndrome in the age of science’ (1987) 39 Arizona Law Review 67–115. Ferzan, KF, ‘Defending imminence: from battered women to Iraq’ (2004) 46 Arizona Law Review 213–62. Finkelstein, CO, ‘Self-defense as a rational excuse’ (1996) 57 University of Pittsburgh Law Review 621– 49. Finnis, JM, Natural Law and Natural Rights (Oxford: Clarendon Press, 1980). Fisher, M, Benn, P, Evans, B, Pozniak, A, Jones, M, MacLean, S, Davidson, O, Summerside, J, and Hawkins, D, ‘UK guideline for the use of post-exposure prophylaxis for HIV following sexual exposure’ (2006) 17 International Journal of STD and AIDS 81–92.

Bibliography

201

Fletcher, GP, ‘Domination in the theory of justification and excuse’ (1996) 57 University of Pittsburgh Law Review 553–78. —— ‘The nature of justification’, in S Shute, J Gardner, and J Horder (eds), Action and Value in Criminal Law (Oxford: Clarendon Press, 1993) 175–86. —— ‘The psychotic aggressor: a generation later’ (1993) 27 Israel Law Review 227–46. —— ‘Defensive force as an act of rescue’, in EF Paul, FD Miller Jr, and J Paul (eds), Crime, Culpability and Remedy (Cambridge, Mass: Blackwell, 1990) 170–9. —— A Crime of Self-Defense: Bernard Goetz and the Law on Trial (Chicago: University of Chicago Press, 1988). —— ‘Rights and excuses’ (1984) 3 Criminal Justice Ethics 17–27. —— ‘The right to life’ (1980) 63 The Monist 135–55. —— Rethinking Criminal Law (Boston: Little, Brown, 1978). —— ‘The right deed for the wrong reason: a reply to Mr Robinson’ (1975) 23 UCLA Law Review 293–321. —— ‘The individualisation of excusing conditions’ (1974) 47 Southern California Law Review 1269–309. Foot, P, ‘The problem of abortion and the doctrine of double effect’, in Foot, Virtues and Vices (Oxford: Basil Blackwell, 1978). Foster, M, A Report of some Proceedings on the Commission of Oyer and Terminer and Goal (sic) Delivery for the Trial of the Rebels in the Year 1746 in the County of Surry, and of other Crown Cases to which are added Discourses upon a few Branches of the Crown Law (Oxford: Clarendon Press, 1762). Funk, TM, ‘Justifying justifications’ (1999) 19 OJLS 631– 47. Gane, CHW, Sexual Offences (Edinburgh: Butterworths, 1992). Gardner, J, ‘Fletcher on offences and defences’ (2004) 39 Tulsa Law Review 817–27. —— ‘The mark of responsibility’ (2003) 23 OJLS 157–71. —— ‘The gist of excuses’ (1998) 1 Buffalo Criminal Law Review 575–98. —— ‘Justifications and reasons’, in AP Simester and ATH Smith (eds), Harm and Culpability (Oxford: Clarendon Press, 1996) 103–29. —— ‘In defence of defences’, in P Asp, CE Herlitz, and L Holmqvist (eds), Flores Juris et Legum: Festskrift till Nils Jareborg (Uppsala: Iustus Forlag, 2002). —— and Macklem, T, ‘Compassion without respect? Nine fallacies in R v Smith’ [2001] Crim LR 623–35. —— and Shute, S, ‘The wrongness of rape’, in J Horder (ed), Oxford Essays in Jurisprudence, Fourth Series (Oxford: Oxford University Press, 2000) 193–217. Gardner, S, ‘Instrumentalism and necessity’ (1986) 6 OJLS 431–8. Gerling, SM, ‘Louisiana’s new “kill the carjacker” statute: self-defense or instant injustice?’ (1999) 55 Washington University Journal of Urban and Contemporary Law 109–34. Getzler, J, ‘Use of force in protecting property’ (2005) 7 Theoretical Inquiries in Law (on-line edition). Glover, J, Causing Death and Saving Lives (Harmondsworth: Penguin, 1977). Goodnough, A, ‘Florida expands right to use deadly force in self-defense’, New York Times, 27 April 2005. Gordon, GH, The Criminal Law of Scotland, 3rd edn, in 2 vols, by MGA Christie: (Edinburgh: W Green, 2000 (vol 1), 2001 (vol 2)). Gorr, M, ‘Private defence’ (1990) 9 Law and Philosophy 241–68.

