Torture and Moral Integrity: A Philosophical Enquiry 0198714203, 9780198714200

Citation preview

TO RT U R E A N D M O R A L I N T E G R I T Y

Torture and Moral Integrity A Philosophical Enquiry By M AT T H E W H .   K R A M E R

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom 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. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © M Kramer 2014 The moral rights of the author have been asserted First Edition published in 2014 Impression: 1 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, by licence 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 work in any other form and you must impose this same condition on any acquirer Crown copyright material is reproduced under Class Licence Number C01P0000148 with the permission of OPSI and the Queen’s Printer for Scotland Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2013957961 ISBN 978–0–19–871420–0 Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

To Norman Geras

Preface My desire to write a book on torture and moral integrity was sparked several years ago by some exchanges with Norman (‘Norm’) Geras, to whom I gratefully and admiringly dedicate this volume. Many other people likewise deserve my thanks for their contributions to my production of this book. I presented early versions of portions of Chapter 1 in lectures at several locations: the University of Bologna in May 2011; the University of Barcelona in February 2012; Tel Aviv University and Hebrew University (Jerusalem) in March 2012; the University of Genoa in June 2012; and the University of Manchester in February 2013. I  am extremely grateful to my hosts and audiences on all those occasions. Without intending to slight anyone else, I  should single out the following people for especially warm thanks:  David Brink, Carla Faralli, Michael Moore, Enrico Pattaro, Tomasz Pietrzykowski, Wojciech Sadurski, and Chiara Valentini (at Bologna); Serena Olsaretti and Andrew Williams (at Barcelona); Chaim Gans, Daniel Klerman, Tamar Meisels, and Amit Pundik (at Tel Aviv); Alon Harel and Re’em Segev (at Jerusalem); Ian Carter, Bruno Celano, Pierluigi Chiassoni, Paolo Comanducci, Diego Moreno Cruz, Jose Juan Moreso, Valeria Ottonelli, and Giovanni Battista Ratti (at Genoa); and Richard Child, Mark Reiff, Liam Shields, and Tom Sinclair (at Manchester). In addition, I am grateful to Tony Ward for supplying me with a copy of his valuable collection of essays on torture. Ian Carter and Richard Child have identified themselves to me as two of the three originally anonymous readers of the typescript. I am greatly obliged to them (and to the third reader) for their valuably perceptive suggestions and queries. I am likewise much obliged to Alex Flach, Natasha Flemming, Clare Kennedy, and their colleagues at the Oxford University Press for their characteristically adept handling of the process of publication. Cambridge, England September 2013

Contents 1. Introduction I: Moral Conflicts and Deontology 1.1. Moral Conflicts 

1.1.1. The disambiguation of some key concepts  1.1.1.1. Two senses of ‘prima facie’  1.1.1.2. Weak permissibility versus strong permissibility  1.1.1.2.1. Two types of obligations and two types of permissibility  1.1.1.2.2. Infringements versus violations  1.1.1.3. Strong justification versus weak justification  1.1.1.4. Two senses of ‘rightness’  1.1.1.5. Two senses of ‘absolute’  1.1.1.6. Overtopping versus overriding 1.1.2. Chariness of moral conflicts 1.1.2.1. Consequentialist balancing 1.1.2.2. The objectivity of morality 1.1.2.3. Logical incoherence 1.1.2.4. Moral conflicts and action-guidance 1.1.2.5. Concerns about fairness or excessive onerousness

1.2. The Deontology/Consequentialism Distinction and the General Structure of Morality

1.2.1. The division between deontology and consequentialism 1.2.1.1. Intrinsic moral statuses 1.2.1.2. Agent-neutrality versus agent-centredness 1.2.1.3. Absolute prohibitions 1.2.2. The general structure of morality 1.2.2.1. Moore’s tripartite account 1.2.2.2. An alternative account of morality 1.2.2.3. A pithy conclusion: some differences between the accounts of morality

2. Introduction II: What is Torture? 2.1. Definitional Ventures

2.1.1. Convention against Torture 2.1.1.1. A few queries 2.1.1.2. Some commendations 2.1.2. The American definition 2.1.3. Amnesty International’s definition 2.1.4. Philosophers’ definitions 2.1.4.1. Michael Davis on the ordeal of torture 2.1.4.1.1. A first query: the helplessness of victims 2.1.4.1.2. A second query: testing the victim’s capacity to endure suffering

1 2 2 2 4

4 6 7 8 8 10 11 11 14 16 18 19

20 20 20 21 22 24 25 26 27

29 30 30 31 33 34 36 37 37 37 39

x

Contents 2.1.4.1.3. A third query: psychological torture revisited 2.1.4.1.4. A fourth query: against the victim’s will 2.1.4.1.5. A final query: indifference to the victim’s welfare 2.1.4.2. Kershnar’s formulation 2.1.4.3. Miller and the complexities of torture 2.1.4.3.1. The third clause 2.1.4.3.2. The second clause 2.1.4.4. Sussman and the difficulties of defining torture 2.1.4.4.1. A preliminary point: a problematic addition 2.1.4.4.2. A matter of consent 2.1.5. Some of the lessons of this survey of definitions

2.2. Varieties of Torture

2.2.1. Interrogational torture 2.2.1.1. Prospective versus retrospective 2.2.1.2. Some varieties of prospective interrogational torture 2.2.1.3. Extreme emergencies 2.2.2. Placatory torture 2.2.2.1. A matter of intentions 2.2.2.2. A certain commitment 2.2.2.2.1. A first difference 2.2.2.2.2. A second difference 2.2.2.2.3. A third difference 2.2.2.2.4. A matter of importance 2.2.3. Intimidatory torture 2.2.4. Extortionate torture 2.2.5. Act-impelling torture 2.2.6. Punitive torture 2.2.7. Sadistic torture 2.2.8. Discriminatory torture 2.2.9. Humiliative torture 2.2.10. Extravagantly reckless torture 2.2.11. Incapacitative torture 2.2.11.1. Lastingly incapacitative torture 2.2.11.2. Ephemerally incapacitative torture 2.2.11.2.1. The act/omission dichotomy 2.2.11.2.2. Fending off the most common objection to Bennett’s analysis 2.2.11.2.3. Sussman’s example of the obese man 2.2.11.2.4. Steinhoff’s example of the rapist 2.2.11.2.5. Steinhoff’s example of the snake bite 2.2.11.2.6. Kamm on torturing from a distance 2.2.11.2.7. From Kamm to Kantians 2.2.11.2.8. Hill and the act/omission distinction 2.2.11.2.9. Hill and the act/omission distinction redux 2.2.12. Edifying torture 2.2.12.1. Salvation-oriented torture 2.2.12.2. Therapeutic torture 2.2.12.2.1. Experimentational torture

42 44 46 48 49 49 50 50 51 51 53

56 56 57 58 59 62 62 63 64 64 65 66 66 68 68 69 72 73 73 74 76 76 77 78 80 84 87 88 90 92 94 96 98 98 98 99

Contents

xi

2.2.12.2.2. Aversion therapy 100 2.2.12.2.3. Averting a coma 101 2.2.12.2.4. A worry about the conflation of distinct issues102 2.2.12.3. Resistance training 103 2.2.13. A table of the main kinds of torture 104

2.3. Conclusion: An Overview

2.3.1. The infliction of severe pain 2.3.1.1. Torture versus attempted torture 2.3.1.1.1. A different question 2.3.1.1.2. A question put aside 2.3.1.1.3. A matter of gravity 2.3.1.1.4. Back to the definition 2.3.1.2. How long? 2.3.2. Against the interests of the victim? 2.3.3. Consent and control 2.3.3.1. Two caveats about control 2.3.3.2. The upshot of the matter 2.3.4. A definition of torture

3. Why Torture is Wrong 3.1. Some Previous Accounts of the Wrongness of Torture

3.1.1. Contractarian approaches 3.1.1.1. Nagel and justifiability 3.1.1.2. Meisels and the social-contract tradition 3.1.1.2.1. A terse critique 3.1.1.2.2. A possible reply 3.1.1.2.3. Another possible reply 3.1.2. Shue on the defencelessness of victims 3.1.2.1. Putting aside an issue 3.1.2.2. On defencelessness as the wrong-making property of torture 3.1.2.3. A loss of one’s ideals? 3.1.2.3.1. A query 3.1.2.3.2. Another query 3.1.3. Shue and others on the spread of torture 3.1.3.1. A weak argument 3.1.3.2. Another weak argument 3.1.3.2.1. A first reply 3.1.3.2.2. A second reply 3.1.3.3. Empirical speculations and slippery slopes 3.1.3.3.1. Two clarifications 3.1.3.3.2. An additional clarification 3.1.3.3.3. A first example 3.1.3.3.4. A second example 3.1.3.3.5. A first objection 3.1.3.3.6. A second objection: preliminary clarifications 3.1.3.3.7. A second objection continued: the meagreness of the evidence 3.1.3.3.8. A closing caveat

104 105 105 105 106 108 109 109 111 112 112 113 114

115 118 118 118 120 122 123 125 127 127 128 130 130 131 132 133 134 135 136 137 138 139 140 142 143 145 146 148

xii

Contents 3.1.4. From consequentialism to Kantianism: torture and agency 3.1.4.1. Waldron on torture and dignity 3.1.4.2. Sussman on the limits of Kantianism 3.1.5. Sussman on torture and self-betrayal 3.1.5.1. Techniques of torturous self-betrayal 3.1.5.2. The source of the self-betrayal 3.1.5.3. Some transitional ruminations on Sussman’s theory

3.2. Why Torture is Wrong

3.2.1. The consumingness of severe pain 3.2.1.1. The consumingness of euphoria 3.2.1.2. Some observations by philosophers and other theorists 3.2.1.2.1. Beccaria on the filling of the sensory field 3.2.1.2.2. Luban on the tyranny of severe pain 3.2.1.2.3. Kreimer on the occupation of the self 3.2.1.2.4. Scarry on the body in severe pain 3.2.1.3. The perils of overstatement 3.2.1.3.1. Preliminary remarks 3.2.1.3.2. Hyperbole best avoided 3.2.2. Combining two insights 3.2.2.1. The two main strands 3.2.2.2. The combination 3.2.2.2.1. Not enough in isolation 3.2.2.2.2. Morally vitiating purposes 3.2.2.2.3. The Minimal Invasion Principle 3.2.2.2.4. Consequentialist calculations 3.2.2.2.5. Agony and oppression: the factors of consent and control afresh 3.2.2.2.6. Some implications: edifying torture and sado-masochism 3.2.2.2.7. Some implications: the problem of animals 3.2.2.2.8. Agony and oppression redux: a recapitulation and a transition 3.2.3. A perpetrator-focused perspective 3.2.3.1. Monstrous victims of torture 3.2.3.2. Being killed versus being tortured 3.2.3.3. An additional perspective 3.2.3.3.1. A perpetrator-focused justificatory basis 3.2.3.3.2. A first query: why is ephemerally incapacitative torture ever permissible? 3.2.3.3.3. Four caveats concerning my response to the first query 3.2.3.3.4. A second query: why is deliberate killing ever morally permissible? 3.2.3.3.5. Clarifying the issue 3.2.3.3.6. A third query: why is highly restrictive confinement ever morally permissible? 3.2.3.3.7. A fourth query: why is punitive torture absolutely wrong?

149 150 152 155 157 158 160

161 161 162 165 165 165 166 166 168 169 169 173 173 174 175 176 177 179 180 182 183 185 187 187 188 189 190 192 194 197 199 201 203

Contents 3.2.3.3.8. Retributivism and the perpetrator-focused perspective 3.2.3.3.9. A fifth query: why is consensual placatory torture impermissible? 3.2.3.3.10. A final query: why is sado-masochistic torture morally wrong?

3.3. Moral Optimality without Moral Permissibility 3.3.1. A first example 3.3.2. A second example 3.3.3. Some general considerations 3.3.3.1. Harshness and protractedness 3.3.3.2. The exigencies of an emergency 3.3.3.3. Threateningness and responsibility 3.3.3.4. Probable efficacy 3.3.3.5. Legal sanctions

xiii 204 209 210

212 212 213 215 215 216 217 217 219

4. The Rationality of Deontological Constraints 4.1. Placatory Torture and the Unremittingness of Deontological Duties

221

4.2. Are Deontological Constraints Irrational?

230

4.1.1. A thought-experiment: sparing someone from a greater wrong 4.1.2. The significance of the thought-experiment 4.1.3. The moral upshot 4.1.3.1. The factor of consent 4.1.3.2. The orientation of the torture 4.1.3.3. Perpetrator-focused reflections 4.1.3.4. Moral optimality 4.2.1. Rationality and maximization 4.2.2. Deontological commitments 4.2.3. Slippage from none-versus-any to fewer-versus-more 4.2.3.1. A first example of the conflation 4.2.3.2. A second example 4.2.3.3. Goals for deontologists: a first example 4.2.3.4. Goals for deontologists: a second example 4.2.4. The maximizing conception of rationality redux 4.2.5. A return to moral optimality

5. Legal Responses to Torture 5.1. Legal Approval Ex Ante?

5.1.1. Dershowitz and torture warrants 5.1.1.1. A first objection to Dershowitz: inapposite comparisons 5.1.1.2. A second objection to Dershowitz: a missing prohibition 5.1.1.3. A third objection to Dershowitz: inordinate narrowing of the options 5.1.1.3.1. A shortcoming in Dershowitz’s reply 5.1.1.3.2. A possible response by Dershowitz 5.1.1.3.3. A second possible response by Dershowitz

223 224 225 226 226 227 228 229 231 232 234 234 235 235 237 238 240

242 243 243 245 247 248 249 250 250

xiv

Contents 5.1.1.3.4. Peculiarly worrisome problems 5.1.1.3.5. A first rejoinder to Dershowitz: torture without warrants 5.1.1.3.6. A second rejoinder to Dershowitz: arguments in tension 5.1.1.3.7. The second rejoinder to Dershowitz continued 5.1.1.3.8. The second rejoinder to Dershowitz completed 5.1.1.3.9. A third rejoinder to Dershowitz: ways of dealing with the problems 5.1.1.3.10. A third rejoinder to Dershowitz continued 5.1.1.3.11. A third rejoinder to Dershowitz completed 5.1.1.4. A fourth objection to Dershowitz: seeking support from a strange quarter 5.1.1.5. A final objection to Dershowitz: a matter of moral principle 5.1.2. Posner and Vermeule on the regulation of torture 5.1.2.1. Posner and Vermeule on deontological absolutism 5.1.2.1.1. Missed distinctions and an inapposite focus 5.1.2.1.2. A charge of fanaticism and an easy target 5.1.2.1.3. Tragic choices as moral conflicts 5.1.2.2. Posner and Vermeule on the Legal Prohibition Thesis 5.1.2.2.1. Empirical conjectures 5.1.2.2.2. Exoneration is tantamount to approval 5.1.2.2.3. Not many operational differences 5.1.2.2.4. An objection of moral principle 5.1.2.2.5. Further remarks on the symbolism of authorizations of interrogational torture 5.1.2.3. The analogy between torture and killing 5.1.3. Legitimate techniques of interrogation 5.1.3.1. Waldron’s distrustful arguments 5.1.3.2. A partial rejoinder to Waldron

5.2. Legal Accountability Ex Post

5.2.1. Which defences? 5.2.1.1. Mitigations rather than justifications or excuses 5.2.1.2. Two preliminary caveats 5.2.1.3. Necessity or protection-of-oneself-or-others? 5.2.1.4. A residual role for the defence of duress 5.2.2. Which sanctions? 5.2.2.1. Criminal sanctions imposed on individual officials 5.2.2.2. Civil sanctions applied to individual officials 5.2.2.3. Institutional sanctions against individual officials 5.2.2.4. Collectively borne sanctions 5.2.3. Private individuals as defendants 5.2.3.1. Collective responsibility 5.2.3.2. Some public/private similarities relating to individually borne sanctions 5.2.3.3. Some public/private differences relating to individually borne sanctions 5.2.3.3.1. An example already encountered

253 253 256 257 259 259 260 261 262 264 266 267 267 268 268 271 271 273 275 277 278 280 282 283 285

287 288 288 290 291 292 293 293 294 295 296 298 298 299 300 301

Contents 5.2.3.3.2. A second example 5.2.3.3.3. Sado-masochistic torture once more 5.2.3.3.4. A rejoinder? 5.2.3.3.5. Sado-masochistic torture and the harm principle

xv 302 305 306 308

5.3. The Complexities of Involvement

309

5.4. Conclusion

316

5.3.1. Extradition, deportation, and extraordinary rendition 5.3.2. Evidence produced by interrogational torture 5.3.3. Article 3 and Article 15 conjoined

309 311 314

References

317

Index

327

1 Introduction I: Moral Conflicts and Deontology We are oft to blame in this, ‘tis too much proved, that with devotion’s visage and pious action we do sugar o’er the devil himself. [Hamlet, III.i.46–9.]

This book is about the wrongness of torture and about the nature of morality. It deals at length with multiple types of torture, and it seeks to explain why the types that figure most prominently in contemporary philosophical discussions are always morally wrong. At the same time, it plumbs the general structure of morality and the intricacies of moral conflicts, and it probes some of the chief grounds for the moral illegitimacy of various modes of conduct. It tackles a concrete moral problem— a problem heatedly debated during recent decades in the governmental and military institutions of many countries as well as in academic circles—and it likewise tackles some very abstract issues in moral and political philosophy. Moreover, as will become apparent at numerous junctures, the abstract ruminations and the concrete prescriptions are closely connected. My abstract reflections on the structure of morality are the vital background for my approach to torture, and my approach to torture is a natural outgrowth of those abstract reflections. In keeping with the overall character of this book, its Introduction is divided into two chapters. This opening chapter will address some important matters in the more abstract reaches of moral philosophy—as it disambiguates several key concepts in order to clarify the import of moral conflicts, and as it elucidates the distinction between deontological obligations and consequentialist obligations. Along the way, it will delineate the general structure of morality. Chapter 2 will then move to a more concrete level of philosophizing, by proceeding to recount many of the complications that surround any attempt to distil the defining characteristics of torture. These discussions in the first two chapters, at quite differing tiers of abstraction and concreteness, will form the basis for the subsequent chapters’ treatment of the wrongness of torture and the value of moral integrity. Chapter 3 will endeavour to explain why torture is morally wrong; Chapter 4 will defend my absolutist stance on torture against a powerful consequentialist critique; and Chapter 5 will prescribe how the law in any jurisdiction should address the matter of torture. Both in those later chapters and in these introductory chapters,

2

Introduction I: Moral Conflicts and Deontology

the mutual supportiveness of the abstract argumentation and the concrete argumentation in this book will become evident.

1.1.  Moral Conflicts To come to grips philosophically and morally with the problem of torture, this book needs to examine the phenomenon of moral conflicts. Throughout the volume, I usually take a moral conflict to be a situation in which some person P is under a moral duty-to-φ and simultaneously under a moral duty-not-to-φ.1 However, in line with most other philosophers who write on these issues, I also apply the phrase ‘moral conflict’ to any situation in which P simultaneously bears moral duties whose contents are contraries rather than contradictories. A situation of the latter sort obtains when P is under a duty to do x and simultaneously under a duty to do y, where (1) ‘P does x’ entails the negation of ‘P does y’ and (2) ‘P does y’ entails the negation of ‘P does x’ and (3) the negation of ‘P does x’ is logically consistent with the negation of ‘P does y’. Though the clashing duties within a moral conflict can perfectly well coexist, they can never be jointly fulfilled; the fulfilment of either of them entails the non-fulfilment of the other.

1.1.1.  The disambiguation of some key concepts Before we directly ponder a number of queries that have been raised about the occurrence of moral conflicts, this chapter should disambiguate some key concepts that figure saliently in discussions of these matters and specifically in discussions of torture. An examination of several overlapping and oft-obscured distinctions will not merely help to avert confusion, but will likewise illuminate a number of the points of contention that surface in the later portions of this book. Attentiveness to those distinctions will greatly enhance one’s understanding of the nature of moral conflicts.

1.1.1.1.  Two senses of ‘prima facie’ As I have observed elsewhere (Kramer 1999a, 267–9; 2004, 290–1; 2011, 53–4), the phrase ‘prima facie’ is construable in two quite divergent ways. On the one hand, it can mean ‘presumptively’ or ‘upon initial examination’. When the phrase is used in this epistemic sense, a person who declares that some state of affairs prima facie obtains is saying that an initial inspection of the matter indicates that the specified state of affairs does obtain. As Susan Hurley wrote: ‘Prima facie reasons are like rules of thumb, that give us reasons provisionally but may turn out not to apply when we learn more about the situation at hand, in which case they have no residual reason-giving force’ (Hurley 1989, 133). This pattern of usage   The variable ‘ φ ’ can stand for any verb(s) or verb phrase(s), denoting any action(s) or omission(s).

1

Moral Conflicts

3

carries over into the workings of legal institutions, where conclusions that have prima facie been demonstrated are conclusions that are tentatively deemed to be demonstrated in light of the incomplete evidence that has been submitted. (In the law, the notion of prima facie evidence is connected with issues relating to the distribution of the burden of proof between plaintiffs and defendants.) Insofar as this pattern of usage is regnant in moral and political philosophy, theorists who discuss prima facie moral duties are referring to properties or states that are believed upon first inspection to be moral duties. No such property or state will qualify as a full-fledged moral duty unless the initial consideration of the factors that bear on its existence as a duty is confirmed by subsequent investigations of all further relevant factors (if any). In short, when things are classified as prima facie moral obligations in this fashion, only some of them are veritable moral obligations. Many are mere appearances that turn out not genuinely to be moral obligations when all things have been considered. On the other hand, ‘prima facie’ can mean ‘pro tanto’ or ‘susceptible to being overtopped in importance by some competing factor’. When the phrase is used solely in this second way, a prima facie moral duty is a genuine moral duty that might be exceeded in stringency or importance by a countervailing moral imperative. In this vein, a merely prima facie obligation is contrasted with an overtoppingly stringent obligation. That is, the relevant dichotomy lies between any moral duty that does not exceed every competing moral requirement in importance and any moral duty that does. Here the prima facie status of a moral duty has nothing to do with first appearances or tentative identifications. If someone is under a prima facie moral obligation in this sense and not in the epistemic sense, he is under it as a result of all relevant considerations. Inasmuch as ‘all things considered’ is construed as an epistemic phrase that means ‘in accordance with what would be found by a thorough investigation’, every moral duty that is prima facie solely in the sense of not being overtoppingly stringent in some possible situations (rather than in the sense of being an initial appearance) is an all-things-considered moral duty. The standing and scope of any such duty as a moral obligation are determined by the full array of circumstances in which the obligation obtains. This point has sometimes been obfuscated, not least by David Ross’s suggestion that any prima facie moral obligation consists in a ‘[t]‌endency to be one’s duty’ (Ross 1930, 28). Such a formulation promotes confusion, for it fails to distinguish between a tendency to be a duty of any sort and a tendency to be an overtoppingly stringent duty.2 Any number of considerations can engender a tendency of the former kind. For example, the pain and indignity that would be suffered by Edward in the event of his undergoing a thrashing by Frank are factors that militate in favour of Frank’s being morally obligated to forbear from thrashing Edward in any particular context. Those factors do not always produce or sustain such an 2   This distinction is missed by Joseph Raz when he draws a parallel between (i) the liberty of courts in some circumstances to overrule precedents which they would in other circumstances be obligated to apply and (ii) the overtoppingly stringent duty of a promisor to decline to abide by her promissory obligation in some circumstances. See Raz 1979, 114.

4

Introduction I: Moral Conflicts and Deontology

obligation, however. If Edward himself launches a serious and unprovoked assault against Frank, then Frank is morally at liberty to use as much force as is necessary to fend off Edward’s attack. He is not under any moral obligation at all—not even a merely prima facie duty, much less an overtoppingly stringent duty—to abstain from landing blows on Edward in such circumstances. At first glance we might think that Frank is under at least a merely prima facie moral duty to restrain himself from using violence against Edward, but a more detailed exploration of his predicament reveals that not even a merely prima facie moral duty is applicable. There are factors that tend toward the existence of a duty, but none of them actually eventuates in it. Such a situation differs markedly, then, from a situation in which a veritable prima facie moral duty does exist while being overtopped by a weightier moral requirement. A veritable prima facie moral obligation does not tend toward the existence of an obligation; it is a moral obligation. What it tends toward is the existence of an overtoppingly stringent moral obligation. If that latter tendency is fulfilled in any particular context, the moral duty in question prevails over all conflicting moral requirements. If the tendency toward overtopping stringency is instead checked by the existence of some competing moral requirement that is more important, the prima facie duty still obtains as such. Even when overtopped in importance, it is indeed a moral obligation rather than a bare appearance or a sheer factor (like Edward’s pain) that contributes toward the existence of such an obligation. Henceforth, whenever this book uses ‘prima facie’ without any explicit signal that those words are to be construed in the first sense above, it is employing the phrase in the second sense. That is, instead of being perceived as an epistemic caveat, the phrase ‘prima facie’ (in the absence of any explicit signal of the sort just mentioned) should be understood herein as an indication that a duty under discussion is susceptible to being surpassed in normative importance by some conflicting moral requirement. When a moral obligation is indeed so surpassed, it is merely prima facie—which, again, does not detract at all from its status as a fully genuine moral obligation that cannot be breached without the incurring of remedial duties.

1.1.1.2.  Weak permissibility versus strong permissibility In application to human conduct, permissibility and wrongness are contradictories. Any type or instance of human conduct is permissible if and only if it is not wrong. Now, if some person X is permitted to perform some action q, then X is not obligated to not perform q. (To avoid any ambiguities in my prose, my placement of ‘not’ in several of the sentences in this subsection will create some ugly split infinitives.) So much is clear, but we need here to take account of two ways in which someone can be morally obligated to do or not do something. 1.1.1.2.1.  Two types of obligations and two types of permissibility The germane distinction, which has already been invoked at a few junctures above, is between overtoppingly stringent and non-overtoppingly stringent moral

Moral Conflicts

5

obligations.3 An overtoppingly stringent moral requirement exceeds in importance all the moral duties that run counter to it, or is unopposed by any competing moral duties. A  non-overtoppingly stringent moral requirement R does not exceed in importance all the moral duties that run counter to it. (Any competing moral requirements might be equal in importance to R, or they might exceed it in importance, or they might be insusceptible to any determinate comparisons with it because of problems of incommensurability.) With reference to these two broad types of obligations, we can apprehend two broad types of permissibility: Weak Permissibility. Some person X is permitted to perform some action q if and only if X is not under any overtoppingly stringent obligation to not perform q.

X’s being weakly permitted to perform q is consistent with the proposition that X is under a non-overtoppingly stringent obligation to not perform q. In that respect, weak permissibility differs from strong permissibility. Strong Permissibility. X is permitted to perform q if and only if X is neither under an overtoppingly stringent obligation to not perform q nor under a non-overtoppingly stringent obligation to not perform q.

X’s being strongly permitted to perform q is inconsistent with the proposition that X is under a non-overtoppingly stringent obligation to not perform q. Note that the proposition ‘X is strongly permitted to perform q’ entails the proposition ‘X is weakly permitted to perform q’, but not vice versa.4 Whenever this book uses the term ‘permissible’ or ‘permissibility’ without any qualification, it is invoking the notion of permissibility in the strong sense. Consequently, the term ‘impermissible’ is to be understood herein as ‘not strongly permissible’. That is, I generally assume that an action q is impermissible for a person X unless X is not under any obligation whatsoever to not do q. For the purpose of gauging whether any type or instance of conduct is impermissible, this book will generally not discriminate between situations in which someone is under an overtoppingly stringent obligation to eschew q and situations in which someone is only under a non-overtoppingly stringent obligation to eschew q. Precisely because my ascriptions of impermissibility will generally not discriminate between those two kinds of situations, this book will contend that morally optimal courses of action can be morally impermissible. If two moral duties clash, and if one exceeds 3   See Kramer 2004, 280–1. Some closely related distinctions are sustainedly brought to the fore in Sinnott-Armstrong 1988. Although the term ‘overriding’ is much more common than ‘overtopping’ in discussions of these matters, I disfavour the former term because it conveys the impression that a less important duty is eliminated or cancelled in any conflict with a more important duty. We shall explore this point shortly. 4   Michael Moore has also posed a distinction between weak permissions and strong permissions (Moore 2007, 42–4), but his distinction is markedly different from mine. Likewise very different from my own distinction is the contrast between weak permissions and strong permissions that was drawn half a century ago by G. H. von Wright (1963, 86). In the text here and elsewhere, I am invoking only my own weak/strong dichotomy. Although I have expounded that dichotomy here with reference to actions, it is equally applicable—mutatis mutandis—to omissions.

6

Introduction I: Moral Conflicts and Deontology

the other in importance, then compliance with the former is a morally optimal but impermissible course of conduct. The moral optimality of such compliance will extenuate but not eliminate the breach of duty involved; it will therefore not eliminate the impermissibility of the compliant course of conduct. These attributions of impermissibility are sustainable because I generally rely on the strong conception of permissibility. (Throughout this book, I use ‘legitimate’ interchangeably with ‘permissible’ and ‘illegitimate’ interchangeably with ‘impermissible’.) 1.1.1.2.2.  Infringements versus violations Note that this exposition of the two types of permissibility can refine one’s understanding of a distinction which was originally expounded by Judith Jarvis Thomson (and slightly later by Joel Feinberg) and which has been quite conspicuous in political and moral philosophy ever since: the distinction between infringements and violations of rights.5 Under the terms of that dichotomy, all violations are infringements, but not all infringements are violations. An infringement occurs when someone has a right that p be the case—where ‘p’ designates a certain proposition—and someone else causes p to be false. For example, if Henry has a right that Ron not bang him in the nose, then Ron infringes the right if he bangs Henry in the nose. A violation is a culpable infringement. If Ron’s banging of Henry’s nose is attributable to malice or recklessness or carelessness, then the infringement of Henry’s right is a violation thereof. Contrariwise, if the banging occurs despite Ron’s careful efforts to avoid it, the infringement is a mere infringement. Proponents of the infringement/violation distinction generally maintain that mere infringements are morally permissible. Only violations are morally impermissible. In other words, these philosophers take culpability to be a necessary condition for impermissibility and thus for wrongdoing. Now, given that an infringement is a contravention of a moral right and is thus a breach of a moral duty, these philosophers may seem clearly mistaken in contending that infringements are morally permissible. Were they adhering to this book’s terminological patterns, they would indeed be mistaken. However, as becomes evident when one peruses their relevant writings, the exponents of the infringement/violation duality have typically relied (at least implicitly) on the weak conception of permissibility when submitting that mere infringements of people’s rights are permissible. They maintain that, so long as any person X is conducting himself in accordance with all the overtoppingly stringent moral duties that are incumbent upon him, X is conducting himself permissibly. The fact that he might be acting athwart a non-overtoppingly stringent moral duty— for example, by breaking an engagement for lunch in order to go to the aid of a seriously injured pedestrian—is compatible with the weak permissibility of his course of action. In their repeated affirmations of the permissibility of mere infringements, then, 5   See Thomson 1986, chaps 3–5; Thomson 1990, 122; Feinberg 1980, 229–32. For some discussions of the infringement/violation dichotomy, see Botterell 2008; Cane 2002, 107; Coleman 1992, 282–3, 299–302; Fletcher 1985, 977; Fletcher 1993, 175, 177; Lee 2012, 144–6; McConnell 1996, 42; Oberdiek 2004; Oberdiek 2008; Parent 1980, 406–8; Rainbolt 2006, chap. 6; Simester 2008, 297–8; Sinnott-Armstrong 1988, 51–2.

Moral Conflicts

7

Thomson and Feinberg and their followers are not really committing any errors. Instead, they are exhibiting their adherence to a conception of permissibility that diverges from my own conception. Still, although the defenders of the infringement/violation contrast are not committing outright errors, their understanding of permissibility as weak permissibility should not commend itself to anyone who wants to grasp the complexities of the moral bearings of conduct such as torture in extreme situations. After all, irrespective of whether a moral duty is overtoppingly stringent or non-overtoppingly stringent, any breach of it will give rise to remedial obligations. In the scenario mentioned just above, for instance, the rescuer of the injured pedestrian will have incurred a moral obligation to remedy his breach of a promissory duty (most likely through an apology and a brief explanation). Of course, to say as much is scarcely to suggest that the division between overtoppingly stringent obligations and non-overtoppingly stringent obligations will play no role whatsoever in my analyses. For example, that division plays a key role in determining the sizeableness of the remedy that is required in the aftermath of a breach of some moral duty. A breach undertaken in order to avoid a transgression of an overtoppingly stringent moral duty will be extenuated more heavily than a breach undertaken in order to avoid a transgression of a non-overtoppingly stringent moral duty. Nevertheless, the need for a remedy of some sort in the aftermath of the contravention of a moral duty is unaffected by the status of the contravened duty as overtoppingly stringent or non-overtoppingly stringent. Because of the irrelevance of the overtopping/ non-overtopping distinction in that crucial respect, this book adheres to the strong conception of permissibility throughout (except when the weak conception is clearly signalled, usually as I am expounding someone else’s views).

1.1.1.3.  Strong justification versus weak justification Closely cognate to the distinction between strong permissibility and weak permissibility is another dichotomy, between strong justification and weak justification (Kramer 2011, 54). Suppose that some person P asserts that the use of torture as a technique of interrogation can be morally justified in certain situations. On the one hand, P might be declaring that the use of interrogational torture can sometimes be strongly permissible as well as morally obligatory. That is, P might be contending that such a technique in some possible circumstances will not be in violation of any moral obligations. On the other hand, P might instead be declaring that moral obligations which sometimes require the use of torture as an interrogational technique are so dauntingly weighty in certain contexts that they can exceed in ethical importance the potently stringent moral obligations which forbid any such use. Construed in this latter way, P’s assertion would not be denying that the use of torture for interrogation is always morally impermissible; P would simply be maintaining that that morally impermissible mode of conduct can in some imaginable circumstances be morally imperative. Understood in the first way just recounted, P’s assertion about torture is a strong justification. Understood in the second way, it is a weak justification. In other

8

Introduction I: Moral Conflicts and Deontology

words, if a mode of conduct is strongly justified, it is both morally obligatory and morally permissible—strongly permissible—in the circumstances in which it occurs. If a mode of conduct is only weakly justified, it is morally obligatory and morally optimal but not morally permissible (that is, not strongly permissible). Whereas both any strongly justified course of conduct and any weakly justified course of conduct are morally optimal,6 only the former is morally permissible. A weakly justified course of conduct is morally wrong, even though its wrongness is not graver than that of any opposing course of conduct in the circumstances.7

1.1.1.4.  Two senses of ‘rightness’ As a moral term, the adjective ‘right’ can be either an antonym of ‘wrong’ or a synonym of ‘correct’. Although those two senses of the adjective are not mutually exclusive, they are not equivalent. If a course of conduct is right in the first sense, it is strongly permissible; engaging in such a course of conduct will not breach any moral duties and will therefore not amount to a wrong at all. If some course of conduct CC is right in the second sense, then it is morally optimal and is thus at least weakly permissible, but it might not be strongly permissible. Suppose that a person is under a duty to engage in CC and simultaneously under a duty not to engage in it, and suppose that CC is the uniquely optimal course of conduct for the person in the circumstances. Albeit the former obligation is more stringent than the latter, the latter duty continues to exist as such. Accordingly, although CC is right in the sense of being the morally correct thing to do, it is not right in the sense of being strongly permissible. Because the person’s adoption of such a course of conduct involves the contravention of his moral duty to refrain from adopting it, CC is morally wrong. It is uniquely optimal—the non-adoption of CC would be an even graver wrong—but it is only weakly permissible.

1.1.1.5.  Two senses of ‘absolute’ In debates over the moral bearings of torture, one of the frequent points of controversy is whether the moral prohibition on torture is absolute. That much-broached question is in need of disambiguation, however. In posing that question, one might be asking whether the duty not to engage in torture is always and everywhere 6   Though the term ‘optimal’ (or the term ‘best’) is evaluative, it is not to be understood along consequentialist lines here or elsewhere in this book. Optimality is instead a function of deontic stringency. If someone’s moral duty-to-φ conflictingly coexists with her moral duty-not-to-φ, and if the stringency of the former duty is greater than that of the latter, her fulfilment of the duty-to-φ is the lone morally optimal course of conduct for her in the circumstances. If the two duties are instead equally stringent, then her fulfilment of either of them is morally optimal in the circumstances. If her duty-to-φ is not countervailed by any conflicting moral requirements, compliance with that duty is both morally optimal and strongly permissible. 7   Note that the division between strong justifications and weak justifications does not encompass all courses of conduct. It obviously excludes morally unjustified courses of conduct, but it also omits all non-obligatory courses of conduct (many of which, such as my scratching of my nose in virtually any ordinary circumstances, are not morally unjustified).

Moral Conflicts

9

binding in all possible worlds, or one might be asking whether the stringency of that ever-binding duty always and everywhere overtops the stringency of any duties that might conflict with it. Although an affirmative answer to the latter version of the question entails an affirmative answer to the former version, there is no entailment between affirmative answers in the other direction. Strong absoluteness entails weak absoluteness, but not vice versa. If the moral prohibition on torture is strongly absolute, it is not only binding at all times in all places in all possible worlds; in addition, it is also always of greater normative importance than any possible countervailing moral requirements. One’s insistence on the absoluteness of that prohibition in this strong sense is perfectly consistent with one’s recognition that there can indeed be countervailing moral requirements that militate against compliance with the prohibition. In other words, a strongly absolute moral obligation can be locked in conflicts with competing moral obligations that are inferior in their normative strength.8 Although those latter obligations are indeed less stringent, they impose genuine moral demands that will give rise to remedial duties if they are left unfulfilled. In such circumstances, then, compliance with the strongly absolute moral prohibition is not strongly permissible; it is only weakly permissible. Any weakly absolute moral prohibition WM is binding everywhere and always in all possible worlds. No circumstances, however exigent, can ever negate or diminish the demands of such a prohibition. Nonetheless, although WM is irrepressibly operative in all possible situations, it can be locked in conflicts with competing moral requirements (just as strongly absolute moral prohibitions can be). Moreover, in some imaginable circumstances, the competing moral requirements are more stringent than WM with which they conflict. In such circumstances, compliance with WM would not be even weakly justified. There is not available any strongly justified course of conduct in such a setting, and the sole weakly justified course of conduct resides in breaching WM. Any such breach is indeed a breach—that is, a moral wrong—and it will thus trigger remedial obligations. Of course, if WM is formidably stringent, there might never materialize any actual situations in which the contravention of WM would be morally optimal. Nevertheless, so long as there could arise some context in which such a contravention would be optimal, WM is only weakly absolute rather than strongly absolute. Even if conformity to WM is always morally optimal in the actual world, there are never any guarantees of that optimality—for there remain counterfactual scenarios in which some actions at odds with WM are the morally best modes of conduct (the least dreadful modes of conduct).

8   This point is altogether missed in Feinberg 1973, 86: ‘For a human right to have this [absolute] character it would have to be such that no conflicts with other human rights, either of the same or another type, would be possible.’

10

Introduction I: Moral Conflicts and Deontology

1.1.1.6.  Overtopping versus overriding In many philosophical writings on the use of torture in extreme situations of desperation, deontological duties are said to be subject to a threshold past which they are overridden by consequentialist duties with which they conflict.9 The notion of overriding, invoked by such a contention, is in need of disambiguation—or, perhaps more accurately, it is in need of differentiation from the notion of overtopping. When philosophers maintain that some duty D1 has been overridden by a conflicting duty D2, they usually mean that D1 has ceased to be operative and that the fulfilment of D2 is therefore strongly permissible. In the eyes of most of those philosophers, overriding involves cancellation or supersession or suspension. Now, as should be evident from what has been said heretofore in this chapter, the notion of overriding—understood in the way just indicated—is unsuitable for any satisfactory account of the nature of moral conflicts. This chapter has already used ‘overtop’ and its cognates many times, and the notion of overtopping is indeed what is pertinent here in contrast with the notion of overriding. An overtopped moral obligation is not cancelled or superseded or suspended by the overtoppingly stringent moral obligation(s) with which it conflicts; rather, it retains its full force as such. It is exceeded in importance or stringency by the obligation(s) with which it clashes, but it continues to impose its requirements. Compliance with the overtoppingly stringent obligation(s) in such circumstances is weakly justified but not strongly justified. (As will be seen presently, this point calls for a reconception of the threshold envisaged by quite a few deontological philosophers.) Just as the language of ‘overriding’ is generally best eschewed in discussions of moral conflicts, so too is the language of ‘defeasibility.’ When philosophers characterize a moral duty as defeasible, they usually mean that it is susceptible to being overridden in the sense recounted above. In the eyes of most of those philosophers, a moral duty D1 that has been defeated by another moral duty D2 is not merely surpassed in stringency by D2 but is also extinguished or suspended outright by it. Given that such an understanding of clashes between moral duties does not accurately capture the persistence of each duty within any clash, we are best advised to put aside the notion of defeasibility when seeking to come to grips with these matters. That notion, like the notion of overriding, tends to foster the impression that a moral duty is deprived of its requirement-imposing force when it conflicts with a more stringent moral obligation. Because any such impression is misguided, characterizations of duties as defeated or overridden will generally be absent from this book when I am presenting my own positions rather than recounting someone else’s views. 9   For some examples of this threshold-deontological approach, see Fried 1978, 10; Gross 2004a, 1511–19; Gross 2004b, 230–1, 250 n4; Kadish 1989, 346; Levinson 2003, 2031–4; May 2007, 232; Moore 1997, chap. 17; Moore 2007; Nagel 1979, 56; Shue 1978, 141–3. The approach is outlined, but neither endorsed nor rejected, in Kamm 2011, 49–50; Nozick 1974, 30 n*. For some wary discussions of threshold deontology, see Coady 2011, § 6; Davis 2005, 170–3; Ginbar 2008, 24–9; Haque 2007, 629, 639–43; Harel and Sharon 2011, 848–51; Kutz 2007a, 254–6; Meisels 2008b, 174–82; Posner and Vermeule 2006, 677–80; Posner and Vermeule 2007, 40–1, 187–90; Waldron 2010a, 216–17; Ward, Johnstone, and Clucas 2009, 4; Wonnell 2011. For an often powerful critique of threshold deontology, see Alexander 2000.

Moral Conflicts

11

Of course, the preceding paragraph is hardly suggesting that moral duties are never defeasible. Frequently, the existence of a concrete moral duty in any given set of circumstances is permanently or temporarily negated by the presence of some defeating factors. When a duty is so negated, it has come to be only prima facie in the epistemic sense of that phrase. Because of the presence of the defeating considerations, the putative duty does not exist as a genuine moral obligation at all. Recall, for instance, my earlier scenario of the self-defensive thrashing administered by Frank to Edward; in that scenario, the occurrence of Edward’s aggressive attack has suspended Frank’s normal duty to refrain from the use of violence (though Frank of course will be morally obligated not to use more force than is reasonably necessary to repel Edward’s onslaught). Nevertheless, as I have already emphasized, the sheer fact of being locked in a conflict with a more stringent moral duty never deprives any moral obligation of its status as such. Hence, when this book trains its attention on moral conflicts involving the moral prohibition on torture, it will not be broaching any ostensible moral duties that turn out to be mere appearances in particular circumstances. Because a number of philosophers and legal theorists have presumed otherwise, this book will invoke the notions of overriding and defeasibility when it summarizes their stances. However, when I elaborate my own contrasting approach to the problem of torture, those notions will disappear—for they generally obfuscate, rather than illuminate, that problem.

1.1.2.  Chariness of moral conflicts Some philosophers have been decidedly reluctant to acknowledge the occurrence of moral conflicts. The factors underlying their reluctance are varied and are in need of critical scrutiny here.

1.1.2.1.  Consequentialist balancing This chapter will shortly explore the division between deontological obligations and consequentialist obligations, in the course of delineating the general structure of morality. What should be noted at present is that a disinclination to recognize the occurrence of moral conflicts is prominent on each side of that division. However, whereas on the deontological side a disinclination of that type is confined chiefly to Kantians, it is much more widespread among consequentialists. Indeed, the very matter of conflicting moral duties has been a salient point of contention between consequentialists and deontologists during the past several decades, not least with regard to the problem of torture. While most consequentialists of sundry stripes have been intent on gainsaying the reality of moral conflicts, many deontologists outside the confines of the Kantian tradition have affirmed or presupposed the reality and importance of such conflicts.10 Deontologists are quite right to do so, 10   For some consequentialists’ denials of the reality of moral conflicts, see Bobbitt 2008, 363–5; Curzer 2006, 45; Gross 2004a, 1498; Hare 1972; Hare 1981, chaps 2–3; Himma 2007, 240–1;

12

Introduction I: Moral Conflicts and Deontology

for—as we shall see—the persistence of deontological duties in the face of consequentialist pressures is a major source of moral conflicts. Though some versions of consequentialism can take account of the occurrence of moral conflicts, the general consequentialist prioritization of the good over the right is largely antithetical to any clear recognition of such conflicts. Every thoroughly consequentialist doctrine takes some overarching desideratum or set of desiderata as an objective that is a touchstone for the rightness or wrongness of any mode of conduct (in accordance with the conduct’s tendency to promote or impede the realization of the objective). Under such a doctrine, one’s sole fundamental moral obligation is to contribute maximally to the realization of the commended objective. Hence, the supporters of such a doctrine are disposed by it to perceive any non-overtoppingly stringent moral duty as merely prima facie in the epistemic sense and thus as not genuinely a moral duty at all. That is, they typically believe that the factors which constitute a non-overtoppingly stringent moral duty are such as to tend toward the existence of a moral duty without actually giving rise to one. Actions in accordance with those factors, at the expense of factors that are more strongly promotive of some consequentialist objective, would produce the net effect of detracting from the realization of that objective. Accordingly, consequentialists are disposed to maintain that such actions are morally non-obligatory as well as morally impermissible. In the eyes of consequentialist theorists, the lone source of moral obligatoriness is the conduciveness of this or that mode of conduct to the maximal attainment of the desideratum or set of desiderata which the theorists favour. If the net effect of some mode of conduct MC would be to detract from the attainment of the specified desideratum or set of desiderata, then MC does not partake of the aforementioned source of moral obligatoriness. By the reckoning of consequentialists, then, MC is neither morally obligatory nor morally permissible. By their reckoning, nobody is under any moral duty to perform MC, and thus nobody faces any moral conflict that involves such a duty. In short, consequentialism, with its emphasis on the balancing of considerations that respectively tend toward the existence of moral duties, is ill-equipped to deal adequately with situations of moral conflict. Of course, to say as much is hardly to reject consequentialism altogether. In regard to manifold sets of circumstances, a consequentialist emphasis on balancing is entirely appropriate; countless sets of circumstances do not pose any moral conflicts. Nevertheless, consequentialism generally obscures the occurrence of moral conflicts in the numerous situations where they do arise. Partly because of the unsatisfactoriness of consequentialism in that respect, this book’s approach to the problem of torture is strongly deontological.

Paskins 1976, 143; Posner and Vermeule 2006, 676–7; Posner and Vermeule 2007, 187. For a discussion of utilitarians’ denials of the reality of moral conflicts, see Sinnott-Armstrong 1988, 74–81. For some deontologists’ affirmations of the importance of moral conflicts, see Ignatieff 2004, passim; Lukes 2005; Nagel 1979, 73–4; Walzer 1973; Williams 1965.

Moral Conflicts

13

Unlike consequentialists, then, this book can properly address an extreme situation in which the use of torture as an interrogational technique by public-safety officials is the only means likely to prove effective in averting some cataclysmically dire crime. Consequentialists such as utilitarians would contend that the officials’ resorting to the use of torture is morally permissible—strongly permissible—as well as morally obligatory (Pettit 1991, 234). Very different is the verdict delivered by this book’s deontological ruminations. Although the public-safety officials are morally obligated to secure the lives of citizens against calamities that are preventable and reasonably foreseeable, and although the officials’ obligations are extremely weighty when the dangers are immense, those obligations do not cancel or suspend any moral duties with which they conflict. In particular, they never cancel or suspend the officials’ moral obligation to refrain from the use of torture. On a prodigiously rare occasion when a moral obligation of the latter type is overtopped in ethical importance by some countervailing moral obligations of officials to protect citizens against extraordinary perils, the officials’ duty to refrain from the use of torture persists as such. Albeit the morally optimal course of conduct for the officials in the terrible circumstances is to act athwart that duty, they will indeed be acting athwart it rather than acting permissibly. Sometimes a morally vital course of conduct is morally impermissible. Though anybody faced with such a moral clash will have acted correctly if he fulfils the weightier of the two conflicting moral obligations, he will have breached the less important moral duty and will thus have incurred a further moral duty to remedy the situation in some way. Were he acting permissibly by fulfilling the weightier duty, he would not incur any remedial obligation; yet, precisely because he will not in fact be acting permissibly (even though he is acting optimally), such an obligation will indeed be incurred. Admittedly, consequentialist theorists such as utilitarians can allow that some measures which resemble genuine remedies should be undertaken by legalgovernmental officials after torture has been employed in a utility-maximizing fashion. If the use of such torture has somehow become known, and if numerous people in the officials’ society do not themselves adhere to utilitarian ways of thinking, then the application of a remedy-resembling measure will most likely tend toward the maximization of utility through its alleviation of people’s distress. Such a measure might be adopted with full sincerity by the officials who implement it, but utilitarian theorists have to perceive it as a simulacrum of a genuine remedy. A  veritable remedy constitutes an acknowledgement that something wrong has been done. That is, it constitutes an acknowledgement that some moral duty has been breached.11 Utilitarian theorists are obliged by their doctrine to deny that any wrong whatsoever has been committed when legal-governmental officials resort to torture in desperate circumstances where such a course of action is foreseeably 11   Thus, for example, a remedial obligation is different from an obligation to make a payment for something that has been legitimately acquired (such as a meal consumed by a customer in a restaurant or a plot of land acquired by a government through its power of eminent domain). A payment for something legitimately acquired does not remedy a wrong that has occurred; rather, it averts the occurrence of a wrong. See Kramer 2004, 286–7.

14

Introduction I: Moral Conflicts and Deontology

utility-maximizing. Those theorists are therefore obliged by their doctrine to deny that the remedy-resembling step broached above is genuinely remedial. Given that there is no wrong to be rectified, the aforementioned step does not play any rectificatory role. That remedy-resembling measure will be applauded by utilitarian theorists if it tends toward the maximization of utility, but they will not view it as a way of correcting a wrong. In their eyes, its only role is consolatory. It mollifies people who erroneously think that they or others have been treated impermissibly. A utilitarian theorist will maintain that, if legal-governmental officials are morally obligated to take any remedy-resembling step in the circumstances envisaged here, their obligation is wholly derivative of their general moral duty to make people feel happy (in this case, by catering to people’s misconceptions).12 Such a theorist will contend that the obligation of the officials has nothing to do with any need for the righting of a wrong, since no wrong has occurred. Utilitarianism’s gainsaying of the reality of moral conflicts is to be rejected, for it entails the conclusion that absolutely any policies—the execution of innocent people, the most brutal varieties of torture, genocide—are morally permissible so long as the utility-promotive considerations that support the policies are strong enough. We should recognize instead that someone faced with a moral conflict will have to act impermissibly regardless of what he does. If legal-governmental officials are ever confronted with a predicament in which they elect to breach their moral duty-not-to-avail-themselves-of-torture in order to comply with some surpassingly important moral duties owed to the general public, they will be justified only in the sense that they will be pursuing the morally best course of action. They will not be justified in the sense of behaving permissibly. On the contrary, because any use of interrogational torture is abidingly impermissible, the officials will trigger stringent remedial obligations if they ever have recourse to such a tactic. Those obligations will be genuinely remedial, rather than a sop to people’s delusions.

1.1.2.2.  The objectivity of morality Bernard Williams helped to bring the topic of moral conflicts to prominence in Anglo-American philosophy through his publication of a couple of incisive and influential essays in the mid-1960s (Williams 1965; 1966). Nonetheless, commendable though his contribution was, he muddied the waters in one significant respect. He suggested that the actuality of moral conflicts is somehow problematic for moral realism; in other words, it is somehow problematic for the thesis that morality is ontologically and epistemically and semantically objective.13 Were Williams correct 12   If the remedy-resembling step is a simulacrum of compensation, then the utilitarians can favour making it mandatory on the ground that requiring such an expenditure will internalize the costs of governmental operations and will thereby help to ensure that those operations in the future do not go ahead unless their expected drawbacks are significantly exceeded by their expected benefits. Many theorists in the law-and-economics movement ascribe a non-remedial role of this type to awards of damages in tort law. 13   I have elsewhere defended at length a minimalist non-naturalistic version of moral realism. See Kramer 2009a.

Moral Conflicts

15

on that point, there would obviously be a strong reason for proponents of moral realism (including me) to deny that moral conflicts do in fact occur. There is no need here for a lengthy engagement with Williams’s arguments about moral realism, since—unlike his far more perceptive exploration of the general structure of moral conflicts—those arguments have been discredited for quite some time. 14 Moreover, as will be seen, his main way of drawing an anti-realist inference from his reflections on moral conflicts is a version of an unsound line of reasoning that will be probed in § 1.1.2.3. Thus, a terse rebuttal of his anti-realist pronouncements will be sufficient here. Williams’s central point in his challenge to moral realism is articulated in the following passage from his 1965 essay: It seems to me a fundamental criticism of many ethical theories that their accounts of moral conflict and its resolution do not do justice to [the persistence of conflicting moral duties]: basically because they eliminate from the scene the ‘ought’ that is not acted upon. A structure appropriate to conflicts of belief is projected on to the moral case; one by which the conflict is basically adventitious, and a resolution of it disembarrasses one of a mistaken view which for a while confused the situation. Such an approach must be inherent in purely cognitive accounts of the matter; since it is just a question of which of two conflicting ‘ought’ statements is true, and they cannot both be true, to decide correctly for one of them must be to rid of error with respect to the other. (Williams 1965, 113)

On the one hand, Williams was undoubtedly correct when he wrote that a number of ethical theories have failed to do justice to the persistence of moral duties in moral conflicts. On the other hand, he went badly astray in submitting that moral realists are at any disadvantage in coming to grips with the nature of moral conflicts. His error becomes apparent in the final sentence of this quotation, which asserts that two conflicting ‘ought’ statements cannot both be true. In fact, both the statement ‘Some person P at some time t is under a moral duty-to-φ ’ and the statement ‘Some person P at some time t is under a moral duty-not-to-φ ’ can be true. In the final sentence of this quoted passage, Williams was presupposing that ‘P at time t is under a moral duty-not-to-φ ’ is equivalent to ‘It is not the case that P at time t is under a moral duty-to-φ.’ His presupposition is untenable, however, as it conflates internal and external negation. Although the proposition ‘It is not the case that P at time t is under a moral duty-to-φ ’ and the proposition ‘P at time t is under a moral duty-to-φ ’ cannot both be true (and cannot both be false), the proposition ‘P at time t is under a moral duty-to-φ ’ and the proposition ‘P at time t is under a moral duty-not-to-φ ’ can perfectly well both be true (and can both be false). What is so peculiar about the misstep by Williams is that elsewhere in his 1965 essay he himself highlighted the distinction which he elided in the passage above. As Philippa Foot aptly commented:  ‘The strange thing about what Williams

14   Of key importance in discrediting those anti-realist arguments was Foot 1983. For another excellent rebuttal of those arguments, see Brink 1994, 242–6.

16

Introduction I: Moral Conflicts and Deontology

wrote in “Ethical Consistency” is that a great deal of it seems designed to show exactly . . . that moral conflict does not imply “contradiction.” It is as if he himself showed the cognitivist how to avoid the very error he thinks the cognitivist must make’ (Foot 1983, 391).

1.1.2.3.  Logical incoherence Some of the philosophers who deny the possibility of moral conflicts are worried by the thought that such conflicts would partake of logical incoherence.15 We have just seen that a thought along those lines is central to Williams’s indictment of moral realism. Williams usually knew better, but some other philosophers have been much more firmly resistant to the notion that moral conflicts are logically possible. For example, Richard Hare, who was staunchly resistant to that notion throughout his career, sometimes conflated internal and external negation in much the same manner as Williams. In an essay published in 1989, he wrote as follows: ‘If I  say “I ought, but there is someone else in exactly the same circumstances, doing it to someone who is just like the person I should be doing it to, but he ought not to do it,” then logical eyebrows will be raised; it is logically inconsistent to say, of two exactly similar people in exactly similar situations, that the first ought to do something and the second ought not’ (Hare 1989, 179, emphasis in original). Seeking to expose a logical inconsistency, Hare placed the negation in the wrong position. Whereas ‘I ought to φ ’ logically contradicts ‘It is not the case that I ought to φ,’ it is logically consistent with ‘I ought not to φ.’ In other words, ‘not ought’ rather than ‘ought not’ is the contradictory of ‘ought’. Given that the structure of a moral conflict comprises ‘I ought to φ ’ and ‘I ought not to φ,’ and given that no role is played in any moral conflict by ‘It is not the case that I ought to φ,’ there are no logical inconsistencies in such a conflict. As a logical matter, the occurrence of moral conflicts is entirely unproblematic. Quite a few other philosophers have committed mistakes similar to that of Hare (Kramer 2009a, 335 n17). One of the central principles of standard deontic logic, the so-called Permissibility Theorem, tends to cloud reflection on these matters.16 The formal rendering of that theorem is as follows:

(∀x )(Ox → Px ) 15   This worry surfaces fleetingly in Coady 2011, § § 3–4; Nagel 1979, 59–60, 74 n12; Wisnewski 2010, 62–3. It is much more prominent in Curzer 2006. It is discussed and rejected in Meisels 2008a; Meisels 2008b, chap. 7. 16   Some sophisticated philosophers have endorsed the Permissibility Theorem. See, for example, Conee 1982; Feinberg 1980, 235, 237; Hill 1996, 177; Hughes and Cresswell 1996, 43; McConnell 2010, § 4; Simester 2008, 292 n10; Steiner 1998, 233, 268 n55; Vallentyne 1987, 119–20; Vallentyne 1989. For a powerful critique, see Sinnott-Armstrong 1988, 156–61. See also Sinnott-Armstrong 1996, 52. Though David Brink (1994, 235–6) appears to endorse the contrapositive of the Permissibility Theorem, he in fact ultimately endorses only the version of that theorem which I myself uphold.

Moral Conflicts

17

What this formulation states in effect is that the obligatoriness of any mode of conduct entails the permissibility thereof. By contraposition, then, the impermissibility of any mode of conduct entails the non-obligatoriness thereof. Now, although there is a way of construing the Permissibility Theorem which renders it necessarily true, that one truth-conferring interpretation of the theorem is fully consistent with the possibility and actuality of moral conflicts. Each of the three other eligible interpretations of the theorem falsifies it. On the whole, the Permissibility Theorem has impeded rigorous thinking about the structure of deontic relations. We are told by the Permissibility Theorem that everything obligatory is permissible. If we construe ‘obligatory’ as ‘overtoppingly obligatory’, and if we construe ‘permissible’ as ‘at least weakly permissible’, then the Permissibility Theorem is true. It is indeed the case that every overtoppingly obligatory course of conduct is at least weakly permissible; if some such course of conduct were not at least weakly permissible, then a person would be under an overtoppingly stringent moral duty-to-do-q and an overtoppingly stringent moral duty-not-to-do-q. Given that an overtoppingly stringent moral duty exceeds in importance any moral duties that conflict with it, each of the two duties just mentioned would be morally more important than the other. Such a state of affairs is impossible, since any coherent relation of superiority (such as ‘more important than’) is strictly asymmetrical. Ergo, every overtoppingly obligatory course of conduct is indeed at least weakly permissible. As has been remarked, this one truth-conferring interpretation of the Permissibility Theorem renders it straightforwardly compatible with the existence of moral conflicts. In any moral conflict, where someone is under a moral duty-to-φ and simultaneously under a moral duty-not-to-φ, no more than one of those obligations is overtoppingly stringent. Accordingly, every such conflict is consistent with the Permissibility Theorem as it has just been glossed. Everyone who recognizes the coherent possibility of moral conflicts should also recognize that the Permissibility Theorem as it has just been glossed is true. Construed in any other way, however, the Permissibility Theorem is false. It is not the case, for example, that every overtoppingly obligatory course of action is strongly permissible. One’s moral duty to aid a badly injured pedestrian might well be overtoppingly stringent in a given context, but the fulfilment of that overtoppingly stringent duty might preclude the fulfilment of one’s less important moral duty to keep one’s promise to meet somebody else for lunch. In that event, one’s going to the aid of the pedestrian is only weakly permissible rather than strongly permissible. Innumerable similar examples could be adduced. Even more plainly, it is not the case that every non-overtoppingly obligatory course of action is strongly permissible. Indeed, no such course of action is strongly permissible. Furthermore, given that some non-overtoppingly stringent moral duties conflict with overtoppingly stringent moral duties, it is not even the case that every non-overtoppingly obligatory course of action is weakly permissible. Many such courses of action are impermissible in every sense. Hence, only under one interpretation is the Permissibility Theorem correct; under the three other available interpretations, it is false.

18

Introduction I: Moral Conflicts and Deontology

Of central importance here, again, is that the lone truth-conferring interpretation of the Permissibility Theorem is consistent in every respect with my emphasis on the possibility and actuality of moral conflicts. That rendering of the theorem is a wholly unexceptionable encapsulation of the logic of the ‘greater than’ relationship. Since every coherent moral conflict naturally conforms to that logic, the Permissibility Theorem poses no obstacles whatsoever to the occurrence of moral conflicts. However, because that theorem is so susceptible to being construed in any of the three ways that render it false, it tends to be highly misleading. It lends itself too readily to egregious misapprehensions. We are therefore best advised to leave it aside, even while recognizing that it is completely unproblematic when it is understood correctly.

1.1.2.4.  Moral conflicts and action-guidance Moral theories and principles are supposed to provide answers to questions about appropriate courses of conduct in multitudinous sets of circumstances. Thus, one complaint sometimes voiced about the spectre of moral conflicts is that we would have no grounds for deciding what to do when confronted with clashing obligations (Dworkin 2011, 90; McConnell 2010, § § 4 and 7). In other words, moral conflicts are thought to leave us bereft of action-guidance. Any anxiety along those lines is largely baseless. In most moral conflicts, the clashing duties are of unequal stringency; hence, the uniquely correct response to such a conflict is to fulfil the more stringent duty. Of course, as should be evident from my earlier discussion of the adjective ‘right’, the term ‘correct’ here does not denote permissibility. Within a moral conflict, no morally permissible course of conduct is available. Nonetheless, although the uniquely correct mode of conduct in such circumstances is itself morally wrong—and although the adoption of that course of conduct will therefore give rise to remedial duties—it is indeed uniquely correct in that the non-adoption of it would be an even more serious wrong. Accordingly, there is no lack of action-guidance in any such moral conflict. Morality determinately prescribes the mode of conduct that is to be undertaken as the best way of dealing with the quandary which such a moral conflict presents. Furthermore, even in a rare moral conflict where the clashing duties are evenly balanced in their stringency or are incommensurably counterpoised, the conclusion follows that the fulfilment of either duty will amount to a breach of the other and will thus impose remedial obligations on anyone who has committed the breach.17 Hence, although neither the fulfilment of the duty-to-φ nor the fulfilment of the duty-not-to-φ in such a conflict is a morally better course of conduct than the other, we are not wholly devoid of guidance in a situation of this sort. Anyone aware of the nature of the situation as a moral conflict has grounds for knowing that, regardless of whether he complies with the former duty or with the latter, he will have incurred a further moral obligation to remedy the wrong that he has thereby done.   This point is rightly emphasized in Nussbaum 2000, 1009.

17

Moral Conflicts

19

1.1.2.5.  Concerns about fairness or excessive onerousness In short, the problem of action-guidance is no more powerful as a basis for some philosophers’ chariness of moral conflicts than are any of the other considerations that we have pondered so far. A  rather different concern appears to underlie a remark by Michael Moore in which he explains why he hopes to show that moral conflicts are very uncommon. On the one hand, Moore does briefly acknowledge the potential for conflicts between deontological duties, and he allows that ‘[i]‌t may not be, as Kant famously proclaimed, that a conflict of such obligations is literally “inconceivable”’ (2007, 37). On the other hand, he declares that ‘it would be unfortunate for us in the extreme if morality often confronted us with choices where we will be “damned if we do and damned if we don’t.” The distinctions we shall examine hold out the possibility of so limiting our stringent obligations as to minimize or even eliminate such situations of moral conflict’ (2007, 37–8). Indeed, the very scenarios with which he grants the possibility of conflicts between deontological obligations—scenarios similar to some of those in Kamm 2007, 27–9, 252, 268–9—are so far-fetched as to make manifest his view that any genuine conflicts in morality are extremely rare.18 Moore’s ambition to establish the infrequency of moral conflicts is focused not on the problem of action-guidance but instead on the onerousness or unfairness of a world in which someone often finds that every mode of conduct open to her is morally wrong. Such a worry is pertinent, of course, but Moore draws an inapt conclusion from it. Instead of trying to expound the general structure of morality in a manner that whisks most moral conflicts out of sight, we should quite frequently seek to act in ways—and to arrange our institutions in ways—that will reduce the incidence of such conflicts. Acting in conformity to a practically oriented conclusion of that sort is the best means of allaying the anxiety felt by Moore and others about the prospect of unavoidable wrongness. To act in such a fashion, one needs to be alert to the possibility and actuality of moral conflicts in a diversity of settings. In other words, the consternation engendered by the spectre of unavoidable wrongness should prompt us to do the opposite of what Moore recommends. Far from trying to delineate the contours of morality in a manner that will obscure the emergence of moral conflicts, we should be seeking to grasp those contours with keen sensitivity to the likelihood of such conflicts. Only thus can we informedly fix upon the practical steps that are best suited to avert predicaments of unavoidable wrongness (insofar as they can and should be averted).

18   For instance, Moore propounds the following scenario (2007, 96): ‘I have begun a boulder rolling down a hill to kill five old enemies of mine, but . . . I repent. Now, however, the only way to stop the boulder is with the corpulent body of one bystander: may I throw him before the boulder, killing him but preventing the deaths of the five? Doing so makes me a killer, but it prevents me from being a killer five times over.’

20

Introduction I: Moral Conflicts and Deontology

1.2.  The Deontology/Consequentialism Distinction and the General Structure of Morality As has been remarked, this book’s approach to torture is embedded in a general account of the nature of morality. Of key importance to that account is my emphasis on moral conflicts, which we have just been considering. Likewise of key importance is the strongly deontological tenor of this book’s outlook. Having already glanced at the distinction between deontology and consequentialism in my ruminations on moral conflicts, we should now examine it more directly and sustainedly. Thereafter, this chapter will conclude by adumbrating the structure of morality. Naturally, both the matter of moral conflicts and the deontology/consequentialism dichotomy will figure saliently in the adumbration.

1.2.1.  The division between deontology and consequentialism My outline of the general structure of morality, along with the rest of this book, will refer prominently to deontological prohibitions and deontological permissions (or prerogatives). Some careful attention here to the deontology/consequentialism distinction will help to elucidate those references. We should mull over two main ways in which that distinction can be drawn.

1.2.1.1.  Intrinsic moral statuses One tack for differentiating between deontology and consequentialism is to concentrate on the moral statuses of modes of conduct. If a course of conduct on the part of any person P is covered by a deontological permission, then P’s engaging in that conduct is not wrong in any respect—regardless of the consequences that it causes or is likely to cause. Within the ambit of the deontological permission, P’s undertaking of the specified conduct does not breach any moral duties. Hence, within that ambit, his behaviour is strongly permissible irrespective of the probable consequences of its occurrence. Suppose, for example, that P is deontologically at liberty to expend a certain portion of his income on his hobby of collecting stamps. Suppose further that any number of consequentialist objectives (such as the maximization of human happiness or the promotion of equal economic opportunities) would be advanced if P were instead to expend that portion of his income on donations to charities. Notwithstanding that his pursuit of his philatelic hobby will bring about a worse state of affairs than he is capable of bringing about through alternative uses of his resources, P is not committing any wrongs so long as he is acting within the scope of the deontological prerogative which encompasses that pursuit. Just as the permissibility of P’s expenditure of funds on his hobby is unaffected (within broad limits) by the consequences of his using his income in that fashion, so too the wrongness of certain modes of conduct is not removed by the benignity of the consequences that are likely to flow from the occurrence of any such modes

The Deontology/Consequentialism Distinction

21

of conduct. If P is under a deontological duty to φ, then his not φ-ing is wrong regardless of how valuable the resultant state of affairs might be. Every deontological duty is at least weakly absolute (in the sense introduced earlier). That is, every deontological duty forbids some course of conduct even if one’s undertaking of that course of conduct in particular circumstances would yield much better consequences than one’s refraining therefrom. A  deontologically prohibited type of conduct is wrong always and everywhere. Whereas deontologists ascribe consequence-independent moral bearings to various actions and omissions, consequentialists of course maintain that the moral character of any type or instance of conduct is fully determined by the probable consequences thereof. In the eyes of the latter theorists, no type or instance of conduct is ever endowed with any inherent moral status. Instead, every action or omission derives its moral status from the effects with which it is associated. Accordingly, as John Finnis has aptly remarked, a proponent of any consequentialist doctrine ‘holds himself ready to do anything’ (Finnis 1980, 121, emphasis in original). For the consequentialist, there are no principled limits on the range of actions that can legitimately be undertaken in sundry circumstances. The breadth or narrowness of that range will depend entirely on the results that are likely to follow from each of the multitudinous modes of conduct that might be adopted by people in any number of contexts. No mode of conduct, however abhorrent it may be, is unconditionally disallowed by a consequentialist theorist.

1.2.1.2.  Agent-neutrality versus agent-centredness One way of differentiating between deontology and consequentialism, then, addresses the question whether the moral statuses of some patterns of behaviour are intrinsic or extrinsic. Another way—which has become especially prominent during the past few decades and which figures conspicuously in some of the philosophical literature on torture—pertains to the manner in which our fundamental moral obligations present their demands. Whereas deontologists contend that those demands are agent-centred,19 consequentialists maintain that they are agent-neutral. In other words, deontological duties are such that they present each agent with reasons specifically for that agent to conduct himself in certain ways. The reasons do not pertain to goals by reference to which every agent is enjoined to produce maximally valuable or minimally disvaluable states of affairs. For example, a deontological duty to refrain from the crime of murder is not a duty to pursue an overarching goal such as the minimization of the number of murders in a society. Rather, it is a duty that requires each person to refrain from committing any murders even when his perpetration of such crimes would reduce their overall incidence or would otherwise be highly valuable. One’s fulfilment of a deontological obligation 19   See, for example, Nagel 1979, 132–4; Scheffler 1985; Darwall 1986. The epithet ‘agent-relative’ is also common in the relevant literature—owing largely to Parfit 1984, 143 and Nagel 1986, 152— but I eschew it in order to make fully clear that a deontological conception of moral duties has nothing to do with moral relativism.

22

Introduction I: Moral Conflicts and Deontology

consists not in one’s venturing alongside other people to maximize the realization of some objective, but instead in one’s compliance with a strict prohibition regardless of how other people might respond. By contrast, consequentialist obligations require each person to strive alongside other people to maximize the realization of some goal or array of goals (which might be highly pluralistic). A consequentialist obligation is agent-neutral in that it presents everyone with reasons-for-action that do not single him or her out as a particular person distinct from everybody else. In that respect, the reasons flowing from any consequentialist duty are the same for everyone—though of course the detailed specificities of what is required of each person will depend on his or her aptitudes and circumstances. In regard to any desideratum or set of desiderata upheld by such a duty, a maximizing quest is prescribed. From a consequentialist perspective, the attainment of the most valuable state of affairs possible is always better than the attainment of any less valuable state of affairs and is therefore both obligatory and strongly permissible. Everyone alike is morally obligated and morally at liberty to contribute to the achievement of that maximally valuable state of affairs, even if someone’s contribution will involve his acting at odds with the content of a deontological principle. (Of course, although any consequentialist theory maintains that the maximization of some desideratum or set of desiderata is morally obligatory and permissible, not every such theory recommends that individuals adopt maximizing outlooks when arriving at decisions. Many consequentialists instead hold that the maximization of a favoured desideratum or set of desiderata is best pursued indirectly through the focusing of individuals’ choices and deliberations on other concerns. For example, the adoption of a satisficing mentality by individuals in relation to some desideratum might be the best way for them to maximize their attainment of a more complex balance of desiderata.)

1.2.1.3.  Absolute prohibitions Such, then, are the two ways of distinguishing between deontology and consequentialism. Now, as is plain, those two ways are not identical. One of them addresses the question whether the basic moral statuses of certain modes of conduct are intrinsic or extrinsic,20 whereas the other addresses the question whether the reasons presented by fundamental moral obligations are agent-centred or agent-neutral. Nonetheless, although the two contrasts are not exactly the same, they are not really separate; the latter is a corollary of the former. If the basic moral status of a mode of conduct is consequence-independent, then—in the manner captured by the notion of agent-centredness—that basic moral status is impervious to any consequentialist balancing that would deem the tokens of an abhorrent act-type AT to be permissible whenever those tokens can lower the overall incidence of AT. Imperviousness to such balancing is a key aspect of the general consequence-independence of an act-type’s wrongness. 20   The phrase ‘basic moral status’ refers to the permissibility or impermissibility of each of the modes of conduct.

The Deontology/Consequentialism Distinction

23

The distinction between deontology and consequentialism has a crucial bearing on the matter of the absoluteness (the weak absoluteness) of some major prohibitions. On the one hand, as this chapter has already observed, not all deontologists are absolutists. Some threshold deontologists believe that the consequence-independent impermissibility of a mode of conduct such as torture is operative only up to a certain point—namely, a point where the consequences of forgoing that mode of conduct would be truly appalling. On the other hand, many deontologists (including me) are indeed absolutists. Such deontologists affirm that certain major moral prohibitions are binding always and everywhere irrespective of the consequences of compliance therewith, and specifically that the prohibitions are binding even when departures from them will serve to lower greatly the overall incidence of such departures. By contrast, consequentialists deny that any prohibitions are absolute (apart from a general prohibition on declining to contribute to the realization of some overriding desideratum or set of desiderata in a maximally promotive manner). In particular, consequentialists deny that any moral prohibition remains binding when a deviation from it will serve to heighten the overall level of conformity to that prohibition. While the distinction between deontology and consequentialism will loom large throughout this book, we shall return to it most sustainedly in Chapter 4—a chapter that seeks to confirm the absoluteness of the moral prohibition on torture. As will become apparent there, some philosophers have mounted powerful challenges to the agent-centredness of deontological prohibitions. Hence, one part of my task in defending the absoluteness of the moral prohibition on torture will be to reaffirm the rationality of deontological agent-centredness in the face of those challenges. A vindication of deontology is essential for a vindication of the thesis that torture is always and everywhere wrong. These present remarks on the deontology/consequentialism division should close with a terse caveat that can help to forestall confusion hereafter. Chapter 3’s discussion of the wrongness of torture will lay emphasis on the distinction between a victim-focused perspective and a perpetrator-focused perspective for the justification of a moral prohibition. That is, it distinguishes between (1) justifications that concentrate on the interests or inviolability of the potential victims of torture and (2) justifications that concentrate on the moral integrity of the potential perpetrators of torture.21 Although a rationale of the first type is perfectly consistent with a rationale of the second type, neither entails the other; either can be invoked to the exclusion of the other, even though they can also be combined. Chapter 3 will argue that a victim-focused rationale is insufficient on its own to support the thesis that torture is always and everywhere wrong. Notwithstanding that the interests and inviolability of potential victims of torture are decisive justificatory factors in most contexts, a victim-focused approach uncombined with a perpetrator-focused approach will fall short of establishing that the moral bar to the permissibility of 21   This distinction between a victim-focused justification and a perpetrator-focused justification is similar to the distinction which I have drawn elsewhere between right-focused justifications and duty-focused justifications; see Kramer 1998, 35–41.

24

Introduction I: Moral Conflicts and Deontology

torture is absolute. A satisfactory engagement with the problem of torture has to concentrate on the moral integrity of people who might resort to the deliberate infliction of excruciating pain. The main reasons for the inadequacy of a purely victim-focused perspective will come to the fore in Chapter 3. For the moment, we should simply note that the victim-focused/perpetrator-focused dichotomy is quite different from the agent-neutrality/agent-centredness contrast which we have briefly explored above. Whereas the distinction between a victim-focused justification and a perpetratorfocused justification pertains to the substantive basis for a moral prohibition on the use of torture (or the substantive basis for some other moral constraint), the agent-neutral/agent-centred dichotomy bears on the question whether any such prohibition is susceptible or insusceptible to consequentialist trade-offs. To be sure, the moral considerations that account for the forbiddenness of torture might well be the same as the moral considerations that account for the fact that the forbiddenness of torture is not subject to consequentialist exceptions. However, any such convergence of considerations is something that has to be shown through moral argumentation, rather than something that can safely be taken for granted through the conflation of the two issues or sets of issues that have just been disentangled here. Even if the factors that underlie the wrongness of torture are also the factors that make the moral prohibition on torture impervious to any consequentialist balancing, the wrongness and the imperviousness are not exactly the same thing. Questions about those matters have to be addressed separately, even if the answers to the questions exhibit close affinities with each other.22

1.2.2.  The general structure of morality Keeping in view the foregoing exposition of the deontology/consequentialism distinction, we can now turn to the general structure of morality—in which that distinction will figure centrally. To highlight the distinctiveness of this book’s conception of that general structure, the present section of the chapter will juxtapose Moore’s account of morality with my own account. Moore’s presentations of his understanding of morality have been helpfully expansive and sophisticated, and his understanding is in accordance with the views of many other philosophers who embrace a threshold-deontological position. Because his perspective (a quite widely shared perspective) is at variance with this book’s version of threshold deontology in several major respects, his delineation of his moral framework is an excellent foil for the elaboration of my competing map of the moral realm.

22   The differences between the agent-centred/agent-neutral dichotomy and the perpetrator-focused/ victim-focused dichotomy are implicitly recognized in Gewirth 1981, 14; and Hill 1991, 81. They are overtly recognized in Kamm 2007, chap. 8. They seem to be overlooked in Gross 2004a, 1494 n48; and Moore 1997, 705 n89. They are definitely overlooked in Waldron 1989, 505 n5. Some related confusion arises when the agent-centred/agent-neutral distinction is perceived as a variant of the act/ omission distinction. For an example of such confusion, see Seidman 2005, 889–91.

The Deontology/Consequentialism Distinction

25

Moreover, as will become evident later in the book, these divergent outlooks on very abstract issues of morality ramify into divergent assessments of the concrete moral matter of interrogational torture. Dissimilarities between my conclusions about interrogational torture and Moore’s conclusions relate not just to the moral bearings of such torture but also to the legal consequences that should ensue if torturous methods of interrogation are ever employed in extreme situations. Underlying the contrarieties between our proposals concerning those legal consequences is the matter of moral conflicts that has loomed so large already in this chapter. To be sure, the connections between the abstract issues about moral conflicts and the concrete issues about the proper legal treatment of torture are not logically necessary. Those connections are substantively moral,23 and the substantive moral values at stake can be explicated plausibly in a number of directions. Nonetheless, as the trajectory of this book will suggest, the paths between the abstract heights and the concrete thickets of moral controversy are sometimes quite straightforward. Theses propounded on the abstract heights can be enrichingly fleshed out and supported by the concrete moral reasoning for which those theses furnish indispensable guidance.

1.2.2.1.  Moore’s tripartite account Moore maintains that the general structure of morality is tripartite. At the first level is a background of consequentialist reasons for each person to seek to bring about valuable states of affairs. Those reasons are always present in varying degrees of weightiness, though they do not always impose obligations. Whereas some consequentialist theories posit a single fundamental desideratum (such as maximal human happiness) that is the ground of the value of everything else, Moore’s conception of the good is pluralistic. Everyone has reasons for promoting sundry types of valuable states of affairs. As has been remarked in the preceding paragraph, consequentialist reasons-for-action do not always impose any obligations. The absence of obligatoriness is quite often due to the normative effects of the second of the three levels in the structure of morality that Moore envisages. At that second tier are deontological permissions (or prerogatives) and deontological duties. Deontological permissions strongly entitle people to conduct themselves in ways that are suboptimal from any consequentialist viewpoint, while deontological prohibitions forbid people to act in certain ways even when their so acting would be optimal from any consequentialist viewpoint. Thus, the second tier in Moore’s triadic account of morality is an expression of his resistance to the notion that consequentialist trade-offs exhaust the moral domain. Nevertheless, the third tier in his account makes clear that his moral outlook is not steadfastly deontological.24 As a threshold deontologist, Moore maintains that 23   Of course, as I have argued elsewhere, the correct basic principles of morality and many correct derivative principles of morality are themselves necessary. See Kramer 2009a, 157–61. My point here is simply that the necessity is moral rather than logical. 24   Chapter 4 will address another respect in which Moore’s adherence to deontology is somewhat tenuous: namely, his uneasiness about the rationality of the ways in which deontological duties check

26

Introduction I: Moral Conflicts and Deontology

the persistence of deontological duties in response to consequentialist buffeting is not unlimitedly tenacious. Past some unspecifiable threshold as the consequences of adhering to a deontological duty D become more and more dire, D ceases to be binding. In a context where the threshold of catastrophic moral horror has been passed, calamity-averting actions at odds with D—which, for the duration of the extreme emergency, is only an erstwhile obligation—are morally permissible. For Moore, then, the third layer in the realm of morality is made up of situations of grave peril where the urgency and immense weightiness of consequentialist requirements override (and not merely overtop) the sway of deontological duties.

1.2.2.2.  An alternative account of morality My alternative theory of the general structure of morality overlaps with Moore’s, but is much more robustly deontological. My theory, like his, recognizes that each person is confronted with an array of consequentialist reasons-for-action. Those reasons-for-action, which are of many different degrees of strength, can constitute moral obligations of varying degrees of stringency. Insofar as endeavours to further some genuine desiderata do not involve any contraventions of deontological constraints, people beyond the bounds of their deontological prerogatives will generally be morally obligated to participate in those endeavours (in specific ways that will depend on their aptitudes and circumstances). In addition to being confronted with an array of consequentialist reasons-for-action, every person is faced with an array of deontological permissions and prohibitions. Of most importance here are conflictual relationships between those reasons-for-action on the one hand and those permissions and prohibitions on the other. When consequentialist considerations clash with the demands of a deontological prohibition, and when the specified considerations are not very weighty, they do not constitute any moral obligations. Instead, the matter to which they pertain is covered by a deontological prerogative, which is thus coupled with the deontological prohibition. In other words, conformity to the deontological prohibition is strongly permissible even though such conformity is inconsistent with what a consequentialist calculation prescribes. When the consequentialist factors that clash with a deontological prohibition are very weighty, those factors are constitutive of a countervailing moral obligation. In such a situation, accordingly, the prohibition is no longer coupled with a deontological permission. Instead, it is locked in a conflict with a diametrically opposed moral requirement; the deontological duty-not-to-φ is countervailed by a consequentialist duty-to-φ. Still, that latter duty is exceeded by the former in stringency. If any

the maximization of desiderata (Moore 1997, 705; Alexander and Moore 2007, § 2.2.2). Moore is worried about the powerful challenge by Samuel Scheffler (1985) to the rationality of agent-centred restrictions—a challenge to which I have referred briefly in § 1.2.1.3. Chapter 4 will seek to rebut Scheffler’s arguments, and will present two complementary rationales for the agent-centredness of the aforementioned restrictions.

The Deontology/Consequentialism Distinction

27

deontological duty D is strongly absolute, then in every possible world D is always and everywhere more stringent than any competing obligations. If a deontological duty is only weakly absolute, its stringency can be exceeded by that of a countervailing moral obligation in a situation of extreme urgency and desperation. Exactly how dire the situation would have to be for the overtopping of the deontological duty’s normative importance is something that depends on the specifics of the duty and on the extent of any requisite contravention. What should be emphasized here, in line with my earlier remarks on the persistence of any moral duties that are locked in a moral conflict, is that an overtopped deontological obligation retains all its normative force despite being surpassed in stringency by a formidably weighty consequentialist requirement that conflicts with it. Its demands are uncancelled and indeed unimpaired by the situation of moral conflict in which it obtains. Admittedly, the remedial duties arising from a breach of the obligation will be considerably less onerous than the remedial duties that would have arisen if the breach had occurred in the absence of any conflict. The gravity of a contravention is substantially mitigated when the contravention is undertaken for the sake of fulfilling an overtoppingly stringent moral requirement. All the same, the very fact that remedial duties will indeed accrue in the wake of a morally optimal course of action is attributable to the undiminished force of the moral obligation that has been transgressed. The status of that deontological obligation as such is unaltered by the presence of an even more important moral demand that calls—in an extreme context—for the obligation to go unfulfilled.

1.2.2.3. A pithy conclusion: some differences between the accounts of morality Moore’s account and my own account of the general structure of morality are both threshold-deontological in character, but our conceptions of the threshold differ markedly. Whereas Moore takes it to be a threshold of moral permissibility, I  take it to be only a threshold of moral optimality. This key difference is directly connected to his discounting of the significance of moral conflicts and to my highlighting of that significance. To be sure, this major divergence between our approaches is accompanied by an array of similarities between them. For example, although we disagree about the nature of the deontological threshold, we concur that the location of that threshold is indeterminate.25 Nevertheless, the affinities between our understandings of morality are overshadowed by the dissimilarities. As has been indicated, those dissimilarities carry over from the abstract reaches of moral philosophy into Moore’s and my divergent assessments of the moral bearings of torture—and into our even more concrete recommendations concerning the legal consequences that should ensue if torture is ever employed. Precisely

25   Thus, I believe that Moore is not vulnerable to several of the criticisms of him in Alexander 2000. Though the remaining criticisms in Alexander’s trenchant article are telling against Moore, they are readily accommodated by my own deontological conception of morality.

28

Introduction I: Moral Conflicts and Deontology

because Moore supposes that the threshold in his deontological account of morality is a threshold of permissibility as well as of optimality, he believes that no wrong whatsoever is committed when interrogational torture is employed to thwart a cataclysmically destructive act of terrorism or criminality. No moral conflict is involved, or so Moore contends. Because no moral wrong has been done, legal sanctions levied for the use of interrogational torture in the specified circumstances would be inappropriate.26 Hence, in combination with his legal-moralist theory of the role of criminal law, the disinclination of Moore to acknowledge the frequency of moral conflicts has impelled him quite smoothly toward his position on the proper legal status of torture. Likewise, of course, my insistent acknowledgement of the frequency of moral conflicts is a key to my position on the legal consequences that should follow when interrogational torture has been employed. Even in an imaginably dreadful situation of urgency where the use by officials of interrogational torture fulfils public-safety obligations that are more stringent than the officials’ duties to eschew any use of such torture, their actions are seriously wrong. The officials may have acted correctly, but the correct course of conduct in a predicament of moral conflict is a breach of a moral duty. Given the stringency of the torture-eschewing obligation, the withholding of legal sanctions in the aftermath of a breach of that obligation would be morally untenable. 27 In other words, in combination with my conception of the moral role of law in a liberal-democratic system of governance, my emphasis on moral conflicts impels me quite smoothly toward my position on the appropriate legal status of torture. These closing paragraphs laconically anticipate matters that will be discussed at much greater length in the rest of this book. The abstract points of moral philosophy expounded in the present chapter will surface again and again in this book as we turn our attention to the complexities of torture and of torture’s wrongness. While the abstract points inform and orient my subsequent chapters’ discussions, they are also crucially reinforced by those discussions. What this book aims to show is that its strongly deontological outlook—with its attentiveness to moral conflicts—is vital for any effort to do justice philosophically and morally to the fiendish problem of torture.

26   Moore—1997, 733–4—does support a blanket legal prohibition on the use of torture against innocents (such as the young children of terrorists). However, he regards as unfair the imposition of sanctions on any officials who have averted calamities by resorting to such torture, and he acquiesces in the punishment of those officials only because he thinks that a policy of legally approving their actions against innocents would be too likely to lead to many instances of unjustifiable torture. 27   As will become clear in Chapter 5, I am not assuming that the appropriate sanctions will always be criminal, nor am I assuming that they will always be imposed on individuals; in some cases, the sanctions should be imposed solely on the collectivities on whose behalf the individuals have resorted to torture.

2 Introduction II: What is Torture? At first glance, the question that forms the subtitle of this chapter may appear quite straightforward and even elementary. When present-day tourists explore the remnants or reproductions of medieval torture chambers at various sites in Europe, there is no great mystery concerning what they expect to behold. Although the specific techniques and implements of torture are multifarious, and although the workings of many of those implements would not be clear to any typical observer until they are explained, the tourists who flock to see them are morbidly fascinated to learn of the ways in which some human beings have deliberately inflicted excruciating pain on others. That general sense of the nature of torture might initially seem to be sufficient for an enquiry into torture’s wrongness. After all, the fact that torture centrally involves the deliberate infliction of agony is obviously a key to its wrongness. The tourists’ general expectations, though not their unsavoury fascination, might therefore seem sufficient as a basis for the discussions that occupy the rest of this book. Any such impression will be dispelled by this chapter. Albeit the sightseers who visit torture chambers are largely correct as far as they go in their understanding of the nature of torture, they do not go nearly far enough. One main reason for the numerousness of the scholarly and legal efforts to define the crime of torture is that the matter is by no means as clear-cut as an everyday conception of the crime might suggest. A rough sense of the lineaments of that crime is enough for someone engaging in touristic voyeurism, but somebody who engages in a philosophical and moral investigation of the problem of torture should be striving for much greater precision and depth. To pin down why torture is wrong, we need to pin down what torture is. Success in that latter endeavour requires alertness to the subtleties and multifacetedness of the phenomenon. Although this chapter is undertaking a conceptual enquiry, there is no sharp distinction between my orientation here and the orientation of the subsequent chapters. Notwithstanding that the enterprise of conceptual demarcation herein is more obliquely and abstractly normative than are the reflections in the later chapters, it is indeed a normative enterprise. The positions taken in those later chapters are not firmly predetermined by the positions taken here, but the relationship between them is far from wholly fortuitous. In a project of specifying what the practice of torture is, judgements about the importance of that ugly practice inevitably figure. Whereas judgements about the importance of various phenomena in some other domains (such as the domain of jurisprudential theorizing) can be evaluative

30

Introduction II: What is Torture?

without being morally laden—since the values that guide investigations of those phenomena can be austerely theoretical-explanatory—no such detachedness is feasible in one’s grappling with the matter of torture. As this chapter refines and deepens the common-sense understanding of the nature of torture, it presents some classifications that facilitate the moral argumentation in the subsequent chapters. Again, to say as much is hardly to say that the conclusions reached by this chapter inexorably preordain the conclusions reached later; someone can quite coherently agree with my analyses of the nature of torture while disagreeing with my subsequent ruminations on torture’s wrongness. Nevertheless, one’s efforts to delineate the ambit and involutions of the practice of torture are informed by the concerns which lead to one’s recognizing that that practice is an especially problematic mode of conduct. The first main section of this chapter will ponder some previous attempts to define the crime of torture. As will be seen, each of those attempts has been underinclusive and overinclusive or has otherwise been unsatisfactory. We shall then move on to ponder the diversity of the patterns of behaviour that are properly classifiable as torture. As will become apparent in Chapter 3, some of the dissimilarities among those sundry types of behaviour are morally pregnant. The present chapter will conclude by advancing its own definition of torture—a considerably more flexible definition than the ones examined in the chapter’s opening main section. At several junctures, we shall mull over the distinctiveness of torture by considering how it differs from some cognate patterns of conduct.

2.1.  Definitional Ventures Jurists and philosophers alike have proposed quite a few definitions of the crime of torture. An exploration of the strengths and shortcomings associated with several of those broached definitions will be an illuminating point of departure for this chapter’s own clarificatory venture. One lesson to be learned is the unwisdom of rigidity. Although an account of the nature of torture can to some extent be framed pertinently as a set of necessary conditions, most of its elements should be formulated as statements of typical conditions rather than necessary conditions.

2.1.1.  Convention against Torture Perhaps the most famous definition of the crime of torture is that contained in the United Nations Convention against Torture, which came into force in 1987.1 The first section of Article 1 of the Convention specifies the nature of torture as follows: For the purposes of this Convention, the term ‘torture’ means any act by which severe pain or suffering, whether physical or mental, is intentionally inflicted on a person for such 1   The text of the UN Convention against Torture and Other Cruel, Inhuman, or Degrading Treatment or Punishment is readily available in many electronic and printed sources. See, for example, . Some of the key provisions of the Convention are reproduced in Levinson 2004, 40–2.

Definitional Ventures

31

purposes as obtaining from him or a third person information or a confession, punishing him for an act he or a third person has committed or is suspected of having committed, or intimidating or coercing him or a third person, or for any reason based on discrimination of any kind, when such pain or suffering is inflicted by or at the instigation of or with the consent or acquiescence of a public official or other person acting in an official capacity. It does not include pain or suffering arising only from, inherent in or incidental to lawful sanctions.

2.1.1.1.  A few queries The opening clause explains why the Convention’s definition covers only instances of torture in which public officials—or people acting in official capacities—are actively or connivingly involved.2 Such a limitation on a definition of the crime of torture would be decidedly peculiar in a philosophical treatment of the subject, but the Convention is not a philosophical disquisition. Rather, it is an international agreement among nation-states as signatories. It directly regulates the behaviour of governments, by obligating them to prevent torture within their respective jurisdictions and by obligating them not to become involved in the perpetration of torture in any way (for example, through the extradition of someone to a jurisdiction where he or she is likely to be subjected to torture). It does not directly regulate the behaviour of private citizens. Accordingly, because everything in the Convention is tailored to its regulatory purposes, its definition of ‘torture’ contains the limitation that has just been noted (Wisnewski 2010, 5; Wisnewski and Emerick 2009, 3). As we shall see in Chapter 5, the distinction between torturous measures carried out by governmental officials and torturous measures carried out by private individuals can be of significance in relation to the legal consequences that should follow when such measures have been undertaken. My focus in this book will lie principally on torture of the former kind, but the latter kind will also receive some attention, and there is no reason whatsoever for my philosophical investigation of the topic to exclude the latter kind of torture from my definition of the crime. A philosophical enquiry does not proceed within the constraints—the political and practical constraints—of an international agreement. Thus, for my purposes, the closing portion of the first sentence in the Convention’s definition is best put aside. More subtle and more serious are some other shortcomings in the definition. On the one hand, the drafters of the Convention were correct in holding that torture can be perpetrated for sundry purposes. By using the construction ‘for such purposes as’, they wisely acknowledged that torture can be employed for purposes apart from those that are listed. They likewise commendably grasped that

2   Admittedly, the final clause in the first sentence of the Convention’s definition begins with ‘when’ rather than with ‘when and only when’. However, once the definition is construed in tandem with the other provisions of the Convention, it becomes clear that the wording at the outset of that final clause should be understood as ‘when and only when’.

32

Introduction II: What is Torture?

the interrogational or intimidatory or punitive force of torture in any particular circumstances can be targeted at someone other than the person who is directly subjected to excruciating pain. On the other hand, the drafters erred in implying that the deliberate infliction of grievous pain is not properly classifiable as torture if it is undertaken for no ulterior purpose but simply for sadistic gratification (Seidman 2005, 896). Torture performed for sadistic pleasure is an especially evil mode of behaviour, and it should certainly not be omitted from the compass of an enquiry into the nature and wrongness of torture. Moreover, although the drafters of the Convention were wise to leave the definition’s list of purposes open-ended, we should not infer therefrom that every type of torture—where types are differentiated by reference to underlying purposes—is broadly on a par morally with every other type. In particular, we shall see later in this chapter and the next chapter that two kinds of torture (ephemerally incapacitative torture and edifying torture) can in certain circumstances be morally permissible. Of course, those varieties of torture are indeed such varieties, and they should not be excluded from the scope of a definition of the term ‘torture’ in any philosophical exploration of the matter. Nonetheless, because that term in the Convention denotes some very serious misconduct that is to be legally proscribed, the document’s definition should have explicitly omitted any morally legitimate instances of the two kinds of torture that can sometimes be morally legitimate. Admittedly, such a refinement might have muddied the clear waters of the Convention and might in any event have lain beyond the philosophical imagination of the drafters, who would perhaps have been disposed not to classify the relevant instances of incapacitative torture or edifying torture as torturous at all. Still, because ephemerally incapacitative torture and edifying torture partake of virtually all the features that underlie the wrongness of other types of torture, their morally legitimate instances should have been explicitly deemed to lie outside the purview of the Convention’s definition. Without such an explicit qualification, the Convention could lead to the imposition of legal penalties for morally impeccable conduct. Another disquieting aspect of the definition in the Convention is that it confines torture to acts, even though omissions—such as the withholding of medical treatment—can also serve to inflict excruciating pain deliberately (Kamenova 2009, 87; Moore 2007, 37 n7; Seidman 2005, 888). Whatever may be the explanation for the Convention’s limitedness in this respect, a philosophical investigation of torture should be more wide-ranging. A definition of the crime is unduly cabined if it is inconsistent with the fact that torture can occur through inaction as well as through actions. One further cause for concern in the Convention’s definition is its final sentence. The import of that sentence hinges on the meaning of the term ‘lawful’. Were that term taken to mean ‘in accordance with the law of the jurisdiction involved’  —quite a plausible understanding of the term in abstracto—the final sentence would amount to a licence for punitive torture whenever such torture is prescribed under a legal system’s schedule of sanctions (Seidman 2005, 895). To avoid that unacceptable implication of the Convention so construed, we should

Definitional Ventures

33

instead take ‘lawful’ to mean ‘not intended to inflict severe pain or suffering, and thus in accordance with the aims of this Convention’. Under such an interpretation, the final sentence does not loosen the Convention’s rigour.

2.1.1.2.  Some commendations Notwithstanding my queries about a few aspects of the Convention’s definition of ‘torture’, the virtues of the definition are considerable. As has already been noted, the sheer fact that the Convention leaves open the possibility of types of torture other than those listed is one of its strengths, for there are indeed some such types of torture (as we shall see presently). Likewise commendable is the drafters’ recognition that, within a definition of the sort which the Convention propounds, the diverse kinds of torture are best differentiated with reference to the purposes for which they are carried out. By contrast, an approach to differentiation with reference to the specific techniques and implements of torture would not have been apposite—largely because those techniques and implements are so dizzyingly multitudinous. As Elaine Scarry repeatedly emphasizes in her reflections on the topic (1985, chap. 1), virtually any ordinary household items can be put to use by a torturer in his efforts to inflict agony on his victims. Accordingly, a catalogue of instruments of torture would not be helpful as a constituent of any definition. Of course, to maintain as much is scarcely to suggest that dissimilarities among techniques of torture can aptly be ignored altogether in this book. On the contrary, because such dissimilarities give rise to variations in the gravity of the wrongs that are committed through instances of torture, we shall have to take account of them at a general level. Taking account of the heterogeneous techniques of torture in that general fashion, however, is quite different from specifying them within a definition of the crime. Another virtue of the definition in the Convention is the reference to severe pain or suffering. Admittedly, the disjunctive phrase ‘pain or suffering’ might be purely appositional and thus pleonastic (as disjunctions often are in legal documents). However, it can alternatively be understood as picking out two main kinds of agony: agony due to overwhelmingly intense pain experienced during a relatively short period of time, and agony due to somewhat less intense misery experienced during a considerably longer period of time. Of course, the distinction between those two types of anguish is not very sharp, but the important point is that the deliberate inducement of anguish of either type can be correctly classified as torture. Thus, for example, whereas deliberately immuring someone in cold and wet and cramped conditions would probably not amount to torture if it persisted for several hours, it would very likely amount to torture if it persisted for a much longer period. Quite appropriately, the Convention extends to an ordeal of the latter sort (Davis 2005, 163–4). In these respects as well as in the respects to which I have briefly adverted in § 2.1.1.1, the definition of ‘torture’ in the Convention is quite solid. My own definition will be more elaborate and more precise (as well as more flexible, in certain ways), and it will avoid the shortcomings that detract from the Convention’s

34

Introduction II: What is Torture?

formulation. Nonetheless, the gist of the Convention’s position is salvageable. It should be amplified and refined rather than jettisoned.

2.1.2.  The American definition In one of the interpretive understandings articulated by the United States Senate in its consent to the Convention against Torture, there appears a definition of the crime of torture with an emphasis on recounting the possible causes of severe mental pain or suffering: [W]‌ith reference to article 1 [of the Convention], the United States understands that, in order to constitute torture, an act must be specifically intended to inflict severe physical or mental pain or suffering and that mental pain or suffering refers to prolonged mental harm caused by or resulting from (1)  the intentional infliction or threatened infliction of severe physical pain or suffering; (2) the administration or application, or threatened administration or application, of mind altering substances or other procedures calculated to disrupt profoundly the senses or the personality; (3) the threat of imminent death; or (4) the threat that another person will imminently be subjected to death, severe physical pain or suffering, or the administration or application of mind altering substances or other procedures calculated to disrupt profoundly the senses or personality.3

Though this definition is not inconsistent with the Convention’s definition, it diverges significantly therefrom in several respects. Whereas the Convention’s definition prominently adverts to several general purposes that can be served by the use of torture, the American definition says nothing about those purposes. Moreover, whereas the Convention’s definition does not expand on the nature of mental pain or suffering, the principal concern of the American definition is to recount the potential sources of such suffering. Furthermore, whereas the Convention’s definition does not touch upon the techniques of torture, the American definition dwells upon those techniques (albeit at a high level of generality) in order to differentiate among types of psychological torture. 3   The text of the reservations, understandings, and declaration attached by the United States to its ratification of the Convention against Torture can be found in many electronic and printed sources. See, for example, . The definition which I have quoted is likewise quoted in Brunnée and Toope 2010, 252; Seidman 2005, 888 n18. I will not here deal with the efforts by some high-ranking lawyers in the administration of President George W.  Bush to incorporate a startlingly restrictive and muddled conception of torture into American policy-making. Because the pronouncements that stemmed from those efforts (circulated in memoranda within the Bush Administration) are so glaringly unsatisfactory as good-faith explications of the concept of torture, my detailing the inadequacies of those pronouncements would be largely superfluous—especially since such a task has been undertaken by many other philosophers and jurists. For some critical accounts of the Bush Administration’s machinations in regard to the use of torture, see Alvarez 2006; Angell 2005; Bassiouni 2006; Bilder and Vagts 2004; Brunnée and Toope 2010, 237–50; Chesterman 2008; Clark 2005; Danner 2004; Donohue 2008, 91–111; Ferzan 2004; Ginbar 2008, chap. 15; Guiora and Page 2006; Ip 2009; Keller 2005, 522–7, 549–56; Kipnis 2007; Koh 2005; Kreimer 2005, 187–201; Kutz 2007a; Kutz 2007b; Levinson 2005; Levit 2005; Lim 2008, 396–8; Luban 2005, 1452–61; Luban 2007; Luban 2009, 182–90; Mayerfeld 2007; Raviv 2004, 180–1; Rouillard 2005; Sands 2008; Scheppele 2005, 295– 305, 326–33; Soniewicka 2013, 191–2; Waldron 2010a, chaps 1 and 7; Wisnewski 2010, chap. 8. For a collection of key documents, see Greenberg and Dratel 2005. For a hard-hitting (and non-partisan) report, see Constitution Project’s Task Force on Detainee Treatment 2013.

Definitional Ventures

35

As has already been contended, a focus on techniques of torture is problematic in a definition of the crime. Notwithstanding that the American definition lessens the problems of such a focus by remaining at a high level of abstraction, the perils of concentrating on techniques are evidenced by some notable omissions in the American definition. Let us glance at two of those omissions. First, if someone suffers from a severe phobia, then torture can be perpetrated through the deliberate exposure of the person to situations that will trigger the phobia. Such exposure can induce unbearable and prolonged anguish. One of the most famous fictional representations of torture, the breaking of the will of Winston Smith in George Orwell’s Nineteen Eighty-Four, involves this sort of technique. No explication of the notion of severe mental suffering in a definition of the crime of torture is satisfactory if it overlooks the agony elicited in Winston Smith—and the agony that can be elicited in others who are in the grip of debilitating phobias.4 Second, the infliction of extreme humiliation on someone can be harrowing even if none of the factors enumerated in the American definition is responsible. Though many modes of conduct that inflict extreme humiliation are best classified as cruel or inhuman or degrading rather than as torturous, some are so extravagant as to fall within the ambit of any proper definition of torture. For example, in Haiti and in some other countries, sons have been forced to rape their mothers (Scholz 2007, 275); each party to such a forced act of intercourse goes ahead with it in order to spare the life of the other party. Now, although threats of lethal violence are operative in bringing about these terrible couplings, the overwhelming mental pain experienced by the parties is due not to the threats but to the couplings themselves. Even if a son can manage to avoid causing his mother any physical pain, the extreme humiliation and despair engendered by his sexual congress with her in front of jeering captors are best classified as agony produced by torture. Further examples arise when devoutly religious people are coerced—through threats of lethal violence against them or against members of their families—into cursing and besmirching sacred items of their faith. Certain aspects of the treatment of some Iraqi prisoners by American troops during the opening decade of the twenty-first century would constitute such examples, but some even more reprehensible instances occurred during the Nazis’ period of governance in Germany. Quite frequently during that period, devout Jews were ordered at gunpoint to spit or urinate or defecate on scrolls of the Torah. Stephen de Wijze records a particularly grotesque occasion on which a Polish rabbi was obliged to spit repeatedly on a Torah scroll (de Wijze 2009, 221). When the rabbi ran out of saliva, a Nazi soldier spat several times into his mouth and then forced him to continue to expectorate on the scroll. Even for an atheist, treatment of this sort would be savagely humiliating. For a devout Jew, the anguish caused would be at a devastating level that 4   Admittedly, the torture of Winston Smith probably falls within the scope of clause (1) of the American definition, since he is threatened with serious physical harm by hungry rats and is not merely exposed to their presence. However, in many cases—perhaps in the case of Winston Smith— the activation of a major phobia can be excruciating even in the absence of any prospect of physical harm. I write as someone afflicted with a severe needle phobia.

36

Introduction II: What is Torture?

warrants our classifying the treatment as torture (Geras 2011, 74). Again, although a threat of lethal violence was operative in the scenario which de Wijze unfolds, the extreme desolation of the rabbi was due not to that threat itself but to the actions that he was obliged to undertake. Given that the treatment of him should be classified as torture, the American definition of that crime is too narrow. Apart from the inordinate narrowness of the American definition’s conspectus of the techniques that bring about extreme mental suffering, the main shortcoming of the definition is its laconicism. It omits some of the illuminating elements of the Convention’s definition, and it does not expand at all on the differences—if any—between the Convention’s use of the term ‘intentionally’ and its own use of the phrase ‘specifically intended’. (As will be maintained later in this chapter, even the former term is slightly too strong. There can be instances of torture in which the infliction of agony is extremely reckless rather than fully intentional.) Thus, although the American definition’s partial unpacking of the notion of severe mental pain or suffering is valuable within its confines, the definition overall is inferior to the Convention’s formulation.

2.1.3.  Amnesty International’s definition In 1973, Amnesty International promulgated a definition of ‘torture’ that has been quite widely cited ever since in the philosophical and juristic literature on the topic: ‘Torture is the systematic and deliberate infliction of acute pain by one person on another, or on a third person, in order to accomplish the purpose of the former against the will of the latter.’5 Despite its iconic status in some quarters, this definition is unsatisfactorily narrow in a number of respects. For one thing, it refers only to acute pain. As can be inferred from my discussion in § 2.1.1.2, the reference to severe pain or suffering in the Convention’s definition is preferable. Moreover, the Amnesty definition is also unduly narrow in implying that torture always consists in the systematic infliction of severe pain. Although the administration of torture can of course be prolonged and carefully calibrated, it can in other contexts be quite short and impromptu. The character of torture as such is not dependent on the systematicity with which the deliberate inducement of agony is carried out. Perhaps most strange is the suggestion in the Amnesty definition that torture is always conducted by one person. If two or more people act in concert to bring about excruciating pain in someone else, the classifiability of their actions as torture is hardly negated by the multiplicity of the individuals involved. Finally, the Amnesty definition’s reference to the accomplishment of a purpose by a torturer is potentially misleading; such a reference is unexceptionable only if it is construed as consistent with the absence of any ulterior purpose. As has already been noted, the purpose underlying the perpetration of torture in any given context can reside in the sheer sadistic pleasurableness of the activity itself. A torturer need not be

5   Amnesty International 1973, 31. Among those who have quoted this definition (sometimes with slight modifications) are Brunnée and Toope 2010, 253; Paskins 1976, 138; Twining 1978, 153.

Definitional Ventures

37

seeking to achieve any end beyond the sadistic gratification that he derives from his vile misconduct. In several important respects, then, the definition propounded by Amnesty International four decades ago is excessively cabined. Like the principal vice of the American definition, however, the prime weakness of Amnesty’s formulation is its terseness. Its brevity makes it unhelpful for a philosophical enquiry into the complexities of torture.

2.1.4.  Philosophers’ definitions During the past few decades, and especially during the past decade, many philosophers have sought to encapsulate the nature of torture in pithy formulations. This chapter’s survey of recent definitions of the crime of torture will conclude here by critically examining several of the philosophical formulations that have been devised.

2.1.4.1.  Michael Davis on the ordeal of torture In a quirky but illuminating article, Michael Davis ponders some previous definitions of torture and then provides his own definition:  ‘[T]‌orture is the intentional testing of a sentient, helpless being’s ability to bear physical suffering against that being’s will and indifferent to its welfare’ (Davis 2005, 167). Whereas the Convention’s definition and the Amnesty definition take as given that every victim of torture is a person, Davis wisely recognizes that non-human animals as well as human beings can undergo torture. As he states:  ‘Nothing in the concept of torture requires the tortured to be human or rational; it is enough if the being can be made to suffer.’ Nevertheless, he immediately goes on to declare that his article will thenceforward ‘ignore the torture of animals, the insane, and the like. My concern is the moral status of torturing ordinary human beings’ (Davis 2005, 167). Throughout this book, I too will concentrate chiefly on torture wielded against human beings of sound mind. However, Chapter 3’s investigation of the wrongness of torture will touch upon the use of torturous techniques against mentally infirm human beings and against non-human animals. Thus, although a definition of the crime of torture should be consistent with the fact that paradigmatic cases of torture do involve human victims of sound mind, it should also leave ample room for cases in which the victims are non-human animals or mentally disabled people. Pro tanto, Davis’s definition is superior to the other formulations that we have scrutinized heretofore. 2.1.4.1.1.  A first query: the helplessness of victims Less commendable is Davis’s definitional affirmation that every victim of torture is helpless. Davis indeed adopts an overweeningly strong stand on the matter. Not long before he advances his definition, he writes: ‘Torture would be (conceptually) impossible if the tortured could protect themselves’ (Davis 2005, 166). On the one hand, Davis’s position is far from ridiculous. In the vast majority of

38

Introduction II: What is Torture?

situations in which torture is employed, the victims have indeed been overpowered and are thus helpless to defend themselves. Moreover, Davis is by no means the only philosopher who has taken the helplessness of any victim of torture to be a defining feature of the phenomenon. Famously, for example, Henry Shue in his classic article on the topic has centred his analysis of the wrongness of torture on the helplessness of its victims. Contending that the use of torture is morally worse in a crucial respect than are some instances of deliberate killing, Shue proclaims: [T]‌he torture victim has exhausted all means of defense and is powerless before the victors . . . In this respect torture is indeed not analogous to the killing in battle of a healthy and well-armed foe; it is a cruel assault upon the defenseless . . . The torturer inflicts pain and damage upon another person who, by virtue of now being within his or her power, is . . . entirely at the torturer’s mercy.6

Michael Ignatieff, among others, has articulated a similar stance: ‘Another way to seize the distinction between torture and killing in combat would be to observe that in combat pain or death is inflicted on those whose job it is to do the same. In the act of torture, pain and possible death are inflicted on a person who is disarmed and helpless’ (2004, 137). As has been remarked, it is quite understandable that so many philosophers and jurists assert or presuppose that the victims of torture are always completely helpless.7 Nearly all such victims are indeed defenceless while they are being subjected to torment. Nonetheless, Davis goes somewhat too far in insisting that the helplessness of victims is conceptually necessary for the classifiability of their travails as torture. As has been suggested by Frances Kamm and Uwe Steinhoff and William Twining (Kamm 2011, 7; Steinhoff 2009, 40–1; Twining 1978, 160), there are possible cases of torture in which the victims are not entirely helpless. Kamm puts forward the following example:  ‘Someone . . . is given a shield as a physical means of defense against a guard who is trying to cause him continuous severe distress in order to get him to give information. This defense, which could have been successful, fails and the guard succeeds in causing him severe distress.’ At some point in Kamm’s scenario, the pain inflicted by the guard presumably 6   Shue 1978, 130. In the elided portion of the final sentence in this quotation, Shue asserts that the victim of torture is ‘no longer a threat’. As Frances Kamm has remarked (2011, 7), Shue does not adequately distinguish between the defencelessness and the unthreateningness of a victim of torture. For another example of such a conflation, see Belvisi 2009, 70: ‘In the ticking bomb case, however, the terrorist is no longer a defenceless victim, but a criminal whose failure to collaborate is tantamount to aiding and abetting a murderous attack by other members of his terrorist group. On this basis, the prisoner has it in his power to avoid torture, or end it, by collaborating.’ Whereas the first sentence of this quotation submits that a terrorist can pose a threat even while undergoing torture, the second sentence adduces a reason—a largely unpersuasive reason—for thinking that a terrorist who undergoes torture is not defenceless. For some further instances of the helplessness/unthreateningness conflation, see Frowe 2011, 202; Gur-Arye 2004, 192; Hill 2007, 396; Kreimer 2003, 305 n95; Kreimer 2005, 224 n124; Kutz 2007a, 241; La Torre 2009, 28; Meisels 2008b, 169–70, 171–2; Seidman 2005, 896 n43; Wisnewski 2010, 111. The difference between helplessness and unthreateningness is recognized in Ignatieff 2004, 137–8; Miller 2005, 180; Sussman 2005, 16. 7   In addition to the theorists already quoted, see, for example, Posner and Vermeule 2006, 678; Scheppele 2005, 287 n3; Sussman 2005, 6–7 et passim, Sussman 2006, 227–8.

