¹ R. Dawkins, ‘Let's All Stop Beating Basil's Car’, Edge, <http://edge.org/q2006/q06_9.html> (2006).

² Greene, J. and Cohen, J., ‘For the Law, Neuroscience Changes Nothing and Everything’, Philosophical Transactions of the Royal Society of London 359 (2004), pp. 1775 Google ScholarPubMed–85.

³ Greene and Cohen, ‘For the Law’, p. 1777.

⁴ Greene and Cohen, ‘For the Law’, p. 1778.

⁵ Greene and Cohen, ‘For the Law’, p. 1780.

⁶ Greene and Cohen, ‘For the Law’, p. 1781.

⁷ Greene and Cohen, ‘For the Law’, p. 1781.

⁸ Greene and Cohen, ‘For the Law’, p. 1777.

⁹ Greene and Cohen, ‘For the Law’, p. 1780.

¹⁰ Greene and Cohen, ‘For the Law’, p. 1784.

¹¹ Brigard, Cf. F. De, Mandelbaum, E. and Ripley, D., ‘Responsibility and the Brain Sciences’, Ethical Theory and Moral Practice 12 (2009), pp. 511 CrossRefGoogle Scholar–24.

¹² Nahmias, E., Morris, S. G., Nadelhoffer, T. and Turner, J., ‘Surveying Freedom: Folk Intuitions about Free Will and Moral Responsibility’, Philosophical Psychology 18 (2005), pp. 561 CrossRefGoogle Scholar–84; Nichols, S., ‘Folk Intuitions on Free Will’, Journal of Cognition and Culture 6 (2006), pp. 57–86 CrossRefGoogle Scholar; Nahmias, E., Coates, D. J. and Kvaran, T., ‘Free Will, Moral Responsibility, and Mechanism: Experiments on Folk Intuitions’, Midwest Studies in Philosophy 31 (2007), pp. 214 CrossRefGoogle Scholar–42; Nichols, S. and Knobe, J., ‘Moral Responsibility and Determinism: The Cognitive Science of Folk Intuitions’, Noûs 41 (2007), pp. 663 CrossRefGoogle Scholar–85; Nahmias, E. and Murray, D., ‘Experimental Philosophy on Free Will: An Error Theory for Incompatibilist Intuitions’, New Waves in Philosophy of Action, ed. Aguilar, J. H., Buckareff, A. A. and Frankish, K. (Basingstoke, 2010), pp. 189–215 Google Scholar; Nahmias, E., ‘Intuitions about Free Will, Determinism, and Bypassing’, The Oxford Handbook of Free Will, 2nd edn., ed. Kane, R. (Oxford, 2011), pp. 555 Google Scholar–75; Nichols, S., ‘Experimental Philosophy and the Problem of Free Will’, Science 331 (2011), pp. 1401 CrossRefGoogle ScholarPubMed–3; Nichols, S., ‘The Indeterminist Intuition’, The Monist 95 (2012), pp. 290–307 CrossRefGoogle Scholar; Shepherd, J., ‘Free Will and Consciousness: Experimental Studies’, Consciousness and Cognition 21 (2012), pp. 915 CrossRefGoogle ScholarPubMed–27; Weigel, C., ‘Experimental Evidence for Free Will Revisionism’, Philosophical Explorations 16 (2012), pp. 31–43 CrossRefGoogle Scholar; Rose, D. and Nichols, S., ‘The Lesson of Bypassing’, Review of Philosophy and Psychology 4 (2013), pp. 599–619 CrossRefGoogle Scholar; Deery, O., Davis, T. and Carey, J., ‘The Free-Will Intuitions Scale and the Question of Natural Compatibilism’, Philosophical Psychology 28 (2015), pp. 776–801 CrossRefGoogle Scholar. Sommers, T., ‘Experimental Philosophy and Free Will’, Philosophy Compass 5 (2010), pp. 199–212 CrossRefGoogle Scholar, provides a good, if already slightly dated, overview

¹³ Morse, Cf. S. J., ‘Compatibilist Criminal Law’, The Future of Punishment, ed. Nadelhoffer, T. A. (Oxford, 2013), pp. 107 CrossRefGoogle Scholar–31, for the former; and Kane, R., ‘Libertarianism’, Four Views on Free Will (Oxford, 2007)Google Scholar; Mele, A. R., Free: Why Science Hasn't Disproved Free Will (Oxford, 2014)CrossRefGoogle Scholar, for the latter. A number of writers have raised even further concerns with Greene and Cohen's argument. e.g. Mele, A. R., ‘Free Will, Science, and Punishment’, The Future of Punishment, ed. Nadelhoffer, T. A. (Oxford, 2014), pp. 177 Google Scholar–92; Pardo, M. S. and Patterson, D., ‘Neuroscience, Normativity, and Retributivism’, The Future of Punishment, ed. Nadelhoffer, T. A. (Oxford, 2014), pp. 133 Google Scholar–54.