202

Bibliography

Green, SP, ‘Castles and carjackers: proportionality and the use of deadly force in defence of dwellings and vehicles’ [1999] University of Illinois Law Review 1–41. Green, TA, ‘The jury and the English law of homicide 1200–1600’ (1976) Michigan Law Review 414–99. Greenawalt, K, ‘Distinguishing justifications from excuses’ (1986) 49 Law and Contemporary Problems 89–108. —— ‘The perplexing borders of justification and excuse’ (1984) 84 Columbia Law Review 1897–927. Gregory, A, ‘To have but not to hold: can “resistance against kidnapping” justify lethal self-defense against incapacitated batterers?’ (2002) 102 Columbia Law Review 729–73. Grisez, GG, ‘Towards a consistent natural law ethics of killing’ (1970) 15 American Journal of Jurisprudence 64–96. Gur-Arye, M, ‘Should the criminal law distinguish between necessity as a justification and necessity as an excuse?’ (1986) 102 LQR 71–89. Hale, M, Historia Placitorum Coronae Volume 1 (1736, 1971 reprint) (London: Professional Books Ltd, 1971). Hampton, J, ‘Defining wrong and defining rape’, in K Burgess-Jackson (ed), A Most Detestable Crime: New Philosophical Essays on Rape (Oxford: Oxford University Press, 1999), 118–56. Harel, A, ‘Efficiency and fairness in criminal law: the case for a criminal law principle of comparative fault’ (1994) 82 California Law Review 1181–229. Harris, J, Violence and Responsibility (London: Routledge & Kegan Paul, 1980). Hart, HLA, Punishment and Responsibility: Essays in the Philosophy of Law (Oxford: Clarendon Press, 1968). Hawkins, W, A Treatise of the Pleas of the Crown, 3rd edn (London: E and R Nutt and R Gosling, 1739). Hay, MA, ‘The criminal law of private defence in England, Scotland and France’ (1990) unpublished PhD thesis, University of Edinburgh. Heberling, PDW, ‘Justification: the impact of the Model Penal Code on statutory reform’ (1975) 75 Columbia Law Review 914–62. Herrmann, J, ‘Causing the conditions of one’s own defence: the multifaceted approach of German law’ [1986] Brigham Young University Law Review 747–67. Hobbes, T, Leviathan (1651, 1991 reprint ed by Richard Tuck) (Cambridge: Cambridge University Press, 1991). Hogan, B, ‘Defence of property’ (1994) 144 NLJ 466–8 and 471. —— ‘The Dadson principle’ [1989] Crim LR 679–86. Hohfeld, WN, ‘Some fundamental legal conceptions as applied in judicial reasoning’ (1913) 23 Yale Law Journal 16–59. Home Office, Domestic Violence: A National Report (London: Home Office, 2005). Horder, J, Excusing Crime (Oxford: Oxford University Press, 2004). —— ‘Killing the passive abuser: a theoretical defence’, in S Shute and AP Simester (eds), Criminal Law Theory: Doctrines of the General Part (Oxford: Oxford University Press, 2002) 283–97. —— ‘Self-defense, necessity and duress: understanding the relationship’ (1998) 11 Canadian Journal of Law and Jurisprudence 143–65. —— ‘Occupying the moral high ground? The Law Commission on duress’ [1994] Crim LR 334–42.

Bibliography

203

——Provocation and Responsibility (Oxford: Oxford University Press, 1992). —— ‘The duel and the English law of homicide’ (1992) 12 OJLS 419–30. Horowitz, D, ‘Justification and excuse in the program of the criminal law’ (1986) 49 Law and Contemporary Problems 109–26. House of Lords Select Committee, Report on Murder and Life Imprisonment (HL Paper 78–1), (London: Home Office, 1988–9). Hume, D, Commentaries on the Law of Scotland Respecting Crimes, 4th edn (Edinburgh: Law Society of Scotland, 1844). —— Commentaries on the Law of Scotland respecting the Description and Punishment of Crimes (Edinburgh: Bell and Bradfute, 1797). Hurd, H, ‘Justification and excuse, wrongdoing and culpability’ (1999) 74 Notre Dame Law Review 1551–73. Hurnard, ND, The King’s Pardon for Homicide before AD 1307 (Oxford: Clarendon Press, 1969). Husak, D, ‘On the supposed priority of justification to excuse’ (2005) 24 Law and Philosophy 557–94. —— ‘The complete guide to self-defense’ (1996) 15 Law and Philosophy 399–406. —— ‘The serial view of criminal law defenses’ (1992) 3 Criminal Law Forum 369–400. —— ‘Justifications and the criminal liability of accessories’ (1989) 80 Journal of Criminal Law and Criminology 491–520. Jefferson, M, ‘Householders and the use of force against intruders’ (2005) 69 JCL 405–13. Jones, P, Rights (Basingstoke: Macmillan Press, 1994). Jones, TH and Christie, MGA, Criminal Law, 3rd edn (Edinburgh: W Green and Son, 2003). Kadish, SH, ‘Excusing crime’ (1987) 75 California Law Review 257–89. —— ‘Respect for life and regard for rights in criminal law’ (1976) 64 California Law Review 871–901. Kalmuss, DS and Straus, MA, ‘Wife’s marital dependency and wife abuse’ (1982) 44 Journal of Marriage and the Family 277–86. Kalven, H Jr and Zeisel, H, The American Jury (Boston: Little, Brown, 1966). Kasachkoff, T, ‘Comment and reply to Suzanne Uniacke’s “a response to two critics” ’ (2000) 19 Law and Philosophy 635–9. —— ‘Killing in self-defence: an unquestionable or problematic defence?’ (1998) 17 Law and Philosophy 509–31. Kates, DB Jr and Engberg, NJ, ‘Deadly force self-defense against rape’ (1982) 15 University of California Davis Law Review 873–906. Kaye, M, ‘Excessive force in self-defence after R v Clegg’ (1997) 61 JCL 448–57. Kilbrandon, Lord, ‘Duress per minas as a defence to crime: I’, in MA Stewart (ed), Law Morality and Rights (Dordrecht: Reidel, 1983) 333–43. Kilpatrick, D, Edmunds, C, and Seymour, A, Rape in America: A Report to the Nation (Arlington, Va: Crime Victims Research and Treatment Center, 1992). Klimchuk, D, ‘Necessity, deterrence and standing’ (2002) 8 Legal Theory 339–58. Kremnitzer, M, ‘Proportionality and the psychotic aggressor: another view’ (1983) 18 Israel Law Review 178–214. —— and Ghanayim, K, ‘Proportionality and the aggressor’s culpability in self-defense’ (2003–4) 39 Tulsa Law Review 875–99.