Definitional Ventures

39

becomes so harrowing and debilitating as to render the victim defenceless. Until that point, however, the torture is taking place even though the victim is capable of fending it off. Perhaps Davis would retort that the victim’s ordeal does not amount to torture until the intensity of the pain has overcome his or her ability to interrupt the ordeal. Such a reply would be lame, however. Unusual though the situation of torture in Kamm’s scenario is, the classification of it as a situation of torture is not a conceptual error. Steinhoff and Twining independently envisage circumstances in which a person is tortured at a distance with some special electronic or laser device that causes severe and continuous pain. (Kamm adduces some broadly similar scenarios, though with a focus principally on the threateningness of the victim rather than on his ability to defend himself.) Steinhoff —2009, 40—depicts a robbery in which the culprit wants to keep a jeweller alive for a while so that he can extract the combination of a safe from her. Thus, instead of threatening to shoot the jeweller, the robber employs a pain-inducing device to afflict her with agony as he demands that she reveal to him the combination of the safe. While writhing with pain, the jeweller manages to reach a gun. Her anguish prevents her from aiming very accurately, but she does manage to shoot in the general direction of the robber, who has to take cover. When the jeweller continues to shoot desperately, the robber eventually flees or is killed. As Steinhoff declares, ‘[t]‌he jeweller is obviously not defenceless. However, it seems that she was tortured nevertheless.’ Summarizing the events in his scenario, Steinhoff pertinently poses a rhetorical question:  ‘Someone was intentionally inflicting pain on [the jeweller] nearly as intense as the pain inflicted by drilling the unprotected nerve of a tooth, and doing so in order to get some information or in order to have the person [let go of the gun]—how could this not be torture?’ Precisely because Steinhoff’s example is more vividly detailed than Kamm’s, it is more powerful as a rebuttal of Davis’s insistence that every victim of torture is defenceless. As a conceptual claim, that insistence by Davis is false. Still, as has already been stated, he has not gone badly astray. There are credibly imaginable situations of torture in which the victims are not wholly defenceless, but those situations are very far from typical. Steinhoff readily acknowledges as much:  ‘I agree that in most cases (perhaps even in all real cases) [a victim of torture] will be defenceless, but this in itself is no reason to make this a definitional requirement’ (2009, 40). 2.1.4.1.2.  A second query: testing the victim’s capacity to endure suffering With his definition, Davis commits himself to the proposition that every instance of torture is aimed at testing and overwhelming the ability of a victim to endure agony. Though most types and instances of torture are indeed so aimed, not all types or instances are. Most notably, punitive torture need not be so aimed. Although somebody can indeed be sentenced to an open-ended course of torture that will be halted only when her will has been broken (or only when she has died), she can alternatively be sentenced to a course of torture that is clearly bounded. For example, she might be sentenced to undergo a fixed number of lashings or electric

40

Introduction II: What is Torture?

shocks. Barbaric though any such sanction is, it is not imposed as a test of the victim’s ability to withstand severe pain. Provided that the sentence is implemented in accordance with its terms, her punitive ordeal will cease at a certain point even if she has not altogether buckled in response to it.8 Davis dissents from the notion that a fixed course of lashings or electric shocks or canings is ever properly classifiable as torture. Earlier in his article, in the following passage (2005, 164), he distinguishes any such course of punishment from torture: Though cruel, such punishment [as flogging or caning] is not torture. The executioner is not required to cause suffering, only to carry out a sentence, say, so many hard strokes of a cane upon the bare back. The suffering is (in the words of the Convention ‘inherent in or incidental to lawful sanctions’). The legislator may have chosen the punishment (in part) because it is painful, and the judge may have imposed it for the same reason. But those intentions are independent of the executioner’s. In contrast, the torturer actually aims at causing the tortured to suffer (either as an end in itself or as a means).

Davis’s reference to the wording of the Convention bears out my anxiety over that wording which I have expressed at the end of § 2.1.1.1. The sheer fact that the law of a jurisdiction prescribes a course of torturous mistreatment as a punishment for some crime(s) does not render such a course of mistreatment non-torturous. Davis’s main argument in this passage is unsound, for there can be many instances of torture—of various types—in which the people who directly apply painful techniques to victims’ bodies do not themselves intend to cause any anguish to the victims. Most obviously, nearly all instances of placatory torture are of this kind. Let us examine here a variant of a scenario plumbed more than three decades ago by Alan Gewirth (1981, 8). Suppose that some terrorists have obtained weapons of mass destruction. They kidnap a leading scientist and show him their weapons and then release him to corroborate their claims about being in possession of such terrible devices. They proceed to announce that, unless a certain prominent conservative politician subjects his own mother to a specified course of torture at a designated time in front of television cameras, they will use their weapons of mass destruction against several major cities. Given that the terrorists have already murdered some other prominent public figures, their threat to unleash their weapons is shuddersomely credible. By inducing the politician to torture his mother, they hope to humiliate him and to demoralize the whole society. We will probe the moral bearings of placatory torture later in this book. For now, the key point is that such torture is indeed torture even though the person who directly applies the lashings or canings or electric shocks or lighted cigarettes to the victim’s body will not intend thereby to evoke any sensations of pain in the victim. If the conservative politician does resort to placatory torture, he will desperately hope that his mother will somehow emerge from the ordeal without 8   For some perceptive remarks on torture of this kind, see Sussman 2005, 30–3. See also Meisels 2008b, 167, 177. For a position on punitive torture similar to Davis’s position, see Miller 2011, § 1. For an extreme version of Davis’s stance, see Seidman 2005, 908. For some brief criticism of Davis’s stance, see Kershnar 2005, 237 n2.

Definitional Ventures

41

having experienced any significant discomfort. He knows of course that his hopes will almost certainly prove to be forlorn, but the eliciting of pain in her is decidedly not one of his objectives. He will be ecstatic if she does miraculously come through the horrible experience with little or no suffering. Still, despite his utter revulsion at the prospect of causing his mother to undergo severe pain, his administering of placatory torture to her is indeed an instance of torture. Unless Davis wishes to argue that all or virtually all placatory torture is not in fact torture, he will have to abandon the line of reasoning that underlies his position on punitive torture. Moreover, even interrogational torture—the kind of torture to which Davis’s claims about the testing of a victim’s capacity for the endurance of agony are most clearly applicable—can credibly occur in circumstances where the person who directly applies violence or noxious substances to a victim’s body does not intend to cause the victim any pain. Suppose that the chief domestic-espionage agency in a repressive regime has conscripted some men into its ranks to carry out the ‘dirty work’ of applying cattle prods or whips or other nasty instruments of torture to the bodies of people from whom the agency’s investigators are seeking information. Each conscript is monitored by the investigators as he performs his grim tasks, to ensure that he does not flinch from the unpleasantness of what he has been assigned to do. In all relevant respects, a conscript in these circumstances is on a par with the person who implements the punitive sentences in Davis’s scenario. A sadistic conscript might relish his assignments and might intend to elicit sensations of agony in his victims, but other conscripts might well feel compunctions about what they are obliged to do. Quite a few of them might hope dearly that their applications of torturous techniques to the victims’ bodies will somehow not cause any serious pain. They naturally recognize that severe pain will in fact almost certainly ensue from their wielding of those techniques, just as the official who implements the punitive sentences in Davis’s scenario recognizes that severe pain will almost certainly ensue from her whipping or caning or burning of the sentenced prisoners. Nevertheless, the more decent conscripts do not intend to induce anguish in their victims; the inducing of anguish is not one of their objectives, either instrumentally or intrinsically. All the same, the interrogational torture which they administer is indeed torture. Furthermore, even if it happens to be that none of the conscripts is morally decent—just as it might happen to be that the official who implements the sentences in Davis’s scenario is a sadist—there could have been some decent people among them. Pain-inducing intentions are not integral to the role that is performed by each of the conscripts. Hence, the line of reasoning that underlies Davis’s position on punitive torture is straightforwardly transferable to some instances of interrogational torture. Instead of inferring that interrogational torture is sometimes not torture when the people who prescribe it are different from the people who directly apply it, we should more sensibly conclude that Davis’s line of reasoning is unsound. Given that Davis does not furnish any other argumentative support for his position on punitive torture, we are well advised to reject that position. Somebody sentenced to a fixed number of lashings or canings or burnings or electric shocks has been sentenced to a course of torture as the punishment for his or her crime. Since

42

Introduction II: What is Torture?

punitive torture of this sort is not a test of the victim’s ability to withstand severe pain, we should likewise demur at Davis’s insistence that every instance of torture amounts to such a test. 2.1.4.1.3.  A third query: psychological torture revisited Another weakness in Davis’s definition is that it adverts only to physical suffering. It therefore implies that psychological torture is not torture at all. In that respect, his formulation is at odds both with the Convention’s definition and with the American definition. Although this chapter has already dealt with the matter of psychological torture in § 2.1.2, we are well advised to ruminate briefly on it again here—since Davis submits that ‘all the examples of “mental torture” seem to me either physical or something other than torture.’9 Let us readily agree with Davis that some torment-producing techniques that have often been classified as psychological torture are equally well classified or better classified as physical torture. Davis concentrates in particular on severe sleep-deprivation, and he is undoubtedly correct in maintaining that the suffering engendered by such a technique can best be regarded not only as psychological but also as physical. (Moreover, given that mental states and processes supervene on physical states and processes, every instance of psychological torment is associated with a physical substrate [Wisnewski 2010, 9]. Still, the supervenience of the mental on the physical is perfectly consistent with the fact that the anguish induced by some techniques of torture is experienced as predominantly psychological.) Though Davis is correct about severe sleep-deprivation, his broader denial of the reality of psychological torture is untenable. His lone argument for that denial is contained in the following passage (2005, 163): Where the body is not harmed, for example, where an interrogator simulates the torture of the child or spouse of the interrogated, or only makes threats, the suffering in question seems to work much as any scheme of extortion (or intimidation) does. Whether or not even this suffering is ‘purely’ mental (that, is suffering only humans are subject to), it is not torture. Torture is not a form of extortion; extortion presupposes rationality, torture does not.

Davis is a perspicacious philosopher, but his argument here is curiously weak. One puzzling aspect of the opening portion of this passage is that it refers only to the simulated torturing of some close relatives of a captive. For Davis’s purposes in this passage, it does not matter whether the torturing of the close relatives is genuine or skilfully simulated. After all, Davis is not addressing the question whether the mistreatment of the relatives is really torture; he is addressing the question whether the captive is undergoing torture by being confronted with a graphic spectacle of egregious physical abuse committed against some members of her family. The

9   Davis 2005, 163. For a somewhat different version of the same position, see Seidman 2005, 909. In a 2005 paper, Seumas Miller declares that ‘I agree with Davis that non-physical mental torture is not really torture as such’ (2005, 179). In an encyclopedia entry revised in 2011, Miller—somewhat reluctantly—accepts that certain types of psychological torture are indeed torture (2011, § 1).

Definitional Ventures

43

correct answer to that latter question does not depend on whether the egregious physical abuse is genuine or merely a clever simulation (nor on whether the people involved are genuinely some members of the captive’s family or merely some cleverly disguised impostors). At any rate, the fatal flaw in Davis’s argument is that it posits a tidy dichotomy in an area where no such tidiness is to be found. Davis presumes that the inducement of severe physical pain operates by overwhelming the deliberative rationality of a victim, whereas the inducement of severe psychological torment operates by trading coercively upon the deliberative rationality of a victim. A  neat division along those lines does not correspond to the actualities and possibilities of the use of torture, however. Although physical torture can of course be aimed at reducing a victim to a state of gibbering idiocy, it can alternatively be aimed at pressuring a victim into choosing coercedly to provide information or sign a confession or embrace some religious doctrine. That is, it can be aimed at prodding her to exercise her deliberative rationality in favour of doing what the torturers want her to do. She faces a choice between being subjected to further pain—probably in an escalating fashion—and capitulating to the torturers’ demands. If she responds by complying with their demands, the objective of the torturers will have been fulfilled through her exercise of her reflective capacities. Conversely, the inducement of severe psychological anguish can overwhelm— and can be aimed at overwhelming—the deliberative rationality of a victim. Someone can be reduced to a state of inarticulate babbling if she is overcome by despair or revulsion or hysteria brought upon her by fiendish techniques of psychological torture. One of the general techniques broached in my earlier discussion of psychological torture, the activating of a phobia, is especially likely to produce such an effect. Quite unsustainable is the proposition that such an effect ensues exclusively from the infliction of severe physical pain and not ever from the infliction of harrowing mental torment. In short, Davis has not supplied any persuasive reasons for denying the reality of psychological torture as such. Quite bootless is his invocation of the distinction between pressures that exploit the deliberative rationality of victims and pressures that obliterate (temporarily or permanently) the deliberative rationality of victims,10 since that distinction does not map smoothly at all onto the distinction between psychological torture and physical torture. Given that he does not advance any other basis for his rejection of the notion of psychological torture, and given that such torture partakes of the key wrong-making properties of physical torture, we should not follow Davis’s position on this matter. By declining to follow his position, one aligns oneself with the principal legal definitions of torture.

10   Though I have impugned Davis’s invocation of that distinction in this context, the distinction itself should scarcely be disregarded—as it is by Louis Seidman, for example, when he suggests that being tortured is tantamount to being ‘confronted with powerful threats’ (Seidman 2005, 898).

44

Introduction II: What is Torture?

2.1.4.1.4.  A fourth query: against the victim’s will Davis’s definition proclaims that torture is perpetrated against the wishes of the victim. Although that element of the definition is correct in application to the vast majority of instances of torture, it implies that the deliberate infliction of severe pain by consenting sado-masochists upon one another is not torture. To be sure, neither the classifiability of the sado-masochists’ activities as torture nor the moral status of those activities is clear-cut. In Chapters 3 and 5, we shall gauge that moral status (without entering into all the complexities of contemporary debates over the legitimacy of paternalism). Here the focus lies instead on the propriety of classifying those activities as torture. Whereas Davis does not directly ponder the interaction of sado-masochists, David Sussman does briefly contemplate whether such interaction amounts to torture (Sussman 2005, 8 n13). On the one hand, Sussman contends that the infliction of severe pain by a sado-masochist upon a willing victim is not torture if they have agreed on a signal or some ‘stop words’ with which the person who undergoes the painful treatment can bring it to a halt. As Sussman writes: ‘Such acts may differ from real torture in that they are (at least sometimes) conducted with a shared understanding that the victim retains an effective power to withdraw consent, and hence that his subjection is merely a kind of pantomime (however real the pain).’ Sussman is of course here using the term ‘pantomime’ in its American sense to denote a simulation. He indeed goes on to characterize the sado-masochistic interaction as ‘pretend-torture’ when the parties have agreed upon a signal for its cessation that can be invoked by the victim. He observes that ‘such pretend-torture may turn into real torture when the tormentor makes it clear that he does not recognize the victim’s right to opt out, taking himself to be as unconstrained morally as he is physically’. On the other hand, as the most recent quotation indicates, Sussman believes that sado-masochistic activity can become real torture if an agreed-upon signal for its cessation is disregarded. What he does not squarely address is a situation in which no such signal has ever been arranged. (Perhaps the absence of an opt-out route is due to the fact that the sexual gratification of the parties would be reduced if any such route were indeed agreed upon.) However, given that a situation of that sort would be marked by all the key characteristics which Sussman attributes to straightforward cases of torture—including the victim’s ‘position of complete vulnerability before another’ (2005, 8 n13)—he could accept that it involves genuine torture, and he would be right to take such a view. Yet when the absence of any opt-out signal is due to the wishes of the victim, who might be seeking to maximize his sexual excitement by doing without the availability of any such signal, the harm inflicted upon him is consensual. Thus, Davis’s characterization of non-consensuality as a defining property of torture is too strong.11 11   For a position similar to Davis’s, see Kershnar 2005, 224. For some opposing positions, see Kamm 2011, 60 n32; Moore 1997, 710–11; Twining 1978, 158–9. For some further remarks on sado-masochistic activity, see Wisnewski 2010, 54–6. In § 2.1.4.4.2, I  will indicate that Sussman would probably not classify any fully consensual sado-masochistic abuse as torture even in the absence of an agreed-upon signal for its cessation. If my ascription of that position to him is correct, then his approach to the matter is misguided.

Definitional Ventures

45

Though Davis does not mull over the activities of sado-masochists, he does adduce a different example of consensual pain-infliction: the inducement of pain in a wrestler W by another wrestler who grabs W in a hold and applies upward pressure on his arm until the aching becomes unbearable and obliges W to surrender (2005, 164). Davis correctly maintains that, in any ordinary wrestling match, the deliberate infliction of severe pain by the opponent of W is not torture. Two aspects of the situation, in combination, militate in favour of such a verdict. First, W has clear-mindedly and uncoercedly consented to take part in the wrestling match which he loses, and he has thereby accepted the possibility—indeed, the likelihood—of his being subjected to arduous holds in that match. Second, W can terminate his opponent’s infliction of severe pain at any time by conceding the match (or else by extricating himself from the hold in which the opponent has enfolded him). These two features of the situation are parallel to the two determinative features of sado-masochistic interaction where a signal for cessation has been fixed upon at the outset. Although the sado-masochistic interaction with the availability of a signal is like the upward pressure on W’s arm by the other wrestler in that each of those patterns of conduct involves the deliberate inducement of acute distress, the two features highlighted here are enough to render inappropriate the classification of those modes of conduct as torture. Conversely, because the second of those two features is missing from a context of sado-masochistic abuse where no opt-out arrangement is in place, such abuse without such an arrangement is torture regardless of how enthusiastically the victim has given his consent. Hence, Davis’s scenario of the wrestlers does not suffice to vindicate his inclusion of non-consensuality—without any qualifications—as an element of his definition of torture. Note, incidentally, that the discussion in the preceding paragraph can also serve to explain why soldiers who consensually undergo extremely painful afflictions in the course of resistance training are thereby participating in a simulation of torture rather than in full-blown torture (Twining 1978, 158). Albeit the people who administer the afflictions are deliberately causing the soldiers to experience severe pain, the process is consensual, and it can be stopped at any time by a soldier who has reached the outermost bound of his capacity to withstand the pain. In such circumstances, then, the treatment of the soldiers is an approximation of torture without quite being torture. (Of course, if the resistance training is such that a soldier is not allowed to withdraw whenever his endurance has reached its limit, then the training amounts to veritable torture. As we shall see later, resistance training along those lines is one species of edifying torture.) This discussion should close with a short caveat, which is applicable to this chapter as a whole. What has been under scrutiny here is the classifiability of various patterns of conduct as torture. When we conclude that a given mode of conduct is not appositely classifiable as torture, we are not perforce concluding that the specified mode of conduct is morally permissible. For example, as Chapter 3 will suggest, we might well arrive at the verdict that sado-masochists are acting wrongly when they inflict serious wounds upon one another, even if they have agreed upon a signal for halting the torment. Although their behaviour with the availability of

46

Introduction II: What is Torture?

an opt-out signal does not quite amount to torture, that fact is hardly in itself a moral licence for their actions. Further disseverable is the matter of the legal consequences that should attach to the sado-masochists’ conduct. One can look askance at their conduct as morally wrong while holding that there is no sufficient moral basis for their being subjected to any legal penalties. We shall return to such questions in Chapter 5. 2.1.4.1.5.  A final query: indifference to the victim’s welfare A final element of Davis’s definition that should be impugned here is its suggestion that the deliberate inducement of severe pain in some person is not torture unless the measures which induce the pain bespeak indifference to the person’s welfare. One minor respect in which Davis’s formulation has to be amplified is the addition of hostility—hostility toward the victim’s welfare—as a possible motivation of torturers. Sadistic torture is always marked by such an attitude, and some other varieties of torture are very frequently so marked. Exceedingly strange indeed would be the notion that a suitable reason for declining to classify the mistreatment of some person as torture is that the mistreatment was motivated by animosity toward the person’s welfare rather than by mere indifference thereto. Davis surely does not subscribe to such a notion; hence, the amplification proposed in this paragraph is a friendly amendment to his definition. More troubling for Davis is the clear possibility of torture that is motivated by a concern to promote the victim’s welfare. Some of the types of torture at which we have glanced in the preceding subsection are classifiable under this heading. For example, if some sado-masochists seek to intensify their sexual gratification by declining to arrange any signal through which the dominant parties’ abusive onslaughts could be brought to a halt by the victimized parties, those onslaughts are torture that is being undertaken by the dominant sado-masochists in order to provide the victimized sado-masochists with enhanced pleasure. Likewise, suppose that a course of resistance training in the military does not include any arrangement whereby the soldiers who undergo harsh treatment can arrest the process, and suppose that the reason for the absence of any such arrangement is that the training would be considerably less effective if the soldiers knew that they could end it whenever it might become unduly arduous. In that event, the soldiers are being tortured on the basis of concern for their welfare. They are being tortured to augment their hardihood and resourcefulness for the future. Perhaps even more troublesome for Davis’s definition is that torture undertaken for the spiritual welfare of heretics and miscreants is very credibly possible. Indeed, such torture has frequently taken place in many countries at various points in history. Torture to save the soul of an apostate or a scoundrel plainly does not proceed out of indifference toward the victim’s welfare. On the contrary, it proceeds from a hideously monomaniacal resolve to uphold an aspect of the victim’s welfare that is perceived by the torturers as supremely important. As Hauke Brunkhorst writes: ‘Torture [in Europe during the late Middle Ages] was supposed also to offer the defendant a chance to salvage his or her immortal soul from eternal condemnation by means of revocation and acknowledgment of Christianity’s objective

Definitional Ventures

47

truth.’12 (Of course, my remarks here scarcely imply that the torture of heretics has always been driven by pure solicitude for the heretics’ spiritual welfare. Struggles for power among competing faiths and competing denominations have quite often played a greater role in impelling such torture than has any fanatically pious solicitude for the souls of the victims. Frequently, the primary objective has been to cow other heretics into abandoning their creeds. Still, that observation is fully consistent with this paragraph’s main point—namely, the point that a concern for a person’s spiritual interests can very credibly motivate the perpetration of torture and other atrocities against that person.) The examples adduced here reveal that the closing words of Davis’s definition are in need of qualification. Neither unintelligible nor even implausible is the idea of torture inflicted with the aim of furthering the welfare of the people who undergo the torture. Still, the reason for Davis’s inclusion of the aforementioned closing words is clear enough. His distillation of the nature of torture appears just after his short discussion of the differences between torture and painful medical care. Most salient among those differences is that medical care aims to improve the overall physical and psychological well-being of a patient whereas nearly every instance of torture aims to impair the overall physical and psychological well-being of a victim. With the closing words of his definition, Davis endeavours to capture that distinction between the profession of torturers and the profession of doctors. (Oddly, Davis does not point out another conspicuous dissimilarity between torture and medical treatment. Whereas the severe pain induced by torture is an intended result—or occasionally a result of extreme recklessness—any pain caused by medical treatment is normally a side-effect, albeit often a readily foreseeable side-effect, rather than something desired as a means or an end. To be sure, there are some unusual situations in which medical treatment does involve the deliberate engendering of pain. Kamm adduces the following example: ‘[S]‌uppose that the only way to stop a patient from falling into a permanent coma is to keep him in intense pain, because only intense pain will get him to try to resist us, and it is this act of resistance that we try to bring forth in order to prevent the coma’ [2011, 59–60 n31]. Twining likewise draws attention to an example of a patient who is undergoing aversion therapy [1978, 159]. Such therapy unfolds in part through the deliberate subjection of the patient to considerable pain in the presence of certain environmental stimuli. Nevertheless, these scenarios do not undermine my generalization about medical treatment, which states what is normally true rather than what is always true. Generally, though not invariably, the painful effects of medical treatment are unintended even when they are amply foreseeable.) Hence, although the reference in Davis’s definition to the welfare of any victim of torture is simplistic, it should be modified and expanded rather than 12   Brunkhorst 2009, 75. See also Améry 1980, 34; Kreimer 2003, 299 n75; Wisnewski 2010, 33–6. For a good historical account of the era of salvation-oriented torture in Europe, see Peters 1989, chap. 2. For some reflections on Saint Augustine’s salvation-oriented justification of the torture of heretics, see Kramer 2003, 98–100.

48

Introduction II: What is Torture?

scrapped. We do indeed need to distinguish between torture and very painful medical or dental procedures. Instead of drawing that distinction as Davis does, we should conclude that the perpetration of torture is typically hostile or indifferent toward the physical and psychological welfare of any victim. In exceptional circumstances, where the aim of some torture is to build up the ability of a victim to withstand arduous onslaughts in the future, or where the aim is a therapeutic objective of the sort mentioned above and explored in § 2.2.12.2 below, the torture can be promotive of the long-term physical and psychological welfare of a victim.

2.1.4.2.  Kershnar’s formulation Stephen Kershnar (2005, 224)  endeavours to explicate the concept of torture with the following biconditional formulation: ‘One person tortures another if and only if the first intentionally imposes great suffering in a short amount of time on the second and the second neither willingly accepts it nor validly consents to it.’ Although this laconic formulation deals only with the torturing of persons, it leaves room for the application of the concept of torture to the mistreatment of non-human animals. It is not explicit on that score in the manner of Davis’s definition, but it does not assert or imply that abusive measures count as torture only if they are undertaken against human beings. One minor point of unclarity in Kershnar’s formulation pertains to the unexplained difference between willingly accepting something and validly consenting to it. Presumably, what Kershnar has in mind is a distinction between retrospective endorsement (or acquiescence) and anticipatory approval. In any case, Kershnar’s inclusion of non-consensuality as a necessary condition for the classifiability of conduct as torture is vulnerable to the same queries that have been raised about Davis’s inclusion of such an element in his own definition. Apart from the uninformative brevity of Kershnar’s formulation, its main weakness lies in its reference to a short amount of time. Though the vagueness of such a reference is unproblematic, there is no real basis for a temporal restriction of this sort. To be sure, the restriction is not as confining as it might initially appear, for Kershnar indicates that ‘[t]‌he duration might include a few days of extreme discomfort’ (2005, 224). The vast majority of instances of torture occur within such a limit. Nevertheless, there is no reason to stipulate that a process of mistreatment cannot amount to torture if its deliberate production of severe suffering has taken full effect—or is likely to take full effect—only after many days or weeks have elapsed. As Davis (2005, 163–4) rightly observes, some techniques of torture can work through the gradual intensification of a victim’s suffering to the point of unbearable agony. Because the temporal limitation in Kershnar’s exposition of the nature of torture leads to the conclusion that the gradual techniques are never properly classifiable as torture, the limitation is quite unwarranted. In that respect, as well as in the other respects touched upon here, Kershnar’s account of the nature of torture is plainly in need of improvement.

Definitional Ventures

49

2.1.4.3.  Miller and the complexities of torture Seumas Miller takes his definition of torture through several layers of refinement and elaboration until he arrives at the following formulation: ‘Torture is: (a) the intentional infliction of extreme physical suffering on some non-consenting, defenceless person; (b) the intentional, substantial curtailment of the exercise of the person’s autonomy (achieved by means of (a)); (c) in general, undertaken for the purpose of breaking the victim’s will’ (2011, § 1, emphasis in original). Albeit most of the weaknesses in this definition are the same as those that mar Davis’s definition, a few points deserve some fresh attention here. 2.1.4.3.1.  The third clause Let us start with the third prong in Miller’s tripartite definition. Despite his italicizing of the phrase ‘in general’, Miller does not precisely and explicitly indicate what that phrase is supposed to mean. It could be construed as ‘typically but not invariably’. Were the phrase to be understood in that fashion, the third prong in Miller’s definition would be tolerably solid. Though in § 2.1.4.1.2 we have probed some types of torture that are not structured by the aim of testing or breaking the capacity of each victim to withstand intense pain, most types and instances of torture are structured by such an aim. Thus, if the phrase ‘in general’ were to be glossed as has just been suggested, the third component of Miller’s definition would not be very objectionable—despite its excessive sweepingness and consequent misleadingness. However, Miller appears to be using the words ‘in general’ to convey something much stronger. That is, he appears to mean that torture always is undertaken for the purpose of breaking a victim’s will; it always is undertaken for that purpose in addition to any other purpose(s) for which somebody might resort to it.13 Insofar as the third clause in his definition is imparting this excessively strong thesis, it is unsustainable. Having already seen as much in § 2.1.4.1.2, we should take note here of another type of torture that belies Miller’s definitional claim. As will be recounted in this chapter’s typology of torture, some instances of torturous mistreatment are aimed at intimidating people other than those who are directly subjected to the mistreatment. Often, the direct victims of such abuse are tortured to death or are tortured savagely before being put to death. In a context of this sort, the objective of the torturer resides not in breaking the will of the victim—whose will no longer exists and therefore no longer matters at the end of the process—but instead in delivering a horrific threat to a much broader group of people. Outlandish is the notion that broadly addressed intimidatory torture is not properly classifiable as torture. It is in fact one of the vilest varieties of the phenomenon.

13   For some other assertions of this excessively strong view, see Gur-Arye 2004, 195 n4; Simester 2008, 309.

50

Introduction II: What is Torture?

2.1.4.3.2.  The second clause Although the chief shortcomings in the opening and closing prongs of Miller’s definition are largely the same as several of the shortcomings in Davis’s formulation, the middle clause of Miller’s definition is somewhat different. Much of what Miller says in justification of that second element of his definition is pertinent and perceptive (2011, § 1): At the very least the torturer is intentionally exercising control over the victim’s body and his attendant physical sensations, e.g. extreme pain. Indeed, in an important sense the victim’s body and attendant physical sensations cease to be his own instrument, but rather have become the instrument of the torturer. Moreover, by virtue of his control over the victim’s body and physical sensations, the torturer is able to heavily influence other aspects of the victim’s mental life, including his stream of consciousness; after all, the victim can now think of little else but his extreme suffering and the torturer.

These quoted remarks are entirely apt, partly because they do not employ the language of ‘autonomy’ to describe the key feature of torture which they are recounting. Albeit such language is of course appropriate when the victim of torture is an adult human being endowed with sound mental faculties, a definition should not imply that the deliberate infliction of severe pain never counts as torture if the victim is a sentient creature of some other type such as a young child or a lunatic. (This point is closely related to the one respect in which the first prong of Miller’s definition is inferior to Davis’s conception of torture: namely, the fact that Miller’s definition presumes that every victim of torture is a person.) Though the controllingness of torture does radically curtail the autonomy of any victim if he or she is possessed of autonomy, torture exerts its controlling force even when a victim is not possessed of reflective agency. Its tendency to concentrate the mind of the victim on her own misery for the purpose of the torturer—a purpose that might consist simply in the derivation of sadistic gratification—is applicable to any victim regardless of whether she is old enough and mentally lucid enough to be autonomous. Hence, notwithstanding that the second component of Miller’s definition points to something important, it should be reformulated with a more ample reach.

2.1.4.4.  Sussman and the difficulties of defining torture In a short law-journal article (2006, 225), Sussman remarks that torture ‘has indeed proved surprisingly resistant to any very clear definition in current debates about its use and justifiability’. Nonetheless, in a slightly earlier article, Sussman himself propounds a definition—even though it is not quite labelled as such. After quoting the definition that appears in the Convention against Torture, Sussman writes: ‘At a minimum, torture involves the deliberate infliction of great pain or some other intensely distressing affective state (fear, shame, disgust, and so forth) on an unwilling person for purposes that person does not and could not reasonably be expected to share’ (2005, 5).

Definitional Ventures

51

2.1.4.4.1.  A preliminary point: a problematic addition Especially in light of the opening clause, we might infer that Sussman believes that the elements specified in his formulation are necessary and sufficient conditions for the occurrence of torture as such. However, he goes on straightaway to add another element in order to distinguish situations of torture from certain other situations in which people are deliberately subjected by assailants to severe pain. For example, ‘I might punch a stranger in the face, breaking his nose, and then run away. The victim here may experience great pain, but we would not normally say he had been subjected to torture’ (Sussman 2005, 6). In an effort to exclude the face-punching scenario and other relevantly similar scenarios from the category of torture, Sussman goes beyond the formulation above by contending—in a manner reminiscent of Davis—that every victim of torture is defenceless. He writes (2005, 6): ‘Victims of torture must be, and must realize themselves to be, completely at the mercy of their tormentors . . . . The victim of torture must be unable to shield herself in any significant way, and she must be unable to effectively evade or retaliate against her tormentor.’ As we have seen in § 2.1.4.1.1, any claims of the sort which Sussman advances here are too strong. Although the vast majority of instances of torture do involve helpless victims, there can credibly be situations of torture in which the victims are able to defend themselves to a considerable degree. Kamm’s scenario and especially Steinhoff’s scenario—each of which we have pondered in § 2.1.4.1.1—establish the credible possibility of such situations. We should thus here leave aside Sussman’s insistence that, in every possible case of mistreatment which is properly classifiable as torture, the victim of the mistreatment is defenceless. Let us concentrate instead on his opening formulation. (Note, incidentally, that the insistence on the defencelessness of victims of torture will in any event not serve to keep every face-punching case out of the category of torture. Suppose that the victim of the punch is asleep or that her arms are filled with bundles of groceries at the time of the unexpected blow, and suppose that she loses consciousness for a while in the immediate aftermath of the punch as the assailant flees. In those circumstances, the victim is helpless both before and after the assailant has struck her. In relation to those circumstances, then, Sussman has left himself unable to explain why the punching of the victim does not amount to torture.) 2.1.4.4.2.  A matter of consent We have seen in § 2.1.4.1.4 that some victims of torture in credible circumstances might willingly undergo the tribulations to which they are subjected. Hence, Sussman’s inclusion of non-consensuality as a defining feature of torture is untenable. However, we should recall that consensual mistreatment counts as torture (rather than as a simulation of torture) only when the victim cannot bring it to a halt at a time of his or her choosing. As my earlier discussion of non-consensuality has remarked with reference to the footnote by Sussman on sado-masochism, he does not squarely address a situation in which a willing victim V and a perpetrator P deliberately decline to arrange a signal that would enable V to terminate P’s severe mistreatment of her. Nonetheless, we can now conclude that Sussman would deny that the subjection of the victim to serious abuse in such a situation

52

Introduction II: What is Torture?

does amount to torture. He would very likely contend that the abuse is a simulation of torture rather than an instance of genuine torture. To see why he would very likely arrive at that mistaken verdict, we should divide the torture of V by P into two components: T1, the torture that occurs up to the point at which V would have signalled for the discontinuation of the abuse if any signal for cessation had been agreed upon; and T2, the torture that occurs past that point. In my discussion of consensual torture in § 2.1.4.1.4, I have regarded both T1 and T2 as consensual. After all, the reason for the absence of any opt-out signal is that V has desired through that absence to intensify her sexual gratification or to toughen herself against hardships. Thus, because all the serious mistreatment inflicted on V by P is consensual, and because the open-endedness of that mistreatment makes it classifiable as torture, I have concluded that there can be instances of consensual torture. Sussman would very likely disagree with me both about T1 and about T2. With regard to T1, he would presumably maintain that the interaction between P and V is consensual and is therefore only a simulation of torture. With regard to T2, he would probably hold that the subjection of V to mistreatment is non-consensual but that it is done for a purpose which V shares—namely, the purpose of enhancing V’s and P’s overall sexual gratification (in the case of the sado-masochists) or the purpose of strengthening V’s capacity to withstand torture in the future (in the case of resistance training in the military). Sussman would of course be correct in thinking that T1 is consensual and that T2 is imposed for a purpose which V shares, but otherwise his analysis of the interaction between P and V—that is, the analysis which I have attributed to him—is misguided. By contending that T1 because of its consensuality is a mere simulation of torture, he would pro tanto be effacing the distinction between the interaction of P and V (‘Case 1’) and the interaction of some other perpetrator and some other victim who agree upon a signal for cessation that is subsequently disregarded by the perpetrator (‘Case 2’). Sussman explicitly mulls over a situation of the latter sort, as I have noted in my earlier discussion of this matter. He declares that the mistreatment of the victim by the perpetrator is transformed from pretend-torture to real torture once the perpetrator pays no heed to the agreed-upon signal for desisting. Until that juncture has been reached, the mistreatment is only pretend-torture. Thus, as far as T1 is concerned, Sussman would obliterate the difference between Case 1 and Case 2; in each case, according to him, T1 is no more than a simulation of torture. His elision of that difference is highly dubious, for V in Case 1 is fully aware that no signal for cessation is available, whereas the victim in Case 2 is confident that such a signal has been agreed upon. (Moreover, the agreement on that signal may well have been sincere on both sides. The perpetrator’s disposition to advance from T1 to T2 might not emerge until some point after T1 has begun.) Insofar as Sussman would hold that T1 is a mere simulation of torture both in Case 1 and in Case 2, he would fail to capture the crucial experiential dissimilarity between those two situations for the respective victims. Equally inapposite is the view that T2 is non-consensual in Case 1.  V has precommitted herself by deliberately declining to arrange any signal for the termination of P’s abusive measures, and she has done so precisely in order to

Definitional Ventures

53

promote her own interests in sexual pleasure or in fortified resistance to pain. The sense in which V consents to P’s infliction of T2 is no more mysterious than the sense in which Odysseus consents to being kept bound to the mast during the song of the Sirens even though he is pleading to be untied. Indeed, given that V is actively pursuing the purpose for which T2 is imposed on her by P, and given that V clear-sightedly understands the implications of that purpose (instead of being under any misconceptions about the need for the occurrence of T2), the fact that T2 is imposed on her for that purpose entails her having consented to T2. If Sussman were to submit that T2 in Case 1 is non-consensual, he would be mistaken. Sussman might well accept that T2 in Case 1 is consensual, since he would undoubtedly accept that T2 in Case 1 is imposed for a purpose which V clear-sightedly shares. Accordingly, he is committed to maintaining that T2 in Case 1 is only a simulation of torture rather than an instance of genuine torture. For Sussman, there is a sharp contrast between T2 in Case 1 and T2 in Case 2; only the latter is correctly classifiable as torture, in his eyes. Quite unclear, however, is what the motivation for such a stance could be. There is of course a major difference between T2 in Case 1 and T2 in Case 2, but it is a concrete moral difference that bears on the wrongness of T2—or on the gravity of the wrongness of T2—rather than on the classifiability of T2 as torture. Though T2 in Case 1 is imposed for a purpose which V clear-sightedly shares, it is the deliberate subjection of V to severe pain which he desperately wishes to terminate throughout the time when the pain is being induced. Hence, the lone point of dissimilarity between T2 in Case 1 and T2 in Case 2 is that only the former is imposed for a purpose that is clear-sightedly shared by the person who undergoes T2. Only the former is consensual. Although that dissimilarity is manifestly of concrete moral importance, its bearing on the status of T2 as torture is obscure. A definitional insistence on the non-consensuality of torture seems to be the only thing that would motivate Sussman to discern or posit such a bearing. In a motivating role of that sort, however, the definitional insistence is an arid dogma. It is a tail wagging a dog, as a theory is distortively shaped to fit an initial stipulation that does nothing to illuminate the nature of torture.

2.1.5.  Some of the lessons of this survey of definitions Various other philosophers and jurists have put forward definitions of the crime of torture, but most of the additional formulations either are virtually identical to some of those examined heretofore or are too dubious to be worthy of perusal. Falling into the latter category, for example, is the following definition propounded by Jeremy Bentham in a posthumously published manuscript on torture: ‘Torture, as I understand it, is where a person is made to suffer any violent pain of body in order to compel him to do something or to desist from doing something which done or desisted from the penal application is immediately made to cease.’14 Even as a definition of interrogational torture or act-impelling 14   Twining and Twining 1973, 309. This definition appears in a manuscript ‘Of Torture’ by Bentham, which was published for the first time in the article by the Twinings. Bentham proposed some even more dubious definitions of torture in another of his manuscripts, ‘Of Compulsion and herein of Torture’, also published in the article by the Twinings. See Twining and Twining 1973, 326.

54

Introduction II: What is Torture?

torture or incapacitative torture—in abstraction from any other species of torture—this formulation would be far too narrow. Dissecting its weaknesses is not worthwhile. At this stage, then, before we move on to my catalogue of the principal types of torture, we should take stock of the chief lessons that can be learned from this survey of juridical and philosophical perspectives on the matter. An enumeration of those lessons will facilitate the transition to the remainder of this chapter, by organizing an array of points to which any satisfactory account of the nature of torture must pay heed. 1.  Perhaps most obvious, but also most important, is that any plausible definition of the crime of torture has to affirm that torture consists in the deliberate infliction of severe pain or suffering (or occasionally in the extravagantly reckless infliction of severe pain or suffering). Any definition that omits the element of severe pain or suffering—whether the inducement of great anguish is undertaken as a means or as an end in itself—would be woefully unsatisfactory. Of course, to say as much is not quite to say that every instance of torture results in the victim’s experiencing of severe pain or suffering. For example, if a victim under interrogation capitulates and provides verifiable information very soon after the administration of torture to her has begun, she might spare herself from the agony that would have afflicted her if the torture had continued. Torture that is interrupted, whether in this manner or in some other manner (perhaps through the intervention of a third party), might not persist long enough to produce the anguish that it would have produced in the absence of the interruption. Still, although some credibly possible instances of torture do not actually lead to severe pain or suffering,15 the defining aim of every such instance resides in the infliction of such pain or suffering—or, on the occasions when torture is extravagantly reckless rather than deliberate, its overwhelmingly likely consequence is the inducement of such pain or suffering. This point may seem obvious. After all, every definition of torture encountered in the foregoing survey has made prominent reference—in one way or another— to the infliction of agony. Nonetheless, as we shall behold in Chapter  3, some major discussions of the wrongness of torture have proceeded without any explicit focus on the fact that torture involves the deliberate or extravagantly reckless causation of agony. The centrality of that feature of torture in the formulation of any adequate definition of the crime should alert us to its centrality in any adequate account of torture’s wrongness. 2.  In constructing a typology of torture, we are best advised to concentrate on the general purposes for which people might resort to the use of torturous techniques—rather than on those endlessly varied techniques themselves. Though no compilation of the sundry kinds of torture should profess to be exhaustive, a

  I examine this matter from a different angle in § 2.3.1.1 below.

15

Definitional Ventures

55

compilation will be much more likely to cover most of the relevant territory fruitfully if it asks why people would ever seek to elicit excruciating anguish in others. What general ends might they thereby be pursuing? Such a question of course allows that some torturers sadistically pursue the evocation of intense pain in other people (or in non-human animals) as an end in itself. 3. A  definition of torture in a philosophical disquisition on the topic should be capacious. For example, although the distinction between torture perpetrated by public officials and torture perpetrated by private individuals (without the acquiescence or cooperation of public officials) will be of some importance in Chapter 5, it should not be invoked here in favour of narrowing this book’s general conception of torture. With regard to the classifiability of certain conduct as torturous—rather than with regard to the law’s proper responses thereto—the public or private status of the person who engages in the conduct is immaterial. Likewise, a definition of torture in a philosophical disquisition should not confine itself to the infliction of physical agony. Overwhelming mental agony that has been inflicted deliberately, or in an extravagantly reckless fashion, is also within the scope of an apt definition. Moreover, a torturer’s deliberate or extravagantly reckless inducement of agony can be effected through omissions as well as through actions. 4.  If we train our attention on the victims of torture, we can detect some further respects in which a philosophical explication of the concept of torture should be expansive. For example, the status of a victim as a reflective agent is not decisive for the classifiability of some mode of conduct as torture. That status of the victim can have a significant bearing on the gravity of the mode of conduct—and on the specific ways in which it is wrong—but severely abusive measures can count as torture even if they are directed against insane people or young children or various non-human animals. While a theory of torture should acknowledge the special importance of the use of torturous techniques against people who are endowed with reflective agency, it should also recognize the possibility of the use of such techniques against other sentient beings. In addition, although the vast majority of instances of torture occur against the wishes of the victims thereof, some people under some circumstances do consent to being subjected to torture. A definition should not deny or ignore the possibility of consensual torture, even while it indicates that any instances of such torture are highly atypical. Similarly exceptional are instances of torture that are undertaken to enhance the physical and psychological well-being of the victims. Rare though such instances of torture are, however, we have observed that they are credibly possible in certain circumstances. Hence, their possibility should not be excluded by any definitional stipulation. Nor should any definition rule out the possibility of situations of torture in which the victims are not entirely defenceless. Any such situations are strikingly

56

Introduction II: What is Torture?

unusual, but they are not unintelligible or utterly fanciful. A theory of torture and of torture’s wrongness should not be unattuned to the possibility of such situations. 5.  Finally, notwithstanding that torturers typically endeavour to break the resistance of any victims on whom they wreak havoc, not all torture partakes of such an aim. Testing the endurance of victims is central to interrogational torture and to some other varieties of torture, but there are still further varieties—such as punitive torture and placatory torture—of which many instances are not characterized by such testing. Once again, then, anyone who develops a general conception of the nature of torture should refrain from suggesting that a typical feature of the phenomenon is an essential feature.

2.2.  Varieties of Torture One’s reflections on the general nature of torture can benefit greatly from a study of the multifariousness of the phenomenon. Although the controversies over the use of torturous measures in recent years have been concerned predominantly with interrogational settings, and although the use of torture in such settings will receive extensive attention herein, one’s conclusions about torture will be quite skewed if one does not also take account of the diverse non-interrogational purposes for which people might deliberately inflict severe pain upon others.16 Even one’s thinking about interrogational torture itself can be both stimulated and sharpened through one’s contemplation of those additional purposes. Such contemplation will reveal some commonalities, as well as dissimilarities, among the ways in which torture is wielded. Hence, let us begin by mulling over the nature of interrogational torture—which itself comprises several varieties—and let us then move on to the other principal species of torture. (At the end of this typology, I will present a chart of the sundry species of torture.)

2.2.1.  Interrogational torture In the contemporary disputation over the moral bearings of torture, the kind of torture that has received far more attention than any other is that which takes place in interrogational contexts. Indeed, some writers proceed as if that type of torture were the only type. Ignatieff, for example, defines torture as ‘the deliberate infliction of physical cruelty and pain in order to extract information’ (2004, 136). 16   For example, while aptly highlighting the ways in which the fear experienced by a victim of torture intensifies her physical agony, David Luban contrasts the situation of such a victim with that of soldiers in resistance training. The soldiers ‘know that within a short, fixed period of time the treatment will stop. They have none of the fear of a torture victim who knows neither of these things, and whose captors tell him that unless he talks he may be in Guantánamo forever’ (Luban 2009, 194). Because Luban purports to be discussing torture generally, he inadvertently implies here that a fixed course of punitive torture is not torture at all.

Varieties of Torture

57

We should not follow Ignatieff’s conflation of the genus with one of its species, but that conflation is understandable against the background of the controversies that have swirled in the United States and other Western countries in recent years. The morality of interrogational torture has been at the centre of those controversies. Indeed, as we shall observe, the morality of only one main sort of interrogational torture has been at the centre of those controversies. Our exploration of this matter will therefore require the drawing of some distinctions.

2.2.1.1.  Prospective versus retrospective When constables or other investigative officials are striving to avert serious crimes or to apprehend major criminals or to disrupt terrorist networks, they might sometimes be disposed to resort to the use of torture in order to extract crucial information from some of the people whom they are interrogating. Similarly, when constables and prosecuting officials are endeavouring to clarify the facts of past criminality and to obtain confessions from people who are believed to have been directly involved, they might be disposed to resort to the use of torture in order to acquire the testimony and confessions and investigative leads that they will need for successful prosecutions. Under repressive regimes, of course, the domestic-surveillance personnel who monitor the activities of members of the public in order to quash the expression of dissident views and the formation of dissident groups will frequently resort to torturous methods of gaining ‘cooperation’ from hapless citizens who might possess information about malcontents. One key distinction to be drawn within the category of interrogational torture, then, is the division between prospective or forward-looking uses and retrospective or backward-looking uses of such torture (Bobbitt 2008, 371–3, 376–9, 392; Bufacchi and Arrigo 2006, 360; Gross 2004a, 1487–8; Kadish 1989, 347; Levinson 2003, 2032; Meisels 2008b, 166–7). On the former side of that distinction are the officials—in benign systems of governance or in malign systems of governance—who seek information to thwart acts of terrorism and criminality or to discover the whereabouts of outlaws or to destroy criminal gangs and terrorist organizations. Also on the prospective side of the distinction are the secret police in an authoritarian regime who seek information to stifle the emergence of opposition to the regime. On the retrospective side of the division are the constables and prosecuting officials who seek confessions from suspects or statements from other people in possession of information that could help in the solving of past crimes. In bygone centuries, brutal methods of extracting confessions (and other information relating to past crimes) were common even in countries that have subsequently become flourishing liberal democracies. For several hundred years in a number of Continental European countries, the judicial use of torture to elicit confessions was routine.17 In the United States through the midpoint of the twentieth century, the use of terrible violence by the police in their interrogation 17   The leading book on this topic—a fascinating book—is Langbein 1977. For a short summary, see Langbein 2004. See also Peters 1996, chap. 2.

58

Introduction II: What is Torture?

of suspects—known as the ‘third degree’ —was dismayingly widespread (Kreimer 2003, 313–15; Kreimer 2005, 221–4; Waldron 2010a, 241–2). Such methods continue to be practised routinely in the many countries of the present-day world that are governed by despotic regimes, but they have largely disappeared from the systems of governance and policing in Western nations. When any incidents of brutality in the efforts by constables to elicit confessions do occur and do become known in those latter nations, they are roundly condemned. Any evidence stemming from them is now strictly inadmissible in judicial proceedings. Hence, in the contemporary debates within some Western nations over the morality of interrogational torture, the ethical bearings of retrospective uses of such torture have not been at issue. Everybody or virtually everybody participating in those debates would accept that such uses are morally illegitimate and that they should be legally proscribed.

2.2.1.2.  Some varieties of prospective interrogational torture Much more contentious has been the matter of prospective uses of interrogational torture. Further distinctions are needed here, however. There is no serious dispute in present-day Western countries over the morality of the torture that is employed by tyrannical regimes to suppress the emergence of political dissidence. Though the use of torture for such a purpose is prospective, neither the torture nor the purpose would be viewed as legitimate by any respectable commentators in a liberal democracy. Moreover, the vast majority of commentators in Western nations would perceive interrogational torture as morally impermissible even when it is undertaken prospectively in pursuit of certain legitimate purposes. The disruption of criminal gangs is a manifestly legitimate end, for example, but most commentators in the West would not contend that the use of torture to gain information for the achievement of that end is morally legitimate. Virtually all philosophers and jurists in liberal democracies would concur that not every instance of prospective interrogational torture in furtherance of a legitimate aim is itself morally permissible. Everyone or virtually everyone would accept that some such instances are morally impermissible. Thus, to fathom the recent controversies over the morality of prospective interrogational torture, we need a further distinction—a distinction relating to the legitimate ends that might be pursued through the use of such torture. At the centre of those controversies, as an intellectual matter, has been the use of torture in extreme emergencies where information is desperately needed to avert impending calamities. Though the employment of torture by the United States and some other Western countries against suspected terrorists in recent years has not been confined to situations of imminent catastrophes, the endeavours to justify the employment of torture have indeed concentrated predominantly on such situations. What have come to be known as ticking-bomb scenarios are pervasively invoked by philosophers and jurists who submit that the use of torture as an effective means to ward off the occurrence of a disaster can be strongly

Varieties of Torture

59

permissible.18 In other words, the type of torture most frequently at issue in contemporary debates is calamity-averting interrogational torture. (Such torture is of course always prospective in its orientation.) In those debates, very few philosophers and jurists attempt to justify any other kind of torture.

2.2.1.3.  Extreme emergencies In a ticking-bomb scenario of the sort that is frequently invoked by proponents of interrogational torture, the calamity which the interrogators are seeking to avert is an atrocity that would take the lives of thousands or even millions of people. Kenneth Himma (2007, 237–8) furnishes an especially hyperbolic version of the ticking-bomb scenario when he envisages a plight in which ten hidden hydrogen bombs will murder nearly all the people in the ten largest metropolitan areas of the United States. Although most other supporters of calamity-averting interrogational torture propound thought-experiments that are less extravagant, the large majority of them conjure up situations in which numerous deaths are in prospect. In a ticking-bomb scenario, the only means of averting those numerous deaths—or the means that is far more likely to succeed than any available alternative—is the use of torture on a suspect in custody who possesses information that is crucial for the defusing of the bombs or for the removal of any other grave dangers that threaten so many lives. In most renderings of the scenario, the suspect in custody is not only possessed of catastrophe-preventing information but is also at least partly responsible for the creation of the dire predicament in which a catastrophe looms. Moreover, the authorities know that he is possessed of the information and that he is responsible for confronting them with a dreadful quandary. He is utterly unwilling to reveal what he knows, perhaps because of his hatred of the authorities and the society over which they preside, or perhaps because of his fear of his accomplices. He stonily declines to yield to his interrogators’ blandishments. The authorities have good grounds for believing that the suspect will not be able to withstand the administration of torture and that he will therefore divulge his information in time for them to act successfully. Such are the outlines of the typical ticking-bomb scenario. (Of course, such a scenario 18   For some of the numerous invocations of the ticking-bomb scenario, see Allhoff 2005; Allhoff 2012, chaps 5–8; Bagaric and Clarke 2005, 583–4; Belvisi 2009; Bobbitt 2008, 361–3, 365, 368, 391–2; Cohan 2007; Curzer 2006; Dershowitz 2002, chap. 4; Gross 2001, 102–5; Himma 2007, 237–8; Kershnar 2005, 224–5; Miller 2005, 182–7; Miller 2011, § 3; Pettit 1991, 234; Raviv 2004; Seidman 2005, 892; Steinhoff 2006; Twining and Twining 1973, 346–8. For some assessments of the ticking-bomb scenario that are sceptical to varying degrees, see Addicott 2003–4, 897–910; Arrigo 2004; Brecher 2007; Bufacchi and Arrigo 2006; Card 2010, 182–204; Chesterman 2008, 328–9; Coady 2008, 88–9; Coady 2011, § 6; Davis 2005; Frowe 2011, 196–205; Ginbar 2008; Golash 2007; Green and Ward 2009, 167; Gross 2004a, 1502–3; Gross 2010, 124–6; Ip 2009; Kreimer 2003, 305–6; Kreimer 2005, 216–17, 228–9; Kutz 2007a, 238–9, 256–7, 263–4, 273–6; Luban 2005; Luban 2009; Lukes 2005, 11–12; Meisels 2008a; Paskins 1976, 141–5; Rodin 2008; Scarry 1985, 352 n160; Scarry 2004; Scheppele 2005; Shue 1978, 141–3; Shue 2006; Soniewicka 2013; Strauss 2003, 265–74; Sung 2003; Van der Vyver 2003, 453–8; Waldron 2010a, 6–7, 217–21, 263–4; Walzer 1973, 166–8; Ward, Johnstone, and Clucas 2009; Wisnewski 2008a; Wisnewski 2010, chaps 5–6; Wisnewski and Emerick 2009, chap. 2; Wolfendale 2006.

60

Introduction II: What is Torture?

need not involve any explosives. For example, the release of deadly toxins into the air-circulation systems of several large skyscrapers—or into the water supply of a major city—could produce disastrous effects akin in quantity to those produced by the detonation of some large bombs.) Many commentators have trenchantly challenged the plausibility of the ticking-bomb scenario. Especially when the scenario is elaborated with the extravagance that Himma introduces into it, the doubts about the plausibility of the situation are amply justified. We shall return in later chapters to the queries about the fancifulness of ticking-bomb predicaments; there I shall maintain that the use of torture in such predicaments is morally wrong even if all the queries about fancifulness are somehow inapplicable. For now, however, we should note that the cruxes posed by extreme emergencies can arise in much more realistic settings. In particular, the fancifulness of one’s depiction of such an emergency tends to disappear if the matters at stake are represented on a more modest scale (Meisels 2008b, 77, 183). Let us consider the following scenario, modelled quite closely on the facts of the American case Leon v. Wainwright.19 Suppose that three or four thuggish criminals kidnap a young man and demand a substantial ransom from his family. They warn credibly that they will torture him to death if their demands are not fulfilled. (Perhaps they lop off one of his fingers and send it to the family to underscore the grim unflinchingness of their intentions.) One of the criminals then arranges to meet with the brother of the young man to collect the ransom. When the encounter takes place, the brother hands over the ransom—at which point the kidnapper pulls out a gun. Some policemen who are hiding nearby shoot the kidnapper in the arm to prevent him from firing his gun, and they proceed to apprehend him. Taking him into custody, they repeatedly offer him highly favourable treatment if he will straightaway reveal to them where the hostage is being held; they are worried that a significant delay in the criminal’s return to that location will induce his accomplices to murder the captive and flee. However, because the arrested criminal loathes the police and is afraid of lethal retaliation by his comrades or their friends if he betrays them, he remains contemptuously silent. After expostulating with the kidnapper several times, the police begin to choke him and to twist his arm behind his back until he divulges the whereabouts of the hostage and the other captors. Acting on the basis of this information, the police are able to save the kidnapped man and to apprehend the remaining criminals. Unlike the case of the multiple hydrogen bombs, the foregoing story of the kidnappers and the police is fully credible. My unfolding of that story has not drawn on any outlandish assumptions about the epistemic capacities of the police or about the feasibility of the threatened harm. On the contrary, quite realistic rather than preposterous are both (i) the policemen’s forming of correct beliefs about the situation and (ii) the likelihood of the occurrence of the threatened harm in the 19   This case from the early 1980s is discussed in Bobbitt 2008, 383–4; Cohan 2007, 1603–5; Dershowitz 2002, 124–5, 253 n31; Dershowitz 2003, 285–6; Dershowitz 2004, 279–80; Ginbar 2008, 319–20; Strauss 2003, 240–1.

Varieties of Torture

61

absence of rapid action by the police. Indeed, the strong resemblance between my scenario and the facts of Leon v. Wainwright is a manifest indicator of the credibility of the situation. In relevant respects, the scenario also resembles an Antipodean case recounted by Miller (2005, 182–3; 2011, § 3) and a much-discussed recent case in Germany.20 Hence, implausible though Himma’s thought-experiment concerning the ten hydrogen bombs undoubtedly is, we should hardly infer that extreme emergencies—in which torture could be an efficacious means of forestalling the occurrence of calamities—can never credibly materialize. Still, even when we are mulling over extreme emergencies on a believable scale, we should remain alert to the ease with which people can overestimate the need for torture or the likely efficacy of torture. For example, after reporting with approbation a scene from the film Dirty Harry in which the eponymous detective tortures a psychopath in order to discover the location of a girl whom the psychopath has abducted, Steinhoff declares: ‘In the next scene, the girl is saved’ (2006, 342). In fact, as is correctly remarked by Davis in an article that antedates Steinhoff’s essay, the girl in the movie is dead when she is eventually found (Davis 2005, 161). She was very likely dead at the time of the act of torture, and indeed the detective had earlier sensed that she was no longer alive. Given that Steinhoff has greatly overestimated the efficacy of the wielding of torture in a cinematic scene which he could watch at leisure, it will hardly be surprising if constables or other legal-governmental officials faced with the pressures of a desperate emergency are frequently prone to similarly exaggerated expectations. At any rate, the category of calamity-averting interrogational torture comprises not only the use of severely and deliberately painful techniques to gain information that can prevent massive catastrophes, but also the use of such techniques to gain information that can prevent terrible harm to lone individuals. Of course, the differences between the fending off of a wide-ranging disaster and the fending off of dire harm to a discrete individual are morally significant. We shall explore some of the concrete moral implications of those differences in the next few chapters. Nevertheless, the focus of the present discussion lies not on those concrete moral implications but instead on the general character of a type of torture. At any rate, the differences between the information-seeking use of torture to avert carnage on a vast scale and the information-seeking use of torture to avert some horrific outcome within a much smaller compass are differences of degree rather than of kind. In each case, the infliction of severe pain is deliberately undertaken to overcome somebody’s resistance and thereby to elicit information that can help the interrogators to ward off some appalling consequences. In that key respect, the sundry instances of calamity-averting interrogational torture belong together under the same heading.

20   For some discussions of the German case, see Bagaric and Clarke 2005, 589, 613; Bobbitt 2008, 391; Ginbar 2008, 320–1; Gross 2004a, 1502–3 n 74; Ip 2009, 54; Jessberger 2005; Moore 2007, 57–8; Simester 2008, 302–3, 306, 308; Soniewicka 2013, 206–10; Steinhoff 2009, 43–4, 53; Ward, Johnstone, and Clucas 2009, 1–3, 6–7.

62

Introduction II: What is Torture?

2.2.2.  Placatory torture Whereas calamity-averting interrogational torture has received abundant attention from philosophers and jurists and legal-governmental officials in recent times, placatory torture has seldom been discussed since Joel Feinberg’s somewhat simplistic reflections on it four decades ago (Feinberg 1973, 87). Furthermore, whereas calamity-averting interrogational torture is aimed at extracting information, placatory torture is aimed at appeasing some opponent(s). Nonetheless, despite those salient dissimilarities, we can profitably examine placatory torture in tandem with calamity-averting interrogational torture—since anyone who resorts to either of those kinds of torture is endeavouring thus to avert some disaster(s). That underlying objective is common to the two kinds of torture, even though they involve quite different ways of achieving it. Moreover, precisely because that underlying objective is the lodestar of each of those types of torture, consequentialists and some threshold deontologists who regard certain instances of calamity-averting interrogational torture as morally permissible are committed also to the proposition that certain instances of placatory torture are morally permissible. In § 2.1.4.1.2 I  have sketched a paradigmatic situation of placatory torture (which quite closely follows the situation recounted by Gewirth in his 1981 article). In that scenario, a band of terrorists credibly threaten to wield weapons of mass destruction against several major cities unless a conservative politician degrades himself by publicly torturing his own mother at some appointed time and place. If the politician capitulates to the terrorists’ demands, he will obviously not be trying to extract information or a confession from his mother. Rather, he will be trying to mollify the terrorists in order to induce them to refrain from wreaking havoc.21

2.2.2.1.  A matter of intentions My brief earlier discussion of this matter has highlighted the fact that the conservative politician will not intend to afflict his mother with any harm or suffering— either as an end in itself or as a means—even if he does resort to torturing her. He will desperately hope against hope that she will not experience any severe pain during his mistreatment of her, even though he knows that the probability of her experiencing such pain is overwhelmingly high. What should be added here is that even the terrorists might not intend to bring about the suffering of the mother either as an end or as a means. They might be utterly indifferent toward her suffering, and might not be dismayed if some bizarre anomaly in the functioning of her nervous system were effectively to shield her from undergoing any sensations of discomfort. The terrorists might be preoccupied solely with the debasement of the politician (and the debasement of the society in which he is an influential 21   My sole significant departure from Gewirth’s scenario lies in my assumption that the torture will be non-lethal or only incidentally lethal. In Gewirth’s scenario, the politician is specifically required to torture his mother to death. I could accommodate that feature of the situation envisaged by Gewirth—a feature also of the situation outlined in Feinberg 1973, 87—but my abandonment of it enables me to make more easily my point about recklessness (in § 2.2.2.1).

Varieties of Torture

63

figure), and they might not care whether his application of torturous techniques to his mother actually causes her any pain or not. By obliging him to subject his mother to mistreatment that is almost certain to cause her excruciating pain, they will have humiliated him and will have led him to sully his moral integrity. They might be fully satisfied with such a result even if the mother miraculously does not experience any agony. Insofar as the outlook of the terrorists is as I have just delineated it, their infliction of severe pain on the politician’s mother is extravagantly reckless rather than intentional. They knowingly act in a way that gives rise to an overwhelmingly high probability of her undergoing dire torment, but her torment is in their eyes a matter of indifference rather than something which they desire as a means or an end. It is not one of their objectives, either instrumentally or intrinsically, since their lone aim is to impose humiliation and degradation on the politician and on the society which he represents. The mother’s actual experiencing of agony is not something on which the success of their terroristic action depends. Yet, although the anguish of the mother is a product of extravagant recklessness rather than of a deliberate quest, the burning or whipping or beating of her which the terrorists callously demand is unmistakably an instance of torture. Hence, we should be wary of any definition which suggests that torture always consists in the deliberate inducement of severe pain or suffering. Most instances of torture do consist in deliberate efforts to induce such pain or suffering, but the possibility of instances of torture that stem instead from extreme recklessness should not be ruled out or overlooked definitionally. Placatory torture is one type of which there can be extremely reckless tokens. (We shall turn later in this chapter, in § 2.2.10, to some further instances of extravagantly reckless torture.)

2.2.2.2.  A certain commitment In the next few chapters, we shall examine the concrete moral bearings of placatory torture along with other types of torture. What should be noted at this stage, in line with what has already been remarked above, is that the consequentialists and threshold deontologists who argue for the moral permissibility of some instances of calamity-averting interrogational torture are obliged to accept that certain instances of placatory torture likewise are morally permissible. Because placatory torture is akin to calamity-averting interrogational torture in being aimed at forestalling the occurrence of horrendous carnage—despite the dissimilarities between the propitiatory approach of the former type of torture and the information-extracting approach of the latter—the general considerations that lead consequentialists and some threshold deontologists to ascribe moral legitimacy to certain instances of calamity-averting interrogational torture are applicable as well to certain instances of placatory torture. Like the extraction of information, placation can sometimes be the only means of avoiding a terrible calamity. To be sure, there are at least three morally pregnant differences between calamity-averting interrogational torture and placatory torture. However, a quick exploration of those differences will reveal that they are insufficient to

64

Introduction II: What is Torture?

keep consequentialists and some threshold deontologists from being committed to the proposition that many possible instances of placatory torture are morally permissible. 2.2.2.2.1.  A first difference First, it may well be that placatory torture is less likely to succeed in warding off catastrophes than is interrogational torture. As Gewirth observed, ‘terrorists who make such demands and issue such threats cannot be trusted to keep their word not to drop the bombs if the mother is tortured . . .; and even if they now do keep their word, acceding in this case would only lead to further escalated demands and threats’ (1981, 10). Considerations of this sort will obviously weigh heavily in the calculations of consequentialists and some threshold deontologists as they judge whether placatory torture is morally permissible and morally obligatory in any given circumstances. Nonetheless, the uncertainties surrounding the likelihood of success for interrogational torture are also formidable, as numerous recent writings on the topic have emphasized. In connection with any torture of either type in a desperate situation, many contingencies can thwart all efforts to avoid a disaster. Consequentialists have to take such contingencies into account as far as possible, but the contingencies are indeed contingencies that can vary considerably in their strength and elusiveness. The prospects of success for placatory torture in some conceivable circumstances can be higher than the prospects of success for interrogational torture in sundry other circumstances—and can, in the eyes of consequentialists, be high enough to justify the perpetration of placatory torture. (For example, if the terrorists have forborne from wreaking havoc in the past when their demands for placatory torture have been followed, and if they have wrought havoc whenever their demands have gone unfulfilled, there will be fairly solid grounds for placing credence in the earnestness of their current threats.) Whether or not those conceivable circumstances are ever actual, there are credibly possible worlds in which they would obtain. At least with reference to any of those credibly possible worlds, consequentialists and some threshold deontologists are committed to the verdict that placatory torture is sometimes morally permissible. 2.2.2.2.2.  A second difference Second, whereas the success of calamity-averting interrogational torture consists in the gleaning of information that leads to the foiling of terrorists’ designs, the success of placatory torture consists in the fulfilment of terrorists’ designs through capitulation to their behests. As is suggested in the quotation from Gewirth in § 2.2.2.2.1, moreover, any administration of placatory torture is apt to strengthen and embolden the group of terrorists who have been propitiated. Still, although the point about the probable strengthening and emboldening of the terrorists’ organization will weigh heavily in the calculations of consequentialists, that contingency is just one factor in those calculations. It will be balanced against the ghastliness of the consequences that are likely to ensue if no placatory torture is undertaken. Furthermore, there are conceivable circumstances in which the perpetration of such torture is not apt to fortify and invigorate the terrorists’ organization. (For

Varieties of Torture

65

example, the authorities might be closing in on the terrorists; although they will not be able to make any arrests in time to avert the threatened disaster if no placatory torture is undertaken, they will be able to apprehend the terrorists if some additional time is gained through the occurrence of the torture. Alternatively, the terrorists as religious fanatics might be overtly disposed to immolate themselves— in some sort of sacred rite—once they have succeeded in their objective of humiliating and degrading their hated enemies.) Again, the requisite circumstances might never be actual, but there are credibly possible worlds in which they would obtain. At least with reference to any of those credibly possible worlds, consequentialists and some threshold deontologists are committed to the verdict that placatory torture is sometimes morally permissible. 2.2.2.2.3.  A third difference Third, whereas a victim of calamity-averting interrogational torture is typically a malefactor (even if his or her guilt has not yet been proved in a criminal trial), a victim of placatory torture is typically an innocent person. However, this observation is doubly inadequate to keep consequentialists from being committed to the proposition that placatory torture is sometimes morally permissible—even if we ignore the obvious fact that interrogational torture can be mistakenly brought to bear on someone who is not in possession of any information that would help to prevent an atrocity. For one thing, the foregoing observation pertains to what is typically the case— rather than to what is always the case—in connection with each type of torture. An innocent person can become apprised of information that would be crucial for the averting of a disaster, and, in response to non-torturous exhortations, she can be too afraid to disclose that information (because of fears of retaliation against herself or her family). Conversely, the terrorists who demand the perpetration of placatory torture might call for it to be wielded against a nefarious member of some rival terrorist network. Indeed, they might call for the torture to be wielded against an erstwhile member of their own network who has defected after making arrangements for the very atrocity which the terrorists are now threatening to carry out. In short, even if we leave aside the possibility of outright mistakes, it is not true that all victims of calamity-averting interrogational torture are guilty. Similarly, it is not true that all victims of placatory torture are innocent. Consequently, the guilt/ innocence distinction does not neatly map onto the interrogation/placation distinction, and it therefore does not constitute any blanket moral difference between calamity-averting interrogational torture and placatory torture. Philosophers and jurists who ascribe moral permissibility to some instances of the former kind of torture are committed to ascribing moral permissibility also to some instances of the latter kind. What is even more important, the guilt/innocence distinction is not something to which consequentialists attach any intrinsic moral significance. That distinction bears on consequentialists’ calculations only insofar as it affects the probability of various outcomes in the future. For example, if the plying of torture against a perceivedly guilty person elicits far less consternation among members of the public

66

Introduction II: What is Torture?

than the plying of torture against a perceivedly innocent person, then pro tanto a consequentialist balance of considerations will be tilted far more strongly toward the moral permissibility of the former than toward the moral permissibility of the latter. A noteworthy feature of the preceding sentence, however, is my twofold use of ‘perceivedly’. In the eyes of a consequentialist, people’s widespread and tenacious beliefs about guilt or innocence trump the underlying realities (Kramer 2011, 45–57). If an innocent person subjected to placatory torture is widely and firmly thought to be guilty of terroristic atrocities—perhaps because of chicanery on the part of the governmental officials who subject him to the torture—then his actual innocence will carry virtually no weight in consequentialist calculations about the moral legitimacy of the way in which he has been treated. Moreover, even when his underlying innocence is widely known, and even when the employment of torture against him will therefore be much more distressing for most members of the public, the additional distress is just one factor to be entered into a consequentialist assessment of the torture. In such an assessment, that factor can certainly be outweighed by the pressing need to avert a cataclysm. When it is outweighed, and when the balance of other consequentialist considerations is likewise in favour of the perpetration of torture, consequentialists and some threshold deontologists will be committed to the verdict that the use of placatory torture in the specified circumstances is morally permissible as well as morally obligatory. 2.2.2.2.4.  A matter of importance Although this chapter is concerned with the general nature of torture rather than with torture’s concrete moral bearings, these reflections on the moral status ascribed by consequentialists and some threshold deontologists to placatory torture have helped to underscore the importance of this type of torture. On the one hand, placatory torture is very rare; it occurs chiefly in the thought-experiments of moral philosophers. On the other hand, despite its rarity, it is important because the features which it shares with interrogational torture—in particular, its role in fending off disasters—will have morally assimilated some of its instances to the instances of interrogational torture that are viewed by consequentialists and some threshold deontologists as morally permissible. Believing that interrogational torture is strongly justified when it is necessary for the avoidance of horrendous calamities, consequentialists and some threshold deontologists have committed themselves to a comparable conclusion about placatory torture. Morally significant differences between calamity-averting interrogational torture and placatory torture do not relieve consequentialists and some threshold deontologists of their commitment to such a conclusion. When philosophers succumb to the temptation to deem interrogational torture morally permissible in veritable ticking-bomb situations, they are plighting themselves to an array of morally objectionable positions.

2.2.3.  Intimidatory torture Intimidatory torture is undertaken for the purpose of frightening the victim or various other people into adopting certain submissive modes of conduct thereafter.

Varieties of Torture

67

Some instances of torture are purely intimidatory, but many instances of intimidatory torture also serve other purposes such as interrogation or punishment or incapacitation or sadistic gratification. Nonetheless, although the purpose of intimidation is combinable with each of those other purposes, it is distinct from each of them (Bobbitt 2008, 373–6; Bufacchi and Arrigo 2006, 360; Gross 2010, 123, 125; Kreimer 2003, 300–1; McMahan 2006, 242; Meisels 2008b, 167; Scarry 1985, 329–30 n7; Shue 1978, 131–4; Twining and Twining 1973, 340 n79). There is no relationship of entailment in either direction between intimidation and any of those other aims. Intimidatory torture is one of the most common modes of torture under repressive regimes. When it is meant to frighten the direct victim into modifying his future behaviour, it is never intentionally lethal. However, this victim-specific intimidatory torture is often especially brutal, and it can quite easily turn out to be lethal. Still, the distinctive objective of such torture is to cow rather than kill the victim. Broadly addressed intimidatory torture is aimed at frightening people other than the direct victim(s). Some instances of this second variety of intimidatory torture are also victim-specific—that is, they are aimed at cowing the direct victims as well as some broader groups of people—but these two main varieties of intimidatory torture are disjoinable. When broadly addressed intimidatory torture is not also victim-specific, it is typically lethal. Littering the streets or fields with the bodies of people who have manifestly been tortured to death is one of the nastiest ways in which the dictatorships in Latin America and elsewhere have exerted their power to terrify members of the public into abject compliance with the dictators’ mandates. In no small part, the vileness of such practices is what lies behind the strength of the revulsion felt in liberal-democratic nations toward the use of torture. One point of clarification should be noted here. As is recognized in the opening section of the United Nations Convention against Torture, interrogational torture is sometimes not wielded directly against a person P from whom information is sought; it is sometimes wielded directly instead against one or more of P’s close relatives. When interrogational torture proceeds in this fashion, it need not be intimidatory. That is, it need not be aimed at causing P to fear that—unless he divulges the information which his captors are seeking—he himself will next be subjected directly to the torturous techniques that are being brought to bear directly on his close relative(s). Instead, interrogational torture of this kind can be aimed simply at causing P such overwhelming anguish that he will capitulate and disclose what he knows. When the torture is indeed so oriented, it is purely interrogational rather than intimidatory. It subjects P to psychological torment rather than directly to physical torment, but the goal that motivates it is the extraction of information from him through the pressure of the psychological agony rather than through the pressure of fear. (Of course, as has already been suggested, those two types of pressure are conjoinable. Though torture can be purely interrogational or purely intimidatory, it can also instead be both interrogational and intimidatory.)

68

Introduction II: What is Torture?

2.2.4.  Extortionate torture Interrogational torture of the sort envisaged in the last paragraph—where P’s close relatives are subjected to brutal mistreatment to impel P to disclose some desired information—is itself a specimen of extortionate torture. Such torture is administered to somebody in order to induce someone else to adopt some mode of behaviour. The mode of behaviour in question might be the provision of information, but it can equally be something quite different such as the payment of a ransom. Suppose that some thugs abduct Joe’s mother and threaten to torture her savagely if Joe has not paid a hefty ransom by a certain time. When Joe discounts the threat and declines to pay the ransom, the thugs film themselves as they administer electric shocks to the mother and as they amputate a couple of her fingers without any anaesthetic. Sending Joe the recording of their atrocities, they warn him that their persecution of his mother will intensify if he continues to resist their behests. In these circumstances, the extortionate torture practised by the thugs is not interrogational. Just as not every instance of interrogational torture is extortionate—in that some such torture (indeed, most such torture) is applied directly to people possessed of information rather than to relatives or friends of those people—not every instance of extortionate torture is interrogational. Some extortionate torture is undertaken to obtain things other than information. Note, moreover, that not every instance of extortionate torture involves the mistreatment of close relatives or friends of the people from whom various things are sought. For example, suppose that some terrorists manage to abduct a political leader (such as Aldo Moro in Italy), and suppose that they film themselves as they torture the leader. Sending some recordings of their flagitious actions to sundry newspapers and broadcasters, they warn that they will inflict even worse tortures if their demands for the release of quite a few of their comrades from prison are not met. Those demands are addressed not primarily to the close relatives and friends of the leader, but instead to the people in charge of the legal and governmental institutions of the country concerned—and also to the country’s citizens at large, who will undoubtedly be dispirited by the humiliatingness and ferocity of the terrorists’ outrages. Such outrages are paradigmatic instances of extortionate torture, even though the close relatives and friends of the political leader are not the primary addressees of the dictates issued by the terrorists.

2.2.5.  Act-impelling torture Several varieties of torture are aimed at impelling victims to perform certain actions (such as the disclosure of information). However, the torture distinctively designated here as ‘act-impelling’ is aimed at inducing a victim to perform some action other than the disclosure of information, where the victim undertakes the action in order to gain immediate relief from the intense pain that has been directly caused to her by the torture. Whereas intimidatory torture achieves its effects through the evocation of fear, and whereas extortionate torture achieves its effects through the triggering of solicitude, act-impelling torture achieves its effects by eliciting in

Varieties of Torture

69

the victim an overpowering desire for a terminus to the harrowing pain that is being directly inflicted on her. To escape from any further anguish, the victim immediately carries out some action (other than the imparting of information) that is demanded of her. Let us consider a slightly modified version of an example offered by Kamm (2011, 13). Suppose that Melvin has activated the detonator of a bomb, which is timed to explode in several minutes. The explosion will slay some hostages whom he has tied up and placed in the vicinity of the bomb. Anti-terrorist police are aware of the situation and are indeed in communication with Melvin, but they are too far away to be able to reach the scene in time. However, they possess a special electronic device that can induce severe pain in any targeted person over long distances. Urging Melvin to press a button that will inactivate the bomb’s detonator, they warn him that they will subject him to excruciating pain if he does not heed their admonitions. (Killing him by shooting him would of course be highly counter-productive—even if it were feasible—since his pressing of the button is the only means by which the explosion can be averted.) When Melvin declines to comply with their behests, they employ the electronic device to afflict him with terrible pain. Writhing on the ground from his agony, he agrees to push the button and does in fact do so when the operation of the electronic device is halted. In these circumstances, the police have resorted to act-impelling torture; through their use of torture, they have forced someone to perform a specified action (other than the divulgence of information) as a means of obtaining relief from his anguish. My designation of this variety of torture as ‘act-impelling’ is hardly incidental. Later in this chapter, in § 2.2.11, we shall examine a kind of torture—incapacitative torture—that is aimed at producing omissions rather than actions. That later section will expound in detail a rigorous distinction between acts and omissions, so that we can differentiate between interrogational torture or act-impelling torture on the one hand and incapacitative torture on the other. For the moment, a pre-theoretical understanding of the difference between acts and omissions will suffice. Under any plausible conception of that difference, Melvin’s pressing of the button (or some relevantly similar instance of conduct) is an action.

2.2.6.  Punitive torture In my discussion of Davis on torture as the testing of a victim’s capacity to endure pain, in § 2.1.4.1.2, we have considered the matter of punitive torture. Hence, my account of it here will be very brief. When torture imposed as a punishment for wrongdoing is open-ended, it will not be discontinued until the victim has died or until the victim has been reduced to gibbering idiocy or until the torturer exercises his discretion and decides that the tribulations of the victim have been sufficient. When torture imposed as a punishment is of a fixed quantity, the specified course of torture will be implemented in full unless it is interrupted or unless the victim has died in the meantime. (In my rejoinders to Davis, I have argued that a fixed

70

Introduction II: What is Torture?

course of severely painful punitive abuse—such as whipping, mutilation without any anaesthetic, caning, or burning—is torture.) Though torture can be purely punitive in character, the objective of retribution can also be pursued in conjunction with some of the other principal objectives which people might seek to attain through torture. For example, punitive torture is typically intimidatory as well. When torturers pursue punishment and interrogation together, the retribution in question is often imposed before the victim has been found guilty in a criminal trial. For instance, long before the occurrence of any formal trial, Khalid Sheikh Mohammed was subjected to waterboarding by his American captors nearly 200 times. Though the numerous instances of waterboarding were undoubtedly conducted in part for interrogational purposes, one gains the strong impression that some of those instances were undertaken to punish a vicious mass-murderer at least as much as to elicit information from him.22 If such an impression is correct, the use of torture against Mohammed for punitive purposes occurred before he had been formally tried and convicted. The point just made should be distinguished from an objection famously raised by the great eighteenth-century Italian jurist Cesare Beccaria against the employment of torture for interrogational purposes. Training his ire especially on the judicially prescribed use of torture for the extraction of confessions from suspects in medieval Europe, Beccaria (1995, 39) wrote as follows: ‘No man may be called guilty before the judge has reached his verdict; nor may society withdraw its protection from him until it has been determined that he has broken the terms of the compact by which that protection was extended to him. By what right, then, except that of force, does the judge have the authority to inflict punishment on a citizen while there is doubt about whether he is guilty or innocent?’ Bentham, who drew heavily upon Beccaria’s discussion of torture at many junctures, concurred with his perception of confession-seeking interrogational torture as a punishment that is levied before any trial and conviction: ‘If the fact of [a suspect’s] guilt were certain, such a proof [that is, a confession extracted through torture] would be unnecessary. If not, there was no sufficient ground for subjecting [the suspect] to so severe a punishment.’23 Beccaria’s substantive point is reaffirmed in the modern day by Scarry (1985, 41), albeit without any reference to Beccaria or Bentham:  ‘In its basic outlines, [interrogational] torture is the inversion of the trial, a reversal of cause and effect. While the one studies evidence that may lead to punishment, the other uses punishment to generate the evidence.’ Beccaria made some incisive observations in his discussion of torture, to which we shall return in Chapter 3. Moreover, he amply deserved plaudits for the moral force of his indignation in response to the grotesque system of torture that prevailed in medieval Europe. Nevertheless, his characterization of confession-seeking interrogational torture as a form of punishment—a characterization echoed by   For a related point, see Luban 2009, 190–1.   Twining and Twining 1973, 326. This passage appears in § 22 of Bentham’s manuscript ‘Of Compulsion and Herein of Torture’, which was published for the first time in the article by the Twinings. 22 23

Varieties of Torture

71

Bentham and Scarry—is simplistic. Whereas my remarks in the penultimate paragraph above have made clear that the punitive character of torture depends on the purpose for which the torture is wielded, Beccaria ascribed a punitive role to confession-seeking interrogational torture simply because of its nature as a hardship that is imposed on someone who is suspected of having committed a crime. Beccaria went astray on that point. If the hardship of torture is imposed solely for the purpose of educing a confession (or some other information relating to an alleged crime), then it is not properly classifiable as a punishment. Torture employed for such a purpose is of course strongly objectionable, but its objectionableness does not stem from its amounting to a punishment that has been levied before the stage at which a punishment of some sort would be appropriate. (Its objectionableness stems instead from the type of hardship that it is, and frequently also from its gratuitousness.) After all, if some person is arrested and detained before being placed on trial, the pre-trial detention is a significant hardship. Nonetheless, if there are solid grounds for suspecting that the person has committed a serious crime, and if the pre-trial detention is necessary for the integrity of an investigation—not least to ensure the safety of witnesses and victims—the detention is not properly classifiable as a punishment, even though the time spent as a detainee might in some jurisdictions be deducted from the length of any term of imprisonment to which the person is ultimately sentenced. The purpose underlying the detention is investigative rather than punitive. Much the same is true, mutatis mutandis, of confession-seeking interrogational torture that is perceived by its perpetrators as necessary for investigative purposes. Although such torture is odious whereas the pre-trial detention in the circumstances recounted here is morally legitimate (since it is not gratuitous and is not intrinsically repugnant), neither the torture nor the detention should be characterized as a punishment. Worth noting here is an objection to confession-seeking interrogational torture which superficially resembles Beccaria’s objection but which does not rely on the claim that such torture is inherently punitive. This alternative line of criticism is suggested by Seth Kreimer, though he advances it in articles that are contributing to debates over the moral legitimacy of calamity-averting interrogational torture; the line of criticism is in fact much more pertinent in the context of debates over the legitimacy of confession-seeking interrogational torture. (Admittedly, there are few if any debates of the latter sort in contemporary Western countries.) Kreimer writes as follows: ‘If there are physical sensations that cannot legitimately be inflicted on prisoners in retaliation for even the most heinous of crimes, presumably the state may not inflict these sensations on individuals whom it has not even prosecuted.’24 As a response to the employment of confession-seeking interrogational torture, Kreimer’s remark is a trenchant indictment that does not make the mistake of classifying such torture as punitive. Given that any punitive torture would be morally illegitimate, and given that the point of eliciting a confession from a suspect is to convict and punish her, the use of torture to extract a confession from a suspect   Kreimer 2003, 297. For a similar contention, see Kreimer 2005, 224 n124.

24

72

Introduction II: What is Torture?

is patently illegitimate. If someone cannot legitimately be tortured once she has been convicted through the verification of her guilt in a trial, then a fortiori she cannot legitimately be tortured for the very purpose of establishing that she is guilty. The point is not that the administration of confession-seeking interrogational torture would punish her before she has been convicted; the point, rather, is that the administration of such torture would consist in ascertaining her guilt or innocence by treating her in a way that would be morally impermissible as a way of treating someone whose guilt has been proved. In making that latter point, Kreimer avoids Beccaria’s error.

2.2.7.  Sadistic torture Torture is usually undertaken for ulterior purposes—when the victims are human beings—but sometimes the sole motivation behind the infliction of severe pain on human beings or animals is the sadistic gratification to be derived from their misery. Such a motivation indeed predominates among wanton children who torture animals. Though purely sadistic torture against human beings is less common, a quest for sadistic gratification often combines with other purposes. Insofar as torture does serve such a quest, the inducement of agony is pursued as an end in itself (even if it is also pursued for other reasons). For example, when the Moors Murderers Ian Brady and Myra Hindley tortured Lesley Ann Downey to death in 1964 in northern England, they did so for the sheer exhilaration of listening to her anguished and terrified screams while they exerted a god-like power to do with her as they pleased. In general, sadistic torture is one of the most nefarious modes of human conduct imaginable. However, we should here briefly re-examine one variety of this species of torture that is less seriously wrong than other such varieties. In §§ 2.1.4.1.4 and 2.1.4.1.5 we have pondered the nature of sado-masochistic torture—which differs not only from other kinds of sadistic torture but also from most kinds of non-sadistic torture, in that it is often consensual. As has been contended in my earlier discussions, sado-masochistic interaction is only a simulation of torture if the victim can bring it to a halt whenever he or she wishes to do so. However, if the parties to the abusive interaction have deliberately declined to arrange any signal whereby the victim can stop the mistreatment, or if a prearranged signal is disregarded by the perpetrator of the mistreatment when the victim indicates that he or she cannot endure any further suffering, the abuse of the victim does amount to genuine torture. Moreover, in a context in which the parties have deliberately declined to agree upon any signal for cessation, the torture is consensual even though it is sadistic. As a consequence, Chapter 3 will maintain that the torture perpetrated in such a context is much less seriously wrong than are other modes of sadistic torture. Nevertheless, even though the torture conducted in such a context is consensual and is therefore less gravely wrong than other modes of sadistic torture, it is indeed itself sadistic torture; it is undertaken for the pleasure of inflicting unbearable agony on some other human being(s).

Varieties of Torture

73

2.2.8.  Discriminatory torture As the UN Convention against Torture recognizes in its opening section, torture is sometimes practised for the purpose of expressing and consolidating the dominance of an ascendant group. Such a purpose can obviously be conjoined with other objectives—especially intimidation and punishment—but it is disseverable from those other objectives. In principle, the administration of torture can be simply a means whereby the subordinate status of a downtrodden group and the superior status of an elite group are reaffirmed. In practice, however, a discriminatory aim of that sort is almost always pursued in combination with some other aim(s) such as intimidation or punishment or sadistic gratification. As Kreimer observes (2003, 295–6; 2005, 210–11), slaveholders in some of the Southern states before the American Civil War were legally permitted to do anything to their slaves short of homicide. In particular, whipping and beating were the most common sanctions imposed on slaves for any behaviour that dissatisfied their masters. Against this legal background, those torturous punishments embodied and perpetuated the racial stratification of the antebellum United States. As Lawrence Friedman writes, ‘the law codified and expressed the . . . massive power of masters and mistresses, and the power of their agents and overseers. The concrete form of this power was the right to administer summary punishment or “correction.” In plain, blunt English, it was the power to beat, to hit, to flog, to whip’ (1993, 85). In the context recounted by Friedman, the law’s authorization of punitive torture against slaves was also an authorization of discriminatory torture. While the legal entitlement to inflict such torture was a badge of one’s membership in the dominant race, the legal susceptibility to being disciplined through such torture was a badge of one’s membership in the subjugated race. As Friedman remarks, ‘any white man (or woman) ranked higher in society than any black, slave or free, in the states of the slave South; and, at least under certain conditions, the law permitted any white to punish a slave who stepped out of line’ (1993, 86, emphasis in original). Torture in the slaveholding states was not only punitive but also integral to the system of racial castes that prevailed in those states.

2.2.9.  Humiliative torture Nearly every instance of torture is humiliating for the victim, as well as excruciating and terrifying.25 Discriminatory torture of the sort recounted just above is especially likely to be humiliating for anyone subjected to it. Still, even if all instances of discriminatory torture are humiliating and are intended to be humiliating, not every possible instance of torture aimed at humiliating the victim(s) is discriminatory along divisions of race or religion or ethnicity or gender or sexual 25   For some accounts of torture that highlight its humiliatingness, see Bufacchi and Arrigo 2006, 357; Davis 2005, 166; Luban 2005, 1431–3; Luban 2009, 194–5; Twining and Twining 1973, 354– 5. As is remarked in Kamenova 2009, 94, some of the Trial Chambers in the International Criminal Tribunal for the former Yugoslavia have recognized humiliation as a distinct purpose that can be pursued through torture.

74

Introduction II: What is Torture?

orientation. Nor does every instance of torture aimed at humiliating the victim(s) derive from a quest for sadistic gratification or intimidation or any of the other objectives examined so far. Thus, despite the very extensive overlap with other types of torture, we are best advised to consider humiliative torture as a distinct type. When repressive regimes employ torture against dissidents, they are typically seeking to intimidate the victims themselves and other malcontents. (They are also of course frequently pursuing other purposes such as the uncovering of information through interrogation.) However, in contexts that are not utterly fanciful, a repressive regime might employ torture against opponents to impair their effectiveness not principally through intimidation or physical incapacitation but instead principally through humiliation. Because the palpable use of torture against a dissident strikingly reasserts the dominance of the regime and the impotence of the dissident, the result can be the undermining of his inspirational authority in the eyes of others and perhaps in his own eyes. Quite credibly, then, torture can be perpetrated with such an upshot in view. In some credibly possible circumstances, the demoralizing effects of humiliation—on the person who directly undergoes the humiliation, and on his or her potential followers—are the effects for which torturers are chiefly striving when they practise their ugly trade. Although a humiliative purpose is rarely if ever pursued in isolation from other purposes that can be furthered through the use of torture, it can plausibly be the paramount objective of torturers in some situations.

2.2.10.  Extravagantly reckless torture Both heretofore and hereafter, this typology differentiates among species of torture by reference to the purposes for which the torture is undertaken. In the present section, however, we should focus not on the torturer’s purpose but on his general state of mind or mode of culpability. Of course, my discussion of sadistic torture has likewise focused on the torturer’s general state of mind. However, a quest for sadistic gratification is itself a purpose that can animate the perpetration of savage torture. By contrast, an extravagantly reckless state of mind is an accompaniment to some purpose(s) rather than itself a generator of purposiveness. My discussion of placatory torture in § 2.2.2.1 has pointed out that the terrorists who demand the commission of such torture on a particular occasion might not take the actual agony of the victim as one of their objectives. They might be intent solely on debasing the person who directly performs the torture (and on thus debasing the society in which he is a leader), and they might not care whether the victim actually experiences severe pain or not. If so, then their mode of culpability in relation to the victim’s pain is that of extreme recklessness rather than intentionality. In that respect, some possible instances of placatory torture differ from the instances of most other types of torture, which always involve the deliberate infliction of severe pain or suffering. Here we should take account of some further types of torture that can involve extravagant recklessness rather than deliberateness. Suppose that some middle-aged man kidnaps a female adolescent and keeps her locked at all times in a windowless room. At several junctures each day, he enters

Varieties of Torture

75

the room and rapes her. Suppose that his sole aim in raping her repeatedly is to satisfy his own concupiscence, and that he does not intend to afflict her with severe pain or suffering. Her undergoing of severe pain or suffering is not something that he seeks either as a means or as an end; it is instead something to which he is indifferent. Let us now vary the scenario. Suppose that a band of rebels overrun a village and repeatedly rape the female inhabitants who have not managed to flee. In so doing, the rebels are simply seeking to satisfy their lust. They are heartlessly indifferent toward the well-being of the women whom they rape multiple times; they are blithely unconcerned whether those women suffer excruciating physical and psychological torment or not. In these two scenarios, torture is occurring even though the perpetrators of the torture are not deliberately inflicting severe pain as a means or as an end. Their mode of culpability, in relation to the infliction of the severe pain, is brutal recklessness rather than intentionality. Although the two examples just propounded are both centred on rapes, the category of extravagantly reckless torture encompasses also some non-sexual iniquities. Suppose for instance that a doctor akin to Josef Mengele abducts some people— with the aid of some henchmen—and carries out hideous medical experiments on them without any anaesthetics or other palliatives. In declining to administer any anaesthetics, he is not deliberately seeking to make his victims suffer excruciating pain. Instead, he is simply opting to save the considerable funding and time that would have to be expended if he were to use anaesthetics. For him, the captives’ undergoing of unbearable agony is not serviceable either as a means or as an end; his grotesque medical investigations are not furthered by their experiencing of such agony. He is stonily indifferent to their anguish, but it is not one of his instrumental or intrinsic objectives. If a weird aberration in the neural functioning of any hostage had somehow spared her from serious pain, the doctor’s ends would not have been set back at all. Hence, his infliction of terrible pain on the hostages is demonically reckless rather than intentional. As these examples indicate, not every instance of torture is marked by the deliberate inducement of agonizing pain. Sometimes the inducement of such pain is outlandishly reckless rather than intentional. In other words, sometimes it is a thoroughly foreseeable side-effect of a torturer’s heinous actions—a side-effect to which the torturer is indifferent—rather than a deliberately pursued means or upshot. This distinction has been recognized by the judges on the International Criminal Tribunal for the former Yugoslavia (ICTY). However, because those judges are working with definitions of torture that deem it to be the intentional infliction of severe pain or suffering, they do not have the latitude to differentiate between extravagantly reckless torture and intentional torture. They instead frame the distinction in terms of ‘motivation’ versus ‘intention’. As one of those judges has recently written: There is an important distinction between ‘motivation’ and ‘intent.’ For instance, in torture of [a]‌sexual nature, the Appeals Chamber has held that even if the perpetrator’s motivation

76

Introduction II: What is Torture?

is entirely sexual, it does not follow that the perpetrator does not have the intent to commit an act of torture or that his conduct does not cause severe pain and suffering, whether physical or mental, since such pain or suffering is a likely and logical consequence of his conduct . . . In view of the definition, it is important to establish whether a perpetrator intended to act in a way which, in the normal course of events, would cause severe pain and suffering, whether physical or mental, to his victims. (Kamenova 2009, 94)

Although the motivation/intention dichotomy on which the ICTY’s judges rely is present in the law of various jurisdictions, a philosophical investigation of the nature of torture—in which we do not have to defer to the excessively restrictive definitions of the crime in international conventions—should rely instead on the distinction between extreme recklessness and intentionality. Instead of tailoring one’s account of torture to conform to a current legal definition, one should tailor one’s definition of torture to tally with one’s philosophical reflections on the issue. Given that certain conduct which amounts to torture does not consist in the deliberate evoking of severe pain as something desired instrumentally or intrinsically, and given that a definition of torture should be sufficiently capacious to encompass that conduct, such a definition should not imply that intentionality is the only mode of culpability that can suffuse the perpetration of torture. In some cases, extreme recklessness (in relation to the eliciting of severe pain) is another such mode of culpability.

2.2.11.  Incapacitative torture I have located incapacitative torture and edifying torture at the end of my typology because some possible instantiations of each of those two kinds of torture are morally legitimate. Notwithstanding that incapacitative torture in some of its manifestations is an especially odious phenomenon, some of its other manifestations are very different. To see as much, we need first to take account of the distinction between lastingly incapacitative torture and ephemerally incapacitative torture.

2.2.11.1.  Lastingly incapacitative torture As Davis remarks (2005, 164), oppressive regimes have sometimes used torture to enfeeble their opponents thoroughly without quite murdering them. By reducing dissidents permanently to inarticulate idiocy or to physical debility, a regime can rid itself of their troublesomeness without having to resort to mass murder. Of course, a regime whose officials stoop to such atrocities would undoubtedly not hesitate to commit mass murder if the interests of the regime would thereby be furthered. Moreover, the outright slaying of dissidents who have undergone torture can be—or can seem to be—a singularly effective way of concealing what has been done to them. For example, as Mark Osiel recounts (2004, 134), the soldiers who perpetrated torture under the Argentinian junta during the 1970s frequently then drugged the victims of the torture and dumped them from helicopters into the ocean. Those soldiers thus disposed of the principal evidence of their recourse to

Varieties of Torture

77

torture. Still, even from the perspective of a vicious regime, a programme of mass murder carries its own costs and drawbacks as well as its advantages. In some situations, such a regime can most efficiently deal with its opponents by using torture to incapacitate permanently their physical and/or mental faculties. Lastingly incapacitative torture is undertaken to deprive people permanently of their abilities to function physically or mentally. By reducing people to shells of their former selves, such torture inflicts pain that ruins their lives sweepingly and irreversibly. Without quite killing them, it effectively removes their capacities to conduct themselves as human beings. As will be argued in Chapter 3, and as is probably evident already, torture of this variety is always and everywhere morally wrongful. Even when practised against despicable terrorists who have themselves gloatingly tortured and murdered numerous people, such torture is a grave wrong that sullies the moral integrity of everyone involved in perpetrating it. Hence, if we are to come upon any examples of incapacitative torture that might be morally legitimate, we shall have to move to the other variety of such torture.

2.2.11.2.  Ephemerally incapacitative torture Ephemerally incapacitative torture consists in the deliberate infliction of severe pain in order to prevent someone from performing an action. The effects of ephemerally incapacitative torture are not permanent, or at any rate are not intended to be permanent. Rather, although the effects might persist for a while past the period during which the victim seeks to carry out some action, they will not last indefinitely—or, in any case, they are not intended to last indefinitely. Insofar as ephemerally incapacitative torture is undertaken to prevent legitimate or mildly wrong actions, it is a type of torture without any morally permissible tokens. However, ephemerally incapacitative torture can alternatively be undertaken to prevent seriously wrongful actions. When this type of torture is so directed, some possible instances of it are morally permissible. Because I am especially interested in ephemerally incapacitative torture as a type of which there can be legitimate tokens, my discussion of it will henceforth concentrate solely on instances of it that are aimed at averting seriously wrongful actions. Not all those instances are morally legitimate, but all morally legitimate applications of ephemerally incapacitative torture are among those instances. (Such instances could be labelled as ‘evil-averting ephemerally incapacitative torture’ to distinguish them from any applications of ephemerally incapacitative torture that are aimed at preventing permissible or mildly wrong actions. However, to avoid the cumbersomeness of the phrase just broached, I shall henceforth instead simply use ‘ephemerally incapacitative torture’ as shorthand for that more ungainly phrase.) Before we consider some examples of ephemerally incapacitative torture, we should ponder the distinction between actions and omissions. As will become apparent—and as has been signalled by my fourfold references to actions in the preceding paragraph—the act/omission distinction is crucial if we are to differentiate between some instances of ephemerally incapacitative torture and some instances of calamity-averting interrogational torture. Given that Chapter 3 will be

78

Introduction II: What is Torture?

maintaining that all instances of the latter type of torture are morally illegitimate whereas some instances of the former type of torture are morally legitimate, one’s ability to differentiate between the two types of torture is important morally as well as analytically. Pivotal for the endeavour of differentiation is the act/omission dichotomy, the explication of which is thus integral to my account of the nature and wrongness of torture. 2.2.11.2.1.  The act/omission dichotomy I originally propounded an analysis of the distinction between actions and omissions in a very different context, in a long book on freedom (Kramer 2003, 324–35). However, with some minor modifications, that analysis is straightforwardly transferable to the present discussion. It draws quite heavily on the work of Jonathan Bennett—1966, 94–7; 1981; 1988, 218–21; 1993; 1995, chaps 4–8—whose explorations of the act/omission divide are unsurpassed in contemporary philosophy for rigour and subtlety. There is no need for me here to reproduce all the layers and involutions of Bennett’s arguments and his responses to critics, but his fundamental line of thought will serve admirably as a point of departure for the exposition of ephemerally incapacitative torture in this chapter.26 (One caveat about my terminology should be entered at the outset. The term ‘omission’ is liable to mislead, since it might be taken to suggest an intentional refusal to act or a failure to act in the presence of a clear opportunity. Although the category of omissions as expounded below does encompass any deliberate refusals and any failures to take advantage of palpable opportunities, it extends further. It includes also every portion of the behaviour of a person whereby he or she does not perform any action of some specified type that is normally within his or her powers.) Refining some ideas put forward by Immanuel Kant and A. J. Ayer, Bennett maintains that the key to the act/omission distinction is the relative numerousness of the ways in which a person P can behave to achieve some result E. If E is attainable only by some sort of action on the part of P, then the different ways in which P can behave while bringing about E are a small proportion of all the ways in which P can behave—much smaller than if E is attainable by an omission on the part of P. Now, if we wish to invoke such a criterion for staking off acts from omissions, we manifestly must rely on some standard for measuring the ways in which P can behave. Unless there exists some metric that will in principle enable us to count the different courses of conduct in which P can engage, we shall not be able to make the comparisons required for distinguishing between acts and omissions. That is, we shall not be able to determine whether the number of ways in which P can behave while bringing about E is small or large as a proportion of the total number of ways in which P can behave.

26   Bennett generally employs the terminology of ‘positive’ versus ‘negative’ rather than of ‘act’ versus ‘omission’, because his discussion of this matter concentrates largely on propositions concerning human conduct rather than on human conduct itself. His orientation and terminology are apposite for his purposes, but I have altered them here.

Varieties of Torture

79

Bennett endeavours to supply the requisite metric (through elaborate arguments that will not be recounted here) with a focus on the sundry movements and the sundry stationary postures that can be adopted by a human body. A four-dimensional grid is necessary for the purpose. We come up with uniformly sized spatial and temporal units by dividing space analytically into finite regions and time analytically into finite stretches. We likewise have to divide objects analytically, most notably the bodies of human beings, into uniformly sized units of matter. (In opting for the size of each unit, we are opting for some level of concreteness or abstraction in our specifications of the movements and positions that can be predicated of a person’s body; whatever the level for which we opt, the spatial and temporal units in our specifications will not be infinitesimal.) With regard to any given person, we then have to ascertain the sets of spatial regions that can be occupied by her body and by concomitant objects during some specified span of time. Some of the available ways in which her body and concomitant objects can be positioned in the spatial regions over time are movements of various sorts, and some are instances of stationariness. Now, if with the use of such an analytical grid we consider all the possible movements and stationary positions among which some person P can select at a particular juncture, and if we consider further whether each of those movements and positions would lead to a certain result E on which P’s conduct has a bearing, we shall be able to distinguish pertinently between acts and omissions. If the movements and positions that would each lead to E are a very small proportion of the total array of movements and positions among which P can select at the specified juncture, then the bringing about of E by P is an action. If on the contrary the movements and positions that would each lead to E are a very large proportion of the total array of movements and positions among which P can select, then E is attainable by dint of P’s omitting to perform any action that is necessary to prevent it. In short, given that E occurs and that P’s conduct has had some bearing on its occurrence, P’s contribution to E is an action if most of the movements and positions available to P would not have led to E; and P’s contribution to E is an omission if most of the movements and positions available to P would have led to E. (Of course, P’s conduct will not have any bearing on the occurrence of myriad events and states of affairs. No act that could ever be performed by P would affect the continuation of the revolution of the planet Jupiter around the sun, for example. If no act within P’s power could ever help to promote or prevent the occurrence of E, then every element of his conduct is neither an act nor an omission in relation to E.) One of the great advantages of Bennett’s account of the act/omission divide is that it fully acknowledges a point which has sometimes been thought to cast doubt on that divide. That is, on the basis of Bennett’s approach, we can easily recognize that every mode of conduct classifiable as an omission is also an action. Typically, when P omits to do whatever is necessary to prevent or bring about E, he pari passu does something else that relates as an action to some other result. Indeed, save perhaps when P has been paralysed or rendered unconscious, he will always be acting as well as omitting to act in various ways. What Bennett’s exposition highlights is that the status of a course of conduct as an omission or an action is not an intrinsic

80

Introduction II: What is Torture?

property. Rather, that status resides in a relation among the specified course of conduct, other available courses of conduct, and some result(s) to which each course of conduct would or would not contribute. Because the aforementioned status is a complex relational property of this sort, and because every stretch of conduct stands in any number of the relevant relations, every stretch of conduct is an act in respect of some outcomes and an omission in respect of numerous others. As Bennett (1995, 86) writes: ‘[T]‌here could not possibly be conduct which was, in itself, negative: it is an error to try to divide items of conduct into those that are and those that are not negative acts or refrainings or forbearances or omissions, as though we had “. . . is negative” or “. . . is a refraining” as a monadic predicate of acts.’ Far from being called into question by the fact that every omission in relation to some E1 is an act in relation to some E2, my exposition of the act/omission distinction is reinforced by that very fact. 2.2.11.2.2. Fending off the most common objection to Bennett’s analysis Because the most commonly voiced objection to Bennett’s analysis of the act/ omission dichotomy has a direct bearing on my application of his analysis to the distinction between calamity-averting interrogational torture and ephemerally incapacitative torture, we should ponder that objection here. Recall that, while bodily movements are key physical components of manifold courses of conduct that are open to a person in various contexts, positions of stationariness are key physical components of other such courses of conduct. In most circumstances and in relation to most consequences, a stretch of conduct that qualifies under Bennett’s test as an action will involve some bodily movement(s). Nevertheless, there are some circumstances in which courses of conduct that qualify under his test as actions do not involve any such movements and instead consist in stark stationariness.27 Let us contemplate two scenarios, the first of which has been broached by Bennett himself and by several of his critics. Suppose that, if Hiram remains completely motionless in the sealed room where he is standing, a fine metallic dust in the air will settle upon the floor. Some of the dust will fall onto a tiny electronic device and will close a circuit, triggering an explosion that will release some hostages from a room where they have been detained. By contrast, if Hiram moves his body in any perceptible way, he will prevent the fine dust from settling and will thereby avert an explosion. In these circumstances, then, his remaining immobile is the only mode of behaviour or just about the only mode of behaviour by which he can set off an explosion near the other room. His remaining immobile will thus qualify as an action by which he can achieve that result. Contrariwise, if he engages in any of the countless perceptible movements of his body that are within his power, he will avert the explosion. Since those movements are huge in number, each of them will be classifiable as an 27   See Bennett 1981, 66–9; Bennett 1993, 82–5; Bennett 1995, 96–100, 112–14. For some of the critics of Bennett who have raised this point, see Dinello 1971; Locke 1982, 463; Quinn 1993, 157–8; Williams 1995, 335. Though Judith Lichtenberg (1994, 212–13) does not criticize Bennett at all, she presupposes the equivalence of the act/omission and movement/stationariness distinctions.

Varieties of Torture

81

omission to do what is necessary for the occurrence of an explosion. In sum, this scenario confronts us with a situation in which a posture of utter stationariness is an action and in which any movement of Hiram’s body is an omission. According to some critics of Bennett, we should be troubled by such an upshot. In the scenario just outlined—which I shall shortly endeavour to defuse—we are not told whether Hiram knows that an explosion will ensue if he keeps his body still, and we are likewise not told whether he views such a result as desirable (or whether he would view such a result as desirable if he knew of it). To be sure, neither of those aspects of the situation is strictly relevant to the classifiability of his stationariness as an action and the classifiability of any movement of his as an omission. Nonetheless, the scepticism of Bennett’s opponents might seem to be especially warranted if Hiram does not know the implications of his conduct. The complaints by those opponents might appear even stronger in respect of a second scenario (of my own invention), where the agency of the person involved is highly attenuated. Suppose that Margaret is walking along when an earthquake erupts around her. While remaining fully conscious, she freezes out of fear. As she stands transfixed, all the ground around her crumbles violently; nothing but the spot on which she stands is spared from the destruction. Had she not frozen into immobility, she would have been killed or at least badly injured. Only because she abruptly desisted from walking, is she on the lone piece of accessible land where she can be safe. Thus, in relation to the objective of remaining unhurt in the unusual circumstances, her motionlessness must apparently be classified as an action—whereas any continuation of her ambling would have been properly classifiable as an omission. Bennett’s approach to the act/omission distinction may seem to entail this conclusion, notwithstanding that Margaret has remained fully stationary out of stark fear rather than as a result of any decision on her part. How can Bennett’s approach be defended against the objection which these scenarios underscore? A small point to be noted initially is something mentioned at the outset of my consideration of that objection. That is, because most courses of conduct classifiable as actions under Bennett’s analysis do involve bodily movements, we are apt to be somewhat perplexed when a course of conduct so classifiable does not consist in any such movements. Being unaccustomed to such a phenomenon, we are naturally prone to find it peculiar. Much more important, the whole objection which we are exploring is based on the very error which Bennett’s analysis is designed to correct. In posing that objection, his critics wrongly presume that the property of being an action or of being an omission is intrinsic—like the property of being immobile—rather than relational. As Bennett (1981, 67) declares, the thesis underlying their complaints ‘is quite worthless because it relies on the notion of negativeness de re, negativeness as a monadic property of concrete actions and events’. In fact, as many of those critics themselves elsewhere emphasize, most omissions (as understood by Bennett or as understood in everyday discourse and legal discourse) are not instances of motionlessness.28 When a non-sleeping person omits to do something that is necessary to 28

  For example, this point is a central theme in Williams 1995.