¹⁴ Arguably the challenge would apply to retributivist theories that base the justification of criminal justice on the reasons of the public. However, I believe that any such theory faces serious difficulties, and even if it does not it will presumably still be subject to most if not all of the considerations I present in this article, so I shall not pursue the point independently; cf. F. K. Thomsen, ‘Why Should We Care What the Public Thinks? A Critical Assessment of the Claims of Popular Punishment’, Popular Punishment, ed. J. Ryberg and J. V. Roberts (Oxford 2014), pp. 119–145.

¹⁵ Note that for the nestled argument to work we have to generalize this premise to concern not merely libertarian free will. It is not entirely clear, it seems to me, whether Greene and Cohen mean to make this stronger claim, and their central argument does not rely on it. There are some indications that they believe that e.g. compatibilism will be similarly revealed by neuroscience to be mistaken, particularly in the discussion of their manipulation-case involving ‘Mr Puppet’. However, unlike the threat of determinism in the shape of a mechanical neuronal system, which libertarians must accept as a threat to free will, the manipulation challenge is not new and is unlikely to sway committed compatibilists. Kane, Cf. R. H., The Significance of Free Will (Oxford, 1996)Google Scholar; Pereboom, D., Living without Free Will (Cambridge, 2001)CrossRefGoogle Scholar; and for responses Fischer, e.g. J. M., ‘Responsibility, History and Manipulation’, Journal of Ethics 4 (2000), pp. 385 Google Scholar–91; Fischer, J. M., ‘Responsibility and Manipulation’, Journal of Ethics 8 (2004), pp. 145 CrossRefGoogle Scholar–77; McKenna, M., ‘A Hard-line Reply to Pereboom's Four-Case Manipulation Argument’, Philosophy and Phenomenological Research 77 (2008), pp. 142 CrossRefGoogle Scholar–59; Mickelson, K., ‘The Soft-Line Solution to Pereboom's Four-Case Argument’, Australasian Journal of Philosophy 88 (2010), pp. 595–617 Google Scholar.

¹⁶ Or, inversely, that criminal justice is unjust if condition Q obtains, and that neuroscience shows that Q does in fact obtain. These are logically equivalent, but the argument will be slightly easier to present using the formulation I have given it in the text.

¹⁷ Honderich, T., Punishment: The Supposed Justifications Revisited (London, 2006)Google Scholar; Berman, M. N., ‘Punishment and Justification’, Ethics 118 (2008), pp. 258 CrossRefGoogle Scholar–90; Boonin, D., The Problem of Punishment (New York, 2008)CrossRefGoogle Scholar; Zimmerman, M. J., The Immorality of Punishment (Toronto, 2011)Google Scholar.

¹⁸ To take just one prominent example, David Boonin argues that we should abolish punishment but preserve a system of criminal justice which requires offenders to provide restitution, and in so doing harms them as a foreseen side-effect. Cf. Boonin, Problem. Critics have responded that this still constitutes what we should conceive of as punishment. Cholbi, Cf. M., ‘Compulsory Victim Restitution is Punishment: A Reply to Boonin’, Public Reason 2 (2010), pp. 85–93 Google Scholar.

¹⁹ H. A. Bedau, ‘Punishment’, Stanford Encyclopedia of Philosophy, <http://plato.stanford.edu/entries/punishment> (2010), provides a good overview of some of the difficulties of defining punishment.

²⁰ We could also conditionalize the argument, that is, hold that if a system of criminal justice involves knowingly inflicting harm upon offenders in response to their wrongdoing, then . . . but the argument will be somewhat easier to present and discuss if we take the route suggested above. It is worth bearing in mind, however, that the boundaries of the qualifier ‘anything like its present shape’ are not vague, since they can be sharpened along precisely this condition if necessary. Note also that I take ‘knowingly causing harm’ to encompass both intentionally and foreseeingly causing harm; even if there is a moral difference between the two, no plausible theory will hold that foreseeingly causing harm is in and of itself morally permissible, and so both raise the problem of punishment.