204

Bibliography

Lacey, N, ‘Partial defences to homicide: questions of power and principle in imperfect and less imperfect worlds’, in A Ashworth and B Mitchell (eds), Rethinking English Homicide Law (Oxford: Oxford University Press, 2000) 107–32. —— State Punishment (London: Routledge, 1988). LaFave, WR, Criminal Law, 3rd edn (St Paul, Minn: West, 2000). Lanham, BP, ‘Defence of property in the criminal law’ [1966] Crim LR 368–79 (part one) and 426–35 (part two). Lanham, D, ‘Offensive weapons and self-defence’ [2005] Crim LR 85–97. —— ‘Death of a qualified defence?’ (1988) 104 LQR 239–49. Law Commission, A New Homicide Act for England and Wales? (Law Com CP No 177) (London: Law Commission for England and Wales, 2005). —— Report on Partial Defences to Murder (Law Com No 290) (London: Law Commission for England and Wales, 2004). —— Consultation Paper on Partial Defences to Murder (Law Com CP No 173) (London: Law Commission for England and Wales, 2003). —— Legislating the Criminal Code: Offences Against the Person and General Principles (Law Com No 218) (London: Law Commission for England and Wales, 1993). —— Criminal Law: A Code for England and Wales (Law Com No 177) (London: Law Commission for England and Wales, 1989). —— Codification of the Criminal Law: A Report to the Law Commission (Law Com No 143) (London: Law Commission for England and Wales, 1985). Law Reform Commission of Canada, Homicide (Working Paper 33) (Ottawa: Law Reform Commission of Canada, 1984). Lee, CKY, ‘The act-belief distinction in self-defense doctrine: a new dual requirement theory of justification’ (1998) 2 Buffalo Criminal Law Review 191–247. —— ‘Race and self-defense: towards a normative conception of reasonableness’ (1996) 81 Minnesota Law Review 367–500. Lee, RV, ‘A further erosion of the retreat rule in North Carolina’ (1976) 12 Wake Forest Law Review 1093–103. Leigh, LH, ‘Manslaughter and the limits of self-defence’ (1971) 34 MLR 685–90. Leverick, F, ‘What has the ECHR done for victims: a United Kingdom perspective’ (2004) 11 International Review of Victimology 177–200. —— A Critical Analysis of the Law of Self-defence in Scotland and England, (2003), unpublished PhD thesis, University of Aberdeen. —— ‘Is English self-defence law incompatible with the ECHR?’ [2002] Crim LR 347–62. —— ‘Mistake in self-defence after Drury’, (2002) Juridical Review 35–48. Levine, S, ‘The moral permissibility of killing a material aggressor in self-defence’ (1984) 45 Philosophical Studies 69–78. Lloyd, L and Freeman, MDA, Lloyd’s Introduction to Jurisprudence, 6th edn (London: Sweet & Maxwell, 1994). Locke, J, Two Treatises of Government (1690, 1988 edn with introduction and notes by Peter Laslett) (Cambridge: Cambridge University Press, 1988). McAuley, F, ‘The grammar of mistake in criminal law’ (1996) 31 Irish Jurist 56–82. —— ‘The theory of justification and excuse: some Italian lessons’ (1987) 35 American Journal of Comparative Law 359–80. —— and McCutcheon, PJ, Criminal Liability: A Grammar (Dublin: Sweet & Maxwell, 2000).

Bibliography

205

McColgan, A, ‘In defence of battered women who kill’ (1993) 14 OJLS 508–29. McConnell, T, ‘The nature and basis of inalienable rights’ (1984) 3 Law and Philosophy 25–59. Macfarlane, LJ, The Theory and Practice of Human Rights (London: Maurice Temple Smith, 1985). McIntyre, A, ‘Doing away with double effect’ (2001) 111 Ethics 219–55. McLean, M, ‘Domestic violence and repeat victimisation’ (2005) 12 International Review of Victimology 51–74. McMahan, J, ‘Self-defense and the problem of the innocent attacker’ (1994) 104 Ethics 252–90. Maguigan, H, ‘Battered women and self-defense: myths and misconceptions in current reform proposals’ (1991) 140 University of Pennsylvania Law Review 379–486. Mahoney, MR, ‘Legal images of battered women: redefining the issue of separation’ (1991) 90 Michigan Law Review 1–94. Martin, M, ‘Bubbins v United Kingdom: civil remedies and the right to life’ (2006) 69 MLR 242–9. Michalowski, S, ‘Sanctity of life: are some lives more sacred than others?’ (2002) 22 LS 377–97. Mihajlovich, M, ‘Does plight make right: the battered woman syndrome, expert testimony and the law of self-defense’ (1987) 62 Indiana Law Journal 1253–82. Mirrlees-Black, C, Domestic Violence: Findings from a New British Crime Survey Self-Completion Questionnaire (London: Home Office, 1999). Model Penal Code and Commentaries Part 1 (General Provisions) (Philadelphia: The American Law Institute, 1985). Montague, P, ‘The morality of self-defence: a reply to Wasserman’ (1989) 18 Philosophy and Public Affairs 81–9. —— ‘Self-defence and choosing between lives’ (1981) 40 Philosophical Studies 207–19. Moore, M, Placing Blame: A General Theory of the Criminal Law (Oxford: Clarendon Press, 1987). Morawetz, T, ‘Reconstructing the criminal defenses: the significance of justification’ (1986) 77 Journal of Criminal Law and Criminology 277–307. Morse, SJ, ‘The misbegotten marriage of soft psychology and bad law: psychological self-defense as a justification for homicide’ (1990) 14 Law and Human Behavior 595–618. Mowbray, A, The Development of Positive Obligations under the European Convention on Human Rights by the European Court of Human Rights (Oxford: Hart, 2004). Murdoch, JB, ‘Is imminence really necessary? Reconciling traditional self-defense doctrine with the battered woman syndrome’ (2000) 20 North Illinois University Law Review 191–218. Muth, RJ and Blumstein, AL, ‘The use of deadly force in the protection of property under the Model Penal Code’ (1959) 59 Columbia Law Review 1212–32. New Zealand Law Commission, Some Criminal Defences with Particular Reference to Battered Defendants (Wellington: New Zealand Law Commission, 2001). Nicholas, S, Povey, D, Walker, A, and Kershaw, C, Crime in England and Wales 2004–05 (London: Home Office, 2005).