82

Introduction II: What is Torture?

avert or attain some result, he or she is typically engaging in bodily movements to do something else; very seldom is he or she remaining immobile like a mummy. However, once we discern that stationariness is not an essential characteristic of omissions that distinguishes them from actions, we need to specify some other characteristic that does so distinguish them. We can scarcely content ourselves with saying that an omission consists in not doing anything that would have prevented a certain outcome. After all, any actions which in the prevailing circumstances are essential for the specified outcome will likewise consist in not doing anything that would have prevented it:  namely, in not doing anything that would have been inconsistent with the performance of those essential actions. We therefore have to look for a more complex relation that can form a line of demarcation between omissions and actions. Bennett’s articulation of a multifaceted relationship—the relationship between (1) any course of conduct for some person P that has some bearing on an upshot E and (2) all the other courses of conduct which are possible for P and which will each promote or not promote E—is a singularly powerful and pertinent solution to this crux. When the opponents of Bennett’s approach purport to discredit his line of thought by pointing out that some instances of behaviour classified under his analysis as omissions will consist in bodily movements and that some instances of behaviour classified under his analysis as actions will consist in strict motionlessness, they damagingly beg the question. Perhaps, as Bennett himself suggests, those opponents in their efforts to refute his approach are slipping back into thinking that a distinctive feature of every omission is the intrinsic property of stationariness. If so, they thereby beg the question against him by upbraiding him for departing from a non-relational conception of omissions (and actions) which his theory robustly rejects. In so doing, moreover, they fail to come to grips with the problem that has motivated the development of his theory. That problem resides in the fact that most omissions do not consist in stationariness; accordingly, stationariness cannot correctly be singled out as the property that distinguishes omissions from actions. Alternatively, the writers who assail Bennett are perhaps simply assuming that a position of immobility cannot be sufficiently ‘active’ to count as an action. Such an assumption on their part is once again a begging of the question, since it takes for granted that the status of an instance of conduct as an action or omission is a corollary of some intrinsic property. As a riposte to Bennett, a claim that all actions must involve bodily movements is no more effective than a claim that all omissions must involve the absence of bodily movements. In each case, the particular claim presupposes that the defining property of every act as an act or of every omission as an omission is non-relational. Any such claim thus begs the question against Bennett, instead of engaging with his arguments. Under his analysis, the retention of a posture of stationariness by some person P can perfectly well count as an action, since it can be just about the only mode of conduct available to P that would lead to some specified result. Furthermore, the notion that all actions must involve bodily movements is at odds even with a quotidian conception of actions. For example, someone who

Varieties of Torture

83

defiantly stands still after being ordered to move along is thereby performing an action. If a black person is told to relinquish her seat near the front of a racially segregated bus and to move toward the back of the vehicle, and if she resolutely remains motionless instead of stirring from her seat at all, she is engaging in an act of resistance to racial prejudice. Likewise, a man is performing an action when he remains firmly stationary for a long period as he presses a board against a hole in the side of a tank to prevent the spilling out of water through the hole. Numerous other comparable scenarios could be imagined. On this point about many instances of motionlessness as actions, and also on most other points, one’s endorsement of Bennett’s rigorous analysis is equally an endorsement of common sense. Let us now return to the scenarios with which I began this discussion of the objection concerning immobility. In particular, we should mull over the importance or unimportance of the outlook of the person involved in each scenario. Let us begin by pondering Hiram’s situation in the room with the descending dust. Will the status of his motionlessness as an action be affected at all by his knowledge or ignorance of the implications of his conduct? On the one hand, to be sure, the status of his conduct as an action is especially clear if Hiram while he remains unmoving is aware that he must so remain in order to bring about an explosion. Given such knowledge, his retention of an immobile posture—which might very well require great doggedness—is directly oriented toward the end which it promotes, the occurrence of an explosion. When the narrative of Hiram’s plight is embellished along these lines, any doubts felt by sceptics about classifying his stationariness as an action should be minimal. Likewise, their doubts about classifying any perceptible movement by Hiram as an omission should be minimal. If he does move, he will have omitted to do what he knows to be necessary for the attainment of an important objective. In that respect, his situation is similar to that of a studio model who is instructed by an artist to remain entirely still while a full-length portrait of her is being painted. If the model scratches her nose or shifts her legs, the irascible artist could quite intelligibly chastise her: ‘You have neglected to do what I told you to do. I instructed you to keep still. How can I work properly if you decline to play your part in the production of this painting?!’ Though such a rebuke can certainly be faulted for its petulance and pomposity, it should not strike anyone as a misapplication of the concept of omissions. On the other hand, despite what has just been said, the retention of a motionless posture by Hiram is an action even if he is ignorant of the full implications of his conduct. Perhaps he remains stationary simply because he is musing on some deep intellectual problem or is recalling some fond memories. Even so, he is behaving in just about the only way available to him that is promotive of the occurrence of an explosion. In relation to that explosive outcome, his stillness is an act, and any movement by him would have been an omission to do what is requisite. His ignorance of the full significance of his remaining immobile does not disqualify the immobility from being an action, any more than various bodily movements are disqualified from being actions when the person who engages in them is comparably ignorant. If Betty presses against a wall and thereby performs just about the only bodily movement that will lead to the disclosure of a secret passage within a

84

Introduction II: What is Torture?

house, her unattunedness to the implications of her act of pressing is fully consistent with the status of that act as such. Many actions are characterized by inadvertence, in that their chief effects have not been envisaged by the people who carry out the actions. In that respect, Hiram’s inadvertent triggering of an explosion (by remaining motionless) is no different from Betty’s inadvertent uncovering of the secret passage (by pressing against a certain part of the wall). If the latter bit of conduct should be classified as an action notwithstanding the unexpectedness of the consequences that flow from it—as it patently should—then the same is true of the former bit of conduct. My scenario involving Margaret and the earthquake is slightly more disconcerting, though not in any way that casts doubt upon the classifiability of some instances of stationariness as actions. Margaret’s terrified retention of an immobile posture is only problematically categorizable as an action, not because of the unmovingness of her stance but because of the attenuation of her agency. It is quite doubtful whether we can correctly describe her transfixed position as a product of her will at all; rather, it is something that has happened to her. Though an omission can simply happen to someone, an action must be volitional. However, these difficulties are not distinctively connected with the fact that Margaret has frozen into motionlessness. Parallel difficulties can surround various bodily movements. For example, in an early scene in the film Schindler’s List, a bulky German soldier bellows truculently at his Jewish victims when he orders them to march in a certain direction. Their first several steps, after he has screamed into their faces, are impelled as much by instinctive reflexes as by any conscious choices. Whether those steps should be classified as actions at all is therefore quite problematic. The dubiousness of the classification stems of course not from the fact that the steps are bodily movements, but from the fact that they are barely volitional. Much the same is true of Margaret’s petrified unmovingness, which saves her life. If her frozenness is only problematically classifiable as an act, the difficulty stems not from the nature of that frozenness as an instance of immobility but from its nature as a largely non-volitional reflex. 2.2.11.2.3.  Sussman’s example of the obese man Various critics of Bennett have broached other objections to his analysis of the act/ omission dichotomy, but I have rebutted those other objections elsewhere (Kramer 2003, 333–6), and in any event they have little or no bearing on the issues relating to torture that are my main concern in this chapter. By contrast, my responses to the queries about the motionlessness of some actions—which we have just been examining—are directly relevant to this chapter’s distinction between ephemerally incapacitative torture and calamity-averting interrogational torture. We can see as much in the course of scrutinizing some examples of ephemerally incapacitative torture that have appeared in the philosophical literature. One of the most thought-provoking examples has been propounded by Sussman in his discussion of the wrongness of torture. Though he is generally keen to highlight the wrong-making aspects of torture, he is rightly doubtful about Shue’s suggestion that a key to torture’s wrongness is the defencelessness and unthreateningness of the victim. Unlike Shue, Sussman distinguishes carefully between

Varieties of Torture

85

unthreateningness and defencelessness, and he maintains that a terrorist who resolutely withholds vital calamity-averting information in response to non-torturous methods of interrogation is posing a dire threat to everyone who is likely to be killed or injured in the calamity. (The threat posed by the terrorist will be especially stark if—unbeknownst to his captors—a high-powered bomb has been surgically implanted into his body with a detonator timed to explode at some point during his interrogation or shortly thereafter.) In order to highlight the threateningness of the terrorist in such a case, Sussman introduces an analogy: Consider a case where the police confront a very obese man who is trying to suffocate another by sitting on his chest. Like the terrorist, the fat man is defenseless before the police, who can wound or even kill him easily. Yet if the fat man dies or loses consciousness, the police lack the strength to shift him off his victim. Only the fat man can end his attack, even though he does not now need to do anything further in order for it to succeed. Here it seems that the police might well be justified in macing the fat man or twisting his arm (or threatening to kill him) in order to get him to derail a train of events that he has intentionally set in motion, even though he no longer has to actively contribute anything to it. The terrorist relies on a bomb’s mechanism to accomplish his goal, the fat man on his weight, and it is hard to see how this difference of method could be of any great moral significance.29

At the end of this quoted passage, Sussman adverts obliquely to the concrete moral significance of the differences between calamity-averting interrogational torture and ephemerally incapacitative torture. Chapter 3 will take up that concrete moral matter. For now, my focus lies on the aforementioned differences themselves. Sussman with his analogy has pointed to some of the chief similarities between those two varieties of torture. Someone who resorts to torturous measures of either type is aiming to thwart the perpetration of serious wrongdoing. He or she is deliberately inflicting severe pain for that very purpose. Moreover, provided that any calamity-averting interrogational torture is wielded directly against someone possessed of vital information (rather than against a close relative of such a person, for example), the direct victim of such torture is like the victim of ephemerally incapacitative torture in posing a serious threat to the basic well-being of others. Sussman is correct in drawing attention to these important similarities between the two kinds of torture. However, Sussman overlooks a crucial disanalogy between the use of torture to extract information from a bomb-planting terrorist and the use of torture to induce the obese man to lift himself off the unfortunate person on whom he is sitting. Whereas the interrogational torture is aimed at forcing the terrorist to perform the act of disclosing what he knows, the incapacitative torture is aimed at preventing the obese man from continuing to perform the act of smothering the person beneath him. In other words, the interrogational torture if successful will result in an action, whereas the incapacitative torture if successful will result in an omission. 29

  Sussman 2005, 16–17. For a short discussion of Sussman’s example, see Frowe 2011, 202.

86

Introduction II: What is Torture?

No doubt, this disanalogy between the two cases is easily overlooked because the omission by the obese man—which the incapacitative torture is designed to produce—will initially seem like an action to most people. After all, that omission consists in the obese man’s lifting of his body off the unfortunate person who is being suffocated underneath his immense bulk. However, when we keep in mind Bennett’s analysis of the act/omission distinction, we can readily recognize that the obese man’s continuing to sit is an action whereas his raising of his body off his squashed victim is an omission. Let us recall that the status of the obese man’s conduct as an act or omission is a complex property determined (i) by the relation between that conduct and the smothering of the victim and (ii) by the relations between that conduct and the other courses of behaviour that could be adopted by the obese man. Because very few courses of conduct open to the obese man will lead to the asphyxiation of the person beneath him, and because far more numerous courses of conduct open to him will not lead to such a result, his opting for one of the former courses of conduct would be an action—in relation to that result—and his opting for one of the latter courses of conduct would be an omission. (Recall that the instances of conduct are individuated as different ways in which the obese man’s body can be positioned or moved among the units of a spatial grid through some specified span of time. Given that the cubic units of the grid are not infinitesimally small, the ways in which the obese man’s body can be positioned or moved among them are not infinitely numerous. Hence, those ways can be counted and compared.) Contrary to initial appearances, then, the conduct which the police are trying to bring about by twisting the obese man’s arm is an omission. They are trying to prevent the man from completing the act of suffocating his victim. Consequently, the twisting of the arm is indeed an instance of ephemerally incapacitative torture; it is the deliberate infliction of severe pain in order to prevent the performance of an action that would amount to a very serious wrong. Because ephemerally incapacitative torture differs from calamity-averting interrogational torture in the respect just indicated, the conclusion drawn by Sussman at the end of the passage quoted above is less significant than he might think. On the one hand, he is right to contend that the difference between murdering somebody through the detonation of explosives and murdering somebody through the pressure of human corpulence is of little moral significance. On the other hand, he has erred if he thinks that that difference between a scenario involving the use of torture for the interrogation of a terrorist and a scenario involving the use of torture for rescuing the obese man’s victim is the only dissimilarity that might be of moral importance. Also potentially of such importance is the fact that the torture in the former scenario is aimed at producing an action whereas the torture in the latter scenario is aimed at forcing an omission. Chapter 3 will consider the moral import of that act/omission dissimilarity between the two situations. (Of course, another morally pregnant difference—cutting in the opposite direction—is that the terrorist seeks to murder far more people than does the obese man. However, as Chapter 3 will maintain, that factor bears on the question of the moral optimality of the interrogational torture rather than on the question of its moral impermissibility.)

Varieties of Torture

87

2.2.11.2.4.  Steinhoff’s example of the rapist Although Steinhoff strongly opposes any institutionalization of the practice of torture, he is equally keen to establish that the use of torture in certain extreme situations is morally permissible. When we plumb the examples of the practice that are regarded by him as morally legitimate, however, we find that some of them consist in ephemerally incapacitative torture. To be sure, Steinhoff also insists that many instances of calamity-averting interrogational torture are morally legitimate; we have already noted, in § 2.2.1.3, his favourable assessment of the use of torture by Dirty Harry to extract information from a murderous psychopath. Nonetheless, some of his scenarios that depict putatively or genuinely legitimate torture are focused on the infliction of severe pain that is aimed at stymieing the efforts of wrongdoers to perform very seriously immoral actions. One such scenario involves a rapist and his victim (Steinhoff 2009, 44). As the rapist assails his victim and begins to tear her clothes off, she manages to grab a pencil which she rams deeply into his ear two or three times in order to prevent him from proceeding further with his nefarious crime. Her objective is to cause him such searing pain that he will not be able to concentrate on what he is seeking to do to her. He will thus not be able to continue with his dreadful act of rape, as he involuntarily lets go of her. In other words, the victim’s plunging of a pencil into the rapist’s ear is a paradigmatic instance of ephemerally incapacitative torture. The general aim of her desperate counterattack is the hallmark of such torture. (We can vary Steinhoff’s example in any number of ways. For instance, suppose that the woman under attack is wearing shoes with stiletto heels. When the rapist seizes her and begins to rip her clothes off, she stamps frenziedly on his foot with her heel a few times in order to overwhelm him with pain so that he will relax his grip on her. Once again, her objective is that of somebody engaging in ephemerally incapacitative torture.) Steinhoff adduces the scenario of the rapist just after declaring that the police in the case of the German kidnapper—mentioned in § 2.2.1.3—would have been acting permissibly if they had resorted to torture against the kidnapper in order to learn from him the whereabouts of his juvenile hostage. As a moral matter, the real-life German case and the imagined case of the rapist are assimilable in Steinhoff’s eyes. However, malgré lui, Steinhoff in fact gestures toward the moral gap between those cases that will be explored in my next chapter. He pungently writes as follows (2009, 44): But, so it is often said by torture opponents, couldn’t the police have talked to the kidnapper longer? Maybe then he would have finally given up the information. Yes, maybe, but as already stated, if milder means—like talking—do not work in a self-defence situation, the defender is allowed to try harsher means. Besides, there are in fact cases where a rapist has been verbally persuaded by his victim to stop. However, hardly anyone would say to a rape victim: ‘Why did you ram the pencil into his ear after only 30 minutes of rape? Why didn’t you endure some more rape, maybe half an hour more? Maybe your begging would finally have worked.’

Steinhoff seeks here to discredit the notion that the police in Germany were under any moral obligation to remonstrate further with the kidnapper before

88

Introduction II: What is Torture?

availing themselves of torture (or of threats of torture). Because he believes that the German case and the scenario of the rapist are morally assimilable, and because he correctly believes that it would be outlandish to suggest that the victim of rape has breached a moral obligation by not putting up with her horrific ordeal for a longer period, he concludes that the German police were likewise not breaching any moral obligations by resorting to sincere threats of torture when they did. Moreover, as he points out, we cannot differentiate between the cases by maintaining that any prolongation of the rape would have added to the hellish suffering of the woman significantly more than any additional time expended on the interrogation of the kidnapper would have added to the suffering of the young hostage. Even if we leave aside how harrowing the terror suffered by any abducted child is, the kidnapped boy in the German case was suffering from more than terror. As Steinhoff remarks (2009, 44, emphasis in original): ‘[T]‌he situation faced by the police [during their interrogation of the kidnapper] is not one where the child is happily playing in a garden and would then, if the kidnapper does not give the required information, suddenly and peacefully die. While the kidnapper is not being tortured by the police, the child is being tortured, namely by the kidnapper. [The child] is suffocating in a box the kidnapper put it in.’ Steinhoff is correct in contending that the German case and the case of the rape are not distinguishable by reference to the intensity of the anguish which the respective victims have undergone. However, he is mistaken in presuming that the two cases are morally indistinguishable, for he fails to take account of some important differences between them. Most notable among those differences is that the German police availed themselves of threats of torture in order to pressure the kidnapper to perform an information-disclosing action, whereas the woman has resorted to torture against the rapist to prevent him from performing or completing an action. Another difference, of less significance but still of some significance, is that the person who employs torture in the case of the rape is a private individual whereas the people who employed the threat of torture in the German case were public officials entrusted with public-safety responsibilities. In this book’s subsequent chapters, we shall ponder the moral import of these divergences between the two cases. For now, we should simply note that there are indeed some such divergences and that any conclusion about the moral assimilability of the two cases is therefore premature. The difference to which Steinhoff himself points—the fact that the police in Germany were morally obligated to try other methods of interrogation, whereas the victim of the rape is morally entitled to inflict pain on her assailant straightaway—may well prove to harbinger the broader inassimilability of the cases on moral grounds. 2.2.11.2.5.  Steinhoff’s example of the snake bite While seeking to establish that calamity-averting interrogational torture can be morally permissible, Steinhoff adduces another example of ephemerally incapacitative torture. He limns the following scenario (2009, 45): Jeanette is in a jungle camp with Bob. A black mamba bites her. The poison will kill her in [a]‌few minutes if she does not immediately get the antidote from the refrigerator. Bob

Varieties of Torture

89

wants Jeanette dead and blocks the refrigerator. She draws her gun and threatens to shoot him if he does not go out of the way. He does not, and so she shoots him in the leg. If for some reason it were necessary to kill him in order to get him out of the way, she would even be allowed to do that. It is a clear case of self-defence.

Let us assume that, by shooting Bob in the leg, Jeanette is endeavouring to cause him so much pain that he will be unable to continue to obstruct her path to the refrigerator. Her act of shooting is thus correctly classifiable as an instance of torture. Moreover, it is ephemerally incapacitative torture. That is, Jeanette is striving to prevent the performance or continuation of an action that would amount to a very serious wrong. She is striving to prevent an action and induce an omission, rather than the other way around. To understand why Bob’s impeding of Jeanette’s progress toward the refrigerator is an action, we need to have recourse again to Bennett’s analysis of the act/omission distinction. On the one hand, Bob is largely or wholly motionless while he is blocking Jeanette’s path, and his getting out of her way would of course involve the movement of his body. On the other hand, in relation to the objective of keeping Jeanette from gaining access to the refrigerator, his blocking of her path is correctly classifiable as an action. In relation to that objective, his moving out of her path would be an omission to do what is necessary. Among the courses of conduct open to Bob, only a small proportion are such that they will deprive Jeanette of access to the refrigerator. By contrast, most of the courses of conduct open to him will leave her with the access that she needs. Hence, under Bennett’s criterion, the latter courses of conduct are omissions whereas the former courses of conduct are actions. Consequently, Jeanette is engaging in ephemerally incapacitative torture when she deliberately afflicts Bob with severe pain by shooting him in the leg. The pain is a means by which she prevents him from continuing to perform his deadly action of obstructing her way to the refrigerator. Yet, given that the moral bearings of ephemerally incapacitative torture can diverge from those of calamity-averting interrogational torture (in ways that will be explored later in this book), Steinhoff cannot safely invoke an example of the former type of torture to clinch a point about the moral status of the latter type of torture. Before we move on, we should note one other dubious feature of Steinhoff’s remarks about Jeanette’s plight. In the penultimate sentence of the quoted passage, Steinhoff indicates that Jeanette would be morally justified in killing Bob if such a measure were the sole tack by which his effort to kill her could be stymied. Steinhoff is correct on that point, but he has gone astray in thinking that the truth of his claim about the moral legitimacy of killing Bob is sufficient to ground a claim about the moral legitimacy of the use of torture. His reasoning is presumably along the following lines. Given that being killed is considerably more harmful than being shot in the leg, and given that the deliberate killing of Bob would be morally legitimate if such a step were the only way of thwarting his lethal conduct, we can conclude that the shooting of him in the leg is a fortiori morally legitimate (since there is not any less harmful means of foiling his deadly designs). So Steinhoff appears to have reasoned.

90

Introduction II: What is Torture?

Although the conclusion about the moral legitimacy of Jeanette’s shooting of Bob in the leg is true, the pattern of reasoning by which Steinhoff has reached that conclusion is suspect. Notwithstanding that intentional killings of people are sometimes morally permissible, and notwithstanding that being killed is more harmful for a person than being subjected to certain techniques of torture, we shall see in Chapter 3 that calamity-averting interrogational torture is always morally impermissible (even if it is sometimes morally optimal). Accordingly, the reasoning that underlies Steinhoff’s correct conclusion about the moral permissibility of shooting Bob in the leg is invalid. If his conclusion were instead about the permissibility of an instance of calamity-averting interrogational torture, it would be false—even though the reasoning which Steinhoff believes to be supportive of his actual conclusion would provide no less support (or apparent support) to such an altered conclusion. In other words, what is doing all the work in vindicating his claim about the permissibility of shooting Bob in the leg is the fact that the shooting is an instance of ephemerally incapacitative torture. Such torture can be morally legitimate in some circumstances, or so Chapter 3 will argue. In that key respect, ephemerally incapacitative torture differs markedly from calamity-averting interrogational torture. Because the reasoning about the harmfulness of being killed versus the harmfulness of being tortured does not in itself distinguish between those two types of torture, that line of reasoning does not genuinely support any inferences about the moral bearings of torture of either type. Hence, the truth of Steinhoff’s claim about the moral permissibility of shooting Bob in the leg has nothing to do with his reasoning about the moral permissibility of killing Bob. We shall return to this matter, at a general level and at greater length, in Chapters 3 and 5—where we shall find that it is squarely connected to the need for a perpetrator-focused perspective in one’s assessments of the moral status of torture. 2.2.11.2.6.  Kamm on torturing from a distance Kamm discusses several overlapping examples of ephemerally incapacitative torture and of act-impelling torture. (In § 2.2.5, we have plumbed a slightly modified version of one of her examples of act-impelling torture.) Each of her examples involves some Person A who is attempting to murder an innocent Person B. Kamm uses first-person plural pronouns to refer to some third parties who will seek to intervene to prevent Person A from carrying out his act of homicide. However, because those parties are far away from the scene of the crime, ‘[t]‌he only way that we can stop him is from a distance, by sending electric shocks through him for an hour, deliberately controlling and gradating the level of pain so that it inhibits his movement. This causes him great distress and temporarily incapacitates him until we can reach the pair’ (Kamm 2011, 7). In some of Kamm’s examples, including the one from which I have just quoted, the torture is incapacitative in that it precludes Person A from performing the murderous action which he is striving to perform. In her other related examples, the torture is act-impelling in that it spares Person B by forcing Person A to perform some action that will avert the demise of B. Although Kamm does not use my terminology, and although her discussion of torture does not supply any rigorous exposition of the distinction between acts and omissions—and although, unlike me, she believes

Varieties of Torture

91

that some instances of act-impelling torture as well as some instances of ephemerally incapacitative torture are morally permissible—she is attuned to the general nature of the difference between ephemerally incapacitative torture and interrogational or act-impelling torture. Kamm usefully elaborates a contrast between torture-wide and torture-narrow. Torture-wide involves the infliction of physical or psychological agony to disable a person from performing an action or to induce non-volitional movements of a person’s body such as twitching and babbling. Torture-narrow ‘involves physical or psychological distress to get someone to decide to do an act under his control, and so tries to affect his will and get him to perform an action’ (2011, 14). Though this contrast is quite closely related to some of my distinctions, it does not map very smoothly onto any of them. Kamm lumps together interrogational torture and act-impelling torture and some intimidatory torture and extortionate torture under the heading of ‘torture-narrow’. Under the heading of ‘torture-wide’ she places not only ephemerally incapacitative torture, but also torture (which might be sadistic or humiliative) that is aimed at stimulating the non-volitional spasming and jerking of a victim’s body. That ‘torture-wide’ heading presumably likewise extends to lastingly incapacitative torture, though Kamm does not specifically deal with it. Her wide/narrow dichotomy does not straightforwardly cover some of the other types of torture, such as punitive or discriminatory torture, that have been discussed in this chapter. Nor does it extend to omission-eliciting instances of intimidatory or extortionate torture or to most instances of sadistic or humiliative torture. Still, although the fit between her dichotomy and my typology is far from perfect, her wide/narrow contrast—along with her sundry examples which are built on that contrast—makes clear her awareness of the general nature of incapacitative torture. One subtle but important dissimilarity between Kamm’s analysis and my own approach relates to her apparent assumption—in the brief quotation above—that only torture-narrow ‘tries to affect [the victim’s] will’. Such an assumption is inapposite, for it does not track the act/omission distinction appropriately. Sometimes a technique of ephemerally incapacitative torture works by affecting the will of a victim. Let us recall Sussman’s example of the obese man. When the police torture the man by twisting his arm, they are seeking to induce him to undertake the volitional process of lifting his body off the person whom he has been smothering. Nonetheless, although that bit of conduct is something which the obese man wills to undertake, it is properly classifiable as an omission (in relation to the smothering of the poor person underneath him). He has omitted to persist in his action of suffocating the person who is squashed beneath him. That very omission is due to his having willed to get up in response to the pain that is being inflicted on him through the twisting of his arm. As the example of the corpulent man illustrates, then, ephemerally incapacitative torture sometimes produces its effects through the volitions of its victims. It does not always induce paralysis or an outright collapse. A successful instance of such torture prevents the performance of an action that would be a serious wrong, but the success can be attained by impelling the victim to will some bodily movement—or some position of motionlessness—that constitutes an omission to perform the action in question.

92

Introduction II: What is Torture?

2.2.11.2.7.  From Kamm to Kantians My disagreement with Kamm on the point just discussed is also to some extent a disagreement with most Kantians, for the act/omission dichotomy on which I draw to distinguish between ephemerally incapacitative torture and certain other types of torture does not map smoothly onto the Kantian division between coercion and non-coercive violence. That latter duality has been explicated illuminatingly by Barbara Herman in her major study of Kant on practical reason. Let us first consider the following passage (1993, 119): Although many violent acts are coercive . . . , it will not do to claim that killing is a limiting case of coercion. A coercive act aims at the control of a person’s will; killing does not (at least not of the will of the person killed). In killing, someone is prevented from doing anything at all, but he is not made to do something against his will. There is a significant difference between threatening pain or twisting your arm (or even threatening to kill) to keep you from joining the opposition party and killing you to achieve the same result. The coercive act looks to alter what will happen by controlling what an agent wills.

One thing to be noted straightaway is that the twisting of an arm described in this passage is not an instance of ephemerally incapacitative torture. Rather, it is an instance of victim-specific intimidatory torture. That is, the infliction of severe pain produces the designed effect not by making it impossible or unbearably agonizing for the victim to persist in the performance of some action now, but by frightening him into declining to perform some action in the future. Victim-specific intimidatory torture is thus rightly classified by Kantians as coercive. Less clear is how Herman and other Kantians would classify ephemerally incapacitative torture. Any successful instance of such torture prevents the performance of some action at the time of that instance, and many of the omissions brought about by such torture are not the products of the wills of the victims; those omissions are brought about willy-nilly. In each such case, the Kantians would presumably classify ephemerally incapacitative torture as non-coercive violence. However, some of the omissions brought about by ephemerally incapacitative torture—such as the obese man’s desisting from his suffocation of the person who is trapped under his immense girth—are indeed products of the wills of the victims. When the obese man stands up, he is moving volitionally. That volitional movement is properly classifiable as an omission (in relation to the smothering of the person beneath him), but it occurs because the twisting of the obese man’s arm has affected his will. Hence, any such instances of ephemerally incapacitative torture will presumably be classified by the Kantians as coercion. In that respect, those instances will be lumped together by the Kantians with victim-specific intimidatory torture. As is evident, then, the Kantian contrast between coercion and non-coercive violence does not correlate neatly with my act/omission dichotomy or with my related distinctions. Let us probe one further passage from Herman’s discussion of the matter (1993, 126): Here is where one might draw the line between violence and coercion . . . [C]‌oercion involves a more direct attack on agency than does any act of (mere) violence. Its intent is to

Varieties of Torture

93

subvert and control the will of another . . . [I]n willing maxims of . . . coercion to be universal laws, one wills a world in which one’s own desires and beliefs are to be available to all as means for their purposes. In willing that our desires or beliefs be so manipulated, we in effect give our agency to the agency of others (we become but one of many who determine what we will).. . Unlike coercion . . ., which involve[s]‌assault on the integrity of willing itself, the object of violent action is not the will but a person’s body. (Threats of violence and threats involving violence are other matters.) Coercion involves an attack on agency; violence, an attack on its conditions. Although violent actions usually prevent an agent from doing what he wills, they do not (they cannot) control willing.

In this passage, again, Herman emphasizes that non-coercive violence differs from coercion in that the latter—unlike the former—exerts control over the will of the person who is affected. Whatever may be the attractions of this Kantian distinction between violence and coercion, it is obviously not tantamount to my distinction between ephemerally incapacitative torture and any of several other types of torture (such as interrogational or act-impelling torture). My distinction is centred on the act/omission dichotomy as it has been explicated in this chapter. That latter dichotomy does not correspond to a division between attacks aimed at manipulating a person’s agency and attacks aimed at undermining the conditions of a person’s agency. In some cases, as in Sussman’s scenario of the obese man, ephemerally incapacitative torture averts a seriously wrongful action—and thus brings about an omission—by constraining the victim to undertake some volitional movements that amount to his abstention from that wrongful action. Very closely related to the foregoing difference between this book’s typology and Kantianism is a moral difference that will be highlighted in my next chapter. As is strongly suggested in the passage from Herman just above, and as is explicitly stated by her somewhat earlier in her book, Kantians are committed to the proposition that coercion is morally worse than non-coercive violence— ‘worse not in terms of consequences, but as forms of willing’ (1993, 118). Hence, since Kantians will classify some instances of ephemerally incapacitative torture as coercion, they are committed to the view that such instances of torture are morally worse than murder or than any torture that is not coercive. Both punitive torture and sadistic torture are frequently non-coercive in the Kantian sense. Hence, Kantians are committed to the proposition that the ephemerally incapacitative torture in Sussman’s scenario of the fat man is morally worse than many instances of sadistic or punitive torture. Chapter 3 will contend that the Kantian position on this moral matter is perverse. Whereas some instances of ephemerally incapacitative torture that are coercive in the Kantian sense are morally permissible, all instances of sadistic torture and punitive torture are heinously wrongful. By maintaining that many instances of the latter two types of torture are less gravely wrong than every coercive instance of ephemerally incapacitative torture, Kantians display moral confusion. As Chapter 3 will submit, their confusion on this point is one manifestation of the skewing that arises from their inordinate emphasis on the import of reflective agency.

94

Introduction II: What is Torture?

2.2.11.2.8.  Hill and the act/omission distinction Writing several years before the publication of Kamm’s discussion of torture, Daniel Hill anticipates some features of her analysis and examples. He rightly wants to distinguish between ephemerally incapacitative torture and calamity-averting interrogational torture (though without using my exact terminology). Near the outset of his article he juxtaposes two scenarios, one of which recounts an instance of ephemerally incapacitative torture and the other of which recounts an instance of calamity-averting interrogational torture. In a scenario that Hill labels as ‘Defense Case’, the police employ an electronic pain-inflicting device against a terrorist who is about to detonate a bomb. By using the device, the police seek to ‘cause no serious lasting harm but so much pain that the terrorist will be paralyzed and unable to detonate the bomb’ (2007, 396). The measure carried out by the police in Defense Case, then, is a paradigmatic instance of ephemerally incapacitative torture. In a second scenario, which Hill labels as ‘Interrogation Case’, some officials from the national-security services administer electric shocks to a captured terrorist in order to induce him to reveal the location of a massive bomb that he has planted with a timed detonator. With the shocks, the officials are ‘intending to cause no serious lasting harm but so much pain that he will say where the bomb is’ (2007, 396). In Interrogation Case, then, the measures plied by the officials from the national-security services are a paradigmatic instance of calamity-averting interrogational torture. So far, Hill’s examples and analyses are aligned with my own typology of torture. Although Hill regrettably conflates unthreateningness and defencelessness by suggesting that the terrorist in Interrogation Case has become unthreatening as a result of being helplessly in custody, that bit of confusion does not seriously detract from his account so far. He grasps that the key dissimilarity between Defense Case and Interrogation Case is that the torture in the former is employed to prevent a seriously wrongful action whereas the torture in the latter is employed to induce a calamity-averting action. However, Hill then goes astray by setting out to impugn the notion that ‘the difference between Defense and Interrogation is merely that in Defense the terrorist is not being caused to perform any action, whereas in Interrogation he is’. In an attempt to show that such a notion is simplistic, Hill introduces a case of morally legitimate torture—Reflex Case—where the victim is supposedly forced to perform an action. In Reflex Case, a terrorist is about to detonate a large bomb. To thwart his heinous misdeed, the police spray him with a substance that will cause his eyes to become severely inflamed. They are ‘intending to cause no serious lasting harm but so much pain in his eyes that he will instinctively start rubbing his eyes in an effort to relieve the pain, and so won’t be able to press the detonator’ (2007, 397). Provided that any feasible alternative tack for countering the commission of an atrocity by the terrorist would be either more injurious or significantly less likely to prove effective, the policemen’s use of the pungent spray in Reflex Case is morally legitimate. Precisely because Hill recognizes that the use of the spray is morally permissible, he thinks that Reflex Case reveals that the key difference between Defense Case and Interrogation Case does not reside in the distinction between acts and

Varieties of Torture

95

omissions. He maintains that in Reflex Case ‘the terrorist is being prevented from detonating the bomb by being caused to perform an action [namely, the rubbing of his eyes], unlike in Defense in which the terrorist had been prevented from performing an action (the action of detonation) without being caused to do anything’. Hill declares that he has recounted ‘a permissible case in which one does cause the terrorist to perform an action’ (2007, 397, emphases in original). Contrary to what Hill presumes, the use of the acrid spray in Reflex Case does not forcibly produce an action; it forcibly produces an omission through the prevention of an action. Hill of course realizes that the measure by the police has prevented the terrorist’s action of pressing the detonator, but he incorrectly thinks that the police have undertaken the prevention by forcing the terrorist to perform an action—the action of rubbing his eyes. Let us assume here that the movement involved in the rubbing of the eyes is volitional rather than purely reflexive, and that it therefore does count as an element of the terrorist’s conduct that could be an action. Even so, it is properly classifiable as an omission rather than as an action. Recall that, under Bennett’s criterion for the act/omission dichotomy, the status of any stretch of conduct as an act or an omission is a complex relational property. That status is determined by reference to some result that might be sought by the person whose conduct is in question. Here the person is of course the terrorist, and the result is the activating of the detonator. What matters to the police is the prevention of that result and thus the prevention of any act-token that would yield that result. What matters to them, then, is the inability of the terrorist to press the detonator. They need to prevent him from moving his hands and body in any of the relatively small number of ways that will activate the detonator, and they consequently need to ensure that he will move or position his hands and body in one of the relatively large number of ways that will not activate the detonator. They hope that their spraying of the irritant into the face of the terrorist will induce him to rub his eyes, but the rubbing of the eyes is important and desired only because it is one of the many modes of conduct that are inconsistent with his pressing the detonator. In other words, it is important and desired only because it is an omission. Suppose that, instead of causing the terrorist to rub his eyes, the spray will cause him to freeze into motionlessness or to flee or to lose consciousness or to fall backward or to grip the top of his head in despair. In each such case the police will have succeeded brilliantly in achieving their objective, for their objective is to bring about an omission rather than an action. Were the success of their calamity-averting efforts dependent on the terrorist’s rubbing of his eyes, the police would be seeking to bring about an action—since only a small proportion of the bodily movements and positionings available to the terrorist would consist in his rubbing his eyes. However, as has just been remarked, the fruitfulness of the calamity-averting endeavours of the police does not depend on the terrorist’s rubbing of his eyes. Those endeavours will succeed so long as the terrorist adopts any of the myriad bodily movements or positionings that do not involve his pressing the detonator. That is, the endeavours will succeed so long as the police can induce the terrorist to omit to press the detonator.

96

Introduction II: What is Torture?

In short, Hill has erred by maintaining—in the final two quotations above— that the prevention of the terrorist from detonating the bomb in Reflex Case occurs through his ‘being caused to perform an action’. On the contrary, the terrorist’s rubbing of his eyes is an omission. Of course, Reflex Case does genuinely differ from Defense Case. If the terrorist’s rubbing of his eyes is volitional rather than purely reflexive, then the omission in Reflex Case is volitional whereas the omission in Defense Case is non-volitional. However, as has been observed in my discussion of Kantianism in the preceding subsection, volitional movements or positionings of one’s body can be omissions. In relation to the pressing of the detonator, the terrorist’s volitional wiping of his eyes is indeed an omission. Consequently, the difference between Reflex Case and Defense Case does not map onto the act/ omission dichotomy; in each of those cases, what happens after the intervention of the police is an omission on the part of the terrorist. (The non-volitionality of the omission in Defense Case is of course no bar to its status as an omission. As was indicated near the outset of § 2.2.11.2.1, an omission need not be volitional. An omission can reside simply in the fact that somebody has not done something that is normally within his or her powers.) 2.2.11.2.9.  Hill and the act/omission distinction redux Hill introduces two further examples, and with them he displays a correct understanding of the distinction between actions and omissions. Nonetheless, as we shall see, a general weakness of his reflections on torture becomes apparent in his comments about those examples. My discussion of ephemerally incapacitative torture can fittingly conclude by drawing attention to that weakness. First, however, let us examine Hill’s two additional scenarios. Holding Case is a variant of Reflex Case. In it, a terrorist is pressing down on the detonator of a large bomb. Unless he keeps the detonator depressed for a minimum of ten seconds, the bomb will not explode. To prevent the terrorist from persisting in his action of holding down the detonator, the police shoot him in one of his feet. By adopting such a tactic, they are ‘intending to cause no serious lasting harm but so much pain that he will make a conscious decision to stop priming the bomb and turn his attention to nursing his foot’ (Hill 2007, 397). Deliberately inflicting severe pain on the terrorist, the police are thereby disrupting his ability to complete an action that would amount to a very serious wrong. Although the terrorist would be motionless or nearly motionless while performing that action, and although the stymieing of the action causes him to move vigorously, those movements are straightforwardly an omission; they are among the numerous movements and stationary positionings available to the terrorist that would not involve his pressing down on the detonator. Hence, the deliberate infliction of severe pain by the police is an instance of ephemerally incapacitative torture. Hill (2007, 398) is entirely correct when he states that ‘in Holding [Case] the terrorist is prevented from performing an action (holding the button down)’. Provided that there is not available to the police any less injurious means of keeping the terrorist from completing his nefarious action, their resort to ephemerally incapacitative torture is morally legitimate—or so Chapter 3 will argue.

Varieties of Torture

97

Importantly different from Holding Case is the more complicated Withholding Case. In the latter scenario, a terrorist has pressed down the detonator of a large bomb. Unless the detonator remains depressed for a minimum of three minutes, the bomb will not explode. However, the terrorist does not need to continue to apply pressure to the detonator; it will remain depressed automatically unless someone with the terrorist’s fingerprints pulls it up. Knowing how the detonator works, the police shoot the terrorist in one of his feet with the aim of inflicting upon him such severe pain that he will comply with their behests. They tell him that they will likewise shoot him in his other foot if he does not immediately pull up the detonator. Here the use of torture by the police is act-impelling rather than incapacitative. It is aimed at inducing an action rather than an omission. In relation to the averting of the bomb’s explosion, the pulling up of the detonator by the terrorist is the only mode of conduct available to him that will produce the specified result. Hence, his compliance with the constables’ commands is straightforwardly an action rather than an omission. Hill is correct in his analysis of this scenario just as in his analys­is of Holding Case. He rightly declares that, in Withholding Case, the terrorist is ‘caused to perform an action, the action of pulling the button up’ (2007, 398). Nevertheless, although Hill is to be commended for correctly characterizing the general nature of the conduct of the terrorist in each of the two cases just considered, his summation of the moral upshot of the chief difference between those cases is disconcerting. Although his conclusions are correct, the reasoning by which he reaches them—or, rather, the lack of any reasoning by which he reaches them—is dismaying. His moral verdict on the act/omission difference between Withholding Case and Holding Case is as follows: ‘It would seem that this [act/ omission difference] makes a moral difference. Intuitively, it is not permissible to inflict pain in Withholding in order to compel the terrorist to pull the button up, but intuitively it is permissible . . . for pain to be inflicted in Holding in order that the terrorist might be prevented from pushing the button down’ (2007, 398, emphasis in original). These ungrounded appeals to intuitions are echoed at quite a few other junctures in Hill’s article. Hill never supplies any arguments of moral principle to back up his intuitions. He is correct in thinking that the use of torture in Withholding Case is morally impermissible and that the use of torture in Holding Case is morally permissible, but intuitions are too wispy to substantiate such a position. After all, the moral import of the act/omission dichotomy is not transparent. It has to be spelled out through moral argumentation of the sort which my third chapter will seek to provide. Hill’s inordinate reliance on intuitions in drawing his moral conclusions is paralleled in his invocation of the act/omission distinction. Nowhere in his article does Hill make any effort to expound that distinction with philosophical rigour; instead, he remains content with an everyday understanding of the matter. (As we have seen, that everyday understanding leads him into error when he tries to specify the nature of the terrorist’s conduct in Reflex Case.) Just as my account of ephemerally incapacitative torture in this chapter has endeavoured to remedy that deficiency in Hill’s discussion, so Chapter 3 will endeavour to present the moral argumentation that is altogether missing from his intuition-strewn ruminations.

98

Introduction II: What is Torture?

2.2.12.  Edifying torture In this chapter’s earlier discussion of Davis’s definition of torture—especially in § 2.1.4.1.5—we have glanced at some varieties of edifying torture. Such torture is undertaken for the benefit of the person who is subjected to it. Though that general purpose is what unifies this category of torture, the more specific aims of the subtypes of edifying torture are diverse.

2.2.12.1.  Salvation-oriented torture As has been observed in this chapter’s earlier discussion, an especially common species of edifying torture in past centuries was the infliction of excruciating pain on heretics to induce them to save their souls by repudiating their unorthodox doctrines. Though the use of such torture was often undertaken insincerely as a means of leverage in struggles among religious/political factions, there is no reason to doubt that it was sometimes undertaken out of genuine though shuddersomely misguided solicitude for the souls of the people on whom the torture was brought to bear. In any event, such a purpose can undoubtedly underlie the perpetration of torture in some credibly possible settings. Insofar as torture is carried out for this purpose, it is aimed at furthering the perceived interests of the people who are subjected to it. Of course the perception of those interests that leads to the use of such torture is hideously warped, but the chief point here is that the perpetrators of salvation-oriented torture (if they are sincere) take themselves to be acting for the good of their victims. In that respect, the torture which they administer is edifying—not in the sense that it is truly beneficial for the victims, of course, but in the sense that its driving purpose is to benefit those very victims. Because this chapter’s typology differentiates among kinds of torture by reference to the aims of the perpetrators, salvation-oriented torture squarely belongs under the heading of ‘edifying torture’ even though it is always wickedly misconceived.

2.2.12.2.  Therapeutic torture As has been remarked in § 2.1.4.1.5, very painful medical care typically differs from torture in some salient respects. First, whereas the infliction of pain by a torturer is always deliberate or extravagantly reckless, the infliction of pain by someone carrying out a medical procedure is usually neither intentional nor reckless even if it is amply foreseeable. In any typical course of medical treatment, the causing of pain is an undesired side-effect rather than something desired as a means or an end. Second, whereas the administration of medical treatment is typically consensual, the administration of torture is typically non-consensual. Partly because the application of medical treatment without the consent of a patient (or without the consent of a patient’s parent or guardian) is generally unlawful in most jurisdictions, such treatment is rarely undertaken without such consent. Even in circumstances where the gaining of consent is not possible—because the patient is unconscious, for example—medical care will typically not be undertaken unless the people

Varieties of Torture

99

providing it can confidently presume that the patient would have consented if he or she had been able to decide on the matter. A third difference is closely related to the point just mentioned, and is of central importance here. Whereas people who administer painful medical treatment to a patient are typically thereby seeking to enhance his or her welfare, torturers very seldom engage in their cruel practice with the aim of promoting the well-being of their victims. Even when the objectives of torturers are not deeply evil, they are rarely focused on the good of the immediate victims. Painful medical care is typically therapeutic in its designs, whereas the infliction of pain by a torturer—in its impact on the direct victim—is typically destructive. To be sure, we have seen already in this chapter that some unusual situations lie outside these broad generalizations. The foregoing differences between medical care and torture obtain typically but not invariably. Still, as broad generalizations rather than as universally quantified propositions, the observations in the last few paragraphs are accurate. However, we should now take account of some varieties of torture that are conducted for medical or quasi-medical purposes; some of these varieties have already been mentioned in this chapter. 2.2.12.2.1.  Experimentational torture In my discussion of extravagantly reckless torture in § 2.2.10, we have considered an example in which a doctor similar to Mengele carries out quasi-medical experiments on people without using any anaesthetics or palliatives. In the example, the excruciating pain suffered by the victims is a fully foreseeable side-effect rather than a deliberately sought means or end. Even so, the flagitious experimentation by the doctor is properly classifiable as torture. Moreover, the element of extravagant recklessness in the example can be replaced here by deliberateness. Suppose that the Mengele-like doctor deliberately afflicts his victims with agonizing pain under various conditions, so that he can study the effects of the pain on their overall physical and mental faculties. For him, their intense pain is a means for the discovery of facts about the biological functioning of human beings. Perhaps he is sadistic—in which case the agony of his victims is also an end which he finds pleasurable in itself—or perhaps he is utterly heartless without actually deriving ghoulish pleasure from his awareness that he is making people suffer severely. In either case, his non-consensual experimentation is flagrantly torturous. Here the torture is edifying not in being aimed at benefiting the direct victims distinctively, but in being aimed at producing medical or quasi-medical knowledge that might be of benefit to others and to the direct victims themselves if they survive their ordeals. Naturally, the experimentational torture might fail even on its own dreadful terms. It might not produce any knowledge that can serve any useful medical purposes, and its only real effect might therefore be to rack its victims with unspeakable anguish. Nonetheless, because the experimentational torture is aimed at producing such knowledge, and because this chapter is differentiating among types of torture by reference to the general objectives of the torturers, the category of therapeutic torture can be stretched to include experimentational afflictions of

100

Introduction II: What is Torture?

the sort envisaged here. After all, as Chapter 3 will make clear, such a classification is hardly an ironclad indication that those afflictions are morally legitimate. On the contrary, the deliberate and non-consensual inducement of unbearable pain is egregiously illegitimate in the circumstances envisioned here. Of course, some medical experiments that involve the deliberate inducement of agony could be less odious than the ones contemplated so far. The doctor conducting the experiments might not be depravedly akin to Mengele, but might instead proceed only with the full and knowledgeable consent of the patients on whom the experimentation is performed. Perhaps the patients are masochists, or perhaps they are strongly altruistic, or perhaps they expect to benefit directly—medically or financially, for example—from the discoveries to which the experimentation is likely to lead. Whatever be their motivations for subjecting themselves to the infliction of excruciating pain, their tribulations should not be classified as torture unless they have no means of terminating the agony when their endurance has been stretched to the breaking point. If a patient can halt the experimentation and its effects whenever he cannot withstand any more, then his ordeal is not torture. It should not even be classified as torture that is morally legitimate. By contrast, if there is no means by which a patient can bring his ordeal to an end when he cannot bear any further pain, his whole ordeal is an instance of experimentational torture. It might be morally legitimate in some circumstances (if the consent of the patient is full and informed), but it is torturous. Whether or not the absence of any means of withdrawal from the experimentation and from its effects is foreseeable at the outset, the experimentation in the absence of such a means is torture. 2.2.12.2.2.  Aversion therapy In § 2.1.4.1.5, I have glancingly referred to Twining’s brief discussion of aversion therapy. During such therapy, a person P is somehow made to experience intense pain whenever he or she is in the presence of a certain deleterious thing (such as narcotics or alcohol or gambling) toward which he or she is strongly attracted. The goal of aversion therapy, of course, is to lessen or eliminate P’s fondness for the thing in question. By building up and entrenching an association in P’s mind between the harmful thing and the disagreeableness of severe pain—regardless of whether the association is fully conscious or at least partly subliminal—the therapy aims to cure P of a potentially fatal attraction. If such a process is torture, it is torture designed to promote the weal of the victim (Kershnar 2005, 237 n2). Like any experimentation that involves the deliberate inducement of severe pain, aversion therapy is torture if it is non-consensual or if there is no route by which the person undergoing the therapy can withdraw from it. On the one hand, the doctors or psychologists who administer the therapy non-consensually might be well-intentioned in that they are endeavouring gamely to further the interests of the victim. They might be helping him to overcome an addiction or a lack of self-control that is afflicting him grievously, and they might be correct in believing that he will never surmount his problem if his consenting to the therapy is treated as a precondition for their proceeding with it. Respect for his withholding of consent in such circumstances—circumstances in which his ability to undertake reflective deliberation is significantly enfeebled—might be seriously detrimental

Varieties of Torture

101

to his broader interests and might therefore be a way of leaving him trapped in a plight over which he can no longer exercise any meaningful control. Given the attenuation of his ability to make informed choices, and given the damage that he is wreaking on his own underlying interests, the doctors or psychologists might be doing him a disservice if they regard him as having a veto over the use of a technique that is essential for extricating him from his predicament. If they have tried any number of alternative methods to no avail, and if the aversion therapy is usually a very effective technique, then the doctors or psychologists might be acting in a morally permissible fashion when they proceed without the patient’s consent. Still, in the absence of such consent, the deliberate eliciting of severe pain as a means of cultivating a salutary aversion is torture. Benignly well-intentioned though the non-consensual administration of aversion therapy might be in some circumstances, its fostering of agony in a patient without his consent is torturous. If instead a course of aversion therapy is conducted with the consent of the patient on whom it is administered, and if he can withdraw from it temporarily or permanently in the event of finding it sustainedly unbearable, then it does not amount to torture. It of course resembles torture in its deliberate inducement of horrific pain, but the opportunities for withdrawal by the patient—together with the consensuality of the process and the benevolence of its orientation toward the patient’s well-being—are sufficient to render it non-torturous. Behavioural conditioning with those features should not be classified as torture, not even as morally legitimate torture. 2.2.12.2.3.  Averting a coma In § 2.1.4.1.5, I have briefly mentioned Kamm’s scenario in which some doctors subject a patient to intense pain in order to keep him from lapsing into a permanent coma. There are three main possibilities to be considered here. First, the need for the infliction of severe pain to avert a coma might be foreseeable, and the patient might consent in advance to that infliction. If so, this coma-preventing technique is not a form of torture. Though the patient does not have any meaningful opportunities for withdrawal from the process while it is occurring, the absence of such opportunities is not due to their being withheld from him but is instead due to his inability to exercise them. Given as much, and given that the eliciting of pain is consensual and is undertaken for the benign purpose of helping the patient, it does not amount to torture. Not only is it morally legitimate; it is also non-torturous, even though it involves the intentional inducement of terrible pain. Second, the need for the infliction of searing pain to prevent a coma might be foreseeable, and the patient might informedly decline to consent to the use of such a tactic. Though he obviously hopes to survive the serious operation that he will be undergoing, he feels that the prospect of being afflicted with dire pain (perhaps for a protracted period) is even more daunting than the prospect of becoming comatose or dying. Hence, he decides to take his chances by expressly instructing the doctors not to subject him to severe pain even if their so subjecting him is the only approach that can save him from slumping into a permanent coma. In these circumstances, the doctors will be acting against his explicit wishes if they do resort to pain-inducing methods in order to avert a coma. Given as much, the use of such

102

Introduction II: What is Torture?

methods would amount to torture. Though the doctors in such a scenario would doubtless be seeking to further what they perceive as the fundamental interests of the patient—in a high-handed fashion—the non-consensuality of their deliberate eliciting of intense pain, in circumstances where the furnishing or withholding of consent by the patient has been feasible, is sufficient to render their actions torturous. Torture practised in the interest (or perceived interest) of a victim is torture nonetheless. A final possibility is that the need for the inducement of intense pain to avert a coma is not reasonably foreseeable. Consequently, the patient in advance has neither provided nor withheld consent to such treatment. At the point when the infliction of pain becomes necessary to keep the patient from falling into a coma, of course, he is not capable of providing or withholding consent. Thus, if the doctors are to proceed with such an approach, they will have to do so without any authorization from the patient. If they do go ahead, the classifiability of their actions as torture will hinge on the matter of presumed consent. If there are no significant grounds for believing that the patient would be opposed to the use of pain-inducing methods to prevent the onset of a coma, and especially if there are grounds for believing that the patient would favour the use of such methods for that purpose, then the actions of the doctors do not amount to torture. They are warrantedly proceeding on the basis of presumed consent in circumstances where the securing of actual consent is wholly infeasible, and they are acting in furtherance of the vital interests of the person on whom they are inflicting great pain. By contrast, if there are any significant grounds for believing that the patient would be opposed to the infliction of great pain even for the purpose of keeping him from slipping into a permanent coma, the doctors will be engaging in torture if they resort to pain-inducing methods for that purpose. To be sure, their actions in such circumstances are not nearly as cavalier as those examined in the penultimate paragraph above, where someone’s express wishes have been disregarded by zealous doctors. Indeed, the pain-inducing actions in the situation envisaged here—actions that do not involve any defiance of someone’s express wishes—might not be morally blameworthy even though they amount to torture. Nevertheless, if there are significant grounds for believing that the patient would be opposed to the use of pain-inflicting methods even for the purpose of forestalling the onset of a coma, the doctors will be committing torture if they avail themselves of such methods. 2.2.12.2.4.  A worry about the conflation of distinct issues Some readers might worry that the foregoing discussion of therapeutic torture has run together two issues or sets of issues: the question whether certain types of conduct amount to torture, and the question whether those types of conduct are morally legitimate. After all, in addressing the former question, this discussion has focused on some morally pregnant factors such as the presence or absence of consent. Some readers might therefore wonder whether my reflections on what counts as torture have metamorphosed into reflections on the moral bearings of certain modes of conduct.

Varieties of Torture

103

On the one hand, as has been stated near the outset of this chapter, the matters addressed herein are not entirely neutral morally. For example, the very fact that torture consists in the deliberate or extravagantly reckless infliction of severe pain— and not in the negligent or innocently unintended infliction of such pain—is of patent moral significance. Because of that aspect of torture, any considerations that ever morally justify its occurrence must be especially potent. More generally, as will become clearer in my next few chapters, the factors that account for the status of some mode of conduct as torture are by no means wholly disjoinable from the factors that account for torture’s wrongness. Those sets of factors are not equivalent, but there is of course a substantial overlap between them. While we obviously cannot fathom the wrongness of torture without pinning down what torture is, we conversely cannot aptly pin down the nature of torture without fathoming why torture is wrong. The interrelatedness of those arrays of questions will emerge more saliently in the next few chapters. On the other hand, torture differs from murder in that a satisfactory definition of it will leave open the possibility that some tokens of the act-type are morally legitimate. Whereas murder (as opposed to killing) is always wrongful—even though the gravity of its wrongfulness is of many different degrees among its instances—some varieties of torture can be instantiated in ways that are not in breach of any moral obligations. To be sure, most varieties of torture are always morally wrong; the distinctive features of those varieties are sufficient to ensure that every one of their instantiations is violative of some major moral obligations. Nonetheless, there can be some morally permissible instantiations of the remaining few varieties of torture. Hence, any encapsulation of the general nature of torture should leave room for such instantiations. As is probably evident already, the analyses in this chapter do leave such room. For example, although the absence or presence of consent does in some contexts determine whether certain conduct is torturous, it does not perforce also determine whether the conduct in question is morally impermissible. Thus, although questions about the character of some conduct C as torture and questions about the moral bearings of C are intertwined in many respects, they are distinct. The fact that the intertwining has been prominent at some points in this chapter is not an indication that the distinctness has become effaced.

2.2.12.3.  Resistance training This chapter has referred at several junctures to the deliberate infliction of severe pain for the purpose of toughening the resistance of someone (usually a soldier or an intelligence official) to the use of torture thereafter. In §§ 2.1.4.1.4 and 2.1.4.1.5, we have pondered the conditions under which the infliction of pain for this purpose should be classified as torture. If the soldiers subjected to the resistance training have not consented to undergoing it—either because they have withheld their consent or because they have not been given any opportunity to furnish or withhold their consent—then the training is torturous. It is not merely a simulation of torture. Likewise, even when a soldier is fully

104

Introduction II: What is Torture?

entitled to embrace or decline the opportunity to undergo resistance training, the process will amount to torture if the soldier lacks any means of halting it when he has reached the limit of his endurance. Even if he has consented from the outset to the unavailability of any such means (because its unavailability makes the process of resistance training much more effective), the subjection of him to intense pain without any route for withdrawal or cessation is more than a simulation of torture. It is itself torturous—notwithstanding its consensuality and the benevolence of its orientation toward the soldier who is undergoing the pain. Albeit the resistance training is salutarily designed to promote the interests of each soldier who participates in it, it promotes his interests by hardening him through torture. Edifying and consensual though the process is, the person undergoing it is deliberately afflicted with terrible pain to an open-ended extent over which he cannot exercise any control. Edifying and consensual though the process is, therefore, it amounts to torture. Its status as torture is consistent with its moral permissibility.

2.2.13.  A table of the main kinds of torture Each of the main types of torture is listed in Table 2.1, which focuses on aspects of each type that are sustainedly probed in the next chapter. Table 2.1  The main kinds of torture Main types of torture

Can the purpose ever be morally legitimate?

Are some instances ever morally legitimate?

Interrogational Placatory Intimidatory Extortionate Act-impelling Punitive Sadistic Discriminatory Humiliative Extravagantly reckless Incapacitative Edifying

Yes Yes Yes Yes Yes Yes No No No No Yes Yes

No No No No No No No No No No Yes Yes

2.3.  Conclusion: An Overview This long chapter should now conclude by drawing together the lessons to be learned from its typology of torture and from my investigation of various definitions of torture. A  salutary consequence of going through the typology is that one becomes keenly alert to the perils of oversimplification in the formulation of a definition. Given the diversity of the kinds of torture which this chapter has

Conclusion: An Overview

105

explored, we can see how simplistic some of the prevailing juridical and philosophical definitions of torture are. We should be wary of any attempt to formulate pithily a set of sweepingly necessary and sufficient conditions for the status of any mode of conduct as torture. An adequate definition will include some necessary conditions, but will also include some conditions that are typically but not always fulfilled when a course of conduct amounts to torture.

2.3.1.  The infliction of severe pain Most obvious among the elements of a satisfactory definition is that torture (if uninterrupted) consists in the infliction of severe pain or suffering. Almost always the pain or suffering inflicted is intended as a means or as an end, though in exceptional cases—such as those recounted in § 2.2.10—the pain or suffering caused by torture is inflicted in an extravagantly reckless fashion rather than deliberately as a means or an end.

2.3.1.1.  Torture versus attempted torture One issue from which we can largely prescind here is the matter of distinguishing between torture and attempted torture. In many credibly possible scenarios, that distinction will be clear-cut. For example, if a would-be torturer connects an electrical-shock device to an outlet in the wall, and if the flow of electricity is abruptly cut off just as he is about to apply the device to the body of his victim, he has attempted to torture her without quite having succeeded in actually carrying out the torture. With regard to such a case, the division between torture and attempted torture is straightforward. In some other possible cases, however, that division is much more elusive. For example, suppose that the man with the electrical-shock device does manage to apply it to the body of his victim Veronica while the flow of electricity is running normally. The device delivers several powerful shocks to sensitive parts of her anatomy, but—because of some strange aberrations in the functioning of her nervous system—Veronica experiences little or no discomfort. Has the man perpetrated torture in these circumstances, or has he attempted unsuccessfully to perpetrate torture? Let us generalize this question as follows: Experience of Pain Inquiry. Given that someone intentionally (or in an extravagantly reckless fashion) takes certain steps against a victim’s body that would almost always be sufficient to induce feelings of severe pain, do those steps amount to torture only if the victim actually undergoes such feelings, or can they amount to torture even if the victim does not experience any great discomfort?

2.3.1.1.1.  A different question Note that the Experience of Pain Inquiry differs from the following question: Achieved Purpose Inquiry. Does torture occur only if a torturer succeeds in achieving the purpose which he is pursuing through the infliction of agonizing pain?

106

Introduction II: What is Torture?

If the Achieved Purpose Inquiry is construed as universally quantified, it should obviously be answered in the negative. On the one hand, some unquantified context-specific versions of the Achieved Purpose Inquiry might be answerable affirmatively. Suppose that Tom the Torturer perceives the eliciting of severe pain in his victim as an end in itself (perhaps because Tom passionately detests the victim, or perhaps because he adheres to a retributivist conception of punishment and believes that the inducement of severe pain is a deserved sanction for some crime). In this case, the occurrence of the pain is a necessary and sufficient condition for Tom’s attainment of his purpose. Thus, if the correct answer to the Experience of Pain Inquiry lies in its first disjunct—that is, if some mode of mistreatment which almost always suffices to produce agony is not torture on a given occasion unless the victim of the mistreatment actually experiences agony—the question whether Tom the Torturer has perpetrated torture will be the same as the question whether he has succeeded in realizing his objective. On the other hand, however, if the correct answer to the Experience of Pain Inquiry lies in its second disjunct, then these two questions about Tom the Torturer are not equivalent. Tom can commit torture without actually eliciting the horrific pain in his victim that would almost always ensue from the steps taken by him against her. In regard to Tom, then, the correctness or incorrectness of an affirmative answer to an unquantified context-specific version of the Achieved Purpose Inquiry depends on the correct answer to the Experience of Pain Inquiry. Most torturers, however, seek to achieve ulterior purposes—such as the extraction of information from terrorists—through the infliction of severe pain. When the Achieved Purpose Inquiry is applied to them, the only tenable answer to it is negative. Suppose for example that a torturer employs various agony-inducing techniques against a terrorist in an effort to obtain information about some nefarious plot. Suppose that the terrorist provides false information, or that he manages to withstand the agony of his ordeal while remaining stonily silent. In these circumstances the torturer has failed to achieve the purpose for which he has afflicted the terrorist with excruciating pain, yet he has manifestly engaged in torture. His not having attained his ulterior purpose is irrelevant to the question whether he has perpetrated torture, regardless of what the correct answer to the Experience of Pain Inquiry is. In sum, the Achieved Purpose Inquiry and the Experience of Pain Inquiry are far from equivalent. The correct answer to the former (in its universally quantified form) is patently negative, whereas the correct answer to the latter is not nearly as obvious. While the former is quite a silly question, the latter is a serious philosophical and moral inquiry. 2.3.1.1.2.  A question put aside Nevertheless, despite the seriousness of the Experience of Pain Inquiry, we can largely set it aside within this book. After all, in this volume I am chiefly investigating why torture is wrong and when it is wrong. Of lesser prominence here is the matter of pinning down exactly how grave the wrongfulness of torture is.

Conclusion: An Overview

107

Let us confine our attention for the moment to the varieties of torture that are always morally wrong. What is evident is that both torture and attempted torture are very serious wrongs. Torture is an even graver mode of misconduct than attempted torture, but the wrongness of the former in all circumstances is paralleled by the wrongness of the latter in all circumstances. Moreover, the morally significant factors that account for the wrongness of the former are the same as the morally significant factors that account for the wrongness of the latter. Hence, my pretermission of the Experience of Pain Inquiry will not detract from this book’s endeavours to address the matters that are its chief concerns. My inclination to prescind from the Experience of Pain Inquiry is greatly strengthened by four further considerations. First, the preceding paragraph’s claim about the relative gravity of torture and attempted torture is hardly uncontroversial. Such a claim rests on the proposition that the moral bearings of a person’s conduct are sometimes affected by moral luck: that is, by some consequences or circumstances over which the person has had no control. I have elsewhere endorsed that proposition about moral luck (Kramer 2011, 205–6, 248–9) and have argued sustainedly in favour of it (Kramer 2004, chap. 8), but some other philosophers have reasoned quite forcefully against it (Enoch and Marmor 2007). A corollary of their arguments is that the gravity of attempted torture is the same as the gravity of torture. Now, although I believe that their lines of reasoning against the moral relevance of resultant luck and circumstantial luck are unsound, the soundness of those lines of reasoning would militate strongly in favour of this book’s declining to pin down the exact distinction between torture and attempted torture. Accordingly, when this book does indeed decline to pin down that distinction—for the reasons indicated in the preceding couple of paragraphs—it is not thereby offering any hostages to fortune. In other words, it is not thereby presupposing the falsity of the proposition that there is no such thing as moral luck. Quite the contrary. A second consideration that supports this book’s pretermission of the issue of attempts is that the distinction between torture and attempted torture is generally clear. Cases of genuine unclarity, such as the case of the anomalies in the functioning of Veronica’s nervous system in § 2.3.1.1, are exceedingly rare. Thus, even if the issue of attempts did bear heavily on the central questions which this book is addressing, there would scarcely ever be a need to pin down the distinction between torture and attempted torture very sophisticatedly and precisely. We could come to grips with virtually all actual cases while relying on an ordinary (largely pre-theoretical) understanding of that distinction. Third, whatever may be the exact difference between the gravity of the actions carried out against Veronica and the gravity of some otherwise indistinguishable actions that succeed in inducing terrible pain in a victim, the difference is modest indeed. It is markedly smaller than the moral disparity between certain techniques of torture: for example, the moral disparity between gouging out someone’s eyes and twisting someone’s arm behind his back. Hence, when this book declines to explore further the extent of the divergence between the gravity of the actions performed against Veronica and the gravity of some closely similar actions that induce agony in another victim, it is putting aside a small matter.

108

Introduction II: What is Torture?

Fourth, even if we did need to explore further the divergence just mentioned, we could perfectly well do so without judging whether the actions performed against Veronica are correctly classifiable as torture or whether they should instead be classified as attempted torture. Though the taxonomical matter of the distinction between torture and attempted torture is manifestly worthy of attention in an analysis of the nature of criminal attempts, it does not have to be settled by theorists who want to gauge the seriousness of the wrong committed when someone has taken steps to afflict Veronica with excruciating pain. Without settling whether the non-occurrence of the pain (because of the anomalies in the functioning of Veronica’s nervous system) is sufficient to ground the taxonomical difference between torture and attempted torture, theorists can take account of that non-occurrence as they ascertain how grievously wrongful the measures carried out against Veronica are. We can prescind from the taxonomical crux while addressing questions about the extent of the wrongfulness—including the question whether the fortuitous non-occurrence of pain attenuates the wrongfulness at all. 2.3.1.1.3.  A matter of gravity Although the division between torture and attempted torture is ancillary to this book’s principal concerns, the matter of the gravity of torture’s wrongness does of course figure in the book’s arguments in some important respects. Two of those respects should be noted briefly here. First, in any particular context relating to torture that might be weakly justified, the location of the threshold of moral optimality—which we have already pondered in Chapter  1 and will ponder further hereafter—obviously hinges on the stringency of one’s duty to abstain from the use of torture. The stringency of that duty, in turn, plainly hinges on the gravity of the wrong that is constituted by any such use. Consequently, we cannot ascertain the location of the threshold of moral optimality in any given situation without knowing how gravely wrong this or that instance of torture would be. Still, the enquiries pursued in this book do not require us to ascertain the location of any threshold of moral optimality. Indeed, as Chapter 1 has remarked, that location is in any event indeterminate. It exists at no particular point in a region (a vaguely demarcated region) of indeterminate borderline cases. Of major importance for this book is the general nature of a threshold of moral optimality, but the exact position of any such threshold is not something that we need to identify or can ever identify—for there is no precise position that it occupies. Second, Chapter 5 will consider the legal consequences that should attach to the perpetration of torture. A schedule of legal sanctions—whether the sanctions are criminal or civil or institutional, and whether they are imposed individually or collectively—should reflect the gravity of the wrongs committed by torturers. The severity of the sanctions should correspond, at least approximately, to gradations of moral wrongfulness. Still, as has been remarked above, we can tackle questions about gradations of moral wrongfulness without mulling over exactly where the demarcation should be drawn between torture and attempted torture. We do not have to settle that demarcation in order to judge whether

Conclusion: An Overview

109

the legal sanctions that should be imposed for the mistreatment of Veronica are lighter than those that should be imposed for the homologous mistreatment of a victim who experiences agony therefrom. Moreover, within the confines of this book, which is pitched at a high philosophical level of abstraction, my ruminations on the appropriate legal responses to the perpetration of torture will not yield any fine-grained prescriptions. Those ruminations will deal with some general types of sanctions rather than with any detailed calibrations of severity. Consequently, although the moral gravity of various torturous actions does obviously bear on the harshness of the legal sanctions that should be levied in response to those actions, the specifics of the correlations between the gravity of the actions and the harshness of the sanctions lie well beyond the scope of this book. 2.3.1.1.4.  Back to the definition Notwithstanding that this book will prescind from the Experience of Pain Inquiry, my definition of torture will appear to take a position on that inquiry by stating that torture (if uninterrupted) consists in the infliction of severe pain or suffering. However, that portion of the definition is not really addressing the Experience of Pain Inquiry; it is not seeking to specify any precise delimitation between torture and attempted torture. Rather, because my definition of torture would become confusingly unwieldy if it were to give expression to a non-committal stance on the Experience of Pain Inquiry, it embraces the first disjunct of that inquiry to the exclusion of the second disjunct. However, nothing of any substance is signalled by the omission of the second disjunct. That omission is a matter of style. If the correct answer to the Experience of Pain Inquiry is encapsulated in its second disjunct, then my definition of torture should be reconstrued accordingly. As has already been suggested, everything to be said in this book about the wrongness of torture is applicable—mutatis mutandis—regardless of how the Experience of Pain Inquiry should be answered.

2.3.1.2.  How long? As we have seen in this chapter’s survey of previous definitions of torture, some philosophers and jurists have submitted that the inducement of agony in a victim of torture is temporally limited in some way. Whereas Kershnar’s definition (examined in § 2.1.4.2) declares that torture consists in the infliction of grievous pain over a short span of time, the Amnesty International definition (examined in § 2.1.3) suggests that torture consists in the infliction of grievous pain over a prolonged span of time. Neither of those durational restrictions is pertinent, for the practice of torture is more diverse than either of them allows. Given the heterogeneity of the purposes for which torture can be employed, we should scarcely be surprised that the periods of time during which it lasts are also highly various. Moreover, even within a single broad type of torture—such as interrogational torture—the lengthiness of the infliction of pain can vary markedly. If the purpose for which someone resorts to torture is achieved quickly and verifiably, the torture

110

Introduction II: What is Torture?

itself will be transitory; contrariwise, if the purpose of the torturer remains unaccomplished for a long period, his use of torture will probably be protracted. Some techniques of torture are designed to induce agony rapidly, whereas others are designed to build up extreme suffering gradually. Thus, any definitional specification of the duration of torture is bound to be an adscititious stipulation. A definition of torture should include at most a capaciously abstract reference to the dimension of time. For example, such a definition might state that the infliction of severe pain will continue until the purpose underlying it has been fulfilled or until the unfulfillability of that purpose through the infliction of severe pain has become manifest. Any number of eventualities can make manifest the unfulfillability of the underlying purpose through the infliction of severe pain. Perhaps the victim has died, or perhaps the victim has displayed unshakeable resoluteness, or perhaps the use of torture has had to be aborted because it has been detected and interrupted by investigative authorities or because the equipment for applying the torture has malfunctioned. Any number of other factors—intrinsic or extrinsic to the perpetration of torture—can likewise render plainly unfulfillable the purpose for which someone has resorted to torturous techniques. Note, moreover, that my capaciously abstract reference to the duration of torture can accommodate a situation in which someone originally makes use of torture for a certain purpose and then continues to employ the torture for some other purpose. For example, suppose that Tom the Torturer originally subjects Veronica the Victim to some pain-inducing techniques in order to obtain some important information from her. Having become convinced that Veronica is not possessed of the information which he is seeking, and having thereby recognized that the purpose behind his interrogational torture is unfulfillable through the affliction of her with severe pain, Tom becomes furious and continues to subject her to agony. His torturous treatment of Veronica has changed from a quest for information to an exercise in sadistic vindictiveness. No longer engaging in interrogational torture, Tom is now engaging in sadistic torture. Because his purpose in deliberately afflicting Veronica with excruciating pain has altered, the duration of the torture to meet that purpose will likewise have altered; it might cease only when Veronica has died or has become utterly incapacitated mentally. Still, even the capaciously abstract reference to time that has been broached here—flexible though it is—would be too confining if it were presented as a necessary condition for the character of any conduct as torture. In regard to some credibly possible instances of torture, the duration is not in accordance with that abstract reference. Let us consider a significantly modified version of the scenario in the preceding paragraph. Suppose that, instead of turning out to be ferociously sadistic, Tom the Torturer has some misgivings about his vile trade. He has been assigned to subject Veronica to interrogational torture, and he embarks on just such a course of action. However, after quite a short time, while he is still firmly convinced that he can succeed in extracting the desired information from her, his awakened sense of compassion leads him to desist from his mistreatment of her. In these credibly possible circumstances, the torture has lasted neither until it has succeeded in the purpose for which it has been undertaken nor until the

Conclusion: An Overview

111

unattainability of that purpose has become clear. All the same, it would be preposterous to maintain that Tom’s administration of pain-inducing techniques to Veronica’s body has not amounted to torture. Ergo, like some of the other elements in a satisfactory definition of the phenomenon of torture, any abstract definitional reference to the duration of each instance of that phenomenon should declare what is typically the case rather than what is invariably the case. Typically (but not always), the administration of torture lasts either until the purpose impelling it has been achieved or until the unattainability of that purpose has become evident to the torturer.

2.3.2.  Against the interests of the victim? In § 2.1.4.1.5, and in my discussions of edifying torture in § 2.2.12 (and perhaps also in my discussions of sado-masochistic torture in §§ 2.1.4.1.4 and 2.1.4.4.2), we have seen that Davis goes astray when he definitionally declares that the infliction of severe pain on someone will count as torture only if it is undertaken out of indifference or hostility toward the well-being of the person against whom the infliction is directed. Laid down as a necessary condition for the classifiability of any instance of conduct as torture, Davis’s claim is too strong. In exceptional cases, instances of torture are conducted for the physical and psychological edification of the victims. However, reformulated as a statement about a typical feature of torture, Davis’s claim is unexceptionable and important. So reformulated, it should be included in a definition of torture. The importance of Davis’s point is squarely connected to the wrongfulness of torture. As Chapter 3 will maintain, the distinctive wrongfulness of torture derives from the way in which the infliction of severe pain uses and abuses the victim of the infliction. Even in the extraordinarily rare situations in which some victims of torture are not utterly defenceless, the affliction of them with excruciating pain will have dominated their minds by fixating them on their own agony. Though a victim might staunchly refuse to comply with the behests of his tormentors— behests for the provision of information, for example—he will be consumingly preoccupied with his own dire pain by the tribulations that are imposed on him. Likewise, even in the exceedingly uncommon circumstances in which a victim of torture is a Stoic who manages to attain to a state of apatheia while undergoing the hideous pain inflicted on her by torturous techniques,30 the need for the recourse to apatheia is forced upon her by the power of her own anguish. Precisely by using people’s susceptibility to the consuming force of agony in this manner, morally wrongful instances of torture embody an aspiration toward an overweeningly god-like sway. They likewise embody an attitude of indifference or arrant hostility toward certain people’s fundamental interests (specifically, their interests in basic physical and psychological well-being). That outlook behind the perpetration of 30   Worth recalling here is Leonato’s famous mockery of Stoicism in Much Ado about Nothing: ‘For there was never yet philosopher |That could endure the toothache patiently, | However they have writ the style of gods | And made a push at chance and sufferance’ (V.i.35–8).

112

Introduction II: What is Torture?

each of those wrongful instances of torture is a key to the absoluteness of the moral prohibition that is contravened by each of those instances, as we shall see in Chapter 3. Hence, the fact that such an outlook impels any morally wrongful instances of torture is something that should be signalled in one’s general definition of the phenomenon.

2.3.3.  Consent and control In several of this chapter’s discussions, the elements of consent and control have figured saliently. Torture is consensual if and only if the person subjected to it has agreed to undergo the experience, in conditions where the voluntariness of the agreement is not vitiated by factors such as coercion or deception or ignorance or immaturity or intoxication or senility or mental retardation. A process of being afflicted with dire pain is under the control of the person subjected to it if and only if she can terminate it (or scale it down) at any point of her choosing, where the genuineness of the control is not undermined by any voluntariness-impairing factors such as those just mentioned. If a victim is in control of a process whereby she is afflicted with severe pain, the process at any given time is consensual—for the victim would have terminated it already if she had been unwilling to continue to undergo it. By contrast, there is no entailment in the other direction between consensuality and control. In a sado-masochistic context or in a context of aversion therapy or in a course of resistance training, someone might consent to undergoing a process of torture over which he will have no control. As this chapter has already remarked, a sado-masochistic person might seek to augment his sexual gratification by consenting to the open-ended administration of torture rather than to a mere simulation of torture which he can halt at a time of his choosing. Similarly, someone undergoing aversion therapy or resistance training might consent to the open-ended administration of torture—rather than to a mere simulation of torture that can be terminated at his request—in order to enhance the effectiveness of the therapy or the training. In such circumstances, consent is present without control. Precisely because the presence of consent does not entail the presence of control, there can occur instances of veritable torture that are consensual. The open-ended infliction of severe pain for the purpose of inducing sexual pleasure or overcoming a harmful addiction or toughening one’s resistance to future hardships is outright torture even when the person who undergoes the pain has readily consented to the ordeal.

2.3.3.1.  Two caveats about control In connection with the notion of control that has been invoked here, two caveats should be entered. First, the element of control is never present in a context of interrogational torture or act-impelling torture. Although a victim of such torture might be able to terminate it by providing the requisite information or by performing the requisite action, the genuineness of the victim’s control over the torture is undermined by the voluntariness-impairing factor of coercion. Whenever that voluntariness-impairing

Conclusion: An Overview

113

factor or some other voluntariness-impairing factor is operative, a victim’s ability to bring her torturous tribulations to a close does not amount to genuine control of the sort envisaged here. Only in some sado-masochistic contexts or in some edifying contexts is any genuine control present. When such control is indeed present in one of those contexts, the deliberate infliction of intense pain is a simulation of torture rather than full-blown torture. Second, control is absent not only when the plying of some agony-eliciting technique is open-ended but also whenever the victim of the plying of such a technique cannot voluntarily halt it at a time of his choosing. For example, a course of punitive torture might be of a fixed duration (20 lashings or canings, for example). Nonetheless, because the victim of the torture cannot terminate it at a time of his choosing, he lacks any control over it. Hence, the punishment of him through the deliberate inducement of excruciating pain is an instance of outright torture.

2.3.3.2.  The upshot of the matter Outside sado-masochistic contexts and some edifying contexts, torture is nonconsensual. Nevertheless, because sado-masochistic torture and some instances of edifying torture are consensual, a definition of the general phenomenon of torture should not specify non-consensuality as a necessary condition for the instantiation of that phenomenon. Contrariwise, the absence of genuine control by a person who undergoes some agony-inducing measures is indeed a necessary condition for the very status of those measures as torture. In the presence of genuine control by the victim, the measures would amount only to a simulation of torture. Consequently, a definition of the general phenomenon of torture should indeed declare that an absence of genuine control by the victim is a necessary condition for the instantiation of that phenomenon. Let us recall that, although this chapter has occasionally touched on the concrete moral bearings of torture, its focus has lain principally on the more abstract matter of the general nature of torture. Thus, although the elements of consent and control are obviously of moral relevance, my current discussion is not squarely addressing questions about the wrongness of torture. Instead, it is squarely addressing questions about the status of any given instance of conduct as torturous or non-torturous. Hence, although the absence of genuine control by the victim is a necessary condition for the status of any instance of conduct as torture, it is not always a wrong-making feature of the conduct. This chapter leaves ample room for attributions of moral legitimacy to some instances of deliberately agony-inducing conduct—some instances of ephemerally incapacitative torture and of edifying torture, for example—where the victims of the conduct lack any genuine control over its continuation or discontinuation.31 Although my account of the nature of torture is of course oriented toward my subsequent account of the wrongness 31   As is evident in § 2.2.12.2.3, this chapter has furthermore left room for the conclusion that some deliberately agony-inducing conduct undertaken to avert a patient’s lapse into a coma can be non-torturous—as well as morally legitimate—even though the patient has no genuine control over the continuation or discontinuation of the conduct.

114

Introduction II: What is Torture?

of torture (as I  have made clear in § 2.3.2 and elsewhere in this chapter), and although the issues tackled by the former account do overlap with those tackled by the latter, the former array of issues is scarcely collapsible into the latter array.

2.3.4.  A definition of torture At last, then, this chapter can present its definition of torture. As is to be expected from everything that has been said in the chapter heretofore, my definition is considerably less tidy than any of the formulations that have previously been propounded by jurists or by other philosophers. Some of the components of my formulation state necessary conditions for the occurrence of any instance of torture, but most of the components instead state typical conditions. (Nonetheless, recall that my definition is tidier than it would be if it were reflective of my non-committal stance on the Experience of Pain Inquiry.) Torture, if uninterrupted, consists in the infliction of severe pain or suffering. Almost always the pain or suffering is induced deliberately as a means or as an end, but in some exceptional contexts it is induced instead through extravagant recklessness. Torture in its many varieties is perpetrated for any of the purposes enumerated in § 2.2 of this chapter and listed in § 2.2.13. Typically, the administration of torture lasts either until the purpose impelling it has been fulfilled or until the unrealizability of the specified purpose through the infliction of torture has become manifest. Save in some edifying contexts—and perhaps also in some sado-masochistic contexts—any administration of torture evinces indifference or hostility toward the basic physical and psychological well-being of its victim. Save in some of the contexts just mentioned, a victim of torture does not genuinely consent to being afflicted with grievous pain or suffering. Whenever the deliberate or extravagantly reckless infliction of severe pain amounts to torture, the victim lacks any genuine control over the duration of the infliction.

3 Why Torture is Wrong Having come to grips with the nature of torture in its sundry manifestations, we should now seek to understand why torture is wrong. Most types of the phenomenon are always and everywhere wrong, and even the two types that can sometimes be morally legitimate—ephemerally incapacitative torture and edifying torture— are also instantiable in countless ways that are morally illegitimate. Among the tasks to be accomplished in this chapter is that of explaining the moral difference between the few permissible instances of torture and the multitudinous instances that are impermissible. Now, when this book declares that nearly all kinds of torture are always and everywhere wrong, it is declaring that those kinds of torture are always and everywhere not strongly permissible. With regard to several of those kinds of torture, this book thus leaves room for some possible instances that are weakly permissible. That is, it leaves room for some possible instances that are morally optimal even though they are morally wrong. Consequently, it leaves room for some possible instances that are morally justified—weakly justified—even though they are morally wrong. Any such instance is overtoppingly obligatory notwithstanding that it is morally impermissible. Hence, this chapter will be seeking to explain why all possible instances of most kinds of torture are morally illegitimate. In so doing, it will allow that some of those morally illegitimate instances are morally obligatory and indeed that some of them in imaginable circumstances are overtoppingly obligatory. This point is in need of emphasis at the outset, since numerous philosophers—both opponents and supporters of the use of torture—have written as if attributions of moral permissibility to any possible instances of torture are more frequently well-founded than are attributions of moral obligatoriness. In other words, those philosophers believe that acts of torture are even more rarely endowed with the property of moral obligatoriness than with the property of moral permissibility. David Luban gives stark expression to this view when he indicates that he is addressing the question whether ‘a system in which any atrocity, no matter how vile, can be permitted (or, worse, required) can count as a morality’. Luban is joined by a medley of other philosophers. For example, Larry Alexander attributes to Michael Moore the thesis that in certain dire situations ‘one is morally permitted, and perhaps required, to

116

Why Torture is Wrong

engage in those acts that are otherwise morally prohibited.’1 Many other philosophers have written in a similar vein, with expressions such as ‘at least permitted if not required’ or ‘permitted and even required’. All such phrases imply that an ascription of moral obligatoriness to any torturous session of interrogation is even less likely to be correct than an ascription of moral permissibility to that session. As should be apparent from what has already been said here as well as from what has been said in Chapter 1, the implication conveyed by the foregoing phrases is misguided. It gets things exactly the wrong way around. Whereas some possible instances of interrogational torture or act-impelling torture are morally obligatory, no possible instance of either of those kinds of torture is ever morally permissible. This chapter will endeavour to explain why all possible instances of each of those two types of torture are morally wrong, but will not deny that some possible instances of each of those types are morally obligatory. Once we are attuned to the occurrence of moral conflicts—especially to the occurrence of such conflicts in dire emergencies—we can readily grasp that the moral wrongness of every possible instance of interrogational torture or act-impelling torture is consistent with the moral obligatoriness of some possible instances of each of those kinds of torture. Contrary to what Luban has suggested in the remark quoted above, an attribution of moral permissibility (strong permissibility) to any torturous interrogational or act-impelling measures would be far worse than an attribution of moral obligatoriness to some such measures. An attribution of the latter sort can derive from a clear-sighted recognition of the occurrence of a moral conflict in some dreadful predicament. By contrast, an attribution of moral permissibility would in effect imply that the moral constraints which forbid the use of any torturous interrogational or act-impelling measures are altogether negated—rather than merely countervailed or overtopped—in some dreadful predicament. It would imply that the predicament does not pose a moral conflict at all. Any such exonerative view of the use of interrogational or act-impelling torture in the circumstances of grim moral quandaries is a principal target of this chapter and Chapter 4. The present chapter will distil some wrong-making features of interrogational or act-impelling torture that are morally operative even in the gravest of emergencies where the torture is employed. In combination with Chapter 4, which will seek to vindicate the agent-centredness of the moral prohibition on the use of interrogational or act-impelling torture, this chapter will venture to affirm that that moral prohibition is absolute. Although the absoluteness of the prohibition is weak rather than strong, the conclusion to be reached herein is that any acts of torture—apart from some acts of ephemerally incapacitative or edifying torture—are in contravention of a formidably stringent moral requirement that is always and everywhere binding in all possible

1   Luban 2009, 202, emphasis added; Alexander 2000, 894, emphasis added. For some of the many writings that use broadly similar language or otherwise express the same general view, see Brunnée and Toope 2010, 221; Curzer 2006, 36, 38, 47, 54 n1; Davis 2005, 162; Harel and Sharon 2011, 848; Kadish 1989, 346; Luban 2009, 184, 198, 201; McMahan 2006, 244; Miller 2005, 185; Miller 2011, § 3.2; Moore 1997, 684; Moore 2007, 92; Nagel 1979, 63; Paskins 1976, 142; Posner and Vermeule 2006, 682 n25; Seidman 2005, 890; Waldron 2010a, 6, 263.

Why Torture is Wrong

117

worlds. Unlike ascriptions of moral obligatoriness (which in extreme cases can be well-founded), ascriptions of moral permissibility to instances of interrogational or act-impelling torture are invariably misconceived. This chapter will begin its account of the wrongness of torture negatively, by pondering some of the previous efforts of philosophers to specify why torture is wrong. Though some of those efforts have produced really valuable insights that will be endorsed in my own account, none of them has been entirely successful in capturing both the distinctiveness and the absoluteness of torture’s wrongness. Thus, while the present chapter will build on some of those past philosophical ruminations, it aspires to go beyond them. Naturally, anyone seeking to provide an account of torture’s wrongness should be careful not to overstate the distinctiveness of that wrongness. As we have observed in Chapter  2, torture shares its key wrong-making properties with attempted torture, and it likewise closely resembles the simulations of torture that take place in some sado-masochistic or therapeutic or resistance-training contexts. Moreover, as is evident from the definition propounded at the end of Chapter 2, the concept of torture is vague (in the technical sense of being embeddable in sorites paradoxes). Differences between torture and cruel or degrading or inhuman treatment that falls short of torture are marked by a range of borderline cases—a vaguely demarcated range—in which the correct classifications are indeterminate. Indeed, even the differences between interrogational torture and some morally legitimate techniques of interrogation are marked by such a range of borderline cases.2 Still, an account of the moral status of torture that highlights the distinctiveness of its wrong-making features can be perfectly consistent with these points about the vagueness of the concept of torture. My emphasis on the distinctiveness of torture’s wrong-making features is not in any way a denial of the vagueness of that concept or a denial of the resemblances between torture and certain other modes of behaviour. Rather, it is a warning against specifying those wrong-making features in a way that too smoothly subsumes torture under some wider category of misconduct. As we shall see, some of the past accounts of the wrongness of torturous measures have indeed lumped torture together unhelpfully with other modes of misconduct—even while some of those same accounts have distilled the wrong-making properties of torture in ways that exclude certain clear-cut instances of the phenomenon. Among the pitfalls to be avoided here are an unilluminatingly high level of abstraction and an excessively narrow focus. A moral assessment of torturous measures that underscores the distinctiveness of the wrongness of those measures will also be able to explain adequately why the moral prohibition contravened by illegitimate uses of torture is absolute. Most of the major insights generated by previous endeavours to encapsulate the wrongness of torture have come from victim-focused approaches to the matter. Such 2   For some (mostly laconic) remarks on the vagueness of the concept of torture, see Kamenova 2009, 87, 93; Levinson 2003, 2025; Levinson 2005, 252; Lukes 2005, 10–11; Strauss 2003, 211 n27, 234; Twining and Twining 1973, 340–1; Twining 1978, 151–61; Waldron 2010a, 200–7; Waldron 2010b, 20–5. I shall return to this matter in Chapter 5.

118

Why Torture is Wrong

approaches are crucial but insufficient for vindicating the absoluteness of the moral prohibition just broached. To uphold the absoluteness of that prohibition, I shall have to supplement the victim-focused assessments with a perpetrator-focused perspective—as Chapter  1 has already briefly suggested. (Whereas the present chapter will supply a key part of the basis for my contention that the aforementioned moral prohibition is absolute, the remainder of that basis will be supplied by Chapter 4 with its defence of the agent-centredness of deontological requirements. Though a perpetrator-focused perspective on torture’s wrongness and an affirmation of the agent-centredness of the deontological prohibition on torture are not mutually entailing, they complement each other nicely as components of a full-blown insistence on the absoluteness of that prohibition.)

3.1.  Some Previous Accounts of the Wrongness of Torture By perusing several of the past attempts of philosophers to explain why torture is wrong, this section of the chapter can alert us both to missteps that should be avoided and to insights that should be gleaned. Attention to the missteps as well as to the insights can serve to sharpen the focus of my own explanation of torture’s wrongness later in this chapter.

3.1.1.  Contractarian approaches Contractarian theories of justice and political obligation have been prominent in Western thought for the past four centuries, and have been especially influential in their modern forms during the past four decades. Contractarian methods of enquiry during those decades have informed various philosophers’ discussions of a wide range of moral issues, including the problem of torture. Hence, my survey of some previous efforts by philosophers to pin down the wrongness of torture can fittingly begin with an examination of a couple of contractarian approaches to the matter.

3.1.1.1.  Nagel and justifiability In a deservedly famous though somewhat uneven essay, Thomas Nagel sets out to provide a contractualist justification for an absolute moral prohibition on the use of interrogational torture. He believes that such torture is distinguishable from certain other harmful modes of conduct, in that it can never be justified adequately to the victim who undergoes it. His initial presentation of his argument is as follows: [L]‌et me try to connect absolutist limitations with the possibility of justifying to the victim what is being done to him. If one abandons a person in the course of rescuing several others from a fire or a sinking ship, one could say to him, ‘You understand, I have to leave you to save the others.’ Similarly, if one subjects an unwilling child to a painful surgical procedure, one can say to him, ‘If you could understand, you would realize that I am doing this to help you.’ One could even say, as one bayonets an enemy soldier, ‘It’s either you or me.’ But

Some Previous Accounts of the Wrongness of Torture

119

one cannot really say while torturing a prisoner, ‘You understand, I have to pull out your finger-nails because it is absolutely essential that we have the names of your confederates.’ (Nagel 1979, 67–8, emphases in original)

This argument as it stands is curiously weak, as some other critics of Nagel have contended (Levinson 2003, 2032; Posner and Vermeule 2006, 676; Posner and Vermeule 2007, 187). Though innumerable instances of interrogational torture could not be justified in any ways that would be plausible to the victims, Nagel is talking about an instance of calamity-averting interrogational torture. If nothing short of torturous measures will extract from a terrorist some crucial information that can be used by the interrogators to spare many innocent people from a calamity, and if the interrogators have good grounds for believing that the torturous measures are likely to succeed, then the justification that can be offered to the terrorist is at least as strong as the justification that can be offered to the bayoneted soldier whom Nagel mentions. (Indeed, it is probably stronger—since the interrogators will probably be acting to save far more lives, and the harm inflicted on the terrorist by the torturous measures will probably be significantly less than the harm inflicted on the soldier by the bayoneting.) Given that Nagel regards as adequate the justification proffered to the bayoneted soldier, there is no apparent basis for his gainsaying the adequacy of the justification that can be furnished to the terrorist. Nagel himself recognizes that his argument in the excerpt above ‘does not take us very far, of course’. He elaborates: [A]‌utilitarian would presumably be willing to offer justifications [to victims of calamityaverting interrogational torture] in cases where he thought they were sufficient. They are really justifications to the world at large, which the victim, as a reasonable man, would be expected to appreciate. However, there seems to me something wrong with this view, for it ignores the possibility that to treat someone else horribly puts you in a special relation to him, which may have to be defended in terms of other features of your relation to him. The suggestion needs much more development; but it may help us to understand how there may be requirements which are absolute in the sense that there can be no justification for violating them. If the justification for what one did to another person had to be such that it could be offered to him specifically, rather than just to the world at large, that would be a significant source of restraint. (Nagel 1979, 68)

Once again, Nagel’s line of reasoning is weak. Were Nagel really managing to show that the moral proscription contravened by interrogational torture is absolute, his remarks would cover a situation of the sort which I have envisaged above. In that situation, the terrorist though defenceless is posing a serious threat to the people (perhaps including the interrogators) who will be adversely affected by the atrocity which she herself has helped to engineer. Her reticence imperils them just as much as the wielding of weapons by an enemy soldier imperils the troops on one’s own side. Yet, given that Nagel believes that the justification extended to the bayoneted soldier is satisfactory, he is committed to accepting that the justification extended to the captured terrorist is satisfactory. That latter justification might be congenial to consequentialists—as might the justification offered to the bayoneted soldier, whose utility-reducing onslaughts have been thwarted—but it is focused on some features of the relationship between the

120

Why Torture is Wrong

terrorist and the interrogators. Specifically, it is focused on the fact that the terrorist is posing a clear threat to other people (perhaps including the interrogators themselves) for whose safety the interrogators are responsible. The terrorist has helped to give rise to a danger which she is now sustaining through her silence. In that respect, she is markedly different from people in the ‘world at large’. Through her actions and her subsequent taciturnity, she has forfeited her right against being subjected to torture that can help to avert the calamity which she has sought to bring about—or so the interrogators can colourably maintain when they explain to her why they have singled her out for the torturous measures which they have employed. Of course, manifold other instances of interrogational torture would not lend themselves to being justified in this individualized fashion. Nagel could succeed in distinguishing any of those instances from each of the other harmful courses of conduct to which he refers in the first passage quoted above. However, as he explicitly states in the second passage, he takes himself to be advancing an argument that accounts for the absoluteness of the moral prohibition on the use of interrogational torture. Thus, if his argument is to be successful, it has to apply as well to the situation of calamity-averting interrogational torture which I have been envisioning in the last couple of paragraphs. As we have seen, the argument does not stand up in application to that situation. With regard to interrogational torture, Nagel has failed to adduce a basis for an absolute moral proscription. This short critique of Nagel should close with a caveat. Here and elsewhere in my survey of some previous efforts by philosophers to explain why interrogational torture is morally impermissible, I am decidedly not denying that the wrongness of such torture is absolute. Interrogational torture is morally illegitimate always and everywhere in all possible worlds. Nagel’s conclusion is correct, notwithstanding that his reasoning which leads toward that conclusion is unsound. As Nagel asserts in the second quoted passage, there can never be any strong moral justification for the perpetration of interrogational torture—even in the catastrophic circumstances that have been broached here. Of course, the use of such torture in those circumstances for the purpose of averting an impending catastrophe might be weakly justified; that is, it might be morally optimal even though it is morally impermissible. To reach any appositely firm judgement on that matter, we would need to know considerably more about the specifics of the situation than I have recounted above. At any rate, even if the perpetration of interrogational torture is indeed weakly justified in the terrible circumstances, any recourse by interrogators to such a measure will be in contravention of a stringent moral duty. On that bottom-line point, my account of torture’s wrongness is in accordance with what Nagel has written in the quoted passages.

3.1.1.2.  Meisels and the social-contract tradition Drawing on the great social-contract tradition in early modern political thought (especially the thought of John Locke) at least as much as on the more recent proliferation of contractualist approaches descending from the work of John Rawls,

Some Previous Accounts of the Wrongness of Torture

121

Tamar Meisels submits that a social-contract perspective supports an absolute ban on the use of interrogational torture. Through the adoption of such a perspective, she contends, we can fathom why interrogational torture is always and everywhere wrong. Torturous techniques of interrogation are at odds with the principles that would be freely chosen by individuals whose decisions are reached from a rigorously impartial viewpoint that is uninflected by any morally arbitrary considerations. Meisels’s ruminations on the use of torture are perceptive and refreshingly nuanced, but her grounding of them in a contractarian methodology is unhelpful and unconvincing. Let us examine some of her contractarian pronouncements concerning the prohibition on interrogational torture. First among them is the following statement (2008b, 81): ‘One advantage of consulting the notion of our social contract, as I  suggest here, is that it places an internal restriction against going . . . far towards abolishing basic procedural rights. One might assume that any original or hypothetical contractors, contracting for their own preservation, would never agree to an arrangement which might place them under torturous investigation.’ Although the wording in this initial assertion is quite tentative, Meisels shifts to a much more robust tone in some comments that amplify her opening remark. Consider, for example, the following passage (2008b, 81): [Torturous] measures [are] entirely ruled out, even for Hobbes, by the logic of contracting in order to protect one’s preservation. While such measures may remain available to Hobbes’s unrestrained sovereign who bears no contractual obligation towards his subject, they cannot be willingly consented to by Lockean contractors. Limiting the powers of the sovereign so as to exclude the use of certain measures, such as torture and coerced self-incrimination, is absolutely essential despite the impediment it places on combating terror. It is . . . a logical cost of even the most minimally Lockean-liberal contractarian justification of government.

As Meisels goes on to declare (2008b, 82): Allowing for the temporary limitation of a degree of liberty under severe circumstances while fervently defending the immunity from torture and the right to withhold self-incriminating evidence, is . . . a consistency requirement of the Lockean-type social contract argument employed above.

As the last quotation indicates, Meisels accepts that some restrictions on civil liberties are legitimate during times of serious violence perpetrated by terrorists. To explain why the parties to a social contract would agree on such restrictions while disallowing any relaxation of the ban on torture, she emphasizes anew the link between that ban and self-preservation (2008b, 83): Resisting the use of outright torture and coerced self-incrimination is admittedly only the bare minimum of what need concern us under present circumstances. Detention without trial, relaxed rules of evidence, non-jury trials, arrests, search and seizure without a warrant, and even applying moderate physical force which does not amount to torture in cases of ‘ticking bombs,’ are all grave issues that cannot be resolved on Hobbesian grounds. None is as closely linked to the right of self-preservation as [is] the prohibition against torture . . . . [T]‌hey are not excluded by the present account so long as they are both necessary for, and reasonably effective in, preserving the lives of innocent citizens.

122

Why Torture is Wrong

3.1.1.2.1.  A terse critique In my brief appraisal of Meisels’s line of reasoning, I  will put aside the general objections which I have raised elsewhere against contractualist accounts of morality (Kramer 2009, 67–9). Here the question is not about the merits of contractualism as a general theory of morality but is instead about the persuasiveness of Meisels’s belief that the parties to a social contract—whose deliberations prescind from any knowledge of their social positions and their natural talents and all the other features that differentiate them one from another—will favour an absolute prohibition on the use of interrogational torture. If her belief is correct, then the parties will have disallowed the employment of such torture even in dire emergencies where torturous techniques of interrogation are likely to succeed in extracting information that can be drawn upon by officials to avert terrible calamities. In accordance with the standard contractarian methodology, we should assume here that each party to the social contract is seeking to maximize the level of well-being that he or she will attain under the legal-governmental arrangements that are enshrined in the contract. (For present purposes, it does not matter whether the parties embrace a Rawlsian maximin mode of reasoning or whether they instead favour a risk-neutral approach in their deliberations.) Given Meisels’s focus on the efforts by the contracting parties to enhance their prospects of preserving their lives, we should assume that their deliberations about the use of torture will be oriented toward the desiderata of survival and basic welfare. Proceeding with such an orientation, the parties will clearly set themselves against many kinds of torture. Even with regard to interrogational torture, they will set themselves against any instances of it that are not aimed at fending off calamities. Moreover, of course, the parties will disallow any instances of calamity-averting interrogational torture that are very unlikely to succeed (owing to any of the numerous factors that can render success improbable). However, parties concerned to maximize their own prospects of preserving their lives and basic welfare will scarcely discountenance the use of calamity-averting interrogational torture in any extreme circumstances where such torture is likely to succeed in preventing the occurrence of horrific carnage—or even in any extreme circumstances where such torture is likely to lead to the rescue of a lone innocent person from being murdered, as in my variant of the case of Leon v. Wainwright in § 2.2.1.3 of Chapter 2. After all, from the strictly impartial perspective of each of the parties to the social contract, nobody can know whether he or she will turn out to be a murderous criminal; nor can anybody know whether he or she will turn out to be a victim of murderous criminality. Given as much, and given that certain techniques of interrogational torture will hardly ever be lethal, and given that the use of some of those techniques in the circumstances just envisaged is likely to save at least one life and perhaps numerous lives, the parties to the social contract would be irrational if they were to rule out the use of such torture in those circumstances. Endeavouring to maximize the chances of retaining their lives and basic well-being, the parties will favour the use of interrogational torture in the extreme situations broached here. For contractarians, the parties’ approval concerning some mode of conduct is constitutive of the moral permissibility of that conduct. Hence, Meisels has committed herself to the conclusion that the use

Some Previous Accounts of the Wrongness of Torture

123

of calamity-averting interrogational torture in some extreme emergencies is morally permissible. 3.1.1.2.2.  A possible reply A defender of Meisels might argue that the use of interrogational torture to deal with extreme emergencies will very likely eventuate in the use of such torture to deal with much less dire situations where no calamities loom. Very likely it will eventuate in the institutionalization of torture, with a bureaucracy of professionals prepared to apply torturous techniques in any number of contexts. So might someone retort in defence of Meisels. (Meisels herself does not resort to this line of reasoning.) A bit later in this chapter, we shall encounter some empirical claims and consequentialist arguments that resemble this imagined retort. Hence, my responses here will be very brief. A first point to be noted, which I shall amplify later with some relevant citations, is that the slippery-slope empirical claims proffered by the defender of Meisels are not well substantiated. For example, the ad hoc use of calamity-averting interrogational torture in the circumstances of Leon v. Wainwright has not led to the institutionalization of torture in Florida (where the facts of the case unfolded), nor has it led to any discernible increase in the incidence of torture. When theorists advance claims about the metastasizing tendency of any use of torture by legal-governmental officials, they usually support those claims by reference to authoritarian regimes—such as the Argentinian junta during the 1970s—that have employed torture primarily to suppress dissidence and to intimidate ordinary citizens. Undoubtedly well-founded are the worries about the metastasizing impetus of any torture that is employed for such purposes by such regimes, but those worries are scarcely transferable in any straightforward way to a context in which the officials of a liberal-democratic regime avail themselves of calamity-averting interrogational torture as a last resort in a dire emergency. Even when torturous methods of interrogation were once quite widespread among the police in the United States, those methods (the ‘third degree’) were plied in retrospective investigations rather than in prospective calamity-averting endeavours. Hence, the lamentable frequency of torturous techniques of interrogation among the police in the United States during bygone decades is of very little weight as evidence for the proposition that torture will spread metastatically if officials in a liberal-democratic regime use it ad hoc to save lives in a desperate emergency. When philosophers assert that any such use of torture is likely to spread, they are voicing conjectures that are not really borne out by the examples that are sometimes adduced in support of those conjectures. Second, the parties to the social contract can attach conditions when they allow the employment of calamity-averting interrogational torture in dreadful emergencies. They can maintain that the ad hoc administration of such torture in a situation of extreme urgency is to be countenanced, provided that no previous measure of that kind has led to any discernible increase in the incidence of torturous methods of interrogation in non-emergency contexts. Their provisional approval of calamity-averting interrogational torture is sufficient to commit Meisels to the

124

Why Torture is Wrong

proposition that the wielding of such torture in a grim emergency can be morally legitimate. Third, even were there a stronger empirical basis for the complaints about the metastasis that will ensue if legal-governmental officials resort ad hoc to calamity-averting interrogational torture, the thought-experiment encapsulated in a ticking-bomb scenario can be tweaked to deflect those complaints. Somebody propounding such a scenario can embellish it with any details that are needed to sweep away the concerns about the untoward effects that might arise. For example, he or she can stipulate that arrangements are in place to monitor the actions of the officials who engage in the torture. Unable to deny credibly what they have done, those officials are required to resign from their positions even though their actions have saved many lives. Their resignations are accepted, and further sanctions are imposed civilly or punitively. Given the upshot of the officials’ recourse to calamity-averting interrogational torture, the probability that their behaviour will increase the incidence of torture overall is extremely low. Many alternative ways of embellishing the ticking-bomb scenario in broadly the same direction are likewise possible. With regard to any plight of the sort depicted in a suitably amplified ticking-bomb scenario, the parties to the social contract will endorse a policy of resorting to calamity-averting interrogational torture. For contractarian theorists such as Meisels, the endorsement by the parties is constitutive of the moral permissibility of calamity-averting interrogational torture in any plight of the sort depicted in a suitably expanded scenario. Of course, my modifications of the ticking-bomb scenario have rendered it less realistic (though the specific modifications suggested here are not wildly fanciful). However, even if my amplifications of the scenario were outlandish, the point of a ticking-bomb narrative does not reside—or need not reside—in confronting us with a situation that is likely to materialize. Rather, the point is to induce people to grant that the use of calamity-averting interrogational torture in some imaginably extreme circumstances would be morally permissible. In other words, the point is to induce people to accept that the moral prohibition on interrogational torture is not even weakly absolute. Having prodded an opponent of torture into granting such a concession, the brandisher of a ticking-bomb scenario can then query whether any minor differences between the imaginably extreme circumstances and some more realistic circumstances would matter morally. As David Luban trenchantly summarizes the strategy of those who peddle the ticking-bomb scenario: The ticking time bomb is proffered against liberals who believe in an absolute prohibition against torture. The idea is to force the liberal prohibitionist to admit that yes, even he or even she would agree to torture in at least this one situation. Once the prohibitionist admits that, then she has conceded that her opposition to torture is not based on principle. Now that the prohibitionist has admitted that her moral principles can be breached, all that is left is haggling about the price. No longer can the prohibitionist claim the moral high ground; no longer can she put the burden of proof on her opponent[s]‌. She is down in the mud with them, and the only question left is how much further down she will go. Dialectically, getting the prohibitionist to address the ticking time bomb is like getting

Some Previous Accounts of the Wrongness of Torture

125

the vegetarian to eat just one little oyster because it has no nervous system. Once she does that—gotcha!3

As Luban suggests, then, the far-fetchedness of a ticking-bomb scenario does not really detract from its anti-absolutist message. His remarks are directly relevant to my present discussion of Meisels, for the key point at issue here is whether her contractarianism can support an absolute moral prohibition on the use of calamity-averting interrogational torture—a prohibition that does not lose its force in the terrible circumstances portrayed by a ticking-bomb scenario. On the one hand, Meisels affirms that she has ‘tried to defend a categorical ban on torture regardless of circumstances’. On the other hand, she acknowledges that her ‘defense may admittedly not cover the most horrendous potential consequences that can be conjured up in the wildest of philosophical examples’. Although she immediately adds that ‘I believe [the defense] holds for every real-world “ticking bomb” scenario we have experienced so far and are likely to experience in the foreseeable future’ (2008b, 194), she has already sold the pass to the anti-absolutists. Indeed, she avows as much:  ‘What if we are holding the only terrorist whom we know for sure can prevent a nuclear incineration of the entire United States, and there is no other way of preventing this than by torturing him? Would I still uphold the ban on torture? Perhaps not’ (2008b, 195). Meisels is to be commended for recognizing that her contractarian approach cannot deliver an absolute moral ban on the use of calamity-averting interrogational torture, but she is to be faulted for adopting that approach in the first place. Seeking to uphold an absolute moral prohibition on the wielding of torture by interrogators, Meisels has gone astray by tying her endeavour to a fundamentally consequentialist doctrine (albeit a doctrine focused on a distributionally sensitive maximand rather than on an aggregative maximand). A key to overcoming the difficulties in which she has landed herself is the adoption of an unremittingly deontological approach—an approach that distinguishes between weak absoluteness and strong absoluteness, and between questions about moral optimality and questions about moral permissibility. 3.1.1.2.3.  Another possible reply A defender of Meisels might try one further tack. Instead of responding with unsubstantiated empirical claims about the institutionalization or metastasis of torturous measures, such a defender might point to the possibility of mistakes in the application of those measures. Such mistakes can emerge in many forms, of course; for example, interrogators might underestimate the fanatical resoluteness of a terrorist, who refuses to disclose vital information even when he is subjected to torture. However, the type of mistake that will most likely be invoked by a defender of Meisels is the application of torturous techniques to somebody who does not possess any calamity-averting information. Particularly distressing, naturally, is the application of such techniques to an innocent person who is incorrectly believed 3   Luban 2005, 1440–1, emphasis in original. For a forcefully articulated view of the matter along the same lines as Luban’s view, see Wisnewski 2009.

126

Why Torture is Wrong

by the interrogators to be involved in the plotting of an atrocity. Given the possibility of missteps of that kind—an abiding possibility that is not in need of any elaborate empirical substantiation—the parties to the social contract will discountenance a policy of resorting to torturous methods of interrogation even in emergencies where catastrophes loom. So a defender of Meisels might argue. Although the posited response to my critique of Meisels is correct in maintaining that the possibility of mistakes can credibly be invoked without any detailed empirical studies, it errs in suggesting that the parties to the social contract will regard that possibility as a sufficient basis for their embracing an absolute ban on the use of calamity-averting interrogational torture. Misdirected applications of such torture are obviously dismaying, but the ineliminable possibility of them is just one factor to be taken into account by the parties. Balanced against the possibility of the misdirected applications are the disasters that might ensue in some extreme situations if torturous methods of interrogation are never employed. No one among the parties will have any grounds for presuming—behind the Rawlsian Veil of Ignorance—that his aim of maximizing the prospects for his own survival and basic well-being will be furthered by a categorical repudiation of those torturous methods. Moreover, to address the possibility of errors in the employment of such methods of interrogation, the parties to the social contract can insist on very high epistemic standards for any occasion when those methods are applied. In other words, they can insist that such methods should never be administered unless the interrogators warrantedly believe beyond a reasonable doubt that the person being questioned is non-innocently possessed of information that can avert a calamity. The parties will have solid grounds for thinking that this stringent epistemic standard can sometimes be met, as in the circumstances of Leon v. Wainwright. Hence, they will not have any grounds for adopting an absolute ban on the use of calamity-averting interrogational torture. (Of course, the parties will impose conditions for the legitimacy of such torture beyond the one epistemic condition broached here. For example, they will insist that all reasonably available non-torturous methods of interrogation be essayed before the police ever resort to torturous methods. Again, however, they will have solid grounds for thinking that the additional conditions can sometimes be satisfied in real-world contexts.) In short, the second posited rejoinder to my critique of Meisels is no more successful than the first such rejoinder. Her contractarian approach, under which each of the parties to the social contract is striving behind the Veil of Ignorance to maximize his own chances of surviving and flourishing, will not support an absolute moral prohibition on calamity-averting interrogational torture. It will underpin a far-reaching prohibition on the use of such torture, but it will not underpin a prohibition that applies always and everywhere. Meisels herself acknowledges as much—as we have seen—despite her aspiration to justify an absolute moral ban on interrogational torture. Thus, since one of the chief objectives of this chapter (and of this book) is to explain why such torture is always and everywhere morally wrong, we shall have to leave Meisels’s contractualism by the wayside. Even if we pretermit the more wide-ranging doubts which I have

Some Previous Accounts of the Wrongness of Torture

127

raised elsewhere about contractualist doctrines, those doctrines cannot serve the objectives of this chapter.