²¹ The classical version of this critique is McCloskey, H. J., ‘Utilitarian and Retributive Punishment’, Journal of Philosophy 64 (1967), pp. 91–110 CrossRefGoogle Scholar. David Boonin provides a good overview of the criticism and responses. Boonin, Problem. Note also that arguably at least some rule-consequentialist theories of punishment avoid this implication. Rawls, Cf. e.g. J., ‘Two Concepts of Rules’, Philosophical Review 64 (1955), pp. 3–32 CrossRefGoogle Scholar. To the extent that they do, they too will be subject to the challenge I set out below.

²² Defining retributivism is neither easy nor uncontroversial and no extant version of retributivism holds only the view I here set out. Wood, Cf. D., ‘Punishment: Nonconsequentialism’, Philosophy Compass 5 (2010)Google Scholar; Matravers, M., ‘Is Twenty-first Century Punishment Post-Desert?’, Retributivism Has a Past – Has it a Future?, ed. Tonry, M. H. (Oxford, 2011), pp. 30–45 CrossRefGoogle Scholar; Berman, M. N., ‘Two Kinds of Retributivism’, The Philosophical Foundations of the Criminal Law, ed. Duff, R. A. and Green, S. (Oxford, 2011)Google Scholar, pp. 433–59; Hanna, N., ‘Retributivism Revisited’, Philosophical Studies 167 (2014), pp. 473 CrossRefGoogle Scholar–84. My argument in the present does not depend, however, on exactly which or how many theories of the justification of punishment can be labelled retributivist, so I shall not pursue that discussion. I mean for the condition offered above only to capture one defining property of a set of theories about the justification of criminal justice.

²³ This is differently phrased and slightly broader than some formulations of the distinction, which would say e.g. that ‘we ought to punish a person as she deserves to be punished (no more, no less)’. I take my formulation here to be consistent with this idea, but it usefully highlights the particular role desert plays in a way that will help show how we can generalize negative retributivism to a broader set of deontological theories below. Note also that this is not to say, of course, that a person's desert is alone sufficient reason to justify harm. Retributivists will impose further conditions, such as e.g. that the harm be inflicted by a proper authority, expresses condemnation, and is proportional. The definitions given here are not an exhaustive statement of any theory of retributivist criminal justice, but they are not intended to be. They establish only the logical distinction between two types of retributivist claims. Note also that the theory can take at least two shapes. One version could hold that desert generates a reason to harm, another that there is a desert-independent reason for harming whose strength nonetheless varies with desert.

²⁴ As above, this is a slightly revisionist definition, where some might hold instead e.g. that ‘we ought to punish a person no more than she deserves’. Again, I take my version to be consistent with this but to highlight the particular role desert plays in a way that will allow us to generalize the point. Note also that typically the reason will be held to ground a desert-sensitive constraint on the instances, forms and/or severity of harm that may justifiably be inflicted on a person. Again, the theory can take at least two shapes. One version could hold that the constraints are themselves based on desert (‘the innocent deserve not to suffer’), another that there are non-desert-derived constraints against inflicting suffering that are nonetheless desert-sensitive (i.e. the desert-independent reasons against harm are weakened or cancelled by desert-generated reasons).

²⁵ Indeed, the negative thesis can form the basis of a number of different challenges. Cf. e.g. Roebuck, G. and Wood, D., ‘A Retributive Argument Against Punishment’, Criminal Law and Philosophy 5 (2011), pp. 73–86 CrossRefGoogle Scholar.

²⁶ My version here is modified to fit the present analysis, but preserves, I believe, the gist of Feldman's claim. Feldman, Cf. F., ‘Desert: Reconsideration of Some Received Wisdom’, Mind 104 (1995), pp. 63–77 CrossRefGoogle Scholar, at 64. Note also that retributive desert is sometimes held to presuppose responsibility as a matter of conceptual necessity, since retributive desert is what is at stake in punishment, and the concept of punishment, it is suggested, entails responsibility. In Geoffrey Cupit's words: ‘it is not that [people who are not responsible] cannot deserve punishment; rather that they cannot deserve punishment’ ( Cupit, G., ‘Desert and Responsibility’, Canadian Journal of Philosophy 26 (1996), pp. 83–100 CrossRefGoogle Scholar, at 92; cf. also J. Kleinig, Punishment and Desert (The Hague, 1973), pp. 57–8). Again, however, in the present context nothing turns on the issues of how we define punishment and retributive desert, so long as we focus our attention on the claims that the theories stake.