206

Bibliography

Norrie, A, Crime, Reason and History: A Critical Introduction to Criminal Law, 2nd edn (London: Buterworths, 2001). Nourse, VF, ‘Self-defense and subjectivity’ (2001) 68 University of Chicago Law Review 1235–308. Omichinski, NM, ‘Applying the theories of justifiable homicide to conflicts in the doctrine of self defence’ (1987) 33 Wayne Law Review 1447–69. Ormerod, D, Smith and Hogan Criminal Law, 11th edn (Oxford: Oxford University Press, 2005). Otsuka, M, ‘Killing the innocent in self-defense’ (1994) 23 Philosophy and Public Affairs 74–94. Pendleton, HA, ‘A critique of the rational excuse defence: a reply to Finkelstein’ (1996) 57 University of Pittsburgh Law Review 651–76. Pickard, T, ‘Culpable mistakes and rape: relating mens rea to the crime’ (1980) 30 University of Toronto Law Journal 75–98. Power, H, ‘Towards a re-definition of the mens rea of rape’ (2003) 23 OJLS 379–404. Price, TL, ‘Faultless mistake of fact: justification or excuse?’ (1993) 12 Criminal Justice Ethics 14–28. Rachels, J, The Elements of Moral Philosophy, 3rd edn (Boston: McGraw Hill College, 1999). Raz, J, Practical Reason and Norms (London: Hutchinson, 1975). Resick, PA, ‘The psychological impact of rape’ (1993) 8 Journal of Interpersonal Violence 223–55. Richards, DA, ‘Rights, resistance, and the demands of self-respect’ (1983) 32 Emory Law Journal 405–35. Ripstein, A, Equality, Responsibility and the Law (Cambridge: Cambridge University Press, 1999). —— ‘Self-defense and equal protection’ (1996) 57 University of Pittsburgh Law Review 685–724. Robinson, PH, ‘Competing theories of justification: deeds v reasons’, in AP Simester and ATH Smith (eds), Harm and Culpability (Oxford: Clarendon Press, 1996) 45–70. —— ‘A functional analysis of criminal law’ (1994) 88 Northwestern University Law Review 857–913. —— ‘Rules of conduct and principles of adjudication’ (1990) 57 University of Chicago Law Review 729–71. —— ‘Causing the conditions of one’s own defence: a study in the limits of theory in criminal law doctrine’ (1985) 71 Virginia Law Review 1–62. —— Criminal Law Defenses (St Paul, Minn: West, 1984). —— ‘Criminal law defenses: a systematic analysis’ (1982) 82 Columbia Law Review 199–291. —— ‘A theory of justification: social harm as a prerequisite for criminal liability’ (1975) 23 UCLA Law Review 266–92. —— and Darley, JH, ‘Does criminal law deter? A behavioural science investigation’ (2004) 24 OJLS 173–205. —— —— ‘Testing competing theories of justification’ (1998) 76 North Carolina Law Review 1095–136. —— Greene, PD, and Goldstein, NR, ‘Making criminal codes functional: a code of conduct and a code of adjudication’ (1996) 57 Journal of Criminal Law and Criminology 304–65.

Bibliography

207

Rodin, D, War and Self-Defense (Oxford: Oxford University Press, 2002). Rogers, J, ‘Applying the doctrine of positive obligations in the European Convention on Human Rights to domestic substantive criminal law in domestic proceedings’ [2003] Crim LR 690–708. Rosen, CJ, ‘The excuse of self-defense: correcting a historical accident on behalf of battered women who kill’ (1986) 36 American University Law Review 11–56. Rosen, R, ‘On self-defense, imminence and woman who kill their batterers’ (1993) 71 North Carolina Law Review 371–411. Ryan, C, ‘Self-defense, pacifism and the possibility of killing’ (1983) 93 Ethics 508–24. Schopp, RF, Justification Defenses and Just Convictions (Cambridge: Cambridge University Press, 1998). Schulhofer, S, ‘The gender question in criminal law’, in EF Paul, FD Miller Jr, and J Paul (eds), Crime, Culpability and Remedy (Cambridge, Mass: Blackwell, 1990) 105–37. Scottish Law Commission, Discussion Paper on Rape and other Sexual Offences (DP 131) (Edinburgh: Scottish Law Commission, 2006). —— A Draft Criminal Code for Scotland with Commentary (Edinburgh: Scottish Law Commission, 2003). Sebok, AJ, ‘Does an objective theory of self defence demand too much?’ (1996) 57 University of Pittsburgh Law Review 725–55. Segev, R, ‘Fairness, responsibility and self-defense’ (2004–5) 45 Santa Clara Law Review 383–460. Seligman MEP and Maier, SF, ‘Failure to escape traumatic shock’ (1967) 74 Journal of Experimental Psychology 1–9. —— —— and Geer, J, ‘Alleviation of learned helplessness in the dog’ (1968) 73 Journal of Abnormal Psychology 256–62. Shad, KA, ‘State v Norman: self-defense unavailable to battered women who kill passive abusers’ (1990) 68 North Carolina Law Review 1159–77. Shestack, J, ‘The philosophical foundations of human rights’, in J Symonides (ed), Human Rights: Concept and Standards (Aldershot: Dartmouth, 2000). Shute, S, ‘Second Law Commission Consultation Paper on Consent: (1) Something old, something new, something borrowed: three aspects of the project’ [1996] Crim LR 684–93. Simester, AP, ‘Mistakes in defence’ (1992) 12 OJLS 295–310. —— and Sullivan, GR, Criminal Law: Theory and Doctrine, 2nd edn (Oxford: Hart, 2004). Simpson, J, Touching the Void (London: Cape, 1988). Singer, R, ‘The resurgence of mens rea: II—honest but unreasonable mistake of fact in self defense’ (1986–7) 28 Boston College Law Review 459–519. Skinner, S, ‘Populist politics and shooting burglars: comparative comments on the Lega Nord’s proposal to reform Italian self-defence law’ [2005] Crim LR 275–84. Smart, BJ, ‘Understanding and justifying self-defence’ (1989) 4 International Journal of Moral and Social Studies 231–44. Smith, JC, ‘The use of force in public or private defence and Article 2’ [2002] Crim LR 958–62. —— ‘Surgical separation: case comment on Re: A (Children)’ [2001] Crim LR 400–5.