3.1.2.  Shue on the defencelessness of victims In a classic article on torture published three-and-a-half decades ago, Henry Shue has maintained that a key to the distinctive wrongness of torture resides in the defencelessness of the victims thereof.4 He is particularly concerned to explain why torture is so flagrantly wrong even though deliberate killings—which, as he accepts, cause more harm to their victims than do quite a few methods of torture—are sometimes morally permissible, especially in military conflicts. Although I shall argue later that the solution to a closely connected puzzle about torture and killing lies in a shift from a victim-focused perspective to a perpetrator-focused perspective, Shue retains a firmly victim-focused perspective in his grappling with the problem. He submits that the special wrongness of torture arises from its unfairness. Specifically, torture differs from intentional killings among armed combatants in that it does not even approximate the conditions of a ‘fair fight’. Because the torturer enjoys total dominance over the victim, the victim is helpless to fend off the torturer’s onslaughts. Though warring combatants are frequently mismatched in their strength and resources, their situation is not marked by the radical disparity of power that obtains in a situation of torture. So Shue contends, as he declares that the gross unfairness of that disparity is what accounts for the unique heinousness of torture: At least part of the peculiar disgust which torture evokes may be derived from its apparent failure to satisfy even this weak constraint of being a ‘fair fight.’ The supreme reason, of course, is that torture begins only after the fight is—for the victim—finished. Only losers are tortured. A ‘fair fight’ may even in fact already have occurred and led to the capture of the person who is to be tortured. But now that the torture victim has exhausted all means of defense and is powerless before the victors, a fresh assault begins. The surrender is followed by new attacks upon the defeated by the now unrestrained conquerors. In this respect torture is indeed not analogous to the killing in battle of a healthy and well-armed foe; it is a cruel assault upon the defenseless. . . . It is in this respect of violating the prohibition against assault upon the defenseless, then, that the manner in which torture is conducted is morally more reprehensible than the manner in which killing would occur if the laws of war were honored. In this respect torture sinks below even the well-regulated mutual slaughter of a justly fought war. (Shue 1978, 130)

3.1.2.1.  Putting aside an issue In the elided portion of the passage above, indicated by the ellipses, Shue runs together the properties of defencelessness and unthreateningness—as he lurches 4   Shue contemplates only a few kinds of torture, and does not take any account of ephemerally incapacitative torture or edifying torture (each of which can be instantiated in morally legitimate ways). Thus, my discussion of him in this section will likewise disregard those latter two types of torture.

128

Why Torture is Wrong

from maintaining that victims of torture are defenceless to maintaining that they are unthreatening. My second chapter has adverted at more than one juncture to Shue’s conflation of those properties, and has impugned his contention that victims of torture are always unthreatening. What can be added now is that his ascription of unthreateningness to victims of torture is an outright distraction. Though the threateningness or unthreateningness of a victim of torture is patently of moral significance, the moral issue posed by it is distinct from the moral matter of fairness on which Shue is concentrating his attention; a captured terrorist can be highly threatening even if he is utterly unable to defend himself against being tortured or killed. Thus, because the argument about fairness is my principal concern at present—since that argument is the means by which Shue seeks to delineate the distinctive wrongness of torture—I shall not here dwell afresh on Shue’s ill-advised remark about the unthreateningness of captured terrorists. Let us simply note the following point. As Chapter  2 has observed, the only type of torture that has recently been advocated by any estimable philosophers in Western countries is calamity-averting interrogational torture. Most instances of such torture involve victims who are threatening, in that those victims are declining to disclose information which—if divulged—could help to forestall some disastrous consequences. (The main exceptions, apart from instances of calamity-averting interrogational torture that are mistakenly applied to people who are not apprised of any information that might help to prevent disastrous consequences, are instances of such torture that are applied to close relatives or friends of people from whom vital information is sought. Though the indirect victims of the latter instances are threatening, the direct victims are not.) Accordingly, anyone who joins Shue in singling out the unthreateningness of victims as the central wrong-making property of torture is failing to provide a basis for ascribing wrongness to most instances of the sole type of torture that has been recommended by any respectable philosophers in recent years.

3.1.2.2.  On defencelessness as the wrong-making property of torture Of course, my second chapter has not only contested Shue’s assumption that victims of torture are always unthreatening. Additionally, and more lengthily, it has contested his assumption that such victims are always defenceless. In § 2.1.4.1.1 we have pondered a couple of scenarios, advanced respectively by Frances Kamm and Uwe Steinhoff, in which some victims of torture are able to defend themselves. Admittedly, the circumstances limned in those scenarios are decidedly unusual (though not utterly fanciful). In the vast majority of situations where people are tortured, they are helpless to ward off the torturers’ afflictions. Still, in some possible situations that are not risibly far-fetched, torture can be wielded against people who are not helpless to defend themselves. Hence, when Shue insists that the key wrong-making property of torture resides in the defencelessness of victims, one reason for demurring is that that very feature is absent from some possible instances of torture that are nonetheless wrong.

Some Previous Accounts of the Wrongness of Torture

129

Moreover, even if every victim in every credibly possible situation of torture were helpless to defend herself, Shue’s argument about defencelessness and unfairness would not withstand scrutiny. Some of the chief shortcomings in his argument have been ably highlighted by Kamm (2011, 5–9) and Steinhoff (2006, 337–8) separately. Of most importance here is that the utter defencelessness of some person or group of people does not always undermine the legitimacy of a lethal attack against the person or group of people. Let us ponder a slightly modified version of an example adduced by Steinhoff (2006, 338). Suppose that the soldiers of Army X are engaged in a just war against the soldiers of Army Y. The weaponry of the former army is markedly superior to that of the latter. As the soldiers from Army X advance toward the trenches in which the soldiers of Army Y are ensconced, they send some clear-cut warnings. Specifically, they warn that they will send a long-range artillery barrage against the soldiers of Army Y unless the latter soldiers have surrendered by a certain time. The soldiers of Army Y will be defenceless against such a long-range barrage of devastating munitions, since they lack any armaments of a comparable range, and they lack any means of deflecting the munitions effectively. Nevertheless, the soldiers of Army Y adamantly refuse to surrender (perhaps because they believe surrender to be dishonourable). Were they to be left in their trenches without being attacked, they would pose a mortal danger to the soldiers of Army X as the latter soldiers advance. Consequently, the soldiers of Army X unleash the long-range barrage about which they have repeatedly warned. Given the justness of their military campaign against Army Y, and given their provision of adequate warnings to the soldiers of Army Y and adequate opportunities for those latter soldiers to surrender, and given the absence of any feasible alternatives that would not involve either the abandonment of a key military objective of Army X or the imperilling of X’s fighters, the long-range attack against Army Y is morally legitimate. In the specified circumstances, the defencelessness of the soldiers of Army Y against the long-range shelling is not per se a wrong-making property of the shelling. A situation of calamity-averting interrogational torture can be relevantly similar to the situation of Army X and Army Y. Suppose that a liberal democracy is engaged in a just campaign against an odious terrorist network, and that a key objective in the campaign is to thwart the perpetration of an atrocity which the members of that network are known to be on the verge of carrying out. Suppose now that a terrorist who has participated in the final steps leading up to the atrocity is apprehended. If the other elements of a suitably realistic version of the ticking-bomb scenario are in place (including the intransigence of the captured terrorist when subjected to ordinary techniques of interrogation and persuasion), then the defencelessness of the terrorist against torturous methods of interrogation is morally on a par with the defencelessness of the soldiers of Army Y.  That is, his defencelessness does not in itself render the torturous methods morally illegitimate. If such methods are indeed morally illegitimate even in the circumstances envisioned here—as this chapter will be maintaining—their wrongness derives from something other than what Shue has suggested.

130

Why Torture is Wrong

3.1.2.3.  A loss of one’s ideals? Shue does not ponder any objection of the sort which I have just raised against his argument about defencelessness and unfairness. However, he mulls over a different objection (1978, 130–7). He notes that, in a situation of interrogational torture where the victim is possessed of the verifiable information which the interrogators are seeking, she can bring an end to her ordeal by divulging the information. Shue rightly poses certain queries about practicalities—concerning mixed motives on the part of torturers and the possibility of innocent victims, for example—but he allows that sometimes a victim of interrogational torture can bring a halt to the proceedings by disclosing what she knows. Indeed, with regard to any victim who is not very fervently committed to the cause which she would betray by revealing the information that is sought from her, Shue (1978, 134) concedes that ‘a betrayal of cause and allies might indeed serve as a form of genuine escape’. Oddly, he then omits to indicate whether or not the torture perpetrated against such a victim is morally legitimate. Given the concession just quoted, which implies that his concerns about helplessness are inapplicable to such a victim, he appears to be obliged to accept that the torture used against her is morally legitimate if it has been undertaken to avert a calamity. Once again, then, we find that Shue’s account of torture’s wrongness fails to come up with any absolute moral prohibition on interrogational torture. With regard to any victim who is fervently committed to the cause that would be betrayed by her imparting what she knows, Shue proceeds quite differently. He submits that the ability of such a victim to terminate her ordeal by divulging her information does not amount to a genuine route of escape: [W]‌hen the torturers succeed in torturing someone genuinely committed to the other side, compliance means, in a word, betrayal; betrayal of one’s ideals and one’s comrades. The possibility of betrayal cannot be counted as an escape. Undoubtedly some ideals are vicious and some friends are partners in crime—this can be true of either the government, the opposition, or both. Nevertheless, a betrayal is no escape for a dedicated member of either a government or its opposition, who cannot collaborate without denying his or her highest values. For any genuine escape must be something better than settling for the lesser of two evils. One can always try to minimize one’s losses—even in dilemmas from which there is no real escape. But if accepting the lesser of two evils always counted as an escape, there would be no situations from which there was no escape, except perhaps those in which all alternatives happened to be equally evil. On such a loose notion of escape, all conscripts would become volunteers, since they could always desert. And all assaults containing any alternatives would then be acceptable. An alternative which is legitimately to count as an escape must not only be preferable but also itself satisfy some minimum standard of moral acceptability. A denial of one’s self does not count. (Shue 1978, 135–6, footnote omitted)

This passage is deeply problematic in a number of respects. 3.1.2.3.1.  A query On the one hand, the second half of the passage is of course correct in contending that the choice made by a victim of interrogational torture to disclose some vital information is a coerced choice. On the other hand, however, the matter of the

Some Previous Accounts of the Wrongness of Torture

131

voluntariness or coercedness of the disclosure bears centrally on the status of the victim’s ordeal as torture (as we have seen in § 2.3.3.1 of Chapter 2) rather than on the wrongness of that ordeal. Although coercion can of course be wrong in many contexts, it is by no means always so. For example, when the prospect of legal penalties has coerced a would-be malefactor into abstaining from the commission of violent misdeeds against other people, the coercedness of his abstention is not a wrong-making feature thereof. Likewise, the sheer fact that a terrorist has been coerced into divulging some calamity-averting information is not in itself sufficient to render the treatment of him morally illegitimate. If the torture used against him is indeed morally wrong—as this chapter will maintain—its wrongness stems not from its coerciveness but from its torturousness as a means of compelling him to perform a certain action of disclosure. 3.1.2.3.2.  Another query Even more troubling in the long excerpt from Shue’s article is his insistence that the abandonment of a victim’s allegiances cannot count as a genuine route of escape. Let us keep in mind two points as we assess that insistence. First, the question whether the abandonment of allegiances can be a genuine route of escape is a normative matter rather than a psychological matter. It pertains to the moral import, rather than to the psychological joltingness, of such a route of escape. Second, if Shue’s claims about the abandonment of allegiances are to shore up his position in a blanket fashion, they have to apply to the situation of a terrorist whose allegiance is to a fascistic and malevolently mass-murdering organization. Let us suppose that a terrorist of the sort just mentioned is in possession of calamity-averting information when she is taken into custody by officials from a liberal-democratic regime. Suppose further that the torturous interrogation of the terrorist will cease whenever she imparts that information to her captors. Shue needs to claim that the terrorist’s divulgence of the specified information will be ethically dubious (rather than just psychologically difficult, as it might well be). Yet a calamity-averting disclosure by the terrorist—with the betrayal of her mass-murdering comrades that will be entailed by such a disclosure—is not ethically dubious at all. It is both morally permissible and morally obligatory. The mass-murdering fascists in the terrorist’s organization have no genuine moral rights to her loyalty, regardless of what they might think about the matter. She does nothing wrong whatsoever when she derails their planned atrocity by revealing what she knows, and she thereby acts in compliance with a potently stringent moral obligation. Shue, in the long quoted passage, acknowledges that some ideals and affiliations are vile; yet he denies that the abandonment of such ideals and affiliations is a veritable means of escape from a process of torture, because any such abandonment is a ‘denial of one’s self’. Shue appears to be conflating the psychological matter and the normative matter between which I have distinguished. As a moral matter, the repudiation of a hideous aspect of oneself is exactly what is permissible and obligatory. Psychologically wrenching though it might be—at least initially—the shedding of one’s viciously immoral attachments is not ethically untoward in any way.

132

Why Torture is Wrong

Indeed, in a footnote, Shue himself reluctantly concedes as much (1978, 136 n14). He allows, albeit very grudgingly, that ‘there are probably rare individuals so wicked as to lack integrity, or anyway to lack any integrity worthy of respect’. He accepts that there is no satisfactory rejoinder to the following rhetorical question: ‘[W]‌hat sort of integrity could one have violated by torturing Hitler?’ To be sure, he straightaway goes on to remark that ‘[a]ny very slight qualification here must not, however, be taken as a flinging wide open of the doors . . . Unfortunately, I  cannot see a way to delimit those who are genuinely beyond the pale which does not beg for abuse.’ Still, having conceded that there are some people whose monstrous goals and affiliations are ‘genuinely beyond the pale’, Shue has in effect acknowledged that the wrongness of any interrogational torture employed against such people is not attributable to their being forced to betray their own ideals. In sum, if anyone seeks to maintain that interrogational torture employed against odious terrorists is indeed morally wrong, then once again he or she cannot rest content with Shue’s account of the wrong-making properties of torture.

3.1.3.  Shue and others on the spread of torture As we have just seen, Shue worries that any concessions concerning the use of torture are liable to be abused. His anxiety on this point is in fact a salient feature of his 1978 article, for—despite his arguments about defencelessness and unfairness and self-betrayal—he does not believe that interrogational torture and intimidatory torture (the only two types of torture that he discusses) are always and everywhere morally wrong in all possible worlds.5 He accepts that, in imaginably extreme circumstances where the use of intimidatory torture or interrogational torture is indispensable for the avoidance of a cataclysm, such a course of conduct would be morally permissible. However, he is keenly concerned to discount the importance of his concession. In addition to emphasizing the prodigious improbability of the emergence of any real-world situation with all the elements of a ticking-bomb scenario, he highlights the perniciousness of the metastasizing effects that are very likely to follow from any recourse to torture. The spectre of those effects will tilt the balance against the permissibility of the employment of torturous techniques in any set of circumstances where the calamity to be averted through such techniques is less than a horrendous catastrophe. Indeed, if empirical studies establish that the probability of the spread or institutionalization of torturous measures through any use of those measures is very high, the balance will be tilted decisively against the permissibility of intimidatory torture in every situation in the actual world (Shue 1978, 140). Though Shue in his 1978 article does not similarly claim that interrogational torture will always and everywhere be morally wrong in the actual world, he does contend that it should always be legally proscribed.

  The final section of his 1978 article is entitled ‘Morally Permissible Torture?’

5

Some Previous Accounts of the Wrongness of Torture

133

3.1.3.1.  A weak argument Shue expresses his anxiety about the spread or institutionalization of torture more stridently in a recent article. He asserts that the situation depicted in a ticking-bomb scenario is ‘like an imaginary alcoholic who drinks two beers only a night’. He elaborates: There are former alcoholics, who do not drink at all, and active alcoholics. To think that there may be rare alcoholics who drink moderately is to fail to understand alcoholism. Similarly, history does not present us with a government that used torture selectively and judiciously. Are we to believe that America is likely to be the first alcoholic in history who can take only one drink? The first state apparatus that will use torture judiciously and selectively? Are American politicians so superior to mere mortals? (Shue 2006, 234, footnote omitted)

This argument by Shue is uncharacteristically weak. Of course a person is not a currently active alcoholic unless he or she drinks alcohol-containing beverages immoderately on a number of occasions, but that obvious point—which is indeed an analytic truth—does not shed any light on the controversies over torture. No participant in those controversies has ever denied anything akin to that analytic truth. Analogous to several of the claims in those controversies, rather, would be the proposition that some erstwhile alcoholics can drink in moderation under certain conditions or the proposition that some people who are predisposed toward alcoholism can nonetheless drink in moderation under certain conditions. Each of those latter propositions is a substantive empirical contention that is pertinently similar to contentions put forward by quite a few of the philosophers and jurists who favour the use of calamity-averting interrogational torture in extreme situations. Many such philosophers and jurists are inclined to argue that the use of torturous methods of interrogation can be confined to those extreme situations if proper safeguards are in place. We shall probe some of their arguments in Chapter 5. For the moment, we should simply note that the conclusions which they seek to sustain are similar in germane respects to the proposition about erstwhile alcoholics or the proposition about people who are predisposed toward alcoholism. Those conclusions are not comparable, in any germane respects, to a denial of the analytically true proposition about currently active alcoholics. Yet Shue assails his opponents by suggesting that their position resembles a denial of that analytic truth. Perhaps I have misinterpreted the second sentence in the passage quoted above. Perhaps Shue is there referring not to currently active alcoholics, but instead to former alcoholics or to people predisposed toward alcoholism. If so, then he is engaging with his opponents; he is not arrantly talking past them. However, he is also advancing a highly contestable empirical claim for which he provides no evidence whatsoever. Moreover, he is presenting that highly contestable empirical claim as a matter of understanding the nature of alcoholism. (Precisely because his assertion appears to be framed as a conceptual thesis rather than as an empirical thesis, I have inferred that it pertains to currently active alcoholics.) In short, regardless of how we choose between the eligible interpretations of the opening portion of the passage which I  have quoted here, that portion does not withstand critical investigation.

134

Why Torture is Wrong

Later in that passage, where Shue focuses directly on torture (though still with reference to the analogy concerning alcoholism), his remarks are less confusing but no more persuasive. By asking rhetorical questions about the use of torture in the United States, he is patently making an empirical claim rather than a conceptual claim. He is insisting that, in light of the experiences of other systems of governance that have made use of torture, we can be confident that any employment of torture by public officials in the United States will get out of hand. Irrespective of the limits that are putatively imposed on the range of circumstances in which the officials will resort to torture, and irrespective of the sincerity of the officials in resolving to abide by those limits, the pressures of desperate emergencies will lead the officials to avail themselves of torture in an ever-widening array of situations. Near the end of his 2006 article, Shue summarizes his stance on this matter as he returns to the analogy concerning alcoholism: ‘You cannot be a little bit pregnant, you cannot—if you are an alcoholic—have a drink only on special occasions, and you cannot—if your politicians are not angels—employ torture only on special occasions’ (2006, 238). Shue recognizes that he is here making some large empirical claims. As he avows: ‘I am certainly not appealing to conceptual necessities, but instead [am] making empirical judgments about the nature of human beings and the effects of power without accountability, secret torture hidden from the citizens in whose name it is inflicted’ (2006, 235). What is disconcerting about his discussion is the dearth of evidence offered in support of his empirical contentions. A fleeting reference to Abu Ghraib and Guantánamo Bay—and to a couple of Eastern European countries—is the only attempt by Shue in his 2006 article to substantiate his far-reaching empirical assertions. In that article, he makes no effort to adduce an example of a system of governance whose officials have initially resorted ad hoc to calamity-averting interrogational torture in some extreme emergency and have then begun to use torture more and more often in non-emergency situations. Still, although the evidence provided by Shue to back up his sweeping empirical contentions is exceedingly wispy, he does present a new line of reasoning in his 2006 article. With that line of reasoning, he endeavours to show that instances of torture in situations of dire desperation cannot realistically be isolated instances. He presumes to have exposed the fancifulness of the idea that interrogational torture can be undertaken in an ad hoc fashion for the avoidance of calamities without any institutionalization of the practice. If he is correct, his failure to substantiate his empirical conjectures with evidence is of minor importance at most. We should therefore now turn to his new line of argument.

3.1.3.2.  Another weak argument According to Shue in his recent article, a scenario of successful ad hoc interrogational torture is preposterous because anyone who resorts to such a desperate course of action on an ad hoc basis ‘would probably . . . be an incompetent torturer, most unlikely to succeed at what might well be his first try at extracting information’. From that point of departure, Shue proceeds to argue that the institutionalization

Some Previous Accounts of the Wrongness of Torture

135

of torture is inevitable under any regime whose officials are to make effective use of torturous methods of interrogation in dreadful emergencies: Successful torturers must avoid sympathy and empathy, or they will go too easy. But they must also avoid anger and cruelty, or they will go too hard and merely knock the victim senseless, or drive him into a dissociative state, and learn nothing useful for the prevention of catastrophe. Torture is not for amateurs—successful torturers need to be real ‘pros,’ and no one becomes a ‘pro’ overnight. At a minimum, one must practice—perhaps do research, be mentored by the still more experienced. In short, torture needs a bureaucracy, with apprentices and experts, of the kind that torture in fact always has. Torquemada was not an independent consultant. Torture is an institution. (Shue 2006, 236)

In support of his argument, Shue draws upon the work of Jean Maria Arrigo. He quotes a passage from Arrigo that should be reproduced in part here: The use of sophisticated torture techniques by a trained staff entails the problematic institutional arrangements I have laid out: physician assistance; cutting edge, secret biomedical research for torture techniques unknown to the terrorist organization and tailored to the individual captive for swift effect; well trained torturers, quickly accessible at major locations; pre-arranged permission from the courts because of the urgency; rejection of independent monitoring due to security issues; and so on. These institutional arrangements will have to be in place, with all their unintended and accumulating consequences, however rarely terrorist suspects are tortured.6

3.1.3.2.1.  A first reply To Shue’s new line of reasoning, two closely related (and mutually reinforcing) replies are apposite. First, as is made clear by the adjective ‘sophisticated’ near the beginning of the passage from Arrigo, the arguments quoted here are about refined and technologically arcane methods of torture. Perhaps the institutionalization of preparations for torture is necessary under a regime whose officials wish to avail themselves of such methods, but there is no basis for the proposition that policemen or intelligence officials in a terrible emergency will have to use sophisticated techniques of torture to overcome the defiant silence of a captive who possesses information that can help to avert a calamity. In the events that led up to Leon v. Wainwright, for example, the policemen twisted the arm and throttled the neck of the kidnapper whom they had apprehended. Their methods were decidedly unsophisticated but were effective in the desperate circumstances. Much the same is true of the Antipodean scenario to which I have referred briefly in § 2.2.1.3 of Chapter 2. In the predicament recounted in that scenario, the recourse of the police to interrogational torture consisted in their punching of a suspect. (In the German case to which I have adverted at a few junctures in Chapter 2, the police did not

6   Arrigo 2004, 564. A  condensed version of Arrigo’s arguments is presented in Bufacchi and Arrigo 2006, 362–7. See also Wisnewski 2010, 100, 125 n20, 146–8. For an assessment of such arguments that partially prefigures my own assessment (albeit from a very different perspective overall), see Allhoff 2012, 147–54.

136

Why Torture is Wrong

have to use any methods of torture at all; a credible threat of torture was sufficient to extract information from the captured suspect.) Although bureaucratic and technologically advanced preparations for torture might be crucial under regimes that employ torture frequently and prolongedly for the consolidation of their own repressive dominance, neither Shue nor Arrigo supplies any grounds for thinking that officials in liberal-democratic regimes who are faced with quandaries of great urgency will likewise need such preparations. The practices of torture which Shue and Arrigo discuss—including the torture perpetrated or arranged by the United States in its campaign against Islamist terrorism during the first decade of the twenty-first century—have virtually nothing to do with the prevention of calamities in ticking-bomb emergencies. Though most of those practices of torture were interrogational, they were not of the calamity-averting variety. Hence, powerful though Shue’s and Arrigo’s accounts of the bureaucratic/scientific infrastructure of torture are, those accounts do not really address the question which Shue purports to have answered negatively. That is, they do not really address the question whether instances of calamity-averting interrogational torture in circumstances of extreme desperation can remain isolated. 3.1.3.2.2.  A second reply A closely related reason for querying the arguments propounded by Shue and Arrigo is that, even in a regime which firmly eschews any institutionalized preparations for the use of interrogational torture, many of the officials will be proficient in various pain-inducing techniques. For one thing, insofar as military and intelligence and policing organizations provide resistance training for some of their members, the people who administer the training will develop expertise in sundry methods of inflicting severe pain. Even more important, numerous ordinary aspects of policing and military operations can be drawn upon by officials who have to resort to calamity-averting interrogational torture in grim emergencies. Constables and soldiers are not naïfs who lack any acquaintance with the use of violence. On the contrary, constables are routinely trained in procedures for apprehending and restraining suspects—procedures that can involve their overcoming the suspects with force if necessary. The manoeuvres which they learn for those tasks (and for other tasks such as crowd control), including the precautions which they learn for preventing their applications of violence from becoming lethal, are quite readily transferable to interrogational settings. Indeed, that very transferability is one chief reason why the employment of violent methods of interrogation was routine in past decades even in liberal-democratic countries. In other words, quite a few of the skills routinely acquired by constables and soldiers in their training and careers can be put to use for the general purpose of afflicting victims with agony during sessions of interrogation. Contrary to what might be inferred from Shue’s and Arrigo’s arguments, many of the ordinary and legitimate techniques of policing or soldiering are not very sharply distinct from some of the sinister techniques of torturing. Thus, another reason for impugning those arguments is that public officials in the absence of any institutionalized

Some Previous Accounts of the Wrongness of Torture

137

preparations for torture are not nearly as ill-equipped to engage in calamity-averting interrogational torture as Shue and Arrigo suggest. Indeed, the adjective ‘ill-equipped’ has signalled a way in which the observations in the preceding couple of paragraphs should be extended. As I have mentioned fleetingly in § 2.1.1.2 of Chapter 2—with reference to Elaine Scarry’s lengthy discussion of torture, where this point is recurrently highlighted—countless mundane items, such as furniture and appliances and cutlery, can be wielded by torturers for the purpose of afflicting their victims with agony. This point applies to diverse routine implements used by constables or soldiers. For example, Shue points out that the United States in the early 2000s was the largest supplier of devices such as stun guns and electric-shock batons to regimes that use those devices for torture (Shue 2006, 237 n23). Those devices are now widely used for potentially legitimate purposes in the ordinary operations of policing in most Western countries. In equipment as well as in skills, then, constables and soldiers can be well positioned to resort to torture in extreme emergencies without any institutionalization of the practice. Admittedly, and fortunately, policemen in their quotidian roles in any liberal democracy will very likely not possess quite the same level of expertise in pain-inducing techniques as would professional torturers. Furthermore, their endeavours in extreme emergencies might fail—just as the endeavours of professional torturers might fail. Still, what the remarks in my last few paragraphs indicate is that Shue has damagingly oversimplified the matters under discussion in passages such as the following: Either ‘torturers’ are just thugs who have no clue what they are doing, in which case we need not allow for exceptional cases in which they rapidly and effectively extract invaluable catastrophe-preventing information, or some can have genuine expertise . . . If expertise is available, we would certainly want it at hand in the exceptional, potentially catastrophic case. If we want it ready, we need to maintain, even nourish, the organizations and networks in which the expertise resides. (Shue 2006, 237)

Despite the confident tone of this passage, the dilemma which it outlines is simplistic. When constables or soldiers in Western countries desperately have recourse to torturous methods of interrogation in grave emergencies after the non-torturous methods available to them have failed, they are neither addle-headed thugs nor professional torturers. Drawing on procedures and implements with which they have become proficient in the course of fulfilling their everyday responsibilities, they have gained adequate expertise without any pernicious institutionalization of the practice of torture. Possessing that expertise, and possessing serviceable equipment, they can come to grips reasonably well with a dire emergency on an ad hoc basis. The effective administration of calamity-averting interrogational torture in such an emergency can proceed without any grounding in organizations and networks of the kind which Shue envisages.

3.1.3.3.  Empirical speculations and slippery slopes Shue is by no means the only philosopher or jurist who has indulged in empirical conjectures and who has worried about the institutionalization of torture that will

138

Why Torture is Wrong

putatively be engendered or presupposed by any use of torturous techniques in an extreme emergency. Similar conjectures and worries abound in the contemporary literature on torture. Of interest in the present context are the worries voiced by theorists whose positions are relevantly similar to Shue’s. That is, the general notion under scrutiny here is that the wrongness of calamity-averting interrogational torture is due to the likelihood that any recourse to torturous measures by public officials will foster the spread of such measures. In other words, the doctrine under examination here is a concern about slippery slopes. It is a fundamentally consequentialist concern. Proponents of that doctrine (such as Shue) subscribe to the substance of the following thesis: Slippery-Slope Illegitimacy Thesis. Although the use of calamity-averting interrogational torture as a last resort in a dreadful emergency is morally legitimate in principle, any such use is wrong in practice because of the high probability that it will lead to the plying of torturous techniques of interrogation in situations where other options are available or where the stakes are too low to warrant such desperate steps.

A supporter of this thesis believes that, if public officials avail themselves of any torturous methods of interrogation as a last resort in some grave emergency, they will become inclined to employ such methods in a host of other contexts as well. Instead of remaining a last-ditch measure of desperation, interrogational torture will become well entrenched in many operations of policing and soldiery. Given as much, and given the consequentialist premise that the moral status of any mode of conduct is determined by its full array of consequences, calamity-averting interrogational torture is always morally illegitimate in practice—even though it would be morally legitimate if the metastasizing tendency of torture were less strong. Such is the position taken by quite a number of the contributors to the disputes over interrogational torture that have raged in Western countries during the past few decades. 3.1.3.3.1.  Two clarifications The position just delineated is quite different from the following deontological doctrine: Deontological Intensification Thesis. Calamity-averting interrogational torture is morally wrong irrespective of its consequences, but any consequences of it that are deleterious— such as the metastasizing spread of torture—intensify the gravity of its wrongness insofar as they have been likely to materialize.

While such a doctrine accepts that the probability or improbability of deleterious consequences will affect the overall moral bearings of any instance of calamity-averting interrogational torture that is undertaken, it stoutly denies that the probability or improbability of those consequences will affect the basic moral status of the torture as illegitimate. Although the magnitude of the wrong constituted by any act of such torture is partly a function of the aforementioned consequences, the sheer wrongness of that act is not; the sheer wrongness obtains whether or not those consequences materialize and whether or not they have been foreseeable. (As

Some Previous Accounts of the Wrongness of Torture

139

is probably evident, the deontological position sketched here is—in a nutshell—an element of the position espoused by this book.) In contrast with the Deontological Intensification Thesis, the Slippery-Slope Illegitimacy Thesis contends that the very wrongness of calamity-averting interrogational torture is a product of the metastasizing tendency of such torture. Were there no such metastasizing tendency, the wielding of calamity-averting interrogational torture as a last resort in circumstances of formidable desperation would be morally permissible—or so the proponents of the Slippery-Slope Illegitimacy Thesis think. Now, those proponents will not be engaging with their adversaries unless they are making a point specifically about calamity-averting interrogational torture rather than about other kinds of torture. That is, we should take them to be holding that a metastasizing tendency is associated specifically with the use of calamity-averting interrogational torture in dire emergencies. After all, nobody doubts that the employment of various other types of torture by public officials can readily spread through a system of governance. Nor does anyone doubt that the employment of various other types of torture by those officials can detrimentally affect many operations of government in profound ways. Hence, what is at issue here is not a claim about torture in general but instead a claim peculiarly about calamity-averting interrogational torture. We are examining the proposition that the use of such torture cannot plausibly be ad hoc. 3.1.3.3.2.  An additional clarification A further point of clarification is advisable. On the one hand, the consequentialist Slippery-Slope Illegitimacy Thesis is different from the following consequentialist doctrine: Legal Prohibition Thesis. Any use of calamity-averting interrogational torture should be legally prohibited ex ante—to prevent the metastasizing effects that would arise from ex ante legal authorizations—even though such torture in this or that terrible emergency can be morally legitimate in practice as well as in principle.

This latter thesis is not about the reasons why calamity-averting interrogational torture is always morally impermissible. On the contrary, the thesis accepts that such torture can be morally legitimate when it is administered in certain circumstances. Rather, as the name suggests, the Legal Prohibition Thesis prescribes how such torture should be legally regulated (whether any particular instance of it is morally legitimate or not). Richard Posner pithily articulates the gist of this position:  ‘If legal rules are promulgated permitting torture in defined circumstances, officials are bound to want to explore the outer bounds of the rules; and the practice, once it were thus regularized, would be likely to become regular.’7 Now, although the final chapter of this book will ponder the legal consequences 7   Posner 2004, 296. For some other writings in which the substance of the Legal Prohibition Thesis is embraced, see Gross 2004a; Gross 2004b; Gross 2008; Gur-Arye 2004; Kadish 1989; Meisels 2008a; Meisels 2008b, chaps 6–7; Parry 2004; Parry and White 2002; Twining and Twining 1973, 348–50.

140

Why Torture is Wrong

that should attach to instances of torture, we are not at present mulling over that matter. Instead, this third chapter is seeking to pin down why calamity-averting interrogational torture—along with nearly every other type of torture—is always morally wrong. Hence, we are not at present exploring the main recommendation of the thesis that has been outlined in this paragraph. On the other hand, although the Legal Prohibition Thesis is not directly the target of my present discussion, many of the arguments propounded by its advocates are serviceable as well for the Slippery-Slope Illegitimacy Thesis which is the target of this discussion. This point might initially seem puzzling because, unlike the champions of the Legal Prohibition Thesis, the supporters of the Slippery-Slope Illegitimacy Thesis believe that calamity-averting interrogational torture is always morally illegitimate on consequentialist grounds. While the proponents of the two theses converge in submitting that such torture should always be legally forbidden, they disagree on the question whether the use of such torture (though legally forbidden) is morally permissible in some circumstances of grave urgency. Adherents of the Slippery-Slope Illegitimacy Thesis contend that calamity-averting interrogational torture in any of its applications is likely to send us down a slippery slope and is therefore always morally impermissible in practice. By contrast, adherents of the Legal Prohibition Thesis contend that the employment of such torture as a last resort in terrible emergencies can be morally permissible in practice if it is legally proscribed ex ante, for they believe that the legal ban will typically be sufficient to keep us from sliding down a slippery slope. Still, notwithstanding the divergence between their views concerning the upshot of a legal prohibition on calamity-averting interrogational torture, the supporters of the two theses concur in thinking that the absence of any such prohibition will give rise to a very serious danger of a tumble down a slippery slope. Because of their concurrence on that concern, a lot of the claims and arguments propounded by the votaries of the Legal Prohibition Thesis are broadly similar to some of the claims and arguments propounded by the votaries of the Slippery-Slope Illegitimacy Thesis. Accordingly, a critique that impugns the former claims and arguments will also have impugned the latter. After all, the champions of the Legal Prohibition Thesis are advancing milder claims about the operativeness of slippery slopes than are the champions of the Slippery-Slope Illegitimacy Thesis. In other words, they are contending that slippery slopes are operative in a narrower range of situations than is envisaged by the advocates of the latter thesis. Hence, if their milder claims about slippery slopes are dubious in certain respects, then a fortiori the bolder claims advanced by the supporters of the Slippery-Slope Illegitimacy Thesis are dubious. 3.1.3.3.3.  A first example Let us glance at a couple of examples of theorists who have subscribed to the substance of the Slippery-Slope Illegitimacy Thesis. These theorists deem calamity-averting interrogational torture to be always morally wrong, and—as consequentialists—they attribute its wrongness to its metastasizing tendency. At least when voicing such a position, they do not join the adherents of the Deontological

Some Previous Accounts of the Wrongness of Torture

141

Intensification Thesis in affirming that calamity-averting interrogational torture is morally wrong irrespective of its consequences. Seth Kreimer is one of the advocates of the Slippery-Slope Illegitimacy Thesis (though, like the other advocates, he of course does not use my appellation for the thesis to which he subscribes). He distances himself from any deontological assessment of calamity-averting interrogational torture as inherently wrong: ‘Indeed, if torture could be confined to a limited situation in which it represented the only way of preventing vast devastation, there are not inconsiderable arguments that, as a matter of both policy and public ethics, an official might choose to violate legal and otherwise peremptory moral norms by ordering such torture’ (2003, 318). Kreimer is a legal scholar rather than a philosopher, and his wording here is regrettably imprecise. He speaks of an official’s violation of moral norms, and he almost conveys the impression that he is discussing a matter of individual psychology— namely, what an official might choose to do. However, when we note that Kreimer characterizes an official’s choices as ‘a matter of . . . public ethics’, and when we further note that he refers to the supposedly violated moral norms as ‘otherwise peremptory’, his rejection of a deontological stance on interrogational torture is quite clear. He thinks that, if the use of calamity-averting interrogational torture as a last resort in an extreme plight were not likely to metastasize, the moral prohibition on such torture would then be overridden (rather than merely overtopped). That moral prohibition would not be binding in such a situation. Although Kreimer’s conditional formulation presents the confinability of calamity-averting interrogational torture as a sufficient condition for the moral legitimacy of such torture, we should construe his conditional as a biconditional that also presents the confinability of such torture as a necessary condition for its moral legitimacy. After all, given the reluctance of the phrasing with which Kreimer expresses his concession in the quoted sentence, he would surely not accept that calamity-averting interrogational torture can be morally permissible even when the use of such torture will send us hurtling down a slippery slope. Non-philosophers, unattuned to the niceties of logical connectives, often frame statements as conditionals when they intend those statements to be understood as biconditionals; the most plausible way to read the quoted sentence from Kreimer, in the context of his overall argument, is to conclude that he has committed such a misstep. (The only passage that might tell against my understanding of Kreimer’s position is in the short final section of his 2003 article: ‘Faced with a threat of mass devastation that can be avoided only through torture, could an American official believe, as a matter of morality and public policy, that she should choose the path of the torturer as the lesser evil? On this question, I am prepared to concede that there is room for debate’ [2003, 324]. Given Kreimer’s consequentialism, we should construe ‘lesser evil’ here not as ‘lesser wrong’ but instead as ‘lesser negative factor’. If we ignore the non-committal character of the second sentence in this quotation, and if we furthermore ignore the fact that the first sentence in the quotation is framed obliquely as a question about what an American official could believe, we can take Kreimer here to be suggesting afresh that calamity-averting interrogational torture is sometimes morally legitimate. What makes the passage

142

Why Torture is Wrong

noteworthy is that no condition about the confinability of such torture to extreme cases is overtly attached. Hence, Kreimer might be understood here as maintaining that calamity-averting interrogational torture is sometimes morally legitimate irrespective of the probability of any slippery-slope consequences. However, given the overall tenor of his article, such an interpretation of his wording is almost certainly incorrect. Much more likely is that Kreimer in this latest formulation has simply neglected to include the confinability condition. At any rate, if the interpretation which I have just oppugned is correct, and if Kreimer therefore ultimately adheres to a variant of the Legal Prohibition Thesis rather than to the Slippery-Slope Illegitimacy Thesis, the reason for his doing so may well be that he has anticipated one of the objections which I will pose against the latter thesis.) Kreimer immediately goes on to deny that officials who resort to calamity-averting interrogational torture will thereafter restrain themselves from using torture in circumstances that are less desperate:  ‘Professor Dershowitz’s analysis, however, diverges from mine in his implicit assertion that torture would, in fact, be so confined’ (2003, 319). Admittedly, Kreimer and Alan Dershowitz are disagreeing principally over the proper legal status—rather than over the moral status—of calamity-averting interrogational torture. Kreimer’s heavy emphasis on slippery slopes might therefore be interpreted as a version of the Legal Prohibition Thesis rather than of the Slippery-Slope Illegitimacy Thesis. Still, although the 2003 article by Kreimer does lend itself to either interpretation, the absence of any clear-cut statement by him that calamity-averting interrogational torture can ever in practice be morally legitimate is what leads me to classify him as an advocate of the Slippery-Slope Illegitimacy Thesis. 3.1.3.3.4.  A second example In an insightful article on ticking-bomb scenarios, Kim Lane Scheppele subscribes to the Slippery-Slope Illegitimacy Thesis. On the one hand, she states near the beginning of her article that ‘I do in fact believe that torture is always and absolutely wrong, given the position we should accord to human dignity, even that of terrorists’ (2005, 287). On the other hand, the deontological stance articulated in that early statement is gradually superseded by the consequentialism of the Slippery-Slope Illegitimacy Thesis. Scheppele begins to move toward that latter thesis several pages after her initial affirmation of a deontological position: ‘[T]‌he farther away we get from the hypothetical [ticking-bomb scenario] in a real-world situation, the more reluctant we should be to condone torture, or even to entertain the possibility of it. Even if the hypothetical persuades us that torture would be justified in some extreme cases where many lives would be saved by immediate action, the anti-terrorism campaign has not yet and most likely never will present such a case’ (2005, 294). By the end of her article, Scheppele claims to have vindicated the Slippery-Slope Illegitimacy Thesis: The hypothetical case of the nuclear terrorist in Manhattan with the bomb set to go off might well persuade even the most principled objector to agree that torture is sometimes justifiable. But the question then is, what does that tell us about the situation that we actually confront?

Some Previous Accounts of the Wrongness of Torture

143

I submit that one can be a consequentialist and say that torture should be permissible in the hypothetical case, while still holding firm to the view that, nonetheless, torture should be absolutely prohibited, even in the present situation of terrorist threat. The consequences one is invited to weigh in the hypothetical pull one toward justifying torture precisely because they are presented in the most artificially isolated way. When it comes to real decisions in real contexts, however, the world will never be as straightforward as it looks in the hypothetical. As a result, one could find the implicit moral calculus in the hypothetical beguiling as a matter of theory and still be an absolutist on torture as a matter of practice. (Scheppele 2005, 338, emphasis in original, footnote omitted)

At an earlier stage in her article, Scheppele contends that one of the perniciously misleading aspects of the ticking-bomb scenario is that it presents calamity-averting interrogational torture as confinable to grave emergencies: ‘The real-world question that arises is not whether you or I would torture the Manhattan nuclear terrorist personally, but instead whether we can design rules for agents in complex organizations (like the military or intelligence communities) that would in fact limit torture to situations like this hypothetical, where we might agree as a political community that torture would be warranted’ (2005, 305). Accordingly, the long passage which I have quoted is a paradigmatic encapsulation of the Slippery-Slope Illegitimacy Thesis. Scheppele there declares that, although calamity-averting interrogational torture might sometimes be morally permissible in principle, it is always in practice endowed with a metastasizing tendency and is therefore always morally illegitimate on consequentialist grounds. Of course, I  am not suggesting here that Scheppele has abandoned her deontological scruples altogether in the course of her article. To be sure, her strong language about ‘persuad[ing] even the most principled objector’ might naturally be taken to indicate that she has indeed somewhat relaxed her deontological convictions in the long passage above. However, we are probably better advised to infer that she has retained those convictions and that she is expounding an alternative rationale (a consequentialist rationale) for the bottom-line prescription to which her convictions lead her. She is seeking to explain why consequentialists can and should agree with her about the moral status—and the recommended legal status—of calamity-averting interrogational torture. She aspires to show that consequentialists, too, should accept that calamity-averting interrogational torture is always and everywhere morally wrong in practice. In any event, although Scheppele probably regards the Slippery-Slope Illegitimacy Thesis as only ancillary to her main deontological position, she develops that thesis elegantly and forcefully. We should now turn to consider whether the Slippery-Slope Illegitimacy Thesis is correct as an account of the wrongness of calamity-averting interrogational torture. Let us ponder two principal objections to such an account. 3.1.3.3.5.  A first objection The Slippery-Slope Illegitimacy Thesis is a consequentialist doctrine. It purports to present a consequentialist justification for the absoluteness of the moral prohibition on calamity-averting interrogational torture. Yet a genuinely consequentialist approach to the matter will not only take account of the drawbacks associated with

144

Why Torture is Wrong

any use of calamity-averting interrogational torture, but will also take account of the benefits associated therewith. If we assume for the moment that any application of calamity-averting interrogational torture will be likely to lead to the unwarranted spread of torture, the disadvantages of that metastasizing spread have to be balanced—in a consequentialist assessment—against the advantages of averting a calamity. Decidedly unpersuasive is the notion that the consequentialist considerations in favour of the use of calamity-averting interrogational torture will always be outweighed in every emergency by the consequentialist considerations against any such course of action. Although the spectre of the metastasis of interrogational torture to non-emergency contexts will dismay most consequentialists, they will also be dismayed by the prospect of untold deaths resulting from an atrocity that could probably have been prevented through the wielding of such torture. Whereas the Slippery-Slope Illegitimacy Thesis rightly gives ample attention to the banefulness of the unwarranted spread of torture, it ignores the grievous harms that will be unprevented if calamity-averting interrogational torture is altogether eschewed. Admittedly, those unprevented harms would be highly isolated. They would occur only in the exceptionally rare contexts of extreme emergencies, whereas the harms caused by the overuse of torturous measures—arising from the metastasizing effects of calamity-averting interrogational torture—will occur more frequently. (Recall that I am temporarily assuming that any employment of calamity-averting interrogational torture will be endowed with a metastasizing tendency.) Still, the greater frequency of the latter harms can be outweighed in a consequentialist calculation by the greater scale of the former harms. Since it is overwhelmingly unlikely that the factor of greater frequency will always exceed the factor of greater scale, it is overwhelmingly unlikely that the balance in a consequentialist calculation will always incline against any use of calamity-averting interrogational torture. This point derives from the fact-sensitivity of a consequentialist approach to torture. Because consequentialists deny that any mode of conduct (such as interrogational torture) partakes of an inherent moral status, and because they insist that instead the moral status of every mode of conduct is determined by the consequences that are likely to flow therefrom, the moral wrongness or rightness of interrogational torture—like the moral wrongness or rightness of any other mode of conduct—is for consequentialists a contingent matter that depends on the specificities of each situation, or each class of situations, in which such torture is carried out. As Eric Posner and Adrian Vermeule, two robustly consequentialist advocates of calamity-averting interrogational torture, have remarked: ‘Here again we emphasize that everything depends on what the facts turn out to be. Because arguments about policies such as coercive interrogation . . . are hostage to what the facts show, in particular domains, there is no slope at all, just a series of discrete policy problems all arrayed on the same level’ (2006, 690; 2007, 203). Given the fact-sensitivity of any consequentialist moral doctrines, they are peculiarly ill-suited to serve as justifications for an absolute moral prohibition. Although in this subsection I have concurred arguendo with the proponents of the Slippery-Slope Illegitimacy Thesis in their empirical claims about the tendency of calamity-averting interrogational torture to metastasize, those claims leave room

Some Previous Accounts of the Wrongness of Torture

145

for a myriad of counterfactual worlds in which the use of calamity-averting interrogational torture would sometimes be strongly justified under any consequentialist reckoning. With reference to any of those counterfactual worlds, a consequentialist assessment of the morality of calamity-averting interrogational torture will generate the conclusion that the use of such torture is sometimes legitimate. No consequentialist doctrine, then, can underwrite an absolute moral prohibition on the use of such torture. That is, no consequentialist doctrine can underwrite the proposition that the use of such torture is always and everywhere morally wrong in all possible worlds. 3.1.3.3.6.  A second objection: preliminary clarifications As has just been noted, the exposition of my first objection to the Slippery-Slope Illegitimacy Thesis takes for granted that the relevant empirical beliefs of the supporters of that thesis are correct. In other words, I have assumed arguendo that any recourse to calamity-averting interrogational torture will be likely to lead to the spreading of torture into situations that fall well short of being extreme emergencies. Let us now query that assumption—not because it is patently false, but because it is unsubstantiated in the writings where it is operative. Though it might be correct, the empirical support for it in those writings is too meagre to vindicate it. If the specified assumption is false, then the Slippery-Slope Illegitimacy Thesis is not sustainable with reference to the actual world any more than with reference to the counterfactual worlds that have been broached at the end of § 3.1.3.3.6. In that event, calamity-averting interrogational torture is not absolutely wrong on consequentialist grounds even in the actual world, much less in all of those counterfactual worlds. To appraise the cogency of the empirical propositions that underlie the Slippery-Slope Illegitimacy Thesis, we need again to understand clear-sightedly the substance of those propositions. Supporters of the Slippery-Slope Illegitimacy Thesis cannot succeed by pointing out that, once torture has become institutionalized, it tends to entrench itself as professional torturers and bureaucrats gain a stake in perpetuating the ugly practice that is their livelihood. Although that observation about the institutionalization of torture is undoubtedly correct, it does not address the question which the champions of the Slippery-Slope Illegitimacy Thesis need to answer. The germane question is not whether institutionalized torture tends to be self-reinforcing; instead, the germane question is whether every recourse to torturous measures in circumstances of desperation will tend to bring about the institutionalization of torture (which will then be self-reinforcing). As I have already argued in response to Shue and Arrigo, the plying of torture by officials in a dire emergency does not depend for its effectiveness on an already-existing set of institutions within which torture has been professionalized. Hence, the point at issue here is whether the use of calamity-averting interrogational torture by officials in a situation of great urgency will tend to foster the growth of just such institutions. Suppose for example that, in their efforts to save the life of a hostage, some constables resort to twisting each arm of a captured kidnapper behind his back in order to ascertain the hostage’s whereabouts after the kidnapper has staunchly

146

Why Torture is Wrong

refused to respond to ordinary questioning. They are acting to prevent an imminent calamity, because they fear that the hostage will be killed by the remaining kidnappers if she is not rescued quickly. Will the actions of the constables in this desperate emergency (an emergency similar to the situation in Leon v. Wainwright) be likely to induce them or their colleagues to administer torture hereafter in circumstances that are not comparably urgent? Votaries of the Slippery-Slope Illegitimacy Thesis need to answer this question affirmatively. As should be abundantly apparent from what has just been said, the question to be addressed by the proponents of the Slippery-Slope Illegitimacy Thesis is about calamity-averting interrogational torture. It is not about torture in general. Those proponents cannot vindicate their position by showing that the use of interrogational torture for purposes other than the avoidance of imminent calamities is likely to metastasize. With perfect consistency, someone can assail the empirical presuppositions of the Slippery-Slope Illegitimacy Thesis while accepting that the employment of interrogational torture for purposes such as the disruption of terrorist networks or the foiling of distant plots or the solving of past crimes or the capture of high-ranking terrorists is likely to metastasize. We can readily grant that, when interrogational torture is wielded outside the pressures of an extreme emergency, it will tend to become routine. We should likewise readily grant that, when torture is employed by repressive regimes for purposes such as the squelching of dissidence and the intimidation of ordinary citizens, it tends to become routine under those regimes. To grant as much is hardly to endorse the empirical presuppositions of the Slippery-Slope Illegitimacy Thesis—for that thesis is precisely about the plying of interrogational torture (by officials in a liberal-democratic regime) in response to the pressures of an extreme emergency. 3.1.3.3.7.  A second objection continued: the meagreness of the evidence Advocates of the Slippery-Slope Illegitimacy Thesis have furnished very little support for the proposition that the use of interrogational torture in the circumstances of a dire emergency is likely to metastasize (Allhoff 2012, 154–7; Posner and Vermeule 2006, 688–90; Posner and Vermeule 2007, 200–3). As I  have remarked in § 3.1.1.2.2, those advocates generally refer to torture that has been administered for purposes other than the prevention of imminent calamities. Moreover, they generally focus on contexts where torture has to varying degrees been institutionalized for those other purposes. As I have sought to emphasize in the last few paragraphs here, conclusions drawn with reference to such contexts are not addressing the question which the devotees of the Slippery-Slope Illegitimacy Thesis should be addressing. For example, in her illuminating discussion of the institutionalization of torture (2005, 307–18), Scheppele has focused entirely on the actions of American soldiers and intelligence officials in Iraq and Afghanistan—and at Guantánamo Bay—during the first decade of the twenty-first century. She convincingly charts many of the difficulties that surround any efforts to control the use of torture when it has been institutionalized as it was in those theatres of war. However, her highlighting of such difficulties does not have any bearing on the matter which the proponents

Some Previous Accounts of the Wrongness of Torture

147

of the Slippery-Slope Illegitimacy Thesis have to address. After all, the institutionalization of torturous methods of interrogation in the aforementioned theatres of war was scarcely a development that mushroomed out of some initial instances of calamity-averting interrogational torture. On the contrary, the institutionalization of those methods was introduced ab initio for all the purposes which I  have mentioned in the preceding subsection (the disruption of terrorist networks, the foiling of distant plots, the uncovering of the facts of past terrorist outrages, and the capture of high-ranking terrorists). Of course, the averting of imminent calamities was another purpose for which the employment of torturous methods of interrogation would be authorized under the memoranda issued by the Department of Defense and the Office of Legal Counsel during the Presidency of George W. Bush. Nevertheless, the countenancing of torture by the Bush Administration was much more wide-ranging from the outset. Far from being a situation in which a few isolated instances of calamity-averting interrogational torture snowballed into a much more extensive practice, any instances of such torture that may have taken place were located in an already-existent practice. Thus, valuable though Scheppele’s reflections on the actions of Americans in Iraq and Afghanistan are, they do not go the slightest distance toward substantiating the empirical presuppositions of the Slippery-Slope Illegitimacy Thesis. They do not go the slightest distance toward showing that every instance of calamity-averting interrogational torture is likely to send a government tumbling down a slippery slope. Kreimer likewise does not provide any evidence that tends to confirm the empirical presuppositions of the Slippery-Slope Illegitimacy Thesis. His largely persuasive concerns about slippery slopes are focused on Dershowitz’s proposal for the institutionalization of torture through a system of torture warrants. Someone who impeaches the empirical underpinnings of the Slippery-Slope Illegitimacy Thesis can, without any inconsistency, agree heartily with Kreimer that the institutionalization of torture along the lines envisaged by Dershowitz is likely to get out of hand. Safeguards imposed on a system of warrants are highly prone to be overrun.8 While agreeing with Kreimer on that point, someone can consistently maintain that rare isolated instances of calamity-averting interrogational torture are unlikely to eventuate in the institutionalization of torture. Hence, forceful though many of Kreimer’s observations and arguments are, they do not support the Slippery-Slope Illegitimacy Thesis. They are observations and arguments about the probable ramifications of any institutionalization of torture, rather than about the metastasizing tendency of torturous measures that are undertaken ad hoc in circumstances of extreme desperation.

8   I shall return to Dershowitz’s torture-warrant proposal in Chapter 5, where I shall consider further the consequentialist arguments against his proposal. Though I regard those arguments as powerful, my main objection to his approach is not focused on slippery slopes or other consequentialist matters. Rather, it is a deontological objection directed against his preparedness to bestow legal approval on serious moral wrongs that involve the infliction of violence by public officials.

148

Why Torture is Wrong

3.1.3.3.8.  A closing caveat In short, the empirical basis of the Slippery-Slope Illegitimacy Thesis is dubious. Once the empirical claims on which that thesis rests are correctly specified, we can see how meagre the evidence in support of them is. Still, this discussion of the empirical presuppositions of the Slippery-Slope Illegitimacy Thesis should close with a caveat. Although I am unaware of any evidence that supports those presuppositions, I  have not endeavoured to adduce any detailed evidence against them. Likewise, in the writings by Allhoff and by Posner and Vermeule which I have cited near the beginning of § 3.1.3.3.7—writings that contest the empirical solidity of slippery-slope worries about the use of torture—the evidence marshalled by those theorists is skimpy. (Though Posner and Vermeule are correct about the scantiness of the evidence presented by their opponents, they have proceeded rather strangely by furnishing such slim evidence in favour of their own position. After all, they are impugning principally the Legal Prohibition Thesis. Given that the empirical claims which underlie that latter thesis are considerably more modest and thus considerably more plausible than the empirical claims which underlie the Slippery-Slope Illegitimacy Thesis, the need for the amassing of disconfirmatory evidence by people who challenge the empirical soundness of the Legal Prohibition Thesis is especially great. Posner and Vermeule have adduced very little such evidence.) Thus, I do not presume here to have rebutted decisively the empirical presuppositions of the Slippery-Slope Illegitimacy Thesis. Rather, the aim of the present discussion has been to query the plausibility of those presuppositions and to underscore the inadequacy of the case that has been made for them. In any event, even if the empirical basis of the Slippery-Slope Illegitimacy Thesis were much more solid than it is, that thesis would still fail as an account of the absoluteness of torture’s wrongness. As has been emphasized in § 3.1.3.3.5, the empirical and normative presuppositions of the Slippery-Slope Illegitimacy Thesis would be false in application to multitudinous counterfactual worlds even if they were always true in application to the actual world. Hence, my rejection of that thesis does not depend on the falsity of its empirical claims (and certainly does not depend on my demonstrating that those claims are false). Regardless of that empirical matter, the thesis does not amount to a satisfactory encapsulation of the nature of torture’s wrongness. Still, although my confutation of the Slippery-Slope Illegitimacy Thesis does not require a decisive rebuttal of the empirical underpinnings of that thesis, I have dwelt on the weaknesses of those underpinnings because they are of some importance to the overall position taken by this book. Notwithstanding that the present chapter and Chapter 4 will be maintaining that the use of interrogational torture is always and everywhere morally wrong in all possible worlds, the wrongness of interrogational torture is weakly rather than strongly absolute. In some imaginably extreme circumstances, the stringency of a person’s duty to refrain from administering such torture can be exceeded by the stringency of his or her duty to take steps that are essential for the averting of a disaster. Now, if the empirical and normative presuppositions of the Slippery-Slope Illegitimacy Thesis were correct,

Some Previous Accounts of the Wrongness of Torture

149

the consequentialist considerations in favour of resorting to interrogational torture would always and everywhere be outweighed in the actual world by the consequentialist considerations against such a step—in which case no one in the actual world would ever be under a consequentialist obligation to resort to such torture. A fortiori, then, no one in the actual world would ever be under an overtoppingly stringent consequentialist obligation to resort to such torture. No emergency, however dreadful, would ever confront anyone in the actual world with such an obligation. Accordingly, given that the use of interrogational torture is never morally optimal in the absence of such an obligation, a corollary of the empirical and normative presuppositions of the Slippery-Slope Illegitimacy Thesis is that the moral prohibition on interrogational torture is always overtoppingly stringent in the actual world. Because that corollary is not morally credible with regard to a situation like the one depicted in Chapter 2’s retelling of Leon v. Wainwright, my contestation of the presuppositions of the Slippery-Slope Illegitimacy Thesis has been motivated by concerns that transcend the present context.

3.1.4.  From consequentialism to Kantianism: torture and agency As we have seen—and as is evident from the general nature of consequentialism— proponents of consequentialist doctrines are unable to account for the absoluteness of the moral prohibition on interrogational torture and most other types of torture. In the eyes of many philosophers, the natural alternative to consequentialism is Kantianism in any of its sundry varieties. Nonetheless, although the present book is robustly deontological in its orientation, it gives Kantianism a wide berth. My wariness of Kantians is focused partly on the misguidedness of their aspiration to derive substantive moral conclusions from formal constraints of rationality (Kramer 1999b, 174–99; 2007, 82–6; 2009a, 289–94; 2009b). However, in the context of this book, an even greater source of dismay is the overemphasis placed by Kantians and others on the attribute of reflective or deliberative agency. When philosophers who write from an agency-overemphasizing perspective denounce the use of torture, they generally ascribe its wrongness to its discomfiture of the rational faculties of its victims. Jeffrie Murphy has delivered an especially arresting version of such a condemnation: Sending painful voltage through a man’s testicles to which electrodes have been attached, or boiling him in oil, or eviscerating him, or gouging out his eyes—these are not human ways of relating to another person. [The victim] could not be expected to understand this while it goes on, have a view about it, enter into discourse about it, or conduct any other characteristically human activities during the process—a process whose very point is to reduce him to a terrified, defecating, urinating, screaming animal . . . We have here a paradigm of not treating a person as a person—and thus an undermining of that very value (autonomous human personhood) on which any conception of justice must rest. (Murphy 1979, 233)

There is nothing in this powerful and incisive passage from which I dissent, but it is not a complete account of torture’s wrongness. Though it is a devastating indictment of the effects of some brutal torturous techniques on the sane human

150

Why Torture is Wrong

adults who undergo them, it does not capture the wrongness of inflicting torture on animals or infants or senile people or mentally deranged people. Because such beings are endowed with deliberative abilities at very low levels if at all, they are not reflective agents in the Kantian sense. That is, they are not persons in the Kantian sense. Accordingly, when Murphy trenchantly ascribes the wrongness of torturous techniques to the fact that they treat persons as non-persons, he has neglected to explain why the administration of such techniques to animals and to mentally incapacitated human beings would be morally wrong. His account of torture’s wrongness does not underwrite an absolute moral ban on the perpetration of sadistic torture. (Connected to this point is a disconcerting feature of the otherwise excellent passage from Murphy. He never explicitly states that the torturous techniques which he graphically recounts are extremely painful for the people who undergo them.) Before we ponder two further examples of the Kantian overemphasis on agency in some philosophers’ ruminations on torture, a couple of terse caveats are in order here. First, as is obvious, the property of reflective agency is a precious attribute that is vested with special ethical significance. In the actual world, insofar as we have become acquainted with it, reflective agency at a high level is unique to normal members of the human species who have passed the stage of infancy. Though some other precious and ethically significant features are common to human beings and animals, the feature of deliberative rationality at a high level is a quality that sets the human species apart. A second caveat relates closely to the first. When an ordinary adult human being is subjected to a wrongful act of torture, the overwhelming or coercion of her rational faculties through the administration of the torture is something that exacerbates the gravity of the wrong which has been committed. Though the sheer wrongfulness of such an act of torture does not hinge on the status of its victim as a reflective agent, the negative effects of the torture on the victim’s deliberative rationality will have intensified the magnitude of the wrongfulness. Someone who applies certain techniques of sadistic or punitive torture to a person has pro tanto behaved even worse than someone who applies similar techniques to a hamster.

3.1.4.1.  Waldron on torture and dignity Jeremy Waldron has written voluminously and perspicaciously on the matter of torture, but he has said surprisingly little to pin down the nature of torture’s wrongness. In the introduction to his principal book on torture, he himself has remarked that ‘I did not say nearly enough about its moral wrongness in the article on torture that became [the pivotal chapter] of this volume’ (2010a, 4). In that introduction he proceeds to offer some laconic but astute reflections on the wrongness of torture, to which I shall return later in this chapter. For the moment, we should concentrate instead on a more recent article by Waldron in which he connects the wrongness of torture to the concept of dignity. He there formulates his conception of dignity in a single long sentence: Dignity is the status of a person predicated on the fact that she is recognised as having the ability to control and regulate her actions in accordance with her own apprehension of

Some Previous Accounts of the Wrongness of Torture

151

norms and reasons that apply to her; it assumes she is capable of giving and entitled to give an account of herself (and of the way in which she is regulating her actions and organising her life), an account that others are to pay attention to; and it means finally that she has the wherewithal to demand that her agency and her presence among us as a human being be taken seriously and accommodated in the lives of others, in others’ attitudes and actions towards her, and in social life generally. (Waldron 2012a, 202)

By tying dignity so tightly to reflective agency, Waldron implies that senile people and insane people and severely retarded people and infants are without any dignity. Perhaps he would embrace that implication with equanimity, but it becomes especially unsettling when he goes on to trace the wrongness of torturous methods of governance to their wanton disregard of the dignity of persons. He contends that the formal ban on the use of such methods in contemporary legal systems is emblematic of the adherence of those systems to ‘a more general non-brutality principle’. Waldron formulates that principle as follows: ‘Law is not brutal in its operation; law is not savage; law does not rule through abject fear and terror, or by breaking the will of those whom it confronts. If law is forceful or coercive, it gets its way by methods which respect rather than mutilate the dignity and agency of those who are its subjects.’ Waldron thus charts an integral connection between the eschewal of torture and ‘respect for human dignity—respect for human dignity even in extremis, even in situations where law is at its most forceful and its subjects at their most vulnerable’. He adds: ‘No one denies that law has to be forceful. But forcefulness can take many forms and not all of it involves, for example, the savage breaking of the human will or a regression of the subject into an infantile state, where the elementary demands of the body supplant almost all adult thought that is the aim of torture’ (2012a, 218). Like Murphy’s hard-hitting comments, the pronouncements by Waldron which I  have quoted here do not contain anything from which I  dissent.9 Particularly insightful is the main sentence in the final quotation, to which we shall return later. However, again like Murphy’s pungent comments, the pronouncements by Waldron fall short of being a complete account of torture’s wrongness. Admittedly, their falling short in that respect might not be perceived by some readers as a flaw, since Waldron is dealing not with torture in general but with torture as a tool of governance. He is focusing on the interaction between a system of governance and those who are properly classifiable as its subjects in the sense that the legal requirements of the system are addressed to them and do not merely pertain to them. Given that focus, his highlighting of the connection between the legal prohibition on torture and the dignity of individual human beings is not inordinate. Nevertheless, although Waldron is undoubtedly correct in his suggestion that legal bans on torturous methods of governance have derived largely from the esteem of many contemporary legal systems for the dignity of individuals in the sense 9   I include the phrase ‘which I have quoted here’ because Waldron’s overall argument in his 2012 article is presented in support of Lon Fuller’s natural-law position against legal positivism. Waldron relies on untenable lines of reasoning which I have rebutted in my critique of Fuller in Kramer 1999a, chap. 3. If Waldron has come up with rejoinders to my critique of Fuller, he does not provide any inkling of them.

152

Why Torture is Wrong

which he delineates, the suggestion is misleading. It tends to obfuscate the reasons why sadistic torture is wrong even when it is undertaken against animals or against human beings who are bereft of dignity in the sense specified by Waldron. Such torture is grievously wrong not only when its excruciating painfulness supplants almost all adult thought in a victim, but also when its excruciating painfulness consumes the attention of a mentally infirm victim who is incapable of any adult thought. Waldron’s preoccupation with dignity tends to obscure the wrongness of torture that is plied against a victim of the latter sort. I have elsewhere criticized Waldron in a broadly parallel fashion for his over­ emphasis on reflective agency in his account of the wrongness of murder (Kramer 2011, 133–4). In a largely excellent article on the lex talionis, he suggests that murder is wrong because it ‘radically disrupts an autonomous life’ (Waldron 1992, 42). As I have objected, such a diagnosis of the wrongness of murder does not go any way toward explaining why the murder of a senile or mentally retarded human being is morally wrong. Although the emphasis on deliberative agency in Waldron’s writings on torture is not as excessive as in his ruminations on murder, his inclination to single out negative effects on autonomy as the pivotal wrong-making feature is still too strong. While we should readily allow that the occurrence of such effects will augment the gravity of any wrong that is perpetrated through torture, we should not locate the very basis of torture’s wrongness in those effects. We should be seeking a more capacious basis that accounts for the wrongness of torturous measures when they are undertaken against animals and mentally infirm people as well as when they are undertaken against normal human adults.10

3.1.4.2.  Sussman on the limits of Kantianism In an article containing many valuable insights, David Sussman (2005, 13–16) has percipiently exposed some of the shortcomings that mar the traditional Kantian approach to torture. In so doing, however, he himself as a Kantian has not fully broken free of the constraints of that approach. I shall slightly later examine his own theory of torture, which does transcend those constraints (while running into some other problems). At present, however, we should concentrate on his engagement with orthodox Kantianism. Sussman first persuasively maintains that utilitarians are unable to explain adequately the special wrongness of torture (2005, 13). By focusing solely on the disutility-engendering effects and utility-engendering effects that ensue from the perpetration of torture, utilitarians cannot distinguish between the production of those effects through torture and the production of those effects through morally legitimate instances of warfare. What can be added here to bolster Sussman’s point, in line with some of the analyses in my second chapter, is that utilitarians likewise 10   For a partly related expression of unease about Waldron on dignity, see Rosen 2012, 95–7. I should here incidentally note that, although I have placed Waldron within the broad tradition of Kantianism, his conception of dignity is not straightforwardly assimilable to Kant’s conception; see Waldron 2012b, 23–7.

Some Previous Accounts of the Wrongness of Torture

153

cannot distinguish between any morally legitimate instances of ephemerally incapacitative torture and any morally illegitimate instances of torture that yield the same net benefits. (Moreover, utilitarian theories of torture’s wrongness partake of numerous other weaknesses. For example, most such theories are committed to the proposition that sadistic torture is morally legitimate if its utility-promotive effects—above all, the pleasure accruing to the torturer—foreseeably exceed the victim’s anguish and any other negative effects of the torture.) Having taken exception to utilitarian perspectives on torture, Sussman reveals that traditional Kantianism is not satisfactory as an alternative. Whereas the utilitarians are preoccupied only with pains and pleasures, the traditional Kantian approach to the practice of torture tends to elide the severe painfulness of that practice. In the eyes of traditional Kantians, the wrongness of any instance of torture consists in its profound disrespect for the agency or autonomy of the victim, who is treated as a mere means for the furtherance of the torturer’s ends. Left out of such an assessment, or relegated to a subordinate place, is the specific way in which the torturer shows disrespect for the victim. As Sussman observes (2005, 14): For the orthodox Kantian, what is fundamentally objectionable about torture is that the victim, and the victim’s agency, is put to use in ways to which she does not or could not reasonably consent. The fact that it is pain that is characteristically involved is of only indirect importance. What immediately matters to the Kantian is that the victim may reasonably and strongly object to such treatment. The use of pain is significant only insofar as pain is something someone may reasonably and strongly refuse to undergo.

What is obscured in the traditional Kantian reckoning is the specificity of torture as the infliction of severe pain. Among the problems for the Kantian reckoning, with its focus on the victim as a reflective agent whose legitimate ends are to be respected, is that it does not cover any torture perpetrated against someone who is profoundly demented or retarded. Such a human being is devoid of reflective agency and is thus incapable of forming and pursuing ends. Such a human being is incapable of providing informed consent to any mode of interaction. Hence, the Kantian/contractualist criterion for wrongness to which Sussman adverts—a criterion that asks whether someone has not consented, and could not reasonably consent, to being treated in a certain manner—will fail to differentiate among the ways in which a profoundly demented or retarded human being can be treated. Insofar as the criterion does enable us to differentiate among ways in which human beings are to be treated, it is applicable only when the human beings in question are possessed of deliberative agency. While Sussman is rightly wary of the orthodox Kantian position, he believes that it can partly take account of the distinctive significance of pain. He writes as follows (2005, 14–15): Unlike other kinds of unwanted imposition, pain characteristically compromises or undermines the very capacities constitutive of autonomous agency itself. It is almost impossible to reflect, deliberate, or even think straight when one is in agony. When sufficiently intense, pain becomes a person’s entire universe and his entire self, crowding out every other aspect of his mental life. Unlike other harms, pain takes its victim’s agency apart ‘from the inside,’

154

Why Torture is Wrong

such that the agent may never be able to reconstitute himself fully. The Kantian can thus recognize that torture is not only a violation of the value of rational agency, but a violation that is accomplished through the very annihilation of such agency itself, if only temporarily or incompletely.

Sussman contends that, even when the orthodox Kantian position is extended in this more sophisticated direction, it remains vulnerable to at least two major objections. In the first place, he submits, the Kantians cannot adequately distinguish between the undermining of a person’s agency through intense pain and the undermining of a person’s agency through intense pleasure. ‘[T]‌he orthodox Kantian seems compelled to say that [the undermining of agency through intense pleasure] is wrong in precisely the same way as the normal [agonizing] forms of torture. The Kantian seems unable to do justice to what we would normally take to be a clearly nonaccidental truth: the fact that torture hurts’ (2005, 15, emphasis in original). A second hurdle for the traditional Kantians, in the view of Sussman, is that they cannot satisfactorily explain why torturing someone is in many contexts worse than killing someone. As he summarizes the matter: ‘We normally think that we may kill in self-defense or as part of the prosecution of a just war. But we are much more reluctant to accept that we may torture in self-defense or as part of proper combat. Although the Kantian might be able to show that torturing a person is not much better than killing him, she seems unable to explain how inflicting torture might actually be more objectionable (or objectionable in a fundamentally different way) than ending his life’ (2005, 15–16, emphasis in original). Now, before I  present my main response to Sussman’s amplification or re-elaboration of orthodox Kantianism, we should think back to my discussion of Barbara Herman on violence versus coercion in § 2.2.11.2.7 of Chapter 2. As is evident from that discussion, the problem broached in Sussman’s second objection is more complicated and in some respects more formidable than Sussman suggests. On the one hand, pace Sussman, Kantians such as Herman do maintain that countless instances of torture are morally worse than killing—because killing is non-coercive violence whereas many instances of torture are coercive. On the other hand, Kantians such as Herman have committed themselves to the proposition that wanton acts of murder are morally less grave than some morally legitimate instances of ephemerally incapacitative torture (such as the torture that is used in Sussman’s example of the obese man). Thus, although Sussman has framed his second objection somewhat simplistically, it touches upon a peculiarly vexing difficulty for the Kantians. Quite preposterous is the notion that the actions of the policemen in Sussman’s scenario of the obese man are morally more problematic than the actions of someone engaging in a vicious murder. At any rate, my chief query about Sussman’s extension of the orthodox Kantian position in the antepenultimate quotation above is that it remains too much within the confines of Kantianism. Highly perceptive and thought-provoking though the quoted passage is, it remains focused on the disruption of autonomous agency. In other words, Sussman there attaches importance to the consumingness of severe pain precisely because the tendency of such pain to fixate a person on her own agony is inimical to her rational deliberation. Hence, even the more

Some Previous Accounts of the Wrongness of Torture

155

sophisticated rendering of the orthodox Kantian position ignores the wrongfulness of sadistic torture that is perpetrated against animals or against human beings who lack autonomous agency. Though the sophisticated Kantian position is very much on the right track as it highlights the way in which the infliction of excruciating pain can take over a victim’s whole psyche, it generates that insight solely out of a concern for the autonomy of rational agents. If we aim to come to grips fully with the wrongness of torture, we have to shed that Kantian preoccupation—while of course recognizing that the disruptive effects of any wrongful instance of torture on the rational autonomy of a victim are a factor that substantially intensifies the gravity of that instance. (Another major shortcoming in the more sophisticated Kantian position expounded by Sussman is that its perspective on the wrongness of torture is inordinately victim-focused. Although it goes some distance toward a perpetrator-focused perspective through its emphasis on the way in which a victim’s ends are fully subordinated to those of the torturer, it concentrates on the physical and psychological degradation thus undergone by the victim. As will become apparent later in this chapter, the moral degradation undergone by the torturer—and by anyone on whose behalf the torturer acts—is instead decisive in some cases. Accordingly, if an account of the wrongness of torture is to be satisfactory, it will have to be more robustly perpetrator-focused than the sophisticated Kantian account.)

3.1.5.  Sussman on torture and self-betrayal Having plumbed the weaknesses of utilitarianism and traditional Kantianism in their accounts of the wrongness of torture, Sussman proceeds to unfold a subtle and fascinating theory of his own. He aims to refine the central Kantian position rather than to jettison it:  ‘My suggestion is essentially an extension of the Kantian thought that torture fails to respect the dignity of its victim as a rationally self-governing agent’ (2005, 19). Nevertheless, although he underscores the continuity between his own theory and the central tenet of the Kantian approach to torture, he develops some insights that surmount the limitations of the Kantian approach. I will raise some doubts about his theory, but it commendably transcends the Kantianism from which it emanates. Sussman’s central idea is that torture induces a peculiar form of self-betrayal. That general idea evolves somewhat in the course of his article, however. Sussman initiates his discussion of the matter rather dubiously by asserting that ‘[t]‌he torturer wants something from his victim, something that the victim would not normally provide (information, confession, pleading, and so on). In response, her captors begin a protracted process of threatening and inflicting pain’ (2005, 19). These claims are of course applicable to interrogational torture and act-impelling torture and victim-specific intimidatory torture, but they are largely or wholly inapplicable to many instances of sadistic torture and punitive torture. Thus, Sussman’s discussion initially appears quite cabined in its scope. Sussman initially appears to have in mind the self-betrayal that consists in acting against one’s hitherto deepest values and allegiances—for example, by divulging the identities

156

Why Torture is Wrong

of one’s fellow participants in a political movement or in some other project of great importance. Self-betrayal in that sense does occur in quite a few situations of torture; but it does not occur, and is often not sought, in many other such situations (including some of the worst such situations). Accordingly, if Sussman were writing solely about self-betrayal in this straightforward and familiar sense, he would be failing to specify any decisive wrong-making property that can accurately be predicated of torture as a general phenomenon. Though the foregoing sense of self-betrayal does figure in Sussman’s subsequent arguments, it is not the principal object of his attention. What Sussman chiefly has in mind is the turning of a victim against herself in a subtler and more profound fashion. When a torturer inflicts excruciating pain on a victim, he uses her body against her to express his own will—even if he is not seeking anything from her other than her experiencing of agony. By inducing severe pain in her, he has enlisted the natural feelings and susceptibilities of her body against her; her body takes his side, as it presents her with urgently clamorous signals and demands. As Sussman writes: What the torturer does is to take his victim’s pain, and through it his victim’s body, and make it begin to express the torturer’s will. The resisting victim is committed to remaining silent, but he now experiences within himself something quite intimate and familiar that speaks for the torturer, something that pleads a case or provides an excuse for giving in. My suffering is experienced as not just something the torturer inflicts on me, but as something I do to myself, as a kind of self-betrayal worked through my body and its feelings. (Sussman 2005, 21, footnote omitted)

Sussman’s reference to the victim’s determination to remain silent is unfortunate, since it can easily be taken to suggest that his analysis is limited to interrogational contexts and to other contexts where victims are resolutely reticent. It thereby fosters the impression that Sussman is still concentrating on self-betrayal in the familiar sense which I have specified in the preceding paragraph. However, the reference to the victim’s determination to remain silent is wholly dispensable; we can and should disregard it. (We should also disregard a number of other comparable references that are scattered throughout Sussman’s discussion.) Whereas self-betrayal in the familiar sense ensues in only some situations of torture, self-betrayal in the more subtle sense explicated by Sussman is a feature of every such situation. A  victim’s own bodily sensations and vulnerabilities are mobilized by the torturer against the victim. If the victim is a reflective agent, then of course the mobilization of her elemental feelings and vulnerabilities against her through her subjection to agony will attenuate or undermine her agency (at least temporarily). However, even if the victim is an animal or a mentally infirm human being, the marshalling of her bodily impulses and susceptibilities against her through her subjection to agony will dominate her consciousness and will gravely impair her well-being. Thus, the victim’s proneness to severe pain serves as a crucial means by which the torturer expresses and exerts his mastery. The bodily impulses of the victim, afflicting her with dire anguish, are emissaries of the torturer’s will. As Sussman states, ‘the victim finds in his pain, and his own immediate responses to that pain, a surrogate for the torturer’ (2005, 24). Having remarked

Some Previous Accounts of the Wrongness of Torture

157

that ‘[p]‌ain resembles a kind of primitive language of bodily commands and pleas that makes the same kind of insistent demands on our attention and response as our children’s shrieking and whining’ (2005, 21), Sussman declares that the ‘victim’s own voice, the voice of his body, has come in part to speak the torturer’s mind’ and that the perpetration of torture ‘puts the victim in the unavoidable position of betraying . . . himself, an experience the victim undergoes whether or not he actually informs or confesses’ (2005, 24). What should be added here is that self-betrayal of this type occurs even if the torturer is not seeking any information or confession, and even if the victim—who might be an animal or a mentally infirm human being—is wholly incapable of communicating any information or confession. When someone is subjected to torture, the wielding of her own primal sensations and susceptibilities against her is not dependent on her status as an autonomous agent.

3.1.5.1.  Techniques of torturous self-betrayal Sussman also adverts to some other ways in which many victims of torture undergo self-betrayal. Numerous techniques of torture use some normally innocuous characteristics of the bodies of victims to rack them with pain. One example not directly recounted by Sussman is the torture inflicted on Jean Améry by the Nazis in 1943. As Améry (1980, 32–3) harrowingly described in his renowned essay on torture published more than two decades afterward, the Nazis connected the shackles on his hands behind his back to a hook that lifted him about a metre off the floor. His own bodily weight then became his nemesis, for, after a short period during which he struggled to keep himself from descending, his weight dragged him downward and thereby caused the excruciating dislocation of his arms and shoulders. Sussman provides some examples where the role of a victim in producing or intensifying her own torment is more active (2005, 22–3). For instance, when someone is subjected to controlled drowning—a mode of torture that in its recent incarnations has come to be known as ‘water-boarding’ —her own desperate urge to inhale is what redounds to her detriment, as each attempt to draw in air simply worsens her plight of drowning. Likewise, when a captive has to remain stationary on a small elevated platform in order to prevent himself from stumbling off and being subjected to electrical shocks through wires attached to his shackles, his desperate determination to avoid the shocks eventually induces agony in him because of his prolonged stationariness. Through such techniques, ‘the sufferer is made to experience himself not just as a passive victim, but as an active accomplice in his own debasement’ (2005, 23). Of course, the techniques just outlined and other relevantly similar techniques are not always operative in situations of torture. Though such techniques in multitudinous forms are remarkably common, there are many ways of administering torture that do not so fiendishly turn victims against themselves. Moreover, the manoeuvring of the body of someone against herself is not confined to situations of torture. For example, when the legs of someone are shackled (non-painfully)

158

Why Torture is Wrong

to the floor, her mobility is acutely limited precisely through the limbs that are normally the mainstays of her bodily mobility. Nevertheless, the self-betrayal that is Sussman’s primary concern does indeed occur in every process of torture that has not been quickly discontinued. That is, the bodily/psychological vulnerabilities and impulses of every victim of such a process are turned against her as they become expressive of the torturer’s will and mastery through the victim’s undergoing of severe pain. Simply by dint of her nature as a sentient and embodied creature who is prone to feelings of pain— whether or not she is also a reflective agent—a victim of torture can be forced into this agonizing self-betrayal.

3.1.5.2.  The source of the self-betrayal Of course, an ordeal of torture is not the only setting in which someone can feel betrayed by her own body or mind. Anyone afflicted by illness or by an accidentally incurred injury can suffer great anguish that surpasses the anguish of many a victim of torture. Someone in such a condition can feel grievously let down by her own body or mind. Sussman readily recognizes as much, while also discerning a key difference between the agony suffered by a victim of illness or injury and the agony suffered by a victim of torture: When inflicted as part of torture, physical pain is importantly different from other kinds of agony that might result from disease or injury. In the case of natural pain, my body may seem to rebel against me, but it does not seem to rebel out of some other allegiance. The pain of disease may transfix me, but it does not seem to do so for any point of its own. Naturally caused pain is relatively inarticulate, manifesting itself as only the blank negation of my will by something that is nevertheless an intimate part of it. Torture is different, in that the sufferer experiences his pain as having a point. There is a will lurking behind his suffering, a will with a project that is somehow meant to be served by all the various torments. (Sussman 2005, 23)

Important though Sussman’s distinction between the agony of torture and the agony of illness or injury is, it stands in need of some qualifications. On the one hand, well-educated people in the modern era generally do recognize that—in the absence of special circumstances involving poisons or the like—illnesses and accidental injuries are due to natural causes or to inadvertencies rather than to anybody’s designs. Such people will therefore generally be inclined to mark the distinction that Sussman draws. Numerous people without any advanced education, and even some animals, will likewise probably be inclined to mark that distinction. On the other hand, however, most people in past ages and many superstitious people in the present age have viewed illnesses and injuries as afflictions sent upon them by gods or evil spirits or their fellow human beings. Most of the healings of maladies in the New Testament, for example, are depicted there as the vanquishing and expulsion of demons. From the perspective of superstitious people who believe that their debilities are due to the machinations of witches or evil spirits, the difference between the agony induced by torture and the agony induced by natural afflictions will have largely disappeared. Such people will tend to perceive any agony of the

Some Previous Accounts of the Wrongness of Torture

159

latter sort as a product of someone’s malevolence and will thus not view it as markedly different from agony of the former sort. Thus, although Sussman is right to maintain that the bodily sensations of a victim of torture become emissaries of the torturer’s will and dominance, he overlooks the fact that the bodily sensations of some sufferers of natural ailments will similarly be perceived by them as emissaries of certain hostile parties. Sussman is not incorrect in highlighting a key experiential difference between the travails of a victim of torture and the travails of a sufferer of a natural malady, but that difference obtains typically rather than always. Still, the perceptions of superstitious people—which sometimes efface the experiential difference just mentioned—do not determine the answer to the question whether those people are undergoing torture when they suffer from natural illnesses or accidental injuries. Except in a very expansive sense that is of no relevance to a discussion of torture as a problem in moral and political philosophy, those people are not being tortured when they are afflicted by such illnesses or injuries. Their sensations of bodily and psychological anguish in such circumstances are not really expressive of anyone’s dominance over them, even if they mistakenly believe otherwise. Consequently, since those sensations are not really products of other people’s designs, the tribulations of these superstitious sufferers do not amount to torture. Contrary to what the sufferers presume, they are not being wronged as they undergo those tribulations (unless, of course, their doctors remissly fail to administer proper curative or palliative care). Not every instance of agony is a moral wrong, even when the person afflicted by the agony erroneously thinks that some ill-disposed party—whether human or demonic—is wronging her. Sussman also fails to notice that mistakes can be made in the opposite direction. Suppose that, unbeknownst to Karen, somebody who holds a grudge against her has poisoned her with a substance which (if ingested) is usually non-lethal but is productive of intense pain that lasts for several days. Her enemy wants to reduce her to misery. As the substance begins to take effect, Karen consults a doctor. Though the doctor examines her with skill and due care, he concludes that her woes are due to some natural ailment. Thus, as Karen languishes in bed for several days while she is racked with severe pain, she believes that her ordeal is not a product of malevolence on the part of anyone else; in her eyes, it is a purely natural affliction. She has no inkling that she is being subjected to sadistically vindictive torture by somebody else. In her case, as much as in the case of a superstitious person who attributes a natural malady to the hostile designs of some foe, Sussman’s experiential distinction between the agony caused by torture and the agony caused by illness or injury has become effaced. Again, however, the misconceptions harboured by someone do not affect the answer to the question whether she is being tortured—or the answer to the closely related question whether she is being wronged. Regardless of what Karen thinks about the source of her misery, she is undergoing torture at the hands of her enemy. Unbeknownst to her, her feelings of debilitation and anguish are expressive of her enemy’s vindictive will. Far from being morally neutral (as Karen mistakenly believes it to be), her ordeal is a serious wrong that has been perpetrated against her.

160

Why Torture is Wrong

3.1.5.3.  Some transitional ruminations on Sussman’s theory Sussman’s account of torture’s wrongness serves as a fine point of transition to my own account, for there is much in his theory on which I shall profitably draw. His 2005 article is a major contribution to the contemporary philosophical literature on torture. Nevertheless, my own diagnosis of the wrongness of torture will seek to avoid the shortcomings that detract from his article. Perhaps most prominent among those shortcomings is his allegiance to Kantianism (albeit not to the orthodox Kantianism which he pertinently criticizes, of course). That allegiance manifests itself in his excessive emphasis on deliberative agency throughout his article. When Sussman expounds the subtle way in which every victim of torture is forced into self-betrayal, he characterizes the process as a perversion of the dignity with which a victim is endowed by dint of her capacities as a rationally self-governing agent. That characterization recurs, in one form or another, at many junctures in his article. Having come up with an insight that illuminates the nature of the travails undergone by every victim of torture, Sussman regrettably treats that insight as if it were confined to the travails undergone by reflective agents who are subjected to torture. By contrast, this chapter will endeavour to re-elaborate Sussman’s insight without any Kantian limitations. Another shortcoming in the 2005 article by Sussman is that he does not sufficiently disentangle the different senses in which a victim of torture can be forced into self-betrayal. In the most familiar sense, whereby somebody betrays herself if she acts in a way (for example, by providing information) that is detrimental to long-standing comrades or inimical to the promotion of values which she has staunchly endorsed, only some victims of torture are forced into self-betrayal. Not all instances of interrogational or act-impelling or intimidatory torture induce self-betrayal in that familiar sense, and most instances of other kinds of torture do not result in any such self-betrayal and are not aimed at inducing it. Though the principal type of self-betrayal which Sussman broaches is significantly subtler, the more familiar type continues to surface in his exposition of his theory (chiefly in passages that I have not quoted). He brings in further aspects of self-betrayal through his discussion of certain techniques of torture. All of these types of self-betrayal are very important, but they are distinct. Only the subtlest kind of self-betrayal delineated by Sussman—whereby the bodily sensations and vulnerabilities of a victim are turned against her through her subjection to severe pain—is realized in every instance of torture that is not quickly aborted or foiled. Given that my own account of torture’s wrongness will advert solely to that always-occurrent kind of self-betrayal, I shall have to be more careful than Sussman to avoid running it together with any of the other kinds of self-betrayal that he depicts. Related to the weakness just outlined is Sussman’s disinclination to ponder modes of torture other than interrogational torture (and punitive torture, in an ancillary discussion toward the end of his article). As Sussman acknowledges forthrightly: ‘My discussion focuses on interrogational torture, i.e., torture that involves a protracted process of inflicting or threatening pain in a context of helplessness and dependence, so as to make its victim provide information, confessions, denunciations,

Why Torture is Wrong

161

and the like’ (2005, 4). Now, as my second chapter has remarked, interrogational torture—especially calamity-averting interrogational torture—is far more frequently debated in contemporary Western societies than is any other type of torture. Hence, although Sussman’s article is entitled ‘What’s Wrong with Torture’ rather than ‘What’s Wrong with Interrogational Torture’, the circumscription of his investigation of the topic is quite understandable. Nonetheless, it skews his analyses toward some conclusions that are peculiar to interrogational torture (though some of his other conclusions are more wide-ranging in their applicability). Moreover, it leads him away from exploring whether and why any non-interrogational modes of torture can ever be instantiated legitimately. My own account of torture’s wrongness proceeds quite differently. Having submitted in Chapter 2 that some possible instances of ephemerally incapacitative torture and edifying torture are morally legitimate, I shall have to explain in this chapter why those kinds of torture differ morally from other kinds. In § 3.1.5.2, we have seen another misstep in Sussman’s reflections on torture: the overstatedness of his distinction between the anguish induced by torture and the anguish induced by natural illnesses or accidental injuries. Albeit the experiential difference to which Sussman’s distinction draws our attention is important, it does not always obtain. Hence, one should be wary of treating it as an integral aspect of torture’s wrongness. More generally, one should be wary of following Sussman in his excessively victim-focused perspective on the wrongness of torture. Crucial though such a perspective is, it has to be supplemented with a perpetrator-focused perspective. Without that latter viewpoint, we cannot fathom the wrongness of certain instances of calamity-averting interrogational torture and act-impelling torture. Similarly, without that latter viewpoint we cannot fully fathom the moral legitimacy of certain instances of ephemerally incapacitative torture and edifying torture.

3.2.  Why Torture is Wrong An account of torture’s wrongness should draw on two of Sussman’s insights that are presented separately from each other in his 2005 article. Central to his article is one of those insights, pertaining to the distinctive form of self-betrayal that is forced upon any victim of torture. Through the affliction of such a victim with severe pain, her own bodily feelings and susceptibilities are turned against her as instruments and expressions of the torturer’s dominance. We have just been mulling over that first main insight, of course.

3.2.1.  The consumingness of severe pain In addition to that first point, which Sussman articulates so prominently in his article, is an equally important point which he articulates less conspicuously and sustainedly. My summary of his critique of orthodox Kantian approaches

162

Why Torture is Wrong

to torture, in § 3.1.4.2, has quoted a passage in which he remarks upon the radically consuming power of intense pain. Although the passage as a whole is itself redolent of Kantianism in its overweening concentration on the destruction of agency, its general theme is disseverable from Kantian preoccupations. Of particular significance are the following two sentences: ‘It is almost impossible to reflect, deliberate, or even think straight when one is in agony. When sufficiently intense, pain becomes a person’s entire universe and his entire self, crowding out every other aspect of his mental life’ (Sussman 2005, 14). The gist of these statements is applicable even to situations of torture in which the victims are animals or mentally infirm human beings whose abilities to reflect and deliberate are extremely limited—far too limited for them to be correctly classifiable as reflective agents. With the quoted statements, Sussman highlights the power of severe pain to fixate upon itself anyone who suffers from it. Whereas somebody can largely disregard minor aches and pains while attending to daily affairs, excruciating pain does not similarly lend itself to being subordinated to other concerns. Instead, as is suggested by Sussman’s analogy between the insistent demands of such pain and the shrieking of children—an analogy which I have quoted in § 3.1.5—one’s excruciating pain thrusts aside one’s other concerns as it clamorously absorbs attention to itself. At its most severe, such pain crowds out everything else in someone’s psyche as its own horrible pulsing stymies her awareness of anything else. Améry memorably recounted the obliterative power of the agony of brutal torture. Among the many relevant statements in his essay are the following: ‘Whoever is overcome by pain through torture experiences his body as never before. In self-negation, his flesh becomes a total reality . . . . [O]‌nly in torture does the transformation of the person into flesh become complete’ (1980, 33). Writing specifically about the strappado-induced dislocation of his arms and shoulders, Améry recalled that the interrogator’s voice faded into a dim hum as ‘[a]ll your life is gathered in a single, limited area of the body, the shoulder joints’ (1980, 32). Of course, the torture to which the Nazis subjected Améry was notably dreadful, and the pain which they elicited in him was correspondingly terrible (though, as he himself remarked near the beginning of his essay, the torturous measures which he underwent—the strappado and horse-whipping—were not the worst such measures that were employed by the Nazis). Still, even when the pain induced by torture is somewhat less severe and is thus not obliterative of everything else in a victim’s consciousness, it dominates everything else by relentlessly fixating the victim on itself. Until it abates, it takes control of the victim as it thwarts him or her from attending sustainedly to other matters.

3.2.1.1.  The consumingness of euphoria To be sure, intense pleasure can also crowd out everything else in a person’s mind. Though Sussman in his rejoinders to the traditional Kantians has made this point with a focus on the disruption of rational self-governance, the point

Why Torture is Wrong

163

can be generalized to animals and human beings who lack the capacities for self-governance. Two remarks in response to that generalized observation are appropriate here. First, as is vividly conveyed by the tale of the lotus-eaters in Book IX of the Odyssey—and by the tale of the Sirens in Book XII of the Odyssey11—the inducement of overwhelming pleasure in someone for a prolonged period can be as sinister and destructive as the inducement of severe pain. As Sussman comments: ‘Certain drugs might induce intense euphoria in someone regardless of what he wants and in a way that, like pain, makes it impossible to think or care about anything else’ (2005, 15). Such a state of euphoria in a normal human adult over a lengthy period would preclude any rational deliberation, and in an animal or a mentally infirm human being it would undermine any sense of self. The non-consensual imposition of such a state on someone for a protracted span of time would be seriously wrong (except in some very unusual circumstances where the inducement of such a state is the only palliative that can spare someone from agonizing pain). Accordingly, the fact that intense pain resembles intense pleasure in its tendency to overwhelm all the other components of one’s consciousness is not a consideration that militates against the ascription of wrongness to the infliction of intense pain. Even more pertinent in the current context is a second remark. Whereas the radical absorbingness of intense pleasure is due to the delight which it elicits, the radical absorbingness of severe pain is due to the misery and aversion which it elicits. Though the grip on anyone’s consciousness exerted by the overpoweringness of ecstasy can in principle be as smothering as the grip exerted by the overpoweringness of anguish and dismay, the inducement of the latter feelings is morally far more problematic than the inducement of the former feeling. Sussman makes this point in a statement which I have quoted in § 3.1.4.2, where he asserts that ‘[t]‌he Kantian seems unable to do justice to what we would normally take to be a clearly nonaccidental truth: the fact that torture hurts’ (2005, 15, emphasis in original). Sussman’s point is that, whereas the orthodox Kantians will focus only on the extent to which a given way of treating someone has subverted his or her autonomy, the means by which the autonomy has been subverted is equally of great moral importance. Admittedly, there can be aberrantly masochistic people in whom the eliciting of severe pain will also evoke intense pleasure, and there can be aberrantly ascetic people in whom the eliciting of intense pleasure will also evoke deep distress. Indeed, somewhat fancifully, a torturer who knows of the strongly ascetic disposition of a victim could even educe sensations of pleasure in her for the purpose of afflicting her with misery. Still, the possibility of extreme masochists and extreme ascetics does not tell at all against the proposition that the deliberate infliction of severe pain is morally worse than the deliberate arousal of intense pleasure. If a torturer were to treat an extremely ascetic victim in the manner just suggested, his conduct would be reprehensible precisely because he would be seeking to rack her with anguish. Conversely, if somebody who knows of the   Also of relevance here is the scenario of the experience machine in Nozick 1974, 42–5.

11

164

Why Torture is Wrong

strongly masochistic disposition of someone else were to afflict her with severe pain in order to arouse keen pleasure in her, the gravity of his conduct would be attenuated precisely because he would be seeking to gratify her with delight. His conduct is morally wrong, but it is less seriously wrong than it would be if he were afflicting her with severe pain in order to make her desolate. (I have not yet expounded the basis for my judgement that the deliberate infliction of severe pain by one sado-masochist upon another is wrong. The basis for that judgement will become clear later, when this chapter introduces a perpetrator-focused perspective on these matters.) Of course, a torturer in an interrogational context might inflict grievous pain on a masochistic victim without knowing of her masochistic propensities. In that event, the torturer has committed a very serious wrong even though he has unwittingly provided his victim with abundant enjoyment. After all, as Sussman has rightly said in his critique of utilitarian approaches to the wrongness of torture, we should not be concentrating solely on the usually dire effects suffered by the victims: ‘There seems to be something about the distinctive structure of the relationship of torturer to victim that is intrinsically objectionable and that goes beyond the badness of its usual effects’ (2005, 13). If Angela subjects Markus to torturous methods of interrogation without knowing that he is a masochist who will derive pleasure from the experience, she will have adopted the outlook of an interrogational torturer in so doing. Unable to foresee the decidedly atypical results of her application of torturous interrogational methods to Markus, she has proceeded on the assumption that those methods will engulf him with searing pain and will thereby reduce him to a state of such desperate misery that he will divulge the information which she is seeking to obtain. Her having proceeded on that assumption is what makes her conduct even worse—considerably worse— than the conduct of a sado-masochist who knowingly subjects Markus to searing pain in order to elicit pleasure in him. (As should be evident, these reflections on a torturer’s unwitting evocation of pleasure in a masochistic victim are relevant to my reasons for prescinding from the Experience of Pain Inquiry in § 2.3.1.1 of Chapter  2. Because the outlook of a torturer who does not succeed in his endeavours to afflict his victim with agony is fundamentally the same as the outlook of a torturer who succeeds in such endeavours, the moral character of each torturer’s conduct is fundamentally the same. Though the actual inducement of agony in a victim intensifies the gravity of the second torturer’s conduct, the fundamental wrongness of the conduct is independent of that actual inducement. Much the same is true here. Although the interrogational torture inflicted on Markus would have been even worse if the sensations of searing pain had educed desolation rather than ecstasy in him, the fundamental wrongness of Angela’s conduct is independent of such an upshot. Given that Angela sets out with the intention of reducing Markus to misery and despair by causing him to experience intense pain, the actual occurrence of the misery and despair would merely be an intensifier of the gravity of her conduct. It would not be indispensable as a wrong-making property thereof.)

Why Torture is Wrong

165

3.2.1.2.  Some observations by philosophers and other theorists The consumingness of severe pain as a central wrong-making or gravity-intensifying feature of the administration of torture has not gone unnoticed by philosophers who have written on the topic. We have already pondered some of Sussman’s and Améry’s pertinent remarks. Let us here glance at a few germane observations by a handful of other writers. 3.2.1.2.1.  Beccaria on the filling of the sensory field Among the theorists deserving of commendation here is Cesare Beccaria. Writing in the eighteenth century, he excoriated judges for their use of torture to extract confessions from suspects. In the course of his denunciation, he highlighted the tendency of severe pain to overwhelm all other aspects of a person’s consciousness of her situation: Every act of our will is always proportional to the force of the sensory impression which gives rise to it; and the sensibility of every man is limited. Therefore, the impression made by pain may grow to such an extent that, having filled the whole of the sensory field, it leaves the torture victim no freedom to do anything but choose the quickest route to relieving himself of the immediate pain . . . And thus the sensitive but guiltless man will admit guilt if he believes that, in that way, he can make the pain stop. (Beccaria 1995, 41)

What is most important in the present context is Beccaria’s recognition that the exquisite pain undergone by a victim of torture is so clamorously preoccupying that it dominates everything else in the victim’s outlook. Notwithstanding that a victim might retain some awareness of other things as well, her attunedness to any of the other things will be mediated through her overwhelming sensations of agony. Those sensations, usually compounded by great fear, screechingly draw attention to themselves. Although a hardy victim might successfully resist any temptations to do what will bring her ordeal to a halt—and although some types of torture do not involve any opportunities for victims to do things that will terminate their hellish ordeals, in any event—the temptations are so powerful because the agony induced by torture is something on which a victim cannot but concentrate harrowingly. 3.2.1.2.2.  Luban on the tyranny of severe pain In one of the best articles on torture published during the last decade, Luban has powerfully recounted the way in which severe pain focuses the consciousness of a victim inexorably on the pain itself: [T]‌orture is a microcosm, raised to the highest level of intensity, of the tyrannical political relationships that liberalism hates the most. I have said that torture isolates and privatizes. Pain forcibly severs our concentration on anything outside of us; it collapses our horizon to our own body and the damage we feel in it. Even much milder sensations of prolonged discomfort can distract us so much that it becomes impossible to pay attention to anything else, as anyone knows who has had to go to the bathroom in a situation where it cannot be done. Ludwig Wittgenstein wrote that the world of the happy is different from the world of the unhappy, and this is not simply a figure of speech when we suffer severe pain.

166

Why Torture is Wrong

The world of the man or woman in great pain is a world without relationships or engagements, a world without an exterior. It is a world reduced to a point, a world that makes no sense and in which the human soul finds no home and no repose. (Luban 2005, 1430–1, footnotes omitted)

Although the closing sentences of this passage rather loosely use ontological language to describe a plight of epistemic shrinkage and misery—in a manner that would have been largely unexceptionable if Luban had refrained from suggesting that the language is ‘not simply a figure of speech’ —the gist of the passage is squarely on target. Excruciating pain does indeed contract the horizons of a victim by fixating her attention grimly on her own distress. It takes control of her outlook by suffusing all of her thoughts with its own strident clamour. Even if a victim of torture can manage to remain attuned to some other things as well, her consciousness will be trained principally on the agony that she is undergoing. The power of that agony to transfix a victim is what Luban eloquently delineates. 3.2.1.2.3.  Kreimer on the occupation of the self Kreimer too has given vivid expression to the consumingness of severe pain. Though he writes with reference to the autonomy of victims of torture, his comments can readily be extended to victims who are incapable of autonomous action in any full-blown sense: The pain of torture by design negates the vision of humanity that lies at the core of a liberal democracy. Justice Kennedy recently set forth the constitutional importance of the ‘autonomy of self ’ in Lawrence v. Texas. Torture seeks to shatter that autonomy. Torture’s evil extends beyond the physical; extreme pain totally occupies the psychic world; the agony of torture is designed to make choice impossible. Effective torture is intended to induce the subject to abandon her own volition and become the instrument of the torturer by revealing information. Such government occupation of the self is at odds with constitutional mandate. (Kreimer 2003, 298–9, footnotes omitted)

Kreimer obviously has in mind only interrogational torture in this passage, but the general tenor of his pronouncements is disseverable from that focus. In other words, the consumingness of the anguish caused by torture is hardly confined to interrogational contexts. Regardless of the type of torture that is carried out, the severe pain inflicted by it will dominate the consciousness of the victim while the pain lasts. 3.2.1.2.4.  Scarry on the body in severe pain Elaine Scarry is the theorist who has most frequently directed attention to the consumingness of agony. She is not a philosopher, and her book The Body in Pain is sometimes disconcertingly imprecise and overblown in its argumentation. All the same, despite its shortcomings, it contains a number of astute observations on the power of severe pain to exert a dismal grip over the psyche of any victim of torture. One of the briefest of her relevant observations does not pertain specifically to the effects of torture, though its applicability to those effects is manifest. Scarry adverts memorably to the painfulness of the debility of old age: ‘As the body breaks

Why Torture is Wrong

167

down, it becomes increasingly the object of attention, usurping the place of all other objects, so that finally, in very, very old and sick people, the world [as they perceive it] may exist only in a circle two feet out from themselves’ (1985, 32–3). My bracketed insertion has forestalled here an ontological/epistemic conflation that surfaces quite frequently in Scarry’s book.12 Still, with that lapse out of the way, her remark insightfully captures the epistemic shrinkage that ensues from the enveloping painfulness of old age. Scarry’s comment on the harrowing debilitation of elderly people occurs in a chapter of her book that focuses sustainedly on the processes and effects of torture. In that chapter, most of her remarks on the consumingness of severe pain do pertain specifically to contexts of torture. At the same time, the parallels between the anguish of victims of torture and the anguish of victims of other adversities often remain in view, as in the following passage: Perhaps only in the prolonged and searing pain caused by accident or by disease or by the breakdown of the pain pathway itself is there the same brutal senselessness as in torture . . . It is the intense pain that destroys a person’s self and world, a destruction experienced spatially as either the contraction of the universe down to the immediate vicinity of the body or as the body swelling to fill the entire universe. Intense pain is also language-destroying: as the content of one’s [experience of the] world disintegrates, so the content of one’s language disintegrates; as the self disintegrates, so that which would express and project the self is robbed of its source and its subject. (Scarry 1985, 35)

Now, as will become evident in § 3.2.1.3, we should be somewhat wary of the tone of this passage. On the one hand, Scarry’s strong wording about the destruction or disintegration of one’s self and one’s language is undoubtedly appropriate with reference to any victim of techniques of torture that are especially brutal or long-lasting. Furthermore, if Scarry allows that the destruction can be temporary and less than comprehensive, her remarks are much more broadly applicable to victims of torture. On the other hand, her strong wording lends itself to being taken at face value. So construed, her assertions are too sweeping. Though numerous victims of torture do indeed undergo the complete dissolution of their epistemic access to the world—as their agony engulfs them to the exclusion of everything else in their consciousness—the dissolution in many such cases is temporary, even though it is typically followed by long-term effects which are often serious but which fall short of the wholesale obliteration of all other mental states by pulses of searing pain. Moreover, even during the very period when torture is wielded against a victim, the power of her severe pain to fixate her thoughts upon itself does not always wipe out every other aspect of her awareness. The disintegration of her connection to the world can be partial rather than thoroughgoing. Though her intense pain dominates everything else in her consciousness during her ordeal, its dominance does not always consist in the outright expunction of everything else. 12   Indeed, the conflation is salient in the subtitle of her volume: The Making and Unmaking of the World. What Scarry in fact describes in her chapter on torture is the unmaking of victims’ epistemic connections to the world.

168

Why Torture is Wrong

Still, although Scarry’s phrasing in the passage above is too strong in application to some instances of torture, the transfixingness of agony is perhaps best conveyed through such phrasing. Let us examine two of the many further germane passages from her book. In the first of these, which prefigures much of Sussman’s later analysis of the matter, Scarry again keeps in view the affinities between the grievous pain produced by torture and the grievous pain produced by some illnesses and accidents: Regardless of the setting in which he suffers (home, hospital, or torture room), and regardless of the cause of his suffering (disease, burns, torture, or malfunctioning of the pain network itself ), the person in great pain experiences his own body as the agent of his agony. The ceaseless, self-announcing signal of the body in [severe] pain, at once so empty and undifferentiated and so full of blaring adversity, contains not only the feeling of ‘my body hurts’ but the feeling ‘my body hurts me.’ This part of the pain, like almost all others, is usually invisible to anyone outside the boundaries of the sufferer’s body, though it sometimes becomes visible when a young child or an animal in the first moments of acute distress takes maddening flight, fleeing from its own body as though it were a part of the environment that could be left behind. (Scarry 1985, 47)

Like the earlier quotation from Scarry on the infirmities of old age, the closing sentence in this excerpt valuably makes clear that the consumingness of agony can engulf not only autonomous human adults but also non-autonomous beings such as animals and young children. Because such beings are vulnerable to the enveloping force of dire pain, most types of torture practised against them are always and everywhere morally wrong in all possible worlds. In the final extract from the book by Scarry that will be considered here, she summarizes the way in which severe pain tends to overwhelm the outlook of anyone who suffers from it (1985, 54–5): Pain begins by being ‘not oneself ’ and ends by having eliminated all that is ‘not itself.’ At first occurring only as an appalling but limited internal fact, it eventually occupies the entire body and spills out into the realm beyond the body, takes over all that is inside and outside, makes the two obscenely indistinguishable, and systematically destroys anything like language or world extension that is alien to itself and threatening to its claims. Terrifying for its narrowness, it nevertheless exhausts and displaces all else until it seems to become the single broad and omnipresent fact of [a victim’s] existence.

Scarry here again vividly recounts the culmination of the workings of severe pain in a victim of torture. Some victims of torture do not experience that full culmination, if the techniques brought to bear upon them are insufficiently brutal or insufficiently protracted. Nevertheless, even when the severity of a victim’s pain does not reach the point of blotting out everything else in her field of awareness, it permeates everything else and pressingly obliges the victim to direct her thoughts to itself in all its ugliness.

3.2.1.3.  The perils of overstatement Central to my account of the wrongness of torture is the consumingness of the agony induced by the use of torturous techniques. However, we should be alert

Why Torture is Wrong

169

to the dangers of overstatement. My account seeks to specify why most types of torture are always and everywhere wrong in all possible worlds; hence, it has to eschew claims (about the effects of agony) that are applicable to only some victims of those types of torture. In § 3.2.1.2.4, I have briefly looked askance at the hyperbole that detracts from some of Scarry’s otherwise evocative descriptions of the effects of severe pain on victims of torture and on other sufferers. Here we should consider in somewhat more detail the inadvisability of overstatement in these matters. 3.2.1.3.1.  Preliminary remarks My account of the wrongness of torture will highlight the self-betrayal induced by torturous mistreatment and the obliterative power of the agony inflicted by such mistreatment, and will bring those points together into a perpetrator-focused perspective. At present we are concentrating on the obliterative power of the agony that is caused by torture. If the torturous techniques applied to a victim are especially harsh or prolonged, they can elicit in her such scorching pain that they block out everything else in her field of awareness. They can of course prove lethal. However, even when the victim remains alive, her sentience can be thoroughly blanketed by the pulsating agony to which she has been reduced. If she is a reflective agent, her agency will have been destroyed either temporarily or permanently. In the event that her agency is not permanently destroyed, it may well be permanently impaired. Moreover, even when techniques of torture are not sufficiently harsh or sufficiently protracted to educe pain of such severity as to wipe out everything else in a victim’s field of consciousness, they will (unless quickly aborted or foiled) educe pain severe enough to dominate everything else in that field of consciousness. In other words, although a victim of such techniques might retain some awareness of various other things in her environment, her perceptions of those things will be filtered through the dominant feature of her consciousness: her terrible pain. As this chapter has already remarked, severe pain is unlike mild pain in that someone suffering from it cannot relegate it to a subordinate level of her attention in order to concentrate principally on other matters. She might be aware of some other matters and might be able to deal with them to some degree, but the shrieking intensity of her pain inexorably draws her attention primarily to the pain itself. In that regard, then, the agony of a victim of torture is consuming even when it is not quite all-consuming. 3.2.1.3.2.  Hyperbole best avoided When the points made in § 3.2.1.3.1 are combined with this book’s exposition of the nature of torture and with my elaboration of a perpetrator-focused perspective and with Sussman’s insights about the self-betrayal forced upon victims of torture, they are sufficient to support my main thesis that most types of torture are always and everywhere wrong in all possible worlds. Accordingly, my account of torture’s wrongness does not presuppose any excessively strong claims of the sort in which Scarry sometimes indulges. Another theorist who has indulged in such

170

Why Torture is Wrong

claims is the philosopher Jeremy Wisnewski, on whose work I will focus briefly here. (Wisnewski uncritically quotes an array of passages from Scarry.) Notwithstanding that the quality of Wisnewski’s philosophical argumentation is uneven, his book makes a number of salutary contributions to the debates over torture. In particular, Wisnewski commendably documents the effects of torture by adducing many pertinent quotations from survivors of torture and from investigators who have studied such survivors (2010, chap. 4). Their testimony vividly underscores the lastingness of many of those effects; long after victims have undergone the administration of torturous techniques to their bodies, they often still experience various physical and psychological ailments from their ordeals. Still, valuable though his writings on torture have been in those respects and in some other respects, Wisnewski goes astray through the exaggeratedness of some of his principal contentions. In particular, Wisnewski goes amiss by maintaining repeatedly that the key to the wrongness of torture (and to the distinctiveness of its wrongness) is that the reflective agency of every victim of torture is destroyed by the measures to which she is subjected. He contends that ‘torture destroys the person, and with this, it destroys the very capacity for rational deliberation,’ and that ‘torture dismantles the agency of the person tortured’. He likewise asserts that the pain experienced by a victim of torture ‘is born of another agent whose very intent is to destroy your agency—to make you incapable of directing your own actions, and even incapable of determining the significance of the things that populate the world around you’ (2010, 64, emphases in original). One shortcoming of these statements, and of the many similar statements that appear in Wisnewski’s two principal chapters on the wrongness of torture, is that they exclude any victims of torture who are not possessed of reflective agency. Consequently, they do not address the plight of animals or mentally infirm people who are subjected to torture. Perhaps Wisnewski would reply that his conception of agency is sufficiently expansive to encompass some non-human animals. Even so, any plausible conception will omit some animals as well as senile people and severely retarded people. Given that such people and animals are potential victims of torture, Wisnewski has failed to explain why torture of any type is always and everywhere wrong. Even more important in the present context is that Wisnewski’s specification of the central wrong-making property of torture is formulated too strongly. On the one hand, as this chapter has already observed, some instances of torture do temporarily or permanently destroy the agency of the people who undergo the torture. Such instances are especially vile. On the other hand, not every instance of torture produces such a devastating effect. Every torturous ordeal that is not quickly aborted or foiled will inflict severe pain on the victim,13 but not every such ordeal goes so far as to demolish the victim’s mental faculties. Some torturous techniques 13   Here and elsewhere, I  use the words ‘or foiled’ to leave aside victims who are congenital analgesiacs.