²⁷ Note that we include the qualification ‘ordinarily’ to accommodate non-absolutist versions that recognize a threshold beyond which competing telic considerations do in fact outweigh the moral constraint on knowingly causing harm. Absolutist theories, of course, will hold the stronger claim that it is ever defeated only by reasons grounded in a person's responsibility. I do not think, however, that anything about the argument at hand turns on the distinction. Note also that a reason is defeated if it is cancelled, overridden or outweighed, but it is not important for present purposes which of these defeats it.

²⁸ Even if one adopts a very wide conception of retributivist theories there are explicitly non-retributivist, non-consequentialist theories of the justification of criminal justice that will encounter the neuroscience challenge because they affirm the criminal responsibility precondition-thesis. David Boonin's previously mentioned theory is one. Another recent candidate is Victor Tadros's theory, which draws on ideas developed in the context of the discussion of just war to argue (roughly) that offenders make themselves liable to be harmed in an extended form of self-defence because they are responsible for moral wrongdoing that impermissibly threatens the interests of their victims. Tadros, V., The Ends of Harm: The Moral Foundations of Criminal Law (Oxford, 2011)CrossRefGoogle Scholar. While harm is here justified as the response to offenders constituting (or having constituted) an impermissible threat, it is important for this theory to distinguish between culpable and innocent threats, which turns again on their being morally responsible and irresponsible, respectively, for constituting such threat.

²⁹ Morse, S. J., ‘New Neuroscience, Old Problems’, Neuroscience and the Law, ed. Garland, (New York, 2004), pp. 157 Google Scholar–98; Morse, S. J., ‘Brain Overclaim Syndrome and Criminal Responsibility: A Diagnostic Note’, Ohio State Journal of Criminal Law 3 (2006), pp. 397–412 Google Scholar; Morse, S. J., ‘The Non-Problem of Free Will in Forensic Psychiatry and Psychology’, Behavioral Sciences and the Law 25 (2007), pp. 203 CrossRefGoogle ScholarPubMed–20; Morse, S. J., ‘Lost in Translation? An Essay on Law and Neuroscience’, Law and Neuroscience, ed. Freeman, (Oxford, 2011), pp. 529 CrossRefGoogle Scholar–62; Morse, S. J., ‘Common Criminal Law Compatibilism’, Neuroscience and Legal Responsibility, ed. Vincent, N. A. (Oxford, 2013), pp. 29–52 Google Scholar; Morse, ‘Compatibilist’. Morse evidently faces the difficulty that also partly motivates this article: that the neuroscience challenge to criminal justice is more talked about than set out. Thus, while his papers repeatedly refer to the challenge as a view commonly presented and debated among legal scholars and cognitive scientists, he does not reference a published version.

³⁰ Morse, ‘Common’, p. 29.

³¹ Note that Morse makes an even wider argument, relying on a distinction between ‘internal’ and ‘external’ critiques of criminal justice: ‘An internal argument accepts that criminal responsibility is a coherent concept and uses free will to explain the positive rules and practices we have or to criticize these rules and practices normatively for the purpose of improving them. An external argument uses free will to demonstrate that the concept of criminal responsibility is incoherent or unjustifiable and therefore it should be abandoned’ (Morse, ‘Non-Problem’, p. 204; cf. also Morse, ‘Brain Overclaim’, pp. 398–9). The gist of his response to the former is to show that the neuroscience challenge to free will and responsibility does not directly affect criminal justice because positive law does not rely on these notions. I shall focus only on the latter. There, much of Morse's response to the two challenges to the justification of criminal justice (hard determinism and dissolution of agency) takes the form of an extended tu quoque setting out the alleged failures and unattractive implications of the views on which the challenges are based. e.g. Morse, ‘Common’, pp. 42–6. This I shall largely ignore in order to focus on Morse's positive argument for the justification of criminal justice in the face of the challenges, although I have critically discussed it previously. See F. K. Thomsen, `There but for the Grace of My Orbitofrontal Cortex…', Criminal Justice Ethics 33 (2014), pp. 220–35.

³² Morse, ‘New Neuroscience’, p. 164; cf. also Morse, ‘Brain Overclaim’, p. 399; Morse, ‘Non-problem’, p. 205; Morse, ‘Lost in Translation?’, p. 530; Morse, ‘Common’, p. 31.