208

Bibliography

Smith, JC, ‘The right to life and the right to kill in law enforcement’ (1994) 144 NLJ 354–6. —— Justification and Excuse in the Criminal Law (London: Stevens, 1989). —— and Hogan, B, Criminal Law (London: Butterworths, 1969). Smith, P, ‘Excessive defence: a rejection of Australian initiative?’ [1972] Crim LR 524–34. Snelling, HA, ‘Killing in self-defence’ (1960) 34 Australian Law Journal 130–8. St John-Stevas, N, The Right to Life (London: Hodder & Stoughton, 1963). Stewart, H, ‘The role of reasonableness in self-defense’ (2003) 16 Canadian Journal of Law and Jurisprudence 317–36. Strube, MJ and Barbour, LS, ‘The decision to leave an abusive relationship: economic dependence and psychological commitment’ (1983) 45 Journal of Marriage and the Family 785–93. Stuart, D, Canadian Criminal Law: A Treatise, 4th edn (Toronto: Carswell, 2001). Stuart, DR, ‘Killing in defence of property’ (1967) 84 SALJ 123–31. Sullivan, GR, ‘Making excuses’, in AP Simester and ATH Smith (eds), Harm and Culpability (Oxford: Clarendon Press, 1996) 131–52. Tadros, V, Criminal Responsibility (Oxford: Oxford University Press, 2005). —— ‘The structure of defences in Scots criminal law’ (2003) 7 Edin LR 60–79. —— ‘Insanity and the capacity for criminal responsibility’ (2001) 5 Edin LR 325–54. —— ‘The characters of excuse’ (2001) 21 OJLS 495–519. —— ‘Practical reasoning and intentional action’ (2000) 20 LS 104–23. Temkin, J, Rape and the Legal Process, 2nd edn (Oxford: Oxford University Press, 2002). Terrance, CA, Matheson, K, and Spanos, NP, ‘Effects of judicial instructions and case characteristics in a mock jury trial of battered women who kill’ (2000) 24 Law and Human Behavior 207–29. Thomson, JJ, ‘Self-defense’ (1991) 20 Philosophy and Public Affairs 283–310. —— Rights, Restitution and Risk: Essays in Moral Theory (Cambridge, Mass: Harvard University Press, 1986). Uniacke, S, ‘Was Mary’s death murder?’ (2001) 9 Medical Law Review 208–20. —— ‘In defence of Permissible Killing: a response to two critics’ (2000) 19 Law and Philosophy 627–33. —— ‘Rights and relativistic justifications: replies to Kasachkoff and Husak’ (2000) 19 Law and Philosophy 645–47. —— Permissible Killing: The Self-Defence Justification of Homicide (Cambridge: Cambridge University Press, 1994). —— ‘The doctrine of double effect’ (1984) 48 The Thomist 188–218. Veinsreideris, ME, ‘The prospective effects of modifying existing law to accommodate pre-emptive self-defense by battered women’ (2000) 149 University of Pennsylvania Law Review 613– 44. Waldron, J, ‘Self-defense: agent-neutral and agent-relative accounts’ (2000) 88 California Law Review 711– 49. Walker, LE, The Battered Woman (New York: Harper Colophon, 1979). —— The Battered Woman Syndrome (New York: Springer, 1984). Wallace, RJ, Responsibility and the Moral Sentiments (Cambridge, Mass: Harvard University Press, 1994). Wallace, S, ‘Beyond imminence: evolving international law and battered women’s right to self-defense’ (2004) 71 University of Chicago Law Review 1749–81.

Bibliography

209

Wallerstein, S, ‘Justifying the right to self-defense: a theory of forced consequences’ (2005) 91 Virginia Law Review 999–1035. Ward, AR, ‘Making some sense of self-induced intoxication’ (1986) 45 CLJ 247–61. Wasserman, D, ‘Justifying self-defense’ (1987) 16 Philosophy and Public Affairs 356–78. Wechsler, H and Michael, J, ‘A rationale of the law of homicide’ (1939) 37 Columbia Law Review 701–61 (part 1) and 1261–325 (part 2). Wells, C, ‘Battered woman syndrome and defences to homicide: where now?’ (1994) 14 LS 266–76. Westen, P and Mangiafico, J, ‘The criminal defense of duress: a justification not an excuse and why it matters’ (2003) 6 Buffalo Criminal Law Review 833–950. Wheatcroft, M, ‘Duty to retreat for cohabitants in New Jersey: a battered spouse’s home is not her castle’ (1999) 30 Rutgers Law Journal 539–67. Williams, B, ‘A critique of utilitarianism’, in Smart, JJC and Williams, B (eds), Utilitarianism: For and Against (Cambridge: Cambridge University Press, 1973) 77–150. Williams, G, ‘Offences and defences’ (1982) 2 LS 233–56. —— Textbook of Criminal Law (London: Stevens, 1978). —— Criminal Law: The General Part, 2nd edn (London: Stevens, 1961). Wilson, M and Daly, M, ‘Spousal homicide risk and estrangement’ (1993) 8 Violence and Victims 3–16. Woodruff, OE Jr, ‘Mistake of fact as a defense’ (1958–9) 63 Dickinson Law Review 319–33. Woods, G, ‘The sanctity of murder: reforming the homicide penalty in New South Wales’ (1983) 57 Australian Law Journal 161–6. Yeo, S, ‘Revisiting excessive self-defence’ (2000–1) 12 Current Issues in Criminal Justice 39–57. —— ‘Applying excuse theory to excessive self-defence’, in S Yeo (ed), Partial Excuses to Murder (Sydney: Federation Press, 1990) 158–77. Young, R, ‘What is so wrong with killing people?’ (1979) 54 Philosophy 515–28. Zipursky, BC, ‘Self-defense, domination and the social contract’ (1996) 57 University of Pittsburgh Law Review 579–613.