Why Torture is Wrong

171

leave those faculties generally intact, even while suffusing the victim’s mental states and processes with exquisite pain. Wisnewski himself, despite all his pronouncements about torture’s destruction of the agency of victims, acknowledges that quite a few instances of torture do not eventuate in such destruction. He accepts that ‘there are instances where torture methods have been applied, and the victim of these methods has not been broken by them, or where victims seem to recover from the unmaking of [their epistemic connectedness to] the world’ (2010, 87). Having acknowledged that point, he immediately impugns the motives of anyone who might invoke it against him: ‘If it is meant to show that torture can be acceptable in certain cases, it is surely a very misleading way of proceeding. The fact that a practice aimed at the destruction of agency might not achieve its goal is hardly evidence that the practice is sometimes acceptable’ (2010, 87). At least in application to me, these strictures by Wisnewski are misconceived. My motives for criticizing him are the opposite of what he envisages here. Far from trying to show that some instances of interrogational or act-impelling or sadistic torture are morally legitimate, I am trying to show that all such instances are morally impermissible. My concern is that, by singling out the destruction of agency as the key wrong-making effect of torture, Wisnewski has left himself unable to indicate why most types of torture are always and everywhere morally wrong. After all, given that the demolition of agency does not ensue in some situations of torture, Wisnewski’s theory does not tell us why those situations are nonetheless wrong. Wisnewski endeavours to square his theory with the fact that not every instance of torture results in the dismantling of the agency of its victim. Instead of outlandishly insisting that torture always does in fact bring about such a result, he retreats to insisting that the aim of every instance of torture is to destroy the agency of its victim (2010, 87–8, 122–3). He draws an analogy to baseball games. Although the aim of the participants in a professional baseball game is to carry on the game for at least nine innings, that aim is not always fulfilled (because of electricity failures or rain-induced abridgements, for example). So too, Wisnewski maintains, the objective of every torturer is to destroy the agency of any victim. That objective is the very hallmark of torture: The aim of torture is to destroy the agent . . . To point out that torture sometimes fails to achieve the goal postulated in torture manuals, however, is irrelevant to determining what torture aims to do. To say otherwise would be as crass as saying that torture does not always destroy the capacity for agency because sometimes a person is executed before the torture is complete. (Wisnewski 2010, 88)

One problematic feature of these pronouncements by Wisnewski is that he again excludes all instances of torture in which the victims are not capable of reflective deliberation. When wanton boys torture a hamster or a senile human adult, they are scarcely seeking to destroy the victim’s status as a reflective agent—for such a creature or human being is not endowed with reflective agency. Instead, the boys are seeking to inflict grievous pain on a creature or human being who is susceptible to feeling such pain. They derive sadistic gratification from their infliction of the

172

Why Torture is Wrong

pain, even though their actions do not impair anyone’s agency and are not aimed at impairing anyone’s agency. Wisnewski appears to have committed himself to the strange conclusion that the delinquent boys are not engaging in torture at all when they subject a hamster or a senile human adult to deliberately agony-inducing mistreatment. Equally problematic is that Wisnewski’s pronouncements also exclude some instances of torture in which the victims are normal adult human beings. Quite implausible is the notion that every such instance is undertaken with the aim of destroying the victim’s agency. Recall, for example, Sussman’s scenario of the obese man. When the police twist the arm of the man behind his back to oblige him to desist from his action of smothering the person who is trapped beneath him, they are hardly endeavouring to dismantle his agency. On the contrary, they want him to exercise his agency by lifting himself up. In no way are the policemen’s actions aimed at terminating the agency of the corpulent man either permanently or temporarily. Of course, the twisting of the obese man’s arm is one of the morally legitimate instances of ephemerally incapacitative torture. Hence, Wisnewski might retort that he is focusing only on the sundry types of torture that can never be instantiated legitimately. Such a retort, however, would be unavailing. Not every morally impermissible instance of torture involving a normal human adult as the victim is aimed at destroying the victim’s reflective agency. Recall, for example, my slightly embellished version of Leon v. Wainwright in § 2.2.1.3 of Chapter 2. When the policemen there resort to choking the captured kidnapper and twisting his arm behind his back in order to prevail upon him to disclose the whereabouts of his hostage, they scarcely are seeking to demolish his status as a reflective agent. Rather, they are attempting to impel him to exercise his agency through his divulgence of the information which they desperately need. Though the severe pain which they elicit in him is such that it dominates everything else in his consciousness, it is not so severe (and is not intended to be so severe) as to destroy his agency by blotting out everything else. Were we to characterize the policemen’s efforts or the aim of their efforts in line with Wisnewski’s pronouncements, the characterization would be hyperbolically distortive rather than illuminating. In short, with reference to Leon v.  Wainwright and any relevantly similar instances of interrogational torture, the hyperbole in which Wisnewski indulges has confronted him with two uninviting alternatives. Either he might choose to deny that any such instances of interrogational torture are torture at all, or he might choose to deny that any such instances of interrogational torture are morally wrong. In either event, he will have left himself unable to explain why actions correctly classifiable as interrogational torture are always and everywhere morally wrong. Thus, although Wisnewski suspects that anyone who objects to his exaggerated pronouncements is thereby seeking to affirm the permissibility of interrogational torture in certain situations, the chief vice of those pronouncements is that they impede efforts to establish that such torture is never morally permissible (even in circumstances like those of Leon v. Wainwright).

Why Torture is Wrong

173

This discussion should close with a caveat. My criticism of Wisnewski’s penchant for overstatement needs to steer clear of overstatement in turn. Many instances of torture throughout the world—especially the brutal instances of torture carried out by officials of thuggishly repressive regimes in parts of Asia and Africa and South America—do temporarily or permanently destroy the agency of the victims. Even when techniques of torture are somewhat less brutal, they can often temporarily destroy the agency of the victims of those techniques (and impair their deliberative abilities thereafter). My point has simply been that such devastating effects do not always occur in situations of torture and are not always sought in those situations. Wisnewski himself acknowledges that such effects do not always occur, but he unsustainably contends that they are always pursued by torturers. Whereas torturers always seek to inflict agony—or are extravagantly reckless about the infliction of agony—they do not always endeavour to go as far as Wisnewski’s excessively strong wording suggests. By recognizing as much, and by not tying the wrongness of torture to the presence of effects or aims that are sometimes absent from situations of torture, my account of that wrongness avoids the pitfall into which Wisnewski has plummeted. In other words, my account covers the situations (such as that in Leon v. Wainwright) which his excessively strong phrasing omits. Consequently, unlike his approach to the matter of torture’s wrongness, my approach can establish that all possible instances of most types of torture are morally wrong.

3.2.2.  Combining two insights Although this chapter will submit that a satisfactory account of torture’s wrongness must incorporate a perpetrator-focused perspective, the considerations adduced so far have been largely victim-focused. Quite understandable is the focus on victims in most indictments of torture, for the interests of victims in being free from excruciating pain are morally determinative in the vast majority of situations where any torturous techniques are employed. In the large majority of such situations, those interests of victims are sufficient to endow them with moral rights against being subjected to torturous techniques. Hence, most applications of those techniques to people or animals are morally wrong because of victim-focused considerations. Accordingly, before we move on to ponder why a victim-focused perspective has to be supplemented with a perpetrator-focused perspective—and why some possible instances of ephemerally incapacitative torture and edifying torture are morally permissible—we should take stock of the two main strands that have heretofore been woven together into my account of torture’s wrongness.

3.2.2.1.  The two main strands The first such strand derives from Sussman’s insight into the subtlest type of self-betrayal induced by torture. That special self-betrayal consists not in the abandonment of allegiances or the imparting of precious information—though of course those consequences often ensue from the administration of torturous

174

Why Torture is Wrong

techniques—but instead in the turning of a victim’s own feelings and vulnerabilities into instruments to be used by her torturer to consolidate and express his ascendance over the victim. When somebody is subjected to a torturous mode of treatment, her susceptibility to feelings of intense pain becomes the chief means by which her torturer exerts his will against her. Whether a torturer is seeking information or submissiveness or sadistic gratification or any other objective that can be attained through the infliction of severe pain on someone else, the agony which he elicits in his victim is both serviceable for the realization of his objective and expressive of the power over her with which he pursues that objective. The intimate sensations of the victim are marshalled against her, as they become implements and messengers of the torturer’s control over her. Even in the extremely rare circumstances where a victim of torture is unaware that her agony has been evoked by torturous measures rather than by a natural ailment, her susceptibility to severe pain has been turned against her by her tormentor. Notwithstanding that the sufferer in those circumstances has failed to recognize that her woes are the products of someone else’s mastery over her, those woes mark the extent to which her body has been employed by someone else to betray her. (Though my remarks here have focused on bodily torment, they can easily be extended to cover psychological agony as well. Likewise, although my remarks here have focused on cases of torture in which the inducement of severe pain is deliberate, they can easily be extended to cases of torture in which the inducement of severe pain is instead extravagantly reckless.) The second strand of my account pertains to the consumingness of severe pain. Whereas mild pain or discomfort can be largely ignored by somebody who carries on with his or her ordinary activities, the severe pain educed by torture—like the severe pain caused by various natural maladies—cannot be put aside in such a fashion. Even if it does not reach the level of intensity where it completely obliterates everything else in a victim’s field of consciousness, it inexorably fixates the victim’s thoughts on itself. It might not wholly blot out her awareness of certain other things, but any such awareness pulsates with the agony that prevents her sustained concentration on any of those other things. Her agony smashes its way into any aspects of her experience that have not been altogether swept from view by it.

3.2.2.2.  The combination In combination, the foregoing two points constitute a powerful victim-focused basis for the proposition that torture is generally wrong as a moral matter. Without a perpetrator-focused perspective, they are not sufficient to establish that the moral prohibition on interrogational or act-impelling or placatory torture is absolute—as we shall see presently—but, together with Chapter 2’s exposition of the nature of torture, they can establish that most applications of torturous measures are morally impermissible. By contrast, neither of the two strands on its own (in isolation from the other or in isolation from Chapter 2’s conception of torture) would suffice to establish that torture is generally wrong.

Why Torture is Wrong

175

3.2.2.2.1.  Not enough in isolation Insufficient in isolation to generate a conclusion about moral wrongness is the fact that torture evokes grievous pain in the people and animals who undergo it. After all, people and animals who suffer any number of natural afflictions (illnesses or injuries) also experience grievous pain; the consumingness of the agony produced by torture can be paralleled by the consumingness of the agony that results from those natural afflictions. Given that the natural ailments are not themselves morally wrong at all, we cannot infer the wrongness of torture from the sheer fact that it elicits severe pain—with the attention-transfixing power of such pain—in its victims. That is, we cannot infer the wrongness of torture from the fact that it is endowed with a feature that assimilates it to sundry natural maladies. Likewise insufficient in isolation to generate a conclusion about moral wrongness is the fact that any instance of torture turns a victim’s sensations into instruments and emissaries of the torturer’s will. After all, quite a few morally legitimate activities can yield a similar effect. Some persuasively enticing blandishments from a salesman can exploit the feelings of a potential customer as the means by which the salesman can realize his objective of selling some merchandise. Provided that the blandishments are not fraudulent or coercive, their effectiveness in marshalling the feelings of the customer as implements and messengers of the salesman’s will is morally unexceptionable. Similarly, if Emily is sincerely attracted toward Richard, and if she speaks to him and touches him in ways that are designed to kindle in him a passionate attraction toward her, she will have mustered his feelings and susceptibilities in the service of her objective of winning him over. Provided that her romantic advances are not dishonest or coercive, her success in trading on Richard’s sentiments and vulnerabilities—her success in putting Richard’s primal feelings to work in behalf of her—is not morally illegitimate at all. Hence, the sheer fact that a torturer employs the elemental sensations of a victim as instruments and expressions of the torturer’s will is not sufficient to ground an ascription of moral wrongness to his actions. Indeed, even when the two main strands of my account of torture’s wrongness are combined, they in isolation from Chapter 2’s exposition of the nature of torture are insufficient to establish the conclusions for which this book is arguing. Detached from that exposition, the two strands recounted here do not suffice to distinguish between the actions of a torturer and the actions of a wrestler who applies a very painful hold to an opponent. As has been discussed in § 2.1.4.1.4 of Chapter 2, wrestlers often enwrap their opponents in holds for the purpose of inflicting pain that is severe enough to induce the opponents to surrender. Unless Walter the Wrestler has cheated in some fashion during a match in which he entangles his opponent Orville in an excruciating hold, his deliberate infliction of agony that will oblige Orville to yield is not morally impermissible. Yet, although Walter has not done anything wrong, he has deliberately exploited the feelings and vulnerabilities of his opponent as instruments for the achievement of his own objective of winning the match. Walter has made those feelings plead in support of the fulfilment of that objective, by imposing on Orville a level of agony that is so consuming as to force him to capitulate. Thus, unless the two principal components of my account of

176

Why Torture is Wrong

torture’s wrongness are conjoined with my specification (or some other credible specification) of the nature of torture, they will fail to sustain the proposition that torture is morally wrong—for they will fail to differentiate between torture and the morally irreproachable activity of wrestling. 3.2.2.2.2.  Morally vitiating purposes Even before we draw together the two chief elements of my account of torture’s wrongness with Chapter 2’s investigation of the nature of torture, that investigation itself (in combination with some elementary moral assumptions) has enabled us to see straightaway that countless possible instances of torture are morally wrong. The table in § 2.2.13 of Chapter 2 indicates as much. For example, given that the purposes pursued through sadistic torture and discriminatory torture and humiliative torture and extravagantly reckless torture are inherently illegitimate, all possible instances of those types of torture are likewise morally illegitimate. Furthermore, although the general purposes behind the remaining types of torture are not inherently wrong, they are wrong in most of their possible instantiations. For example, intimidatory torture—whether victim-specific or broadly addressed—is typically used by vile regimes to consolidate their ascendance over the people whom they tyrannize. Such an evil purpose is sufficient in itself to render illegitimate every instance of torture that pursues it. Similarly, interrogational torture and act-impelling torture are frequently employed by despotic regimes to obtain information and bring about patterns of behaviour that can help to maintain the evil sway of those regimes. Again, such a purpose morally invalidates any instances of torture that are aimed at serving it. Without having to penetrate beyond the level of purposes, we can know that those instances are morally impermissible. (Of course, even though an enquiry that penetrates more deeply is not necessary for our knowing that the aforementioned instances of torture are morally impermissible, such an enquiry would be necessary for a full moral assessment of the gravity of the wrong that has been committed in each case.) In diverse other situations, however, the purposes sought through torture are not themselves morally illegitimate. In such situations, then, any assessments of torture as morally wrong have to go beyond the objectives that are being pursued through it. For example, in the ticking-bomb scenarios that have preoccupied so many philosophers and jurists and politicians in Western liberal democracies during recent years, we encounter predicaments in which some officials resort to interrogational or act-impelling torture for the morally worthy purpose of saving numerous lives. In the more realistic settings of small-scale emergencies, the use of interrogational or act-impelling torture can again sometimes serve the morally worthy purpose of saving lives. Much the same is true of the use of placatory torture in the scenario by Alan Gewirth which I have recounted (with some modifications) in § 2.1.4.1.2 of Chapter 2. Intimidatory torture can likewise play such a role in some imaginably extreme situations, and ephemerally incapacitative torture can play such a role on many possible occasions. Moreover, even when torture is undertaken for purposes other than the saving of people’s lives, the ends sought can be morally worthy. For example, if constables

Why Torture is Wrong

177

or adjudicative officials administer interrogational torture to gather information that can help to identify the perpetrator of a serious crime, the objectionableness of their actions does not lie in the purpose which they are pursuing. Similarly, if the officials in a system of criminal justice administer punitive torture to visit a heinous malefactor with a form of hard treatment which he deserves for his crimes, the retributive principles with which they seek to comply are not morally illegitimate. Even more clearly, the aims that underlie quite a few instances of edifying torture are not morally dubious. For example, toughening the endurance of a soldier through open-ended resistance training is a way of pursuing a commendable goal. In all the situations mentioned in the last couple of paragraphs, and in any other relevantly parallel situations, torturous techniques are brought to bear for the attainment of ends that are morally unexceptionable in themselves. To be sure, this chapter will argue—in line with what has already been suggested in Chapter 2—that some possible instances of ephemerally incapacitative torture and edifying torture are morally permissible. However, in regard to other possible instances of those two types of torture and in regard to all possible instances of the sundry remaining types of torture, this chapter will contend that the moral prohibition on torture is firmly applicable. In relation to each of those remaining types of torture, that moral prohibition is absolute. Given as much, and given that some instances of several of those remaining types are oriented toward objectives and concerns that are morally unexceptionable in themselves,14 this chapter has to look beyond the objectives which torturers pursue in their pain-inflicting endeavours. We have to turn our gaze instead on those endeavours themselves. 3.2.2.2.3.  The Minimal Invasion Principle In my book The Ethics of Capital Punishment (2011), I relied heavily at many junctures on a basic principle of political morality: the Minimal Invasion Principle. According to that precept, the pursuit of any morally legitimate public end by legal-governmental officials is itself morally illegitimate if it is undertaken through measures that are more invasive than some other feasible measures that would have been sufficient to achieve the specified end. On the basis of that principle, we can judge many further instances of torture to be morally wrong. Even a superficial inspection of the practice of torture in interrogational contexts, for example, will reveal that it is a highly invasive method of obtaining information from uncooperative suspects or witnesses. Whenever any less invasive methods of interrogation are feasible and are at least as likely or nearly as likely to succeed in uncovering the desired information, legal-governmental officials in pursuit of a legitimate public 14   As has already been remarked here, and has been indicated in § 2.2.13 in Chapter 2, the objectives pursued through sadistic torture, discriminatory torture, humiliative torture, and extravagantly reckless torture are morally illegitimate in all cases. For each of the other eight types of torture that I have delineated in Chapter 2, some possible instantiations are oriented toward objectives that are themselves morally legitimate.

178

Why Torture is Wrong

purpose are morally obligated under the Minimal Invasion Principle to refrain from resorting to torture. (Of course, if the officials are not in quest of a legitimate public purpose, that fact on its own is sufficient to vitiate the moral status of their use of torture.) This point about the availability of less invasive alternatives is quite far-reaching, for—as I have suggested in Chapter 2—constables and other legal-governmental officials (along with members of the general public) are oft inclined to overestimate both the need for torture and the probable efficacy of torture in interrogational contexts. If a less invasive means of attaining the officials’ information-seeking ends is feasible, the Minimal Invasion Principle obligates the officials to eschew any recourse to torture. Much the same is true, mutatis mutandis, in non-interrogational contexts. For example, if the thwarting of a very serious criminal act in a given situation can be accomplished through some method that is less invasive than ephemerally incapacitative torture, the police or other officials in that situation are morally obligated to abstain from such torture. Although ephemerally incapacitative torture can be morally permissible in some of its possible instantiations, it is never permissible when its salutary effects can be achieved through an approach that is less harmful. Although here and in my book on capital punishment I  have presented the Minimal Invasion Principle as a precept of political morality, its requirement is generalizable to private individuals. Whenever someone is pursuing a morally legitimate end, he is morally obligated to forgo any tactics that are more invasive or restrictive than some other available tactic for attaining that end. If a private individual in an extreme emergency is confronted with a quandary in which his resorting to torture might resolve the emergency, and if some other feasible course of action might resolve the emergency and would be significantly less harmful, the generalized Minimal Invasion Principle morally obligates the individual to decline to wield any torture. In that respect, the public/private divide is not of any fundamental moral importance. (Chapter 5 will return to the moral significance of the public/private distinction.) In every case where the Minimal Invasion Principle requires the eschewal of torture because of the availability of some less harmful approach to a predicament, its applicability is sufficient to render morally illegitimate any recourse to torturous measures. A myriad of possible instances of torture are indeed proscribed by the Minimal Invasion Principle. However, not all possible instances of interrogational torture run afoul of that principle. In a grave emergency, such as the situation depicted in Chapter 2’s retelling of Leon v. Wainwright, the use of interrogational torture can be the only feasible course of conduct that is likely to avert the occurrence of some calamity. In a plight of that kind, the use of such torture would not contravene the requirements of the Minimal Invasion Principle. In a plight of that kind, moreover, the objective sought is itself morally worthy. Consequently, if this chapter is to show that interrogational torture (along with most other types of torture) is always morally impermissible even in dire emergencies, it cannot simply invoke the Minimal Invasion Principle or the morally vitiating force of wrongful objectives.

Why Torture is Wrong

179

3.2.2.2.4.  Consequentialist calculations On the one hand, this chapter has expressed some doubts in response to consequentialists who argue that a metastasizing spread of torture will inevitably occur—or very likely occur—in the event of any recourse to torture in a dreadful emergency. On the other hand, those consequentialist concerns and other consequentialist concerns can clearly be pertinent in some possible contexts. On some conceivable occasions, if legal-governmental officials resort to torture to avert calamities, they will foreseeably render more probable the employment of torture in future circumstances where it is not needed or where it is unlikely to be efficacious or where in some other way it will be more harmful than beneficial. Insofar as the use of torture in any given emergency will probably engender such an effect, and insofar as the future drawbacks of that effect outweigh any present benefits that are apt to be gained through torture in the particular emergency, consequentialist considerations weigh decisively against the permissibility of resorting to torturous techniques. Notoriously, however, consequentialist calculations are resistant to absolute moral prohibitions. Consequentialist considerations can on balance indeed weigh against the employment of torture in this or that emergency, but in various other dire emergencies such considerations can militate decisively in favour of the employment of torture. This point becomes especially evident when we recall that, among the philosophers and jurists in Western countries who affirm the moral permissibility of the use of torture for the extraction of calamity-averting information in terrible emergencies, most are firmly opposed to any institutionalization of the practice. That is, most of those philosophers and jurists contend that the wielding of torture for the extraction of such information is ever morally permissible only on an ad hoc basis. Whereas consequentialist concerns about the institutionalization of torturous methods of interrogation are quite far-reaching, the consequentialist considerations against the use of torture on an ad hoc basis in some desperate emergencies are much more variable. Though it would be implausible to maintain that the consequentialist factors against the wielding of ad hoc torture are always outweighed by consequentialist factors that cut in the opposite direction, it would likewise be implausible to maintain that the latter factors are always outweighed by the former. (To be sure, as we have seen in § 3.1.3.2, some consequentialists submit that any ad hoc employment of torture would not really occur ad hoc—since it would have to draw on the skills of torturers who had undergone institutionalized training for the task. However, as I have argued in response, that consequentialist line of reasoning is unfounded.) In short, although consequentialist considerations will of course sometimes militate against the use of torturous measures even in this or that dreadful emergency, such considerations will in other cases militate in favour of those measures. As some earlier portions of this chapter have indicated—and as is similarly indicated by the voluminous philosophical literature on consequentialism, including the relevant parts of Chapter 1 of this book—consequentialist doctrines are decidedly unreliable as a basis for the justification of any absolute moral prohibition on torture. Accordingly, this book’s effort to vindicate just such a prohibition on most types of torture has to proceed from a different basis.

180

Why Torture is Wrong

3.2.2.2.5.  Agony and oppression: the factors of consent and control afresh When the two main strands of my account of torture’s wrongness are brought together with Chapter  2’s explication of the nature of torture, they furnish the materials for a vindication of an absolute moral prohibition on most types of torture (including interrogational torture—the type that has received far more attention than any other in recent years). Admittedly, if the absoluteness of that prohibition is to be fully upheld, those elements of my account of torture’s wrongness will have to be subsumed into a perpetrator-focused perspective. Still, even before we explore the need for such a perspective, we can begin to understand from those elements why torture in nearly all of its instantiations is an especially abhorrent mode of conduct. Because a process of torture induces in the victim the self-betrayal highlighted by Sussman, it is a process in which the victim’s very bodily feelings and vulnerabilities become the vehicles and expressions of the torturer’s ascendance. As the torturer turns the victim against herself by making her primal sensations the offshoots of his will, his power over her is singularly intimate in its workings. Because that power exerts itself through the evocation of exquisite pain in her, it subjects her to the radical absorbingness of agony. Even in the exceptionally rare circumstances where a victim is unaware that her tribulations are due to her having come within the power of someone else, her attention cannot but be fixated on those tribulations. Agony screeches for attention. It does so, furthermore, because it is overwhelmingly disagreeable (outside sado-masochistic contexts); its clamours for attention are pari passu clamours for relief. Hence, the ascendance of the torturer over the victim not only is peculiarly intimate in its reach and operations, but also is peculiarly consuming. So long as the excruciating effects of that ascendance last, they dominate the psyche of the victim. Although those effects might not completely obliterate everything else in the outlook of the victim, they permeate her awareness of everything else. Even in an extraordinarily uncommon situation where a victim of torture is able to defend herself to some degree (as in Steinhoff’s scenario of the jeweller and the robber), she does so while in the thrall of the terrible pain that the torture is eliciting in her. She does so, in other words, while her own sensations are serving as vehicles of the torturer’s will by engulfing her in the torment which insistently racks her and which thereby advances his purposes. Now, as has been suggested at the end of § 3.2.2.2.1, the remarks in the preceding paragraph have not yet furnished a basis for distinguishing between situations of torture and some situations of the morally legitimate activity of wrestling. However, the basis for a germane distinction emerges from Chapter 2’s explication of the nature of torture. Whereas a process of torture is very rarely consensual and whereas it never involves control by the victim over the duration of the process, a wrestling match in any typical circumstances is consensual, and—unless it is itself a form of torture—each participant can terminate it when he so chooses by conceding the match to his opponent. Of course, if a wrestling contest is such that a participant P and his opponent know that P will incur a severe penalty (such as the loss of his life) in the event of his surrendering to the opponent, then P does not have any genuine control over the duration of the contest. Any control exercisable by him

Why Torture is Wrong

181

is egregiously coerced rather than genuine. In such a situation, the opponent’s enfolding of P in a grievously painful grip would amount to a form of torture. Fittingly, then, a nightmarishly grotesque contest of that sort is hardly what people have in mind when they contrast the infliction of pain by a torturer with the infliction of pain by a wrestler. Rather, they have in mind an ordinary match in which each participant enjoys genuine control over the duration of any agony that he might undergo at the hands of his opponent. In such a wrestling match, where each participant does possess genuine control over the lengthiness of his own suffering, any pain inflicted—however searing it may be—is not a product of torture. My focus on consent and control in the preceding paragraph not only enables us to distinguish between the deliberate inducement of severe pain in a process of torture and the deliberate inducement of severe pain in the non-torturous activity of wrestling. In addition, that focus alerts us to a key moral difference. Let us assume for now that, in any contrast between a process of torture and a match of wrestling, the torture is of a type that is always and everywhere morally wrong. Hence, the distinction between torture and a non-torturous pastime is here conjoined with a distinction between moral illegitimacy and moral legitimacy. In the present context, the factors of consent and control underlie the latter dichotomy as well as the former. Why do they make a moral difference? Situated within my account of torture’s wrongness, the answer to this question becomes plain. When a victim is used for a torturer’s purposes by being subjected to excruciating pain that turns her feelings into servants of the torturer, her subordination is both distinctively profound and distinctively consuming. Though her ability to move her body might in some cases be less tightly restricted than the ability of someone who is non-torturously shackled, her inner impulses are being marshalled against her to a degree that is not equalled by any non-torturous methods of confinement. Now, given that a wrestler W has consented to participate in a match, and given that he enjoys genuine control over the prolongedness of any agony that he will suffer during the match, the subjection of W to severe pain by his opponent does not reduce him to an instrument or offshoot of the opponent’s purposes. His opponent is of course pursuing the purpose of winning the match and is making use of W’s elemental feelings and vulnerabilities in furtherance of that purpose, but the two wrestlers also share the central purpose of putting each other to a test of strength and dexterity. That central purpose is, indeed, what structures their whole contest. When W is subjected to dire pain by the opponent, that shared purpose is being fulfilled just as is the opponent’s unshared purpose of prevailing over W. As W retains genuine control over the duration of any agony that he will experience, and as he shares the central purpose that is realized through the occurrence of the contest with his opponent (whether he wins or loses), the opponent’s exploitation of W’s susceptibility to severe pain does not treat W as a mere plaything or as a mere vehicle of the opponent’s objectives. Unlike a victim of torture, then, W has been treated with the respect and concern that are owed to him by his opponent and by everyone else as a fundamental matter of morality.

182

Why Torture is Wrong

3.2.2.2.6.  Some implications: edifying torture and sado-masochism In the past couple of paragraphs, I have assumed that the type of torture under discussion is any of the types that are always and everywhere wrong. However, the remarks in those paragraphs help to clarify why some possible instances of edifying torture are morally permissible. Suppose for example that a soldier or an intelligence official chooses to undergo a course of resistance training in which he will not have any control over the protractedness of the torturous hardships to which he will be subjected. His reason for forgoing any control is that he will thereby make the resistance training considerably more effective in toughening him against the physical and psychological adversities that he might subsequently encounter in his dangerous line of work. Thus, although the resistance training in the open-endedness of its deliberate inducement of severe pain is a process of veritable torture rather than just a simulation of torture, it is not morally impermissible. As Chapter 2 has suggested, the moral legitimacy of such a course of resistance training derives from its edifying and consensual character: that is, from the benevolence of its orientation toward the victim’s bodily and psychological well-being, and from his genuine consent to the process. With those factors present, the resistance training does not treat the victim as a toy or as a mere implement of someone else’s purposes. On the contrary, it is one of his projects (or an element in one of his broader projects), and its exploitation of his susceptibility to agony is centrally aimed at enhancing his physical and mental hardihood. Like W in the wrestling match, then, the soldier or intelligence official has been treated with the elementary respect and concern that are morally due to him. My remarks in the last few paragraphs might also appear to imply that some possible instances of sado-masochistic torture are morally permissible. In any sado-masochistic interaction that amounts to veritable torture rather than just to a simulation of torture, the infliction of severe pain is open-ended; the victim cannot terminate it at a time of his choosing. However, if the open-endedness is something to which the victim eagerly consents (in order to augment his sexual pleasure), the eliciting of severe pain in him might appear to be morally assimilable to the inducement of severe pain in a soldier or an intelligence official during resistance training. Although the process of afflicting the sado-masochistic victim with agony is not aimed at strengthening his physical and psychological fortitude, it is aimed at providing him with sexual gratification. It therefore counts as one of his projects, on which he has very willingly embarked. Still, despite those similarities between the situation of the sado-masochistic victim and the situation of the soldier or intelligence official, we should not leap to the conclusion that sado-masochistic torture can sometimes be morally permissible. Although there are no victim-focused considerations that undermine the permissibility of such torture, this chapter will later adduce some perpetrator-focused considerations that do indeed vitiate the moral status of torture that is undertaken for sado-masochistic excitement. A  victim of such torture does not commit a wrong by consenting to undergo it, but the perpetrator of the torture commits a wrong by taking advantage of the victim’s consent. We shall return to this matter presently.

Why Torture is Wrong

183

3.2.2.2.7.  Some implications: the problem of animals Even knottier than my discussion’s implications concerning resistance training and sado-masochism are its implications concerning animals. Like infants and lunatics, non-human animals are incapable of the sophisticated understanding and evaluation and reflection and communication that would enable them to furnish genuine consent to processes of edifying torture. (Likewise, they are incapable of exercising informed control over the duration of any processes in which they are subjected to severe pain. They of course can and would exhibit attitudes of aversion in response to such processes, and they would escape from any pain-inducing measures at the earliest opportunities open to them. However, that primitive form of control over the lengthiness of those measures would not be grounded in any informed understanding of the benevolence of the purpose which the edifying inducement of pain is designed to achieve.) Now, given that meaningful consent is never operative in any case of edifying torture in which the victim is a non-human animal, some readers might presume that the application of such torture to any non-human creature is always morally illegitimate. However edifying the torture might be, it would not amount to a project undertaken by the animal who is its victim. Consequently, some readers might conclude, the administration of such torture would not show due respect and concern toward an animal as a sentient creature with a life of its own and with a constitution susceptible to the overwhelming unpleasantness of severe pain. A conclusion to that effect, however, would be hasty. Very rare are any circumstances in which the administration of torture is truly edifying for an animal, but some such circumstances are conceivable. Suppose for example that, before some particular animal can be safely released into the wild, it has to become inured to certain major adversities that it will face there. Given as much, a process of subjecting the animal to grievous pain—akin to the pain that it will experience through those adversities—can be an instance of genuinely edifying torture. Though the animal is incapable of assenting (or informedly refusing to assent) to such a process, the absence of any consent is due solely to that inability; the torturers have not declined to seek consent whenever it can be meaningfully rendered or withheld. In such circumstances, if the animal will benefit greatly from being released into the wild despite the hardships that it will encounter there, and if the only adequate way of preparing it for those hardships is to put it through a process of edifying torture of the kind envisaged here, the administration of such a process will be morally legitimate. The orientation of the whole endeavour toward the animal’s physical and mental well-being is benign, and the lack of consent by the creature derives not from any high-handedness on the part of the torturers but from the sheer infeasibility of the provision or refusal of such consent. Accordingly, the torturers do not exhibit any disrespect or unconcern for the creature when they subject it to agony. By contrast, the subjection of an animal to experimentational torture—of the sort described in § 2.2.12.2.1 of Chapter 2—is generally not morally permissible. Though such torture can be aimed at the morally legitimate end of adding to medically valuable knowledge, that end does not render permissible the morally

184

Why Torture is Wrong

illegitimate means by which it is pursued. That is, it does not render permissible the deliberate and non-consensual infliction of severe pain on an animal solely for the benefit of others. Such a way of treating an animal does not show proper concern for its susceptibility to agony or proper respect for its status as a creature with a life of its own. To be sure, in some imaginable circumstances, a process of experimentational torture could be aimed at benefiting greatly the very animal on which it is brought to bear. In such circumstances, although the experimentational torture could never be truly consensual, it could be morally permissible on grounds similar to those on which the edifying torture described in the last paragraph is morally permissible. Still, notwithstanding that such circumstances are imaginable, they are rarely if ever actual. Usually, the deliberate inducement of severe pain in an animal for the purpose of expanding medical knowledge is not aimed at benefiting the particular animal in whom the pain is elicited. Any such experimentation is morally impermissible. This short discussion of animals and torture should close with two caveats. First, the experimentational torture assessed in the preceding paragraph involves the deliberate infliction of severe pain on an animal. I  have not there ruminated on experimentation in which the affliction of an animal with severe pain is an undesired though amply foreseeable side-effect rather than something sought as a means. If some experimentation with such a side-effect will predictably be of little or no medical value—like the Mengelian experimentation envisaged in § 2.2.10 of Chapter 2—then its amply foreseeable infliction of agony on an animal is an instance of extravagantly reckless torture. As has already been remarked in this chapter, extravagantly reckless torture is never morally permissible. Similarly, even if the experimentation is likely to be of significant medical value, its amply foreseeable inducement of agony in an animal is an instance of extravagantly reckless torture if the agony could have been avoided through reasonable precautions. However, if the experimentation is likely to be of considerable medical value, and if its affliction of an animal with terrible pain is unavoidable through any measures short of the outright cessation of the experimentation, then the affliction is not an instance of extravagantly reckless torture. Because the process that induces the pain is undertaken for ethically weighty reasons rather than without any substantial justificatory grounding, it is not reckless (even though the experimenters who undertake the process have knowingly hazarded a very high risk of causing exquisite pain). It therefore does not amount to an instance of extravagantly reckless torture. Moreover, whereas the eliciting of severe pain through the experimentation is abundantly foreseeable, it is not intentional; the pain is not sought either as a means or as an end. Thus, since every instance of torture inflicts agony either intentionally or in an extravagantly reckless fashion, the medically valuable experimentation contemplated here is not an instance of torture at all. Hence, there is no occasion for this book to take a position on the question whether such experimentation can ever be morally legitimate.15 15   Among the challenges to be faced by someone who wants to answer that question affirmatively is the task of differentiating an animal such as a chimpanzee or a dog from a severely retarded or demented human being.

Why Torture is Wrong

185

A second caveat pertains to a different kind of limit on the present discussion. My remarks here have focused on animals and edifying torture. (Recall that § 2.2.12.2.1 of Chapter  2 construes the category of edifying torture expansively enough to include the sorts of experimentational torture that I  have pondered in the current subsection.) My present discussion has therefore left open the possibility of morally legitimate instances of ephemerally incapacitative torture wielded against animals. On the one hand, the moral legitimacy of the use of such torture against a normal human adult partly depends on whether the adult is performing a very seriously wrongful action at the time. Given that animals are not moral agents and are thus not capable of performing very seriously wrongful actions, the conclusion might seem to follow that they can never permissibly be subjected to ephemerally incapacitative torture. On the other hand, although animals are not capable of performing very seriously wrongful actions, they are obviously capable of performing extremely harmful actions (actions that would be very seriously wrongful if the animals were possessed of moral agency). If the infliction of ephemerally incapacitative torture is the least injurious feasible way of preventing an animal from wreaking major harm—a less injurious way than killing the animal outright, for example—then the plying of such torture against the creature, for the specific purpose of preventing or minimizing the harm, can be morally permissible. Much the same is true of the plying of such torture against a mentally infirm human being. A bit later in this chapter, after I have elaborated a perpetrator-focused perspective on these matters, we shall return to the topic of ephemerally incapacitative torture to explore why the use of such torture can ever be morally permissible. 3.2.2.2.8.  Agony and oppression redux: a recapitulation and a transition Key among the wrong-conducive characteristics of torture, visible from a victim-­ focused perspective, are the considerations that have been adduced heretofore: the marshalling of the primal feelings and vulnerabilities of a victim against her as expressions and instruments of the ascendance of the torturer; the infliction of intense pain which, in its overpowering disagreeableness, fixates the victim on itself and hence dominates her consciousness; and the absence of any genuine control by the victim over the duration of the process of torture. Torture is also usually non-consensual, and the employment of it usually bespeaks the torturer’s hostility or indifference toward the bodily and psychological well-being of the victim. In most contexts those properties of torture are jointly sufficient to undermine its moral legitimacy, for in most contexts those properties are inconsistent with the elementary respect and concern that are due to a victim as a sentient and vulnerable creature. Given those features of torture, the infliction of it on anyone treats her—with her susceptibility to excruciating pain—as a plaything or as a mere vehicle for the realization of the torturer’s purposes. In most contexts, being treated in such a way is at odds with a victim’s fundamental dignity (her human dignity as a reflective agent, or her basic dignity as a creature capable of feeling agony and of undergoing positive experiences). As is suggested by the qualification ‘in most contexts’, however, the victim-focused considerations just adduced are not always sufficient to undermine

186

Why Torture is Wrong

the moral legitimacy of torture. In particular, even when all those considerations are operative, ephemerally incapacitative torture can be morally permissible in some contexts. Of course, not all those considerations are operative on some of the occasions when such torture is wielded. If the prevention of a very serious criminal act is possible not only through ephemerally incapacitative torture but also through the outright slaying of the malefactor, and if the police opt to employ such torture because it is the less harmful of the two feasible ways in which they can stymie the performance of the very serious criminal act, then they are not displaying hostility or indifference toward the malefactor’s bodily and psychological well-being. In some other cases, however, ephemerally incapacitative torture is the only feasible way of preventing the performance or culmination of a very serious criminal act. In Sussman’s scenario of the obese man, for example, the slaying of the man would be woefully counter-productive. Given that the man does not respond to desperate exhortations, the plying of ephemerally incapacitative torture against him is the only way of saving the person who is squashed beneath him; such torture is not chosen as the less harmful of two available alternatives. In those circumstances, then, the ephemerally incapacitative torture does bespeak hostility or indifference toward the bodily and psychological well-being of the corpulent man—yet, as a means of thwarting his continuation of his very serious criminal act, the use of the torture is morally legitimate. Hence, in some contexts of ephemerally incapacitative torture, all the considerations adduced so far are not sufficient to render the torture morally wrong. In a sense, the fact that ephemerally incapacitative torture can in certain circumstances be morally permissible—despite the presence of all the wrong-conducive factors that we have pondered so far—is attributable to the forfeiture of various rights by people who commit or seek to commit grave wrongs. Such people forfeit their rights against being fined or imprisoned. In cases of extreme evil, as I have argued elsewhere (Kramer 2011, 258–9), they even forfeit their rights against being executed. Accordingly, given that ephemerally incapacitative torture can be morally permissible when it is the least invasive means of foiling the commission of a very serious crime, we should conclude that someone who commits or endeavours to commit a crime of that sort has forfeited his right against being subjected to such torture. Still, that conclusion gives rise to more questions than it resolves. After all, this chapter is maintaining that interrogational torture and punitive torture are always and everywhere morally wrong. Nobody can ever forfeit his right against being subjected to punitive torture or interrogational torture, yet some of the people who might undergo torture as a method of interrogation or punishment are guilty of crimes at least as grave as those that could legitimately be prevented through the use of ephemerally incapacitative torture. Why does such a difference obtain? Why can major malefactors sometimes be subjected legitimately to ephemerally incapacitative torture, whereas no such malefactor (or anyone else) can ever be subjected legitimately to punitive or interrogational torture? Neither the doctrine of forfeiture nor the several victim-focused considerations recapitulated in the opening paragraph of this subsection will suffice to answer this question. An appeal to the possibility of the forfeiture of rights is unavailing in

Why Torture is Wrong

187

itself, since we need an explanation of why that possibility applies to ephemerally incapacitative torture but not to punitive or interrogational torture. An appeal to the victim-focused factors rehearsed in this subsection’s opening paragraph is likewise unavailing in itself, since all of those factors are operative in many contexts where the use of ephemerally incapacitative torture is morally permissible. The applicability of all of those victim-focused considerations to any torturous methods of punishment or interrogation does not differentiate those methods from many morally legitimate impositions of ephemerally incapacitative torture. Quite plain, then, is the need for my account of torture’s wrongness to go beyond the victim-focused perspective that has prevailed hitherto. Although the insights gained from that perspective are vital, and although they will continue to be invoked herein, they have to be supplemented and reconfigured by the introduction of a perpetrator-focused perspective. Only thus will this chapter be able to indicate why some possible instances of ephemerally incapacitative torture are morally permissible whereas all possible instances of punitive or interrogational torture are morally impermissible.

3.2.3.  A perpetrator-focused perspective We have just found one reason why this chapter has to supplement its victim-focused perspective with a perpetrator-focused perspective. We should here mull over two further reasons why the former perspective on its own is inadequate for an account of the wrongness of torture. Each of those additional reasons is closely related to the reason that has already been broached, and each of them pertains to situations that do not tally with the central premise of any victim-focused account of the wrongness of torture: namely, the premise that the justificatory basis for the moral prohibition on torture resides in the interests or inviolability of the potential victims thereof.

3.2.3.1.  Monstrous victims of torture In the standard presentation of the ticking-bomb scenario, the person under interrogation is a mass-murdering terrorist (such as Khalid Sheikh Mohammed) who is seeking to add to the string of atrocities for which he has been responsible. This chapter, with its insistence that the moral prohibition on interrogational torture is absolute, has to maintain that the use of such torture against a mass-murdering terrorist is always morally wrong. Now, if my account of the wrongness of torture were exclusively victim-focused, I would be committed to the proposition that the impermissibility of the use of interrogational torture against any mass-murdering terrorist is attributable to his interests or inviolability. His interests or inviolability would constitute the justificatory basis for the absolute moral ban on the wielding of such torture against him. In fact, however, the interests of a mass-murdering terrorist in being free from excruciating pain are of no positive ethical weight. Were such a person to experience agonizing pain as a result of purely natural causes while in isolation from any society

188

Why Torture is Wrong

whose medical resources might alleviate his suffering,16 the world would not be ethically inferior to an otherwise identical alternative world in which he never undergoes any agony. His interest in being free from terrible pain does not weigh in favour of a world where that interest is realized. (If it were necessary for me to do so here, I would argue for a bolder proposition. That is, I would contend that a world in which a mass-murdering terrorist suffers from agonizing pain through purely natural causes is ethically superior—on grounds of ‘just deserts’  —to an otherwise identical alternative world in which the terrorist never experiences any such pain. However, the present context requires only the milder proposition that the former world is not ethically inferior to the latter.) Hence, if we concentrate solely on the interests of potential victims of torture when we explain why interrogational torture is wrong, we shall be unable to say why the use of interrogational torture against Khalid Sheikh Mohammed or any other heinous murderer is morally wrong. As can be inferred from my remarks about the forfeiture of rights in § 3.2.2.2.8, we shall face a parallel difficulty if we concentrate solely on the inviolability of the potential victims of torture. Although every person is endowed with absolute inviolability (weakly absolute inviolability) by dint of his or her status as a reflective agent,17 an extravagantly evil criminal will have forfeited many of the rights that are held by any ordinary human being. If we are to ascertain whether his right-not-to-be-subjected-to-interrogational-torture has been forfeited along with his right-not-to-be-deprived-of-his-freedom-through-incarceration (for example), we have to go beyond appealing to his inviolability. Because the protective sway of his inviolability encompasses only the rights which he has retained despite his responsibility for iniquities, one’s invocation of that inviolability will go no way toward telling us whether he still enjoys a right-not-to-be-subjected-tointerrogational-torture. By contrast, a shift to a perpetrator-focused perspective will enable us to answer that latter question affirmatively. It will enable us to apprehend the correctness of John McCain’s insistence that the moral basis for our abstention from the use of torture against wicked terrorists is not about the terrorists but is instead about us (McCain 2006).

3.2.3.2.  Being killed versus being tortured A further reason for embracing a perpetrator-focused perspective (along with a victim-focused perspective) is that the deliberate killing of people is morally permissible in some contexts. For example, if a gunman has taken several people hostage, and if he begins to murder them when the demands which he has issued are not met, sharpshooters from the police will be acting permissibly if 16   My reason for including this qualification about the evildoer’s isolation is that one’s deliberate sustainment of another person’s pain through one’s inaction can itself amount to torture and can thus be wrongful. For an account of an instance of torture-through-deliberate-inaction, see Moore 2007, 37 n7. 17   I am not here implying that such a status is the only basis for weakly absolute inviolability. Rather, I am simply indicating that it is a sufficient basis.

Why Torture is Wrong

189

they resort to lethal force against him as the only feasible means of saving the remaining hostages. Similarly, if Joe and his young daughter are attacked by a murderous assailant as they are walking down some street, and if the only effective way in which Joe can fend off the attack is to kill the assailant, he will be acting permissibly in so doing. Furthermore, as I have argued elsewhere (Kramer 2011, 256–60), executions conducted on the basis of the purgative rationale for capital punishment can be morally legitimate in cases where verdicts of guilt are returned against defilingly evil offenders who have been placed on trial fairly for the iniquities which they have committed. Now, being killed is more harmful than being subjected to certain techniques of interrogational torture (such as the twisting of one’s arm behind one’s back for a limited period). It amounts to a much greater invasion of the victim’s interests. Given as much, and given that the deliberate killing of a person can be morally permissible in some circumstances, we cannot explain the absoluteness of the impermissibility of interrogational torture while focusing exclusively on the interests of potential victims. A shift to a perpetrator-focused viewpoint will enable us to understand that, although the use of interrogational torture is not always more harmful or more gravely immoral than deliberate killing, it differs from such killing in that it is always and everywhere morally wrong.

3.2.3.3.  An additional perspective In sum, by supplementing its victim-focused perspective with a perpetrator-focused perspective, this chapter can resolve at least three cruxes: why ephemerally incapacitative torture is sometimes morally permissible whereas interrogational torture and punitive torture are always morally impermissible; why the use of interrogational torture for the extraction of calamity-averting information from monstrous terrorists in circumstances of extreme desperation is morally impermissible; and why the use of interrogational torture is always morally impermissible even though deliberate killings are sometimes morally permissible. Another crux that can be resolved here, outlined in § 3.2.2.2.6 above, is the question why all possible instances of sado-masochistic torture are morally wrong even though some other possible instances of edifying torture are morally permissible. Were we to cleave to a purely victim-focused perspective, we would not be able to come to grips adequately with any of these cruxes. As a consequence, this book would not be able to vindicate fully the absoluteness of the moral prohibition on interrogational torture and punitive torture. A switch to a perpetrator-focused viewpoint will enable us to see why that prohibition is absolute, and will also reveal that that absolute prohibition extends to placatory torture and act-impelling torture and extortionate torture and intimidatory torture. Although some possible instances of each of those types of torture are undertaken in pursuit of morally worthy ends, certain perpetrator-focused considerations tell decisively against the moral legitimacy of every such instance. (As has already been noted in this chapter, the ends sought through sadistic torture and discriminatory torture and humiliative torture and extravagantly reckless torture are always themselves morally wrongful. Hence,

190

Why Torture is Wrong

a vindication of the absoluteness of the moral ban on each such kind of torture can simply advert to those ends. Of course, if we were to seek to gauge the gravity of the wrong committed by someone who engages in any of those kinds of torture, we would need to go beyond the general purpose of the act of torture in order to examine its specifics.) 3.2.3.3.1.  A perpetrator-focused justificatory basis What, then, is the perpetrator-focused viewpoint that should supplement the victim-focused viewpoint on which this chapter has relied heretofore? As has been suggested in Chapter  1’s terse remarks on the matter, we supplement the latter viewpoint with the former by moving away from an exclusive concentration on the interests of the potential victims of torture (most notably their interests in being free from excruciating pain and from the controlling ascendance of others). We have to concentrate additionally on the moral integrity of the potential perpetrators of torture. Of course, this redirection of attention does not mean that the victim-focused considerations simply get left by the wayside. On the contrary, those considerations importantly inform the perpetrator-focused concerns that here come to the fore. The latter concerns, which bear on the moral integrity of the people who prescribe or undertake acts of torture, pertain to the outlooks through which those acts occur. However, this chapter is not embarking on a psychological or sociological study of the mental states and ulterior motivations of individuals who perpetrate torture. Insofar as those mental states and motivations are ascertainable through the methods of the social sciences, they are multifarious empirical phenomena. My philosophical and moral enquiry here prescinds from most of the empirical minutiae that would come under scrutiny in a social-scientific approach to the matter. Instead, the enterprise in this discussion is one of moral and philosophical assessment. What are the moral bearings of the general outlooks that are credibly ascribable to people who inflict the agony of torture on others? An answer to this question can become clearer if we begin by contemplating a morally permissible instance of edifying torture. Suppose that Brenda is a soldier who has consented to undergo the open-ended infliction of severe pain as a component of her resistance training, and suppose that Daniel is one of the soldiers entrusted with the task of eliciting that pain. What is the moral character of the outlook of Daniel in relation to Brenda? He knows that she has consented to undergo the ordeal that is to be administered, and he further knows that the ordeal is designed to enhance her overall physical and mental well-being. Thus, notwithstanding that his inducement of severe pain in her is deliberate rather than an unintended though foreseeable side-effect, it is salutary rather than destructive in its orientation toward her bodily and psychological welfare. It manifests Daniel’s ascendance over Brenda during the open-ended period of the torture, but it involves no animosity on his part, and it does not stem from any effort by Daniel to treat Brenda as a plaything or to exploit her susceptibility to exquisite pain as a mere vehicle for the realization of his own ends (however laudable those ends might be). Consequently, it is not an endeavour that morally degrades him by requiring his

Why Torture is Wrong

191

adoption of an outlook that would be suitable for an exploitative effort. On the contrary, the outlook that best fits with his conduct is benevolently supportive of one of Brenda’s own projects. Suppose now, by contrast, that Sylvia is a mass-murdering terrorist—akin in her iniquity to Khalid Sheikh Mohammed—and suppose that Walter is one of the interrogators assigned to extract information from her that can help to avert an imminent calamity. After Walter tries several non-torturous techniques of interrogation without any success, the urgency of the situation and the absence of any effective alternatives lead him to resort to certain torturous methods of interrogation against Sylvia. His general purpose in availing himself of those methods, in circumstances of dire desperation, is per se morally unexceptionable. In fact, his pursuit of that morally worthy purpose in those circumstances might render his conduct weakly justified (in the sense specified by Chapter 1). Nevertheless, although his torturous actions might be morally optimal in the terrible predicament in which he performs those actions, they are not morally permissible. Even if the consequentialist duty fulfilled through his recourse to such conduct is more stringent than the deontological duty which he thereby breaches, that latter duty has indeed been breached. Of key importance here is that my absolutist position on the wrongness of Walter’s conduct does not attach any positive ethical weight to the interests of Sylvia. Because of her responsibility for past atrocities and for the impending atrocity, her interests in being free from searing pain do not contribute to any justification for a ban on the use of interrogational torture against her. Quite baseless is the allegation that someone who adopts an absolutist position on the wrongness of such torture has failed to grasp that ‘the putative victims of an atrocity deserve more care than the would-be perpetrators’ (Bobbitt 2008, 380). My insistence on the wrongness of Walter’s employment of interrogational torture does not derive from any solicitude whatsoever for the interests of Sylvia. No victim-focused considerations are sufficient to undermine the moral legitimacy of Walter’s wielding of torture against her for the purpose of averting a calamity. Rather, the illegitimacy of his use of interrogational torture derives from his sullying of his moral integrity through his adoption of a self-aggrandizing course of conduct. Walter has sullied his moral integrity—even if he has acted optimally—because of the differences between the outlook associated with his conduct and the outlook associated with Daniel’s subjection of Brenda to edifying torture. Whereas the latter outlook is neither hostile nor indifferent toward the victim’s bodily and psychological well-being, the former outlook partakes of just such hostility or indifference. Whereas Daniel’s inducement of intense pain in Brenda does not exploit her susceptibility to such pain as a mere vehicle for the realization of his own objectives, Walter’s inducement of intense pain in Sylvia does indeed exploit her susceptibility to such pain as a mere vehicle for the realization of his own objectives. His calamity-averting objectives are of course highly commendable and morally pressing, and Sylvia’s terroristic objectives are of course thoroughly odious. Still, because Walter has had to pursue his morally worthy objectives by treating another human being as a locus of vulnerability to severe pain that is to be employed as an instrument and an expression of his ascendance over her, and

192

Why Torture is Wrong

because he has had to exert that ascendance to impel her toward the performance of the type of action that suits his purposes—highly commendable purposes—he has taken (or has sought to take) control over her in a way that is inconsistent with his own elementary humility as a moral agent. That elementary humility would involve his showing basic concern even for a loathsome terrorist. Of central importance here, again, is that Walter’s owing of basic concern to Sylvia is not attributable to her interests in receiving such concern. As has already been emphasized, her interests are of no positive ethical weight. Instead, Walter’s owing of basic concern to her is attributable to his own moral integrity. Sylvia has led a profoundly evil life devoted to the unleashing of atrocities, but she remains a human being whose existence as such has to be acknowledged by Walter if his interaction with her is to be morally legitimate. In this context, his acknowledging her existence as a human being (or, indeed, as a sentient creature) would consist chiefly in his not afflicting her with excruciating pain as a means of taking control of her to direct her performance of some action. His taking control of her in that way is morally wrong not because it redounds to the detriment of her interests— interests that carry no positive ethical weight—but because it elevates him to a position of overweening dominance. His elevation of himself to such a godlike position morally degrades him, for it is incompatible with the fundamental modesty that is incumbent upon him. It is incompatible with his treating Sylvia as a separate human being (or sentient being) whose proneness to agony is not to be exploited as a vehicle for reducing her to an instrument or a marionette. 3.2.3.3.2. A first query: why is ephemerally incapacitative torture ever permissible? We should now ponder several queries, the first of which pertains to ephemerally incapacitative torture. Why is torture of that sort ever morally permissible, given that Walter’s recourse to interrogational torture against Sylvia is morally impermissible? On the one hand, every instance of ephemerally incapacitative torture (as understood throughout this book) is aimed at preventing a very serious criminal action.18 On the other hand, however, Walter likewise resorts to interrogational torture for the purpose of preventing a calamitous crime. His torture of Sylvia is of course non-consensual, but so too is any instance of ephemerally incapacitative torture against some malefactor. Walter’s torturous methods of interrogation bespeak his hostility or indifference toward Sylvia’s physical and mental well-being, but some legitimate instances of ephemerally incapacitative torture are similar in that respect. Furthermore, any instance of ephemerally incapacitative torture exploits a criminal’s vulnerability to severe pain, just as Walter’s plying of interrogational torture against Sylvia exploits her vulnerability. Why, then, is interrogational torture always morally impermissible—even in the circumstances of a dire emergency like the circumstances in which Walter acts—whereas some possible instances of ephemerally incapacitative torture are morally permissible? 18   Recall, from § 2.2.11.2 of Chapter 2, that this book employs the phrase ‘ephemerally incapacitative torture’ as shorthand for the more cumbersome phrase ‘evil-averting ephemerally incapacitative torture’.

Why Torture is Wrong

193

Before I present my main reply to this question, we should remember that the moral impermissibility of interrogational torture in all of its possible instantiations is consistent with its moral optimality in some of those possible instantiations. While this chapter is arguing that the moral prohibition on interrogational torture is weakly absolute, it is not arguing that that prohibition is strongly absolute. On the contrary, the chapter’s final main discussion will suggest that the use of such torture in some extreme emergencies (whether actual or only possible) can be morally optimal even though it is morally illegitimate. That is, in some conceivable situations, the wielding of such torture can fulfil a duty even more stringent than the duty which it breaches. Consequently, although my insistence on the absoluteness of the moral impermissibility of interrogational torture does differentiate torture of that kind from torture of the ephemerally incapacitative kind, the divergence between them is somewhat less sharp than it might initially appear. In extremely rare circumstances, torture of either kind can be morally optimal. My chief response to the question at the end of the penultimate paragraph above, however, does not reside in playing down the divergence between ephemerally incapacitative torture and calamity-averting interrogational torture. To understand the key moral difference between those types of torture, we should return to the act/ omission distinction that has been expounded at length in Chapter 2 (§ 2.2.11.2). Let us recall that acts are distinguished from omissions by reference to the numerousness of the ways in which someone can move or position her body and any concomitant objects. Given that an effect E occurs and that the conduct of someone S has had some bearing on E’s occurrence, S’s contribution to E is an action if most of the movements and positions available to S would not have led to E. Conversely, S’s contribution to E is an omission if most of the movements and positions available to S would have led to E. Now, any instance of ephemerally incapacitative torture is aimed at preventing the performance of a very serious criminal action, whereas any instance of calamity-averting interrogational torture is aimed at preventing an omission by impelling the performance of an act of disclosure. Accordingly, the dominion exerted through an instance of calamity-averting interrogational torture is much more minutely controlling than the dominion exerted through an instance of ephemerally incapacitative torture. An instance of torture of the latter sort is undertaken to induce the victim to forgo only a small proportion of the bodily movements and positionings available to her—some bodily movements or positionings that would amount to a very serious criminal action. By contrast, an instance of calamity-averting interrogational torture is undertaken to induce the victim to forgo most of the bodily movements and positionings available to her (namely, all such movements and positionings that do not involve her divulging the information that is demanded of her). Thus, although ephemerally incapacitative torture induces the distinctive self-betrayal described by Sussman, and although it is expressive and promotive of the ascendance of the torturer over the victim through the exploitation of the victim’s susceptibility to agony, and although the victim has not consented to the torture and does not enjoy any genuine control over its duration, and although the torturer in his determination to thwart the victim’s performance of a very serious criminal action is firmly

194

Why Torture is Wrong

disposed to impair the victim’s physical and psychological well-being if necessary, the torture does not betoken the overweening dominance that undoes the moral legitimacy of every instance of interrogational torture. It does not reduce the victim to being an instrument or a marionette. Instead, it is aimed simply at preventing her from engaging in some act-type that is very seriously wrong. In circumstances where the torture is the least invasive feasible means of keeping her from engaging in that act-type, it can be morally permissible. Any instance of calamity-averting interrogational torture is quite different. It does reduce the victim to an instrument or a marionette at least for a certain period, as it seeks to force (rather than prevent) the performance of some act-type. Through the use of searing pain, it aims to steer the victim away from the vast majority of the bodily movements and positionings that are available to her. It pushes her toward the relatively small array of bodily movements and positionings that consist in her performance of the act-type of revealing the information which her interrogators are trying to ascertain. While so doing, it also partakes of all the wrong-conducive properties that have been attributed in the preceding paragraph to ephemerally incapacitative torture. That is, it brings about the peculiar self-betrayal highlighted by Sussman, and it both expresses and implements the ascendance of the torturer over the victim through the exploitation of the victim’s vulnerability to severe pain; the infliction of it is non-consensual, and the victim does not possess any genuine control over the length of the infliction; and the torturer is hostile or indifferent toward the victim’s physical and mental welfare. In combination with the minutely controlling character of the measures plied against the victim, the properties just recounted are not only wrong-conducive but also wrong-making. In combination, those features of calamity-averting interrogational torture render it morally impermissible even though the purpose for which it has been undertaken is morally commendable. Notwithstanding that such torture in extreme circumstances can be morally optimal, it is never morally legitimate. 3.2.3.3.3.  Four caveats concerning my response to the first query In sum, the moral difference between ephemerally incapacitative torture and calamity-averting interrogational torture—that is, the reason why torture of the former kind can sometimes be morally permissible whereas torture of the latter kind is always and everywhere morally impermissible—resides in the act/omission distinction. Because ephemerally incapacitative torture is aimed at preventing certain actions whereas interrogational torture is aimed at preventing certain omissions, a minimally invasive instance of the latter is much more heavy-handedly directive than a minimally invasive instance of the former. Four caveats or clarifications should be appended at this juncture. First, this point about the moral difference between the two types of torture is focused on their characteristic objectives rather than on their effects. Someone who resorts to ephemerally incapacitative torture is seeking to avert an action, while someone who resorts to interrogational torture is seeking to avert an omission. Some methods of ephemerally incapacitative torture might be as far-reaching in their injurious effects as some methods of calamity-averting interrogational

Why Torture is Wrong

195

torture, but the aim of any of those former methods is to stymie the performance of a wrongful action, whereas the aim of any of the latter methods is to impel the performance of a calamity-averting act of disclosure. Given the way in which acts and omissions are distinguished, the aim of thwarting a certain action is much less minutely directive than is the aim of forcing the performance of a certain action. That former aim is much less fine-grained in its controllingness. For that reason, some possible instances of ephemerally incapacitative torture are morally permissible even if their injurious effects might match or exceed the injurious effects of some possible instances of interrogational torture. When we ask whether all possible instances of some kind of torture are overweeningly dominative, the answer lies in the general purpose for which the torture is conducted rather than in the extent of the injurious effects that might be produced by a given instance. Second is a related point. My discussion above has compared minimally invasive instances of ephemerally incapacitative torture with minimally invasive instances of calamity-averting interrogational torture. Such a comparison is uniquely germane because no instance of ephemerally incapacitative torture can ever be morally legitimate if it is not minimally invasive; in other words, no such instance can ever be morally legitimate if it is more invasive than some other feasible means of achieving the same end. My discussion above has revealed that, although every instance of calamity-averting interrogational torture is morally impermissible even when the use of such torture is no more invasive than any feasible alternative, some minimally invasive instances of ephemerally incapacitative torture are morally permissible. We do not need to go ahead with any further comparisons, since we can know from the outset that every instance of torture of either kind is morally illegitimate if it is not the least invasive approach available in the circumstances. Third, Jonathan Bennett, on whose work I  have drawn in differentiating between acts and omissions, has contended that the act/omission division is without any inherent moral significance (Kramer 2003, 342–3). Bennett acknowledges that that division is strongly correlated with some other dichotomies that are of inherent moral significance, but he maintains that the act/omission distinction itself is morally neutral. Given as much, and given that I  have concurred with his way of drawing that distinction, my invocation of it to differentiate morally between ephemerally incapacitative torture and calamity-averting interrogational torture might seem problematic. In fact, however, the difficulty just broached is illusive. There is no need for me here to contest Bennett’s assertions about the moral neutrality of the act/omission dichotomy, since his focus is markedly different from mine. Bennett is assessing the conduct of somebody who seeks to bring about a specified result either through an act or through an omission; Bennett is contending that the act/omission duality is morally neutral in the context of any such assessment. My own assessments have been very differently oriented. I have not principally been judging the moral status of the conduct of a terrorist who is subjected to calamity-averting interrogational torture. Nor have I principally been judging the moral status of the conduct of a criminal who is subjected to ephemerally incapacitative torture. Instead, I have chiefly been judging the moral status of the measures undertaken by the police or soldiers or intelligence officials who

196

Why Torture is Wrong

respond to the conduct of a terrorist or a criminal. Thus, even if Bennett is correct in suggesting that the act/omission dichotomy does not per se affect the moral bearings of the terrorist’s conduct or the criminal’s conduct, my invocation of that dichotomy to differentiate morally between a torturous response to the terrorist’s conduct and a torturous response to the criminal’s conduct is apposite. Because the terrorist’s conduct is an omission whereas the criminal’s conduct is an action—and thus because a torturous response to the former is aimed at inducing an action whereas a torturous response to the latter is aimed at inducing an omission—a minimally invasive use of torture against the terrorist is far more minutely controlling in its objectives than is a minimally invasive use of torture against the criminal. That difference grounds the division between a type of torture that is absolutely wrong and a type of torture that can sometimes be morally permissible. Fourth, my way of distinguishing morally between ephemerally incapacitative torture and calamity-averting interrogational torture is robustly perpetratorfocused. What underlies the absolute impermissibility of the latter type of torture is hardly the fact that every mass-murdering terrorist has an interest in not being treated as a marionette or a mere instrument through the exploitation of his susceptibility to agony. As has been remarked several times, the interests of any such evildoer are of no positive ethical weight. Instead, what carries the justificatory burden in this context is the moral integrity of any perpetrators of the calamity-averting interrogational torture. If they do indeed avail themselves of such torture, they will have sullied their moral integrity by using another human being’s vulnerability to severe pain as a means of minutely directing his conduct. When they arrogate to themselves such fine-grained direction through the subjection of another person (indeed, another sentient creature) to agony, they deal with that other person from a position of godlike dominion. They gain and exert that dominion for a highly worthy and pressingly urgent purpose, but their recourse to it renders their actions morally illegitimate even in circumstances where those actions are morally optimal. When they assume a position of quasi-divine ascendance, they are departing from the elementary modesty that is required of them as agents interacting with another sentient creature. In this context, the justificatory basis for the requirement of elementary modesty resides not in the interests of the victim of the perpetrators’ behaviour but instead in the integrity of the perpetrators themselves. Their integrity depends on their not using the misery of someone else to elevate themselves to a level that transcends their own humanity. (Note that the point made in this paragraph is applicable whether or not the use of interrogational torture succeeds in eliciting the information that is being sought. What is of key importance morally is not the actual attainment of godlike dominance, but the aim of attaining such dominance. By bringing to bear torturous techniques on somebody who is undergoing interrogation, the torturers seek to control her conduct minutely. Whether or not they achieve and exert the fine-grained control which they are hoping to achieve and exert through the imposition of unbearable pain, their very quest for such control through such pain renders their conduct wrongful as a mode of overweening self-aggrandizement.)

Why Torture is Wrong

197

3.2.3.3.4. A second query: why is deliberate killing ever morally permissible? In §§ 3.2.3.3.2–3, with my discussion of ephemerally incapacitative torture versus calamity-averting interrogational torture, I have submitted that the latter type of torture differs from the former in being absolutely wrong and that it is absolutely wrong because its aim is so much more minutely controlling than the aim of the former. Having maintained as much, however, I have also earlier maintained that some deliberate killings are morally permissible. In addition, some highly restrictive modes of confinement (involving straitjackets, for example) can be morally permissible in circumstances where such measures are necessary to prevent very dangerous people from inflicting grievous injuries on themselves or others. With a focus in the present subsection chiefly on deliberate killings, then, let us address the following question: Legitimacy of Killing Inquiry. If deliberate killings (and close confinement) can be morally legitimate in certain situations, why is the fine-grained directiveness of interrogational torture morally wrong in quandaries where the use of such torture is necessary for the prevention of calamities?

Subsequently, in § 3.2.3.3.6, I shall address squarely the matter of highly restrictive confinement. Just before this discussion tackles the Legitimacy of Killing Inquiry directly, we should briefly recall afresh that the absolute moral wrongness of interrogational torture is consistent with the moral optimality of such torture in some circumstances of desperate emergencies. Morally illegitimate though such torture always and everywhere is, the perpetration of it can be less gravely wrong than the eschewal of it in some imaginable predicaments. Accordingly, although calamity-averting interrogational torture differs from deliberate killings and highly restrictive confinement in that it is always and everywhere morally impermissible, it shares with them the property of being sometimes morally optimal. Let us now return to the Legitimacy of Killing Inquiry. Two replies are pertinent here. First, whereas killings and close confinement involve severe pain only incidentally, the intentional eliciting of such pain is central to calamity-averting interrogational torture and most other types of torture. My account of the wrongness of torture focuses above all on the fact that torturers exploit the susceptibility of individuals to feelings of agony. In so doing, a torturer designedly takes advantage of a fundamental and precious property of human beings by using it to afflict them with excruciatingly disagreeable sensations.19 Any fine-grained control that a torturer gains over somebody else’s conduct is acquired by him through his arousal of such sensations. By contrast, if any morally legitimate killings or instances of highly restrictive confinement ever give rise to excruciating pain, their doing so is extrinsic 19   I characterize the susceptibility to severe pain as precious because it plays a vital role in keeping people from incurring serious injuries and—even more—in keeping people from exacerbating serious injuries that have already been incurred. See, for example, Wisnewski 2010, 52–3. See also Scarry 1985, 333 n62.

198

Why Torture is Wrong

to the effects which they are supposed to produce. Although the intended effects of such measures are far-reaching—more far-reaching than the effects of some instances of torture—the production of those effects does not proceed through any deliberate evocation of severe pain. My first reply to the Legitimacy of Killing Inquiry has highlighted a major way in which the killing or close confinement of someone differs from the torturing of someone for any purpose. Second, and even more important, is a reply that underscores the divergences between ephemerally incapacitative torture and other types of torture; the killing or restrictive confinement of someone is assimilable to the former type of torture rather than to any of the latter types. If the deliberate killing of some captors or assailants is necessary to rescue oneself or anyone else, or if the deliberate killing of enemy troops in combat is necessary to eliminate the threat which they pose to oneself or to one’s fellow soldiers, the acts of killing can be morally legitimate. In such circumstances, however, those acts of killing are closely akin to acts of ephemerally incapacitative torture. Of course, in one obvious respect (in addition to the respect noted in the preceding paragraph) the former acts are very different from the latter, for their intended effects are permanent rather than transitory. Nonetheless, in the respect that is decisive here, the two types of acts are assimilable. Whenever ephemerally incapacitative torture is morally legitimate, it is undertaken as the least invasive means of preventing the occurrence or completion of a very serious criminal act. Similarly, whenever a deliberate killing is morally legitimate (in the circumstances envisaged above), it is undertaken as the sole feasible means of preventing the occurrence or completion of a very serious criminal act—or, in the context of military combat, as the sole feasible means of eliminating a mortal threat posed to oneself or to one’s fellow soldiers by enemy troops who have declined to surrender. In other words, the aim of any morally legitimate instance of deliberate killing is relevantly similar to the aim of any morally legitimate instance of ephemerally incapacitative torture. That is, the objective is to thwart some injurious action(s) rather than to impel some beneficial action(s). As a consequence, the objective of a morally legitimate instance of deliberate killing is far less minutely controlling than the objective of any instance of calamity-averting interrogational torture. From a perpetrator-focused perspective, therefore, we can discern that some instances of deliberate killing do not partake of the hubris that is characteristic of every instance of calamity-averting interrogational torture. For that reason, in combination with the reason expounded in the preceding paragraph, the absolute wrongness of calamity-averting interrogational torture is not paralleled by the moral status of deliberate killing. Some possible tokens of the act-type of deliberate killing are indeed morally permissible. In the last paragraph, I have concentrated on killings that are undertaken in defence of oneself or others. Nothing is said there about executions. Yet, as has been remarked in § 3.2.3.2, I have elsewhere argued at length that executions of defilingly evil offenders can be morally legitimate in liberal democracies. How can any executions be morally permissible while all instances of calamity-averting interrogational torture are morally impermissible? The answer to this question broadly resembles my answer to the more general Legitimacy of Killing Inquiry.

Why Torture is Wrong

199

When someone is executed on the basis of the purgative rationale for capital punishment, the aim of such a sanction obviously does not consist in minutely controlling him by prodding him to perform some action. Corpses do not undertake actions. Rather, the aim is to terminate a life that would constitute a standing affront to humanity if it were to continue (Kramer 2011, chap. 6). Of course, I cannot here present the long and complex arguments that support the purgative rationale for the death penalty. Still, there is not any need for me here to re-elaborate those arguments, since the dispositive point at present is that no purgative execution is designed to achieve any fine-grained control over someone’s behaviour. Nor does any such execution, conducted in the most humane manner that is technologically feasible, deliberately exploit a person’s vulnerability to severe pain. Hence, although the effects of a purgative execution are more wide-ranging than the effects of most techniques of torture, the minutely controlling and agony-exploiting features of interrogational torture (and most other types of torture) are not present in such an execution. Accordingly, readers who disagree with me about the moral permissibility of executions of the most egregiously iniquitous criminals will have to adduce alternative grounds for condemning those executions as morally wrong. No such condemnation can warrantedly advert to the key wrong-making properties of calamity-averting interrogational torture, since purgative executions do not partake of those properties. In other words, even readers who disagree with me about the morality of capital punishment can recognize that my affirmation of the moral permissibility of purgative executions (in appropriate circumstances) is perfectly consistent with the thesis that all instances of calamity-averting interrogational torture are morally wrong. 3.2.3.3.5.  Clarifying the issue Before we move on to the matter of highly restrictive confinement, this chapter should clarify the issue addressed by my contrast between deliberate killings and calamity-averting interrogational torture. Let us note the following two propositions: Proposition 1. Some deliberate killings are morally worse than many wrongful acts of torture. Proposition 2. The victims of any deliberate killings (apart from certain acts of euthanasia) are harmed more than the victims of some acts of calamity-averting interrogational torture. Those two propositions are fully consistent with the following two further propositions: Proposition 3. Calamity-averting interrogational torture is always and everywhere morally wrong. Proposition 4. Some deliberate killings are morally legitimate. These two further propositions are likewise consistent with each other, notwithstanding that the victims of any morally legitimate killings—apart from certain

200

Why Torture is Wrong

acts of euthanasia—are harmed more than are the victims of many instances of calamity-averting interrogational torture. Some philosophers, writing from a firmly victim-focused perspective, have supposed that Proposition 3 and Proposition 4 are inconsistent with each other and have concluded that Proposition 3 is false. Helen Frowe, for example, invokes Proposition 2 in support of just such a claim about inconsistency and just such a conclusion about Proposition 3: It is certainly proportionate (and permissible) to kill in self-defence a person who will otherwise kill me. Unless torture is a significantly worse harm than death, it looks as though torture should also be a proportionate form of defence against a lethal harm . . . [E]‌ven if inserting needles under someone’s fingernails is very painful, it’s hard to imagine that anyone would rather die than suffer, say, an hour of that sort of pain. People undergo painful operations precisely because suffering some level of pain (especially temporary pain) is better than dying. Of course, as we discussed above, the wrongness of torture might not be limited to the physical pain inflicted, but might also include the wrong of violating autonomy. But even if by causing you an hour’s worth of pain, I (might) get you to reveal your plans to me and thereby act against your own autonomous preferences, it still seems a stretch to say that what I did was worse than if I had killed you. After all, I may have subdued your will to my ends for an hour, but you still have the rest of your life to go about forming and pursuing autonomous preferences. There would be something odd about a view that urged respect for autonomy, but held temporary violations of autonomy to be worse than its permanent elimination.20

Frowe is correct in endorsing Propositions 2 and 4, but she errs in thinking that those propositions are inconsistent with Proposition 3. Her error stems principally from her implicit assumption that the wrongness of calamity-averting interrogational torture must be attributable to the injurious effects of such torture on its victims. She has reasoned as follows: I. The deliberate killing of an assailant in order to stave off his lethal attack can be morally permissible. II. The harm suffered by an assailant who has been killed in self-defence is typically greater than the harm suffered by a terrorist who has been subjected to calamity-averting interrogational torture. III. The calamity which such a terrorist has sought to bring about is typically even more heinous than the murder which the assailant has sought to commit. IV. Ergo, some instances of calamity-averting interrogational torture are morally permissible.

20   Frowe 2011, 204–5, emphasis in original. For some views broadly similar to Frowe’s, see Cohan 2007, 1602; Dershowitz 2002, 148–9; Himma 2007, 236, 241; Kamm 2011, 3–4, 15–26; Kershnar 2005, 231; Posner and Vermeule 2006; Posner and Vermeule 2007, chap.  6; Seidman 2005, 897; Steinhoff 2006; Steinhoff 2009.

Why Torture is Wrong

201

Although the premises in her line of reasoning are true, the conclusion therein does not follow from those premises. As should be evident by now, her conclusion is a non sequitur because it ignores the perpetrator-focused considerations that nullify the permissibility of calamity-averting interrogational torture even in cases where no victim-focused considerations are sufficient to disallow such torture. That is, her line of reasoning disregards the ways in which all instances of interrogational torture (including morally optimal instances that help to avert the occurrence of atrocities) morally degrade the people who prescribe and administer the torture. Those people morally degrade themselves because they aim with their torturous measures to achieve fine-grained control over the conduct of others through the subjection of the others to agony. Striving for such control through the infliction of harrowing pain, the torturers sully their moral integrity by endeavouring to elevate themselves to a position of minutely directive dominance—godlike dominance—that exceeds what can legitimately be sought by human beings in their interaction with one another and with sentient beings more generally. Although the measures undertaken by the torturers might be morally optimal in some dire emergencies, and although the outlooks of the torturers might be desperately well-intentioned, their outlooks are besmirched by the hubristic overreaching to which they have to resort in fulfilment of their good intentions. Their worthy ends might weakly justify, but never strongly justify, the torturous means through which those ends are pursued. In short, the question addressed here is not whether some deliberate killings are morally worse than most instances of torture, nor whether the victims of deliberate killings are harmed more than are the victims of many instances of torture. Albeit the answer to each of those questions is affirmative, the point at issue in the present discussion is not a matter of comparative gravity or comparative harmfulness. Accordingly, the correctness of an affirmative answer to each of the questions just broached has no bearing on the answer to the question that is my central concern here: namely, whether the act-type of deliberate killing differs from the act-type of calamity-averting interrogational torture in that some tokens of the former but no tokens of the latter are morally permissible. When we supplement our victim-focused perspective with a perpetrator-focused perspective—that is, when we pay attention to the moral integrity of torturers even in predicaments where their victims’ interests are of no positive ethical weight—we can see that the answer to the relevant question here is affirmative. Unlike deliberate killing, calamity-averting interrogational torture is wrong always and everywhere in all possible worlds. 3.2.3.3.6. A third query: why is highly restrictive confinement ever morally permissible? Let us return to the matter of highly restrictive confinement. Some instances of such confinement are morally legitimate when they are necessary to keep extremely dangerous people from harming themselves or others. If the need to avert the commission of some very serious crimes is what can render permissible the use of

202

Why Torture is Wrong

such confinement, then why is interrogational torture not also permissible when law-enforcement officials need to avail themselves of it in order to avert calamitous crimes? The framing of this question is tendentious, for it conveys the impression that calamity-averting interrogational torture is akin to ephemerally incapacitative torture in its workings. Contrary to any such impression, the former type of torture is markedly different from the latter in that the direct effect of any successful instance of it is the prevention of an omission (specifically the prevention of the withholding of information) rather than the prevention of an action. Of course, the information extracted through the use of interrogational torture can then be employed to prevent a very serious criminal action. However, the direct effect of the torture resides in the prevention of an omission through the forced inducement of an act of disclosure. Consequently, the administration of such torture involves the minute control of someone else’s behaviour through the exploitation of her susceptibility to agony. As has already been contended, the aim of exerting such control by such a means is hubristically corrosive of the moral integrity of everyone responsible. Highly restrictive confinement is significantly different. It is not imposed with the aim of evoking severe pain either as a means or as an end, nor is it designed to prevent an omission and induce an action. On the contrary, it is designed to be incapacitative. It prevents some very seriously wrongful action(s) that would have been carried out by the confined person if he had not been restrained. Seeking to induce omissions rather than actions, the officials who impose the close confinement are not engaged in the fine-grained direction of someone else’s behaviour. In that respect, as well as in the absence of any intention on their part to inflict severe pain, their recourse to the imposition of close confinement is very different from any recourse to interrogational torture. To be sure, although the dominion (over someone else’s behaviour) that ensues from highly restrictive confinement is not fine-grained, it is far-reaching. Such confinement, in addition to precluding the performance of the very seriously wrongful action(s) which it is designed to preclude, rules out the performance of many other actions that would have been possible in its absence. Still, much the same is true of some morally legitimate instances of ephemerally incapacitative torture. Although the aim of such torture is to thwart the performance of some very seriously wrongful action(s), its effects are often considerably more far-reaching; for a certain period, the pain that the torture causes can be sufficient to stymie the performance of many other actions as well. In regard to restrictive confinement, just as much as in regard to these morally legitimate instances of ephemerally incapacitative torture, the distinction between the aim and the effects is crucial. While the objective of each such measure is to preclude the quite small proportion of bodily movements and positionings that amount to some very seriously wrongful action(s), an incidental effect of any given measure—which, though incidental, is often amply foreseeable—will typically be the preclusion of many other bodily movements and positionings as well. The moral status of restrictive confinement is differentiated from the moral status of

Why Torture is Wrong

203

calamity-averting interrogational torture by the modesty of the confinement’s aims despite the sweepingness of its effects. Of course, I am hardly suggesting that the probable effects of close confinement or of ephemerally incapacitative torture are morally irrelevant. Those effects are of great importance, for, under the sway of the Minimal Invasion Principle, no instance of restrictive confinement or of ephemerally incapacitative torture is ever morally permissible unless it is the least invasive means of preventing the very serious crime(s) which it is undertaken to prevent. Furthermore, any such instance is morally impermissible if its own harmfulness significantly exceeds the harmfulness of the crime(s) which it foils. Still, when those conditions of minimal invasiveness and proportionality are satisfied, the stymieing of very seriously wrongful actions through the use of tight confinement or ephemerally incapacitative torture is morally permissible. Because the objective of such confinement or torture is to preclude certain actions rather than to impel certain actions, the officials who desperately resort to the confinement or torture—in circumstances where the conditions of minimal invasiveness and proportionality are satisfied—will not have sullied their moral integrity through a quest for superhuman ascendance. 3.2.3.3.7.  A fourth query: why is punitive torture absolutely wrong? My way of distinguishing morally between ephemerally incapacitative torture and calamity-averting interrogational torture might lead some readers to wonder why punitive torture is always and everywhere wrong. Though punitive torture is carried out on some occasions as a form of interrogational or act-impelling torture, it is on other occasions purely punitive. On those latter occasions, then, the perpetrators of punitive torture do not aim to impel the performance of any actions; while eliciting agony in their victims, they are not thereby seeking to exert any minute control over the victims’ behaviour. In that respect, the character of such torture on those occasions is markedly different from that of interrogational torture and is assimilable to the character of ephemerally incapacitative torture. Given that the last-mentioned type of torture is sometimes morally permissible, why is purely punitive torture always and everywhere wrong? This query overlooks a key way in which ephemerally incapacitative torture is similar to calamity-averting interrogational torture and different from punitive torture. On any of the rare occasions when ephemerally incapacitative torture is a morally permissible tack, an official or a private citizen is confronted with an emergency in which a miscreant is seeking to perform a very seriously wrongful action. On an occasion of that kind, the use of such torture is the least harmful way of keeping the miscreant from performing the action in question, and the infliction of searing pain through the administration of the torture is not significantly more injurious than the effects which the thwarted crime would probably have caused if it had occurred. In other words, the moral permissibility of ephemerally incapacitative torture on such an occasion does not derive solely from the absence of any effort on the part of the torturer to gain fine-grained control over the victim’s conduct. In addition, it derives from the absence of any superior alternatives in a pressingly grim emergency. By contrast, typically, the exigencies of

204

Why Torture is Wrong

an emergency are not operative in a criminal-justice system where punishments are imposed on malefactors. Even if the officials in a given system adhere to a consequentialist rationale for their levying of punishments on criminals—such as the deterrence-oriented rationale which I have critically assessed elsewhere in the context of capital punishment (2011, chap.  2)—their endeavours to lower the future incidence of crimes will very rarely if ever take place in circumstances of grave urgency that sharply limit their options. To be sure, there can in extremis arise social upheavals that render the imposition of punishments a matter of grim urgency. In Herman Melville’s famous novella Billy Budd, for example, Captain Vere reluctantly imposes the death penalty on the eponymous seaman because he believes that his declining to do so would invite mutinous conduct among the sailors on his own ship and in other parts of the British fleet. Let us suppose that Vere’s perception of the danger of insubordination is accurate. In a quandary of that kind, an official has levied a punitive measure in response to an outright emergency. However, even if we leave aside the rarity of any relevantly similar predicaments in the workings of an ordinary criminal-justice system, the prospect of a mutiny on Vere’s ship would not render morally permissible the wielding of torture against Billy Budd. A torturous sanction imposed as a deterrent, which would be an instance of both punitive torture and broadly addressed intimidatory torture, would differ starkly from an instance of ephemerally incapacitative torture. Were Vere to inflict severe pain on Billy as a punishment, he would scarcely be seeking to prevent the young man from performing a very seriously wrongful action. Rather, he would be using Billy’s susceptibility to agony as a means to discourage the spread of refractoriness among the crew. Any such instrumentalization of somebody’s proneness to terrible pain—that is, any measure that uses somebody and his proneness to terrible pain purely as a means for influencing the conduct of others—would be morally illegitimate, regardless of the exigencies of the emergency to which the instrumentalization would be a response. Though those exigencies might conceivably render the employment of torture morally obligatory on consequentialist grounds, the moral obligatoriness would be unaccompanied by moral permissibility. No demands of any emergency, however dire, could bestow moral legitimacy on the exploitation of someone’s vulnerability to agony as a mere means for channelling the behaviour of other people. (For cognate reasons, the execution of Billy by Captain Vere is morally wrong even if Vere is under a moral obligation to impose such a punishment. Though the execution does not trade on the susceptibility of the young man to sensations of exquisite pain, it uses him and his life in an emergency purely as a means for the furtherance of discipline on Vere’s ship and in the British fleet.) 3.2.3.3.8.  Retributivism and the perpetrator-focused perspective In sum, if punitive torture were ever undertaken in an emergency to realize some consequentialist goal such as the deterrence of unrest, it would be a morally illegitimate use of the victim purely as a means for the advancement of others’ ends. Can punitive torture ever be morally permissible if it is instead conducted on

Why Torture is Wrong

205

retributivistic grounds? Insofar as the rationale for torturous punishments is retributivistic, the analogy between any such punishment and a morally permissible instance of ephemerally incapacitative torture is plainly unsustainable. Whereas a permissible instance of ephemerally incapacitative torture is aimed at preventing the performance of a very serious criminal action that will occur if the torture is not wielded, a punishment levied for retributivistic purposes is aimed at rectifying the commission of a crime that has already occurred. Whereas ephemerally incapacitative torture is forward-looking, retributivistic punishments are backward-looking. Hence, although the pressures of an emergency can render the employment of ephemerally incapacitative torture morally legitimate in certain circumstances, those pressures are not applicable to the imposition of punishments for retributivistic purposes. Such punishments are not ways of dealing with emergencies or of foiling very seriously wrongful actions. In short, although retributivistic torture is similar to ephemerally incapacitative torture in that it is not aimed at exerting fine-grained control over anyone’s behaviour, it is morally inassimilable to the latter type of torture in that its exploitation of the susceptibility of people to agony is not aimed at preventing those people from performing very serious criminal actions. Still, even though punitive torture for retributivistic ends is never morally permissible on the basis of any analogy to ephemerally incapacitative torture, some readers might think that it can sometimes be morally permissible independently of such an analogy. Let us ponder a situation in which the use of punitive torture on retributivistic grounds might seem peculiarly apposite. Suppose that Theodore has committed a number of grisly murders during which he has tortured his victims mercilessly, and suppose that he is eventually apprehended and tried and convicted in a system of criminal justice that is oriented toward the principles of retributivism. Would the imposition of punitive torture—either on its own or in combination with some other sanction—be morally legitimate? Such a punishment would be consistent with the main tenets of retributivism. Specifically, it would accord with the principle of commensurateness (quantitative equivalence) between the severity of a punishment and the seriousness of a crime, and it would likewise conform to the lex talionis principle of a qualitative homology between the wrong-making features of a crime and the equality-restoring features of a punishment. Would such a sanction also be consistent with the broader demands of morality? Worth noting straightaway is that the consistency (in certain cases) between the use of punitive torture and the chief tenets of the doctrine of retributivism has quite often been invoked by opponents of that doctrine in order to discredit it. What has particularly drawn the ire of those opponents is the principle of lex talionis (which is often misrepresented as a principle of revenge). Supporters of that principle, such as Jeffrey Reiman, have correspondingly felt a need to distance it from the notion that torture is ever a legitimate sanction: ‘[S]‌trict adherence to the lex talionis amounts to allowing criminals, even the most barbaric of them, to dictate our punishing behavior . . . [T]here are at least some crimes, such as rape or torture, that we ought not to try to match . . . [E]ven though it would be just to

206

Why Torture is Wrong

rape rapists and torture torturers, other moral considerations weigh against doing so’ (Reiman 1998, 94–5). Now, on the one hand, my arguments elsewhere concerning the principle of lex talionis—in the context of debates over the death penalty (Kramer 2011, 128–35)—strongly suggest that that principle does not ever require the imposition of punitive torture. Hence, regardless of what Reiman has in mind when he writes of ‘strict adherence’, justice never requires the levying of a torturous sanction. On the other hand, as I have stated above, the lex talionis is indeed consistent with the imposition of punitive torture on villains who have themselves made use of torture in their crimes. Given as much, Reiman is correct in thinking that the lex talionis principle does not provide reliable guidance on what can legitimately be done to punish those villains. When we contemplate the example of Theodore sketched in the penultimate paragraph above, a victim-focused perspective will not enable us to perceive why the wielding of punitive torture against him is morally impermissible. Because of the numerousness and vileness of his terrible crimes, his interests in being free from excruciating pain are of no positive ethical weight. He has forfeited his right against being imprisoned and also his right against being executed. If he has not similarly forfeited his right against being tortured, the reason for his retention of that right is not discernible from a purely victim-focused point of view. When we shift to a perpetrator-focused perspective, we concentrate not on the interests of a potential victim of punitive torture—which, in the case of Theodore, are justificatorily weightless—but on the moral integrity of the officials in a system of criminal justice who would prescribe and administer the torture. We also concentrate on the moral integrity of the society in which the torture would be administered as a sanction. From such a perspective, we know that calamity-averting interrogational torture is always and everywhere wrong partly because it involves a quest for minute control over a victim’s behaviour through the exploitation of her susceptibility to intense pain; a self-aggrandizing quest of that sort, through such an agony-inducing route, sullies the moral integrity of anyone who directs or undertakes it. Punitive torture, like ephemerally incapacitative torture, does not aim at any comparably fine-grained control. Accordingly, the moral wrongness of such an aim is not a factor in determining the moral status of punitive torture. Should we conclude that punitive torture, like ephemerally incapacitative torture, can be morally permissible in certain situations? Here we return to the gist of the opening paragraph of this subsection. Precisely because any punishment inflicted in conformity to the tenets of retributivism is aimed at rectifying some past wrong(s) rather than at forestalling the occurrence of some impending wrong(s), the use of torture in conformity to those tenets is not rendered morally permissible as the least invasive way of thwarting a calamitously wrongful action. Someone responsible for prescribing or administering punishments on retributivistic grounds is not thereby coming to grips with an emergency in which he faces no other options that could achieve what the torture can achieve. On the contrary, anyone who performs such a role can select among a number of options; some of the sanctions that can serve as retributivistic punishments are fines and imprisonment and executions. When dealing with a depraved

Why Torture is Wrong

207

mass-murderer and torturer like Theodore, the officials in a retributivistic system of criminal justice have to impose some especially harsh sanction(s). The death penalty, humanely implemented, would tally with the principles of retributivism—and would be independently justifiable on the basis of the purgative rationale for capital punishment which I have expounded elsewhere—but also consistent with the tenets of retributivism would be lifelong imprisonment-without-parole in solitary confinement with few amenities. (Since the death penalty would be more invasive than the lifelong imprisonment, the Minimal Invasion Principle disallows the former sanction within a purely retributivistic system of criminal justice.) Proponents of retributivistic torture will undoubtedly grant that it is not the only sanction that can suitably be imposed on someone like Theodore, but they will insist that it is one further option available to the officials who have to prescribe and administer his punishment. Moreover, since certain techniques of torture are less injurious than the death penalty or lifelong imprisonment-without-parole in austere conditions, the Minimal Invasion Principle in a retributivistic system of criminal justice forbids the use of either of those non-torturous punishments. So, at any rate, the proponents of retributivistic torture will be apt to contend. One relatively minor weakness in the position of the proponents of retributivistic torture is that any techniques of torture that are significantly milder than lifelong imprisonment-without-parole will be too lenient to serve satisfactorily as a retributivistic punishment in application to a monstrous evildoer like Theodore. For example, if Theodore were sentenced to undergo ten blows with a cane or the twisting of his arms behind his back for half an hour or the drilling of the unanaesthetized roots of some teeth for half an hour, the enormity of his murders and rapes and grisly acts of torture would not be reflected in the punishment meted out to him. In application to him, such sanctions would not adequately give effect to the principles and requirements of retributivism as a doctrine of deserved punishments. Of course, the problem just broached could be overcome through a substantial intensification of the harshness and prolongedness and frequency of the torture to be inflicted on anyone of Theodore’s turpitude. However, that very ‘solution’ to the first problem brings us to a far more formidable weakness in the position of the proponents of retributivistic torture. Their position fundamentally misconstrues the character of retributivism. Contrary to what numerous opponents and some supporters of the principles of retributivism have contended, those principles do not constitute a doctrine of revenge. Retributivism does not call for the wreaking of vengeance or the satisfaction of vindictive impulses. If it were a doctrine of revenge along those lines, it would fail to provide a morally worthy rationale for punishment. Its moral credibility as such a rationale ensues from its character as a doctrine of impersonal justice. It prescribes the imposition of punishments to vindicate the dignity of victims of crimes and to uphold the moral order of a community and to offset the self-indulgence to which malefactors have stooped in their criminality (Kramer 2011, chap. 3). Pursuing those morally worthy objectives in the context of a liberal-democratic society, a retributivistic system of criminal justice can invest its punishments with moral legitimacy. Retributivism, in other

208

Why Torture is Wrong

words, can serve as one of the underpinnings of a morally upright system of governance. Were it a doctrine of revenge, by contrast, it could never serve as such an underpinning. Now, let us recall that some punishments other than torture can fulfil the requirements of retributivism in application to a heinous criminal such as Theodore. Given as much, a decision by judges to sentence Theodore to a regimen of brutal torture in a retributivistic system of criminal justice would partake of vindictiveness. Regardless of what the conscious motivations of the judges might be, the imposition of such a sentence would not be plausibly construable except as an act of revenge. Subjecting Theodore to harsh torture, in contrast with sentencing him to lifelong imprisonment-without-parole in solitary confinement or to a humanely conducted execution, would squarely cater to impulses of vindictiveness harboured by members of the general public and especially by people who have suffered directly from his iniquities. There is no other plausible reason for choosing such a sanction in preference to one of the available alternatives (or in combination with one of those alternatives). Perfectly understandable though the impulses of vindictiveness are, the gratification of them is not among the purposes that can legitimately be pursued by any system of criminal justice. It is not among the purposes prescribed by any morally tenable version of retributivism.21 Even if mirabile dictu a judge were not consciously endeavouring to gratify any feelings of vengeance when imposing a regimen of torture on Theodore, she would be culpably remiss in failing to recognize that such a sanction so directly marshals and reaffirms those feelings. It marshals and reaffirms sadistically vindictive urges, and is morally impermissible as a consequence. What should be emphasized again is that the moral illegitimacy of the use of torture as a punishment to be imposed on Theodore is not due to the justificatory weight of his interest in being free from excruciating pain. As has already been stated, the justificatory weight of his interests is nil. Having led a flagitious life whose ethical bearings are overpoweringly determined by the atrocities which he has gleefully committed, Theodore deserves no better than the agony that would be inflicted on him through a regimen of harsh torture. Only from a perpetrator-focused perspective can we see why the imposition of such a regimen as a sanction would be morally impermissible. Undeserving though Theodore is of any restrainedness in his punishment, the use of torture as a sanction would sully the moral integrity of a criminal-justice system and its officials. If they avail themselves of such a sanction, they will have turned their punitive endeavours into a vehicle for the pursuit of vengeance (even in the exceedingly unlikely event that their choice of a torturous sanction has not been consciously motivated by such a pursuit or by the desire to pander to such a pursuit). Officials who so grossly cater to sadistically vindictive 21   A  quite distinct theory of punishment, the denunciatory theory, does attribute to sanctions the role of satisfying ordinary people’s urges for revenge against criminals. See my critical assessment of the denunciatory theory in Kramer 2011, chap. 5. Because the denunciatory theory is thoroughly consequentialist in its orientation, it suffers from most of the same weaknesses that afflict the deterrence-oriented conception of punishment.

Why Torture is Wrong

209

instincts are adulterating the retributivism of their system of criminal justice and are pro tanto depriving the system of its moral uprightness. Their indulgence of such ignoble instincts is morally unworthy, even in application to someone like Theodore who deserves no better. Also worth noting here is a closely related point. That is, the moral forbiddenness of the use of torture as a punishment is not attributable (or is not always attributable) to the Minimal Invasion Principle. In other words, the moral wrongness of such a sanction is not attributable—or is not always attributable—to its excessive harshness. Even if some course of torture would be less harsh for the victim than every alternative sanction that is retributivistically suitable, it would contravene the moral requirements that are incumbent on the officials in a system of criminal justice. Notwithstanding that it would be less severe than the other retributivistically suitable punishments, it would tie the officials’ system too tightly to dubiously base impulses from which the system must be distanced if the endeavours of the officials therein are to be morally legitimate. Although the victim-focused Minimal Invasion Principle does not per se disallow the plying of such a course of torture against Theodore as a punishment for his reprehensible crimes, the employment of such a sanction is excluded by a perpetrator-focused delimitation of the promptings with which a system of criminal justice can permissibly and closely associate itself. 3.2.3.3.9. A fifth query: why is consensual placatory torture impermissible? In § 2.1.4.1.2 of Chapter 2, I have recounted Alan Gewirth’s scenario of a mother who is to be tortured by her son (a prominent politician) in order to appease some terrorists who will otherwise detonate several catastrophic explosions in major cities. Let us now mull over an embellished version of that scenario. Suppose that the mother learns of the terrorists’ nefarious threats, and that she implores her son to go ahead with the torture in order to spare the lives of hundreds of thousands of people. She repeatedly adjures him that he must put the elementary safety of those people ahead of his own compunctions and her own comfort. She firmly indicates that she is prepared to undergo the agony of brutal torture in the dire circumstances that confront them. Will the politician be acting permissibly if he accedes to the entreaties of his mother by subjecting her to placatory torture? Manifestly, the terrorists are behaving impermissibly and indeed evilly in this modified scenario as much as in the original scenario. However, does my embellishment of Gewirth’s thought-experiment alter the moral status of the conduct of the politician if he now goes ahead with the placatory torture? One reason why the answer to this question is negative is that the modified scenario does not involve any genuine consent by the mother to the administration of torture. When she beseeches her son to avert the calamitous explosions by subjecting her to torture, she is pleading with him under conditions of extreme coercion (where the coercion is exerted not by him, of course, but by the terrorists). Hence, her impetrations do not amount to genuine consent. Consequently, any legitimizing force that attaches to the presence of genuine

210

Why Torture is Wrong

consent in some edifying contexts is not operative in this scenario of placatory torture. Moreover, the orientation of the placatory torture in my embellished scenario— like its orientation in Gewirth’s original scenario—is markedly different from the orientation of any morally legitimate instance of edifying torture. In any instance of the latter sort, the severe pain is inflicted to further the physical and psychological well-being of the victim. In my extension of the Gewirthian thought-experiment, by contrast, the son tortures his mother not to enhance her bodily and mental hardihood but to mollify the terrorists. Of course, in the respect which I  have highlighted in Chapter 2, the son’s orientation toward his mother is benign. That is, the son shudders with dismay at the prospect of torturing his mother, and he desperately hopes that his administration of torturous techniques to her body will not elicit any sensations of agony in her. Nonetheless, while harbouring such a hope, he naturally recognizes that the torturous techniques will almost certainly afflict her with excruciating pain. He further recognizes that the terrible pain will do nothing to enhance her physical and psychological well-being and will instead detract dreadfully therefrom. In that regard, despite the horrified yearning and anguish of the son who has to apply the torturous measures to his mother, the orientation of those measures in relation to her is decidedly inimical rather than benevolently wholesome. While the foregoing victim-focused considerations are sufficient to indicate why my embellishment of Gewirth’s scenario has not altered the moral status of the placatory torture from impermissible to permissible, they can and should be supplemented with some brief perpetrator-focused reflections. Although the politician is acting both for the good of the broader public and in response to the exhortations of his mother, and although the gravity of the wrong committed by him is greatly extenuated by those factors, he is indeed committing a wrong when he applies torturous techniques to an innocent person for any objective other than her own physical and mental edification. He is not seeking to gain fine-grained control over her behaviour, nor is he associating himself with instincts for revenge, but he is demonstrating that he is ready to do whatever the terrorists bid him to do. He is thereby using himself, as well as his mother, for their ends. He is of course doing so not for any discreditable purposes, but instead for the commendable purpose of saving hundreds of thousands of lives. He is fulfilling a stringent duty owed to the general public even while he is transgressing some stringent duties owed to himself and to his mother. Still, notwithstanding that his behaviour might be morally optimal, he is behaving wrongly by using himself and his mother as playthings of the terrorists. He morally degrades himself by complying with their behests—behests squarely aimed at psychologically humiliating him and morally debasing him—even though his self-degradation is perhaps the least terrible course of conduct available to him in conditions of extreme duress. 3.2.3.3.10.  A final query: why is sado-masochistic torture morally wrong? The factor of the mother’s consent (albeit coerced consent) in my modified Gewirthian scenario brings us back to the matter of sado-masochistic torture, on

Why Torture is Wrong

211

which I have already remarked in this chapter and the preceding chapter. Why is the infliction of sado-masochistic torture always morally wrong? After all, such torture is typically consensual, and in a salient respect it is oriented toward the physical and psychological well-being of the victim. That is, the perpetrator of the torture aims to provide the victim—as well as himself—with sexual gratification. Moreover, the perpetrator is not hubristically seeking to exert minute control over the behaviour of the victim and is not acting in a fashion that gives vent to vindictive urges. Nor is he treating himself as a plaything of his enemies; he does not engage in his peculiar pastime as a way of complying with humiliating behests from anybody else. Why, then, is his perpetration of sado-masochistic torture morally wrong? Let us note that this question is indeed about the perpetration of sado-masochistic torture. More doubtful is whether the consensual undergoing of such torture is always morally wrong. At any rate, that latter issue will be left aside here as we ponder why the act-type of wielding sado-masochistic torture against a consenting victim is always morally impermissible. Let us further note that the matter under scrutiny here is the moral status of such torture rather than its appropriate legal status. As Chapter 2 has intimated, and as Chapter 5 will maintain at greater length, sado-masochistic torture should not be legally proscribed unless the administering of the torture is non-consensual or unless it results in permanent and very serious harm. Nonetheless, although sado-masochistic ordeals should be legally unforbidden in most circumstances, they are morally wrong in all circumstances. A sado-masochist Samuel who inflicts searing pain consensually on a similarly inclined person Peter is seeking to elicit sexual pleasure in Peter but is also seeking to derive sexual pleasure for himself from the experience of causing Peter to undergo dire pain. That latter aspect of his conduct—his derivation of sadistic gratification from the knowledge that he is causing someone else to suffer grievously—is what renders his behaviour morally illegitimate. If Samuel were instead participating very reluctantly in response to numerous importunate requests from his friend Peter, and if his sole aim were to satisfy Peter’s craving for sexual arousal, then his conduct might not be morally impermissible. (Its moral permissibility would hinge on the specifics of the torture, especially on the extent of the harm that is likely to ensue from it.) Still, if Samuel were a reluctant participant with the sole objective of indulging Peter’s masochistic predilections, his infliction of severe pain would not amount to sado-masochistic torture at all. Hence, the permissibility of his conduct in those circumstances would be consistent with the proposition that the perpetration of sado-masochistic torture is always morally wrong. To apprehend why the infliction of sado-masochistic torture is wrong in all of its instantiations, we have to adopt a perpetrator-focused perspective. If the victim of the torture such as Peter has eagerly consented to the process, and if the techniques of torture are not likely to result (and do not actually result) in death or in very serious and permanent injuries, then there are no victim-focused considerations that would undermine the moral permissibility of the actions of

212

Why Torture is Wrong

the perpetrator. What does undermine the permissibility of those actions is the sadism with which they are undertaken. Samuel’s derivation of pleasure from Peter’s misery is morally debasing—it demeans both Samuel and Peter—regardless of how enthusiastically Peter has chosen to participate as a masochistic victim. To be sure, the sadism in Samuel’s outlook is coupled with his desire to stimulate sexual euphoria (as well as severe pain) in Peter. His outlook is not unalloyedly self-centred or ghoulish. Still, the delight that Samuel experiences at the prospect of another person’s anguish and at his knowledge of his own role in bringing about that anguish is morally unworthy. It is a feature of his conduct that places him in contravention of moral duties which he owes to himself and to Peter.

3.3.  Moral Optimality without Moral Permissibility This lengthy chapter will close by contemplating briefly a matter which Chapter 1 has explored in a more general form. Given that most types of torture are always and everywhere morally impermissible, can any instances of those types of torture be morally optimal despite being morally wrong? In other words, is the wrongness of those types of torture strongly absolute or only weakly absolute? Let us ruminate on this question with reference to calamity-averting interrogational torture, the kind of torture that has received far more attention than any other in the contemporary philosophical investigations of the topic. The moral impermissibility of such torture is only weakly absolute rather than strongly absolute; there are some possible instances (and even some credible instances) of calamity-averting interrogational torture that are morally optimal.22 Before we mull over several of the main considerations that bear on the matter of moral optimality, we should take account of a couple of examples.

3.3.1.  A first example In § 2.2.1.3 of Chapter 2, I have presented a slightly modified account of the facts in Leon v. Wainwright. Given the facts as I have recounted them there, the use of calamity-averting interrogational torture by the police to extract vital information from the kidnapper is morally optimal. On the one hand, by resorting to such torture, the police have acted wrongly in that they have exploited the kidnapper’s susceptibility to severe pain and have done so for the purpose of exerting fine-grained control over his conduct. Gaining and exercising such dominance over the kidnapper, the police arrogate to themselves the role of gods. On the other hand, they do so in a dire emergency where their attainment of a position of overweening ascendance vis-à-vis the kidnapper is the sole effective means of saving the life of his captive. Although they have contravened the absolute moral 22   In a thought-experiment that is wildly fanciful but not unintelligible, even sadistic torture can be morally optimal (though such torture is of course odiously wrongful even in the conditions delineated by the fanciful thought-experiment). See Kramer 2009, 158.

Moral Optimality without Moral Permissibility

213

prohibition on any recourse to interrogational torture—a prohibition whose moral force, in its applicability to their conduct, derives from perpetrator-focused considerations—they have fulfilled some even more stringent moral duties which they have owed to the captive and to the general public. They have achieved that feat, moreover, while employing techniques of torture that will not permanently injure the kidnapper. Hence, notwithstanding that their use of interrogational torture is morally impermissible (on perpetrator-focused grounds), it is morally optimal.