³³ Morse, ‘New Neuroscience’, p. 163. Cf. also: ‘It could not be otherwise. Laws could not guide people ex ante and ex post unless people were the types of creature who could use laws as premises in their practical reasoning. . .. Unless human beings were rational creatures who could understand the good reasons for action, including the relevant facts and rules, and could conform to legal requirements through intentional action, the law would be powerless to affect human action. Legally responsible agents are therefore people who have the general capacity to grasp and be guided by good reason’ (Morse, ‘Non-problem’, pp. 205–6).

³⁴ Morse, ‘New Neuroscience’, p. 168; cf. also Morse, ‘Brain Overclaim’, pp. 401–2; Morse, ‘Lost in Translation?’, pp. 543–5.

³⁵ The phrasing here is also imprecise. Morse clearly has in mind the claim that bodily movements absent consciousness that causes the bodily movements are not actions.

³⁶ To pre-empt misunderstanding, I do not mean here to exclude wrongdoing through negligence, that is, failing to perform a morally required act. I take it that agents could, if they met other conditions of moral responsibility, be responsible for such ‘omissions’ (although I should prefer to say simply that they are responsible for whatever act that they did perform, and that performing this act was wrong because they ought to have performed another act that was required). Note also that the argument may still work, although it will require additional premises and as a result be slightly more complicated, if agents can be responsible for certain things apart from their actions (e.g. character traits that they develop as a result of their actions).

³⁷ Morse, ‘New Neuroscience’, p. 167.

³⁸ Cf. Libet, B. W., ‘Unconscious Cerebral Initiative and the Role of Conscious Will in Voluntary Action’, Behavioral and Brain Sciences 8 (1985), pp. 529 CrossRefGoogle Scholar–66; Libet, B. W., ‘Do We Have Free Will?’, Journal of Consciousness Studies 6 (1999), pp. 47–57 Google Scholar. Similar arguments have been prominently presented in D. M. Wegner, The Illusion of Conscious Will 2003). and Gazzaniga, M. S., Who's in Charge? Free Will and the Science of the Brain (New York, 2011)Google Scholar. Here, although Morse does not fully develop an argument, his view is supported by persuasive criticism e.g. in Levy, N., ‘Libet's Impossible Demand’, Journal of Consciousness Studies 12 (2005), pp. 67–76 Google Scholar; Mele, A. R., Effective Intentions: The Power of Conscious Will (Oxford, 2009)CrossRefGoogle Scholar; Mele, Free.

³⁹ Morse, ‘Non-Problem’, p. 214.

⁴⁰ Morse, ‘New Neuroscience’, p. 175. Note that this badly misrepresents the relation between consequentialist considerations (which, I take it, are what Morse has in mind), and hard determinism. Hard determinism cannot produce this system, or any system, since it is simply a theory of free will. At most it has implications such that if hard determinism is true we can know that certain theories are false, but not, importantly, that any particular theory is true. Similarly, it is unsurprising that it cannot tell us ‘which goals we should pursue’, since this is not what it is meant to do, and for the same reason its inability in this respect cannot plausibly count against it.

⁴¹ Good overviews illustrating the complexity and controversy are M. McKenna, ‘Compatibilism’, Stanford Encyclopedia of Philosophy, <http://plato.stanford.edu/entries/compatibilism/> (2009), and Levy, N. and McKenna, M., ‘Recent Work on Free Will and Responsibility’, Philosophy Compass 4 (2009), pp. 96–133.CrossRefGoogle Scholar I elaborate on the implications of this complexity for the neuroscience challenge in the next section.

⁴² Morse, ‘New Neuroscience’, p. 179.

⁴³ Morse recognizes that reasons-responsiveness must come in degrees when he writes that: ‘syndromes and other causes do not have excusing force unless they sufficiently diminish rationality in the context in question’ (Morse, ‘New Neuroscience’, p. 180). Note too that for this claim to be plausible, rationality must encompass not only an agent's ability to respond to reasons, but also her ability to shape her actions accordingly. Else the claim that only the diminishment of rationality suffices for decreasing responsibility would be subject to strongly intuitive counterexamples of decreased agent-control, e.g. by a compulsion that, contrary to Morse's interpretation, overrode reasons rather than blocking access to them.

⁴⁴ Dennett, D., Elbow Room: The Varieties of Free Will Worth Wanting (Oxford, 1984)Google Scholar; Fischer, J. M. and Ravizza, M., Responsibility and Control: A Theory of Moral Responsibility (Cambridge, 2000)Google Scholar. Morse himself credits Wallace, R. Jay, Responsibility and the Moral Sentiments (Cambridge, MA, 1994)Google Scholar, whose work is in turn indebted to P. F. Strawson's famous discussion of the problem. See Strawson, P., ‘Freedom and Resentment’, Freedom and Resentment and Other Essays (Abingdon, 2008).Google Scholar

⁴⁵ Morse, ‘Non-Problem’, p. 206.