This page intentionally left blank

Index ‘absence of an element’ defences 16 n. 23 adjudication rule 28 African Charter on Human and People’s Rights 57 aggressor: culpable 5, 44, 47 innocent 5, 37, 46, 47, 50–3, 63–4, 67 multiple 47 terminology 5 violating right to life 165 Alexander, LA 47 American Convention on Human Rights 57 American Declaration of Independence 1776 56 American Declaration of the Rights and Duties of Man 57 Anscombe, GEM 53–4 Appel, S 98 Aquinas, St Thomas 53, 56 Archbold 137, 138 Article 2, see European Convention on Human Rights Ashworth, A irremediable harm 152 rape 147 retreat 81 right to life 58, 59 self-generated self-defence 125, 126 assistance, right of 38, 39 Austin, JL 18 Australia defence of property 133 excessive force as partial defence 173, 174, 175 killing to prevent rape 146 mistake 161 retreat rule 73 self-generated self-defence 117 automatism 120 Ayyildiz, E 93 Baron, M 30 battered woman syndrome (BWS) 103, 104, 105 battered women court orders 106–7 defences available 108 diminished responsibility 108 imminence of harm 87–8, 90–1, 92, 99 inevitability of harm test 102–8

justification defence 108 killing in non-confrontational situations 9–10, 87, 90–1, 92, 99 police response 106 proportionality requirement 91–2 provocation 108 psychological self-defence 154 ‘reasonable battered woman’ perspective 105 retreat rule 84 separation assault 106 use of term 87 Beale, JH 77 behaviour-guiding argument 27–9 Blackstone, W 115 Blumstein 136 Brandt, RB 48 breach of the peace 128, 129 British Crime Survey 1996, domestic violence 91 burglary, violence and 139–40 BWS (battered woman syndrome) 103, 104, 105 Byrd, SB 33 bystander forfeiture 67 innocent 44, 50, 60 no direct threat to accused 7, 8 terminology 5 Canada defence of property 132 excessive force as partial defence 172 imminence of harm 96–7 killing to prevent rape 146 mistake 161 rape, meaning 148 retreat rule 74 self-generated self-defence 117–18 Canadian Charter of Rights and Freedoms 177 cannibalism 7 Christie, MGA 147 claim-right 59–60 Clarkson, CMV 10 coercion, see duress Cohen, JM 93 compulsion, see duress conduct rule 28 conjoined twins 8–9

212

Index

consequentialism culpable aggressor 47 definition 45 discount approach 46–7 future harm approach 48 innocent aggressors 46, 47 innocent bystander 50 multiple aggressors 47 passive threats 46, 47 permissibility of self-defensive killing 45–50 rule 48–9 Cranston, M 58 criminal defence classification 16–23 excuse 17–20 failure of proof defence 16 justification 17–20 meaning 13–15 offence modifications 16–17 culpable aggressor 5, 44, 47 culpable homicide 172 Dancy, J 24 Darley, JH 33–5 Davis, N 48, 51 de minimus defence 17 death as foreseen side-effect 53 deeds theory arguments against 35–7 consequences 37–41 development 31–3 justificatory circumstances 31 Robinson’s arguments for 26, 31–5 rule-articulation function 35 third party 34, 36 unknowingly justified accused 35–7 defence, meaning 14 defence of property approaches to 132–4 degree of force 132–7 ECHR position 181–3 home, in the 137–42 intruder 138–9 justification of killing 134–7 lethal force 133–5, 137, 138, 181–3 preservation of life, essential to 142 privacy argument 140–1 property within home 136, 139–40 proportionality 138 psychological harm argument 156 reasonable force 132 retreat from the home 83–5 right to life 131, 181–3 trespasser 137 valuable property argument 136, 141 defences fair labelling of 11 partial, see partial defences

‘defences of mental impairment’ 20 n. 41 Diamond, GA 95 Dicey, AV 126 Dillof, AM 25, 26 diminished responsibility 21, 108, 176, 195 diplomatic immunity 22 discount approach 46, 47–8 double effect doctrine 53–4 Dressler, J behaviour-guiding argument 27–9 duress 9 imminence of harm 20, 90 insanity 20 killing to protect property 140–1 moral argument 29–31 dual requirement theory 24–6, 37–41 see also reasons theory duress circumstances, pressure of 81 excuse defence, as 1, 5, 19, 21, 39, 121 foreseeable risk 120 necessary action, as 10–11 objective test 11 self-defence, distinguished from 9–10 terminology 6 duress of circumstances, see necessity East, EH 113, 115 ECHR, see European Convention on Human Rights Engberg, NJ killing to prevent rape 147, 150, 151, 152, 153–4, 157, 158 England defence of property 132, 137, 138–9 ECHR violation 192–5 excessive force as partial defence 172 excusable homicide 74, 75 gross negligence 193–4 justifiable homicide 74–5 mistake 161–3 retreat rule 70, 72, 73, 74–5 self-generated self-defence 114 unreasonable mistaken belief 171, 192–5 English Law Commission partial defences to murder 172 self-generated self-defence 114 unreasonable mistaken belief 171 entrapment 22, 23 Epps, G 43–4 error of law 21 Eser, A 49 European Convention on Human Rights breach of the peace 129 freedom from inhuman and degrading treatment 177, 181, 192–5 positive obligation 191–5

Index protection of Convention rights by state 191–5 right to respect for private and family life 192 European Convention on Human Rights – Article 2 (right to life) Andronicou case 185–6 Caraher case 186–8 duty to give reasons 178–9 exceptions 179, 180 generally 57, 177–81 Gül case 188 Kelly case 183–4 landfill site explosion 191 McCann case 184–5, 188, 189 mistaken belief 179, 183–90, 192–5 private citizens 191–5 procedural obligation 178 property, defence of 3, 179, 181–3, 197 self-defensive killing 179–81 state officials 190–5 unreasonable mistaken belief 192–5 violation 3, 188, 190, 191 evidential defences 16 n. 23 Ewing, C 94, 154 excessive force as partial defence 169–76 excusable homicide 74, 75 excuse see also duress; lack of capacity defences application of defence 21, 22, 121, 165 battered women 108 classification 21 compared with justification 17–20, 37, 39–41 innocent aggressor 5 insanity as 20 lack of capacity defences distinct from 19–20 learned helplessness 104 Faigman, DL 104 failure of proof defences 16, 22 Finkelstein, CO 44 Finnis, JM 57–8 Fletcher, GP criticism of deeds theory 32 insanity 20 justification and excuse, distinction between 17–18, 19 justified conduct 25–6 unknowing justification 32, 36, 37 unreasonable mistaken belief 166 forfeiture circumstances 66, 67, 68, 79 conditional right 61–2 conditions to be satisfied 66 conduct, association with 67, 68, 79 definition 67