3.3.2.  A second example Though I have slightly modified the facts of Leon v. Wainwright, my version of those facts (which I have sketched in Chapter 2) is highly credible. It reveals that calamity-averting interrogational torture can be morally optimal—though still morally wrongful, of course—in circumstances that are not outlandishly fanciful. For an even more vivid example of calamity-averting interrogational torture that is morally optimal, however, we should turn to a far-fetched scenario propounded by Steinhoff. The unreality of his scenario is offset by its piquancy, and is in any event untroublesome because I could have broached here many other possible situations of morally optimal interrogational torture that are far more credible than Steinhoff’s lively thought-experiment. Steinhoff’s presentation of his example is long and complicated, and there is no need for me to quote it in full. The opening portion of his presentation should be quoted directly, whereas the rest of it can aptly be paraphrased: Bill works for a company that has a lot of trolleys on its enormous property to transport different goods. He is in charge of the maintenance of the trolleys. There is some kind of animal in the region that often enters the trolleys from below and bites through the wires. Therefore, Bill planted several foot traps, which, however, can also trap humans. The traps have combination locks, and Bill knows the combination. In order to set in motion certain trolleys, one has to hold on to a lever well above one’s head. Since Bill is very small, he has to jump to reach the lever. One day, Jeanette and Paolo, two completely innocent persons, cross the tracks and both accidentally step into a foot trap. Jeanette shouts to Bill: ‘Help us!!’ ‘You wish’, he shouts back. ‘I prefer to kill you.’ And he jumps up to a lever and sets in motion a trolley, which is slowly but fatally moving in Jeanette’s and Paolo’s direction. If not stopped, it will crush them. Jeanette has with her both her explosive projectile gun (these projectiles can blow people into small pieces but do not much affect trolleys) and her pain-infliction ray gun. Bill, for whatever reasons, would rather die than let the two escape. Fearing that they might shoot at him with normal guns so that he lets go of the lever, which would stop the trolley, he handcuffs himself to the lever and throws away the keys, and shouts sneeringly: ‘I know the combination of your traps—but I won’t tell you. I will watch you die.’ Even if they shoot him dead, that would not stop the trolley since Bill would still be hanging on to the lever by the handcuffs. Jeanette draws her pain-inflictor and shows it to Bill: ‘If you do not tell me the combination of the traps, I will torture you! This gun inflicts pain like a dentist drilling on an unprotected nerve.’ Bill remains silent. Jeanette sadly aims the pain inflictor gun at him. (Steinhoff 2009, 51, italics removed)

As Jeanette prepares to operate her pain-inflicting device, Paolo vigorously expostulates with her. He urges her not to stoop to the perpetration of torture,

214

Why Torture is Wrong

and he reminds her that killing in defence of oneself or others against a lethal attack can be morally permissible. Paolo implores Jeanette to use her explosive projectile gun to blow Bill to pieces and thereby prevent him from completing his fatal attack against them. She resists any such exhortation, by pointing out that her pain-inflicting device will harm Bill far less severely than her projectile gun. Instead of blowing him into small pieces, she will cause him for a fairly short period to experience pain which—though harrowingly intense—will not produce any permanent injuries or debility. However, because Paolo is so incensed by her intention to employ torture, he knocks her out and grabs her projectile gun and fires it squarely at Bill. Having pulverized Bill’s body into tiny fragments, Paolo manages to stop the trolley from its lethal progress toward Jeanette and himself. He is delighted at his success and especially at his having achieved that success without any recourse to torture. As he brushes off some bits of Bill’s corpse that have been strewn over his jacket by the explosion of the projectile, he commends himself for having upheld human dignity. Steinhoff believes that his scenario recounts a situation in which the use of calamity-averting interrogational torture would be morally permissible. In fact, however, his scenario recounts a situation in which the use of such torture would be morally optimal despite being morally impermissible. We should attend to the subtleties of that situation (subtleties paralleled in some of the examples of torture adduced independently by Kamm). Although the wielding of deadly force in defence of oneself or others against a wrongful and lethal attack can be morally legitimate, it is not always so. In particular, the wielding of such force is not morally legitimate if a significantly less harmful tactic is feasible as an alternative. In the situation outlined by Steinhoff, a less harmful tactic is indeed feasible and is known to be feasible. Hence, contrary to what Paolo thinks, his firing of the explosive projectile gun at Bill is a serious breach of a moral duty. Still, the impermissibility of Paolo’s employment of deadly force does not mean that the use of torture by Jeanette would be morally permissible. Rather, the predicament portrayed by Steinhoff is a situation of moral conflict. Under the Minimal Invasion Principle as applied to private individuals, Jeanette and Paolo are morally obligated not to use the explosive projectile gun as their means of fending off Bill’s fatal attack. On independent grounds, furthermore, each of them is morally obligated to do what is necessary and feasible to fend off the fatal attack in order to save the life of the other as well as his or her own life. The latter obligation requires them to employ either the explosive projectile gun or the pain-inflicting device to thwart Bill’s onslaught, while the Minimal Invasion Principle requires them to forgo any use of the explosive projectile gun. However, they are likewise morally obligated to refrain from any recourse to interrogational torture with the pain-inflicting device. For the reasons already discerned through this chapter’s adoption of a perpetrator-focused perspective, any plying of interrogational torture is morally illegitimate as a quest for overweeningly fine-grained control of another person’s conduct through the exploitation of his or her susceptibility to feelings of severe pain. Such torture is morally illegitimate even when it is undertaken to avert a calamity, for it is then a morally improper means to a morally

Moral Optimality without Moral Permissibility

215

worthy end. Hence, given that Jeanette and Paolo are each morally duty-bound to do what is necessary and feasible to stave off Bill’s deadly assault, and given that they are morally obligated not to wield the projectile gun and also morally obligated not to wield the pain-inflicting device, they are morally duty-bound to commit a moral wrong. They are faced with a plight in which some stringent moral duties incumbent on them are in conflict. In that plight, the morally optimal course of action for them lies in resorting to interrogational torture. Steinhoff’s scenario, like a few of Kamm’s examples, nicely brings out the moral complexities that arise in certain emergencies. Were Jeanette possessed only of the explosive projectile gun without the pain-inflicting device, she or Paolo would be morally permitted (strongly permitted) to use the gun in the circumstances to blow apart Bill’s body. However, because of the availability of the pain-inflicting device, neither Jeanette nor Paolo is morally permitted to resort to the projectile gun. Still, although the availability of the pain-inflicting device produces such an effect—through its triggering of the Minimal Invasion Principle as applied to private individuals—any employment of that device against Bill would itself be morally impermissible. The sheer fact of its moral impermissibility is unaffected by its moral obligatoriness, though of course the gravity of the wrong committed through the subjection of Bill to interrogational torture would be markedly extenuated by the exigencies of the situation.

3.3.3.  Some general considerations Let us conclude by glancing at some general considerations that bear on the matter of moral optimality. My terse discussion of some of those considerations will take for granted that, in any situation under scrutiny here, the use of interrogational torture is the sole feasible course of action—or the mildest feasible course of action—that is likely to avert a calamity. In other words, this discussion will take for granted that the employment of such torture in any situation broached here would not contravene the Minimal Invasion Principle.

3.3.3.1.  Harshness and protractedness One major factor that affects the gravity of the wrong committed through the use of interrogational torture is the specific nature of the torture. As has been noted en passant at several junctures in this book, some techniques of torture are much worse than others. Although the recent efforts by some politicians and scholars to distinguish between torture and ‘torture-lite’ (or ‘enhanced interrogational techniques’) are to be discountenanced,23 the correct point underlying those dubious efforts is that the methods by which torture can be inflicted on victims are multifarious and are of many different degrees of injuriousness. In addition, the 23   For one of the efforts to which I refer, see Elshtain 2004, 85–8. For some critical treatments of such efforts, see Green and Ward 2009, 167; Kreimer 2005; Luban 2005, 1436–8; Meisels 2008b, 186–8; Miller 2011, § 4; Waldron 2010a, 7–10. See also Moore 1997, 726; Moore 2007, 77–8.

216

Why Torture is Wrong

prolongedness of the application of any such methods is obviously something that can vary markedly. While every instance of interrogational torture is a breach of a stringent moral duty, the gravity of the breach generally increases in direct proportion to the lengthiness of the torture and to the harshness or brutality of the techniques. Among the reasons for the moral optimality of the interrogational torture in my variant of Leon v. Wainwright is that the techniques of torture applied to the kidnapper—the twisting of his arm and the choking of his neck—are undertaken for only quite a brief period and are unlikely to produce any permanent and very serious injuries. Though the pain induced by those techniques is certainly severe enough to be correctly classifiable as torturous, it is not likely to be physically or psychologically debilitating over the long term. Similarly, the torture to be wielded by Jeanette against Bill in Steinhoff’s scenario would not be very protracted. Though the pain elicited by it would be agonizing, it would not be likely to damage Bill’s bodily and mental well-being over the long term or even over the medium term. Partly as a consequence, Jeanette’s use of the pain-inflicting device against Bill would be morally optimal even though it would be morally impermissible.

3.3.3.2.  The exigencies of an emergency Among the contemporary philosophers and jurists who believe that interrogational torture can sometimes be morally justified, nearly everyone is focused on contexts in which the use of such torture is aimed at averting the occurrence of calamities. Those writers err in thinking that interrogational torture can ever be morally legitimate, but their concentration on dire emergencies is apposite—for only in such emergencies can the plying of interrogational torture ever be morally optimal. Only in such terrible circumstances can a duty that conflicts with the absolute duty-to-eschew-interrogational-torture be sufficiently stringent to exceed the latter duty in moral importance. While the recent endeavours to justify the employment of torturous interrogational techniques have most frequently adverted to ticking-bomb quandaries in which the lives of thousands or even millions of people are imperilled, the employment of some such techniques can be morally optimal even in certain emergencies where the lives at stake are numerically on a much smaller scale. Steinhoff’s scenario of Jeanette and Paolo recounts an emergency of that kind, and so does the scenario which I have modelled on Leon v. Wainwright. In each of those cases, someone desperately needs information in order to rescue one person or a couple of people from a dire fate. Because of that pressing need, and because the techniques of torture to be employed are toward the milder end of the spectrum, the wielding of those techniques can be morally optimal in either situation despite the small number of lives at stake—provided, of course, that in each situation there are reasonable grounds for believing that the use of those techniques is likely to be effective in extracting the vital information. No recourse to interrogational torture is ever morally permissible, but the recourse by the police in my scenario and the recourse by Jeanette in Steinhoff’s scenario (if she had not been stymied from acting) are morally optimal. In each

Moral Optimality without Moral Permissibility

217

predicament, the use of such torture can fulfil a moral duty even more stringent than the moral duty which the use of such torture contravenes.

3.3.3.3.  Threateningness and responsibility In my variant of Leon v. Wainwright and in Steinhoff’s scenario of Jeanette and Paolo and Bill, the person to be subjected to interrogational torture is directly responsible for the extremely dangerous quandary that has given rise to the need for calamity-averting information. In each case, moreover, the person to be subjected to interrogational torture is posing a continuing threat by withholding the vital information. Those two closely related factors do not render the interrogational torture morally permissible, of course, but they do mitigate the gravity of its wrongfulness. Partly because of those factors, the interrogational torture in each case is (or would be) morally optimal. The properties of threateningness and responsibility are scalar; people can partake of them to many different degrees. Hence, the extent to which those properties attenuate the gravity of the wrongfulness of interrogational torture in any number of settings is highly variable. What can be said at a general level, however, is that someone who does not partake of either of those properties in any given emergency is vested with an overtoppingly robust moral right against being subjected to interrogational torture therein. For example, if legal-governmental officials subject the young child of a terrorist to torture in the hope of extracting some calamity-averting information from the terrorist, they will have breached a moral duty that is even more stringent than any moral duty (relating to the safety of the public) with which they are seeking to comply. Though the dire circumstances that have prompted the perpetration of the torture do mitigate its enormity, any moral duties fulfilled are less important than the moral duties that are transgressed. No recourse to interrogational torture against an unthreatening and innocent person can ever be morally optimal.

3.3.3.4.  Probable efficacy As has been stated at the outset of this discussion of some general considerations that bear on the matter of moral optimality, I am assuming that the use of interrogational torture in any emergency under examination here would be in compliance with the Minimal Invasion Principle. Unless that principle is satisfied, no resort to calamity-averting interrogational torture (or any other kind of torture) can ever be morally optimal. Likewise, the employment of such torture is never morally optimal if there are hardly any reasonable grounds for believing that it will be efficacious. Though the grounds in question do not have to be highly specific, the absence of any reasonable basis for attaching a significant probability of success to the use of some envisaged technique of torture will hugely accentuate the gravity of the wrong that is constituted by any such use. Naturally, certainty is not required for the moral optimality of a decision to use torture—since well-founded certainty relating to any area of human interaction

218

Why Torture is Wrong

is utterly unattainable—but reasonable grounds for a significant level of confidence are indeed required. This point deserves emphasis because Steinhoff has ill-advisedly contended that even a tiny probability of success is enough to render permissible the use of interrogational torture for the purpose of averting a calamity (2006, 342–3). Of course, one reason for rejecting his contention is that the wielding of such torture is never morally permissible. However, even if he were referring to moral optimality rather than to moral permissibility, his position concerning the probable efficacy of torture would be untenable. Steinhoff conjures up a situation in which the probability that the use of torture will succeed in averting a calamity is 0.000001. Though his example involves ephemerally incapacitative torture with a pain-inflicting gun, he adduces the example to make a point about interrogational torture. As he affirms, ‘the application to the torture of ticking bomb terrorists and Dirty Harry kidnappers is obvious’ (2006, 343). Squarely addressing the question whether any torture should be undertaken with such a vanishingly low probability of success, Steinhoff declares that ‘I cannot even begin to make sense of the suggestion that, given these odds, the defender should not use the [pain-inducing] gun against the aggressor. Yes, the pain inflicted by the weapon on the aggressor is extremely unlikely to secure the survival of the defender, but there still is a chance that it will, so why should the defender forgo this chance for the benefit of the aggressor? Obviously, there is no reason’ (2006, 343, emphases in original). Despite Steinhoff’s closing suggestion that his view of the matter is obviously correct, the appropriate retort to his view is in fact obvious. Let us assess his argument with reference to interrogational torture. Given the odds which Steinhoff specifies, the overwhelmingly likely upshot of any recourse to interrogational torture is not only the sullying of the moral integrity of the perpetrators but also the occurrence of the calamity which the perpetrators have sought to prevent. Furthermore, the perpetrators will have administered the torture while knowing that it stands virtually no chance of attaining the purpose for which they are resorting to it. Hence, although the employment of interrogational torture would have been morally illegitimate even if it had been morally optimal as a desperate tack with a significant prospect of success in averting a calamity, the gravity of its wrongfulness is greatly augmented by the overwhelming odds against such success. Any recourse to interrogational torture under these circumstances will fall radically short of being morally optimal. Indeed, any recourse to such torture in these circumstances is not morally obligatory at all. There is no moral conflict. In such circumstances the abiding duty to eschew the use of interrogational torture is not countervailed by any conflicting moral duty, since—even in the context of a grim emergency—the officials responsible for public safety are not under any moral duty to proceed with torture that will predictably prove to be futile. Very different is the situation recounted in my variant of Leon v. Wainwright, and very different as well is the situation recounted in Steinhoff’s narrative of Jeanette and Paolo and Bill. When the police in my scenario twist the arm and choke the neck of the kidnapper in their custody, they have reasonable grounds for believing

Moral Optimality without Moral Permissibility

219

that the severe pain inflicted by those methods of torture is likely to loosen his tongue. Their grounds for harbouring that belief are probably not very specific in relation to the particular suspect, but their general familiarity with the forcible restraint of wrongdoers is sufficient to provide a basis for the belief. Because of the significant likelihood that the methods of torture carried out by the constables will prove to be efficacious, and because of the other features of the situation that have been highlighted in the last few subsections, the recourse to those methods is morally optimal. Such techniques of torture are morally wrong, but they are less gravely wrong in the dreadful circumstances than would be the remissness of not resorting to them after the other feasible interrogational approaches have failed. Similarly, Jeanette has a reasonably solid basis for thinking that her use of the pain-inflicting device against Bill will induce him to reveal the information that will save her life and Paolo’s life. Though she probably does not have any detailed knowledge of Bill’s psychology, her general awareness of the ways in which human beings respond to the drilling of the unanaesthetized roots of teeth is sufficient to ground her expectation that the pain-inflicting device will extract the vital information from Bill. In that respect, her situation differs pronouncedly from the situation of someone who knows that a torturous measure’s chances of success are 0.000001. Unlike the latter measure, then, her use of the pain-inflicting device in a dire emergency can be morally optimal even though it is morally wrong.

3.3.3.5.  Legal sanctions Chapter  5 will explore at length the proper legal responses to any instances of torture that are undertaken by public officials or private individuals. For the moment, the present chapter will conclude by emphasizing afresh that the moral optimality of some such instances does not entail their moral permissibility. Though ephemerally incapacitative torture and edifying torture can sometimes be morally permissible, every other type of torture is morally wrong in all of its possible instantiations. Consequently, although calamity-averting interrogational torture can sometimes be morally optimal, it is in breach of a redoubtable moral duty even when it is indeed optimal. Given as much, and given that it involves some major physical or psychological violence, any wielding of such torture should trigger a reaction from the system of legal governance in the jurisdiction where it occurs. This book’s closing chapter will ponder in more detail the legal sanctions that are appropriate in the aftermath of acts of torture or attempted torture. At present, the key point is that the moral optimality of an instance of torture should not be legally exonerative any more than it is morally exonerative. Even when an instance of ephemerally incapacitative torture is morally permissible, it should evoke a legal reaction in the form of an enquiry into the circumstances that have led to the use of the torture (just as shootings of suspects by constables are routinely investigated in liberal democracies even when the shootings have been morally legitimate applications of force). A fortiori, when a morally optimal instance of torture is not morally permissible, it should evoke a legal reaction. It should be met with

220

Why Torture is Wrong

the imposition of some sanction(s), whether collective or individual, and whether criminal or civil or institutional. Moral optimality uncombined with moral permissibility is no shield against such an imposition. Though the moral optimality of a morally impermissible act of torture does considerably extenuate the gravity of the act’s wrongness and does therefore warrant a considerable reduction in the heaviness of any sanction(s) imposed, it does not eliminate altogether the need for some sanction(s). The torture has sullied not only the moral integrity of anyone who has perpetrated it, but also the moral integrity of the society on whose behalf the torture has been conducted. That society would be further tarnished if its legal system were to acquiesce in the commission of such a wrong. Hence, while this chapter accepts that calamity-averting interrogational torture can in some extreme circumstances be morally optimal, I am hardly suggesting that those circumstances are fully exculpatory. The closing pages of this chapter have not been aimed at letting anyone ‘off the hook’.

4 The Rationality of Deontological Constraints My third chapter has argued that most types of torture are morally wrong in all of their possible instantiations. That absolutist position on the morality of torture is in need of some further defence here, for quite a few philosophers have queried whether such a position is reconcilable with the requirements of practical rationality. These doubters have in mind especially the agent-centredness of our duties to refrain from perpetrating acts of torture. While granting that those duties are operative in any ordinary contexts, the opponents of deontological constraints insist that the duties would be perverse if they continued to be binding in circumstances where the perpetration of acts of torture can lower the overall incidence of such acts. That charge of perverseness—amounting to a charge of irrationality—is what the present chapter will seek to rebut. Chapter 1 has presented two ways in which deontological doctrines can be distinguished from consequentialist doctrines. First, whereas the former doctrines hold that the moral permissibility or impermissibility of many types of conduct is independent of the consequences that are likely to ensue from any instances of those types of conduct, the latter doctrines contend that the moral permissibility or impermissibility of every type (or instance) of conduct is determined fully by the consequences that are likely to ensue from the occurrence of the conduct. Second, whereas deontologists maintain that many of the moral duties incumbent on us are agent-centred, consequentialists declare that every moral duty is agent-neutral. Now, as Chapter 1 has remarked, these two ways of framing the deontology/consequentialism division are not really distinct; the second way is comprehended within the first as one of its chief aspects or implications. Nonetheless, the separate presentation of the agent-centred/agent-neutral dichotomy is salutary, since the insistence of deontologists on the agent-centredness of many of our moral duties is the leading point of contention in the debates between deontologists and their consequentialist opponents. It is the very point of contention that will be addressed sustainedly in this chapter. As Chapter 1 has acknowledged, the consequence-independence of the impermissibility of various act-types is not per se tantamount to the absoluteness of our duties to refrain from those act-types. More specifically, then, the agent-centredness of those duties is not per se tantamount to the absoluteness of their sway. Threshold deontologists such as Michael Moore affirm that our duties to refrain from

222

The Rationality of Deontological Constraints

interrogational torture are agent-centred, while also contending that those duties cease to be binding past some high threshold level of terrible consequences that will ensue from compliance with them. Deontologists of Moore’s stripe are not absolutists. Still, although not all deontological approaches to morality are absolutist (in regard to torture and in regard to other types of conduct), the moral prohibition on any given mode of conduct MC cannot be absolute unless the wrongness of MC is consequence-independent. Accordingly, the agent-centredness of our duties to abstain from MC is indispensable for the absoluteness of those duties. Albeit the agent-centredness of those duties is not a sufficient condition for their absoluteness, it is a necessary condition. Thus, by defending the rationality of the agent-centredness of many key duties—including our duties to forgo any recourse to most types of torture—this chapter is contributing crucially to my insistence on the absolute wrongness of most types of torture. In its espousal of an absolutist position, this book accentuates the difficulties to be faced and overcome here. After all, given that agent-centred prohibitions strike some philosophers as perverse or irrational, the problem will seem especially acute if there is no threshold past which those prohibitions cease to be binding. In addition to taking an unflinchingly absolutist stance that poses the problem of deontological rationality in its starkest form, this chapter will explore a moral conflict that involves an intrapersonal trade-off rather than an interpersonal trade-off. Typically, trade-offs of the former sort are morally far less problematic than trade-offs of the latter sort. Hence, by seeking to establish that the moral prohibition on placatory torture continues to be binding even in the context of an intrapersonal trade-off where a breach of that prohibition would redound to the benefit of the victim of the breach, this book is deliberately confronting itself with a challenge that leaves no easy routes of escape. Both in its countering of some doubts about the rationality of agent-centred constraints and in its efforts to vindicate the absoluteness of the moral ban on placatory torture (and, by extension, on most other types of torture), this chapter takes the hard and narrow path. It aims to surmount obstacles that are encountered here in their most troublesome guises. In line with what has just been suggested, the first main section of this chapter will recount a situation in which the perpetration of placatory torture against a certain victim would lower the incidence and severity of the acts of torture carried out against that very same victim. As will be argued, the perpetration of placatory torture is morally wrong even in such a situation. We will then move on to ponder whether the agent-centredness of deontological duties—such as our duties not to engage in acts of torture—is rationally tenable. Among the philosophers who have impugned the rationality of agent-centred constraints, Samuel Scheffler has mounted a particularly prominent and thought-provoking attack.1 He tries to show that agent-centred restrictions clash not only with the tenets of consequentialist moral theories but also with a conception of practical rationality that extends beyond those theories into much of our ordinary thinking about an array 1   His critique initially appeared as Scheffler 1985, and has been reprinted as Scheffler 1988 and Scheffler 1994, 133–51. In this chapter, my citations are to the 1988 reprinting of his essay.

Placatory Torture

223

of matters. By so arguing, Scheffler presumes to have exposed the unreasonableness of deontological requirements without begging any questions along the way in favour of consequentialist theorizing. Hence, my rebuttal of his influential critique is essential for the sustainment of one of this book’s cardinal contentions: namely, my contention that most types of torture are always and everywhere morally wrong in all possible worlds. Those types of torture are morally wrong even when instances of them are undertaken to avert the occurrence of many other such instances.

4.1.  Placatory Torture and the Unremittingness of Deontological Duties In most of the philosophical discussions where deontologists and their opponents contemplate situations in which certain acts of torture will lower the overall incidence of such acts, the trade-offs are interpersonal. That is, some people are to be subjected to torture for the sake of sparing numerous others from comparably dire mistreatment. (Much more often in the scenarios that are broached in these philosophical discussions, the other people are to be spared from sudden death. However, for the purposes of this chapter, we need to concentrate on situations in which acts of torture are undertaken to avert the occurrence of other acts of torture that are more numerous or more severe.) Though scenarios that involve interpersonal trade-offs are valuable, an investigation of them does not shed light on the most difficult cruxes that could arise. After all, when the person who is to undergo the torture in a scenario involving an interpersonal trade-off is not a malefactor—and, thus, when contexts of ephemerally incapacitative torture have been put aside2—we can very easily find a ground for deeming the envisaged use of torture to be morally wrong. If an innocent person P is subjected to excruciating pain for the furtherance of the well-being of others, then P has been used purely as a means for the advancement of the ends of others. Accordingly, the administration of the torture is straightforwardly condemnable on moral grounds, even if it is morally optimal in the circumstances. Seemingly, one way to keep the endeavour of moral condemnation from being too easy is to tweak the thought-experiment by adding to it the factor of consent or apparent consent. In other words, we can embellish the thought-experiment in much the same way in which I have amplified Alan Gewirth’s scenario of placatory torture in § 3.2.3.3.9 of Chapter 3. If the victim of the torture has consented to undergoing it in order to save some other people from being afflicted with agonizing pain, then the subjection of her to the torture might appear to be classifiable as 2   Contexts of edifying torture have of course been left aside as well, since a situation in which some person P gets subjected to torture in order to save other people from dreadful mistreatment is manifestly not a situation in which P gets subjected to torture for the enhancement of her own bodily and psychological well-being. Even if P happens to be a masochist who will derive gratification from the agony-inducing techniques that are brought to bear on her, those techniques are undertaken for the purpose of placation rather than for the purpose of providing P with masochistic pleasure.

224

The Rationality of Deontological Constraints

one of her projects—in which case she is not being treated solely as a means for the furtherance of others’ ends during that subjection. However, as I have observed in Chapter 3, the provision of consent under conditions of extreme duress does not significantly alter the moral character of the situation. Sincere and selfless though the consent may very well be, it is tendered under conditions that deprive it of genuineness. Hence, the legitimizing force bestowed by genuine consent is not bestowed by the coerced consent of a person who braves torture in order to spare other people from a similar fate. Her preparedness to sacrifice her own well-being for theirs does ethically alter some aspects of the situation, but it does not alter the fundamental moral status of the torture from impermissible to permissible. Of course, as my remarks in § 3.2.3.3.9 have made clear, quite a bit more can be said about my embellished version of Gewirth’s scenario. Nevertheless, the foregoing paragraph is sufficient to indicate that my amplification of the scenario—which leaves in place the interpersonal character of the trade-off that is depicted—does not really pose a stiff challenge to an absolutist position on the morality of torture. Detecting the wrongness of the torture administered in that scenario is not a very difficult feat. Let us, then, probe a more formidable example that replaces the interpersonal trade-off with an intrapersonal trade-off.

4.1.1. A thought-experiment: sparing someone from a greater wrong Fifteen thugs burst into the house of Melvin and Miranda one evening. Several of them tie Miranda firmly to a plinth on which they have splayed her. Brandishing guns and explosives, the thugs order Melvin to subject Miranda to torture by applying some lighted cigarettes and an electric-shock device to sensitive parts of her body repeatedly. They admonish Melvin that, if he shrinks from this ghastly assignment, they themselves will apply those methods of torture to Miranda’s body far more protractedly along with several other methods of torture that are even more excruciating. Melvin is faced with a stark choice, then. Either he subjects his wife to torture himself or he lets her be subjected to a considerably harsher ordeal of torture by the thugs. (In the event of his non-compliance with their behests, he will of course be securely restrained by several of the thugs while some of their comrades administer the torture to Miranda’s body.) Although Melvin of course will harbour some uncertainty concerning the thugs’ intentions, his epistemic situation is not racked with doubts. He knows that the thugs have carried out a number of similar raids on the homes of married couples in various parts of the country. On each occasion when a husband has engaged in the terrible acts of torture which he has been ordered to perform, the thugs have then left the premises without further ado. Contrariwise, on each occasion when a husband has refused to do what the thugs have directed him to do, they have fulfilled their threat by applying torture prolongedly and brutally to his wife. Melvin knows these facts, which have been reported widely in the news media over the months when these atrocious incidents have taken place. He is therefore confident—not absolutely certain, of course, but confident—that the thugs mean

Placatory Torture

225

just what they say when they bid him to torture Miranda in order to spare her from an even more hellish process of mistreatment at their hands. Miranda likewise knows about the past raids conducted by the thugs, and therefore she too is confident that her tormentors mean exactly what they say. She consequently urges her husband to obey their behests. She hardly relishes the prospect of being afflicted with severe pain by the acts of burning and shocking to which he will have to stoop, but she wants to avoid the grimmer ordeal that will be inflicted on her if Melvin declines to perform those dismaying acts. She recognizes that he is filled with revulsion at the thought of torturing her, but she implores him to overcome that revulsion. With huge reluctance and consternation, Melvin heeds her entreaties.

4.1.2.  The significance of the thought-experiment Although this tale of Melvin and Miranda is in many respects similar to my modified version of Gewirth’s scenario of the politician and the mother, the key difference of course is that Melvin has undertaken the placatory torture not with the aim of saving the lives of numerous other people but instead with the aim of minimizing the agony that will be inflicted on Miranda. The harrowing trade-off is intrapersonal rather than interpersonal. As a result, a vindication of this book’s absolutist position on the wrongness of placatory torture cannot rest content with Robert Nozick’s account of the inviolability of individuals: Side constraints [that is, deontological duties] express the inviolability of . . . persons. But why may not one violate persons for the greater social good? Individually, we each sometimes choose to undergo some pain or sacrifice for a greater benefit or to avoid a greater harm: we go to the dentist to avoid worse suffering later; we do some unpleasant work for its results; some persons diet to improve their health or looks; some save money to support themselves when they are older. In each case, some cost is borne for the sake of the greater overall good. Why not, similarly, hold that some persons have to bear some costs that benefit other persons more, for the sake of the overall social good? But there is no social entity with a good that undergoes some sacrifice for its own good. There are only individual people, different individual people, with their own individual lives. Using one of these people for the benefit of others, uses him and benefits the others. Nothing more. What happens is that something is done to him for the sake of others. Talk of an overall social good covers this up. (Intentionally?) To use a person in this way does not sufficiently respect and take account of the fact that he is a separate person, that his is the only life he has. He does not get some overbalancing good from his sacrifice.3

Nozick’s remarks on inviolability are fine as far as they go, and indeed my third chapter has made similar remarks about some of the interpersonal trade-offs on 3   Nozick 1974, 32–3, footnote omitted, emphases in original. My narrative of Melvin and Miranda differs not only from Gewirth’s and Nozick’s examples, but also from Bernard Williams’s famous scenario of Jim and the Indians. See Williams 1973, 98–9. Williams envisaged a situation in which a traveller can save the lives of 19 Indians by killing one other Indian. Though the one Indian will not be made worse off by the traveller’s act of slaying him (since he would have been shot in any event), he will likewise not be made any better off by that act.

226

The Rationality of Deontological Constraints

which I have ruminated there. However, his account of inviolability does not go far enough. For one thing, as Chapter 3 has noted, a focus on inviolability will not in itself enable anybody to explain why heinous wrongdoers such as Khalid Sheikh Mohammed can never legitimately be subjected to interrogational torture. Much more important in the present context is that Nozick’s reflections on inviolability—powerful though they are—do not address the matter of coerced intrapersonal trade-offs. That is, they do not address the question whether someone can legitimately be mistreated to spare her from even worse mistreatment. What makes the scenario of Melvin and Miranda problematic is its posing of that very question.

4.1.3.  The moral upshot In one respect, of course, the situation depicted in my tale of Melvin and Miranda is morally clear-cut. Any sensible person would agree that the actions of the thugs are morally reprehensible. However, the interesting question about the situation obviously pertains not to their actions but to the actions of Melvin. Has he behaved permissibly when he resorts to torturing his wife in the dire circumstances? Has he behaved permissibly by complying with her own expostulations as well as with the bidding of the thugs?

4.1.3.1.  The factor of consent Although Miranda has consented to being tortured by Melvin, her consent is not genuine. Her consent is of course fully sincere and indeed desperately exhortative, but it is provided under conditions of overwhelming duress. Its genuineness is thoroughly vitiated by its coercedness. Hence, the permissibility-engendering effect produced in many circumstances by genuine consent is not produced by Miranda’s impetrations. In other words, as far as the moral effects of the victim’s consent to mistreatment are concerned, the plight of Melvin and Miranda is indistinguishable from the plight of the politician and the mother in my amplified version of Gewirth’s scenario. In each case, the extreme coercion under which the victim gives her consent is destructive of any legitimizing force which the consent might otherwise have possessed. That destructive upshot of the coercion is independent of the distinction between an interpersonal trade-off and an intrapersonal trade-off. Notwithstanding that Miranda furnishes her consent for her own good whereas the mother in the Gewirthian scenario furnishes her consent for the good of others, Miranda’s consent does not confer permissibility on Melvin’s acts of torture any more than the mother’s consent confers permissibility on the politician’s acts of torture. Though the intrapersonal/interpersonal distinction might make a moral difference in some other respects, it does not here make a moral difference with regard to the factor of consent. The legitimacy-engendering effect of any instance of consent depends both on the genuineness of the consent and on the legitimacy of the purpose for which the consent has been tendered. While Miranda’s purpose of sparing herself

Placatory Torture

227

from additional agony (through an intrapersonal trade-off) and the mother’s purpose of saving other people’s lives (through an interpersonal trade-off) are both legitimate, the consent in each case is deprived of its voluntariness—and thus of its genuineness—by the terrible and credible threats to which it is a response. In neither case, then, does the consent render morally permissible the use of placatory torture. Still, although Miranda’s fervent exhortations to Melvin do not alter the basic moral status of his acts of torture from impermissible to permissible, they do mitigate the gravity of the wrongness of those acts. Whereas Miranda’s remonstrations would be bizarrely unreasonable in any ordinary circumstances, they are quite understandable in the appalling predicament in which they are uttered. Hence, when Melvin proceeds in accordance with those remonstrations, he is fulfilling some earnestly expressed wishes of his wife that are reasonable in the circumstances. In that respect, his outlook is very different from that of a typical torturer. Though the coercedness of Miranda’s expressions of her wishes is enough to deprive them of any legitimizing force, it does not deprive them of extenuative force. The fact that Melvin does what his wife desperately and understandably wants him to do is a consideration that manifestly lessens the culpability of the wrong which he has to commit against her.

4.1.3.2.  The orientation of the torture Any morally legitimate instance of edifying torture is benign in its orientation toward the bodily and psychological well-being of the victim. By contrast, the torture administered to the mother by her son in my amplified version of Gewirth’s scenario is inimical to the mother’s bodily and psychological well-being. Were the torture not to take place, the mother would be much better off (except in her distress over the harm inflicted on other people by the explosions which the terrorists would detonate in response). Less straightforwardly classifiable is the orientation of the torture undertaken by Melvin against Miranda; here the problem of fixing upon the germane baseline of comparison looms large. On the one hand, if we ask whether Miranda will be better off if Melvin declines to perform any acts of torture, the answer is negative. Given that the thugs are resolutely disposed to perpetrate far worse acts of torture against her in the event of Melvin’s non-compliance with their behests, his engaging in placatory torture redounds to the benefit of her bodily and mental well-being. Injurious to her though his acts of burning and shocking are, they forestall the considerably worse onslaughts that would have been inflicted by the thugs in response to his defiance of their orders. Thus, if the appropriate baseline of comparison is the course of events that would have unfolded in the presence of such defiance, the orientation of Melvin’s placatory torture toward Miranda’s physical and psychological welfare is benevolent. On the other hand, if the appropriate baseline of comparison is instead a course of events in which nobody contravenes the moral prohibition on the perpetration of torture against Miranda, the acts of placatory torture by her husband make her

228

The Rationality of Deontological Constraints

considerably worse off. The exquisite pain and wounds caused by his acts of burning and shocking are obviously detrimental to her elementary well-being, if they are compared with the absence of any such untoward effects in the baseline course of events that has just been broached. Because his acts of torture foreseeably cause such deleterious effects, the orientation of his conduct toward Miranda’s bodily and mental soundness is destructive rather than benignly supportive. Of course, Melvin engages in those acts not with relish but with the utmost repugnance. Still, the impact of his torturous techniques on Miranda—in comparison with the morally required situation in which no one afflicts her with any such techniques—is severely harmful. Now, if our moral assessment of Melvin’s behaviour is to be well-founded, we need to take account of the full array of circumstances in which the behaviour has occurred. Given as much, the appropriate baseline of comparison for judging the harmfulness of his actions is the first rather than the second. The context in which Melvin resorts to torture is not the morally required situation in which everyone else is disposed to abstain from any torturous measures; he adopts his harrowing course of conduct precisely in recognition of that fact. Faced with a plight in which he cannot keep Miranda from undergoing agony, he does his best to minimize the harm and pain that will be inflicted on her—even though the only tack open to him that will produce such a result is the subjection of her to torture. In the dreadful circumstances, the orientation of his placatory torture toward her bodily and mental well-being is benevolent.

4.1.3.3.  Perpetrator-focused reflections Although the victim-focused considerations that we have pondered so far are somewhat inconclusive, they largely militate in favour of the proposition that Melvin’s recourse to placatory torture is morally legitimate. Miranda has not genuinely consented to the torture, but she has fervently consented in a context in which no genuine consent is possible. Notwithstanding that the burning and shocking of her body by Melvin will greatly impair her welfare, the orientation of those measures is benevolent in light of the horrific harm which they fend off. Hence, even though the victim-focused considerations are not entirely clear-cut, they do not on the whole supply a basis for deeming Melvin’s actions to be wrong. From a perpetrator-focused perspective, however, we can descry the wrongness of Melvin’s resort to placatory torture. Though Melvin is desperately well-intentioned as he takes the steps that are necessary to spare his wife from the viciousness of the thugs, he morally degrades himself by taking those steps. From a perpetrator-focused perspective, we can see that his conduct is fundamentally on a par with the conduct of the politician in my embellished version of Gewirth’s scenario. Melvin uses himself and Miranda as playthings of the thugs, as he shows himself ready—however reluctantly and disconsolately—to stoop to an especially demeaning and terrible mode of conduct for the sake of keeping the thugs at bay. Albeit his conduct is in conformity to his wife’s entreaties, the only people whose uncoerced wishes he obeys are the thugs. They gleefully want Miranda to

Placatory Torture

229

be tortured, and they particularly want Melvin to administer the torture. When he abases himself by submitting to their directives, he is ignominiously providing them with exactly what they desire. He has been forced by them into becoming an active participant in his own psychological humiliation and moral degradation, as he turns his agency into the vehicle through which the thugs afflict Miranda with excruciating pain. He indelibly stains the relationship between himself and his wife, as he subjects her to modes of mistreatment to which nobody can ever legitimately be subjected (outside certain edifying contexts). Though Melvin and Miranda might recover from their ordeal psychologically and physically, the ethical tenor of their relationship will henceforward be tarnished by his demonstration of his willingness to afflict her with such mistreatment. He has demonstrated that he is prepared to place his agency in the service of the thugs as an instrument of their torturous designs—designs that are centred on debasing him as just such an instrument.

4.1.3.4.  Moral optimality Melvin by resorting to placatory torture has sullied his moral integrity and his relationship with his wife. Albeit the inconclusive victim-focused considerations pertaining to the torture are not in themselves sufficient to deprive it of moral legitimacy, the perpetrator-focused considerations just outlined do indeed render Melvin’s conduct morally wrong. His conduct breaches a duty owed to himself and a duty owed to Miranda; the former duty requires Melvin to sustain the conditions for self-respect, while the latter duty requires him to abstain from gross forms of mistreatment. However, although his torturous actions are not strongly permissible, they are weakly permissible. That is, in the grim predicament in which Melvin performs those actions, they are morally optimal. His having recoiled altogether from performing them would have been even worse morally. Melvin has owed and contravened the moral duties mentioned above, but those are not the only moral duties that are incumbent on him in his plight. He owes a duty to Miranda to take all feasible and necessary steps to defend her against any foreseeable infliction of very serious harm.4 In the grotesque circumstances, the only way in which he can discharge that duty is to subject her to placatory torture. He will thereby be causing severe harm to her, of course, but he will be protecting her from the even worse harm that she would undergo at the hands of the thugs directly. Moreover, his recourse to placatory torture occurs not least as a response to her own supplications. If she had not urged him to accede to the behests of the thugs—and especially if she had adjured him to preserve his moral integrity and the integrity of their relationship by refusing to treat the two of them as playthings for the amusement and mollification of the thugs—his recourse to torture would not have been morally optimal. It would not have been even weakly permissible (though its gravity would still have been considerably mitigated by the horrific 4   She owes him a cognate duty, of course, though no steps are feasible for her after the thugs have tied her securely to the plinth.

230

The Rationality of Deontological Constraints

context of its occurrence). However, given that Miranda has in fact beseeched Melvin to overcome his scruples in order to lessen her misery, his torturous actions are morally optimal. While running athwart some stringent moral duties that continue to be binding upon him, his actions fulfil some even more stringent moral obligations that are in conflict with the duties which he has breached. Although the moral optimality of Melvin’s conduct does not render that conduct morally legitimate, it does markedly attenuate the gravity of the wrongs that Melvin has committed. He has wronged both himself and Miranda, but—given the full circumstances, including her robust expostulations—his declining to carry out the placatory torture would have wronged her even more gravely. In a terrible quandary that has confronted him with a choice between two morally impermissible courses of conduct, he has chosen correctly (even though his correct choice has not enabled him to avoid the commission of some wrongs). Consequently, the remedies morally required of him to repair the impermissibility of his actions will be relatively mild. In particular, as Chapter 5 will indicate, he should not be subjected to any criminal sanctions or other legal sanctions. If he is ever brought to trial on any criminal charges for his administration of torturous measures to Miranda, he should be acquitted with a full excuse of duress.

4.2.  Are Deontological Constraints Irrational? In the thought-experiment relating to Melvin and Miranda, as in any other setting where deontological constraints are applicable, the moral permissibility or impermissibility of someone’s conduct is not determined by any consequentialist balance of considerations. More specifically, the fact that a person can lower the overall frequency of some dreadful mode of conduct by engaging in an instance of that mode of conduct is not something that renders the instance morally legitimate. Melvin perpetrates acts of torture against Miranda in order to keep her from being subjected to even worse acts of torture. By any consequentialist reckoning, then, his conduct is impeccable. Under any such reckoning, his behaviour is not only morally optimal but also strongly permissible. By contrast, a deontological theorist can recognize that the moral impermissibility of placatory torture is consequence-independent. Irrespective of the fact that Melvin can lower the overall incidence of acts of torture by performing such acts, his resorting to any such measures is morally illegitimate. He wrongs both himself and Miranda when he burns and shocks her, even though the trade-off which he thereby accomplishes is intrapersonal rather than interpersonal. His recourse to placatory torture is morally optimal but not strongly permissible. My scenario of Melvin and Miranda thus raises, in an especially acute form, a problem that has troubled many proponents and critics of deontological doctrines. Is the agent-centredness of deontological constraints at odds with practical rationality? How can acts of torture be morally wrong if they foreseeably avert some far more numerous and brutal acts of torture that would otherwise have taken place? To philosophers of a consequentialist bent, any such ascription of moral

Are Deontological Constraints Irrational?

231

wrongness seems irrational. These philosophers do not fathom how we can ever be morally required to bring about states of affairs that are worse (perhaps far worse) than the states of affairs that we would foreseeably have brought about by acting contrary to the terms of deontological restrictions. Their doubts about the rationality of deontological prohibitions will be particularly intense with regard to situations which relevantly resemble that of Melvin and Miranda. How can Melvin be under a moral obligation to abstain from perpetrating any placatory torture against Miranda, given that his abstention from the placatory course of action would foreseeably bring about a state of affairs that is far worse for her (as well as worse more generally) than the state of affairs brought about through his perpetration of the torture? In the eyes of consequentialists, such an obligation is perverse and is therefore morally unsustainable.

4.2.1.  Rationality and maximization These consequentialist worries about the irrationality of deontological constraints have been voiced with trenchancy by Scheffler, who asks (1988, 244): ‘[H]‌ow can it be rational to forbid the performance of a morally objectionable action that would [foreseeably] have the effect of minimizing the total number of comparably objectionable actions that were performed and would have no other morally relevant consequences? How can the minimization of morally objectionable conduct be morally unacceptable?’ He endeavours to parlay these queries into a general critique of deontology by maintaining that agent-centred limits on the maximizing of good consequences or the minimizing of bad consequences run counter to demands of practical rationality that are widely and deeply embraced. He believes that, because those demands are routinely recognized as such in a myriad of contexts, the invocation of them against deontological doctrines does not damagingly beg any questions in favour of consequentialism. He writes:  ‘[A]lthough the conception of rationality that generates the appearance of paradox [in deontology] lies at the heart of consequentialism, it is not peculiar to consequentialism. On the contrary, it is a fundamental and familiar conception of rationality that we accept and operate with in a very wide and varied range of contexts.’ Claiming as much, Scheffler thus contends that the allure of consequentialism—an allure which, according to him, is felt by all or most opponents of consequentialism as well as by its champions— ‘derives from the fact that it appears to embody a notion of rationality which we recognized from myriad diverse contexts, and whose power we have good independent reason to respect’ (1988, 251). Scheffler proceeds to recount the nature of the rationality which he perceives as central to multitudinous everyday practices as well as to consequentialism: The kind of rationality that consequentialism seems so clearly to embody, and which makes so much trouble for views that incorporate agent-centered restrictions, is what we may call maximizing rationality. The core of this conception of rationality is the idea that if one accepts the desirability of a certain goal being achieved, and if one has a choice between two options, one of which is certain to accomplish the goal better than the other, then it is, ceteris paribus, rational to choose the former over the latter. Consequentialism seems

232

The Rationality of Deontological Constraints

to embody this kind of rationality because it starts from a conception of what is desirable (the overall good) and then tells us always to promote as much of it as we can. (1988, 252, emphasis in original)

In anticipation of my rejoinders to Scheffler’s critique of deontology, we should here glancingly note two points. First, Scheffler does not adduce any evidence to substantiate his repeated assertions that the maximizing conception of rationality is operative in a vast medley of quotidian contexts. That omission is significant, for—as I will contend later in this chapter—there are strong grounds for doubting that that conception in any unrestricted form is indeed widely operative in everyday settings. Second, if the ‘ceteris paribus’ clause in the extract above were to be construed expansively, it would leave ample room for a reconciliation between deontological restrictions and the maximizing conception of rationality. Scheffler recognizes as much, and he therefore goes on to argue that the ‘ceteris paribus’ clause cannot properly be invoked to effect such a reconciliation. I will impugn his argument on that point in due course.

4.2.2.  Deontological commitments In some ruminations which this chapter will scrutinize shortly, Scheffler explores the inconsistency between deontological constraints and the maximizing conception of rationality. However, before we probe those ruminations, we should take note of his contention that deontologists themselves are committed to that conception of rationality. His main explanation of their being so committed is contained in the following passage: [Deontological doctrines] seem committed to the idea that violations of [agent-centred] restrictions are morally objectionable or undesirable, in the sense that there is a moral point of view from which it is preferable that no violations should occur than that any should. Defenders of deontological views are typically happy to say things like this, and with good reason. For on standard deontological views, morality evaluates actions from a vantage-point which is concerned with more than just the interests of the individual agent. In other words, an action will be right or wrong, on such a view, relative to a standard of assessment that takes into account a number of factors quite independent of the interests of the agent. And defenders of such views are unlikely to claim that the relevant standard of assessment includes agent-centered restrictions, but that it is a matter of indifference, from the vantage-point represented by that standard, whether or not those restrictions are violated. For if it is not the case that it is preferable, from that vantage-point, that no violations should occur than that any should, it is hard to see how individual agents could possibly be thought to have reason to observe the restrictions when doing so did not happen to coincide with their own interests or the interests of those they cared about. In other words, deontological views need the idea that violations of the restrictions are morally objectionable or undesirable if the claim that people ought not to commit such violations when doing so would be in their own interests is to be plausible. (1988, 253–4, emphasis in original)

Now, on the one hand, proponents of deontological doctrines will tend to be wary of the language in this latest extract. Scheffler inappositely conveys the impression that those doctrines are grounded on evaluations of states of affairs. On the other

Are Deontological Constraints Irrational?

233

hand, the gist of what is said in the extract is unexceptionable—provided that it is amplified or qualified in one respect. Although deontologists rightly insist that every contravention of a deontological prohibition is morally wrong, they are not quite committed to the proposition that it is always better that no contraventions occur than that any do. Suppose for example that Edgar, who is a meteorologist or a seismologist, has to break a promise in order to take some steps that are necessary to avert a natural disaster. If we compare a situation in which Edgar keeps his promise and a situation in which he averts the disaster by departing from his promise, we are comparing a course of events in which no transgressions of deontological prohibitions occur and a course of events in which at least one such transgression occurs. (As a product of meteorological or seismic developments, the natural disaster would not consist in any contraventions of deontological restrictions if it were to proceed.) Deontologists should nonetheless allow that, because the consequentialist considerations in favour of Edgar’s taking the steps necessary to avert the catastrophe are extremely weighty, his breaking of his promise in order to take those steps is morally optimal. To be sure, his breach of his promise is morally wrong; it is only weakly permissible rather than strongly permissible, and it therefore places him under a moral obligation to apologize to the promisee or to undertake some other suitable remedial measure. Still, the wrongness of the non-fulfilment of the promise is entirely consistent with the moral optimality of such a course of conduct—and it is indeed a morally optimal course of conduct in the circumstances. Recognizing as much, deontological absolutists are not committed to the proposition that it is always better that no contraventions of deontological prohibitions occur than that any do. Many other examples could be adduced, not all of which would involve the prevention of natural disasters. When a deontological obligation (such as a minor promissory obligation) is not formidably stringent, and when a breach of that obligation can avert a very substantial detriment or bring about a very substantial benefit, and when the non-occurrence of the breach would not involve any contraventions of deontological duties, the situation can be such that a breach of a deontological obligation—despite its wrongness—is better in the circumstances than no such breaches at all. Still, Scheffler obviously does not have in mind quandaries of that kind. If we pretermit such quandaries, and if we think instead only of predicaments in which non-compliance with deontological constraints is necessary for the prevention of some even more numerous or serious instances of non-compliance, then the passage quoted above is correct in its account of one of the commitments of deontologists. Within the confines just indicated, deontologists are indeed committed to the proposition that it is always better that no transgressions of deontological prohibitions occur than that any do. Let us designate that proposition as the ‘No Contraventions Principle’. My ripostes to Scheffler’s critique of deontology will henceforth endorse the No Contraventions Principle and will not mention again the requisite qualification discussed in this paragraph and the preceding paragraph. (Any further mention of that qualification would be gratuitous, since it will play no role in my rebuttal of Scheffler’s critique.)

234

The Rationality of Deontological Constraints

4.2.3.  Slippage from none-versus-any to fewer-versus-more The fatal misstep that undoes Scheffler’s critique of deontology is his recurrent shifting from the No Contraventions Principle—a principle focused on the distinction between no transgressions of deontological prohibitions and some such transgressions—to a quite different thesis about the distinction between fewer transgressions and more. That latter thesis, which can be designated here as the ‘Minimization Principle’, consists in the claim that it is always morally permissible for someone to act in a way that will foreseeably bring about fewer violations of some deontological restriction(s) than if he or she does not so act. Unlike the No Contraventions Principle, the Minimization Principle is a thesis which any deontologist would and should reject.

4.2.3.1.  A first example of the conflation Scheffler’s conflation of none-versus-any with fewer-versus-more is first occurrent in his exposition of the maximizing conception of rationality: ‘Views that incorporate agent-centered restrictions . . . seem troubling, relative to this [maximizing] notion of rationality. For they appear to identify certain kinds of actions as morally objectionable or undesirable, in the sense that it is morally preferable that no such actions should occur than that any should, but then tell us that there are situations in which we must act in such a way that a greater rather than a lesser number of these actions are actually performed’ (1988, 252). In the closing portion of this quotation, the auxiliary verb ‘must’ is clearly to be construed as deontic. Hence, the quoted remark is composed of three main claims. First, the opening sentence correctly maintains that agent-centred restrictions are at odds with the maximizing conception of rationality which Scheffler expounds. Second, the initial half of the next sentence rightly asserts that deontologists embrace the No Contraventions Principle. Third, the closing portion of that next sentence accurately indicates that deontologists reject the Minimization Principle. In other words, each of the three main claims in the quotation is unexceptionable. What is puzzling is Scheffler’s assumption—signalled by ‘but’—that deontologists’ embrace of the No Contraventions Principle is in tension with their rejection of the Minimization Principle. His perception of some tension derives from his conflation of the none-versus-any contrast and the fewer-versus-more contrast, for the No Contraventions Principle pertains to the former contrast whereas the Minimization Principle pertains to the latter. Suppose that, instead of adhering to the No Contraventions Principle, deontologists adhered to a variant of that principle focused on the fewer-versus-more contrast. That is, suppose they were committed to the thesis that it is always morally better that a smaller number of violations of deontological prohibitions occur than that a greater number occur. In that event, their position concerning what is always morally better would be in tension with their repudiation of the Minimization Principle. In fact, however, deontologists are not committed to the smaller-versus-greater thesis that has just been broached. They are instead committed to the No Contraventions

Are Deontological Constraints Irrational?

235

Principle, which Scheffler accurately attributes to them at many junctures. Given as much, there is not even a superficial contrariety between their position concerning what is always morally better and their repudiation of the Minimization Principle. Deontologists reject the Minimization Principle because it implies that contraventions of deontological prohibitions are sometimes morally permissible. Their position on that point is straightforwardly consistent with their endorsement of the No Contraventions Principle. After all, any situation in which the Minimization Principle would deem a transgression of a deontological prohibition to be morally permissible for some person P is a situation where the non-scalar ideal encapsulated in the No Contraventions Principle cannot be realized through any mode of conduct available to P. Thus, when deontologists reject the implications of the Minimization Principle in regard to any such situation—and when they thereby insist that P is morally obligated in the specified circumstances to behave in a way that will result overall in more rather than fewer violations of deontological constraints—there is not the slightest tension with their espousal of the No Contraventions Principle. The No Contraventions Principle does not draw a favourable comparison between fewer violations and more; rather, it draws a favourable comparison between no violations and some. It articulates the underlying attitude not of a person who minimizes violations, but of a person who eschews them.

4.2.3.2.  A second example Scheffler again rides roughshod over the distinction between the none-versus-some contrast and the fewer-versus-more contrast, in the following passage: ‘Yet if [deontological] views do regard violations [of deontological restrictions] as morally objectionable or undesirable, in the sense that it is morally preferable that none should occur than that any should, it does then seem paradoxical that they tell us there are times when we must act in such a way that a larger rather than a smaller number of violations actually takes place’ (1988, 254). This passage declares that deontologists are adopting a paradoxical set of positions when they embrace the No Contraventions Principle while gainsaying the Minimization Principle. As should be evident from my response to Scheffler in the preceding subsection, however, the deontologists’ combination of positions is not paradoxical at all. On the contrary, because the No Contraventions Principle encapsulates the underlying attitude of someone who seeks to avoid any violations of deontological constraints—rather than the underlying attitude of someone who seeks to minimize such violations—an endorsement of that principle goes together very smoothly with a denial of the Minimization Principle. Only by eliding the difference between the none-versus-some contrast and the fewer-versus-more contrast has Scheffler presumed to find any incongruity between the positions which the deontologists combine.

4.2.3.3.  Goals for deontologists: a first example Writing more than a decade after Robert Nozick inveighed against any efforts to formulate deontological doctrines in terms of goals, Scheffler nonetheless

236

The Rationality of Deontological Constraints

contends that the proponents of those doctrines cannot overcome the following problem: ‘[V]‌iews incorporating agent-centered restrictions actually do present as desirable [a] goal whose maximum accomplishment they then prohibit. They assign each person the agent-relative goal of not violating any restrictions himself . . . . Thus in forbidding the minimization of overall violations, they are . . . thwarting the achievement of [a] goal whose desirability they recognize’ (1988, 254). He slightly later reaffirms that ‘defenders of standard deontological views do not appear to be in a position to make the claim that, in forbidding us to minimize the violations of those restrictions they insist on, they are not thwarting the achievement of any goal whose desirability they recognize’ (1988, 255). Despite Nozick’s apt animadversions on the use of teleological language in the formulation of deontological constraints, we can accept as harmless the notion that a deontological theory assigns to each person the agent-centred goal of not contravening any deontological restrictions himself. That goal is strictly non-scalar; the attainment of it is an all-or-nothing matter. What is remarkable in Scheffler’s comments is the suggestion that the realization of that non-scalar goal is undermined by deontologists’ rejection of the Minimization Principle. On the contrary, any transgression of a deontological prohibition that is undertaken by someone to lower the overall incidence of transgressions of the prohibition will ensure that he or she fails to achieve the goal of not contravening any deontological restrictions. Only by avoiding all such transgressions can he or she attain that goal. Immanent in the No Contraventions Principle is the overarching ideal of thorough compliance by everyone with deontological prohibitions. Let us designate that agent-neutral ideal—an ideal to which deontologists are genuinely committed—as the ‘Thorough Compliance Objective’. Such an ideal will go unfulfilled in any situation where some person P breaches a deontological restriction DR in order to avert some other violations of DR. The Thorough Compliance Objective does not call for a minimized level of contraventions of DR. Rather, it calls for no contraventions of DR; it is focused on the contrast between none and some rather than on the contrast between fewer and more. Hence, the realization of the Thorough Compliance Objective is stymied whenever P strives to minimize the level of transgressions of DR by engaging in a transgression. Of course, the Thorough Compliance Objective will also go unfulfilled if P refrains from breaching DR and thereby allows some other breaches of DR to occur. My point is not that that agent-neutral ideal will always be given effect by P’s attainment of the agent-centred goal of avoiding any breaches of DR. My point, rather, is that P’s attainment of that latter goal is a necessary condition for the fulfilment of the agent-neutral ideal. Whenever P deviates from his agent-centred goal—even if his aim is to increase the overall level of conformity to DR—he thwarts the realization of the Thorough Compliance Objective. Perhaps, when Scheffler refers to a goal thwarted by deontologists, he has in mind neither the agent-centred goal of avoiding any transgressions of DR nor the agent-neutral Thorough Compliance Objective. Perhaps he instead has in mind the agent-neutral goal of minimizing the number of transgressions of DR. If so, he is correct in maintaining that the achievement of that agent-neutral goal

Are Deontological Constraints Irrational?

237

can sometimes be thwarted by one’s pursuit of the agent-centred goal of not acting at odds with DR. However, we should then be baffled by Scheffler’s assertion that deontologists are somehow committed to recognizing the desirability of the accomplishment of the minimizing goal. Deontologists are not committed in any way to being consequentialists. The only agent-neutral ideal that can accurately be ascribed to them—an ideal repeatedly ascribed to them by Scheffler himself—is the Thorough Compliance Objective. That ideal is not a minimizing goal; it is centred on the contrast between no contraventions and some, rather than on the contrast between fewer contraventions and more. In the first of the two quotations at the outset of this subsection, Scheffler speaks of a ‘goal whose maximum accomplishment [deontological doctrines] then prohibit’. Similarly, he later states that ‘deontology is [committed] to presenting as desirable [a]‌non-relative goal whose maximum accomplishment it then prohibits’ (1988, 254). Given these references to the maximum accomplishment of a goal, and given that the goal of minimizing the number of violations of DR is itself non-scalar, Scheffler here presumably has in mind the goal of lowering the number of the violations. The latter goal is scalar; the maximum accomplishment of it would consist in minimizing the incidence of violations of DR. Again, however, Scheffler is quite mistaken in claiming that deontologists are committed to regarding as desirable a consequentialist goal of the sort just mentioned. Though deontologists will undoubtedly accept that such a goal is desirable within the constraints of any deontological prohibitions, they are certainly not committed to accepting the desirability of such a goal insofar as the attainment of it requires any breaches of those prohibitions. As has been stated, the only agent-neutral ideal which deontologists are committed to regarding as desirable is the Thorough Compliance Objective. That non-scalar ideal consists in no violations of DR rather than in a lowered level of violations.

4.2.3.4.  Goals for deontologists: a second example I have already remarked that the ‘ceteris paribus’ clause in Scheffler’s exposition of the maximizing conception of rationality could be an opening for a reconciliation between that conception and the requirements of deontology. Scheffler seeks to close off that opening, in the following passage where he again attributes to deontologists a teleological outlook (1988, 256–7): [S]‌omeone who wanted to show that there was no conflict between agent-centered restrictions and maximizing rationality might point out that, if the ceteris paribus clause in the formulation of maximizing rationality were fully cashed out, one of its main features would be a provision to the effect that it can sometimes be rational to act in such a way as to worse achieve one goal if that will make it possible to better achieve another. Since that is so, it might be said, views that include agent-centered restrictions need not come into conflict with maximizing rationality when they tell us to further the agent-relative goal of not violating the restrictions ourselves at the expense of the non-relative goal of minimizing violations of the restrictions. By itself, however, this claim is not fully persuasive. The problem is that the agent-relative goal and the non-relative goal appear to be related to each

238

The Rationality of Deontological Constraints

other in such a way as to make the insistence on giving priority to the relative goal puzzling, from the standpoint of maximizing rationality. Since, as our earlier discussion suggested, the fact that violations of the restrictions are objectionable from a moral point of view constitutes at least part of the basis for claiming that individual agents ought not ordinarily to commit such violations, the agent-relative goal looks as if it is derivative from, and given life by, the non-relative objection, and does not appear to represent something independently desirable. Rather, the desirability of achieving the agent-relative goal seems contingent on its serving to advance the non-relative goal of minimizing the morally objectionable. And if that is so, then the insistence that one must satisfy the agent-relative goal even when doing so will inhibit achievement of the non-relative goal is incompatible with considerations of maximization.

In many respects, the most revealing word used here by Scheffler is ‘ordinarily’ in the antepenultimate sentence of this passage. Whereas consequentialists maintain that individual agents ought not ordinarily to transgress deontological restrictions—because conformity to those restrictions is ordinarily promotive of good consequences— deontological absolutists insist that individual agents ought not to transgress deontological restrictions, tout court. There is no ‘ordinarily’ qualification in what deon­tological absolutists claim about the requisiteness of abiding by deontological ­constraints. Once the unqualified nature of any deontological constraints is duly recognized, they do not at all appear to be derivative of the goal of minimizing the frequency with which those constraints are violated. Rather, they appear to be what they are: namely, limitations on the extent to which (or the ways in which) that agent-neutral goal of minimization can legitimately be pursued. In the antepenultimate sentence of the passage quoted above, Scheffler refers to one of the earlier discussions in his article. That earlier discussion, which I have quoted in § 4.2.2, correctly contends that deontologists are committed to the No Contraventions Principle. Because the No Contraventions Principle is centred on the distinction between none and some rather than on the distinction between fewer and more—in other words, because it is an absolutist principle rather than a minimizing principle—the commitment of deontologists to it scarcely renders puzzling their insistence that agent-centred prohibitions impose moral obligations even when the fulfilment of those obligations will increase the incidence of violations of the prohibitions. Of course, an insistence on that point is at odds with the Minimization Principle. Were deontologists committed to that latter principle, as Scheffler clearly appears to be assuming in the extract quoted in this subsection, they would indeed be muddled. However, as this chapter has sought to underscore, deontologists squarely repudiate the Minimization Principle. Scheffler himself has intimated as much in the earlier portion of his article, by correctly attributing to deontologists an embrace of the No Contraventions Principle.

4.2.4.  The maximizing conception of rationality redux Scheffler fails in his efforts to show that deontologists have espoused some positions that are in tension with each other, and specifically he fails in his efforts to show that deontologists are committed to any consequentialist precepts such as

Are Deontological Constraints Irrational?

239

the Minimization Principle. However, his central worry about the inconsistency between deontological doctrines and the maximizing conception of rationality might seem to have withstood my ripostes. While granting that deontologists are not mired in the confusion of which Scheffler accuses them, a defender of Scheffler might nonetheless contend that deontological doctrines are deeply problematic because they clash with a mode of practical rationality that pervades our everyday lives. Such a defender might feel that I have not yet parried Scheffler’s critique on that vital point. Two main replies to such a line of defence are appropriate; the present subsection will advance one of those replies, and § 4.2.5 will advance the other. First, then, we need to pin down what the maximizing conception of rationality is. In particular, we need to pin down the import of the ‘ceteris paribus’ clause in Scheffler’s exposition of that conception of practical rationality. On the one hand, if the ‘ceteris paribus’ clause indicates that any maximizing quest is morally impermissible when such a quest contravenes some deontological restriction(s), then the maximizing mode of rationality does indeed pervade our everyday lives. However, when the ‘ceteris paribus’ clause is so construed, and when the permissibility of maximization is thus hemmed in by deontological constraints, the maximizing conception of rationality is fully consistent with any deontological doctrine. Contrary to what Scheffler proclaims, there is no tension whatsoever between the morality of common sense and the practical rationality of common sense. On the other hand, if the ‘ceteris paribus’ clause does not incorporate any deontological constraints into the maximizing conception of rationality, and if the proponents of that conception therefore hold that the permissibility of maximizing quests is not bounded by any deontological restrictions, then it is far from clear—to say the least—that the maximizing conception of rationality does encapsulate a mode of reasoning that is widely endorsed in our everyday practices. Scheffler does not offer one whit of evidence to support the view that an unrestricted approach to maximization is a mode of reasoning that permeates those practices. Moreover, at the outset of his article he concedes that ‘the common-sense morality of our culture is substantially deontological in content’. He subsequently refers to deontological constraints as ‘the morality of common sense’, and he suggests that ‘agent-centered restrictions are congenial to the common-sense morality of our culture’ and that ‘the restrictions thus embody constraints on practical reasoning that seem to us natural and intuitively appealing’ (1988, 243, 251, 259). Although he then goes on to assert that the unrestricted approach to maximization is likewise ‘a very powerful form of thought which itself occupies a central place within what we recognize as human practical rationality’ and that it is ‘a process of reasoning that itself seems natural and intuitively compelling’ (1988, 259), he provides no substantiating evidence for these claims about unconstrained maximization. Hardly in need of confirmatory evidence is the notion that a maximizing orientation within the limits of deontological restrictions is a form of practical rationality that suffuses our everyday practices. Far more dubious is the notion that those practices are similarly suffused by a maximizing orientation without any such

240

The Rationality of Deontological Constraints

limits. Unless people in Western societies are so obtuse as to participate frequently in practices where they forget how congenial agent-centred restrictions are to the common-sense morality of their culture, they are not likely to embrace an unrestricted approach to maximization in a wide range of everyday contexts. In other words, whereas Scheffler contends that the morality of common sense clashes with the rationality of common sense, the real clash lies between the common-sense morality and his misrepresentation of the common-sense rationality. If his exposition of the rationality of common sense is to be credible, its ‘ceteris paribus’ clause has to be construed as indicating that the permissibility of any maximizing quests is bounded by agent-centred restrictions.

4.2.5.  A return to moral optimality Though interrogational torture and most other types of torture are always and everywhere morally wrong in all possible worlds, some possible instances of interrogational torture are morally optimal. Those instances would occur under truly extreme circumstances where the consequentialist considerations in favour of the use of torture are redoubtably potent and urgent. Notwithstanding that the perpetration of interrogational torture is morally wrong even in those circumstances, the gravity of the wrongness is so greatly extenuated that the resort to torturous techniques of interrogation is morally optimal. In a dire emergency where interrogational torture is likely to be efficacious in eliciting information that can avert a calamity—and where all the available non-torturous methods of ascertaining the information have failed—the use of some torturous techniques can be less grave as a breach of a moral duty than the wholesale eschewal of those techniques. In sum, despite my insistence on the absoluteness of the moral impermissibility of interrogational torture, I  have not lost sight of the force of consequentialist pressures. Although those pressures can never suspend the moral prohibition on such torture, they can substantially mitigate the gravity of transgressions of that prohibition. Their mitigating influence can sometimes (albeit very rarely) render such a transgression morally optimal as the least terrible course of conduct that is feasible in some dreadful predicament. Hence, although the maximizing conception of rationality has to encompass deontological restrictions to be credible as an encapsulation of a mode of reasoning that pervades everyday contexts, the maximizing impetus that is at the core of the conception does not disappear even when those restrictions are applicable. Its force is felt not at the basic level of the impermissibility of any conduct that contravenes those restrictions, but instead at the more fine-grained level of the gravity of such conduct. While deontological constraints do significantly limit the range of circumstances in which any maximizing quests can legitimately be pursued, there is not a simple clash between those quests and those constraints. One reason for rejecting Scheffler’s critique of deontology is that it posits a simple clash in which the demands of deontology have to give way if the demands of a maximizing orientation are to be duly acknowledged. By highlighting the ways in which consequentialist pressures can mitigate the gravity of transgressions of deontological prohibitions, and by

Are Deontological Constraints Irrational?

241

indicating that those pressures can even render such transgressions morally optimal in certain extreme situations, this book acknowledges the force of the consequentialist factors without retreating at all from its insistence on the absoluteness of the deontological prohibitions. Through attentiveness to the occurrence of moral conflicts and to the intricacies of the dynamics in those conflicts, we can discern that deontological doctrines are far more resourceful than Scheffler allows. Proponents of those doctrines can register the weightiness of consequentialist considerations—and can register the power of the maximizing conception of rationality that underpins those considerations—without flinching from the strictness of deontology.

5 Legal Responses to Torture This book has been concerned with the nature and moral status of torture, and also with some fundamental aspects of morality. In this closing chapter, we shall ponder the appropriate legal responses to the perpetration of torture. Because so many aspects of law are variable across jurisdictions (even if the range of jurisdictions under consideration is confined to liberal democracies), anyone addressing this chapter’s topic at a philosophical level of abstraction should be circumspect. At that level of abstraction, there are significant limits on what can usefully be said about the proper legal regulation of torture. Many practical difficulties—concerning the likelihood of official mendacity, for example—are better handled by experts in juridical procedures than by philosophers. Still, despite the need for modesty in any philosophical treatment of this chapter’s topic, an investigation of the nature and wrongness of torture would be seriously incomplete if it offered no general reflections on how the law should deal with the problem. Of course, Chapters 2 and 3 have already made some scattered remarks about the legal regulation of torture. Here, however, we shall focus much more sustainedly on some of the questions that have been broached there in passing. Nevertheless, although this chapter will be concentrating on legal issues, its ruminations are firmly grounded on the previous chapters’ enquiries into the moral bearings of torture. Time and again, my recommendations here will advert to arguments propounded and conclusions reached in the preceding chapters. Hence, notwithstanding that we are turning our attention from the nature of morality to matters of legal remedies, the present chapter is a natural offshoot or culmination of the earlier chapters. At a philosophical level of abstraction, which prescinds from the nuts-and-bolts questions that have to be answered by lawyers and social scientists, my aim is to outline how the law in a liberal-democratic country—and in the international legal community—can best track the morality of torture. In the contemporary debates over the legal regulation of torture, interrogational torture draws far more scrutiny than any other kind. Accordingly, this final chapter will devote attention principally to the use of torture in interrogational settings. However, it will not altogether disregard some of the other types of torture that have been explored in previous chapters. One of the shortcomings in the contemporary debates is precisely the inadequacy of the attention paid to those other types of torture. Though this chapter’s remarks on any of the non-interrogational varieties of torture will generally be terse, they will help to indicate that the ways

Legal Approval Ex Ante?

243

in which the law should come to grips with the perpetration of torture are more intricately nuanced than is often presumed.

5.1.  Legal Approval Ex Ante? One major set of issues pertaining to the legal regulation of torture is centred on the question whether the use of torture should ever be legally authorized ex ante. Though punitive torture is no longer officially practised in any liberal democracy, the matter of interrogational torture is still a live point of contention. One of the most widely discussed and frequently condemned proposals in recent years, floated chiefly by Alan Dershowitz, has submitted that certain torturous techniques of interrogation should indeed be legally authorized—provided that warrants for the plying of such techniques are sought and obtained beforehand by the relevant officials. Dershowitz’s torture-warrant proposal has been trenchantly criticized by quite a few of the other philosophers and legal theorists who write on these matters,1 but we shall examine it at some length here. We shall then turn to a proposal advanced by Eric Posner and Adrian Vermeule, who maintain that the use of torture for interrogation should be allowed and regulated in broadly the same manner as the use of deadly force by the police for the apprehension of suspects. After this chapter has taken issue with Posner’s and Vermeule’s advocacy of the institutionalization of torture, it will consider whether torturous techniques of interrogation should be legally differentiated ex ante from other interrogational techniques. Jeremy Waldron has argued that any such differentiation is misconceived; I will contend that Waldron’s arguments on that point are overbroad.

5.1.1.  Dershowitz and torture warrants Dershowitz is often simplistically denounced as a proponent of the use of torture for interrogational purposes, but his main concern is in fact to eliminate the hypocrisy that surrounds the matter.2 He believes that legal-governmental officials—even in liberal democracies—have availed themselves of torture and will continue to 1   For some of the previous critiques of Dershowitz’s proposal, see Addicott 2003–4, 905–8; Allhoff 2012, 180–5, 193–4; Arrigo 2004, 558; Bobbitt 2008, 382–3; Brecher 2007; Bufacchi and Arrigo 2006, 358–9, 365–6; Card 2010, 177–9, 186–92; Cohan 2007, 1602–3; Elshtain 2004, 83; Frowe 2011, 200–1; Ginbar 2008, chap. 13; Gross 2004a; Ip 2009, 41–2, 52–4; Kreimer 2003; La Torre 2009, 26–7; Meisels 2008a, 159–62; Meisels 2008b, 209–13; Miller 2005, 189–90; Miller 2011, § 4; Parry and White 2002, 745, 747; Posner 2004, 295–7; Raviv 2004, 161–75; Scarry 2004; Scheppele 2005, 308–9; Soniewicka 2013, 197–9, 210–11; Steinhoff 2006; Strauss 2003, 271–3; Sung 2003; Van der Vyver 2003, 454–8; Waldron 2010a, 217–21, 250–1; Wisnewski 2008b; Wisnewski 2010, 93, 162–3, 172–3; Wisnewski and Emerick 2009, 40–54. For some wary defences of Dershowitz, see Levinson 2003, 2044–6, 2048–50; Levinson 2004, 37–8; Posner and Vermeule 2006, 699, 702; Posner and Vermeule 2007, 208–9, 212; Seidman 2005, 884–5. For some more enthusiastic defenses, see Bagaric and Clarke 2005, 614–16; Moher 2004. 2   Dershowitz has written and lectured voluminously in support of his torture-warrant proposal. The main writings of his on which I have drawn are Dershowitz 2002, chap. 4; 2003; 2004.

244

Legal Responses to Torture

avail themselves of torture in extreme situations. His frequent invocations of ticking-bomb scenarios are aimed not at supporting the use of torture in dire emergencies, but at underscoring why officials responsible for public safety (even in liberal democracies) will continue to be attracted to the use of torture in such emergencies. Dershowitz submits that, given the allure of torturous techniques of interrogation in the eyes of officials who are striving to avert calamities in desperate situations of terrorism or criminality, the wielding of those techniques should be subjected to overt legal regulation. Instead of pretending that torture does not occur and will not ever occur, we should acknowledge its actuality by bringing it tightly under the control of the law. So Dershowitz contends, as he writes: ‘We cannot evade our responsibility by pretending that torture is not being used or by having [governments in other nations] use it for our benefit’ (2002, 139). Dershowitz thinks that, if legal-governmental officials who wish to resort to torture are required to obtain warrants from judges or magistrates before they go ahead, two salutary consequences will follow. First, he maintains, the overall incidence of interrogational torture will very likely diminish. Second, the wishful ignorance of the citizenry and the mendacity of legal-governmental officials will be largely overcome, as the use of torture by those officials in extreme interrogational contexts will be faced squarely. For Dershowitz, that second point is indeed the heart of his proposal: If torture is in fact being used and/or would in fact be used in an actual ticking bomb mass terrorism case, would it be normatively better or worse to have such torture regulated by some kind of warrant, with accountability, record-keeping, standards, and limitations. This is an important debate, and a different one from the old, abstract Benthamite debate over whether torture can ever be justified. It is not so much about the substantive issue of torture, as it is over accountability, visibility, and candor in a democracy that is confronting a choice of evils. (2003, 277–8, footnote omitted, emphases in original)

As Waldron, a harsh critic of Dershowitz, has observed: ‘Dershowitz really is concerned that the alternative to his proposal is not “no torture” but torture conducted sub rosa, torture conducted beneath legal notice and with law’s complicity or silence’ (2010, 250–1). Snide dismissals of Dershowitz as a modern-day Torquemada are glibly misconceived. Although this chapter will assail his torture-warrant proposal as a badly misguided idea, the proposal is well-intentioned, and the problems of public hypocrisy and official mendacity which it confronts are pressingly real. Dershowitz is undoubtedly sincere when he proclaims that ‘my [torture-warrant] argument is not in favor of torture of any sort. It is against all forms of torture without accountability. Let us continue to reaffirm not only our opposition to torture but our opposition to the kind of hypocrisy that loudly denounces torture while discreetly closing our eyes to its increasing use’ (2004, 275–6). Albeit Dershowitz’s torture-warrant proposal does in some respects hark back to the era of judicially supervised torture in late medieval and early modern Europe, there are also a number of major differences between his envisaged scheme and the bygone system of torture. Most notable is the difference of purpose. Torture was employed by the judiciary in much of medieval Europe to elicit confessions which—if

Legal Approval Ex Ante?

245

subsequently reaffirmed in court—would be admissible against the defendants who had succumbed to the torture. By contrast, Dershowitz emphatically denies that confessions or other evidence obtained through torture can ever properly be admissible in any judicial proceedings. He holds that the role of the interrogation countenanced under his scheme of torture warrants is solely prospective. Warrants can be issued under that scheme for the purpose of extracting information that will help to avert calamities or to foil future plots or to disrupt terrorist networks. No information so extracted will be admissible as evidence in any adjudicative forum. Before we turn to my objections to Dershowitz’s arguments, we should note an important caveat that has similarly been issued by most of the other main critics of his proposal (who have also anticipated some of my chief objections to his ideas). Whereas my foremost criticism of Dershowitz will pertain to a matter of moral principle, my initial few criticisms will concentrate on his empirical hypotheses. This chapter’s rejoinders to those hypotheses will not purport to be dispositive, for this book is not a work of social science. I have not amassed any reams of data to rebut his empirical stance. Instead of marshalling any data to disconfirm the empirical conjectures which Dershowitz has propounded, I  will point to some egregious weaknesses in the arguments with which he tries to bolster those conjectures. Well-intentioned though his work on torture warrants is—and well-founded though his worries about the problems of official mendacity and public hypocrisy are—he attempts to vindicate his proposal with some disconcertingly feeble lines of reasoning. Though my exposure of the shakiness of his lines of reasoning will not conclusively falsify the empirical conclusions for which he argues, it will reveal that those conclusions are unsubstantiated in his writings.

5.1.1.1. A first objection to Dershowitz: inapposite comparisons Dershowitz draws some strange comparisons in his efforts to sustain his contention that a system of torture warrants would reduce the incidence of interrogational torture. He adduces two examples which he believes to be supportive of that contention: [T]‌wo examples demonstrate why I think there would be less torture with a warrant requirement than without one. First, recall the case of the alleged national security wiretap being placed on the phones of Martin Luther King by the Kennedy administration in the early 1960s. This was in the days when the attorney general could authorize a national security wiretap without a warrant. Today no judge would issue a warrant in a case as flimsy as that one. Indeed, no law enforcement agent would even request one. Second, when Zaccarias Moussaui was detained after trying to learn how to fly an airplane, without wanting to know much about landing it, the government did not even seek a national security wiretap because its lawyers believed that a judge would not have granted one. If Moussaui’s computer could have been searched without a warrant, it almost certainly would have been.3

3   Dershowitz 2003, 281–2, footnotes omitted. Nearly identical passages are in Dershowitz 2002, 159–60; and 2004, 270–1.

246

Legal Responses to Torture

The opening and closing sentences of this quotation make clear the unsuitability of the examples which Dershowitz invokes here. In the first example, he recounts a situation in which national-security wiretapping can be lawfully pursued without the obtention of a warrant, and he then contrasts that state of affairs with a situation in which such wiretapping can be lawfully pursued only if a warrant has been obtained. Hardly surprising is his conclusion that wiretapping is more likely to go ahead in the former situation than in the latter, but that glaringly obvious conclusion is of no avail whatsoever to Dershowitz in his disputation with the opponents of his proposal. After all, those opponents are scarcely suggesting that the preferred alternative to Dershowitz’s scheme of torture warrants is a legal system in which interrogational torture can be perpetrated lawfully without any warrants. Rather, they are insisting that the preferred alternative is a legal system in which the perpetration of interrogational torture is not legally permitted at all. Hence, Dershowitz needs to compare a liberal-democratic legal system of that latter type with a liberal-democratic legal system in which interrogational torture can lawfully be performed whenever a warrant has been properly obtained. He does not shed the slightest light on the probable outcome of such a comparison, when he invokes the example of the wiretaps placed on the telephones of Martin Luther King. Dershowitz fares no better with his example of Zacarias Moussaoui. Although the baseline for his comparison is not made explicit until the closing sentence of the quotation,4 he is plainly drawing a contrast between a situation in which wiretapping is legally permitted with warrants (Situation A) and a situation in which wiretapping is legally permitted without any warrants (Situation B). Again the outcome of such a comparison is hardly surprising, but again the comparison itself is beside the point. Nobody doubts that, ceteris paribus, the incidence of wiretapping will tend to be lower in Situation A than in Situation B. What Dershowitz’s opponents do doubt is that the incidence of wiretapping will tend to be lower in the former situation than in a context where the practice of wiretapping is legally forbidden altogether. Dershowitz should be posing a comparison between a comprehensive legal prohibition and a scheme of warrants, yet his example of Moussaoui does not involve such a comparison. (Suppose that Dershowitz were to amend the example by refocusing it on such a comparison. He will then have to assume implausibly that the officials who refrained from seeking a national-security warrant would have gone ahead with an unlawful wiretap if and only if there had been no provision for the issuance of any warrants under a blanket legal ban on the practice of wiretapping. An assumption along those lines is not self-contradictory, but it is very far from compelling. If Dershowitz were to present his example of Moussaoui on the basis of such a far-fetched assumption without any supportive data or argumentation, the example would be worthless for his purposes. It would certainly not enhance the cogency of his empirical claims.)

4   The auxiliary verb ‘could’ in that closing sentence is clearly to be construed as deontic. In other words, it is the equivalent of ‘could lawfully’.

Legal Approval Ex Ante?

247

5.1.1.2.  A second objection to Dershowitz: a missing prohibition Dershowitz makes largely the same error when he later ponders the rationale for requiring the police to obtain warrants before they search any private premises. He writes that ‘the policies underlying the warrant requirement [for searches by the police] are relevant to the question whether there is likely to be more torture or less if the decision is left entirely to field officers, or if a judicial officer has to approve a request for a torture warrant. As Abraham Maslow once observed, to a man with a hammer, everything looks like a nail. If the man with the hammer must get judicial approval before he can use it, he will probably use it less often and more carefully’ (2002, 160–1). In the closing sentence of this quotation, the auxiliary verb ‘can’ is clearly to be construed as deontic; it denotes legal permissibility. Dershowitz is here drawing a contrast between a situation in which the lawfulness of interrogational torture by constables or intelligence officials is unconditional and a situation in which the lawfulness of such torture is conditional on the obtention of warrants. Unsurprisingly, he predicts that more torture will occur in the former situation than in the latter. Such a prediction is of course accurate but is also irrelevant and thus obfuscatory. Resistance to Dershowitz’s torture-warrant proposal does not come from people who wish to see the use of torture by officials legally unrestricted. It comes from people who wish to preserve a strict legal prohibition on the use of interrogational torture. Hence, Dershowitz should be comparing a situation in which acts of torture are strictly forbidden legally and a situation in which acts of torture by officials are legally permitted through the obtention of warrants. Such a comparison gets left aside in his remark on the hammer and nails. Much the same misstep mars the following line of reasoning by Dershowitz: ‘[A]‌ double check is always more protective than a single check. In every instance in which a warrant is requested, a field officer has already decided that torture is justified and, in the absence of a warrant requirement, would simply proceed with the torture. Requiring that decision to be approved by a judicial officer will result in fewer instances of torture even if the judge rarely turns down a request’ (2002, 158). If this remark by Dershowitz is to be relevant, then it has to be construed as posing a comparison between what will happen under a strict legal prohibition on the use of interrogational torture and what will happen under a system of torture warrants. Dershowitz is therefore assuming that, under a blanket legal prohibition on the use of interrogational torture, officials will ‘simply proceed with the torture’ whenever they deem it to be appropriate. He is further assuming that those same officials under a system of torture warrants will seek warrants whenever they deem torture to be appropriate, and that they will refrain from going ahead with the torture if their applications for warrants are rejected. He provides no evidence for these extraordinarily dubious empirical suppositions. Instead, in the opening sentence of the quotation, he dresses those suppositions up as a conceptual thesis. What most likely underlies Dershowitz’s apparent misrepresentation of some outlandish empirical speculations as a conceptual thesis is that the implicit point of comparison in his remark is not in fact a society with a strict legal prohibition on the use of interrogational torture. Rather, his implicit point of comparison is a

248

Legal Responses to Torture

society in which there is neither a legal prohibition on the use of interrogational torture nor a torture-warrant requirement. Everyone can agree that the incidence of torturous interrogations in a society of that latter kind will almost certainly exceed the incidence of such interrogations in a society with a torture-warrant requirement. However, that unexcitingly obvious conclusion does not help Dershowitz at all—because nobody among his opponents has ever contested that conclusion. His opponents have instead sought to contest the proposition that the incidence of interrogational torture will tend to be higher in a society with a strict legal prohibition on such torture than in a society that authorizes such torture through the issuance of warrants. Dershowitz’s remark about the greater protectiveness of a double check scarcely vindicates the proposition which his opponents have actually impugned; as a putative vindication of that proposition, his remark has to rely on the implausible and unsubstantiated empirical claims which I have recounted in the preceding paragraph.

5.1.1.3. A third objection to Dershowitz: inordinate narrowing of the options Dershowitz suggests that the competing considerations which relate to the use of interrogational torture have confronted us with ‘a triangular conflict unique to democratic societies’. He elaborates: If these horrible practices [of interrogational torture] continue to operate below the radar screen of accountability, there is no legitimation, but there is continuing and ever-expanding sub rosa employment of the practice. If we try to control the practice by demanding some kind of accountability [through a system of warrants], we add a degree of legitimation to it while perhaps reducing its frequency and severity. If we do nothing, and a preventable act of nuclear terrorism occurs, then the public will demand that we constrain liberty even more. There is no easy answer. (2003, 278; 2004, 267)

Most problematic in this passage is its penultimate sentence. What does the phrase ‘If we do nothing’ mean? Should it be construed as ‘If we do not alter the current practices of interrogation’ or ‘If we cease to combat terrorism’ or ‘If we cease to employ interrogational torture’? If the first of these three interpretations were exegetically correct, then the final horn of Dershowitz’s trilemma would be essentially the same as the opening horn. If the second of the three proffered interpretations were exegetically correct, then the final horn of Dershowitz’s trilemma would be ridiculous; nobody has ever suggested that the preferred alternative to the use of interrogational torture is the discontinuation of efforts to thwart terrorism. Accordingly, we can quite safely infer that ‘If we do nothing’ is Dershowitz’s extravagantly misleading way of saying ‘If we cease to employ interrogational torture’. One problem for Dershowitz is that many of the opponents of his torture-warrant proposal—such as the adherents of the Legal Prohibition Thesis, which I have discussed in § 3.1.3.3.2 of Chapter  3—believe that interrogational torture can be morally permissible as a last resort in extreme emergencies. While they strongly

Legal Approval Ex Ante?

249

disfavour any ex ante legal authorization of interrogational torture, they do not think that the perpetration of such torture is morally wrong in the direst of predicaments where all other available methods of coming up with calamity-averting information have failed. Dershowitz will reply as he does reply to Richard Posner, one of the premier exponents of the Legal Prohibition Thesis (Dershowitz 2003, 279–81; 2004, 267–9). That is, Dershowitz will maintain that the supporters of that thesis run aground on the first horn of his trilemma. He will insist that their position, which recommends the ex post exoneration of officials who have managed to avert calamities by resorting to interrogational torture in dreadful emergencies, is in effect an endorsement of the current practices of subterranean and unaccountable torture. 5.1.1.3.1.  A shortcoming in Dershowitz’s reply Although I shall later criticize the Legal Prohibition Thesis, Dershowitz’s rejoinder to it—in the form of his riposte to Posner—is simplistic at best. To be sure, when Posner writes that the strict legal prohibitions on torture should be retained ‘with the understanding that of course they will not be enforced in extreme circumstances’ (2004, 296), he does leave himself somewhat vulnerable to Dershowitz’s charge that he is conniving at the perpetration of torture without accountability. However, whether or not that charge is to some degree accurate against Posner, it does not tell against quite a few other proponents of the Legal Prohibition Thesis. They are scarcely calling for the unaccountability of officials who resort to torture in extreme emergencies. Albeit they believe that the actions of the officials are in some cases covered by legal justifications or excuses and are therefore not properly subject to legal sanctions, the applicability of any such justifications or excuses is to be ascertained ex post in adjudicative proceedings that will hold the officials accountable for what they have done. Proponents of the Legal Prohibition Thesis set themselves against Dershowitz’s torture-warrant scheme partly because the issuance of warrants is so likely to have the effect of shielding officials from accountability after the fact. As Elaine Scarry observes: ‘[S]‌hould [we] understand the warrant as a temporary grant of permission that is, upon review, subject to retrospective revocation, at which point the torturer’s exemption from punishment would dissolve? Long experience with search warrants suggests the opposite: search warrants, far from facilitating review, historically have tended to close the door on review’ (Scarry 2004, 287). Advocates of the Legal Prohibition Thesis contend that, if any legal-governmental officials do employ interrogational torture in desperate circumstances, they should have to justify their actions in criminal or civil proceedings. Those advocates believe that the retrospective scrutiny undertaken in such proceedings is a more exacting form of accountability than is the process of applying for a torture warrant. The latter process, occurring under severe constraints of time, is all too likely to eventuate in perfunctory approval on nearly every occasion.5 5   Scarry, writing in the early 2000s, remarks that ‘[t]‌he court set up to issue warrants under the Foreign Intelligence Surveillance Act (FISA) has declined only one requested warrant in twenty-five years: the estimated number of warrant requests is twenty-five thousand’ (2004, 286).

250

Legal Responses to Torture

5.1.1.3.2.  A possible response by Dershowitz Of course, the belief just ascribed to the supporters of the Legal Prohibition Thesis is an empirical claim and is thus open to rebuttal on empirical grounds. Dershowitz can challenge the notion that the processes for considering torture-warrant applications will be largely ineffective as a filter, or he can challenge the notion that ex post legal proceedings will be a rigorous form of scrutiny. Given his work as a criminal-defence attorney who is well aware of the cursoriness with which search-warrant applications are frequently approved, he would not be well advised to mount the former challenge. Only in the following comment does he attempt such a tack: ‘The argument that judges today hand out national security warrants in every instance in which one is sought ignores the fact that the warrant requirement may reduce the number of instances in which agents seek a warrant’ (2003, 282 n34). The wording of this comment by Dershowitz is absurd, since there will obviously not be any applications for warrants if there is no warrant requirement. However, what he has presumably endeavoured here to say is that intelligence officials will tend to refrain from seeking a warrant whenever they anticipate that an application would not be ratified. Such a conjecture by Dershowitz might be correct, though he does not adduce any evidence in support of it. However, even if his conjecture is correct, it will be of little pertinence unless two other points are also true. (I will state those two points here with reference to the use of torture rather than to the use of surveillance, because the former activity is of chief interest to Dershowitz and me.) First, in the circumstances in which the officials refrain from seeking warrants, they will not resort to torture without any warranted authorization. Second, in circumstances like those in which the officials refrain from seeking warrants, they would resort to unlawful torture in the presence of a blanket legal ban of the sort favoured by the votaries of the Legal Prohibition Thesis. Unless both of those two points are true, the quoted conjecture by Dershowitz—even if true— will not be sufficient to establish that a torture-warrant requirement can lower the incidence of interrogational torture effectively. Now, although those two points and Dershowitz’s conjecture might all be true, their joint truth is hardly evident. Dershowitz has not even begun to substantiate any of them. 5.1.1.3.3.  A second possible response by Dershowitz Dershowitz will probably be inclined to opt instead for the other approach which I have mentioned at the outset of the preceding subsection. That is, he will be inclined to contest the proposition that ex post legal proceedings constitute a rigorous mode of accountability for officials who have breached a strict legal prohibition on the use of interrogational torture. He does indeed clearly take issue with that proposition, in more than one way. For example, he raises the problem of chicanery on the part of officials who have resorted to torturous techniques of interrogation:  ‘The real issue, therefore, is not whether some torture would or would not be used in the ticking bomb case—it would. The question is whether it would be done openly, pursuant to a previously established legal procedure, or whether it would be done secretly, in violation of existing law’ (2002, 151).

Legal Approval Ex Ante?

251

Mendacious furtiveness is one means by which legal-governmental officials can escape any genuine accountability for their recourse to interrogational torture. Another means is jury nullification—or, in the absence of a jury, judicial nullification. That is, a jury or judge can acquit officials who are brought to trial for having manifestly employed interrogational torture in the circumstances of a terrible emergency. Through such acquittals, juries or judges can decline to give effect to a strict legal prohibition on interrogational torture. Formal processes for holding officials accountable are thus substantively eviscerated by the connivance of juries or judges. Moreover, even the activation of a formal process can be stymied by the exercise of prosecutorial discretion, which is yet another route of escape from meaningful accountability for officials who have availed themselves of interrogational torture. Executive pardons can be a further such route. Dershowitz as a canny criminal-defence lawyer is alert to these ways in which any officials who perpetrate torture can be shielded from legal responsibility for their actions. In one of his most caustic retorts to Scarry, he invokes the spectre of prosecutorial discretion and jury nullification: What would Professor Scarry have us do instead [of introducing a system of torture warrants]? She would want torture to be used if it could save multiple lives, but she would leave the initial decision to the ex ante decision of ‘the torturers’ and then leaves the post facto decision about whether the torturer did the right thing to ‘a jury of peers.’ This is extraordinarily naïve, as anyone with any experience in criminal justice will understand. No prosecutor would prosecute and no jury would convict if it turned out that the torturer was right, even if the basis on which he acted was weak or bigoted. (2004, 274, emphasis in original)

In a broadly similar vein, Dershowitz adverts to the case of Leon v. Wainwright (which I have discussed for other purposes at several junctures in Chapters 2 and 3). No charges were brought against the policemen who extracted the life-saving information from the captured kidnapper through torture in that case.6 Moreover, although the judges on the Eleventh Circuit Court of Appeals indicated that they did not wish to condone the use of violent techniques of interrogation by the police, they asserted that ‘this case does not represent the typical case of unjustified force. We did not have an act of brutal law enforcement agents trying to obtain a confession in total disregard of the law. This was instead a group of concerned officers acting in a reasonable manner to obtain information they needed in order to protect another individual from bodily harm or death’ (Leon v.  Wainwright 1984, 773). Dershowitz remarks: ‘If an appellate court would so regard the use of police brutality—torture—in a case involving one kidnap victim, it is not difficult to extrapolate to a situation in which hundreds or thousands of lives might hang in the balance’ (2002, 253 n31; 2004, 280 n26). He subsequently adds that ‘no FBI agent who tortured a suspect into disclosing information that prevented an act of mass terrorism would be prosecuted—as the policemen who tortured the kidnapper into disclosing the whereabouts of his victim were not prosecuted’ (2002, 162). 6   Instead of involving any criminal charges against the policemen, the case addressed the question whether certain statements made by the kidnapper could be admitted as evidence against him.

252

Legal Responses to Torture

Worries about the means by which officials can escape ex post accountability are presumably what lead Dershowitz to postulate a tendentious dichotomy in his formulations of the topic on which he and his opponents are concentrating. He repeatedly professes to be engaged in a ‘debate over whether, if torture would in fact be used in a ticking bomb case, it would be worse to make it part of the legal system, or worse to have it done off the books and below the radar screen’ (2002, 156). As has already been stated, most of his opponents do not view the debate in these terms. When they reject his proposal for the ex ante authorization of interrogational torture, they are not submitting that such torture should go ahead ‘off the books and below the radar screen’. Rather, they are contending that any officials who perpetrate such torture should be held to account ex post in criminal or civil or institutional proceedings. Dershowitz presumably believes that such proceedings will not take place at all—because mendacious officials will conceal what they have done, or because prosecutors will elect not to press charges—or that any proceedings which do take place will be hollowly ineffective because of jury nullification. That underlying belief is presumably what informs his tendentious characterizations of the disputes between himself and his critics. (Of course, a few of Dershowitz’s critics do think that interrogational torture should be conducted ‘below the radar screen’. Dershowitz refers to them over and over. Nonetheless, such a view is not endorsed by most of the people who condemn his envisaged system of torture warrants.) Concerns about the efficacy of ex post accountability have likewise troubled Sanford Levinson, one of the few jurists or philosophers to support Dershowitz’s torture-warrant proposal. He has been led to his uneasy embrace of that proposal largely because of his lack of confidence in juries and judges as enforcers of prohibitions on torture. He writes: [V]‌ictims of torture will often indeed be unattractive, across one or another dimension, and the alleged torturers will often be able sincerely to argue—in front of jurors or judges far more inclined to identify with them than with the tortured—that they believed they were acting to protect society. Can anyone genuinely imagine, for example, that an alleged terrorist claiming to have been tortured by or with the knowledge of American officials could win a jury trial for damages in the United States today, even assuming that one could surmount various immunity defenses that might prevent ever getting to the jury? If one believes in the probability of what might be termed the underenforcement of the norm against torture—and any other view is willfully naïve—then we should at least consider, and not merely dismiss, Dershowitz’s suggestion that we look at torture before it occurs, so to speak, when the state pleads for a warrant, rather than from the after-the-fact perspective when the torturer, even assuming that he is charged with a criminal offense or sued by his victim, might be all too able to persuade a judge or jury of his ‘acceptable’ motives.7

7   Levinson 2004, 36–7. Much of the quoted passage also appears in Levinson 2003, 2048. For some other remarks on jury nullification and kindred matters, see Bobbitt 2008, 383–4; Brunnée and Toope 2010, 233–4; Chesterman 2008, 317–18; Davis 2005, 173; Gross 2008, 65–9, 73–4, 81–5; Koh 2005, 659–60; Posner and Vermeule 2006, 694–5, 703; Posner and Vermeule 2007, 172–3, 186.

Legal Approval Ex Ante?

253

5.1.1.3.4.  Peculiarly worrisome problems Before this chapter responds to these concerns of Dershowitz and Levinson, we should note that their concerns are even more pressing for me than for the adherents of the Legal Prohibition Thesis. After all, the proponents of that thesis believe that the perpetration of calamity-averting interrogational torture in circumstances of extreme desperation should sometimes not trigger any sanctions. Though they insist that the officials who resort to such torture should be held accountable in criminal or civil or institutional proceedings, they contend that those proceedings should sometimes eventuate in acquittals through the invocation of fully exonerative defences—even if no collectively borne sanctions are to be imposed. Thus, although the theorists who embrace the Legal Prohibition Thesis do of course have to come to grips with the problems that have just been discussed, those theorists will be inclined to regard some instances of jury nullification as legitimate ways of recognizing the applicability of exonerative defences. They will probably even regard some instances of prosecutorial inaction and executive pardons in a similar light (Gross 2008, 65–6). By contrast, as has already been suggested, this chapter will argue that all acts of interrogational torture perpetrated by public officials—and nearly all acts of interrogational torture perpetrated by private citizens—should be met with legal sanctions (whether the sanctions are individually borne or collectively borne, and whether they are criminal or civil or institutional). Thus, by the reckoning of this book, every instance of jury nullification in a trial of a governmental organization whose officials have availed themselves of such torture is morally improper. Likewise, every instance of prosecutorial inaction in such a case is morally improper, if the inaction derives not from doubts about the strength of the evidence but instead from doubts about the actual or perceived wrongness of the torture. Such responses to acts of interrogational torture not only fail to rectify the original wrongs, but also commit new breaches of moral duties by condoning those wrongs. Unlike the Legal Prohibition Thesis, this book’s position on the moral impermissibility of interrogational torture is absolutist. Accordingly, unlike the proponents of the Legal Prohibition Thesis, this chapter maintains that fully exonerative defences (as opposed to mitigative defences) are never appropriate in trials of collectivities whose officials have perpetrated interrogational torture. Hence, the spectre of jury nullification and prosecutorial inaction is especially worrisome for me. We should thus now turn to some rejoinders to Dershowitz and Levinson. My rejoinders will initially focus on their concerns in the context of the debates over torture warrants, and will then address their concerns beyond the confines of those debates. 5.1.1.3.5.  A first rejoinder to Dershowitz: torture without warrants Real and important though the problems of official mendacity and jury nullification and prosecutorial inaction are, they are difficulties for Dershowitz as well as for his opponents. Albeit the overall impact of a scheme of torture warrants on the incidence of those problems is an empirical matter to be charted by social scientists, some general considerations support the conclusion that such a scheme would

254

Legal Responses to Torture

do little to alleviate those problems. For one thing, while Dershowitz’s envisaged scheme would effectively shield officials from ex post accountability for interrogational torture conducted with warrants, it would scarcely eliminate the occurrence of warrantless instances of such torture. Officials might engage in interrogational torture without warrants on occasions when they have sought warrants and have failed to obtain them—if there are any such occasions—or in emergencies where they have not sought warrants (because of the urgency of the circumstances). Moreover, even when officials do seek and obtain warrants, they might do so on the basis of hyperbolic or disingenuous statements in their applications, and they might carry out interrogational torture beyond what has been specified in the warrants. The incentives for the officials to prevaricate about their unauthorized torture will be approximately as strong in the presence of a system of torture warrants as in the absence of such a system. What should be noted here is that warrantlessness is likely to be typical rather than exceptional. After all, Dershowitz purports to be addressing the problem of ticking-bomb emergencies. He invokes ticking-bomb scenarios repeatedly in support of his torture-warrant proposal, and the title of the relevant chapter from his 2002 book is ‘Should the Ticking Bomb Terrorist be Tortured?’ Yet the severe constraints of time that are operative in ticking-bomb emergencies will typically render infeasible the obtention of any warrants. For example, in Leon v. Wainwright—a case discussed frequently by Dershowitz—there was not enough time for the constables to comply with procedures for seeking a warrant, even if such procedures had been available. Likewise, in the Antipodean case recounted by Seumas Miller (2005, 182–3; 2011, § 3) and mentioned by me in § 2.2.1.3 of Chapter 2, the desperate pressures of time left no opportunity for the pursuit of a warrant even if a torture-warrant scheme had been in place. Dershowitz himself readily acknowledges that exceptions to the requirement for search warrants in the United States are commonplace, and he specifically mentions ‘exigent circumstances’ as the first class of exceptions (2002, 254 n41). Exceptions on grounds of exigent circumstances will virtually swallow up the requirement for torture warrants in any jurisdiction where Dershowitz’s proposal has been adopted, since those warrants are supposed to be issued only in circumstances of extreme desperation: the very circumstances in which there will not usually be sufficient time to get warrants. Whereas warrant-authorized searches of people’s premises by the police are often morally legitimate outside the contexts of dire emergencies, Dershowitz propounds his torture-warrant scheme precisely as a means of dealing with such emergencies in an open fashion. In other words, his scheme would be instituted to come to grips with cases of acute desperation in which the scheme will almost always be unworkable. Even in a jurisdiction where Dershowitz’s proposal has been adopted, then, nearly all acts of interrogational torture will occur without warrants (unless, of course, the issuance of warrants is extended to non-emergency situations—in which the moral wrongfulness of interrogational torture is especially grave). Dershowitz does fleetingly take account of the possibility of warrantless torture, but his terse remark on that possibility is woefully inadequate: ‘[A]‌lthough it is

Legal Approval Ex Ante?

255

certainly possible that some individual agents might torture without a warrant, they would have no excuse, since a warrant procedure would be available. They could not claim “necessity,” because the decision as to whether the torture is indeed necessary has been taken out of their hands and placed in the hands of a judge’ (2002, 159). Remarkably, Dershowitz appears not to have recognized that a scheme of warrants will typically be unworkable in the very circumstances of dreadful urgency which the scheme is supposed to address. In most such circumstances, there will be insufficient time for officials to avail themselves of a warrant procedure. Hence, in most such circumstances, the existence of a warrant procedure will not take the key decisions out of the hands of constables or soldiers or intelligence officials. Dershowitz’s unattunedness to the problem discussed here is apparent from his quoting of the US Supreme Court’s statement of the rationale for the search-warrant requirement:  ‘The informed and deliberate determinations of magistrates . . . are to be preferred over the hurried action of officers’ (Dershowitz 2002, 160). Given that informed and deliberate determinations of magistrates are feasible in quite a few of the situations in which constables legitimately wish to undertake searches of premises, those sober determinations in such situations are indeed to be favoured over the hurried actions of officers. However, in the horrific predicaments that might lead to the perpetration of calamity-averting interrogational torture, there will typically not be sufficient time for the magistrates’ informed and deliberate determinations. Under the dire constraints of time that are operative in ticking-bomb quandaries, there are typically no realistic alternatives to the hurried actions of officers. Because of the absence of alternatives to those hurried actions in ticking-bomb plights, the incidence of warrantless interrogational torture might well increase in a jurisdiction where Dershowitz’s torture-warrant proposal has been adopted. When constables or other officials in that jurisdiction are confronted with a grim emergency in which there are no opportunities to seek warrants for interrogational torture, the fact that their society in its general law has authorized such torture (albeit with a warrant requirement) might well embolden them to resort to it.8 Of course, the probability of such an increase in the incidence of warrantless interrogational torture is an empirical matter on which nobody can very safely pronounce in abstracto. Still, the empirical conjecture advanced in this paragraph is at least as plausible as Dershowitz’s surmises about the torture-reducing effect of a system of torture warrants. At any rate, whether or not the incidence of warrantless interrogational torture would actually increase in a jurisdiction where Dershowitz’s scheme of torture warrants has been introduced, most instances of interrogational torture that occur in such a jurisdiction will be warrantless (unless the issuance of warrants is extended beyond the terrible emergencies to which the scheme is supposed to be confined). 8   Likewise, the fact that the society in its general law has authorized the perpetration of interrogational torture might well lead juries to be especially indulgent toward officials who have resorted to such torture in urgent circumstances where the obtention of warrants has not been feasible.

256

Legal Responses to Torture

Accordingly, the problems which Dershowitz associates with legal systems that have not embraced his torture-warrant proposal—the problems of official mendacity, jury nullification, and prosecutorial inaction—will be comparably troublesome in any legal systems where his proposal has been adopted. Even if his proposal were sound in most other respects, it would do little if anything to alleviate those problems. 5.1.1.3.6.  A second rejoinder to Dershowitz: arguments in tension As we have seen in § 5.1.1.3.3, Dershowitz frequently asserts that the problems of official mendacity and jury nullification and prosecutorial inaction will loom large—and will undermine the accountability of constables and other officials who employ interrogational torture—in any jurisdiction where no scheme of torture warrants is in place. However, that line of argument is in tension with another line of argument which he advances at several junctures. He submits that the scheme of ex post legal accountability favoured by the proponents of the Legal Prohibition Thesis is unfair to the constables and intelligence officials who might have to resort to interrogational torture in desperate emergencies. He contends that the ex post approach ‘leaves each individual member of the [police or] security services in the position of having to guess how a court would ultimately resolve his case. That is extremely unfair to such investigators.’ Dershowitz declares: ‘Individual interrogators should not have to place their liberty at risk by guessing how a court might ultimately decide a close case. They should be able to get an advance ruling based on the evidence available at the time’ (2002, 252 n26; 2004, 263). So far, the line of argument about uncertainty and unfairness appears to be starkly inconsistent with Dershowitz’s claim that ex post accountability will prove to be chimerical because of jury nullification and prosecutorial inaction. If Dershowitz is so confident that prosecutors and juries will exonerate officials who have availed themselves of calamity-averting interrogational torture, why is he insisting that a scheme of ex post accountability will leave such officials racked with uncertainty concerning the legal consequences of their actions? Why does he think that those officials will have to guess how the courts might decide their cases, if he believes that the courts’ decisions are foregone conclusions? However, a later passage indicates that there is a tension—rather than a stark inconsistency—between Dershowitz’s positions. Let us recall his response to Scarry which I have quoted in § 5.1.1.3.3. Dershowitz there accuses her of naivety for suggesting that juries would ever convict officials who have managed to avert calamities through recourse to interrogational torture. His response to her continues: ‘But some juries might well convict if the torturer turned out to be wrong, even if he or she had a very strong basis on which to act. Our legal . . . systems should make accountability turn on a defendant’s mens rea (state of mind) at the time he or she acted, not on fortuities beyond his or her control’ (2004, 274). In a digression, let us glancingly note that the final sentence in this latest quotation is little more than a flourish of hyperbole. Dershowitz is well aware that the guilt or innocence of someone under the criminal law often quite legitimately hinges on fortuities beyond his or her control. For example, suppose that two

Legal Approval Ex Ante?

257

motorists drive with equal hazardousness down the same road at approximately the same time. Suppose that a pedestrian happens to be crossing the road during the second motorist’s journey, whereas nobody is in the road during the first motorist’s journey. If the negligence of the second driver results in his hitting and killing the pedestrian, he will be guilty of vehicular homicide under the criminal law of the jurisdiction in which the mishap occurs. By contrast, the first driver is not guilty of vehicular homicide even though his mens rea was the same as that of the second driver. Quite legitimately, the distinction between the innocence of the first motorist and the guilt of the second—in connection with the crime of vehicular homicide—depends on the fortuity of the pedestrian’s presence in the road. At any rate, the main point of my including this latest quotation from Dershowitz is that it presents a more sophisticated version of his complaint about uncertainty and unfairness. He is maintaining that, if officials who have resorted to calamity-averting interrogational torture are placed on trial for their actions, nothing succeeds like success. When officials have employed such torture precipitately in circumstances where it is unlikely to be efficacious or where alternative methods of interrogation are still feasible, they will nonetheless be acquitted if they have extracted some information that has helped to avert a calamity. Conversely, when officials have employed interrogational torture in desperate circumstances where it is likely to be efficacious as a means of averting a calamity and where all feasible alternative methods of interrogation have failed, they will probably be convicted if their efforts have come to naught. Such is Dershowitz’s contention. Hence, although there is a tension between his insistence on the illusiveness of ex post accountability and his worries about the uncertainty of officials concerning the legal consequences of their perpetration of torture, there is not an outright inconsistency. Dershowitz is not suggesting that arrangements for ex post scrutiny will never hold perpetrators of calamity-averting torture meaningfully accountable. He is instead contending that such arrangements will hold the perpetrators meaningfully accountable when and only when their endeavours to prevent disasters have turned out to be unsuccessful. 5.1.1.3.7.  The second rejoinder to Dershowitz continued Although Dershowitz talks only about criminal sanctions, we can construe his remarks more broadly to encompass other types of sanctions as well. His position, then, consists of two main claims: the claim (Claim A) that officials and/or their organizations will probably undergo sanctions if their efforts to avert calamities through the use of interrogational torture have failed; and the claim (Claim B) that the officials and their organizations will be exonerated if their efforts to avert calamities through the use of interrogational torture have succeeded. Let us ponder these two claims. Dershowitz’s Claim A should elicit applause rather than dismay. As this book has argued, interrogational torture is morally wrong even when it extracts information that helps to forestall the occurrence of some calamity. A fortiori, then, such torture is morally wrong when it fails to achieve the calamity-averting purpose for which it has been undertaken. Because interrogational torture not only is morally

258

Legal Responses to Torture

wrong but also involves serious violence, a legal system’s connivance at it through the exoneration of the officials and organizations responsible for it would in turn amount to quite a serious wrong. Of course, the circumstances of the perpetration of the torture have a significant bearing on the gravity of its wrongness. To gauge the gravity of what has been done, and thus to gauge the severity and nature of the sanctions that are appropriate in response to what has been done, juries or judges will have to look carefully at the circumstances of the perpetration of the torture. If officials have undertaken the interrogational torture as a last resort in a dreadful emergency, and if there were reasonably solid grounds for them to expect the torture to be efficacious, the gravity of its wrongfulness is considerably extenuated by those aspects of its occurrence—even when the officials have not succeeded in fending off some catastrophe. Dershowitz does not suggest that juries or judges will be disinclined to take account of such considerations, and he certainly does not provide any evidence that they will be so disinclined. Rather, he simply asserts that they might well decide to impose sanctions despite the extenuating factors. That assertion is entirely in keeping with my position in this chapter, which is that juries or judges (or other decision-makers) should impose sanctions in such cases. The extenuative factors bear on the heaviness and nature of the sanctions that are due, rather than on the question whether any sanctions at all are due. Thus, although Dershowitz seems to regard Claim A with consternation, it is actually to be welcomed. Moreover, even the proponents of the Legal Prohibition Thesis—who believe that some instances of interrogational torture by officials in extreme circumstances should not trigger the imposition of any sanctions—can welcome Dershowitz’s Claim A.  Although those proponents maintain that sanctions should not be levied in truly extreme cases, they hold that sanctions are appropriate and important in every case where the circumstances that have prompted officials to wield interrogational torture are less than horrendously urgent. Likewise, they hold that sanctions are appropriate and important in every case where interrogational torture has been undertaken with no reasonably solid grounds for expecting it to be successful, and in every case where such torture has been undertaken before other methods of interrogation have been exhausted. The adherents of the Legal Prohibition Thesis contend that the torture in each such case is morally wrong. Hence, they can heartily welcome the deterrent force of the state of affairs that is recounted in Claim A. That is, they can be pleased that jurors and judges are disposed to levy sanctions in cases where interrogational torture has failed to avert calamities, for the prospect of sanctions will deter officials from administering such torture except in the direst of emergencies. As Sanford Kadish writes: ‘Would not the burden on the official be so great that it would require circumstances of a perfectly extraordinary character to induce the individual to take the risk of acting? The answer is of course yes, that’s the point’ (1989, 355). Oren Gross has repeatedly and emphatically stated a similar view. He declares that his model of ex post accountability ‘seeks to add uncertainty to the decision-making calculus of state agents which, in turn, raises both the individual and national costs of pursuing an extra-legal course of action [such as

Legal Approval Ex Ante?

259

interrogational torture] and, at the same time, reinforces the rule of obedience’ (2008, 72). 5.1.1.3.8.  The second rejoinder to Dershowitz completed Thus, Dershowitz’s opponents can hail Claim A. Claim B, concerning the likelihood of exonerations in cases where calamity-averting interrogational torture has been plied successfully, is obviously more troublesome. Still, two brief observations in response to Claim B will suffice at this stage. One point to be noted, in keeping with what has been argued in § 5.1.1.3.5, is that the worries raised by Dershowitz about exonerations are as troublesome (or nearly as troublesome) for him as for his opponents. Given the infeasibility of obtaining warrants for interrogational torture in the dreadful emergencies which his system of torture warrants is supposed to handle, a society that adopts his system will have to deal with warrantless instances of interrogational torture just as will a society that does not adopt his system. Quite speculative is any answer to the question whether the level of exonerations will be higher in the one society than in the other. Perhaps the existence of a system of warrants will more strongly incline jurors and judges to impose sanctions on officials who have resorted to interrogational torture without warrants in circumstances of great urgency where the obtention of warrants was infeasible. Or perhaps instead the existence of such a system, which legally authorizes the perpetration of interrogational torture (albeit with warrants), will make jurors and judges more indulgent toward officials who have had recourse to interrogational torture in emergencies where there was insufficient time for the seeking of warrants. Though the latter conjecture seems to me much more plausible than the former, there is no need here for a choice between them. My aim in assailing Dershowitz’s empirical surmises does not lie in disproving them by vindicating some contrary empirical hypotheses. Rather, my aim in assailing his surmises is to show how weak and speculative his arguments are. That aim can be fulfilled without my having to argue that the latter of the two empirical conjectures broached in this paragraph is correct. The other main point to be noted in response to Claim B is that the concern raised in that claim would have to be addressed in any event. Even if Dershowitz had never propounded his ideas about torture warrants, this chapter with its advocacy of a blanket legal prohibition on interrogational torture would have needed to come to grips—at least briefly—with the problems of official mendacity and jury nullification and prosecutorial inaction. We should therefore now move to a short exploration of those problems beyond the context of the debates over torture warrants. 5.1.1.3.9.  A third rejoinder to Dershowitz: ways of dealing with the problems At the philosophical level of abstraction occupied by this book, not very much can usefully be said about the best ways of dealing with the problems just mentioned. Many factors that vary across jurisdictions will affect the appropriateness of different tacks that might be adopted. However, a few laconic observations can suggest some paths forward.

260

Legal Responses to Torture

First, the problems of official chicanery and jury nullification and prosecutorial inaction are hardly unique to contexts involving interrogational torture. Those problems have surfaced, separately or in combination, in a number of settings. For example, the lynchings of black people which so shamefully stained American history in the first several decades of the twentieth century—and which often involved sadistic torture—were frequently carried out with the cooperation of local officials, who then routinely lied about their actions (Skolnick 2004, 106–7). On the rare occasions when prosecutions were pursued against perpetrators of the lynchings, juries almost always declined to convict anyone (Lyons 2008). Those egregious problems were very gradually overcome through civil-rights campaigns against the evils of racism, and through coordinated efforts to alter the prevailing attitudes and practices among police forces (Waldron 2010, 241–2), and through interventions by federal law-enforcement agencies and federal courts (Skolnick 2004, 112–16). As is suggested by the scale of the endeavours to tackle the problems of official wrongdoing and inadequate law-enforcement in this context, the prospects for success in those endeavours are greatly enhanced when multiple agencies and levels of government work together. Another matter in regard to which jury nullification has been a problem in the United States is that of deaths caused by drunken driving. In the early decades of automotive travel, prosecutors lacked confidence in the willingness of jurors to return guilty verdicts on charges of homicide against motorists who had caused fatal collisions while driving under the influence of alcohol. One of the chief developments during that period, in various American states, was the establishment of the crime of vehicular homicide as an offence with which an inebriated driver could be charged if he or she had caused a deadly accident. Typically, the elements of the crime were somewhat more easily proved than the elements of other types of homicide, and the punishments and stigma attached were somewhat lighter (Jacobs 1989, 84). Those features of the new offence, in combination with public-education campaigns, helped to bolster the leverage of prosecutors in pressing charges and in plea bargaining. Some comparably innovative reclassifications might be salutary in the legal regulation of torture. 5.1.1.3.10.  A third rejoinder to Dershowitz continued A further observation moves back from these other contexts to the context of brutal interrogation itself. As Jerome Skolnick discusses (2004, 119), some imaginative procedural devices in civil litigation can help to surmount the obstacles posed by prosecutorial inaction and jury nullification. Civil litigation obviously circumvents the problem of prosecutorial inaction. However, under the traditional arrangements for such litigation in the United States, the problem of jury nullification remains. After all, the matter considered here is that of legal remedies for calamity-averting interrogational torture perpetrated by constables or soldiers or intelligence officials. Especially if the torture in a given case has proved to be effective in extracting some life-saving information from a culprit, the defendants in any civil litigation that arises from the administration of the torture will tend

Legal Approval Ex Ante?

261

to be viewed sympathetically by jurors, and the plaintiff(s) will tend to be viewed with disdain. Accordingly, an ordinary civil lawsuit for a large award of damages to be paid to the plaintiff(s) will not stand a good chance of success, even if the evidence of the use of torture is clear-cut. Hence, the US Congress passed a law in 1976 that provides for departures from the traditional arrangements for civil litigation. In cases where violations of civil rights by law-enforcement officers are alleged, judges have the leeway to award reasonable fees to the lawyers of successful plaintiffs irrespective of the levels of damages awarded. Thus, a defendant in such a case can be saddled with quite a hefty bill for the plaintiff’s legal costs even if the jurors have awarded a very small amount of damages in order to signal their lack of sympathy for the plaintiff. As Skolnick observes, for example, ‘Los Angeles attorney Stephan Yagman was awarded $29,137 in attorney’s fees in a case where his clients won a two-dollar verdict against the Los Angeles police department’ (2004, 119). Given such an arrangement, lawyers have incentives to pursue litigation in the civil-rights cases, and jurors can give effect to the law while evincing their low esteem for the plaintiffs who have been subjected to calamity-averting interrogational torture. Moreover, although the award of damages in such a scenario will be derisory, the overall remedy will be meaningful—since the defendants will be landed with the full expenses of the lawsuit that has been brought against them. They must bear the costs of the legally authoritative process that deems them to have breached the legal prohibition on torture. 5.1.1.3.11.  A third rejoinder to Dershowitz completed Over the longer term (probably the much longer term), one of the best ways to address the problems of jury nullification and prosecutorial inaction is to strengthen the international institutions that are responsible for the enforcement of the strict international-law prohibitions on the use of interrogational torture. Because those institutions are not rooted in the societies whose constables or soldiers or intelligence officials might resort to such torture, they—when functioning properly— are considerably less likely than domestic institutions to connive indulgently at instances of torture. Detached both organizationally and perspectivally, judges in properly functioning international institutions can unflinchingly give effect to the aforementioned prohibitions even while they also recognize the extenuative force of emergencies. Of course, I  am hardly suggesting that international institutions at present are performing adeptly in the role of enforcers of international-law prohibitions. A number of them have become notorious for ponderous inefficiency, corruption, grotesquely illiberal biases, and selectivity in their endeavours of enforcement. Nevertheless, given the potential importance of such institutions in overcoming the problems of jury nullification and prosecutorial inaction, the best tack for a liberal democracy resides not in disengaging from them but instead in striving for their improvement. If international institutions do eventually achieve the impartiality and effectiveness that befit their supplementation of the efforts of domestic institutions to enforce legal prohibitions on the use of torture, the

262

Legal Responses to Torture

concerns raised by Dershowitz and Levinson—concerns that abide even when their misconceived proposal for torture warrants is rejected—will have been significantly allayed.9 Also of crucial importance, especially in lowering the incidence of official mendacity, are human-rights organizations such as Amnesty International and Human Rights Watch and Freedom House and the American Civil Liberties Union. Those and other non-governmental organizations are themselves not beyond reproach in some of their activities, but their scrutinizing of the conduct of governmental officials throughout the world is invaluable. Although the use of interrogational torture is not the only type of wrongdoing that draws attention from these groups, campaigns against such torture are central to their mission. The people who work in those organizations are willing to brave the unpopularity and dangers that attend their endeavours to expose official chicanery. Rightly and admirably, they do not shrink from intervening in cases where victims of torture are rebarbative and where torturous techniques of interrogation have been undertaken in desperate circumstances for the purpose of saving lives. Employees of these organizations are prepared to take an absolutist stand on the wrongness of interrogational torture even in such cases. Accordingly, their activities play a vital role—within liberal democracies as well as within societies ruled by tyrannical regimes—in holding officials accountable for any breaches of legal prohibitions on torturous techniques of interrogation. Human-rights organizations are a mainstay of efforts to ensure that the wrongness of interrogational torture (even calamity-averting interrogational torture) is reflected in the sanctions levied by an ex post scheme of accountability. Irritating and unbalanced though the pronouncements of many of these groups sometimes are, their monitoring of the actions of legal-governmental officials is amply deserving of support from everyone who hopes to see legal prohibitions on torture effectively enforced. Similarly deserving of such support is the striving of investigative journalists to uncover acts of torture and other misconduct by governmental officials. Keenly motivated to expose the misdeeds of officials, those journalists work in tandem with human-rights organizations to reinforce indispensably the mechanisms by which torturers are brought to account for what they have done.

5.1.1.4. A fourth objection to Dershowitz: seeking support from a strange quarter Dershowitz (2003, 279–81; 2004, 268–9) strangely seeks to support his torturewarrant proposal by referring to the infamous case of Korematsu v. United States, in which the US Supreme Court held that the removal of Japanese-Americans from certain regions of the western United States for the purpose of detaining them in internment camps during the Second World War was constitutional. He focuses on the dissenting opinion by Justice Robert Jackson, who argued 9   For some general remarks on the role of international law as a basis for the imposition of sanctions in response to acts of torture by public officials, see Gross 2008, 75–7, 90–1; Lim 2008, 398–406; Mayerfeld 2007; Van der Vyver 2003.

Legal Approval Ex Ante?

263

that the military orders which applied to the defendant should not have been treated as laws enforceable by courts. Jackson further maintained that the matter of the validity of those orders as military directives should have been deemed to be non-justiciable. He worried that the Supreme Court’s affirmation of the constitutionality of the orders established a precedent that would subsequently be invoked to justify invidious discrimination in other spheres of government. In his eyes, the Court’s countenancing of the military directives as constitutional was worse than the promulgation of the directives themselves, because it was much more far-reaching in its implications. Dershowitz takes Jackson to have favoured an ‘off the books’ approach that broadly resembles the approach favoured by Posner in connection with calamityaverting interrogational torture. Having quoted Jackson’s prophecy that the discrimination-condoning import of the Supreme Court’s decision would ramify into many other areas of the law, Dershowitz retorts: Experience has not necessarily proved Jackson’s fear or Posner’s prediction [of the metastasizing effects that will ensue from a system of torture warrants] to be well founded. The very fact that the Supreme Court expressly validated the detentions contributed to its condemnation by the verdict of history. Today, the Supreme Court’s decision in Korematsu stands alongside decisions such as Dred Scott, Plessy v. Ferguson, and Buck v. Bell in the High Court’s Hall of Infamy. Though never formally overruled, and even occasionally cited, Korematsu serves as a negative precedent—a mistaken ruling not ever to be repeated in future cases. Had the Supreme Court merely allowed the executive decision to stand without judicial review, a far more dangerous precedent might have been established: namely, that executive decisions during times of emergency will escape review by the Supreme Court. That far broader and more dangerous precedent, espoused by Posner, would then lie about ‘like a loaded weapon’ ready to be used by a dictator without fear of judicial review. (2003, 280–1; 2004, 268–9, footnotes omitted)

Two replies to Dershowitz’s contentions are appropriate here. First, even if Dershowitz is correct in claiming that Posner favours an ‘off the books’ approach to calamity-averting interrogational torture, such an approach is not favoured by most of the people who oppose a system of torture warrants. Most such people insist that any officials who resort to interrogational torture should be held accountable ex post for their actions. Hence, the position of most of Dershowitz’s opponents does not resemble Justice Jackson’s position. Far from asserting that the courts should not be involved in gauging the lawfulness of calamity-averting interrogational torture, most of the opponents of Dershowitz’s torture-warrant proposal contend that the courts should be centrally involved in passing judgment on every instance of such torture. Second, it is decidedly odd that Dershowitz adduces the Korematsu case in support of his proposal for the issuance of torture warrants by judges. Although the Supreme Court’s assessment of the military policy of relocating and interning Japanese-Americans was ex post rather than ex ante, it was otherwise relevantly similar to the judiciary’s authorization of interrogational torture under the scheme of torture warrants envisaged by Dershowitz. Just as the Supreme Court scrutinized and approved the treatment of Japanese-Americans by the US military

264

Legal Responses to Torture

during the Second World War, so judges under Dershowitz’s scheme will scrutinize and approve applications for warrants to perpetrate interrogational torture. Yet, as Dershowitz himself emphasizes, the Korematsu decision has been salutary only because it has come to be so roundly execrated as a grievous mistake on the part of the Supreme Court. Moreover, most modern-day commentators who view the Korematsu judgment with revulsion do not submit that the Supreme Court’s error lay in deeming the constitutionality of the military relocation policy to be a justiciable matter. Rather, the Court’s error lay in affirming the constitutionality of that policy. Correct though the Court was in deeming the matter to be justiciable, the majority of the Justices erred terribly when they failed to hold that the military orders were an unconstitutional abuse of power. Dershowitz himself acknowledges that ‘[t]‌he major downside of any warrant procedure would be its legitimation of a terrible practice’ (2003, 283; 2004, 271). He in effect acknowledges, then, that his system of torture warrants is relevantly similar to the widely despised Korematsu decision. He acknowledges as much while also rightly maintaining that that lamentable decision has been valuable only because it serves as a negative precedent: that is, only because it is recognized as an egregious misjudgment that should never be repeated. What is puzzling is that his clear-sighted perception of the iniquity of the Korematsu decision does not deter him from championing a scheme of torture warrants that is so closely analogous to that decision.

5.1.1.5.  A final objection to Dershowitz: a matter of moral principle Especially given how shaky the empirical presuppositions of Dershowitz’s torture-warrant proposal are—though even if those presuppositions (and his arguments in support of them) were not quite so feeble and conjectural—his proposal should be rejected as a matter of moral principle. More objectionable than every other aspect of his proposal is that the incorporation of it into the law of any jurisdiction would formally authorize the perpetration of interrogational torture under certain conditions. As Chapter  3 of this book has argued, the perpetration of such torture is always morally wrong. Even when any torturous techniques are undertaken against a mass-murdering terrorist for the purpose of averting a calamity which he has endeavoured to bring about, there are perpetrator-focused considerations that undermine the moral legitimacy of those techniques. Thus, if the law of a particular jurisdiction were to establish a scheme of torture warrants, it would be formally bestowing its imprimatur on the performance of morally wrongful actions. In that respect, torture warrants would differ saliently from most of the other licences that are frequently issued by governments: for example, licences to drive automotive vehicles, licences to engage in certain trades, licences to purchase and keep guns, licences to export products of certain types. None of those normally licensed activities is morally wrong per se (though of course any of them can be pursued in abusive ways). By contrast, interrogational torture is always morally wrong. Moreover, it is a serious moral wrong that involves violence. In that latter regard, it differs from some other morally dubious activities—such as

Legal Approval Ex Ante?

265

prostitution—that are occasionally licensed by governments. A legally established system for the issuance of torture warrants would be distinctive in its countenancing of seriously wrongful actions that involve major violence.10 This objection of moral principle to Dershowitz’s proposal for a scheme of torture warrants does not depend on the claim that such a scheme will inevitably result in the issuance of warrants. On the one hand, some previous portions of my critique of his proposal strongly suggest that the introduction of a system of torture warrants would indeed very likely result in the issuance of warrants (notwithstanding that most torture would still be warrantless). Dershowitz himself does not presume that applications for warrants will never be approved under such a system, and his arguments are confused or dubiously conjectural when he tries to establish that his scheme of warrants will tend to lower the incidence of interrogational torture. On the other hand, even if mirabile dictu the introduction of such a scheme would not eventuate in the issuance of any warrants, the very existence of the scheme within the relevant jurisdiction would lend the imprimatur of the law to a seriously wrongful mode of conduct. The authorization of that mode of conduct at a general level would tarnish the moral standing of the law, even if (amazingly) no warrants for torturous sessions of interrogation would ever be granted in the processes prescribed for giving effect to the general authorization. Although this objection to Dershowitz’s proposal is concerned partly with the symbolic import of legal prohibitions and authorizations, the symbolic dimension of law is not the sole matter—or even the principal matter—at issue. Even more important than the dismal message conveyed by the existence of a scheme that formally authorizes the perpetration of interrogational torture is the sheer wrongfulness of such a scheme. Consider, as an analogy, a law that provides for the authorization of the enslavement of people from certain racial groups under specified conditions. Any such law if promulgated would of course be horrendous symbolically, in that it would convey the message that the subjugation of people on the basis of their racial characteristics is sometimes to be countenanced. That message would be conveyed even if the programme of enslavement were never concretely implemented. Suppose, however, that the law would not ever be promulgated and that it would instead be kept strictly secret (while also not ever being concretely implemented). Its symbolic import in those circumstances would be negligible, yet the enactment of such a law would still be a serious wrong. It would besmirch the moral standing of any legal system that includes it, for it would associate that system with racial discrimination and enslavement. It would direct the judges or administrators of the system to authorize a nefarious mode of conduct under certain conditions, and—even if never promulgated or concretely implemented—it would thereby pro tanto vitiate the moral tenor of the system. An individual who 10   I here presuppose that killings undertaken lawfully in emergencies or in combat can be morally legitimate. I  have discussed some of the key moral differences between killing and interrogational torture in §§ 3.2.3.2, 3.2.3.3.4, and 3.2.3.3.5 of Chapter 3. I am also here assuming that hunting and fishing at moderate levels can be morally legitimate—though I am not wholly unsympathetic to the suggestion that such activities are always morally illegitimate and that they should thus never be licensed.

266

Legal Responses to Torture

adopts a bigoted attitude is tarnished in his character even if he never discloses or expresses his attitude (perhaps out of fear of the negative reactions that would be elicited by any such disclosure or expression); likewise, a system of legal governance that adopts a bigotry-condoning and slavery-condoning law is sullied in its moral fibre even if the law is never publicized or effectuated. Much the same is true of a system of legal governance that adopts a law which condones the perpetration of interrogational torture. Even if that law is kept hidden and is never implemented, its presence as one of the authoritative norms of the system is morally corrosive. Those norms in any legal regime should never include a law which authorizes a practice that is always seriously wrongful and violent. This point about the intrinsic wrongness of condonations of interrogational torture is missed by Louis Seidman in his attack on Dershowitz’s critics: ‘[G]‌ood faith empirical disagreement alone hardly explains the level of anger generated by [Dershowitz’s torture-warrant] proposal. That anger suggests that many torture opponents find his proposal abhorrent even if [Dershowitz is correct in his empirical conjectures]. Apparently, these [opponents] are unwilling to trade the symbolic satisfaction produced by the legal prohibition of torture for a reduction in the incidence of the practice itself ’ (2005, 884–5). With the phrase ‘symbolic satisfaction’, Seidman rather fatuously insinuates that the critics of Dershowitz’s torture-warrant scheme are concerned chiefly with good appearances and emotional comfort. Although the symbolic impact of the establishment of a torture-warrant scheme would indeed be dismaying, the objectionableness of such a scheme goes well beyond matters of symbolism. In line with what has just been argued here, the opponents of Dershowitz’s proposal are right to be angry about his preparedness to corrupt the moral tenor of the American legal system. They are right to be angry about his efforts to commit that system—in principle and almost certainly in practice—to the authorization of actions by public officials that are always morally wrongful. His opponents’ anger is undoubtedly intensified by the weakness and confusedness of the arguments and empirical surmises with which Dershowitz tries to show that the adoption of his scheme would be salutary.

5.1.2.  Posner and Vermeule on the regulation of torture Dershowitz is by no means the only theorist to have written during the past decade in support of the ex ante authorization of interrogational torture. Though the writings of Posner and Vermeule on the topic have received far less attention outside scholarly circles than have Dershowitz’s ideas, their proposals for the regulation of interrogational torture are much more wide-ranging and are more sophisticatedly argued. Posner and Vermeule endorse Dershowitz’s torture-warrant proposal, albeit tepidly. However, they also propound a considerably broader regulatory framework for the use of interrogational torture (2006, 699–703; 2007, 208–13). Their main contention is that the laws which regulate the wielding of interrogational torture by officials in the police or the military or the intelligence agencies should be essentially the same as the laws which regulate the wielding of deadly force by those officials. As they write: ‘[W]‌e think that the regulation of the use of

Legal Approval Ex Ante?

267

deadly force provides a model for regulating [interrogational torture]’ (2006, 700; 2007, 211). Like the exerting of deadly force, the plying of interrogational torture should be legally authorized ex ante under certain conditions. Such is the chief claim advanced by Posner and Vermeule. Posner and Vermeule marshal a number of dubious arguments as well as some sound arguments in the course of presenting their regulatory framework for the use of interrogational torture. My brief engagement with their work will examine only a few of their lines of reasoning. We shall first look at their dismissal of deontological absolutism. We shall then mull over their criticisms of the Legal Prohibition Thesis; as will become apparent, I  agree with some but not all of their queries about that thesis. While casting doubt on the operational significance of the difference between the ex post orientation of the proponents of the Legal Prohibition Thesis and the ex ante orientation of Posner and Vermeule, my discussion will maintain that the former orientation is morally superior in some key respects. My engagement with the work of Posner and Vermeule will then close by impugning their pivotal analogy between interrogational torture and deliberate killing. This chapter’s exploration of that analogy will return us to some matters pondered in Chapter 3.

5.1.2.1.  Posner and Vermeule on deontological absolutism In their extremely terse and dismissive appraisal of deontological absolutism, Posner and Vermeule present virtually no argumentation at all. Indeed, they initially state that ‘we bracket and ignore the claim that coercive interrogation is deontologically impermissible per se, whatever the facts’ (2006, 674; 2007, 185). They explain their intention to ignore deontological absolutism by maintaining that ‘[w]‌ith a very few exceptions, [such absolutism] is a view nobody holds; most mainstream philosophers—both consequentialists and deontologists—agree that [interrogational torture] may be morally justified under certain conditions’ (2006, 674). Here and elsewhere, Posner and Vermeule draw none of the key distinctions which I have drawn in Chapter 1. However, given their robustly consequentialist outlook, they believe that interrogational torture is strongly justified in certain situations. That is, they believe that such torture in some extreme situations does not breach any abiding moral duties, because the moral duties that would normally be operative are cancelled—overridden rather than merely overtopped—in those situations. Hence, when Posner and Vermeule sneer at deontological absolutism, they are rejecting the thesis that the use of interrogational torture is never strongly justified. They are rejecting a central thesis of this book. 5.1.2.1.1.  Missed distinctions and an inapposite focus Despite their initial assertion that they intend to ignore deontological absolutism altogether, Posner and Vermeule subsequently devote a couple of paragraphs to it. They renew their allegation that such absolutism is shuddersomely unpopular: ‘One might hold that [interrogational torture] is absolutely impermissible, as a violation of rights rooted in human dignity or autonomy. This position is held by a

268

Legal Responses to Torture

very few moral philosophers’ (2007, 187). Regrettable, once again, is the failure of Posner and Vermeule to draw any of the chief distinctions which I have drawn in Chapter 1. They differentiate neither between strong absoluteness and weak absoluteness nor between strong permissibility and weak permissibility. Furthermore, they take as given that any considerations invoked by proponents of deontological absolutism in condemnation of interrogational torture will be victim-focused. My own position in this book, of course, is that the administering of interrogational torture is always and everywhere morally wrong in all possible worlds. To be sure, in some imaginable emergencies the impermissibility of such torture is only weak rather than strong; that is, in some imaginably extreme circumstances, the use of interrogational torture is a less grave wrong than is the remissness involved in abstaining from it as a last resort. Still, although the moral prohibition on interrogational torture is weakly absolute rather than strongly absolute, it does indeed extend to every instance of such torture. As Chapter 3 has argued, the justificatory basis for that absolute prohibition has to be perpetrator-focused as well as victim-focused. An unremittingly victim-focused perspective would not lead to the conclusion that interrogational torture is always and everywhere morally impermissible. 5.1.2.1.2.  A charge of fanaticism and an easy target Posner and Vermeule advert to ticking-bomb scenarios and assert that ‘it is fanatical to argue on deontological grounds that rights against [interrogational torture] should not be overridden to prevent serious harms to others. That [absolutist] position denies that there can ever be such a thing as a justified violation of rights, or a necessary evil’ (2006, 676; 2007, 187). Because the term ‘overridden’ in the first sentence of this quotation is to be construed as ‘cancelled’ or ‘suspended’, the term ‘justified’ in the second sentence is to be construed as ‘strongly justified’. Hence, Posner and Vermeule are here dismissing the weakly absolutist position espoused by this book—as well as any strongly absolutist position, of course. Their dismissal proceeds through their bandying of the epithet ‘fanatical’, in lieu of argumentation and in lieu of any of the distinctions which I have drawn in Chapter 1. Posner and Vermeule do go on to offer a couple of quick arguments. They first rightly take exception to the effort by Thomas Nagel to supply a contractualist justification for the moral prohibition on the use of interrogational torture. However, given that Nagel ultimately defends a threshold-deontological position rather than an absolutist position on the matter of torture, and given that I too have mounted a critique of his contractualist lines of reasoning—in § 3.1.1.1 of Chapter 3—the tellingness of some strictures against his contractualism does not betoken the tellingness of any strictures against deontological absolutism. An absolutist position on the problem of interrogational torture scarcely needs to rely on contractualist lines of reasoning. 5.1.2.1.3.  Tragic choices as moral conflicts In a second attempt to expose the inadequacies of deontological absolutism, Posner and Vermeule submit that absolutists have failed to grasp that certain predicaments

Legal Approval Ex Ante?

269

confront us with tragic choices: ‘[Interrogational torture] presents a “tragic choice.” A view holding that [interrogational torture] is sometimes permissible need not deny that [interrogational torture] is a grave moral evil; of course it is. But sometimes evils, even grave ones, are also necessary. The absolutist deontological view fails to come to grips with the inevitability of tragic choices.’ They straightaway conclude: ‘In what follows, then, we will put the absolutist deontological view off the table. Anyone who genuinely holds it may ignore our argument, but we do not think there are many such people’ (2006, 677; 2007, 187). Quite ironic is the allegation that deontological absolutists do not come to grips with tragic choices, for it is hard-core consequentialists like Posner and Vermeule who are committed to denying the occurrence of any veritable moral conflicts. When they characterize interrogational torture as a ‘grave moral evil’, they are not thereby indicating that the perpetration of it in a desperate emergency is morally wrong. Rather, they are indicating that it is a decidedly negative or undesirable factor which has to be balanced against some decidedly negative or undesirable factors that will materialize in the event that the interrogational torture does not go ahead. If the perpetration of the interrogational torture will be less detrimental to human well-being than would the calamity which the torture can help to avert, then—in the eyes of Posner and Vermeule—the perpetration of the torture is morally permissible despite its status as an ‘evil’. Though the deliberations leading up to the use of such torture might be extraordinarily arduous (not least because the probability of the inefficacy of the torture has to be taken into account), no moral conflict is involved. When an official resorts to calamity-averting interrogational torture as the less terrible of the two grievously deleterious courses of conduct that are open to him, he is not thereby choosing to undertake a wrong; rather, by minimizing the impairment of human well-being that will occur, he is opting for a course of conduct that is morally permissible as well as morally obligatory. Such is the position of staunch consequentialists like Posner and Vermeule. Their position on this matter becomes evident in a footnote where they make reference approvingly to the work of Martha Nussbaum. They write as follows: ‘Where tragic choices are involved, Nussbaum suggests, decisionmakers should at a minimum take pains to commemorate the values, rights, or interests that are overridden in the service of other commitments. That commemoration can presumably occur in a variety of ways, from compensatory payments to public apologies and memorials’ (2006, 676 n9; 2007, 296 n9). By talking about the relevant values or rights as overridden, and by suggesting that those overridden values or rights should be commemorated, Posner and Vermeule express their view that those values or rights cease to be operative in any dire emergency where the upholding of them would produce calamitous consequences. If Posner and Vermeule did not believe that those values or rights cease to be operative in such an emergency, they would not perceive any need for commemoration. Instead, they would perceive a need to rectify some contraventions of those values or rights. When Posner and Vermeule refer to compensatory payments and public apologies as means of commemoration, they are not referring to genuine remedies. Instead, they are referring to the simulacra of genuine remedies which I have discussed in § 1.1.2.1

270

Legal Responses to Torture

of Chapter 1. In the eyes of consequentialists, those remedy-resembling measures do not rectify any wrongs—for, in their eyes, there are no wrongs to be rectified in the aftermath of any tragic choice that has maximized good consequences or minimized bad consequences. Because there are no wrongs to be rectified, the role of a simulacrum of compensation (or a simulacrum of an apology) is consolatory rather than remedial. Such a measure assuages people who have undergone hardships as a result of the tragic choice. In light of its consolatory power, a remedy-resembling measure can win the favour of consequentialists in the aftermath of a tragic choice even though they do not believe that there is any role to be played by a genuine remedy in such circumstances. Far from agreeing with Nussbaum, then, Posner and Vermeule have distorted and obscured her insights into the occurrence of moral conflicts. Unlike them, Nussbaum recognizes that many situations of tragic choice consist in such conflicts. For her, the ‘tragic question’ in any situation of that sort is whether every course of conduct available to someone is morally illegitimate. She writes that ‘[t]‌he tragic question registers not the difficulty of solving the [question concerning which available course of conduct is best] but a distinct difficulty: the fact that all the possible answers to the [latter] question, including the best one, are bad, involving serious moral wrongdoing’ (2000, 1007, emphasis added). She goes on to state the matter emphatically, with reference to someone who has chosen the best course of conduct available in a grim plight: [T]‌here is a point to the tragic question. It keeps the mind of the chooser firmly on the fact that his action is an immoral action, which it is always wrong to choose. The recognition that one has ‘dirty hands’ is not just self-indulgence: it has significance for future actions. It informs the chooser that he may owe reparations to the [detrimentally affected people] and an effort to rebuild their lives after the disaster that will have been inflicted on them. When the recognition is public, it constitutes an acknowledgment of moral culpability. [2000, 1009.]

Unlike Posner and Vermeule, Nussbaum recognizes that measures such as compensation and apologies in the aftermath of tragic dilemmas are genuinely remedial rather than just consolatory. Instead of commemorating rights that have been suspended or cancelled, those measures serve to acknowledge rights which have never ceased to obtain as such and which have been transgressed. A genuinely remedial measure will be needed to rectify a wrong that is perpetrated in any situation where even the best course of conduct available to someone is morally illegitimate. Whereas consequentialists like Posner and Vermeule cannot accept that the best course of conduct available to someone is ever morally illegitimate, Nussbaum grasps that moral optimality does not entail moral permissibility. She grasps that, in any situation of moral conflict, somebody is confronted with a predicament in which he cannot avoid the perpetration of a wrong. Thus, although Posner and Vermeule assert that deontological absolutists are unattuned to the inevitability of tragic choices, it is Posner and Vermeule themselves who do not fathom the profundity of tragic quandaries. They can and do grant that such quandaries involve trade-offs that are often fiendishly unpleasant,

Legal Approval Ex Ante?

271

but they as consequentialists cannot grant that any such trade-off is a moral conflict in which every available course of conduct—including the morally optimal course of conduct in the circumstances—is morally impermissible. By contrast, a deontological absolutist can readily recognize that grim emergencies are frequently situations of moral conflict. In Chapter 2’s version of Leon v. Wainwright, for example, the police are faced with a plight in which they cannot avoid the commission of a wrong. If they resort to interrogational torture against the kidnapper, they will have breached their abiding moral duty to refrain from any such torture. If they do not resort to such torture after their attempts to elicit the crucial information from the kidnapper through several alternative methods have failed, they will have breached the moral duties which they owe as public-safety officials to the victim. In the dire emergency that confronts them, they are not engaged merely in a trade-off or balance of unappealing factors. Rather, they are immersed in an arrant moral conflict within which they will breach moral duties regardless of what they choose to do. Their recourse to interrogational torture might well be the morally optimal course of conduct in the desperate circumstances, but its wrongfulness sullies their moral integrity and the moral integrity of their society. Because deontological absolutists can readily grasp that the constables’ plight is a moral conflict of this kind, it is they—rather than consequentialists such as Posner and Vermeule—who come to grips with the searingness of tragic choices.

5.1.2.2.  Posner and Vermeule on the Legal Prohibition Thesis Although Posner and Vermeule do not entirely ignore deontological absolutism, their disdain for it leads them to be laconic in their treatment of it. They bestow far more attention on the Legal Prohibition Thesis, which is indeed their principal target (albeit not under that name, of course). Like Posner and Vermeule themselves, the proponents of the Legal Prohibition Thesis believe that the use of interrogational torture in circumstances of great urgency can be morally legitimate. They also agree that, in some extreme situations, no sanctions should be imposed in the aftermath of the wielding of such torture by public officials. However, whereas Posner and Vermeule favour the ex ante authorization of interrogational torture under certain conditions, the supporters of the Legal Prohibition Thesis favour a blanket ex ante legal prohibition combined with ex post accountability. 5.1.2.2.1.  Empirical conjectures Posner and Vermeule criticize the adherents of the Legal Prohibition Thesis for relying on empirical surmises. Those adherents call for a blanket ex ante legal ban on interrogational torture largely because they fear that any ex ante legal authorization of such torture would be abused. They hold that the sway of a blanket prohibition will help to ensure that officials never resort to interrogational torture except in the harrowing predicaments where the use of such torture is justified. Because the officials can expect to be held accountable ex post for what they have done, they will be loath to engage in torture in any situations that fall short of extreme desperation.

272

Legal Responses to Torture

Such is one of the paramount contentions advanced by the champions of the Legal Prohibition Thesis. In this chapter’s critique of Dershowitz, we have already encountered one set of queries that could be posed against such a position. That is, a critic of the Legal Prohibition Thesis might worry that the key role of ex post accountability will be undermined by phenomena such as jury nullification and prosecutorial inaction and official mendacity. Posner and Vermeule do broach such concerns (2006, 694–5, 703; 2007, 172–3, 186), but they are more worried that a blanket ex ante ban will deter officials from having recourse to interrogational torture in quandaries where the use of such torture would be justified. As they write (2006, 695; 2007, 173): ‘Excessive caution is the most likely result [of such a ban].’ At any rate, because the concerns about jury nullification and prosecutorial inaction and official mendacity would be applicable to Posner’s and Vermeule’s regulatory scheme as well as to the alternative approach favoured by the advocates of the Legal Prohibition Thesis, and because I have already addressed those concerns at some length in this chapter, we can leave them aside here. Posner and Vermeule are much more inclined to raise queries about the insolidity of the empirical claims that have been propounded by the exponents of the Legal Prohibition Thesis. They contend that the worries raised by those exponents about the effects of any ex ante authorization of interrogational torture are ‘speculation unencumbered by serious empirical support’ (2006, 687; 2007, 199). On that score, Posner and Vermeule are largely correct. Many of the empirical claims advanced by the supporters of the Legal Prohibition Thesis are conjectural. However, two points should be noted here in response. First, as has been suggested in § 3.1.3.3.8 of Chapter 3, Posner and Vermeule have left themselves vulnerable to a ‘tu quoque’ riposte. Although they are correct in taking exception to the scantiness of the evidence adduced by the champions of the Legal Prohibition Thesis, they themselves do not do any better. All of their empirical suppositions in their rejoinders to the Legal Prohibition Thesis are supported by very modest evidence indeed. For example, the claim about excessive caution which I have quoted in the penultimate paragraph above is a surmise for which Posner and Vermeule offer no evidence. Second, as has also been suggested in § 3.1.3.3.8 of Chapter 3, the empirical presuppositions of the Legal Prohibition Thesis are more plausible than those of the Slippery-Slope Illegitimacy Thesis. Posner and Vermeule are especially scathing in their comments on the meagreness of the empirical basis of the latter thesis, and the mordancy of their comments is warranted (as is suggested by my own dissection of the empirical presuppositions of that latter thesis in Chapter  3). By contrast, the empirical basis of the Legal Prohibition Thesis is quite credible. Far from implausible is the notion that the law’s inclusion of ex ante authorizations of interrogational torture would entice officials to employ such torture in an ever-expanding range of cases. Of course, that notion is an empirical hypothesis in need of empirical substantiation. Most of the proponents of the Legal Prohibition Thesis are open to censure for not having done nearly enough to substantiate that hypothesis. Nonetheless, Posner and Vermeule are at least as susceptible to censure

Legal Approval Ex Ante?

273

for the inadequacy of their own efforts to substantiate the empirical suppositions with which they contest the Legal Prohibition Thesis. Given the credibility of the empirical claims which underlie that thesis, the burden of proof resides on those who wish to contest its empirical soundness. Posner and Vermeule have not discharged that burden of proof. 5.1.2.2.2.  Exoneration is tantamount to approval Much more powerful is a different objection which Posner and Vermeule press against the Legal Prohibition Thesis. This latter objection pertains not to the empirical underpinnings of that thesis, but to the belief of its proponents that the wielding of interrogational torture by public-safety officials should sometimes not lead to the imposition of any sanctions. Notwithstanding that most of those proponents insist that any officials who have recourse to such torture should be held accountable ex post in legal proceedings, they allow that the proceedings should in some cases exonerate the officials and the relevant governmental agencies. After all, theorists who subscribe to the Legal Prohibition Thesis believe that some instances of interrogational torture in extreme emergencies can be strongly justifiable. Whenever an instance of such torture is strongly justified, they contend, the levying of legal sanctions would be unconscionable. In their view, no individually borne or collectively borne penalties should be imposed for actions— calamity-averting actions—that have not breached any moral duties whatsoever. All or most supporters of the Legal Prohibition Thesis presume that the upshot should sometimes be fully exonerative in legal proceedings which hold officials and governmental agencies accountable for any torturous sessions of interrogation. Posner and Vermeule trenchantly seize upon this corollary of the Legal Prohibition Thesis. Although they themselves of course believe that some instances of interrogational torture by officials should not trigger the imposition of any legal sanctions, they contend that such a position undermines the point of a blanket ex ante legal prohibition on such torture. Whereas the point of a comprehensive prohibition is to forbid legally the perpetration of interrogational torture—and to convey the clear message that the perpetration of such torture is legally forbidden—the availability of fully exonerative defences will in effect render the perpetration of such torture legally permissible in the cases where torturers and their governmental agencies can successfully invoke those defences. In other words, the rationale for the Legal Prohibition Thesis is at odds with the results favoured by its advocates in certain extreme cases. As Posner and Vermeule write with specific reference to the legal defence of necessity, ‘reliance on the necessity defense would not avoid legitimizing [interrogational torture]; it would avoid legitimating [such torture] only when it is not “necessary.” . . . The defense of necessity, like the defense of official immunity, renders legitimate those actions that fall within its scope’ (2006, 698; 2007, 176, footnote omitted). When legal defences such as that of necessity are successfully invoked by governmental agencies whose officials have resorted to interrogational torture, the exonerative outcomes of the legal proceedings in question will be matters of public knowledge. Officials and citizens—or their lawyers, at any rate—will know

274

Legal Responses to Torture

that interrogational torture can sometimes be practised with impunity even on occasions when its occurrence is not denied or concealed by anyone in the legal proceedings that follow. They will know, in short, that such torture is sometimes legally permitted. Despite the presence of a ban that putatively applies in a wholesale fashion, the use of interrogational torture is legally countenanced in certain extreme situations. That legal ban is thus hollow if not downright disingenuous. Posner and Vermeule state this point pungently, again with reference to the legal defence of necessity: The implicit theory of the advocates of the necessity defense [in the context of the Legal Prohibition Thesis] is that a statute that creates liability sends the message to the public, while the statute that provides a defense against liability remains silent. But as the public does not usually pay attention to the law on the books—and when it does, never discriminates between statutes that create liability and statutes that provide defenses—but instead observes police officers either being convicted of crimes or not being convicted of crimes, this theory is dubious. If the public is paying no attention to legal rules, and only looks at outcomes, it will just see interrogators going unpunished in a range of cases. If the public does pay attention to legal rules, why will it only pay attention to the ex ante prohibition and not the ex post license?11

What tends to support Posner’s and Vermeule’s animadversions here is that Fritz Allhoff and Adam Raviv, who independently and robustly support the use of interrogational torture in ticking-bomb cases, are in alignment with the proponents of the Legal Prohibition Thesis in favouring ex post legal accountability for torturers through proceedings where the necessity defence is available as a route of exoneration (Allhoff 2012, 188–94; Raviv 2004, 161–78). Largely because Posner and Vermeule are correct in holding that exoneration on grounds of necessity (or self-defence) is a form of condonation, full defences should not be invocable to block the imposition of collectively borne sanctions for acts of interrogational torture performed by constables or other officials. This chapter will say more on that point later. However, we should here note one significant operational difference between the ex ante authorizations commended by Posner and Vermeule and the ex post proceedings commended by the proponents of the Legal Prohibition Thesis. That operational difference pertains to the allocation of the burden of proof in some jurisdictions (such as that of California). Under the approach favoured by the advocates of the Legal Prohibition Thesis—and by Allhoff and Raviv—the burden of proof in the aforementioned jurisdictions will lie on officials to establish that the defence of necessity is applicable to their acts of interrogational torture. Under the approach favoured by Posner and Vermeule, contrariwise, the burden of proof lies on prosecutors or tort plaintiffs to establish that acts of interrogational torture carried out by officials are not covered by any ex ante authorizations. Although this operational difference between the former approach and the latter does not detract from the queries raised here by Posner 11   Posner and Vermeule 2006, 698. A nearly identical passage is contained in Posner and Vermeule 2007, 176.

Legal Approval Ex Ante?

275

and Vermeule about the partial hollowing out of the ex ante legal prohibition that is associated with the former approach, it can be quite significant as a practical matter. In a number of borderline cases, the direction of the burden of proof can determine the outcomes of legal proceedings. Posner’s and Vermeule’s approach would significantly lower the likelihood of the imposition of any sanctions in some borderline cases of interrogational torture. Still, notwithstanding that the champions of the Legal Prohibition Thesis are less supportive of torturers than are Posner and Vermeule, they are vulnerable to the objection recounted in this subsection. By contrast, the position espoused in this chapter is not touched by Posner’s and Vermeule’s exposure of the porousness of the ex ante legal ban that is envisaged by the proponents of the Legal Prohibition Thesis. As has already been indicated, this chapter will insist that sanctions should always be imposed in response to any acts of interrogational torture performed by legal-governmental officials. Though defences should be available to those officials and their governmental organizations in criminal or civil proceedings, the defences for the organizations should always be mitigative rather than exonerative. This position on the suitable legal responses to interrogational torture follows morally from Chapter  3’s account of the absolute wrongness of such torture. Because interrogational torture is morally wrong in all of its instances (including its morally optimal instances), and because it involves serious violence, the officials in a legal-governmental system cannot properly be allowed to perpetrate such torture with impunity. Were a legal-governmental system to forbear from imposing any sanctions in the aftermath of the wielding of interrogational torture by its officials, it would sully its moral integrity. Now, because this chapter adopts an absolutist stance on the need for sanctions in the wake of any acts of interrogational torture by officials, I am not susceptible to Posner’s and Vermeule’s indictment of the Legal Prohibition Thesis. Whereas the blanket legal ban commended by the advocates of that thesis is partly hollowed out by the availability of full defences for governmental agencies whose officials have resorted to interrogational torture in extreme emergencies, there is no similar hollowing out of the blanket legal ban commended by this book. Mitigative defences can take account of the extreme circumstances under which the perpetration of interrogational torture by officials may have occurred, without condoning the perpetration of such torture altogether in any instance. Thus, the comprehensive legal prohibition favoured by this book is indeed comprehensive. It leaves room for some flexibility in the specific means by which it is implemented, but it does not leave any room for its going entirely unimplemented through outright exonerations. Accountability without sanctions (whether individually borne or collectively borne) is never enough to preserve the moral integrity of a legal-governmental system whose officials have had recourse to interrogational torture. 5.1.2.2.3.  Not many operational differences In § 5.1.2.2.2 we have observed one significant respect in which the scheme of ex ante authorizations proposed by Posner and Vermeule would—in some jurisdictions— be operationally different from the scheme of ex post accountability advocated by

276

Legal Responses to Torture

most of the supporters of the Legal Prohibition Thesis. The direction of the burden of proof can be determinatively important in some cases. However, what should also be noted here is that the operational differences between the scheme of ex ante authorizations and the scheme of ex post accountability are otherwise very limited. After all, despite Posner’s and Vermeule’s emphatic preference for ex ante authorizations over ex post accountability, their scheme of ex ante authorizations clearly has to include ex post proceedings. Those proceedings will ascertain whether particular acts of interrogational torture by officials have occurred in compliance with the terms of the authorizations. As Posner and Vermeule assert: ‘Officers who employ [interrogational torture] measures in violation of these rules [which authorize the use of such measures within specified limits] should be punished in the same way that officers who violate the rules [concerning the use of deadly force] are punished’ (2006, 702; 2007, 212). Posner and Vermeule also refer to administrative sanctions that can be appropriate even when punishments are not imposed. Central to the implementation of Posner’s and Vermeule’s regulatory scheme for interrogational torture, then, are ex post proceedings to determine when punishments and administrative sanctions should be levied. Particularly noteworthy here is that those ex post proceedings would have to focus on essentially the same matters that are addressed in the ex post proceedings envisaged by the supporters of the Legal Prohibition Thesis. For example, when Posner and Vermeule recount the general ways in which their ex ante authorizations would place restrictions on the brutality of the methods of torture that are to be employed, they write that ‘a good rule would limit agents to the minimal amount of coercion that is necessary’ (2006, 701; 2007, 211). They likewise tentatively suggest the following rule to pick out the occasions on which the use of interrogational torture would be authorized: ‘[P]‌olice may use [interrogational torture] only when they are reasonably certain that an individual possesses information that could prevent an imminent crime that will kill at least n people.’12 In the ex post proceedings that would ascertain whether rules of these sorts have been followed, the inquiries posed would be essentially the same as those posed in ex post proceedings to determine whether instances of interrogational torture are covered by any full legal defences such as that of necessity. Unlike the individualized authorizations that are issued ex ante under a scheme of torture warrants, the authorizations established ex ante by Posner’s and Vermeule’s regulatory framework are general. Only in ex post proceedings would any individualized judgements about the lawfulness or unlawfulness of interrogational torture be reached. Thus, whereas ex post proceedings within a scheme of torture warrants are very likely to be perfunctory—at least in the absence of manifest transgressions of the warrants’ terms—the ex post proceedings contemplated by Posner and 12   Posner and Vermeule 2006, 701; 2007, 211. In their 2006 article, Posner and Vermeule assume that this rule—with its reference to ‘an imminent crime’ —would supplement a scheme of torture warrants that could deal with the averting of non-imminent crimes (Posner and Vermeule 2006, 702). In their 2007 book, however, they are considerably more wary of Dershowitz’s torture-warrant proposal (Posner and Vermeule 2007, 212).

Legal Approval Ex Ante?

277

Vermeule are likely to be considerably more exacting. Though the direction of the burden of proof in the latter proceedings will be favourable to the officials who have wielded interrogational torture, the proceedings are otherwise likely to be closely similar to those for which the proponents of the Legal Prohibition Thesis have called. Operationally, there is not a very sharp contrast between the Posner/ Vermeule regulatory scheme and the ex post accountability championed by the votaries of the Legal Prohibition Thesis. 5.1.2.2.4.  An objection of moral principle To be sure, the direction of the burden of proof is not the only point on which the Posner/Vermeule approach diverges operationally from the approach advocated by the supporters of the Legal Prohibition Thesis. Posner and Vermeule propose the institutionalization of interrogational torture in various respects. For example, they favour the training of constables in torturous techniques of interrogation, and they submit that monitoring by public-watchdog review boards will help to forestall abuses in the plying of such techniques (2006, 702–3; 2007, 212–13). Those proposals for the institutionalization of torture serve to underscore the key difference—a difference that is a matter of moral principle, rather than an operational difference—between the ex ante authorizations envisaged by Posner and Vermeule and the ex post accountability envisaged by the proponents of the Legal Prohibition Thesis. Here we return to the principal objection that I have raised against Dershowitz’s torture-warrant proposal in § 5.1.1.5. Like Dershowitz’s system of torture warrants, the regulatory scheme commended by Posner and Vermeule will formally bestow approval on techniques of torture as methods of lawful interrogation. Posner and Vermeule expressly design their scheme to facilitate the perpetration of interrogational torture under prescribed conditions, but my objection here is applicable even if the introduction of their regulatory framework would somehow miraculously not increase the incidence of such torture. Their framework lends the imprimatur of a legal-governmental system to a mode of conduct that is always and everywhere seriously wrong and violent. It lends the imprimatur of a legal-governmental system to a mode of conduct that is always degradingly hubristic in the manner recounted by Chapter 3. Formal authorizations of that mode of conduct are incorporated by their scheme into the standing law of a jurisdiction. As a consequence, the standing law as a whole will be sullied—just as it would be by the inclusion of formal authorizations of invidious racial or ethnic discrimination. Regardless of whether the formal authorizations championed by Posner and Vermeule will eventuate in more frequent acts of interrogational torture by a regime’s officials, they will alter the complexion of the regime. In a country whose system of governance adopts the set of regulations proposed by Posner and Vermeule, the system will have committed itself to the proposition that interrogational torture is a type of conduct that can properly be undertaken by public officials with impunity. Posner and Vermeule derisively ponder some arguments about the symbolic importance of laws that comprehensively forbid the use of interrogational torture. In response, we should first note that the objection outlined here in the preceding

278

Legal Responses to Torture

couple of paragraphs does not pertain exclusively—or even predominantly—to the symbolic import of the ex ante authorizations for which Posner and Vermeule call. As I have remarked in connection with Dershowitz’s torture-warrant proposal in § 5.1.1.5, the morally corrupting effect that has been highlighted here would occur even if the ex ante authorizations were never publicized and never implemented. In those circumstances the symbolic import of the authorizations would be negligible, for they would not convey any unsavoury message to members of the public. Nonetheless, the incorporation of those authorizations into the law of a jurisdiction would taint the law in broadly the same way in which a person’s outlook is tainted by her harbouring of morally vicious attitudes (even if those attitudes never gain expression during her interaction with others). The chief wrong is that of authoritatively committing a system of legal governance to the approval of a seriously immoral mode of conduct; the system’s moral integrity is degraded by its being committed to the approval of such conduct, whether or not its degrading laws have become publicly known and have thus acquired symbolic significance. Hence, even if Posner and Vermeule were correct in deriding their opponents’ worries about the symbolic import of ex ante authorizations of interrogational torture, my objection of moral principle to those authorizations would stand. 5.1.2.2.5. Further remarks on the symbolism of authorizations of interrogational torture In fact, moreover, Posner’s and Vermeule’s criticisms of the complaints about the deleterious symbolism of their proposed laws are far from unanswerable. For example, some of their criticisms depend squarely on the proposition that the plying of interrogational torture for the purpose of fending off calamities can be morally legitimate. Admittedly, their reliance on that proposition is quite understandable in the dialectical context, for their immediate opponents (namely, the advocates of the Legal Prohibition Thesis) agree that calamity-averting interrogational torture can be morally legitimate. However, insofar as Posner and Vermeule presuppose the permissibility of such torture, their lines of reasoning are otiose against deontological absolutists. For instance, in responding to the claim that a blanket legal prohibition on the use of interrogational torture can play a salutary educative role, Posner and Vermeule retort: ‘But if [interrogational torture]—like imprisonment, or police shootings—is justified, and thus consistent with our values, then we shouldn’t want to teach people that [interrogational torture] is wrong; quite the contrary’ (2006, 691). Though such a retort might to some degree be effective against the proponents of the Legal Prohibition Thesis, it is altogether ineffective against a deontological absolutist. A  comprehensive legal ban on the use of interrogational torture can indeed play a salutary educative role, because it correctly conveys the message that such torture is always and everywhere morally wrong. My discussion of Posner and Vermeule is not focusing principally on the symbolic or educative function of a ban, but the pertinence of such a function is readily recognizable within my robustly deontological account. That pertinence is due to the congruity between the absoluteness of the message communicated

Legal Approval Ex Ante?

279

by a comprehensive ban and the absoluteness of the impermissibility of interrogational torture. In short, this book’s deontological absolutism can parry most of the lines of reasoning with which Posner and Vermeule oppugn the symbolic import of any legal mandates that altogether proscribe the use of interrogational torture. Their remaining lines of reasoning on the topic damagingly conflate moral optimality and moral permissibility. Posner and Vermeule assume that, whenever a theorist maintains that the use of interrogational torture is always morally impermissible, she is thereby committed to maintaining that the use of such torture is never morally optimal. They assume in effect that the absolutism of an unflinchingly deontological stance on torture must be strong rather than merely weak. Their assumption on that point becomes especially clear when they are retorting to Waldron’s thesis that a comprehensive legal prohibition on interrogational torture epitomizes the eschewal of brutality in the law of the United States: The commitment to minimize law’s brutality is on both sides of this argument. Where [interrogational torture] can save lives, not engaging in it might seem the more brutal choice, especially to those whose lives are at stake. Those people might reasonably hold that there is a sort of brutal callousness, a self-absorbed moral preciosity, in the decision to preserve the law’s archetypal integrity by permitting third-party deaths to go unprevented. (2006, 692; 2007, 205, footnote omitted)

Worth noting briefly here is the way in which Waldron’s deontological focus on the avoidance of brutality is transmuted by Posner and Vermeule into a consequentialist focus on the minimization of brutality. Also remarkable is the Orwellian way in which the placing of limits on the range and harshness of lawful techniques of interrogation is itself portrayed by Posner and Vermeule as brutal. (Their Orwellian distortion trades on a baseless slur, as they declare that anyone who acknowledges the moral obligatoriness of limits on the range and harshness of techniques of interrogation is thereby displaying indifference toward the interests of people who might be murdered by terrorists or criminals.) However, the main feature of this passage that should be assailed here is its implicit elision of the distinction between moral optimality and moral permissibility—along with its concomitant elision of the distinction between strong absoluteness and weak absoluteness. Given that deontological absolutists contend that the use of interrogational torture is never morally permissible and should never be legally permissible, Posner and Vermeule infer that those absolutists must also contend that the use of such torture is never morally optimal. Yet, as is evident from Chapter 3 and some other parts of this book, Posner and Vermeule are quite mistaken in drawing such an inference. Someone who espouses a weakly absolutist position on the morality of interrogational torture will maintain that such torture is always and everywhere morally impermissible, and that it should always and everywhere be legally impermissible; but he or she will also accept that the perpetration of such torture can be morally optimal in certain imaginably extreme situations. In any of those imaginably extreme situations, refraining from the administration of such torture would be an even graver wrong than administering it. This point about the possibility of the moral optimality of interrogational torture in

280

Legal Responses to Torture

circumstances of dire desperation is consistent with an absolutist stance (a weakly absolutist stance) on the wrongness of such torture. It is thus consistent with the proposition that the use of interrogational torture should always and everywhere be legally forbidden, and with the proposition that every act of such torture by public officials should trigger the imposition of legal sanctions. By embracing that latter proposition, of course, my absolutist stance goes beyond what the advocates of the Legal Prohibition Thesis recommend. It does so while also squarely engaging with Posner’s and Vermeule’s concerns about the interests of people who might be murdered by terrorists or criminals. In short, once we take account of the distinction between strong absoluteness and weak absoluteness, we can recognize how Posner’s and Vermeule’s concerns should be addressed without any discontinuation of a comprehensive legal ban on the wielding of interrogational torture. Waldron is correct to insist on such a ban and to emphasize its great symbolic significance. Moreover, pace the champions of the Legal Prohibition Thesis, that ban on interrogational torture should always be given effect when it is contravened by public officials. Any legal defences available to the officials’ organizations should be mitigative rather than fully exculpatory. Still, although the defences should never be fully exculpatory, they should be heavily mitigative in every case where the recourse by officials to interrogational torture as a desperate last resort has been morally optimal. One’s alertness to the moral conflict that has confronted the officials in such a case—one’s awareness that, by perpetrating a serious moral wrong through their recourse to interrogational torture, the officials have avoided an even more serious moral wrong of fatal remissness—is the key to grasping the importance of extenuative defences in legal proceedings that deal with instances of calamity-averting interrogational torture. Posner and Vermeule and the proponents of the Legal Prohibition Thesis have all written as if any such proceedings must lead either to maximal sanctions or to outright exonerations for the organizations involved. We should reject that stark dichotomy, since outright exonerations would not do justice to the absoluteness of the moral prohibition on interrogational torture, and maximal sanctions would not do justice to the moral optimality of some transgressions of that prohibition. Mitigated sanctions, in appropriate cases, reflect both the absoluteness of the prohibition and the exigencies of some of the situations in which the prohibition is breached by law-enforcement officials.

5.1.2.3.  The analogy between torture and killing Posner and Vermeule repeatedly invoke an analogy between interrogational torture and deliberate killing, to support their proposed regulatory scheme for the employment of such torture by public officials. They deliberately model their scheme along the lines of existing frameworks for the regulation of the exertion of lethal force by constables and other officials. On that very point, they make clear their intentions near the outset of their discussion of interrogational torture: Police are allowed to use deadly force in order to prevent dangerous suspects from harming other people. Killing a person is . . . a serious harm to dignity and autonomy; although

Legal Approval Ex Ante?

281

we will see arguments holding that [interrogational torture] is worse than killing in some respects, there are other respects in which killing is worse than [interrogational torture]. To prevent officials from engaging in unjustified killings, governments take the conventional route of enacting laws that describe the conditions under which a police officer may use deadly force, making the police liable only if they violate these rules in bad faith. Why shouldn’t the same system be used for [interrogational torture]? (2006, 673–4; 2007, 184)

Posner and Vermeule hark back to this theme throughout their discussion. With their analogy between interrogational torture and deliberate killings, Posner and Vermeule are making the same general missteps which I have criticized in the work of Helen Frowe in § 3.2.3.3.5 of Chapter  3. Most notably, they are conflating the matter of the relative gravity (or relative harmfulness) of certain wrongs and the matter of the range of circumstances in which certain modes of conduct are wrong. Posner and Vermeule are correct in thinking that some instances of interrogational torture are less gravely wrongful than some instances of deliberate killing. They are likewise correct in thinking that acts of interrogational torture often are less harmful than are acts of deliberate killing. However, when we are pondering whether interrogational torture and deliberate killing should be legally regulated in parallel ways or in divergent ways, we should not be focusing on the factor of comparative gravity or comparative harmfulness. If we do concentrate on either of those factors, we will be led to conclude with Posner and Vermeule that the scheme of regulation for interrogational torture should be largely the same as that for deliberate killing. Such a conclusion is inapposite because it derives from posing an inapposite question. Instead of asking whether instances of interrogational torture are more gravely wrong (or more harmful) than instances of deliberate killing, we should be asking whether the wrongness of the act-type of interrogational torture is absolute and whether the wrongness of the act-type of deliberate killing is absolute. Now, as has been argued in Chapter 3, the answer to the first of the latter pair of questions is affirmative while the answer to the second of that latter pair of questions is negative. Whereas the employment of interrogational torture is always and everywhere morally wrong (even when it is morally optimal), some acts of deliberate killing are morally permissible. Consequently, a legal framework that can suitably regulate deliberate killings by constables and other officials will not be properly transferable to the context of interrogational torture. That framework, which authorizes ex ante the use of lethal violence by law-enforcement officials in certain situations, correctly presupposes that the exertion of such violence in those situations is morally legitimate and that it should therefore be legally permissible. Given that the administering of interrogational torture is never morally legitimate, the law’s handling of the perpetration of such torture should not be modelled on the framework just mentioned. A legal-governmental system should not introduce a scheme of regulation with ex ante authorizations which presuppose that the administering of interrogational torture is sometimes morally permissible and that it should therefore sometimes be legally permissible. Posner and Vermeule of course disagree with me about the morality of interrogational torture. Like other consequentialists, they conflate moral optimality

282

Legal Responses to Torture

and moral permissibility. They believe that, on any occasion when the wielding of such torture is morally optimal, it is morally permissible as well as morally obligatory. Accordingly, they believe that the wielding of interrogational torture should be legally permissible on any such occasion. Hence, in their view, the analogy between interrogational torture and deliberate killing is fully sustainable. Yet, whenever Posner and Vermeule seek to vindicate that analogy, they refer to the matter of comparative harmfulness or comparative gravity—and they thus neglect the question of the range of circumstances in which the act-type of interrogational torture or the act-type of deliberate killing is morally wrong. Moreover, when pondering the considerations that underlie the wrongness of interrogational torture, Posner and Vermeule take account only of victim-focused factors (and of causally concomitant factors such as the dismay and apprehension that might be felt by members of the public). They do not take account of any perpetrator-focused considerations, because those considerations do not register in a utilitarian calculation of harms and benefits. When we move beyond the confines of the consequentialism in which Posner and Vermeule are ensconced, and when we attend to the moral integrity of the officials who perpetrate interrogational torture and to the moral integrity of the society on whose behalf the officials act, we can recognize that the analogy between interrogational torture and deliberate killing is untenable. Similarly, when we are alert to the occurrence of moral conflicts, we can recognize that some morally optimal courses of conduct are morally impermissible. Whereas all possible instances of interrogational torture (including any morally optimal instances) are morally wrong, some possible instances of deliberate killing are morally legitimate. Such is one of the main conclusions for which Chapter 3 and other parts of this book have argued. That conclusion and the arguments underpinning it are not addressed to the matter of comparative gravity or comparative harmfulness. Rather, they are addressed to the matter that is decisive here: namely, the range of circumstances in which the act-type of interrogational torture or the act-type of deliberate killing is morally illegitimate. Because that range is all-encompassing for the former act-type and is less than all-encompassing for the latter, we cannot properly extrapolate from the legal regulation of the latter act-type to the legal regulation of the former.

5.1.3.  Legitimate techniques of interrogation As has just been argued, interrogational torture should never be legally countenanced through any ex ante authorizations. However, what should indeed be subject to ex ante legal approval are sundry morally legitimate techniques of interrogation. After all, the ends pursued through interrogation are often themselves legitimate and very important, and interrogation is not per se a morally illegitimate activity. Accordingly, constables and soldiers and other officials should of course be legally entitled to interrogate suspects. Also entirely sensible is the adoption of guidelines which commend certain techniques of interrogation as permissible and which disallow various other techniques as impermissible. Any such guidelines

Legal Approval Ex Ante?

283

should of course be drawn up by people who are attuned to the strict impermissibility of torturous techniques of interrogation. Hardly any theorists would dispute the need for officials to exercise interrogational powers, or the need for guidelines to direct interrogators toward certain methods of interrogation and away from other methods. Somewhat more controversial, however, are questions about the degree of specificity with which the drafters of guidelines should seek to differentiate between legitimate methods of interrogation and torturous methods. On the one hand, there are clearly some respects in which very little specificity is required. A general definition of torture, such as the one presented at the end of Chapter 2 of this book, should of course be included in any instructions for interrogators. In addition, quite a few techniques of interrogation should be forbidden outright. No manual of interrogation could ever properly specify some amount of whipping or beating or water-boarding as an acceptable amount, for example. Any such specification would corrupt, rather than enhance, the guidelines in which it appears. In that vein, the endeavours by some contemporary scholars and politicians to distinguish between torture and torture-lite have been corrupting. On the other hand, there are a number of respects in which a high degree of specificity can be salutary. Some techniques of interrogation are morally legitimate at low levels, even though they would become torturous if their severity were greatly augmented. For example, mild sleep-deprivation can be a legitimate tack through which interrogators keep a suspect on edge and thus induce him to be less adroitly guarded in his answers to questions. By contrast, severe sleep-deprivation is a classic technique of torture. Given as much, a clear limit—at a mild level— should be placed on the extent to which any suspect can be lawfully deprived of sleep. Likewise, the guidelines should prescribe a ceiling on the duration of the interrogational sessions to which anyone can be subjected over rolling periods of time.

5.1.3.1.  Waldron’s distrustful arguments Waldron has doubted whether specificity in steering interrogators away from torturous measures—and from cruel, inhuman, or degrading measures—is ever advisable (2010a, 200–7; 2010b, 20–5). Writing in the aftermath of the attempts by lawyers in the administration of President George W. Bush to loosen the legal prohibition on torture through the attachment of greater specificity to the terms of that prohibition, Waldron is distrustful of any moves toward concreteness in the laws that regulate the harshness of interrogational techniques. He initially and somewhat tentatively contends that interrogators in their roles as interrogators are not entitled to the same degree of solicitude as ordinary citizens whose felt leeway in performing certain activities will be chilled if legal prohibitions that might apply to those activities are formulated very abstractly. At any rate, he submits, the government on whose behalf the interrogators carry out their labours is not entitled to such solicitude. However, Waldron relies principally on two additional arguments.

284

Legal Responses to Torture

First, he maintains that the soldiers and law-enforcement officials who administer interrogations cannot credibly claim to be surprised if brutal methods of dealing with suspects are deemed by courts to be torturous. If the soldiers or officials indulge in practices such as water-boarding or setting dogs upon prisoners or squeezing the prisoners’ genitals, they realize that they are afflicting their interrogatees with severe pain and fear. They perform those actions precisely because they recognize that their interrogatees will undergo severe pain as a result. Hence, unless the soldiers or officials are too obtuse to be entrusted with their responsibilities, they have ample grounds for knowing that they are contravening the terms of a legal mandate which prohibits the deliberate inducement of severe pain for a purpose such as interrogation. Although the mandate is formulated abstractly, its applicability to the pain-inducing measures just mentioned is palpable. Second, Waldron argues that a high degree of concreteness in the terms of an interdictory legal mandate LM is desirable only when the people covered by LM have legitimate interests in knowing quite precisely the point of transition between lawful behaviour and the unlawful behaviour which transgresses LM. If the mode of conduct forbidden under LM is on a spectrum with conduct that is unexceptionable, then people disposed to engage in the latter conduct will have legitimate interests in being told how far they can go. Waldron offers an example of a taxpayer who wants to arrange her business in ways that will minimize her tax liability; she needs to know quite precisely where her lawful tax-avoidance would become unlawful tax-evasion. Another example is that of a motorist who has a legitimate interest in being apprised of the speed at which her driving would be classifiable as unlawfully fast. Waldron contrasts those two examples with two others in which the people involved do not have any legitimate interests in being told exactly how far they can take their conduct without running afoul of any legal mandates. One of the latter examples is that of a husband who wants to push his wife around without violating any laws against domestic battery, and the remaining example is that of a professor who wants to interact with his students in a sexually charged manner while not transgressing any rules against sexual harassment. Neither the husband nor the professor has any legitimate interest in knowing precisely where he will encounter a legal limit on his behaviour. As Waldron writes:  ‘There are some scales one really shouldn’t be on, and with respect to which one really does not have a legitimate interest in knowing precisely how far along the scale one is [legally] permitted to go’ (2010a, 205, italics omitted; 2010b, 23–4). Waldron holds that any bright-line rules to differentiate between torturous techniques and lawful techniques of interrogation would be as misconceived as any bright-line rules that inform husbands how far they can lawfully go in pushing their wives around (or professors how far they can lawfully go in flirting with their students). Someone who calls for much greater specificity in the legal prohibition on torturous methods of interrogation is presupposing a scale or continuum where ‘[i]‌nterrogators have an interest in being as coercive as possible and in being able to inflict as much pain as possible short of violating the prohibition on torture’ (Waldron 2010a, 205). Waldron declares that interrogators’ endeavours should not be located on such a scale or continuum at all.

Legal Approval Ex Ante?

285

Waldron does recognize, however, that the relevant scale or continuum could be characterized quite differently. He posits the following line of reasoning that might be advanced by people who believe that the terms of the legal ban on torture should be endowed with greater specificity: [T]‌he point of interrogation is to get people to do what they don’t want to do and for that reason pressure of some sort is necessary, to elicit information that the subject would rather not reveal. Since interrogation as such is not out of bounds, it may be thought interrogators obviously do have a legitimate interest in being on a continuum of pressure and it is just a question of how far along that continuum we ought to allow them to go. If we fail to specify that point, we might chill any use of pressure in interrogation, even what might turn out to be legitimate pressure. (2010a, 205, emphases in original)

Waldron realizes of course that anyone who undergoes a session of interrogation will thereby have been pressured to impart information which he or she would prefer not to disclose. However, he retorts to the line of reasoning in this quoted passage by distinguishing between pressure that involves non-painful coercion and pressure that involves brutality and the deliberate infliction of pain. He maintains that that distinction is qualitative rather than quantitative. Techniques that bring to bear pressure of the former kind—techniques such as threatened or actual incarceration, or lengthy and unremitting sessions of interrogation—are not on a continuum with techniques that bring to bear pressure of the latter kind. Whereas methods of interrogation that bring to bear pressure of the former kind are consistent with the human dignity of the interrogatees, methods that bring to bear pressure of the latter kind are violative of that human dignity. Waldron thus argues as follows. Because interrogational approaches are separated into two principal categories by a qualitative divide, the spectrum that leads into the truly torturous approaches is not a continuum on which the efforts of any interrogators should ever be located. Accordingly, there are no grounds for the introduction of any legal standards that would specify quite precisely a point of transition between the lawful and the unlawful on that spectrum. Instead, the legal prohibition on the use of interrogational torture should be formulated abstractly—in the manner of the definition at the end of Chapter 2—to indicate that any interrogational technique in the vicinity of such torture is to be eschewed. Legislators should not help a husband to pinpoint the degree of violence that he can lawfully exert against his wife, and they should likewise not help interrogators to pinpoint the severity of any pain that can lawfully be inflicted on suspects.

5.1.3.2.  A partial rejoinder to Waldron As should be evident from the opening paragraphs in § 5.1.3, I concur with much of what Waldron says on this matter. There are some continua on which the labours of interrogators should never be located; the legal standards concerning torture should not be drafted in ways that invite interrogators to proceed along those continua. For example, the law should never specify some amount of whipping

286

Legal Responses to Torture

or beating or water-boarding as a lawful level that falls short of torture. A general definition of torture in the law can serve to steer interrogators away from such dubious practices. However, Waldron sweeps too broadly with his remarks, and his dichotomy between non-painful coercive pressure and painful pressure is somewhat simplistic. Though that dichotomy is pertinent in application to tactics such as whipping or beating or water-boarding—tactics that should be altogether eschewed rather than used in moderation—some other methods of interrogation can be morally legitimate at low levels even though they would be torturous at higher levels. As has already been indicated, sleep-deprivation is one of those methods. If those latter techniques are to be controlled appropriately, an abstract definition of the crime of torture should be supplemented in the law with some much more concrete guidelines to fix clearly a ceiling on the extent to which each of those potentially legitimate techniques may be employed. By indicating that sleep-deprivation (or any relevantly similar tack) is lawful only at the low levels where it is clearly non-torturous and non-cruel, the guidelines can underscore the forbiddenness of interrogational torture and can help to ensure that the regulated method is employed only with the moderation that renders it morally legitimate. In short, Waldron errs by contending that specificity is never desirable in the legal regulation of torture. Probably because he is preoccupied with the machinations of the Bush Administration during the first decade of the twenty-first century, he presumes that specificity in this area of the law is always aimed at attenuating the restrictions on the leeway of interrogators. That wary assumption is correct with reference to many interrogational practices, but it is not correct across the board. Sometimes specificity is the best way of shoring up the distinction between non-painful pressure and torturous pressure; it is sometimes the best means of restraining the former from growing into the latter. Consequently, pace Waldron, the legal standards for averting the use of interrogational torture should be a mixture of generality and specificity. On the one hand, a general definition of torture is essential (in combination with similarly general formulations concerning cruel, inhuman, or degrading treatment). The embedding of that definition in a legal prohibition can salutarily disallow altogether many practices that might be adopted by interrogators. Its chilling effect is to be welcomed rather than regretted. On the other hand, the general prohibition on the use of interrogational torture should be accompanied by legally specified limits on certain practices that can be legitimate if they are strictly confined. Not every pressure-applying technique should be disallowed altogether, even if a technique would become torturous at high levels. Instead of embracing Waldron’s broad-brush approach, the law in any jurisdiction should differentiate between interrogational methods which at very low levels are already torturous (albeit perhaps incipiently torturous) and interrogational methods which at very low levels are not torturous—or cruel or inhuman or degrading—at all. In regard to the latter methods, some precise and stringent ceilings should supplement the general prohibition on the wielding of torture for interrogational purposes.

Legal Accountability Ex Post

287

5.2.  Legal Accountability Ex Post Interrogational torture and most other types of torture should never be authorized ex ante. How should a system of legal governance deal ex post with any acts of torture that do occur? To a large degree, my answers to this question are apparent from earlier chapters and from earlier portions of this chapter. In the first place, we need to contemplate the two types of torture—ephemerally incapacitative torture and edifying torture—that can sometimes be instantiated legitimately. For each of those two kinds of torture, ex ante authorizations are sensible.13 Those authorizations should indicate the general sorts of circumstances under which the use of ephemerally incapacitative torture or edifying torture will not trigger any legal sanctions. (In most cases to which the authorizations are applicable, the imposition of sanctions would be inappropriate because the use of ephemerally incapacitative torture or edifying torture is morally permissible. In cases of sado-masochistic torture to which the authorizations are applicable, however, the inappropriateness of the levying of sanctions is instead due to the illiberal intrusiveness of subjecting people to punishments for engaging in fully consensual interaction that has not produced any permanent and very serious injuries.) Ex post accountability is crucial as well, of course, but in each case the legal proceedings that provide such accountability should concentrate—at least initially—on the question whether the conditions for the applicability of any ex ante authorizations are satisfied. If those conditions are indeed satisfied, then no sanctions are to be imposed. By contrast, if the conditions for the applicability of the ex ante authorizations are not fulfilled, the legal proceedings should turn to the question of the gravity of the wrong that has been perpetrated through the undertaking of the torture. Sanctions should be imposed for any acts of ephemerally incapacitative torture or edifying torture to which the ex ante authorizations are inapplicable, and they should likewise be imposed for virtually all instances of every other kind of torture that has been discussed in this book. The other kinds of torture are morally wrong in every one of their instances, and they are not within the scope of the considerations that militate against the imposition of legal sanctions in certain cases of sado-masochistic torture. Nonetheless, as will be argued, some calamity-averting instances of a few of those other kinds of torture carried out by private individuals are not properly subject to legal sanctions even though they are morally wrong. Still, all instances of those other kinds of torture perpetrated by public officials and nearly all instances of those other kinds of torture perpetrated by private individuals are properly subject to such sanctions.

13   The authorizations need not be legislative in origin. They can emerge as judicially or administratively formulated doctrines. Their contents, if they are suitably drafted, can be gathered from what I  have already said in Chapters  2 and 3 about permissible instances of ephemerally incapacitative torture and edifying torture.

288

Legal Responses to Torture

As has already been remarked, the specifics of the suitable sanctions will vary considerably from case to case. Obviously, the severity of the sanctions should be proportional to the gravity of the wrongs that have been perpetrated. In addition, the kinds of sanctions that are most pertinent—the fittingness of criminal punishments or civil liability or institutional penalties or some combination thereof— will hinge on the gravity of the wrongs and on other variable factors in a myriad of circumstances. Similarly variable are the answers to the question whether the sanctions should be imposed only on individuals or only on collectivities or both on individuals and on collectivities. Although the details of the multitudinous considerations that bear on these features of sanctions are well beyond the scope of the present book, this chapter will provide some brief general reflections on those considerations.

5.2.1.  Which defences? As has become apparent in my discussions of the Legal Prohibition Thesis in this chapter and in Chapter 3, the proponents of that thesis go astray when they submit that some instances of interrogational torture by public officials should not trigger the imposition of any sanctions. By maintaining that the governmental organizations whose officials have avowedly resorted to such torture should sometimes be fully exonerated in the legal proceedings through which they are held accountable ex post, the proponents of the Legal Prohibition Thesis in effect condone the use of such torture by those officials. Posner and Vermeule are correct in their critique of the Legal Prohibition Thesis on that point, notwithstanding that their own preferred approach to the matter is even worse.

5.2.1.1.  Mitigations rather than justifications or excuses Like other consequentialists, the advocates of the Legal Prohibition Thesis conflate moral optimality and moral permissibility. Having grasped that the recourse of officials to interrogational torture in certain extreme emergencies can be morally optimal, they presume that the use of such torture by the officials in those emergencies is morally permissible. They consequently conclude that nothing properly subject to legal sanctions will have occurred. Having resisted their conflation of moral optimality and moral permissibility throughout, this book naturally rejects their conclusion about the inappropriateness of legal sanctions. Although the officials who wield interrogational torture in an extreme emergency have acted to avert an especially grave breach of their public-safety responsibilities, they have thereby perpetrated a very serious and violent wrong; the moral optimality of their conduct does not eliminate the wrongness of what they have done or the requisiteness of legal sanctions in response. They have had to adopt a morally wrongful means to a morally worthy and vital end. Hence, if no legal sanctions are imposed— at least on the organizations in which the officials are employees—the presiding system of governance will have countenanced the seriously wrongful means to which the officials have resorted. The moral integrity of the officials, and the moral

Legal Accountability Ex Post

289

integrity of the system of governance on behalf of which they have acted, will remain tarnished. Still, although the legal proceedings in a case of this sort should never be fully exonerative (provided, of course, that the perpetration of interrogational torture by some officials has been shown through evidence that meets the prevailing standard of proof ), the direness of the predicament that has led to the use of interrogational torture should of course be taken into account—especially at the final stage of the proceedings, when a punishment or some other remedy is prescribed. The exigencies of the aforementioned predicament are to be deemed a substantially mitigating factor that should lessen the heaviness of any punishment or other remedy. In addition, those exigencies can decisively bear on the question whether certain types of remedies are suitable at all. Because the defences properly available to the employers of public officials in these cases are extenuative rather than fully exculpatory, the traditional distinction between justifications and excuses is not really applicable to the pleadings of such parties. Crucial though that distinction is in some other contexts,14 the mitigative defences that are operative in cases of interrogational torture are such as to straddle the divide between justifications and excuses. That divide, as it has generally been understood, runs as follows. If conduct is justified, then it is at least permissible (morally and legally) and is perhaps also laudable; someone engaging in such conduct has not thereby done anything wrong. Contrariwise, if the conduct of some person is excusable, he has done something wrong (morally and legally) but is absolved of responsibility for it. Justifiable conduct consists in responsibility without wrongdoing, whereas excusable conduct consists in wrongdoing without responsibility. An example of justified conduct is the action of a policeman in restraining a culprit who has committed a misdeed, while an example of excusable conduct is the action of some person P who punches another person because a third party has credibly threatened P with death or severe harm if he does not inflict the blow. Justifications and excuses, as they have traditionally been understood, are fully exonerative. Somebody who successfully invokes a justification or an excuse as a defence is to be acquitted. (Elsewhere—in Kramer 2004, chap. 8—I have argued that excuses as fully exculpatory defences are operative only in the domain of law and never in the domain of morality. However, because the present chapter is concentrating solely on legal defences, my references to excuses here are unproblematic.) Although some jurists and philosophers differentiate between full excuses and partial excuses, we are best advised to embrace the traditional understanding of excuses as fully exculpatory defences. Justifications and excuses, as fully exonerative defences, are thus to be distinguished from mitigative or extenuative defences. A key reason for forgoing the distinction between full excuses and partial excuses, and for classifying partial excuses instead as mitigative defences, is that 14   For a good recent discussion, see Simester 2012. See also Simester 2008. The distinction between justifications and excuses is of central importance in a previous essay of mine; see Kramer 2004, chap. 8. I there cite many other writings that deal with the justification/excuse dichotomy.

290

Legal Responses to Torture

those latter defences cut across the justification/excuse division. On the one hand, in a quasi-excusatory role, the circumstances of extreme desperation that lead to the use of calamity-averting interrogational torture will have lessened the culpability of any of the officials involved. Although the officials have resorted to a seriously wrongful mode of conduct, they have done so under extreme duress and with benign objectives. Their responsibility for what they have done is attenuated by the urgency of the plight in which they have done it. On the other hand, in a quasi-justificatory role, the very features of the emergency that diminish the responsibility of the officials will also have reduced the gravity of the wrongfulness of their actions. Of course, the wrongfulness is not eliminated entirely; after all, the calamity-averting interrogational torture is an impermissible means to a commendable end. However, the magnitude of the wrongfulness is diminished by the ghastliness of the quandary in which the officials have acted. Because the mitigating features of an extreme emergency play both a quasi-excusatory role and a quasi-justificatory role, the legal defences that pick out those features are best not classified on either side of the justification/excuse dichotomy. They are best understood as belonging to a third category of defences. Thus, when we ponder which legal defences can appropriately be invoked by organizations whose officials have resorted to calamity-averting interrogational torture, we should not be seeking to squeeze those defences into a justificatory mould or an excusatory mould. We should be considering them solely as pleas in mitigation.

5.2.1.2.  Two preliminary caveats Before we turn to the matter of the legal defence that can most pertinently be invoked as a plea in mitigation, we should note two important preliminary points. First, my discussion here pertains only to the actions of public officials in their public capacities. This chapter will later deal separately with the actions of private individuals. Second, my discussion in § 5.2.1.3 pertains only to situations in which public officials have had recourse to interrogational torture out of desperation— after any feasible alternatives have failed—for the purpose of averting calamities. The mitigative defence delineated in that subsection does not cover any situation in which interrogational torture has been plied for a purpose other than the avoidance of some imminent calamity. Nor does it cover any situation in which interrogational torture has been administered before every less invasive alternative has been tried. Nor does that defence cover a situation in which torture has been administered despite the absence of any significant likelihood that it will succeed in achieving the objective for which it is undertaken. If public officials or private individuals have availed themselves of interrogational torture in circumstances where they are not seeking to fend off an imminent calamity or where some less invasive alternatives have been left untried or where the probability of the torture’s efficacy is very low, the mitigative defence outlined in § 5.2.1.3 is inapplicable to their actions. Even if the objectives pursued through the interrogational torture are benign, the extenuative force of the exigencies of a dire emergency is not operative in any of the circumstances just mentioned. Still,

Legal Accountability Ex Post

291

we should not conclude that extenuation occurs only in cases where the use of calamity-averting interrogational torture by officials is morally optimal. Although the pressures of an emergency in which any officials ply such torture are more powerfully mitigative when the torturous actions are morally optimal, the urgency of the situation can carry some mitigative force even when the recourse to torture falls short of being morally optimal. Given that the stringency of each of the clashing duties in a moral conflict is a scalar property, the stringency of the duty breached by the perpetration of calamity-averting interrogational torture in any particular case might exceed the stringency of the public-safety duty that has been fulfilled through the perpetration of such torture. Even so, the fact that the torture has served to fulfil the latter duty is a consideration that extenuates the gravity of the violation of the former duty.

5.2.1.3.  Necessity or protection-of-oneself-or-others? Because mitigative defences do not fit within the justification/excuse dichotomy, we can ruminate on the specifics of those defences without pondering whether they should be placed on one side or the other of that dichotomy. In discussions of calamity-averting interrogational torture among proponents of the Legal Prohibition Thesis and others, the two most frequently mentioned legal defences are necessity and self-defence. Each of those defences is normally classified as a justification and is thus treated as fully exonerative. However, insofar as either defence can suitably be invoked by organizations whose officials have resorted to interrogational torture, it is only mitigative rather than fully exonerative. Accordingly, as we mull over the two defences briefly here, we should of course take each of them to be mitigative rather than justificatory. (In addition to reconstruing the defences substantively in this fashion, I here introduce an appellational alteration. Because references to the legal defence of self-defence are stylistically awkward, and because that defence is generally understood to include the protection of other people from imminent harm, I henceforth designate it as ‘protection-of-oneself-or-others’.) Which of the two aforementioned defences is to be preferred? As several other commentators have observed (Bobbitt 2008, 389–90; Gur-Arye 2004, 191–5; Kutz 2007a, 244; Moore 1997, 711–19; Simester 2008, 303–10), the defence of protection-of-oneself-or-others is singularly appropriate. Its orientation and its limits salutarily differentiate it from the defence of necessity. For one thing, the defence of protection-of-oneself-or-others is applicable only when somebody has acted to save himself or other people from undergoing some imminent and very serious harm. Consequently, in a case where interrogational torture has been employed, that defence is applicable only if the purpose of the torture was calamity-averting. Interrogational torture conducted for any other purpose—such as the investigation of past crimes or the disruption of terrorist networks—is not within the scope of this mitigative defence. Furthermore, the defence of protection-of-oneself-or-others is applicable only if the interrogational torture has been wielded against a malefactor who has brought about the extremely dangerous predicament in which the torture has been administered as a last resort. That mitigative defence is inapplicable if

292

Legal Responses to Torture

the direct victim of the torture is instead somebody (such as the innocent child of a terrorist) who has not been involved in causing the predicament. Although an interrogatee can be answerable for a dangerous situation simply by dint of his deliberately concealing some vital calamity-averting information, he is scarcely answerable for that situation through the sheer fact that he is related to somebody else. Interrogational torture perpetrated against an innocent relative of a terrorist, or against any other innocent person, should be treated by the law as unmitigatedly wrongful. It will be thus treated insofar as the limits on the defence of protection-of-oneself-or-others are duly heeded.

5.2.1.4.  A residual role for the defence of duress In some imaginable though rather fanciful circumstances, the defence of duress can appropriately be invoked by public officials who have resorted to interrogational torture that is not aimed at averting a calamity. The defence of duress is normally classified as an excuse, but often as only a partial excuse; hence, my understanding of it here as a mitigative defence is largely in keeping with the traditional understanding. Though it is never a fully exculpatory defence when invoked by a public official who has had recourse to interrogational torture, it can be suitable as a mitigative defence in some imaginable cases. Suppose that Orville is a policeman who is ordered by his superiors to wield interrogational torture against a suspect for the purpose of ascertaining the facts of a past crime (or for the purpose of disrupting some network of criminals or terrorists). Orville has a young daughter, whom the superior officers credibly threaten to murder if he does not comply with their directives. To spare his daughter, Orville carries out the interrogational torture. Obviously, no mitigative defence is applicable to the actions of the superior officers. However, the mitigative defence of duress is applicable to the actions of Orville himself. Although he in his capacity as a public official has availed himself of interrogational torture for a purpose other than the avoidance of a calamity, he has done so in response to a dreadful threat. His recourse to interrogational torture does not fall within the scope of the mitigative defence of protection-of-oneself-or-others—because the direct victim of the torture has not posed the threat to which Orville desperately reacts—but it does fall within the scope of the mitigative defence of duress. Notwithstanding that Orville should incur legal sanctions for having committed a very serious wrong, the severity of the sanctions should be lightened to reflect the terribleness of the quandary in which he has acted. A similar conclusion would follow in any number of other possible examples along the same lines. Though it is very unlikely that an official in a liberal- democratic regime would ever be coerced as domineeringly as Orville has been, an official in a despotic regime might sometimes face such coercion. If an official is indeed comparably bullied into performing some act(s) of interrogational torture, she can properly invoke the mitigative defence of duress in any legal proceedings that subsequently hold her accountable for her role as a torturer. Even if that defence is successfully invoked, it will of course not exculpate her entirely; nonetheless, it will significantly alleviate the penalties to which she is subject.

Legal Accountability Ex Post

293

5.2.2.  Which sanctions? As has already been stated, the multitudinous factors that bear on the fittingness of various legal sanctions in the aftermath of acts of torture by public officials are largely beyond the ambit of this book. Even if the remarks here are confined to sanctions for acts of interrogational torture, the aforementioned factors are too intricately detailed and diverse to be charted fruitfully at a philosophical level of abstraction. Some of the factors are practicalities like those which we have pondered in this chapter’s earlier discussions of jury nullification and prosecutorial inaction and kindred problems, whereas other considerations that bear on the suitability of various sanctions are more profound matters of moral principle. For example, among those deeper matters are the underlying justifications for the imposition of any legal sanctions. The elaboration of such justifications is an endeavour that far transcends the confines of this book. (One can get a sense of the magnitude of that endeavour from the fact that I have previously published a long volume on the justification for a single type of criminal sanction: the death penalty. An exploration of the justifications for other types of criminal sanctions and for non-criminal sanctions would extend across several further volumes.)

5.2.2.1.  Criminal sanctions imposed on individual officials Thus, instead of futilely trying to engage in the unmanageable task of prescribing specific types of sanctions for the sundry types of torture, I will here simply broach some of the chief possibilities that are available to the institutions of justice in any society. Most obvious among those possibilities is the imposition of criminal sanctions. Whenever any non-optimal instance of torture has been carried out by an official (or with the cooperation of an official) in his or her public capacity, the legal response should include the levying of criminal sanctions on the official. Though other sanctions are requisite in addition, the imposition of some criminal punishment on the official is essential as a means of holding him properly accountable for the serious and violent wrong that he has perpetrated or directed. To be sure, this opening contention is inapplicable to some of the very cases in which contemporary philosophers and jurists are especially interested: namely, some cases that arise from the use of calamity-averting interrogational torture. One reason why criminal proceedings against individual officials would not be germane in some such cases is that the practical problems of jury nullification and prosecutorial inaction are likely to undermine the effectiveness of the proceedings. Another reason why criminal charges against individual officials would not be apposite in certain cases that arise from the use of calamity-averting interrogational torture—a reason of moral principle—is that, in some such cases, the imposition of criminal sanctions on the officials directly involved in the torture would unduly individualize the responsibility for their actions. We are concentrating at present on cases in which the officials have acted in their public capacities and have thus resorted to torture on behalf of the societies in which they bear their public-safety responsibilities. In some such cases (for example, in my amplified version of Leon

294

Legal Responses to Torture

v. Wainwright), the recourse to torture is morally optimal even though it is morally wrong. In a case of that sort, the subjection of the direct perpetrators of the torture to criminal sanctions would not fairly reflect the extent to which their actions should be attributed to their whole community. In at least some cases where the wielding of calamity-averting interrogational torture by officials has been morally optimal, the role of the officials in fulfilling their public-safety responsibilities will be best recognized if the sanctions imposed for the torture are all borne collectively. (Collectivities can themselves be subject to criminal sanctions. Hence, even in the extremely rare cases of calamity-averting interrogational torture where the resultant penalties should be solely collective, the penalties can still be criminal.) What has just been said about some cases of calamity-averting interrogational torture will apply as well to some possible cases of placatory or act-impelling torture. In some possible circumstances—though perhaps no actual circumstances—the perpetration of placatory or act-impelling torture by public officials is morally optimal as a means of fending off the occurrence of calamities. Although the officials are breaching moral duties by engaging in such torture, they can thereby be fulfilling their even more stringent public-safety duties. In at least some of the possible cases where their recourse to such torture has been morally optimal, that aspect of their conduct will best be recognized if all the sanctions levied for the perpetration of the torture are borne collectively. In any situation where someone carries out a morally legitimate act of ephemerally incapacitative torture or edifying torture, no sanctions should ensue. A fortiori, no criminal sanctions should ensue in response to any such situation. However, as is evident from the relevant discussions in Chapters 2 and 3, not every act of ephemerally incapacitative torture or edifying torture is morally legitimate. Quite the contrary. Whenever an official in his public capacity has acted illegitimately by performing an act of such torture, criminal sanctions—adjusted in severity to correspond to the seriousness of the misdeed—should be imposed on him. More generally, whenever an official in his public capacity has performed or directed an act of torture that is not morally optimal, criminal sanctions should be imposed on him. Mitigative defences might be pertinent in some of the cases where such sanctions are to be levied on officials; as has been remarked, those defences can be applicable in any given case even if the torture perpetrated was not morally optimal. However, if such defences are applicable, they should indeed be mitigative rather than exonerative.

5.2.2.2.  Civil sanctions applied to individual officials As has been recounted in § 5.1.1.3.10, civil proceedings can in some cases help to overcome the problems of prosecutorial inaction and jury nullification that might stymie the effectiveness of criminal prosecutions as means of holding officials and organizations accountable for acts of calamity-averting interrogational torture. Civil proceedings skirt the problem of prosecutorial inaction outright. They can likewise go quite a long way toward surmounting the problem of jury nullification, if nominal awards of damages to unsavoury plaintiffs are allowed

Legal Accountability Ex Post

295

while the incentives for pursuing civil litigation are sustained through the awardability of plaintiffs’ legal costs against defendants. In cases that pertain to the use of calamity-averting interrogational torture, awards of damages against individual officials are very unlikely. Far more likely is that litigation will be pursued against the employers of the officials. In the very unlikely event that civil proceedings are pursued successfully against individual officials in cases that pertain to morally optimal instances of calamity-averting interrogational torture, the officials should be indemnified by their employers. Much the same can be said about cases that pertain to morally optimal instances of calamity-averting placatory torture or act-impelling torture. In other cases, where the administering of torture has not been morally optimal and especially where it has not been aimed at averting calamities, civil proceedings against individual officials can salutarily supplement criminal proceedings as means of holding the officials responsible for their actions. To be sure, the plaintiffs in some of those civil lawsuits might be decidedly unappealing in the eyes of jurors; for example, such lawsuits might be pursued by dangerous criminals who have been tortured into providing information about some recent crimes in which they have been involved. However, as has just been indicated, suitable approaches to the awarding of damages and attorneys’ fees can enable litigators to overcome the problem of jury nullification even in many cases brought by rebarbative plaintiffs. In litigation undertaken by plaintiffs whom jurors are apt to view sympathetically, awards of damages against individual officials can be highly effective in helping to remedy the wrongness of any torture that has been perpetrated. If any litigation arises from a morally legitimate instance of ephemerally incapacitative torture or edifying torture, then the claim for compensation should fail—on the ground that no wrong has been committed. Compensatory sanctions are no more warranted than criminal sanctions in such a case. Of course, to say as much is hardly to say that every instance of ephemerally incapacitative torture or edifying torture is morally legitimate. If lawsuits arise from any morally illegitimate tokens of either of those types of torture, the claims for compensation should succeed if there is sufficient evidence in support of them. In such lawsuits, as in other lawsuits that ensue from the perpetration of serious torts, compensatory awards of damages are of key importance in the rectification of wrongs.

5.2.2.3.  Institutional sanctions against individual officials In addition to criminal punishments and compensatory awards of damages, institutional sanctions of sundry kinds can serve to register the seriousness of the wrongdoing that has been perpetrated when officials have availed themselves of torture. Among the institutional sanctions that can be germane are outright dismissal, suspension, demotion, fines, and frozen opportunities for promotion. Some such sanctions, at the harsher end of the spectrum, will be pertinent whenever criminal sanctions against individual officials are pertinent. Even when criminal sanctions would not be appropriate against individual officials who have perpetrated acts of torture, the subjection of those officials to some institutional sanctions at the

296

Legal Responses to Torture

milder end of the spectrum can aptly serve as a means of authoritatively recognizing the wrongness of the acts of torture which they have performed. Of course, if the acts under consideration are morally permissible instances of ephemerally incapacitative torture or edifying torture, institutional sanctions are no more warranted than are sanctions of any other kind. Indeed, some morally permissible acts of ephemerally incapacitative torture—which avert the occurrence of very serious crimes—can suitably be treated as bases for institutional rewards rather than institutional sanctions. Whereas the use of torture of any other kind for the purpose of averting a calamity is always wrongful even when it is morally optimal, the use of ephemerally incapacitative torture in a dire emergency can be strongly permissible as well as commendable. On an occasion of that sort, when the moral status of the wielding of such torture is unequivocally favourable, the institutional responses to the torture should likewise be unequivocally favourable (though of course a formal investigation must be conducted to determine whether the actions taken were indeed necessary for the prevention of a very serious crime).

5.2.2.4.  Collectively borne sanctions As has already been suggested, collectively borne sanctions can sometimes be a superior alternative to individually borne sanctions in criminal or civil proceedings that arise from emergencies in which officials have had recourse to calamity-averting interrogational torture (or to calamity-averting placatory torture or act-impelling torture). In some such emergencies, the administration of the torture is morally optimal. In resorting to the morally optimal torture, the officials are fulfilling public-safety obligations which they bear on behalf of the whole system of governance in which they occupy their positions. Thus, although interrogational torture or act-impelling torture or placatory torture is morally wrongful even in such an emergency, the responsibility for the wrongfulness of the morally optimal torture is best attributed to the system of governance as a whole. More specifically, the responsibility should usually be attributed to the governmental organization—a police force or an intelligence agency or a military division, for example—in which the officials involved are employees. That organization should incur criminal fines and/or civil liability, to register the wrongfulness of the torture that has been perpetrated in fulfilment of the organization’s public-safety mission. Were no such sanctions to be levied (in a situation where no criminal or civil sanctions have been imposed on the individual officials), the system of governance would in effect be countenancing the use of interrogational or act-impelling or placatory torture as a means to a calamity-averting end. Such an end-justifies-the-means stance would leave the moral integrity of the system besmirched. In a situation where ephemerally incapacitative torture or edifying torture has been wielded by officials in a morally legitimate fashion, no collectively borne sanctions should be imposed. Such sanctions would be just as clearly inappropriate as would individually borne sanctions. Nothing morally impermissible has been done in such a situation, and therefore no legal penalties are required in order to sustain the moral integrity of the system of governance in which the

Legal Accountability Ex Post

297

torture has occurred. By contrast, in a situation where torture of any other kind has occurred at the hands of officials in their public capacities, collectively borne penalties should be imposed—either in exclusion of individually borne sanctions (in a case arising from an instance of interrogational or act-impelling or placatory torture that is morally optimal) or as a supplement to such sanctions. The levying of the collectively borne penalties is crucial if the moral integrity of the system of governance is to be redeemed in the aftermath of a serious and violent wrong perpetrated by public officials. Even if the torture carried out by an official in his public capacity is not only morally wrongful but also squarely at odds with the express policy and prevailing ethos of the organization in which he is employed, the relationship of responsibility between an employer and an employee—with the appurtenant fact that the official in his public capacity acts on behalf of his whole society—is such that the official’s transgression sullies not only his own moral integrity but also the moral integrity of his organization. Having stained the moral integrity of that organization, the official has pro tanto tarnished the overall system of governance in which he holds his position. Just as the employer/employee relationship of responsibility is to the benefit of an organization when the employees behave permissibly and commendably, it is to the detriment of an organization when the employees behave wrongfully. Anglo-American tort law routinely recognizes as much through the imposition of vicarious liability on employers for torts committed by any members of staff within their roles. Collective responsibility for acts of torture perpetrated by public officials is very much in keeping with moral principles that have long been established as legal principles in a number of jurisdictions. Of course, there are also some largely practical considerations that militate in favour of collective responsibility for the wrongful acts of torture perpetrated by officials in their public capacities. In the context of tort-law claims for compensation, the organizations that employ the officials will typically have much ‘deeper pockets’ than the officials themselves and will thus be more attractive as defendants in the eyes of plaintiffs (or their lawyers). Claims for compensation—and for lawyers’ fees—would often be partly unsatisfiable if those deeper pockets were not available as sources of funding to meet the claims. Moreover, both in the context of tort-law claims and in the context of criminal prosecutions, the positioning of organizations as defendants can help to overcome the reluctance of jurors to arrive at judgments of liability or guilt. Especially when torture has been perpetrated against thuggish criminals or terrorists (for calamity-averting purposes or otherwise), jurors might be quite strongly disinclined to hold that the individual officials directly involved are guilty of crimes or liable to pay compensation. In some credibly possible cases, they will feel less strongly disinclined to return verdicts of guilt or liability in trials where the chief defendants are organizations; the attenuation of their reluctance to return such verdicts will help to ensure that justice is done in such cases through the imposition of sanctions which serve to acknowledge the wrongness of the torture that has been inflicted. Hence, these practical considerations reinforce the principled considerations that tell in favour of collective responsibility for the wrongful use of torture by officials.

298

Legal Responses to Torture

5.2.3.  Private individuals as defendants Heretofore, this chapter has focused on the appropriate legal responses to torture that has been perpetrated by officials in their public roles. We should now consider how the law ought to deal with acts of torture that have been perpetrated by private individuals (including officials who are not acting in their public roles at all).

5.2.3.1.  Collective responsibility Unlike officials in their public capacities, private individuals in any society are not generally acting on behalf of the society as a whole or in the service of the system of governance which presides thereover.15 On the one hand, some possible incidents of torture perpetrated by private individuals will encumber their societies with moral obligations to make amends even though the individuals have not acted on behalf of their societies at all. For example, suppose that a private individual Thomas from Country X captures some tourists or diplomats from Country Y and subjects them to sadistic torture during a prolonged ordeal. When Thomas is eventually apprehended, the institutions of justice in Country X will of course be morally obligated to ensure that (after a fair trial) he is punished with suitable harshness. They will also be morally obligated to provide opportunities for civil proceedings against Thomas insofar as such proceedings will be fruitful. In addition, however, the institutions of governance in Country X will be morally obligated to make amends to Country Y in other ways—especially if the visitors from Country Y are diplomats, but even if they are only private tourists. The institutions of governance in Country X will at least be morally obligated to extend formal apologies to the victims and to representatives of Country Y, and they may well be further morally obligated to provide some funding for compensation if Thomas himself does not have ‘deep pockets’ from which such funding can come. Although Thomas has not been acting on behalf of Country X in any way, his status as a citizen of that country is sufficient to trigger some degree of collective responsibility for his terrible crimes of torture against the citizens of a foreign country. On the other hand, even in the scenario just sketched, the level of collective responsibility for Thomas’s torturous misdeeds is significantly lower than the level of collective responsibility that would have been triggered if the misdeeds had been carried out by someone working as a public official for Country X. When the victims of torturous wrongdoing are from the same country as the private citizens who perpetrate the wrongdoing, there is no collective responsibility whatsoever for the wrongdoing itself. Of course, the institutions of governance in the country will be morally obligated to take all reasonable measures to bring the perpetrators of torture to justice. However, the obligations incumbent on those institutions are not remedial in character, in that they are not reflective and derivative of wrongdoing that can properly be attributed to the institutions themselves (or to the officials 15   In any situation where a private individual does indeed act on behalf of his society, his conduct should be construed as tantamount to that of a public official.

Legal Accountability Ex Post

299

who run the institutions). Instead, they are obligations to respond—through appropriate legal proceedings and sanctions—to wrongdoing that has occurred at the hands of some private individuals in the jurisdiction over which the institutions of governance preside. Since those individuals have decidedly not been acting on behalf of their society as a whole when they have engaged in their crimes of torture, their society has not become implicated in those crimes. Because its moral integrity has not been sullied by the crimes, the society has not incurred any moral obligations to pay fines or compensatory awards for the purpose of redeeming its integrity. To be sure, the officials who run the system of governance in a jurisdiction might well be morally duty-bound to operate a criminal-injuries support scheme that can provide financial assistance to victims of torturous violence and other crimes (especially when the victims are unable to recover adequate compensation from the perpetrators of the crimes). However, although the relevant organization in the United Kingdom is designated as the ‘Criminal Injuries Compensation Authority’, its role is not really compensatory. Its financial support is provided in fulfilment of a government’s general moral responsibility to sustain the well-being of people who have suffered major adversities, rather than in fulfilment of any moral obligations to rectify wrongs for which the government has been responsible. No obligations of the latter sort exist, since—in the absence of malfeasance or negligence on the part of public officials—the perpetration of torture by private citizens against other people from the same country is not correctly attributable to the system of governance which presides in that country. In regard to the underlying moral principles, then, the payments made through a criminal-injuries support scheme are very different from the genuinely remedial payments that would be due from a government if its own officials were involved in the infliction of torturous mistreatment.

5.2.3.2. Some public/private similarities relating to individually borne sanctions Collectively borne moral obligations of a remedial character (as opposed to collectively borne moral obligations of other kinds) very rarely ensue from wrongful acts of torture that have been perpetrated by private individuals. In that respect, the perpetration of such acts by private individuals is significantly different from the perpetration of such acts by officials in their public capacities. Nonetheless, there are other important respects in which the former and the latter are similar. Let us for the moment leave aside calamity-averting instances of torture, to which I will return shortly. Accordingly, any people directly involved in performing or prescribing wrongful acts of torture will thereby have incurred some individually borne moral obligations of a remedial character, whether those people are private individuals or public officials. Because of the seriousness and violence of the wrongful acts of torture, the individually borne remedial obligations generated by those acts will include obligations to undergo legal punishments—regardless of whether the perpetrators of the torture are private individuals or public officials. Correspondingly,

300

Legal Responses to Torture

of course, the prevailing institutions of justice will be morally obligated to levy such punishments on the perpetrators. All of these moral obligations should be paralleled by legal obligations, the enforcement of which can help to ensure the fulfilment of the moral obligations. Conversely, if some action is a morally legitimate instance of ephemerally incapacitative torture or edifying torture, no moral obligations of a remedial character are engendered by it—irrespective of whether the person who has performed the action is a private individual or a public official. Because no wrong has been committed through a morally legitimate instance of such torture, there is no moral requirement for the rectification of any wrong in the aftermath of that instance. Consequently, no legal penalties should be imposed on anyone who has engaged in such conduct (though of course some legal proceedings for the purpose of ascertaining the nature and circumstances of the conduct could be entirely proper, especially in connection with ephemerally incapacitative torture). Again, this point about the unwarrantedness of any legal penalties is applicable regardless of whether the person who has wielded the torture is a private citizen or a governmental official.

5.2.3.3. Some public/private differences relating to individually borne sanctions Notwithstanding that the appropriateness or inappropriateness of individually borne legal sanctions for acts of torture by private individuals does correspond in many respects to the appropriateness or inappropriateness of individually borne legal sanctions for acts of torture by public officials, the correspondence does not always obtain fully (though, even when it does not obtain fully, there are salient similarities as well as differences). In particular, the correspondence becomes somewhat tenuous in certain cases of calamity-averting torture. Although interrogational or act-impelling or placatory torture is always morally wrong even when it is wielded in an extreme emergency for the purpose of averting a calamity, the wielding of it can be morally optimal in such a situation. If an instance of torture of any of those kinds is indeed morally optimal in such a situation, the people involved in perpetrating or directing the torture should not be subjected to individually borne legal sanctions. They should not be imprisoned or fined, and, if they are sued successfully by the victim(s) of the torture, they should be indemnified. That conclusion about the inappositeness of individually borne legal sanctions is sound irrespective of whether the people involved are private citizens or public officials. However, the considerations which support that conclusion are different across the public/private divide. As has been argued in § 5.2.2.1, the reason for the inappositeness of the levying of individually borne legal sanctions on public officials for morally optimal instances of calamity-averting interrogational torture (or act-impelling or placatory torture) is that such sanctions would unduly concentrate the responsibility for the use of the torture. Because the torture in such a case has been wielded as a last resort in an extreme emergency for the purpose of preventing the occurrence of a catastrophe,

Legal Accountability Ex Post

301

and because the officials who carry out the torture have done so on behalf of their whole society and in fulfilment of public-safety obligations which they owe to their society, the sanctions that acknowledge and rectify the wrongness of their actions are best borne by institutions of governance that represent everyone. Though the officials directly involved should perhaps incur some institutional sanctions, the burden of any criminal and civil sanctions should not be concentrated on them. Instead, that burden should fall on the entire community that has been tainted by the torturous measures which have been undertaken on its behalf. Quite different is the reason for the inappositeness of the levying of individually borne legal sanctions on private individuals who have engaged in morally optimal instances of calamity-averting interrogational torture (or act-impelling or placatory torture). Such individuals in any jurisdiction are not typically acting on behalf of the general public when they resort to torturous measures in circumstances of grave desperation. Nor do they owe special public-safety obligations to their community, which those measures serve to fulfil. Hence, the rationale for shielding them from any individually borne criminal sanctions—and for indemnifying them in the event of any awards of damages against them—has to be quite different from the rationale that is operative when the perpetrators of torture are public officials. Indeed, the rationale applicable to private individuals should highlight the differences between them and public officials. Unlike such officials, private individuals are not acting within positions of special responsibilities and powers when they resort to torture in grim emergencies. Their recourse to interrogational or act-impelling or placatory torture is morally wrong, but it does not implicate their society as a whole and therefore does not besmirch the moral integrity of their society. Of course, any private individual who avails himself of such torture will have besmirched his own moral integrity and will thus have incurred moral obligations to remedy his wrongdoing (at least by feeling regret over the fact that he has stooped to such behaviour). Still, not all moral wrongs committed by private individuals—not even all serious moral wrongs committed by private individuals—are properly subject to legal sanctions. In some cases where private individuals have had to use interrogational or act-impelling or placatory torture out of desperation in extreme emergencies, they should be able to exempt themselves from legal sanctions through the invocation of full excuses. Let us briefly ponder two examples. 5.2.3.3.1.  An example already encountered As I  have indicated already in Chapter  4’s discussion of the placatory torture administered to Miranda by Melvin, his desperate actions to spare her from the ravages of the thugs should not result in his being subjected to any legal sanctions. Were he to be prosecuted for the placatory torture which he has undertaken, he should be able to exonerate himself with a full excuse. In his situation, of course, the defence of protection-of-oneself-or-others would not be pertinent. After all, he has performed his acts of torture against his wife rather than against the thugs who are responsible for the extremely dangerous predicament with which he has been confronted. However, although the defence of protection-of-oneself-or-others would not be germane, the excuse of duress is fully applicable. Melvin plies his

302

Legal Responses to Torture

torturous measures under horrific circumstances in which those measures are the only means by which he can save Miranda from the thugs while complying with her own urgent entreaties. Although his torturous measures are morally wrong for the reasons indicated in Chapter 4, the emergency in which he has resorted to them is so dire that he should not be subjected to any legal sanctions for his wrongdoing. He should be allowed to invoke the full excuse of duress as a shield against any such sanctions (in the very unlikely event of a prosecution pursued against him). 5.2.3.3.2.  A second example In the scenario of Melvin and Miranda, the calamity-averting torture is placatory and the applicable excuse is that of duress. Let us now consider a scenario in which the calamity-averting torture is interrogational and the applicable defence is that of protection-of-oneself-or-others. Suppose that several vicious criminals abduct three members of a family—a husband, a wife, and a son—and take them to an apartment from which the only means of exit is a door sealed with an electronic combination lock. The lock can be opened only through the typing of nine digits in a correct sequence. There are no telephones in the apartment. Having raped the wife and beaten the husband and son, the criminals proceed to tie them up tightly. Intending first to demand a ransom and then to murder the family members after the ransom has been collected, the criminals undertake some preparations for the completion of their terrible crime. All but one of them leave for approximately 90 minutes in order to eat dinner at a nearby restaurant, while the one remaining kidnapper stays behind to ensure that the hostages do not escape. Thinking that the family members are securely bound and dazed, the thug who is standing guard over them relaxes and begins to watch television. Meanwhile, the father and son furtively endeavour to undo the knots in the cords with which they have been tied up. Before much time has elapsed, each of them succeeds in freeing himself from the cords. Creeping up silently behind the seated kidnapper, the father and son manage to grab him and bind him with the cords. After the son unties the mother, they and the father desperately seek to escape from the apartment. They know that they have to get away before the return of the other thugs, whose onslaughts they would not be able to withstand. Because there are no telephones in the apartment, they cannot summon the police. In any event, they do not know the location of the apartment building and would thus not be able to specify the address even if they could communicate with the police. Hence, their only hope lies in exiting through the door that is sealed with the combination lock. None of the members of the family is familiar with the relevant nine-digit sequence, of course, whereas they know that the thug whom they have tied up is familiar with that sequence. They saw him type in the nine digits by memory when the other kidnappers left for dinner. Proceeding under severe constraints of time, the members of the family adjure their captor to apprise them of the sequence. They offer to provide a subterfuge for him in order to spare him from the wrath of his fellow criminals; to obscure the fact of his supplying them with the crucial

Legal Accountability Ex Post

303

information, the members of the family will testify that they overpowered him after he had opened the door for some other purpose. However, he sneeringly refuses to cooperate with them in any way. After remonstrating with him further to no avail, the father and the son grab him. The father pulls one of the thug’s fingers back until it breaks, and he indicates fiercely that he will continue with each of the other fingers unless the man divulges the vital information. After the father has broken two of the fingers and has started to pull on a third, the kidnapper is no longer able to endure the exquisite pain. He blurts out the sequence of digits, which the wife types into the combination lock. With the door opened, the members of the family scramble out of the apartment and escape from the site of their captivity. Now, when the facts of the case become known, the police will almost certainly exercise their discretion to refrain from arresting the father and son for participating in interrogational torture. If the police do arrest the pair, the prosecutor in the jurisdiction will almost certainly decline to proceed any further with the case. Such an exercise of prosecutorial discretion, like the exercise of discretion by the police in forbearing from even arresting the pair, would be entirely proper. If the prosecutor instead very surprisingly chooses to place the two men on trial, each of them should be able to invoke a full excuse of protection-of-oneself-or-others. That legal excuse is straightforwardly applicable to their situation; they have wielded violence only as a last resort to save themselves and the mother from considerably worse violence, and the kidnapper against whom they have used force is one of the people directly responsible for the extreme peril which the father and son have endeavoured to fend off. In any prosecution that might be brought against them, the conditions for a fully exonerative excuse of protection-of-oneself-or-others are satisfied. As is indicated by the very need for the invocation of an excuse (in the highly unlikely event of a prosecution), the propriety of the legal exoneration of the father and son is not due to the moral permissibility of their torturous techniques of interrogation. On the contrary, as Chapter 3 has argued at some length, any recourse to interrogational torture is morally wrong even in a grim emergency where the use of such torture is morally optimal. Even in such an emergency, a commendable end does not strongly justify a wrongful means. Hence, the inappropriateness of the levying of any legal sanctions on the father and son is due not to the absence of any wrongdoing on their part but instead to the nature of their wrongdoing. The father and son have not been acting as public officials on behalf of their society, and they have therefore not begrimed their society with the taint of the torture to which they have resorted. Unlike constables and other law-enforcement officials, they have not been acting in positions of special powers which they have accepted in fulfilment of special responsibilities which they have likewise accepted. Rather, they have been acting as private individuals who are desperate to save the lives and elementary well-being of themselves and their family members. As a consequence, their actions are fully understandable—and morally optimal— even though those actions are morally wrong. As Andrew Simester rightly suggests (2012, 100, 105), an exonerative legal excuse is available when someone has

304

Legal Responses to Torture

behaved wrongfully but in a thoroughly understandable fashion toward which any reasonable observer should be sympathetic. Given the direness of the plight in which the father and son have acted, that condition for the applicability of a full excuse is satisfied in relation to their torturous measures. Of course, as has already been noted, the legal defence of protectionof-oneself-or-others is normally treated as a justification. However, when it is invoked as a fully exonerative defence by somebody who has engaged in interrogational torture, it should be construed instead as an excuse. It should be available as an excuse only when the recourse to calamity-averting interrogational torture has been morally optimal—that is, only when the torture has been undertaken as a last resort against someone directly responsible for the peril which the torturous measures are aimed at heading off. As an excuse, the defence does not indicate that the excused conduct was morally permissible; such a defence instead indicates that, although the conduct was morally impermissible, the circumstances of its occurrence were such that no legal sanctions are appropriate in response to it. As has already been stated, not all moral wrongs—not even all serious moral wrongs that involve violence—are properly subject to legal sanctions. This point is important because there is a significant difference between the invocation of an excuse by a public official and the invocation of an excuse by a private individual. If a public official successfully invokes an excuse to shield himself from any individually borne sanctions for an instance of calamity-averting interrogational torture which he has carried out, the wrongness of his torturous conduct can and should still be legally rectified through the imposition of criminal or civil sanctions on the collectivity that has been sullied by the torture (given that his conduct has been undertaken on behalf of the collectivity). By contrast, when a private individual successfully invokes an excuse to shield himself from any individually borne sanctions for an instance of calamity-averting interrogational torture which he has carried out, the wrongness of the torture will not be legally rectified. No collectively borne legal sanctions would be apposite in such a case, since the private individual has not been acting on behalf of his community and has thus not besmirched its moral integrity through his behaviour. Hence, in the absence of any individually borne legal sanctions, no legal sanctions at all will be imposed in response to his morally wrongful act of torture. However, as has been stated, not every moral wrong can suitably be rectified through the punitive or compensatory workings of a legal system. Although the father and son in my scenario have behaved impermissibly by resorting to interrogational torture, the dreadfulness of their predicament—along with the relative mildness of the technique of torture which they have employed—is such as to render unreasonable any subjection of them to legal sanctions for their actions. Like certain other wrongs committed by private individuals, their torturous measures are properly rectifiable only through pangs of conscience. Notwithstanding the exigencies of the situation in which they have adopted those measures, they should feel consternation over the degradingness of such a tack; in the circumstances, their feelings of consternation are sufficient to remedy the wrongness of what they have done.