213

innocent aggressor 63–4, 67 passive threat 63–4, 67 qualification, based on 61 right to life and 60–8 specification and 61, 67 temporary state, as 61 unnecessary self-defensive killing 64 use of term 67 Foster, M 113, 115, 143, 150 Gardner, J classification of defences 22 dual requirement theory 24–6 guiding reasons 33 moral argument 30 rape, dehumanizing effect of 156–7, 158 genocide 79 Germany defence of property 135 retreat rule 76, 79 Glover, J 43, 58 Gordon, GH 147 Greenawalt, K 20, 27, 29 Grisez, GG 64 gross negligence 193–4 Gur-Arye, M 49 Hale, M 77, 113 Hampton, J 156, 157 Harel, A 112 Hart, HLA 18, 43 Hawkins, W 143 Herald of Free Enterprise case 7–8 hierarchy of defences 22–23 Hobbes, T 60–1 Hogan, B 147 Hohfeld, WN 59–60 home, defence of, see defence of property honest belief 184, 185, 186–8 honest belief principle extension to indecent assault 162 rejection 162 Horder, J 9, 81, 90, 98, 168 hostage 97, 98, 101 n. 86, 104, 185–6 human shield 185–6 Hume, D imminence of harm 88 mistake 161 proportionality requirement 144–5 rape 150 Hurd, H 31, 33 Husak, D 13, 27, 28 imminence, definition 87 imminence of harm abolishing imminence and necessity 93–4 alternatives to requirement 93–9 battered women 87, 90–1, 92, 99, 107

214

Index

imminence of harm (cont.) conjoined twins 96 force to be immediately necessary 97 highly likely but not imminent 92–3 hostage situations 97, 98, 101 n. 86, 104 inevitability 102–8 interpretation, wide 94–6 jury decision-making 100–1 necessity test only 96–9 non-confrontational situations 87, 90–1, 92, 99, 100, 107 purpose of rule 88–9 reasonable belief 88 reinterpretation of term 95–6 replaced by necessity 99–102 requirement 88–9 strict approach, 88 terrorism 95–6, 97, 98, 100 time measurement 89 impossibility defence 16 India excessive force as partial defence 173 indifference argument 50 innocent aggressor consequentialism 46, 47 definition 5 forfeiture 63–4, 67 justification of self-defence 37, 44 personal partiality approach to 50–3 innocent bystander 44, 50, 60 insanity 5, 20, 21 intentional killing 43 International Covenant on Civil and Political Rights 57 intoxication 120, 121 intruder 138–9 Jones, P 57, 58 Jones, TH 147 jury 40, 82, 83 justifiable homicide 74–5 justification defences battered women 104, 108 behaviour-guiding argument 27–9 classification 21 conduct-guiding function 39–40 consequentialist approaches 45–50 deeds theory 31–7 defence, as 22, 164–5 double effect doctrine 53–4 dual requirement theory 24–6 excuse, compared with 17–20, 37, 39–41 hierarchy, position in 22–3 moral argument 29–31 permissibility 43–5 personal partiality approach 50–3 reality and 30 reasons theory 26–31

right to life approach 54–68 unknowing 23–37 Kadish, SH 20, 48–9 Kasachkoff, T forfeiture 63–4 justification, approaches to 50, 52 self-defence distinguished from necessity 7 Kates Jr, DB killing to prevent rape 147, 150, 151, 152, 153–4, 157, 158 kidnapper 151, 152 King, R 28 kneecapping 151, 152 Kremnitzer, M 49 lack of capacity defences 19–20, 21, 22, 23, 108 LaFave, WR 147 landlord ejecting drunken and abusive customer 170 learned helplessness 104 lethal force absolute necessity 182, 183, 184 mistaken belief 183–90 private citizens and 191–5 property and 181–3, 192 state officials 180–1, 183–90 Locke, J 56 McConnell, T 62 Macfarlane, LJ 55, 56 McMahan, J 47, 51 Maguigan, H 91 Mahoney, MR 106 mandatory life sentence, abolition 174 Mangiafico, J 33 manslaughter excessive force as partial defence 172 killing to prevent rape 149 mistake case law 160–3 deeds theory 31–7 defence of others 160 dual requirement theory 24–6 excessive force 169–70, 171–6 excuse defence, as 5 good reasons for belief 165–6, 184, 185–6, 188, 189 honest belief 161–3, 184, 185 honest but unreasonable belief 163 justified act 23, 24 labelling issue 174–5 legal options 160–3 necessary force 159, 169–76 perception 159 public officials 168, 184–90 reasonable 161–3, 164, 184

Index reasonableness requirement 168, 171 reasons theory 26–31 unknowingly justified 23 unreasonable 161–3, 166 unreasonable belief, types 167–8, 179, 183–90, 192–5 mistaken belief, see mistake Montague, P 46, 47, 52 moral argument 29–31 motivational reasons for behaviour 24 mountaineering 7 murder offence and defence, distinction 14–15 self-defence as partial defence 14, 173 wicked intent 163–4, 173 Muth, RJ 136 necessary action 10–11 necessity excuse defence 21 generally 1 proportionality, distinction between 5–6 self-defence, distinguished from 6–9, 10 use of term in self-defensive force 5 Netherlands breach of Convention obligations 192 non-age 2, 5, 20, 21, 23 non-exculpatory public policy defences: classification 22 entrapment 22, 23 non-fatal injury 195 Northern Ireland honest belief 186–8 imminence of harm 100 self-generated self-defence 115–16 offence, meaning 14 offence modification defence 16–17, 22 partial defences diminished responsibility 176 excessive force 169–176 provocation 14, 21, 108, 173, 175, 176, 195 passive threat forfeiture 63–4, 67 justification 44 personal partiality approach to 50–3 terminology 5 personal partiality approach 50–3 post traumatic stress disorder 155 privacy 140–1 property, defence of, see defence of property proportionality necessity, distinction between 5–6 use of term in self-defensive force 5–6 provocation 14, 21, 108, 173, 175, 176, 195

psychological harm argument 154–6 public order offence 128 pure self-defence 109 racism 167 rape case law 143–6 commission 145 n.14 dehumanizing, as 156–8 four types of harm 153 gender 144–5 meaning 147–8 permissibility of killing 146–9 physical impact 153–4 pregnancy as a result 153–4 principled argument 146–58 proportionality 148 psychological impact 154–6 serious bodily harm, equivalent to 151–2 suicide 155 without threat of death 148–9, 150 Raz, J 24 reasonable chastisement defence 192 reasons theory see also dual requirement theory behaviour-guiding argument 27–9 consequences 37–41 generally 26–7 moral argument 29–31 Resick, PA 154–5 resistance, right of 38, 39 responsibility 18 retreat absolute retreat rule 69–71, 76–7 battered women 84 Bird case 71, 72–3 generally 69 home 83–5 Julien case 70, 73, 76 jury, role of 82, 83 McBrearty case 71 no retreat rule 74–6, 77–80 rule-utilitarian argument 80 strong retreat rule 71–2, 76, 82–3 weak retreat rule 72–4, 75, 76, 80–82 right to life see also ECHR conditional, as 61–2 definition 55 forfeiture 60–6 fundamental, as 55, 58–9 generally 45, 54–68 innocent bystander and 50 loss is irremediable 58–9 natural law 57–8 self-defensive killing and 59–60 United Nations 56–7

215

216

Index

rights-based account, see right to life Ripstein, A 95 Robinson, PH classification of criminal defences 16–17 criminal defence definition 13 deeds theory 31–5 discount approach 46 excuses 17 failure of proof defences 16–17 imminence requirement 97–8, 101 n.86, 104 insanity 20 justification 17, 26, 28 non-exculpatory public policy defences 16, 22 rape 147 rule consequentialism 49 self-generated self-defence 114 Rosen, R 89, 100 rule-utilitarian argument 80 Ryan, C 65, 66 Schopp, RF 33 Scotland defence of property 132–3 excessive force as partial defence 172 imminence of harm 88 insanity 20 mistake 161, 163–4 murder requiring wicked intent 163–4, 173 proportionality requirement 144–5 rape 144–5, 150 retreat rule 71–2, 76 special defences 16 n.20 Sebok, AJ 94 self-defence duress, distinguished from 9–10 excuse defences as 21 mistake, see mistake necessity, distinguished from 6–9 subjective test 11 self-generated self-defence agreeing to fight 111, 123–5 behaviour likely to provoke violence 128–9 categories 110–11 communication of withdrawal 122–3, 124 competing values 125–7 conduct of defendant 116 conduct of initial aggressor 117 consensual fight 124–5 deliberate generation of situation 110, 113–14 denial of defence 124–5 disproportionate response 110, 118–21 extent of self-defence 115 failing to avoid trouble 111, 125–9

legal options 111–13 meaning 109 Northern Ireland case 115–16 proportionate response 110, 122–3 starting the trouble 110, 114–23 terminology 109–10 self-preservation 7–8 Seligman, M 103–4 separation assault 106 shipwreck 7 shooting defensive 167 mistaken belief 168 n.66, 184–90 Shute, S 15, 156–7 Simester, AP 136, 147, 164 Singer, R 166, 168 Smart, BJ 65 Smith, JC 147, 180, 195 Smith & Hogan 73 South Africa killing to protect property 133–5 South African Constitution 177 special defences 16 n.20 Sullivan, GR 136, 146 superior orders 21 Tadros, V 29 terminology, explanation of use 5–6 terrorism duress defence 11 imminence of harm 95–6, 97, 98, 100 suspected terrorist 188 value of life 136 third party assistance 34, 39 harm to 36 Thomson, JJ consequentialism 50 forfeiture 61, 62, 64, 65 right to life 50, 57, 58, 59 time-bar 22 trespasser 137 Uniacke, S consequentialism 50 double effect doctrine 54 forteiture 61–2, 64–5 right to life 2, 45, 57, 58, 59 unified defence 10–11 United Nations Universal Declaration of Human Rights 57 United States of America American Convention on Human Rights 57 American Declaration of Independence 1776 56 American Declaration of the Rights and Duties of Man 57

Index battered women 90, 91, 94, 95, 97 Bill of Rights 177 defence of property 83–5, 133, 137, 139, 141 home 83–5, 133, 137 imminence of harm 88, 90, 91, 95, 97 mistake, 160, 161, 163 no retreat rule 77–8 rape 145, 155 retreat rule 72, 73–4, 75, 76, 77–8, 80, 83–5 self-generated self-defence 113, 118, 123, 124 workplace 83–5 unreasonable mistake acquittal 161–3 appropriate punishment 166 English law 161–3 unreasonable mistaken belief see also mistake ECHR violation 192–5

non-fatal injury 195 partial defence, as 194–5 utilitarianism 45 Veinsreideris, ME 99 victim, status 10 violence burglary and 139–40 domestic 91 voluntary intoxication 120, 121 Walker, L 103, 104 Wallerstein, S 63, 64 Wasserman, D 46 Westen, P 33 wicked intention to kill 163–4, 173 Williams, G 14 workplace attack 83–5 Wright, AJ 104 Young, R 59

217