Legal Accountability Ex Post

305

5.2.3.3.3.  Sado-masochistic torture once more As has been suggested in my discussions of sado-masochistic torture in Chapters 2 and 3, the perpetration of such torture should not trigger any legal sanctions unless it is non-consensual or unless it foreseeably results in death or permanent and very serious injuries. In other words, in the absence of very serious and permanent harm, the exonerative defence of consent should be available in any criminal proceedings that might be pursued against a perpetrator of such torture. Now, although acquittals are appropriate in some cases of sado-masochistic torture just as they are in some cases of calamity-averting interrogational torture, the rationale for the acquittals in the former cases is quite different from the rationale in the latter cases. Likewise different, of course, is the defence through which the acquittals are effected. Sado-masochistic torture is not undertaken as a last resort in a dire emergency for the purpose of averting a calamity, nor is it ever realistically optimal. That is, the chief considerations that militate against the imposition of legal sanctions on the father and son in the preceding subsection’s scenario are not operative in cases of sado-masochistic torture. Determinative instead are some considerations pertaining to the intimacy of sexual activity that is conducted in private between consenting adults. If a sado-masochist inflicts severe pain on another sado-masochist in interaction that is fully consensual and private and aimed at achieving intense sexual gratification for both parties, then—unless the infliction of pain has predictably caused the victim to suffer death or permanent and very serious injuries—the sole factor that makes the conduct of the perpetrator morally wrongful is the attitude of cruelty which he harbours. His sadistic glee arising from his affliction of someone else with excruciating pain is antithetical to the respect and concern which he should always show toward himself and toward anyone else. That attitude of sadistic relish in response to the misery of others, an attitude paradigmatically connected to evil, is what morally vitiates the conduct of a perpetrator of sado-masochistic torture even in the absence of any permanent and very serious harm to his victim. Hence, if sado-masochistic torture does trigger criminal sanctions even when no such harm has occurred or been likely to occur (and even when the torture is straightforwardly consensual), the sanctions are morally illegitimate unless a system of governance can properly criminalize the attitude that underlies the perpetrator’s conduct. Yet, although that attitude is morally wrongful, the imposition of legal sanctions on someone for harbouring an unsavoury attitude—in the absence of other aspects of his conduct that would be properly criminalizable—is illiberally overweening. A criminal prohibition that provides for sanctions in the circumstances envisaged here is especially dubious because its purpose is so patently the discouragement of distasteful sexual deviance. Such a prohibition, when applied to fully consensual interaction in which no permanent and very serious harm has occurred or been likely to occur, would criminalize conduct not because of its harmfulness but because of its marked divergence from normality. To be sure, that divergence from normality is a moral failing insofar as it is impelled by an attitude (on the part of the perpetrator of the sado-masochistic torture) that falls short of the elementary respect and concern which morality requires. However, as has just been remarked,

306

Legal Responses to Torture

a wrongful attitude is not in itself a fit object of criminal sanctions. In any event, the criminal bans on sado-masochistic activities have been aimed less at dodgy attitudes than at conduct that is driven by those attitudes. What manifestly underlies those bans, insofar as they are applied to consensual conduct that does not result in death or permanent and very serious injuries, is concern over the extent to which sado-masochistic activity deviates distastefully from normal patterns of behaviour. When we recognize that hostility toward deviance typically underpins the criminalization of consensual sado-masochistic conduct, we can understand why such conduct has in fact frequently been criminalized even while dangerous sports such as boxing and American football have generally been permitted. Those sports involve deliberately rough contact that generates quite high risks of serious and permanent injuries suffered by the participants. Moreover, sadistic attitudes— whereby players and spectators derive gratification not only from winning but also from the despair of their opponents who have lost—are held by quite a few of the people who play or follow those sports. Nonetheless, such athletic activities are rarely banned altogether. What explains the rarity of laws which proscribe those sports and the frequency of laws which proscribe sado-masochistic torture is that participation in the sports is perceived as normal whereas participation in sado-masochistic torture is perceived as decidedly abnormal. Accordingly, if the criminalization of sado-masochistic torture in the circumstances envisaged here is to be morally tenable, the discouragement of sexual deviance must itself be an end that is legitimately pursuable through the punitive power of government. However, in itself, the discouragement of sexual deviance is not such an end. Disconnected from worries about serious injuries and non-consensuality, the objective of discouraging sexual deviance through the levying of criminal sanctions is at odds with the tenets of liberal democracy. In situations where permanent and very serious injuries are neither occurrent nor likely to occur, the levying of punishments to coerce adults into abstaining from the consensual pursuit of sexual pleasure is objectionably officious and intrusive as a form of nannying. It is also discriminatory, in that other relevantly similar pastimes are treated very differently. The prospect of criminal sanctions might steer people away from engaging in the wrongful activity of sado-masochistic torture, but not every wrong committed by private individuals is suitably avoided or discouraged through the imposition of such sanctions. Sane adults engaging consensually in intimate relations should not be hectored by the punitive mechanisms of a system of criminal justice, unless the intervention of the system is necessary to deal with the infliction of very serious and permanent harm. 5.2.3.3.4.  A rejoinder? Someone who advocates the imposition of punishments on sado-masochistic torturers might resist my analogy between sado-masochistic torture and certain sports. Let us recall here Chapter 2’s distinction between the wielding of actual torture in sado-masochistic contexts and the simulation of torture in those contexts. If sado-masochists have agreed upon a signal whereby the victim of some extremely painful techniques can halt the administration of them at a time of his

Legal Accountability Ex Post

307

choosing, the victim’s control over the process means that the application of those techniques is a simulation of torture rather than an instance of veritable torture. Hence, whenever sado-masochistic torture occurs, the victim has no control over the process. Given as much, there is a morally pregnant difference between such torture and any typical match of boxing or game of American football. A participant in such a match or game can withdraw from it at any point if he has injured himself or is otherwise experiencing significant pain. Each participant enjoys control over his involvement in the match or game, precisely because he can terminate that involvement at any time of his choosing; thus, his role in the match or game differs in a morally significant fashion from the role of a victim in sado-masochistic torture. On the one hand, the difference just recounted is genuine and is morally pregnant. Indeed, the fact that each participant in a match of boxing or a game of American football can control his involvement in the way just specified is one main reason for the moral legitimacy of such a match or game in any ordinary context. That feature of boxing matches and American football games does contribute to distinguishing them favourably from sado-masochistic torture, which is always morally illegitimate. On the other hand, the question currently under consideration is not whether sado-masochistic torture always differs from morally permissible activities such as boxing or American football in ways that render such torture morally wrong. That question has already been answered in the affirmative, but it is not the point at issue here. Rather, we are at present seeking to ascertain whether the features of sado-masochistic torture are such that the criminalization of it is appropriate even when the torture is fully consensual and is not productive of death or serious and permanent injuries. In the preceding paragraph, the suggestion from an imagined opponent of my position is that a victim’s lack of control over a process of sado-masochistic torture is sufficient to warrant the levying of criminal sanctions on the perpetrator of such torture. Is that suggestion correct? Let us suppose that, in a prosecution for aggravated assault, the victim’s lack of control over a process of sado-masochistic torture is specifically perceived by the deciding judge or jury as a key factor that militates in favour of a guilty verdict against the perpetrator of the torture. Would the judge or jury be correct in regarding the lack of control as a sufficient ground for a conviction? What should be kept in mind here is that, in the circumstances on which I have concentrated in § 5.2.3.3.3, the victim’s lack of control over the process of torture is itself consensual. That is, the victim has readily consented to the absence of control in order to intensify his own sexual elation. Thus, his lack of control does not detract from the consensuality of the process. Accordingly, if the lack of control is to render the perpetration of sado-masochistic torture properly susceptible to criminal sanctions, that upshot must be due to a high likelihood of death or permanent and very serious injuries. Now, the likelihood of death or major injuries is of course increased by the absence of control on the part of the victim. If that likelihood increases to a high level, and if very serious and permanent harm actually occurs, then the conduct of the perpetrator of the torture can appropriately be classified and punished as aggravated assault. However, in the circumstances on which § 5.2.3.3.3 has

308

Legal Responses to Torture

concentrated, the probability of death or permanent and very serious injuries is not high, and no consequence of that magnitude occurs. In those circumstances, then, the victim’s lack of control over the process of torture does not affect that process in any way that renders the perpetrator of the torture legitimately subject to criminal sanctions. Hence, the rejoinder broached in the present subsection does not succeed as an objection to my stance on the criminalization of sado-masochistic torture. In short, the position espoused here (and in Chapter  3) in relation to sado-masochistic torture is hardly an ‘anything goes’ doctrine. Regardless of how eagerly a victim consents to being tortured, the perpetrator of the torture can rightly be subjected to criminal sanctions if his actions foreseeably result in death or permanent and very serious injuries. However, if the torture is fully consensual and if no grievous harm actually occurs, the imposition of criminal sanctions on the perpetrator is morally illegitimate. Morally wrongful though his conduct is, the criminalization of its attitudinal wrongfulness is irreconcilable with the principles of liberal democracy. Any such criminalization is irreconcilable with the proposition that the punitive mechanisms of the state should not be activated to coerce adults into refraining from consensual interaction that does not very seriously injure anyone. 5.2.3.3.5.  Sado-masochistic torture and the harm principle John Stuart Mill famously articulated a precept of political morality that has come to be known as the harm principle: ‘[T]‌he only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others’ (1956, 13). We should ask here whether the levying of criminal sanctions on a perpetrator of sado-masochistic torture is consistent with the harm principle, which lays down a necessary condition for the moral legitimacy of any punishment. One’s application of that principle has to draw on some understanding of what is to count as ‘harm’ within the formulation which Mill propounded. For example, as H. L. A. Hart contended (1963, 46–8), the harm principle would be hollow as a constraint on the scope of legitimate governmental regulation if the notion of harm within it encompassed the feelings of consternation or distress that are experienced by some people when they know that other people are engaging in certain sexual practices in private. Such feelings of dismay lie outside the referential reach of the term ‘harm’ in Mill’s formulation. Should we arrive at a similar conclusion about the effects of sado-masochistic torture undergone by a consenting adult who does not die or suffer permanent and very serious injuries from the torture? On the one hand, given that the harm principle sets forth a necessary condition rather than a sufficient condition for the moral legitimacy of any punishment, my general position on the criminalization of sado-masochistic torture does not depend on an affirmative answer to the question just posed. On the other hand, that question should indeed be answered affirmatively. Although the fact that someone has informedly consented to being treated in a certain manner does not always establish that the effects of his being so treated lie outside the extension of the word ‘harm’ in Mill’s formulation, it does establish as much when

The Complexities of Involvement

309

those effects do not include death or permanent and very serious injuries. Being afflicted with the excruciating pain of torture is harmful in an ordinary sense even when the undergoing of the pain is fully consensual, but the notion of harmfulness in the harm principle is more restrictive. We would be blinding ourselves to the strongly anti-paternalist tenor of the harm principle if we failed to recognize that the presence or absence of fully informed consent is frequently determinative of the character (the harmlessness or harmfulness) of various experiences which people undergo. More specifically, the provision of informed consent is determinative if an experience undergone by a consenting adult has not substantially and permanently impaired her functioning as a human being. When no such substantial and permanent impairment has occurred, the provision of informed consent renders harmless what would otherwise be aptly classifiable as harmful under the terms of the harm principle. Were a system of criminal justice not to treat consent as determinative in such circumstances, it would be showing disrespect for the victims of sado-masochistic torture in the course of punishing the perpetrators.

5.3.  The Complexities of Involvement Heretofore, this chapter has considered how a legal system in any given jurisdiction should respond to the perpetration of torture within that jurisdiction. In recent years, however, some especially knotty issues in the United States and the United Kingdom (and certain other Western countries) have pertained to a number of more oblique connections between the use of torture and the operations of systems of legal governance. This chapter, and this book, will close by looking very briefly at a handful of those issues. A lot of the gnarled problems just mentioned are contemplated in the UN Convention against Torture. For example, Article 3 of the Convention decrees that ‘[n]‌o State Party shall expel, return (“refouler”) or extradite a person to another State where there are substantial grounds for believing that he would be in danger of being subjected to torture.’ Similarly, Article 15 decrees that ‘[e]ach State Party shall ensure that any statement which is established to have been made as a result of torture shall not be invoked as evidence in any proceedings, except against a person accused of torture as evidence that the statement was made.’ These two provisions do not cover all the complexities of the ways in which systems of legal governance can become indirectly linked to the perpetration of torture, but they do cover many of those ways.

5.3.1.  Extradition, deportation, and extraordinary rendition Like Article 15 of the Convention, Article 3 is in accordance with the moral obligations that are incumbent on governments. Just as a state and its officials are under moral obligations never to administer interrogational torture or most other types of torture, they are likewise under moral obligations never to abet any use of such

310

Legal Responses to Torture

torture by foreign regimes. Complicity in torture practised by foreign regimes is especially stark when detainees are transferred to those regimes for the very purpose of having the detainees subjected to no-holds-barred interrogations. That is, any process of ‘extraordinary rendition’ constitutes a flagrant breach of a government’s moral duties, and it should trigger the imposition of sanctions under domestic law and international law (though of course the heaviest sanctions should be imposed on the nefarious regimes to which the detainees are transferred). However, as is correctly suggested by Article 3, complicity in torture practised by foreign regimes is also morally illegitimate when it is less cynical. Sometimes, detainees are transferred not for the purpose of having information extracted from them through torture, but instead for the purpose of getting rid of them. When a dangerous suspect has committed his crimes principally or exclusively in another jurisdiction, there will typically be strong pressures—both domestically and externally—for the extradition of him to that other jurisdiction. Nevertheless, a government will be breaching its moral obligations and its Convention-based legal obligations if it extradites someone to a country where he will probably be subjected to torture. Similarly, of course, a government will be breaching its moral obligations and its Convention-based legal obligations if it returns an asylum-seeker to a country where he will probably be subjected to torture. These moral duties are quite often grounded solely on the perpetrator-focused considerations that have been delineated in my third chapter. Although some of the people whom Article 3 of the Convention spares from extradition or deportation are innocents—or, at any rate, are not guilty of any serious crimes—a number of others are mass-murdering terrorists who are wanted abroad to stand trial for their roles in sundry atrocities. If we approach these matters from a purely victim-focused perspective, we will be unable to grasp why people of the latter sort are morally and legally shielded from the prospect of torture; their interests in being free from excruciating pain are of no positive ethical weight. However, when we approach these matters also from a perpetrator-focused perspective, we can grasp why governments are morally obligated to spare the mass-murdering terrorists as well as the innocents from the likelihood of being tortured. Albeit the extradition or deportation of someone to a land ruled by a torture-practising regime does not sully a government’s moral integrity as gravely as does any direct involvement in the perpetration of interrogational torture (or other absolutely impermissible types of torture), all actions that foreseeably and significantly facilitate the occurrence of such torture are morally wrong. Given that such torture is always and everywhere morally wrong—even when it is plied against mass-murdering terrorists—actions that foreseeably and significantly facilitate its occurrence are likewise always and everywhere morally wrong. Because the considerations that justify my claim about the absolute moral wrongness of interrogational torture (and most other types of torture) are perpetrator-focused as well as victim-focused, the considerations that justify my claim about the absolute moral wrongness of the abetment of such torture are similarly perpetrator-focused as well as victim-focused.

The Complexities of Involvement

311

5.3.2.  Evidence produced by interrogational torture If any of the officials in a system of governance have resorted to interrogational torture themselves and have thus acquired some information or other evidence, or if they have acquired some information or other evidence that has been extracted through interrogational torture practised in a foreign jurisdiction, the information or other evidence should be inadmissible in any legal proceedings except in proceedings against the torturer(s) for the purpose of confirming that the information or other evidence was gained through torturous methods. Article 15 of the Convention correctly declares as much. An exclusionary rule of this sort is justifiable partly on grounds of deterrence; by greatly reducing the law-enforcement value of statements that have been elicited through interrogational torture, the exclusionary rule discourages the use of such torture by law-enforcement officials. However, the chief grounds for such a rule are deontological. The moral integrity of a system of legal governance is tarnished not only if the officials of the system direct or perpetrate interrogational torture, but also if they knowingly or negligently bestow formal recognition on any evidence that emerges from such torture. Even if none of the officials within a system of governance has prescribed or undertaken any torturous techniques of interrogation, the system will be besmirched insofar as its officials in legal proceedings rely on evidence gathered through such techniques in some foreign jurisdiction(s).16 Unless the officials in those proceedings have no grounds for knowing that the evidence on which they rely is tainted by the use of interrogational torture, their admission of that evidence will be in breach of their moral obligations and their legal obligations under Article 15 of the Convention. Worth noting here is that the inadmissibility of evidence gathered through interrogational torture is acknowledged by most of the contemporary philosophers and jurists who contend that the use of calamity-averting interrogational torture can sometimes be morally permissible or that it should sometimes be legally permissible. When officials have recourse to torturous techniques of interrogation for the purpose of forestalling a calamity, they are seeking to extract crucial information that can help to realize their disaster-averting objective in circumstances of desperation. If they are carrying out their endeavours in a liberal democracy, they know that the statements extracted will not be admissible as evidence in any adjudicative proceedings. Given the exigencies of the plight with which the officials are confronted, however, the inadmissibility of those statements is greatly overshadowed by the possibility that the elicited information will enable the prevention of a catastrophe. Correspondingly, when various contemporary theorists write in support of the legal authorization of calamity-averting interrogational torture, most of them readily accept that any evidence gained through the wielding of such torture should not be admissible in legal proceedings. For example, as this chapter has already pointed out, Dershowitz firmly denies that any confessions or other 16   This assertion, along with any relevantly similar assertions in the present discussion, is of course subject to the one exception which I  have already mentioned and which is likewise mentioned in Article 15 of the Convention. To avoid stylistic clottedness, I will not make any further explicit references to that exception.

312

Legal Responses to Torture

evidence obtained through the use of interrogational torture can ever properly be admissible in such proceedings. Thus, in the present-day disputation over the morality and legality of calamityaverting interrogational torture, Article 15 of the Convention against Torture does not figure prominently. Its exclusionary implications are taken for granted by virtually all the participants in the disputation. Still, the exclusionary rule articulated in Article 15 does figure conspicuously in some other lively controversies pertaining to evidence that has been procured through interrogational torture.17 As has already been noted, Article 15 does not distinguish at all between interrogational torture employed domestically and interrogational torture employed in foreign jurisdictions, and it likewise does not distinguish between interrogational torture that has been abetted or encouraged by officials of any particular country and interrogational torture that has been conducted entirely independently of those officials. In sum, Article 15 (or any cognate law) straightforwardly forbids the admission of torture-produced evidence into judicial proceedings, regardless of whether the torture has occurred at home or abroad and regardless of whether a given nation’s officials have been involved in any way in the torture. That exclusionary mandate, which so clearly goes beyond any deterrence-oriented concerns, is rightly grounded on deontological qualms about the moral integrity of any system of legal governance. However, what is less straightforward morally and legally is whether torture-produced evidence—that is, evidence produced by torture in which the officials of one’s own nation have not been involved either directly or indirectly—can be relied upon in deliberations that are not adjudicative. Questions of this sort have arisen in the United Kingdom and elsewhere in recent years, as Islamist terrorists and other international terrorists have fled from their native lands and have arrived in Western nations. Frequently, the handing over of such people to the regimes in their native countries is forbidden under Article 3 of the Convention against Torture (and under cognate laws that emanate from other sources). Still, although the deportation of the terrorists to countries ruled by torture-wielding governments would be both illegal and immoral, these refugees are dangerous fanatics who will seek to perpetrate atrocities if they are allowed to roam freely. Hence, British officials and other Western officials have had to resort to various modes of detention and control in dealing with such people. While the initial decision to subject a refugee to detention or control is an executive determination in the United Kingdom, the detention or control is subject to review in adjudicative proceedings if it is appealed or challenged. Quite frequently, the basis for the decision of the executive in favour of detention or control includes evidence that has been obtained through interrogational torture practised abroad (without any encouragement or cooperation on the part of British officials). Now, such evidence should clearly not be admissible into any adjudicative proceedings through which the determinations of the executive are reviewed. Notwithstanding that British officials have been uninvolved in the use of the torture, the moral integrity of their system of legal governance will be besmirched   For a good exploration of some of these controversies, see Birkinshaw 2009.

17

The Complexities of Involvement

313

if any evidence produced by the torture is knowingly or negligently invoked by British judges as bases for their rulings. If the judges know or have good grounds for knowing that the evidence has been extracted through torturous techniques of interrogation, they will in effect be endorsing those techniques if they rely on the evidence in some adjudicative proceedings. Pro tanto, they will be saddling their system of governance with their approval of a practice that is always and everywhere morally impermissible. What is of principal concern is not the unreliability of the evidence, but the wrongness of knowingly or negligently endowing it with formal acceptance. Morally, the situation is akin to a situation in which some important evidence has been unearthed through depredations that have been undertaken by foreign officials for the purpose of persecuting individuals from certain religious or racial groups. If the judges in any legal system have grounds for knowing of the tainted means by which the evidence has been uncovered, they will sully their system if they nonetheless admit the evidence and thus treat it as an authoritatively invocable basis for a ruling. They will be committing their system to a morally benighted end-justifies-the-means doctrine. Such a doctrine is no more tenable in relation to evidence produced through interrogational torture than in relation to evidence produced through racist persecution. Less clear-cut, however, is whether the initial executive determination to impose detention or control or deportation can properly draw upon evidence from abroad that has been obtained through interrogational torture (conducted without any cooperation or encouragement on the part of British officials). Legally, the executive officials in the United Kingdom are indeed permitted to make use of such evidence when arriving at their initial determinations. They will of course not be able to adduce that evidence in judicial proceedings if their determinations are challenged, but their initial actions—which will be dispositive if they go unchallenged—are performed under less stringent evidentiary standards. Executive officials draw on many types of inadmissible evidence when reaching their decisions; evidence attributable to interrogational torture that has been administered abroad without any involvement of British officials is just one of those types. Provided that British officials have indeed been entirely uninvolved in the plying of the torture, and provided that there are some independent grounds for thinking that the evidence obtained through the plying of the torture is credible, the lawfulness of the use of such evidence by executive officials for their initial decisions is in accordance with the morality of the situation. Unlike adjudicative proceedings, the deliberations that lead to those initial decisions do not bestow formal recognition on any sources of evidence. When a British Home Secretary relies on some foreign intelligence (subject to the two qualifications just stated) without any expectation that the intelligence will ever be admissible in formal proceedings, she is not thereby enshrining an end-justifies-the-means doctrine within the British system of law and governance. A number of further complexities would have to be pondered in any thorough exploration of these matters (Birkinshaw 2009, 109–14). For example, given that the Home Secretary can legitimately draw informally on foreign intelligence that may have been gained through the use of interrogational torture, can she

314

Legal Responses to Torture

legitimately advert to that intelligence in defence of herself against a lawsuit for false imprisonment? In which direction does the burden of proof lie, concerning whether any particular elements of evidence from abroad are attributable to the use of interrogational torture? How stringent is the standard of proof? If some evidence E2 is uncovered partly because of the earlier receipt of some evidence E1 that was extracted by foreign torture, can E2 properly be admitted into legal proceedings if there is a substantial likelihood of its having come to light even without the presence of E1? These and other intricate questions are the minutiae of the ethical quandaries that confront judges and other officials in modern liberal democracies. Because the use of interrogational torture remains so extensive in many parts of the world, and because intelligence relating to international terrorism and criminality can be directly relevant to investigations or prosecutions undertaken in countries far beyond the borders of the land from which the intelligence has emanated, difficult questions pertaining to torture-tainted evidence must be addressed even by officials in regimes that scrupulously refrain from torturous methods of interrogation.

5.3.3.  Article 3 and Article 15 conjoined Some of the most vexing predicaments that have emerged during the past few decades, exemplified by the efforts of successive British governments to deport the Islamist terrorist Abu Qatada to Jordan, are traceable to the combined implications of Article 3 and Article 15 of the Convention against Torture (or to similar implications of other sources of law such as the European Convention on Human Rights). Having fled from Jordan in the early 1990s and been granted asylum in the United Kingdom in 1994, Qatada was convicted in absentia in Jordan in 1999 on terrorism-related charges. After his arrest in England in the early 2000s for his roles in various extremist activities, he was repeatedly imprisoned and released and re-imprisoned as British Home Secretaries sought from 2005 onward to deport him to Jordan. At several junctures, the endeavours to deport him were blocked by British courts or the European Court of Human Rights. The principal concern of the courts was not that Qatada would be likely to undergo torture in Jordan, but instead that he would be placed on trial there in proceedings where some of the key evidence against him had been obtained in the past through the use of interrogational torture. In response to the sundry judicial rulings, British Home Secretaries elicited stronger and stronger assurances from the kingdom of Jordan—culminating in a treaty adopted in June 2013 that finally cleared the way for the return of Qatada to that kingdom in July 2013. Qatada is a menacingly unscrupulous fanatic, but the British and European courts acted rightly by insisting on safeguards against the introduction of torture-tainted evidence into any legal proceedings that might be undertaken by the Jordanians to deal with his crimes. On the one hand, the wrong which the courts sought to avert is less serious than the wrongs that are principally addressed by Articles 3 and 15 of the Convention against Torture. Deporting someone to a country where

The Complexities of Involvement

315

he will probably undergo torture is a graver wrong than deporting someone to a country where torture-obtained evidence will probably be adduced against him in a trial. Similarly, admitting torture-obtained evidence into the proceedings of the courts in one’s own system of justice is a graver wrong than deporting someone to a country where the courts will probably admit such evidence into the proceedings whereby he is tried and convicted. On the other hand, the deportation of Qatada to Jordan in the absence of solid assurances from the Jordanians about the inadmissibility of torture-extracted evidence would have sullied the British system of legal governance. It would perhaps not have amounted to outright approval of the use of such evidence, but it would have amounted to a display of indifference. Though the British courts and the European Court of Human Rights are not generally responsible for the upstandingness of the procedures of the Jordanian courts, they become responsible when they are called upon to determine whether the deportation of someone to stand trial in Jordan is lawful. They are not morally obligated to gain assurances that the trial will be punctiliously fair in every respect, but they are morally obligated to gain assurances that it will comply with elementary standards of propriety. Because interrogational torture is always seriously wrongful (especially when it is wielded with the brutality that has been common in Jordan), the admission of evidence acquired through such torture would be a violation of those elementary standards. British courts would not be upholding the moral integrity of their own operations if they were to exhibit lackadaisicalness about the prospect of the use of such evidence in proceedings which they have effectively been called upon to authorize. Their vigilance has been in fulfilment of their moral duties as well as their legal responsibilities. The moral requisiteness of such vigilance derives from perpetrator-focused considerations. Qatada is a heinous extremist whose interests in not being confronted with torture-extracted evidence are ethically of little or no positive weight. Those interests could not render morally obligatory the endeavours that were undertaken by the British courts and Home Secretaries to ensure that he would not in fact be confronted with torture-extracted evidence. His interests were not sufficient to render morally obligatory the expenditure of the many hundreds of thousands of pounds that were bestowed on those endeavours. Nonetheless, those endeavours were indeed morally obligatory. Their requisiteness arose from the fact that Qatada’s interests were not the only matters of ethical significance at stake in the legal proceedings that eventuated in his deportation. Also directly at stake was the moral integrity of the British system of legal governance. In circumstances where the British courts were vested with a pivotal role in determining whether a trial of Qatada in Jordan would go ahead, the preservation of the moral integrity of the British system of governance required those courts to insist that any such trial be free of torture-tainted evidence. Qatada himself did not deserve such solicitude, but the factor of moral integrity was not focused on his deserts. It was focused instead on the propriety of the actions of British judges and other officials as they steered clear of implicating themselves (even indirectly and post hoc) in the use of interrogational torture.

316

Legal Responses to Torture

5.4. Conclusion Having become immersed in some of the minutiae of the legal issues that surround the problem of torture, this book has continued to invoke the broad philosophical themes that are expounded in its earlier chapters: the absolute wrongness of interrogational torture and most other types of torture; the importance of perpetrator-focused considerations; the occurrence of moral conflicts, and the consequent fact that moral optimality does not entail moral permissibility. Those broader themes have persistently surfaced as this book has come to grips with the nature of torture and the wrongness of torture and the germane legal responses to torture. Still, although several main philosophical ideas have served as lodestars throughout, they have not predetermined the details of my discussions. For example, my deontological insistence on the absoluteness of the impermissibility of interrogational torture and most other types of torture has not in itself explained why those sundry types of torture are indeed always and everywhere morally impermissible. Chapter 3 has marshalled an array of moral arguments to provide the pertinent moral explanations. Similarly, the aforementioned deontological insistence has not in itself explained why Dershowitz’s proposal for torture warrants or Posner’s and Vermeule’s proposal for the institutionalization of interrogational torture is morally unsound. The present chapter has marshalled some further moral arguments to substantiate my claims about the unsoundness of each of those proposals. While the arguments in this chapter have been crucially informed by the more abstract ideas developed in earlier chapters, they have in turn amplified and supported those ideas. They have supplied the more concrete ethical reasoning that links the abstractions to the policies which this book recommends. Hence, as has been stated at the close of Chapter 1, the layers of argumentation in this book—moving from levels of high abstraction to levels of considerably greater concreteness—are mutually reinforcing. Without the distinctions that are drawn and the general positions that are expounded in Chapter 1, the arguments in the later chapters would have been largely adrift. They would have lacked the framework that situates and fortifies them. Likewise, the lines of reasoning in the present chapter would have been adrift in the absence of the arguments advanced in Chapters 2–4. Throughout, each layer in the unfolding of this book has served as a basis for subsequent layers. In return, those subsequent portions of the volume have vindicated the earlier portions by showing why their theses are correct in application to an array of concrete moral cruxes. Overall, the book has aimed to establish that abstract matters of moral conflicts and deontology have a decisive bearing on the problem of torture and moral integrity.

References Addicott, Jeffrey. 2003–4. ‘Into the Star Chamber:  Does the United States Engage in the Use of Torture or Similar Illegal Practices in the War on Terror?’ 92 Kentucky Law Journal 849–912. Alexander, Larry. 2000. ‘Deontology at the Threshold’. 37 San Diego Law Review 893–912. Alexander, Larry, and Moore, Michael. 2007. ‘Deontological Ethics’. In Edward Zalta (ed.), Stanford Encyclopedia of Philosophy (Winter 2007 Edition). URL = . Allhoff, Fritz. 2005. ‘A Defense of Torture: Separation of Cases, Ticking Time-Bombs, and Moral Justification’. 19 International Journal of Applied Philosophy 243–64. Allhoff, Fritz. 2012. Terrorism, Ticking Time-Bombs, and Torture. Chicago:  University of Chicago Press. Alvarez, José. 2006. ‘Torturing the Law’. 37 Case Western Reserve Journal of International Law 175–223. Améry, Jean. 1980. At the Mind’s Limits. Translated by Sidney Rosenfeld and Stella Rosenfeld. Bloomington, IN: Indiana University Press. Originally published in German in 1966. Amnesty International. 1973. Report on Torture. London: Duckworth. Angell, Julie. 2005. ‘Ethics, Torture, and Marginal Memoranda at the DOJ Office of Legal Counsel’. 18 Georgetown Journal of Legal Ethics 557–69. Arrigo, Jean Maria. 2004. ‘A Utilitarian Argument against Torture Interrogation’. 10 Science and Engineering Ethics 543–72. Bagaric, Mirko, and Clarke, Julie. 2005. ‘Not Enough Official Torture in the World? The Circumstances in which Torture is Morally Justifiable’. 39 University of San Francisco Law Review 581–616. Bassiouni, M. Cherif. 2006. ‘The Institutionalization of Torture under the Bush Administration’. 37 Case Western Reserve Journal of International Law 389–425. Beccaria, Cesare. 1995. On Crimes and Punishments. In On Crimes and Punishments and Other Writings. Edited by Richard Bellamy. Translated by Richard Davies, Virginia Cox, and Richard Bellamy. Cambridge: Cambridge University Press. Originally published in Italian in 1764. Belvisi, Francesco. 2009. ‘The Ticking Bomb Scenario as a Moral Scandal’. In Bev Clucas, Gerry Johnstone, and Tony Ward (eds), Torture:  Moral Absolutes and Ambiguities. Baden-Baden: Nomos, pp 61–72. Bennett, Jonathan. 1966. ‘Whatever the Consequences’. 26 Analysis 83–102. Bennett, Jonathan. 1981. ‘Morality and Consequences’. In Sterling McMurrin (ed.), The Tanner Lectures on Human Values: 1981. Cambridge: Cambridge University Press, pp 45–116. Bennett, Jonathan. 1988. Events and Their Names. Oxford: Oxford University Press. Bennett, Jonathan. 1993. ‘Negation and Abstention:  Two Theories of Allowing’. 104 Ethics 75–96. Bennett, Jonathan. 1995. The Act Itself. Oxford: Oxford University Press. Bilder, Richard, and Vagts, Detlev. 2004. ‘Speaking Law to Power: Lawyers and Torture’. 98 American Journal of International Law 689–95.

318

References

Birkinshaw, Patrick. 2009. ‘English Law and Evidence Obtained by Torture: Vindication of Basic Principle or Judicial Abnegation? Implications of A v. Secretary of State for the Home Department’. In Bev Clucas, Gerry Johnstone, and Tony Ward (eds), Torture:  Moral Absolutes and Ambiguities. Baden-Baden: Nomos, pp 96–117. Bobbitt, Philip. 2008. Terror and Consent. London: Allen Lane. Botterell, Andrew. 2008. ‘In Defence of Infringement’. 27 Law and Philosophy 269–92. Brecher, Bob. 2007. Torture and the Ticking Bomb. Oxford: Blackwell. Brink, David. 1994. ‘Moral Conflict and Its Structure’. 103 Philosophical Review 215–47. Brunkhorst, Hauke. 2009. ‘Torture and Democracy’. In Bev Clucas, Gerry Johnstone, and Tony Ward (eds), Torture: Moral Absolutes and Ambiguities. Baden-Baden: Nomos, pp 73–82. Brunnée, Jutta, and Toope, Stephen. 2010. Legitimacy and Legality in International Law. Cambridge: Cambridge University Press. Bufacchi, Vittorio, and Arrigo, Jean Maria. 2006. ‘Torture, Terrorism and the State: A Refutation of the Ticking-Bomb Argument’. 23 Journal of Applied Philosophy 355–73. Cane, Peter. 2002. Responsibility in Law and Morality. Oxford: Hart Publishing. Card, Claudia. 2010. Confronting Evils: Terrorism, Torture, Genocide. Cambridge: Cambridge University Press. Chesterman, Simon. 2008. ‘Deny Everything: Intelligence Activities and the Rule of Law’. In Victor Ramraj (ed.), Emergencies and the Limits of Legality. Cambridge: Cambridge University Press, pp 314–33. Clark, Kathleen. 2005. ‘Ethical Issues Raised by the OLC Torture Memorandum’. 1 Journal of National Security Law and Policy 455–72. Coady, C. A. J. 2008. Messy Morality. Oxford: Oxford University Press. Coady, C. A.  J. 2011. ‘The Problem of Dirty Hands’. In Edward Zalta (ed.), Stanford Encyclopedia of Philosophy (Summer 2011 Edition). URL = . Cohan, John Alan. 2007. ‘Torture and the Necessity Doctrine’. 41 Valparaiso University Law Review 1587–632. Coleman, Jules. 1992. Risks and Wrongs. Cambridge: Cambridge University Press. Conee, Earl. 1982. ‘Against Moral Dilemmas’. 91 Philosophical Review 87–97. Constitution Project’s Task Force on Detainee Treatment. 2013. The Report of the Constitution Project’s Task Force on Detainee Treatment. Washington, DC: Constitution Project. Curzer, Howard. 2006. ‘Admirable Immorality, Dirty Hands, Ticking Bombs, and Torturing Innocents’. 44 Southern Journal of Philosophy 31–56. Danner, Mark. 2004. Torture and Truth. New York: New York Review Books. Darwall, Stephen. 1986. ‘Agent-Centered Restrictions from the Inside Out’. 50 Philosophical Studies 291–319. Davis, Michael. 2005. ‘The Moral Justifiability of Torture and Other Cruel, Inhuman, or Degrading Treatment’. 19 International Journal of Applied Philosophy 161–78. Dershowitz, Alan. 2002. Why Terrorism Works. New Haven: Yale University Press. Dershowitz, Alan. 2003. ‘The Torture Warrant:  A  Response to Professor Strauss’. 48 New York Law School Law Review 275–94. Dershowitz, Alan. 2004. ‘Tortured Reasoning’. In Sanford Levinson (ed.), Torture: A Collection. Oxford: Oxford University Press, pp 257–80. De Wijze, Stephen. 2009. ‘Recalibrating Steiner on Evil’. In Stephen de Wijze, Matthew H. Kramer, and Ian Carter (eds), Hillel Steiner and the Anatomy of Justice. New York: Routledge, pp 214–32. Dinello, Daniel. 1971. ‘On Killing and Letting Die’. 31 Analysis 83–6.

References

319

Donohue, Laura. 2008. The Cost of Counterterrorism. Cambridge: Cambridge University Press. Dworkin, Ronald. 2011. Justice for Hedgehogs. Cambridge, MA: Harvard University Press. Elshtain, Jean Bethke. 2004. ‘Reflection on the Problem of “Dirty Hands”’. In Sanford Levinson (ed.), Torture: A Collection. Oxford: Oxford University Press, pp 77–89. Enoch, David, and Marmor, Andrei. 2007. ‘The Case against Moral Luck’. 26 Law and Philosophy 405–36. Feinberg, Joel. 1973. Social Philosophy. Englewood Cliffs, NJ: Prentice-Hall. Feinberg, Joel. 1980. Rights, Justice, and the Bounds of Liberty. Princeton:  Princeton University Press. Ferzan, Kimberly Kessler. 2004. ‘Torture, Necessity, and the Union of Law & Philosophy’. 36 Rutgers Law Journal 183–90. Fletcher, George. 1985. ‘The Right and the Reasonable’. 98 Harvard Law Review 949–82. Fletcher, George. 1993. ‘The Nature of Justification’. In Stephen Shute, John Gardner, and Jeremy Horder (eds), Action and Value in Criminal Law. Oxford:  Oxford University Press, pp 175–86. Foot, Philippa. 1983. ‘Moral Realism and Moral Dilemma’. 80 Journal of Philosophy 379–98. Fried, Charles. 1978. Right and Wrong. Cambridge, MA: Harvard University Press. Friedman, Lawrence. 1993. Crime and Punishment in American History. New  York: Basic Books. Frowe, Helen. 2011. The Ethics of War and Peace. London: Routledge. Geras, Norman. 2011. Crimes against Humanity. Manchester:  Manchester University Press. Gewirth, Alan. 1981. ‘Are There any Absolute Rights?’ 31 Philosophical Quarterly 1–16. Ginbar, Yuval. 2008. Why Not Torture Terrorists? Oxford: Oxford University Press. Golash, Deirdre. 2007. ‘Torture and Self-Defense’. In Steven Lee (ed.), Intervention, Terrorism, and Torture: Contemporary Challenges to Just War Theory. Dordrecht: Springer, pp 263–71. Green, Penny, and Ward, Tony. 2009. ‘Torture and the Paradox of State Violence’. In Bev Clucas, Gerry Johnstone, and Tony Ward (eds), Torture: Moral Absolutes and Ambiguities. Baden-Baden: Nomos, pp 163–75. Greenberg, Karen, and Dratel, Joshua (eds). 2005. The Torture Papers:  The Road to Abu Ghraib. Cambridge: Cambridge University Press. Gross, Emanuel. 2001. ‘Legal Aspects of Tackling Terrorism:  The Balance between the Right of a Democracy to Defend Itself and the Protection of Human Rights’. 6 UCLA Journal of International Law and Foreign Affairs 89–168. Gross, Michael. 2010. Moral Dilemmas of Modern War. Cambridge: Cambridge University Press. Gross, Oren. 2004a. ‘Are Torture Warrants Warranted? Pragmatic Absolutism and Official Disobedience’. 88 Minnesota Law Review 1481–555. Gross, Oren. 2004b. ‘The Prohibition on Torture and the Limits of the Law’. In Sanford Levinson (ed.), Torture: A Collection. Oxford: Oxford University Press, pp 229–53. Gross, Oren. 2008. ‘Extra-Legality and the Ethic of Political Responsibility’. In Victor Ramraj (ed.), Emergencies and the Limits of Legality. Cambridge: Cambridge University Press, pp 60–93. Guiora, Amos, and Page, Erin. 2006. ‘The Unholy Trinity: Intelligence, Interrogation and Torture’. 37 Case Western Reserve Journal of International Law 427–47.

320

References

Gur-Arye, Miriam. 2004. ‘Can the War Against Terror Justify the Use of Force in Interrogations? Reflections in Light of the Israeli Experience’. In Sanford Levinson (ed.), Torture: A Collection. Oxford: Oxford University Press, pp 183–98. Haque, Adil Ahmad. 2007. ‘Torture, Terror, and the Inversion of Moral Principle’. 10 New Criminal Law Review 613–57. Hare, R. M. 1972. ‘Rules of War and Moral Reasoning’. 1 Philosophy and Public Affairs 166–81. Hare, R. M. 1981. Moral Thinking. Oxford: Oxford University Press. Hare, R. M. 1989. Essays in Ethical Theory. Oxford: Oxford University Press. Harel, Alon, and Sharon, Assaf. 2011. ‘“Necessity Knows No Law”: On Extreme Cases and Uncodifiable Necessities’. 61 University of Toronto Law Journal 845–65. Hart, H. L. A. 1963. Law, Liberty, and Morality. Oxford: Oxford University Press. Herman, Barbara. 1993. The Practice of Moral Judgment. Cambridge, MA:  Harvard University Press. Hill, Daniel. 2007. ‘Ticking Bombs, Torture, and the Analogy with Self-Defense’. 44 American Philosophical Quarterly 395–404. Hill, Thomas. 1991. Autonomy and Self-Respect. Cambridge: Cambridge University Press. Hill, Thomas. 1996. ‘Moral Dilemmas, Gaps, and Residues:  A  Kantian Perspective’. In H. E. Mason (ed.), Moral Dilemmas and Moral Theory. Oxford: Oxford University Press, pp 167–98. Himma, Kenneth. 2007. ‘Assessing the Prohibition against Torture’. In Steven Lee (ed.), Intervention, Terrorism, and Torture:  Contemporary Challenges to Just War Theory. Dordrecht: Springer, pp 235–48. Hughes, G. E., and Cresswell, M. J. 1996. A New Introduction to Modal Logic. London: Routledge. Hurley, Susan. 1989. Natural Reasons. Oxford: Oxford University Press. Ignatieff, Michael. 2004. The Lesser Evil. Edinburgh: Edinburgh University Press. Ip, John. 2009. ‘Two Narratives of Torture’. 7 Northwestern Journal of International Human Rights 35–77. Jacobs, James. 1989. Drunk Driving: An American Dilemma. Chicago: University of Chicago Press. Jessberger, Florian. 2005. ‘Bad Torture—Good Torture? What International Criminal Lawyers may Learn from the Recent Trial of Police Officers in Germany’. 3 Journal of International Criminal Justice 1059–73. Kadish, Sanford. 1989. ‘Torture, the State, and the Individual’. 23 Israel Law Review 345–56. Kamenova, Tsvetana. 2009. ‘Survey of the Crime of Torture in the Jurisprudence of the ICTY’. In Bev Clucas, Gerry Johnstone, and Tony Ward (eds), Torture: Moral Absolutes and Ambiguities. Baden-Baden: Nomos, pp 83–95. Kamm, Frances. 2007. Intricate Ethics. Oxford: Oxford University Press. Kamm, Frances. 2011. Ethics for Enemies. Oxford: Oxford University Press. Keller, Linda. 2005. ‘Is Truth Serum Torture?’ 20 American University International Law Review 521–612. Kershnar, Stephen. 2005. ‘For Interrogational Torture’. 19 International Journal of Applied Philosophy 223–41. Kipnis, Ken. 2007. ‘Prisons, POW Camps, and Interrogation Centers:  Reflections on the Juridic Status of Detainees’. In Steven Lee (ed.), Intervention, Terrorism, and Torture: Contemporary Challenges to Just War Theory. Dordrecht: Springer, pp 289–98. Koh, Harold. 2005. ‘A World without Torture’. 43 Columbia Journal of Transnational Law 641–61.

References

321

Kramer, Matthew. 1998. ‘Rights without Trimmings’. In Matthew H. Kramer, N. E. Simmonds, and Hillel Steiner, A Debate over Rights. Oxford: Oxford University Press, pp 7–111. Kramer, Matthew. 1999a. In Defense of Legal Positivism. Oxford: Oxford University Press. Kramer, Matthew. 1999b. In the Realm of Legal and Moral Philosophy. London: Macmillan Press. Kramer, Matthew. 2003. The Quality of Freedom. Oxford: Oxford University Press. Kramer, Matthew. 2004. Where Law and Morality Meet. Oxford: Oxford University Press. Kramer, Matthew. 2007. Objectivity and the Rule of Law. Cambridge:  Cambridge University Press. Kramer, Matthew. 2009a. Moral Realism as a Moral Doctrine. Oxford: Wiley-Blackwell. Kramer, Matthew. 2009b. ‘Consistency is Hardly Ever Enough:  Reflections on Hillel Steiner’s Methodology’. In Stephen de Wijze, Matthew H. Kramer, and Ian Carter (eds), Hillel Steiner and the Anatomy of Justice. New York: Routledge, pp 201–13. Kramer, Matthew. 2011. The Ethics of Capital Punishment: A Philosophical Investigation of Evil and Its Consequences. Oxford: Oxford University Press. Kreimer, Seth. 2003. ‘Too Close to the Rack and the Screw: Constitutional Constraints on Torture in the War on Terror’. 6 University of Pennsylvania Journal of Constitutional Law 278–325. Kreimer, Seth. 2005. ‘“Torture Lite,” “Full Bodied” Torture, and the Insulation of Legal Conscience’. 1 Journal of National Security Law and Policy 187–229. Kutz, Christopher. 2007a. ‘Torture, Necessity and Existential Politics’. 95 California Law Review 235–76. Kutz, Christopher. 2007b. ‘Causeless Complicity’. 1 Criminal Law and Philosophy 289–305. Langbein, John. 1977. Torture and the Law of Proof. Chicago: University of Chicago Press. Langbein, John. 2004. ‘The Legal History of Torture’. In Sanford Levinson (ed.), Torture: A Collection. Oxford: Oxford University Press, pp 93–103. La Torre, Massimo. 2009. ‘“Jurists, Bad Christians”: Torture and the Rule of Law’. In Bev Clucas, Gerry Johnstone, and Tony Ward (eds), Torture: Moral Absolutes and Ambiguities. Baden-Baden: Nomos, pp 10–38. Lee, Steven. 2012. Ethics and War: An Introduction. Cambridge: Cambridge University Press. Leon v. Wainwright. 1984. United States Court of Appeals, Eleventh Circuit. 734 Federal Reporter, Second Series 770–3. Levinson, Sanford. 2003. ‘“Precommitment” and “Postcommitment”: The Ban on Torture in the Wake of September 11’. 81 Texas Law Review 2013–53. Levinson, Sanford. 2004. ‘Contemplating Torture: An Introduction’. In Sanford Levinson (ed.), Torture: A Collection. Oxford: Oxford University Press, pp 23–43. Levinson, Sanford. 2005. ‘In Quest of a “Common Conscience”:  Reflections on the Current Debate about Torture’. 1 Journal of National Security Law and Policy 231–52. Levit, Kenneth. 2005. ‘The CIA and the Torture Controversy: Interrogation Authorities and Practices in the War on Terror’. 1 Journal of National Security Law and Policy 341–56. Lichtenberg, Judith. 1994. ‘The Moral Equivalence of Action and Omission’. In Bonnie Steinbock and Alastair Norcross (eds), Killing and Letting Die. New  York:  Fordham University Press, pp 210–29. Lim, C. L. 2008. ‘Inter Arma Silent Leges? Black Hole Theories of the Laws of War’. In Victor Ramraj (ed.), Emergencies and the Limits of Legality. Cambridge:  Cambridge University Press, pp 385–407. Locke, Don. 1982. ‘The Choice between Lives’. 57 Philosophy 453–75.

322

References

Luban, David. 2005. ‘Liberalism, Torture, and the Ticking Bomb’. 91 Virginia Law Review 1425–61. Luban, David. 2007. ‘Torture and the Professions’. 26 (2) Criminal Justice Ethics 2, 58–66. Luban, David. 2009. ‘Unthinking the Ticking Bomb’. In Charles Beitz and Robert Goodin (eds), Global Basic Rights. Oxford: Oxford University Press, pp 181–206. Lukes, Steven. 2005. ‘Liberal Democratic Torture’. 36 British Journal of Political Science 1–16. Lyons, David. 2008. ‘The Legal Entrenchment of Illegality’. In Matthew Kramer, Claire Grant, Ben Colburn, and Antony Hatzistavrou (eds), The Legacy of H. L. A. Hart: Legal, Political, and Moral Philosophy. Oxford: Oxford University Press, pp 29–43. McCain, John. 2006. ‘It’s about Us’. 33 (1) Human Rights 20–2. McConnell, Terrance. 1996. ‘Moral Residue and Dilemmas’. In H. E. Mason (ed.), Moral Dilemmas and Moral Theory. Oxford: Oxford University Press, pp 36–47. McConnell, Terrance. 2010. ‘Moral Dilemmas’. In Edward Zalta (ed.), Stanford Encyclopedia of Philosophy (Summer 2010 Edition). URL = . McMahan, Jeff. 2006. ‘Torture, Morality, and Law’. 37 Case Western Reserve Journal of International Law 241–8. May, Larry. 2007. ‘Humanity, Prisoners of War, and Torture’. In Steven Lee (ed.), Intervention, Terrorism, and Torture: Contemporary Challenges to Just War Theory. Dordrecht: Springer, pp 221–34. Mayerfeld, Jamie. 2007. ‘Playing by Our Own Rules: How U.S. Marginalization of International Human Rights Law Led to Torture’. 20 Harvard Human Rights Journal 89–140. Meisels, Tamar. 2008a. ‘Torture and the Problem of Dirty Hands’. 21 Canadian Journal of Law and Jurisprudence 149–73. Meisels, Tamar. 2008b. The Trouble with Terror. Cambridge:  Cambridge University Press. Mill, John Stuart. 1956. On Liberty. Edited by Currin Shields. Indianapolis: Bobbs-Merrill Company. Originally published in 1859. Miller, Seumas. 2005. ‘Is Torture Ever Morally Justifiable?’ 19 International Journal of Applied Philosophy 179–92. Miller, Seumas. 2011. ‘Torture’. In Edward Zalta (ed.), Stanford Encyclopedia of Philosophy (Summer 2011 Edition). URL  =  . Moher, Andrew. 2004. ‘The Lesser of Two Evils? An Argument for Judicially Sanctioned Torture in a Post-9/11 World’. 26 Thomas Jefferson Law Review 469–89. Moore, Michael. 1997. Placing Blame: A General Theory of the Criminal Law. Oxford: Oxford University Press. Moore, Michael. 2007. ‘Patrolling the Borders of Consequentialist Justifications:  The Scope of Agent-Relative Restrictions’. 27 Law and Philosophy 35–96. Murphy, Jeffrie. 1979. Retribution, Justice, and Therapy. Dordrecht: D. Reidel Publishing. Nagel, Thomas. 1979. Mortal Questions. Cambridge: Cambridge University Press. Nagel, Thomas. 1986. The View from Nowhere. Oxford: Oxford University Press. Nozick, Robert. 1974. Anarchy, State, and Utopia. New York: Basic Books. Nussbaum, Martha. 2000. ‘The Costs of Tragedy:  Some Moral Limits of Cost-Benefit Analysis’. 29 Journal of Legal Studies 1005–36. Oberdiek, John. 2004. ‘Lost in Moral Space: On the Infringing/Violating Distinction and Its Place in the Theory of Rights’. 23 Law and Philosophy 325–46.

References

323

Oberdiek, John. 2008. ‘What’s Wrong with Infringements (Insofar as Infringements are not Wrong): A Reply’. 27 Law and Philosophy 293–307. Osiel, Mark. 2004. ‘The Mental State of Torturers:  Argentina’s Dirty War’. In Sanford Levinson (ed.), Torture: A Collection. Oxford: Oxford University Press, pp 129–41. Parent, W. A. 1980. ‘Judith Thomson and the Logic of Rights’. 37 Philosophical Studies 405–18. Parfit, Derek. 1984. Reasons and Persons. Oxford: Oxford University Press. Parry, John. 2004. ‘Escalation and Necessity:  Defining Torture at Home and Abroad’. In Sanford Levinson (ed.), Torture: A Collection. Oxford: Oxford University Press, pp 145–64. Parry, John, and White, Welsh. 2002. ‘Interrogating Suspected Terrorists: Should Torture be an Option?’ 63 University of Pittsburgh Law Review 743–66. Paskins, Barrie. 1976. ‘What’s Wrong with Torture?’ 2 British Journal of International Studies 138–48. Peters, Edward. 1989. Inquisition. Berkeley: University of California Press. Peters, Edward. 1996. Torture. Expanded Edition. Philadelphia: University of Pen­n­syl­vania Press. Pettit, Philip. 1991. ‘Consequentialism’. In Peter Singer (ed.), A Companion to Ethics. Oxford: Blackwell, pp 230–40. Posner, Eric, and Vermeule, Adrian. 2006. ‘Should Coercive Interrogation be Legal?’ 104 Michigan Law Review 671–707. Posner, Eric, and Vermeule, Adrian. 2007. Terror in the Balance. Oxford: Oxford University Press. Posner, Richard. 2004. ‘Torture, Terrorism, and Interrogation’. In Sanford Levinson (ed.), Torture: A Collection. Oxford: Oxford University Press, pp 291–8. Quinn, Warren. 1993. ‘Actions, Intentions, and Consequences: The Doctrine of Doing and Allowing’. In Morality and Action. Cambridge: Cambridge University Press, pp 149–74. Rainbolt, George. 2006. The Concept of Rights. Dordrecht: Springer. Raviv, Adam. 2004. ‘Torture and Justification:  Defending the Indefensible’. 13 George Mason University Law Review 135–81. Raz, Joseph. 1979. The Authority of Law. Oxford: Oxford University Press. Reiman, Jeffrey. 1998. ‘Why the Death Penalty Should be Abolished in America’. In Louis Pojman and Jeffrey Reiman, The Death Penalty: For and Against. Lanham, MD: Rowman & Littlefield, pp 67–132. Rodin, David. 2008. ‘Torture, Rights, and Values:  Why the Prohibition of Torture is Absolute’. Carnegie Council Transcripts. URL  =  . Rosen, Michael. 2012. ‘Dignity Past and Present’. In Jeremy Waldron, Wai Chee Dimmock, Don Herzog, and Michael Rosen, Dignity, Rank, and Rights. Edited by Meir Dan-Cohen. Oxford: Oxford University Press, pp 79–98. Ross, W. D. 1930. The Right and the Good. Oxford: Oxford University Press. Rouillard, Louis-Philippe. 2005. ‘Misinterpreting the Prohibition of Torture under International Law: The Office of Legal Counsel Memorandum’. 21 American University International Law Review 9–41. Sands, Philippe. 2008. Torture Team:  Deception, Cruelty, and the Compromise of Law. London: Allen Lane. Scarry, Elaine. 1985. The Body in Pain. Oxford: Oxford University Press. Scarry, Elaine. 2004. ‘Five Errors in the Reasoning of Alan Dershowitz’. In Sanford Levinson (ed.), Torture: A Collection. Oxford: Oxford University Press, pp 281–90.

324

References

Scheffler, Samuel. 1985. ‘Agent-Centered Restrictions, Rationality, and the Virtues’. 94 Mind 409–19. Scheffler, Samuel. 1988. ‘Agent-Centered Restrictions, Rationality, and the Virtues’. In Samuel Scheffler (ed.), Consequentialism and Its Critics. Oxford:  Oxford University Press, pp 243–60. Scheffler, Samuel. 1994. The Rejection of Consequentialism. Revised Edition. Oxford: Oxford University Press. Scheppele, Kim Lane. 2005. ‘Hypothetical Torture in the “War on Terrorism”’. 1 Journal of National Security Law and Policy 285–340. Scholz, Sally. 2007. ‘War Rape’s Challenge to Just War Theory’. In Steven Lee (ed.), Intervention, Terrorism, and Torture:  Contemporary Challenges to Just War Theory. Dordrecht: Springer, pp 273–88. Seidman, Louis. 2005. ‘Torture’s Truth’. 72 University of Chicago Law Review 881–918. Shue, Henry. 1978. ‘Torture’. 7 Philosophy and Public Affairs 124–43. Shue, Henry. 2006. ‘Torture in Dreamland:  Disposing of the Ticking Bomb’. 37 Case Western Journal of International Law 231–9. Simester, Andrew. 2008. ‘Necessity, Torture, and the Rule of Law’. In Victor Ramraj (ed.), Emergencies and the Limits of Legality. Cambridge: Cambridge University Press, pp 289–313. Simester, Andrew. 2012. ‘On Justifications and Excuses’. In Lucia Zedner and Julian Roberts (eds), Principles and Values in Criminal Law and Criminal Justice. Oxford:  Oxford University Press, pp 95–112. Sinnott-Armstrong, Walter. 1988. Moral Dilemmas. Oxford: Basil Blackwell. Sinnott-Armstrong, Walter. 1996. ‘Moral Dilemmas and Rights’. In H. E. Mason (ed.), Moral Dilemmas and Moral Theory. Oxford: Oxford University Press, pp 48–65. Skolnick, Jerome. 2004. ‘American Interrogation: From Torture to Trickery’. In Sanford Levinson (ed.), Torture: A Collection. Oxford: Oxford University Press, pp 105–27. Soniewicka, Marta. 2013. ‘How Dangerous Can the Sterilized Needle Be? Torture, Terrorism, and the Self-Refutation of the Liberal-Democratic State’. In Bartosz Wojciechowski, Piotr Juchacz, and Karolina Cern (eds), Legal Rules, Moral Norms, and Democratic Principles. Frankfurt am Main: Peter Lang, pp 185–212. Steiner, Hillel. 1998. ‘Working Rights’. In Matthew H. Kramer, N. E. Simmonds, and Hillel Steiner, A Debate over Rights. Oxford: Oxford University Press, pp 233–301. Steinhoff, Uwe. 2006. ‘Torture—The Case for Dirty Harry and against Alan Dershowitz’. 23 Journal of Applied Philosophy 337–53. Steinhoff, Uwe. 2009. ‘Justifying Defensive Torture’. In Bev Clucas, Gerry Johnstone, and Tony Ward (eds), Torture: Moral Absolutes and Ambiguities. Baden-Baden: Nomos, pp 39–60. Strauss, Marcy. 2003. ‘Torture’. 48 New York Law School Law Review 201–74. Sung, Chanterelle. 2003. ‘Torturing the Ticking Bomb Terrorist:  An Analysis of Judicially Sanctioned Torture in the Context of Terrorism’. 23 Boston College Third World Law Journal 193–212. Sussman, David. 2005. ‘What’s Wrong with Torture?’ 33 Philosophy and Public Affairs 1–33. Sussman, David. 2006. ‘Defining Torture’. 37 Case Western Reserve Journal of International Law 225–30. Thomson, Judith Jarvis. 1986. Rights, Restitution, and Risk. Cambridge, MA:  Harvard University Press.

References

325

Thomson, Judith Jarvis. 1990. The Realm of Rights. Cambridge, MA:  Harvard University Press. Twining, William. 1978. ‘Torture and Philosophy—I’. 52 Proceedings of the Aristotelian Society, Supplementary Volumes 143–68. Twining, William, and Twining, P. E. 1973. ‘Bentham on Torture’. 24 Northern Ireland Legal Quarterly 305–56. Vallentyne, Peter. 1987. ‘Prohibition Dilemmas and Deontic Logic’. 117–18 Logique et Analyse 113–22. Vallentyne, Peter. 1989. ‘Two Types of Moral Dilemmas’. 30 Erkenntnis 301–18. Van der Vyver, Johan. 2003. ‘Torture as a Crime under International Law’. 67 Albany Law Review 427–63. Von Wright, G. H. 1963. Norm and Action. London: Routledge & Kegan Paul. Waldron, Jeremy. 1989. ‘Rights in Conflict’. 99 Ethics 503–19. Waldron, Jeremy. 1992. ‘Lex Talionis’. 34 Arizona Law Review 25–51. Waldron, Jeremy. 2010a. Torture, Terror, and Trade-Offs. Oxford: Oxford University Press. Waldron, Jeremy. 2010b. ‘Torture, Suicide and Determinatio’. 55 American Journal of Jurisprudence 1–29. Waldron, Jeremy. 2012a. ‘How Law Protects Dignity’. 71 Cambridge Law Journal 200–22. Waldron, Jeremy. 2012b. ‘Lecture 1: Dignity and Rank’. In Jeremy Waldron, Wai Chee Dimmock, Don Herzog, and Michael Rosen, Dignity, Rank, and Rights. Edited by Meir Dan-Cohen. Oxford: Oxford University Press, pp 13–46. Walzer, Michael. 1973. ‘Political Action: The Problem of Dirty Hands’. 2 Philosophy and Public Affairs 160–80. Ward, Tony, Johnstone, Gerry, and Clucas, Bev. 2009. ‘Introduction’. In Bev Clucas, Gerry Johnstone, and Tony Ward (eds), Torture:  Moral Absolutes and Ambiguities. Baden-Baden: Nomos, pp 1–9. Williams, Bernard. 1965. ‘Ethical Consistency’. 39 Proceedings of the Aristotelian Society (Supplement) 103–24. Williams, Bernard. 1966. ‘Consistency and Realism’. 40 Proceedings of the Aristotelian Society (Supplement) 1–22. Williams, Bernard. 1973. ‘A Critique of Utilitarianism’. In J. J.  C. Smart and Bernard Williams, Utilitarianism:  For and Against. Cambridge:  Cambridge University Press, pp 77–150. Williams, Bernard. 1995. ‘Acts and Omissions, Doing and Not Doing’. In Rosalind Hursthouse, Gavin Lawrence, and Warren Quinn (eds), Virtues and Reasons. Oxford: Oxford University Press, pp 331–40. Wisnewski, Jeremy. 2008a. ‘It’s about Time: Defusing the Ticking Bomb Argument’. 22 International Journal of Applied Philosophy 103–16. Wisnewski, Jeremy. 2008b. ‘Unwarranted Torture Warrants: A Critique of the Dershowitz Proposal’. 39 Journal of Social Philosophy 308–21. Wisnewski, Jeremy. 2009. ‘Hearing a Still-Ticking Bomb Argument: A Reply to Bufacchi and Arrigo’. 26 Journal of Applied Philosophy 205–9. Wisnewski, Jeremy. 2010. Understanding Torture. Edinburgh: Edinburgh University Press. Wisnewski, Jeremy, and Emerick, R. D. 2009. The Ethics of Torture. London: Continuum. Wolfendale, Jessica. 2006. ‘Training Torturers:  A  Critique of the “Ticking Bomb” Argument’. 32 Social Theory and Practice 269–87. Wonnell, Christopher. 2011. ‘Deontology, Thresholds, and Efficiency’. 17 Legal Theory 301–17.

Index absoluteness and agent-centredness  22–3, 222 consequentialism inadequate as basis for  179 of deontological constraints  221–2, 230–41 and moral optimality without permissibility 212–20 of moral prohibition on interrogational torture  116–27, 189, 199, 201, 253, 316 and rejoinders to Posner/Vermeule  275, 278–9, 281–2 and Slippery-Slope Illegitimacy Thesis  148–9 weak versus strong  8–9, 116–17, 125, 148–9, 212, 268, 279–80 Wisnewski’s inadvertent obscuring of 171, 172 Abu Ghraib  134 accountability ex post 287–309 favoured by Allhoff and Raviv  274 perceived by Dershowitz as unfair  256–7 under Legal Prohibition Thesis  249–59, 263, 267, 271–80 under the Posner/Vermeule regulatory scheme 276–7 Achieved Purpose Inquiry  105–6 act-impelling torture  68–9 action-guidance 18–19 actions and stationariness  80–6, 88–9, 91 volitionality of 84 act/omission distinction  24 n22, 69, 77–97 Bennett on morality of  195–6 and wrongness of torture  193–6, 202 Addicott, Jeffrey  59 n18, 243 n1 Afghanistan 146, 147 agent-centredness  21–4, 118, 221 and absoluteness  22–3, 222 rationality of  221–41 agent-neutrality 21–4 alcoholism 133–4 Alexander, Larry  10 n9, 26 n24, 27 n25, 115–16 Allhoff, Fritz on Dershowitz’s torture-warrant proposal 243 n1 ex post accountability favoured by  274 on institutionalization of torture  135 n6 on Slippery-Slope Illegitimacy Thesis 146, 148 on ticking bombs  59 n18 Alvarez, José  34 n3 American Civil Liberties Union  262 American definition of torture  34–6 American football  306–7 Améry, Jean  157

on consumingness of agony  162 on salvation-oriented torture  47 n12 Amnesty International  262 definition of torture by  36–7 on prolongedness of torture  109 Angell, Julie  34 n3 animals 55 and consumingness of agony  162, 168 Davis on  37, 50 and edifying torture  183 and ephemerally incapacitative torture  185 and extravagantly reckless torture  184 and overemphasis on autonomy  50, 150, 152, 155, 170–2 and self-betrayal  156 Antipodean emergency  61, 135, 254 apatheia 111 Argentinian junta  76–7, 123 Arrigo, Jean Maria  57 on Dershowitz’s torture-warrant proposal 243 n1 on humiliatingness of torture  73 n25 on institutionalization of torture  135–7, 145 on intimidatory torture  67 on ticking bombs  59 n18 attempted torture  105–9, 117, 164 Augustine of Hippo  47 n12 autonomy distinctively human at high level  150 overemphasis on  50, 55, 93, 149–55, 160, 162, 170 aversion therapy  47, 100–1, 112 Ayer, A.J.  78 Bagaric, Mirko  59 n18, 61 n20, 243 n1 baseball 171 basic moral status explication of  22 n20 impermissibility as 240 permissibility as 227 Bassiouni, M. Cherif  34 n3 Beccaria, Cesare  70–2, 165 Belvisi, Francesco  38 n6, 59 n18 Bennett, Jonathan  78–97, 195–6 Bentham, Jeremy  53–4, 70–1 biconditionals 141 Bilder, Richard  34 n3 Billy Budd 204 Birkinshaw, Patrick  312 n17, 313 Bobbitt, Philip  191, 252 n7 on Dershowitz’s torture-warrant proposal 243 n1 on German emergency  61 n20

328

Index

Bobbitt, Philip (Cont.) on interrogational torture  57 on intimidatory torture  67 on legal defences  291 on Leon v. Wainwright 60 n19 on moral conflicts  11 n10 on ticking bombs  59 n18 Botterell, Andrew  6 n5 boxing 306–7 Brady, Ian 72 Brecher, Bob  59 n18, 243 n1 Brink, David  15 n14, 16 n16 broadly addressed intimidatory torture  49 Brunkhorst, Hauke  46–7 Brunnée, Jutta  34 n3, 36 n5, 116 n1, 252 n7 Bufacchi, Vittorio  57 on Dershowitz’s torture-warrant proposal 243 n1 on humiliatingness of torture  73 n25 on institutionalization of torture  135 n6 on intimidatory torture  67 on ticking bombs  59 n18 burden of proof, under Legal Prohibition Thesis 274–7 Bush, George W.  34 n3, 147, 283, 286 calamity-averting interrogational torture  116, 134–49, 176, 179, 212–20 and abandonment of allegiances  130–2 contractualism ineffective as basis for a critique of  119–20, 122–7 and defencelessness/unthreateningness distinction 128–9 ephemerally incapacitative torture contrasted with  85–90, 94–7, 152–3, 192–7, 202, 219 general overview of  58–61 and inadmissibility of evidence  311–12 legal responses to  293–6, 300–4, 311 placatory torture similar to  62–6 potential optimality of  212–20 self-aggrandizement through  191–7, 201–3, 212, 214–15 and torture-warrant proposal  243–4, 254–9, 264 Cane, Peter  6 n5 capital punishment  189, 198–9, 207, 293 Card, Claudia  59 n18, 243 n1 ceteris paribus clause  232, 237–40 Chesterman, Simon  34 n3, 59 n18, 252 n7 civil-rights campaigns  260 civil sanctions  294–5 Clarke, Julie  59 n18, 61 n20, 243 n1 Clark, Kathleen  34 n3 Clucas, Bev  10 n9, 59 n18, 61 n20 Coady, C.A.J.  10 n9, 16 n15, 59 n18 coercion brutality contrasted with  285–6 potential legitimacy of  130–1

violence contrasted with  92–3, 154 Cohan, John Alan  59 n18, 60 n19, 200 n20, 243 n1 Coleman, Jules  6 n5 collective responsibility  298–9 collectively borne sanctions  296–7, 300–1 comas, painful averting of  47, 101–2, 113 n31 Conee, Earl  16 n16 confession-seeking interrogational torture  57– 8, 70–2, 244–5 Beccaria on  70–1, 165 opposed by Dershowitz  244–5 congenital analgesiacs  170 n13 consent in averting of a coma  101–2 to edifying torture  112–14 to experimentational torture  100 and harm principle  308–9 to placatory torture  223–4, 226–7 in resistance training  103–4, 182 to sado-masochistic torture  305–9 to therapeutic torture  102–3 vitiated by coercion  209–10 in wrestling contrasted with torture  180–2 consequentialism contrasted with deontology  1, 20–8, 221–3, 230–41 fact-sensitivity of  144–5, 179 and guilt/innocence distinction  65–6 indirect versions of  22 and maximization  12, 22 moral absolutes denied by  23, 179 and moral conflicts  11–14 and placatory torture  63–6 in Posner’s and Vermeule’s work  269–71, 279, 281–2 and punishment  204 and remedy-resembling measures  13–14 and slippery slopes  138, 140–5, 179 Constitution Project’s Task Force  34 n3 consumingness of agony  111–12 Beccaria on 165 Kreimer on 166 Luban on  165–6 Miller on 50 Scarry on  166–8 Sussman on  153–7, 161–4 and wrongness of torture  169, 174–6, 180–1, 185, 197 consumingness of euphoria  162–4 contractualist accounts of morality general doubts about  122, 126–7 Meisels’s version of  120–7 Nagel’s version of  118–20, 268 torture’s wrongness diagnosed by 118–27, 153 control by victim of pain absent in averting of a coma  101–2 absent in experimentational torture  100

Index absent in sado-masochistic torture  306–8 absent in therapeutic torture  102–3 absent in torture  112–14, 180–2, 185 in aversion therapy  100–1 in resistance training  45, 103–4 in wrestling  45 Convention against Torture  67 definition of torture in  30–4, 42, 50 on discriminatory torture  73 ephemerally incapacitative torture overlooked by 32 focus on purposes in  31–3 on indirect involvement in torture  309–15 only public officials covered by  31 Cresswell, M.J.  16 n16 criminal-injuries support scheme  299 criminal sanctions imposed on collectivities  294, 296–7 imposed on individual torturers  293–4, 299–300 Curzer, Howard  11 n10, 16 n15, 116 n1 on ticking bombs  59 n18 Danner, Mark  34 n3 Darwall, Stephen  21 n19 Davis, Michael  116 n1, 252 n7 on defencelessness of victims  37–9 and definition of torture  37–48, 50, 98 on humiliatingness of torture  73 n25 on lastingly incapacitative torture  76 on nonconsensuality of torture  44–5 on psychological torture  42–3 on punitive torture  39–42, 69 on sleep-deprivation  42 on suffering  33 on threshold deontology  10 n9 on ticking bombs  59 n18 on torture’s hostility toward victim 46–8, 111 ‘deep pockets’  297, 298 defeasibility of duties  10–11 defencelessness of victims  55–6 Davis on  37–9 Kamm on  38–9, 128–9 Shue on  127–31 Steinhoff on  38–9, 128–9 Sussman on 51 unthreateningness distinguished from  38 n6, 84–5, 94, 119–20, 127–8 definition of torture  30–56 by Amnesty International  36–7 in Convention against Torture  30–4 by Davis  37–48, 50, 98 flexibility needed in  104–14 formulation of 114 by Kershnar  48 by Miller  49–50 by philosophers  37–54 by Sussman  50–3

329

denunciatory theory of punishment  208 n21 deontic auxiliaries  246 n4, 247 deontological absolutism, as target of Posner/ Vermeule 267–71 Deontological Intensification Thesis  138–41 deontological permissions  25, 26 deontological requirements agent-centredness of 118 rationality of  221–41 deontology and agent-centredness  21–4 contrasted with consequentialism  1, 20–8, 221–3, 230–41 and intrinsic moral statuses  20–2 and moral conflicts  11–14 rationality of  25–6 n24, 221–41 threshold version of  10, 23–8 Dershowitz, Alan  59 n18, 142 confession-seeking torture opposed by 244–5, 311–12 on inadmissibility of torture-extracted evidence 311–12 on Justice Jackson  262–3 on killing versus torture  200 n20 on Korematsu v. United States 262–4 on Leon v. Wainwright 60 n19 on search warrants  247–8, 250, 255 and torture-warrant proposal  147, 243–66, 276, 277, 316 and warrantless torture  253–6, 259 on wiretapping  245–6 destruction of self as aim of torture  171–3 Scarry on  167–9 torture as  167–8, 170–3 Wisnewski on  169–73 de Wijze, Stephen  35 dignity 150–2 Dinello, Daniel  80 n27 Dirty Harry 61, 87 discriminatory torture  73 domestic battery  284, 285 Donohue, Laura  34 n3 Downey, Lesley Ann  72 Dratel, Joshua  34 n3 drunken driving  260 duress, defence of  292, 301–2 Dworkin, Ronald  18 edifying torture  46–8, 55, 98–104, 111–14, 177 animals as victims of  183 control by victim absent in  45 disregarded by Convention  32 disregarded by Shue  127 n4 moral integrity sometimes consistent with 190–1 no legal sanctions for legitimate instances of  287, 294–7, 300

330 edifying torture (Cont.) placatory torture contrasted with 210, 223 n2 potential legitimacy of  161 efficacy of torture, probability of  217–19, 290 Steinhoff on 218 Elshtain, Jean Bethke  215 n23, 243 n1 Emerick, R.D.  31, 59 n18, 243 n1 empirical inadequacies of Legal Prohibition Thesis  271–3 in Posner’s and Vermeule’s work  272–3 of Slippery-Slope Illegitimacy Thesis  145–9 of torture-warrant proposal  245–64 employer/employee relationship  297 Enoch, David  107 ephemerally incapacitative torture  77–97, 176 animals as victims of  185 disregarded by Convention  32 disregarded by Shue  127 n4 Hill on  94–7 interrogational torture contrasted with  85– 90, 94–7, 152–3, 161, 186–7, 192–7, 202, 219 Kamm on  90–1 killing similar to  198 no legal sanctions for legitimate instances of  287, 294–7, 300 potential legitimacy of  186–7 punitive torture contrasted with  203–5 euphoria, consumingness of  154, 162–4 evidence uncovered through torture  311–14 ex ante legal authorization of torture  243–82 of edifying or ephemerally incapacitative torture 287 exclusionary rule  311–14 executive pardons  251, 253 exoneration of torturers under Legal Prohibition Thesis  273–5, 280, 288 mitigation contrasted with  253, 258, 273–4, 279–80, 288–92, 300–4 Experience of Pain Inquiry  105–9, 114, 164 experimentational torture  99–100 animals as victims of  183–4 ex post accountability  287–309 extortionate torture  68 extradition and deportation  309–10, 312 extraordinary rendition  309–10 extravagant recklessness  36 animals as victims of  184 and placatory torture  62–3 torture committed through  74–6, 99, 105, 174 extreme emergencies, mitigative force of 289–91 fact-sensitivity of consequentialism  144–5 Feinberg, Joel on absoluteness  9 on infringements and violations  6, 7

Index on Permissibility Theorem  16 n16 on placatory torture  62 Ferzan, Kimberly Kessler  34 n3 fewer-versus-more, contrasted with none-versus-any 234–8 Finnis, John  21 fishing 265 n10 Fletcher, George  6 n5 Foot, Philippa  15–16 forfeiture of rights  186–8 Freedom House  262 Fried, Charles  10 n9 Friedman, Lawrence  73 Frowe, Helen  59 n18 on defencelessness  38 n6 on Dershowitz  243 n1 on killing versus torture  200–1, 281 on Sussman’s example of obese man  85 n29 victim-focused perspective of  200–1 Fuller, Lon  151 n9 Geras, Norman  vii, 36 German emergency  61, 135–6 Steinhoff on  87–8 Gewirth, Alan  24 n22 on placatory torture  40, 62, 64, 176, 209, 210, 223–7 Ginbar, Yuval  59 n18 on Bush Administration  34 n3 on Dershowitz  243 n1 on German emergency  61 n20 on Leon v. Wainwright 60 n19 on threshold deontology  10 n9 Golash, Deirdre  59 n18 gravity of torture’s wrongness  106–9, 190 bearing of consequences on  138 general considerations that bear on  215–20 mitigated by optimality  108 and specificities of cases  176 Green, Penny  59 n18, 215 n23 Greenberg, Karen  34 n3 Gross, Emanuel  59 n18 Gross, Michael  59 n18, 67 Gross, Oren  57, 252 n7, 253 on agent-centredness  24 n22 on Dershowitz  243 n1 on German emergency  61 n20 on international law and torture  262 n9 on Legal Prohibition Thesis  139 n7 on legal sanctions for torturers  258–9 on moral conflicts  11 n10 on threshold deontology  10 n9 on ticking bombs  59 n18 Guantánamo Bay  134, 146 Guiora, Amos  34 n3 Gur-Arye, Miriam  38 n6, 49 n13, 139 n7, 291 Hamlet 1 Haque, Adil  10 n9 Hare, R.M.  11 n10

Index on moral conflicts  16 Harel, Alon  116 n1 on threshold deontology  10 n9 harm principle  308–9 Hart, H.L.A.  308–9 Herman, Barbara  92–3, 154 Hill, Daniel  94–7 on defencelessness  38 n6 and excessive reliance on intuitions  97 Hill Jr, Thomas  16 n16, 24 n22 Himma, Kenneth on killing versus torture  200 n20 on moral conflicts  11 n10 on ticking bombs  59–60, 61 Hindley, Myra  72 Hitler, Adolf  132 Hughes, G.E.  16 n16 human-rights organizations  262 Human Rights Watch  262 humiliation 35–6 humiliative torture  73–4 hunting 265 n10 Hurley, Susan  2 hyperbole in Scarry’s work  167–9 in Wisnewski’s work  169–73 hypocrisy 243–5 Ignatieff, Michael on defencelessness of victims  38 on moral conflicts  12 n10 on torture as interrogational  56–7 illness as source of agony  158–9, 161 Scarry on  166–8 contrasted with torture  175 impermissibility absoluteness of  116–27, 201 as basic moral status  240 optimality consistent with  86, 90, 125, 193, 197, 212–20, 233, 240–1, 279–82, 288, 316 presupposed by weak justification  115, 120, 191 unavoidability of, in moral conflicts  13, 14, 18, 28, 229–30, 270–1 inadmissibility of torture-extracted evidence 311–14 incapacitative torture  69, 76–97 and act/omission distinction  77–97 ephemeral variety of  77–97 Kantian perspective on  92–3 lasting variety of  76–7 indirect consequentialism  22 indirect involvement in torture  309–15 infringement/violation distinction  6–7 institutional sanctions  295–6 institutionalization of torture  123–5, 132–49, 179, 272–3, 277, 316 instruments of torture  137 intentionality versus recklessness  62–3

331

intentions versus motives  75–6 International Criminal Tribunal for former Yugoslavia  73 n25, 75–6 international institutions  261–2 interpersonal trade-offs  222–6 interrogational techniques, non-torturous 117, 282–6 limits on protractedness of  283 specificity in guidance for  283–6 Waldron on  283–6 interrogational torture  41, 56–61 absolute wrongness of  278–9, 281–2 applied to close relatives or friends of interrogatees  42–3, 67–8, 128, 217, 291–2 confession-seeking variety of  70–2 mistaken applications of  125–6, 128 prospective versus retrospective  57–9 punishment distinguished from  70–2 as self-aggrandizement  191–7, 201–3, 212, 214–15 as sometimes punitive  70 intimidatory torture  49, 66–7, 204 intrapersonal trade-offs  222–31 intrinsic moral statuses  20–2 intuitions 97 investigative journalists  262 inviolability 187–8, 225–6 Ip, John  34 n3, 59 n18, 61 n20, 243 n1 Iraq 146, 147 Jackson, Robert  262–3 Jacobs, James  260 Jessberger, Florian  61 n20 Johnstone, Gerry  10 n9, 59 n18, 61 n20 Jordan 314–15 jury nullification  272, 293–5 Dershowitz on  251–62 lessened in trials of collectivities  297 in relation to drunken driving  260 in relation to lynchings  260 justification, strong versus weak  7–8 justification/excuse distinction Simester on  289 n14 straddled by mitigative defences  289–92 Kadish, Sanford  57, 116 n1, 258 on Legal Prohibition Thesis  139 n7 on threshold deontology  10 n9 Kamenova, Tsvetana  32 on humiliative torture  73 n25 on intentions versus motives  75–6 on vagueness of concept of torture  117 n2 Kamm, Frances  214, 215 on act-impelling torture  69 on agent-centredness  24 n22 on averting of comas  47, 101 on defencelessness of victims  38–9, 51, 128–9

332 Kamm, Frances (Cont.) on ephemerally incapacitative torture 90–1, 94 on killing versus torture  200 n20 on moral conflicts  19 on nonconsensuality of torture  44 n11 on threshold deontology  10 n9 Kant, Immanuel  19, 78 Kantianism autonomy overemphasized in  149–55, 160, 162 and incapacitative torture  92–3 and moral conflicts  11, 19 Sussman on  152–5 torture’s painfulness obscured by 153–4, 163 and wrongness of torture  149–61 Keller, Linda  34 n3 Kershnar, Stephen  40 n8, 59 n18 on aversion therapy  100 on brevity of torture  109 on killing versus torture  200 n20 on nonconsensuality of torture  44 n11, 48 torture defined by  48 killing analogized by Posner/Vermeule to torture 266–7, 280–2 contrasted with murder  103 contrasted with torture  89–90, 127, 154, 188–9, 197–201, 265 n10 similar to ephemerally incapacitative torture 198 King, Martin Luther  245–6 Kipnis, Ken  34 n3 Koh, Harold  34 n3, 252 n7 Korematsu v. United States 262–4 Kreimer, Seth  34 n3, 59 n18 on confession-seeking torture  71–2 on consumingness of agony  166 on defencelessness of victims  38 n6 on discriminatory torture  73 on intimidatory torture  67 on salvation-oriented torture  47 n12 Slippery-Slope Illegitimacy Thesis advocated by 141–2, 147 on ‘third degree’  58 on ‘torture-lite’  215 n23 on torture-warrant proposal  147, 243 n1 Kutz, Christopher  34 n3, 59 n18 on defencelessness of victims 38 n6 on legal defences  291 on threshold deontology  10 n9 Langbein, John  57 n17 last resort, torture as  126, 129, 198, 290, 291, 303 Kreimer on 141 and Legal Prohibition Thesis  249, 257, 258

Index and Minimal Invasion Principle  177–8, 195, 203 lastingly incapacitative torture  76–7 La Torre, Massimo  38 n6, 243 n1 lawyers’ fees, awardability of  260–1 Lee, Steven  6 n5 legal defences  288–92 legal positivism  151 n9 Legal Prohibition Thesis  139–40, 142, 148, 291 accountability under  249–59, 263, 267, 271–80 and burden of proof  274–7 empirical conjecturalness of  271–3 and exoneration of torturers  273–5, 280, 288 opposed to torture-warrant proposal 248–59, 263 Posner and Vermeule on  271–80 legal responses to torture  28, 242–315 and gravity of torture  108–9 and inefficacy of torture  257–9 legal sanctions  293–309 absolutist position on  275, 280 for non-optimal acts of torture  293, 294, 299–300 legitimacy, equivalent to permissibility  6 Legitimacy of Killing Inquiry  197–201 Leon v. Wainwright  60–1, 122, 123, 126, 135, 146, 149 Dershowitz on 251 destruction of victim’s agency not sought in 172, 173 and inappropriateness of individually borne criminal sanctions  293–4 and Minimal Invasion Principle  178 as moral conflict  271 and moral optimality  212–13, 216–19, 293–4 and torture warrants  254 Levinson, Sanford  30 n1, 57 on Bush Administration  34 n3 on Nagel  119 on threshold deontology  10 n9 on torture-warrant proposal  243 n1, 252–3, 262 on vagueness of concept of torture  117 n2 Levit, Kenneth  34 n3 lex talionis 152, 205–6 licences, contrasted with torture warrants 264–5 Lichtenberg, Judith  80 n27 Lim, C.L.  34 n3, 262 n9 Locke, Don  80 n27 Locke, John  120 logical coherence and moral conflicts  16–18 lotus-eaters 163 Luban, David  70 n22 on Bush Administration  34 n3 on consumingness of agony  165–6

Index on fear induced by torture  56 n16 on humiliation induced by torture  73 n25 on permissibility and obligatoriness  115–16 on ticking-bomb scenarios  59 n18, 124–5 on ‘torture-lite’  215 n23 Lukes, Steven  12 n10, 59 n18, 117 n2 lynchings 260 Lyons, David  260 Marmor, Andrei  107 Maslow, Abraham  247 masochists as victims of torture  163–4, 223 n2 maximin reasoning  122 maximization and consequentialism  12, 22, 231–41 maximizing conception of rationality  231–41 May, Larry  10 n9 Mayerfeld, Jamie  34 n3, 262 n9 McCain, John  188 McConnell, Terrance  6 n5, 16 n16 on action-guidance  18 McMahan, Jeff  67, 116 n1 medical care, contrasted with torture 47–8, 98–9 medieval torture  46–7, 57, 98, 244–5 Beccaria’s denunciation of  70 Meisels, Tamar  16 n15 on absoluteness of torture’s wrongness 120–7 on defencelessness of victims  38 n6 on Dershowitz  243 n1 on interrogational torture  57 on intimidatory torture  67 on Legal Prohibition Thesis  139 n7 on punitive torture  40 n8 on threshold deontology  10 n9 on ticking-bomb scenarios  59 n18, 60 on ‘torture-lite’  215 n23 Melville, Herman  204 Mengele, Josef  75, 99–100, 184 metastasizing tendency of torture  123–5, 132–49, 179, 272–3 Mill, John Stuart  308–9 Miller, Seumas  116 n1 on consumingness of torture  50 on defencelessness of victims  38 n6 on Dershowitz  243 n1 on psychological torture  42 n9 on punitive torture  40 n8 on ticking-bomb scenarios  59 n18, 61 on ‘torture-lite’  215 n23 torture defined by  49–50 Minimal Invasion Principle  177–8, 195, 203, 207, 214–15, 217 private individuals covered by  178 victim-focused character of  209 Minimization Principle  234–8 mistakes in applications of torture  125–6, 128 mitigation

333

attributable to consent  227 attributable to dire circumstances  229–30 exoneration contrasted with  219–20, 253, 258, 273–5, 279–80, 288–92, 294 and justification/excuse distinction  289–92 in legal proceedings  280, 289–92, 294 in moral conflicts  27 without optimality  290–1, 294 Mohammed, Khalid Sheikh  226 as monstrous victim of torture  187–8, 191 water-boarding of 70 Moher, Andrew  243 n1 monstrous victims of torture  187–8, 191–2, 196, 206, 208–9, 310, 314–15 Moore, Michael  5 n4, 115–16, 215 n23 on general structure of morality  24–8 on German emergency  61 n20 on legal defences  291 on moral conflicts  19 on nonconsensuality of torture  44 n11 on threshold deontology  10 n9, 25–8, 221–2 on torture by omission  32, 188 n16 moral conflicts  115–17, 191, 193, 197, 219, 291, 316 and action-guidance  18–19 chariness of  11–19 and consequentialism  11–14 and critique of Scheffler  240–1 and deontology  11–14 general account of  1–19 Hare on 16 and inappropriateness of individually borne criminal sanctions  293–4 Kant on 19 Kantians on 11 logical coherence of  16–18 logical form of  2 and Melvin’s placatory torture against Miranda 229–30 and mitigation  27 Moore on 19 and moral realism  14–16 onerousness of 19 and overall structure of morality  25–8 and Permissibility Theorem  16–18 persistence of duties in  10–13, 27 and rejoinders to Posner/Vermeule  268, 270–1, 279–82 and Slippery-Slope Illegitimacy Thesis 148–9 and Steinhoff’s example of optimal torture 214–15 and wrongness of torture  25–8 moral duties conflicts between  1–19, 219, 270–1, 279–82, 291, 293–4, 316 defeasibility of  10–11 and general structure of morality  25–8

334

Index

moral duties (Cont.) persistently existent in conflicts  10–13, 27, 191 potential optimality of contraventions of  197, 229–30, 240–1, 268 unavoidability of contraventions of  214–15 weak absoluteness of  148–9, 193 without permissibility  115–17 moral integrity  190–203, 206–12, 282, 288–9 moral luck  107, 256–7 moral necessity  25 n23 moral realism, and moral conflicts  14–16 moral relativism  21 n19 morality, general structure of  24–8 Moro, Aldo  68 Moussaoui, Zacarias  245–6 Much Ado about Nothing 111 n30 murder contrasted with killing  103 contrasted with torture  103 Waldron on 152 Murphy, Jeffrie  149–50 Nagel, Thomas  116 n1 on agent-centredness  21 n19 on moral conflicts  12 n10, 16 n15 Posner and Vermeule on  268 on threshold deontology  10 n9 on wrongness of torture  118–20 Nazis and psychological torture  35–6 and torture of Améry  157, 162 necessity, defence of  273–4, 291 New Testament  158 Nineteen Eighty-Four 35 No Contraventions Principle  233–8 nonconsensuality of torture  55, 98 in aversion therapy  100–1 Davis on  44–5 Kershnar on 48 in resistance training  103–4 Sussman on  51–3 none-versus-any, contrasted with fewer-versus-more 234–8 Nozick, Robert on experience machine  163 n11 on inviolability  225–6 teleological reconstruals of deontology opposed by  235–6 on threshold deontology  10 n9 Nussbaum, Martha on action-guidance  18 n17 Posner and Vermeule on  269–70 on tragic choices  270–1 Oberdiek, John  6 n5 obese man, Sussman’s example of  85–6, 91–3, 154, 172, 186 obligatoriness without permissibility  115–17, 191, 193, 197, 219

and critique of Posner/ Vermeule 270–1, 279–82 and critique of Scheffler  240–1 and Melvin’s placatory torture against Miranda 229–30 and Slippery-Slope Illegitimacy Thesis 148–9 and Steinhoff’s scenario of optimal torture 214–15 Odysseus 53 Odyssey 163 official mendacity  244–5, 250–62, 272 omissions sought through incapacitative torture 69, 84–97 as torturous measures  32, 188 n16 as volitional  91, 93, 95–6 ontological/epistemic conflation by Luban  166 by Scarry  167 optimality and critique of Scheffler  233, 240–1 as function of deontic stringency  8 n6 and gravity of torture  108 mitigation without  290–1, 294 in moral conflicts  8, 13, 14, 18, 27–8 of placatory torture  229–30 and rejoinders to Posner/ Vermeule 270–1, 279–82 of torture in extreme emergencies  149 without permissibility  86, 90, 115, 120, 125, 191, 193, 197, 201, 212–20, 288, 316 Orwell, George  35 Orwellianism 279 Osiel, Mark  76 overriding versus overtopping  10–11, 267–9 overtoppingness  3–7, 10–11, 26, 141 and Permissibility Theorem  17 Page, Erin  34 n3 pain, salutary roles of  197 n19 pain versus suffering  33 painfulness of torture  54 obscured by Kantians  153–4, 163 Parent, William  6 n5 Parfit, Derek  21 n19 Parry, John  139 n7, 243 n1 Paskins, Barrie  12 n10, 36 n5, 59 n18, 116 n1 permissibility absent in moral conflicts  13, 14, 18, 27–8 as basic moral status  22 n20, 227 equivalent to legitimacy  6 optimality contrasted with  86, 90, 115, 125, 191, 193, 197, 201, 212–20 strong versus weak  4–7, 268 Permissibility Theorem  16–18 perpetrator-focused perspective  118, 127, 173, 174, 316

Index and absoluteness of wrongness of torture 23–4 contrasted with agent-centredness  23–4 essential as supplement to victim-focused perspective  187–203, 206–12, 214 on indirect involvement in torture  310–15 on intrapersonal trade-off  228–9 and moral integrity of torturer  155, 161 neglected by Frowe  200–1 neglected by Posner/Vermeule  282 and sado-masochism  182 Peters, Edward  47 n12, 57 n17 Pettit, Philip  13, 59 n18 phobias 35 placatory torture  62–6, 301–2 calamity-averting interrogational torture similar to  63–6 and consequentialism  63–6 edifying torture contrasted with 210, 223 n2 and extravagant recklessness  62–3, 74 Feinberg on 62 Gewirth on  40–1, 62, 64, 176 and intrapersonal trade-off  222–30 wrongness of  209–10, 222–30 Posner, Eric  116 n1, 243, 252 n7, 316 on defencelessness of victims  38 n7 on deontological absolutism  267–71 empirical inadequacies in work of 148, 272–3 and ex ante authorization of torture  266–82 on exoneration of torturers under Legal Prohibition Thesis  273–5, 288 on fact-sensitivity of consequentialism  144 on killing versus torture  200 n20 on Legal Prohibition Thesis  271–80 on moral conflicts  12 n10 on Nagel  119 on Nussbaum  269–70 on remedy-resembling measures  269–70 on Slippery-Slope Illegitimacy Thesis 146, 148 on threshold deontology  10 n9 on torture warrants  243 n1, 276 n12 on tragic choices  268–71 Posner, Richard  139, 243 n1, 249, 263 Posner/Vermeule regulatory framework  316 as formal authorization of serious misconduct 277–8 fundamental moral objection to  277–8 operationally similar to Legal Prohibition Thesis 276–7 symbolic importance of  277–80 pre-trial detention  71 ‘prima facie’ epistemic sense of  2–3, 11, 12 ontological sense of  3–4 private individuals’ acts of torture  253, 287, 290, 298–309 covered by Minimal Invasion Principle  178

335

not usually collective responsibility for 298–9 withholding of legal sanctions for optimal instances of  300–4 proportionality 288 prosecutorial inaction  272, 293, 294 appropriate instance of  303 in relation to lynchings  260 and torture-warrant proposal  251, 253, 260–2 prospective interrogational torture  57–61 prospective/retrospective distinction  57–9 protection-of-oneself-or-others 291–2, 303–4 psychological torture  55 covered by American definition  34–6 Davis on  42–3 Miller on  42 n9 public officials as perpetrators of torture  55 covered by Convention against Torture  31 legal responses to  253, 275, 280, 287, 290–7, 299–301, 304 punitive torture  32, 69–72, 155 Davis on  39–42 ephemerally incapacitative torture contrasted with 203–5 Kantian view of  93 potential legitimacy of aim of  177 vindictiveness of  207–9 wrongness of  203–9 purposes of torture  54–74, 76–104, 110–11 alterations in 110 effects contrasted with  194–5 morally vitiating types or instances of 176–7, 189–90 omitted by American definition  34 specified in Convention against Torture 31–3 Qatada, Abu  314–15 Quinn, Warren  80 n27 Rainbolt, George  6 n5 rape 35, 74–5 rationality of deontology  25–6 n24, 221–41 Raviv, Adam  34 n3, 59 n18, 243 n1, 274 Rawls, John  120 Rawlsianism 122, 126 Raz, Joseph  3 n2 Reiman, Jeffrey  205–6 relational properties  81–2, 95 remedy-resembling measures  13–14, 269–70 resistance training  103–4, 112–13, 177 and consent  182 and control by victim  45 Sussman on  52–3 as torturous  46 responsibility for danger  291–2 restrictive confinement  197, 201–3 retributivism 204–9 retrospective interrogational torture  57–8, 123

336 revenge 205, 207–9 rightness, two senses of  8 Rodin, David  59 n18 Rosen, Michael  152 n10 Ross, David  3 Rouillard, Louis-Philippe  34 n3 sadistic torture  32, 36–7, 55, 72, 155 fanciful optimality of  212 n22 Kantian view of  93 sado-masochistic torture  46, 112–13 attitudinal wrongness of  305–6, 308 control by victim absent in  306–8 as deviance  305–8 and harm principle  308–9 legal responses to  211, 305–9 other types of sadistic torture contrasted with 72 Sussman on  44, 51–3 wrongness of  45–6, 163–4, 182, 189, 210–12 salvation-oriented torture  46–7, 98 Sands, Philippe  34 n3 satisficing 22 Scarry, Elaine on consumingness of agony  166–8 on Dershowitz  243 n1, 249 Dershowitz on  251, 256 on illness  166–8 on interrogational torture as punitive  70–1 on intimidatory torture  67 on ordinary items as instruments of torture 137 on techniques of torture  33 on ticking bombs  59 n18 Scheffler, Samuel  21 n19, 26 n24, 222–3, 231–41 Scheppele, Kim Lane  34 n3 on defencelessness of victims  38 n7 on Slippery-Slope Illegitimacy Thesis 142–3, 146–7 on ticking bombs  59 n18 on torture warrants  243 n1 Schindler’s List 84 Scholz, Sally  35 search warrants  247–8, 250, 255 Seidman, Louis  24 n22, 32, 116 n1 on defencelessness of victims  38 n6 on Dershowitz  243 n1, 266 on killing versus torture  200 n20 on psychological torture  42 n9 on punitive torture  40 n8 on ticking bombs  59 n18 on torture as coercion  43 n10 self-aggrandizement, in interrogational torture  191–7, 201–3, 212, 214–15 self-betrayal as acting against allegiances  130–2, 155–6 Améry on 157

Index as manoeuvring of victim’s body against itself 157–8 Shue on  130–2 Sussman on  155–61, 169, 173–5, 180, 193–4 as victim turned into instrument of torturer  156–8, 160–1, 169, 173–6, 180–1, 185, 193–4 sexual harassment  284 sexual intimacy  305–9 Shakespeare, William  1, 111 n30 Sharon, Assaf  10 n9, 116 n1 Shue, Henry on alcoholism  133–4 on defencelessness of victims  38, 84, 127–31 defencelessness and unthreateningness conflated by  127–8 on institutionalization of torture 132–8, 145 on intimidatory torture  67 on killing versus torture  127 on professionalization of torture  134–7, 145 on self-betrayal  130–2 on threshold deontology  10 n9 on ticking bombs  59 n18 victim-focused perspective of  127 on wrongness of torture  127–38 Simester, Andrew on excuses  303–4 on German emergency  61 n20 on infringement/violation distinction  6 n5 on justifications and excuses  289 n14 on legal defences  291 on Permissibility Theorem  16 n16 on torture as test of will  49 n13 simulacra of remedies  13–14, 269–70 simulation of torture  44, 45, 51–3, 112–13, 117 in aversion therapy  101 and control by victim  182 resistance training more than  103–4 by sado-masochists  306–7 Sinnott-Armstrong, Walter  5 n3, 6 n5, 12 n10, 16 n16 Sirens 53, 163 Skolnick, Jerome  260–1 sleep-deprivation mild versus severe  283, 286 physical as well as psychological  42 Slippery-Slope Illegitimacy Thesis  138–49 consequentialism of  138, 140–5, 179 Kreimer on  141–2, 147 empirical inadequacies of  145–9 Posner and Vermeule on  272 Scheppele on  142–3, 146–7 Soniewicka, Marta  34 n3, 59 n18, 61 n20, 243 n1 sorites paradoxes  117 stationariness as action  80–6, 88–9, 91

Index Steiner, Hillel  16 n16 Steinhoff, Uwe on defencelessness of victims  38–9, 51, 128–9, 180 on Dershowitz  243 n1 on Dirty Harry 61, 87 on efficacy of torture  218 on German emergency  61 n20 on killing versus torture  200 n20 on optimal-torture example  213–19 on rapist example  87–8 on snake-bite example  88–90 on ticking bombs  59 n18 Stoicism 111 strappado 162 Strauss, Marcy  59 n18, 60 n19 on Dershowitz  243 n1 on vagueness of concept of torture  117 n2 stringency, scalar character of  291 strong absoluteness  8–9 strong justification  7–8, 267, 268 strong permissibility  4–7, 17 suffering 33 Sung, Chanterelle  59 n18, 243 n1 Sussman, David autonomy overemphasized by  154–5, 160, 162 on consumingness of agony  153–7, 161–4 on defencelessness of victims  38 nn6–7, 51 on defencelessness/unthreateningness distinction 84–5 focused solely on interrogational and punitive torture  160–1 on Kantianism  152–5, 160 on nonconsensuality of torture  51–3 on obese-man example  84–6, 91, 93, 154, 172, 186 on punitive torture  40 n8 on sado-masochistic torture  44, 51–3 on self-betrayal  155–61, 169, 173–5, 180, 193–4 torture defined by  50–3 on utilitarian account of torture’s wrongness 152–3, 164 on wrongness of torture  152–61 symbolic importance of Posner/Vermeule regulatory framework 277–80 of torture warrants  265–6 table of types of torture  104 tax-avoidance 284 techniques of interrogation, legitimate versus torturous 282–6 techniques of torture differing gravity of  107, 215–16 diversity of 33 teleology 235–8 temporal restrictions on occurrence of torture  36, 48, 109–11

337

therapeutic torture  98–103 ‘third degree’  57–8, 123 Thomson, Judith Jarvis  6, 7 Thorough Compliance Objective  236–8 threshold deontology  10, 23–8, 221–2, 268 ticking-bomb scenarios  116, 129, 176, 179, 187, 191 and abandonment of allegiances  131, 132 and contractualist accounts of the wrongness of torture  119–20, 122–6 general introduction to  58–63 and institutionalization of torture  134–49 and mitigative defences  288–90 Posner and Vermeule on  268 and torture-warrant proposal  243–4, 254–9 Toope, Stephen  34 n3, 36 n5, 116 n1, 252 n7 Torquemada 244 torture act-impelling type of  68–9 animals as victims of  37, 50, 55, 170–2, 183–5 attempted torture contrasted with  105–9 capital punishment contrasted with 189, 198–9 consumingness of  50, 111–12, 161–8 defined by Amnesty International  36–7 defined by Bentham  53–4 defined by Convention against Torture 30–4, 50 defined by Davis  37–48, 50 defined by Kershnar  48 defined by Miller  49–50 defined by Sussman  50–3 definitions of  30–56 as destruction of self  167–8, 170–3 discriminatory type of  73 distinctiveness of  117–18 diversity of techniques of  33 edifying type of  98–104 extortionate type of  68 extravagantly reckless type of  74–6 extreme painfulness of  54 of heretics  46–7 humiliative type of  73–4 illness contrasted with  158–9, 161, 175 incapacitative type of  76–97 as indifferent or hostile to victim’s interests 46–8 indirect involvement in  309–15 institutionalization of  123–5, 132–49, 179, 272–3, 277 instruments of 137 interrogational type of  56–61 intimidatory type of  66–7 killing analogized by Posner/Vermeule to 266–7, 280–2 killing contrasted with  89–90, 127, 154, 188–9, 197–201 as last resort  126, 129, 141, 177–8, 195, 198, 203, 249, 257, 258, 290, 291, 303

338

Index

torture (Cont.) legal responses to  28, 242–315 medical care contrasted with  47–8, 98–9 metastasizing tendency of  123–5, 132–49, 179, 272–3 and Minimal Invasion Principle  177–8, 195, 203, 207 and moral conflicts  27–8 murder contrasted with  102 nature of  29–114 of non-autonomous beings  50, 55, 149–53, 155, 160, 162, 168, 170–2 and nonconsensuality  44–5 omissions as 32 perpetrated by private individuals  287, 290, 298–309 perpetrated by public officials  31, 55 and phobias  35 placatory type of  62–6 probable efficacy of  217–19, 290 professionalization of  134–7, 179 punitive type of  69–72 purposes of  31–3, 54–5 relatives or friends of interrogatee as victims of 42–3 restrictive confinement contrasted with 197, 201–3 sadistic type of  72 simulated by sado-masochists  306–7 table of types of  104 temporal restrictions on  36, 48, 109–11 as test of victim’s will  39–42, 49, 56 wrestling contrasted with  45, 175–6, 180–2 wrongness of  115–220 ‘torture-lite’ 215, 283 torture-narrow 91 torture-tainted evidence  311–14 torture-warrant proposal Dershowitz on  243–66, 276, 277, 316 distinctive as authorization of violent misconduct 264–5 empirical inadequacies of  245–64 fundamental moral objection to  264–6 Levinson on  252–3, 262 opposed by Kreimer  147 symbolic importance of  265–6 torture-wide 91 tragic choices Nussbaum on  269–70 Posner and Vermeule on  268–71 Twining, P.E.  53 n14, 59 n18, 67, 73 n25, 117 n2, 139 n7 Twining, William  36 n5, 53 n14 on aversion therapy  47, 100 on defencelessness of victims  38–9 on humiliatingness of torture  73 n25 on intimidatory torture  67 on Legal Prohibition Thesis  139 n7

on nonconsensuality of torture  44 n11, 45 on ticking bombs  59 n18 on vagueness of concept of torture  117 n2 unthreateningness versus defencelessness  38 n6, 84–5, 94, 119–20, 127–8 utilitarian account of torture’s wrongness 152–3, 164 Vagts, Detlev  34 n3 vagueness of concept of torture  117 Vallentyne, Peter  16 n16 Van der Vyver, Johan  59 n18, 243 n1, 262 n9 Veil of Ignorance  126 Vermeule, Adrian  116 n1, 243, 252 n7 on defencelessness of victims  38 n7 on deontological absolutism  267–71 on Dershowitz  243 n1, 276 n12 empirical inadequacies in work of 148, 272–3 and ex ante legal authorization of torture 266–82 on exoneration of torturers under Legal Prohibition Thesis  273–5, 288 on fact-sensitivity of consequentialism  144 institutionalization of torture advocated by 266–82, 316 on killing versus torture  200 n20 on Legal Prohibition Thesis  271–80 on moral conflicts  12 n10 on Nagel  119 on Nussbaum  269–70 on remedy-resembling measures  269–70 on Slippery-Slope Illegitimacy Thesis 146, 148 on threshold deontology  10 n9 on tragic choices  268–71 vicarious liability  297 victim-focused perspective  23–4, 117–18 of Frowe  200–1 insufficiency of  185–90, 268 on intrapersonal trade-off  228, 229 of Posner and Vermeule  282 of Shue  127 of Sussman  155, 161 victim-specific intimidatory torture  92 violence versus coercion  92–3 von Wright, G.H.  5 n4 Waldron, Jeremy  116 n1, 243 on agent-centredness  24 n22 on archetypal import of legal ban on torture 279, 280 on Bush Administration  34 n3 on coercion versus brutality  285–6 on Dershowitz  243 n1, 244 on guidance for interrogators  283–6 on legal positivism  151 n9 on murder  152

Index Posner and Vermeule on  279, 280 on ‘third degree’  58, 260 on threshold deontology  10 n9 on ticking bombs  59 n18 on ‘torture-lite’  215 n23 as unorthodox Kantian  152 n10 on vagueness of concept of torture  117 n2 on wrongness of torture  150–2 Walzer, Michael  12 n10 Ward, Tony on German emergency  61 n20 on threshold deontology  10 n9 on ticking bombs  59 n18 on ‘torture-lite’  215 n23 warrantless torture  253–6, 259 water-boarding  70, 157, 283, 286 weak absoluteness  8–9, 116–17, 148–9, 212, 268 and agent-centredness  23 weak justification  7–8, 115, 120, 191, 193, 197, 201, 212–20 weak permissibility  4–7, 17 White, Welsh  139 n7, 243 n1 Williams, Bernard on act/omission distinction  80 n27, 81 n28 on Jim and the Indians  225 n3 on moral conflicts  12 n10

339

on moral realism  14–16 wiretapping 245–6 Wisnewski, Jeremy  34 n3, 42 on aim of torture  171–3 autonomy overemphasized by  170 on Convention against Torture  31 on defencelessness of victims  38 n6 on Dershowitz  243 n1 on destruction of self in torture  169–73 on institutionalization of torture  135 n6 on moral conflicts  16 n15 on sado-masochism  44 n11 on salutary roles of pain  197 n19 on salvation-oriented torture  47 n12 on ticking bombs  59 n18, 125 n3 Wolfendale, Jessica  59 n18 Wonnell, Christopher  10 n9 wrestling, contrasted with torture  45, 175–6, 180–2 wrongness of torture  115–220 and act/omission distinction  193–6, 202 Kantian perspective on  149–61 Meisels on  120–7 Murphy on  149–50 Nagel on  118–20 Shue on  127–38 Waldron on  150–2