⁴⁶ Morse, ‘Non-Problem’, p. 214. Although of lesser importance it is perhaps worth noting that the division of ‘answers’ here is muddled. The first is actually two claims in one, the first half of which is presumably a version of the third answer. That is, Morse's compatibilism really holds: (1) we have good reasons to adopt certain responsibility practices simply because they are ‘normatively desirable’, (2) these practices do not need to conform to any particular metaphysical facts, and (3) nevertheless, it just so happens that the practices which we ought to adopt for other reasons are compatible with determinism.

⁴⁷ Morse, ‘New Neuroscience’, p. 177.

⁴⁸ Smart, J. J. C., ‘Free Will, Praise and Blame’, Mind 70 (1963), pp. 291–306 Google Scholar; Arneson, R., ‘The Smart Theory of Moral Responsibility and Desert’, Desert and Justice, ed. Olsaretti, S. (Oxford, 2007), pp. 233 Google Scholar–58.

⁴⁹ Vargas, M., ‘The Revisionist's Guide to Responsibility’, Philosophical Studies 125 (2005), pp. 399–429 CrossRefGoogle Scholar; Vargas, M., Building Better Beings: A Theory of Moral Responsibility (Oxford, 2013)CrossRefGoogle Scholar.

⁵⁰ Morse, ‘Brain Overclaim’, p. 402.

⁵¹ Roughly two thirds of respondents to the Philpapers survey affirm or lean towards compatibilism; a proportion which remains more or less constant when respondents are narrowed to specialists in metaphysics, philosophy of action, applied ethics, or normative ethics. Cf. <http://philpapers.org/surveys/results.pl>. Recall also that we saw in connection with reviewing Greene and Cohen's argument that there is conflicting evidence on what the intuitions of the broader public are on this topic.

⁵² A different and arguably more promising possibility is to show that for every extant theory of the conditions of moral responsibility there is at least one condition that can be demonstrated by neuroscience to be unattainable. Given the complexity of the free will and moral responsibility debate, to complete this argument will require enormous neuroscientific labour, but even partially completing it might be worthwhile: presumably we should adjust our credence in the possibility of moral responsibility in the light of the elimination of candidate theories, since even if not every theory of moral responsibility has been falsified, the fact that some have gives us more reason to think that agents cannot be morally responsible.

⁵³ Bode, S., He, A. H., Soon, C. S., Trampel, R., Turner, R. and Haynes, J.-D., ‘Tracking the Unconscious Generation of Free Decisions Using UItra-High Field fMRI’, PLoS ONE 6 (2011)CrossRefGoogle ScholarPubMed; Soon, C. S., Brass, M., Heinze, H.-J. and Haynes, J.-D., ‘Unconscious Determinants of Free Decisions in the Human Brain’, Nat Neurosci 11 (2008), pp. 543 CrossRefGoogle ScholarPubMed–5. Evaluating this literature is beyond the scope of this article, though it is probably fair to say that the difficulty of experimentally undermining free will and moral responsibility was initially underestimated by many participants, that difficulty resulting at least partly from the complexity of and disagreement about theories of free will and moral responsibility that I note below. For good overviews of the literature on experimental challenges to free will and moral responsibility, Roskies, cf. A. L., ‘Neuroscientific Challenges to Free Will and Responsibility’, Trends in Cognitive Sciences 10 (2006), pp. 419 CrossRefGoogle ScholarPubMed–23; Sie, M. and Wouters, A., ‘The BCN Challenge to Compatibilist Free Will and Personal Responsibility’, Neuroethics 3 (2010), pp. 121 CrossRefGoogle ScholarPubMed–33; Shepherd, J., ‘Scientific Challenges to Free Will and Moral Responsibility’, Philosophy Compass 10 (2015), pp. 197–207 CrossRefGoogle ScholarPubMed.

⁵⁴ Consider differences between compatibilists and non-compatibilists, volitionists and attributionists, leeway and source-theories, historical and non-historical theories, reasons-responsive, dispositionalist, and agent-consistent views, the transfer of non-responsibility and the ultimacy condition, event- and agent-causality, etc. Good overviews of the range of extant theories and conditions are provided by K. Timpe, ‘Free Will’, Internet Encyclopedia of Philosophy, <http://www.iep.utm.edu/freewill/> (2006); K. Vihvelin, ‘Arguments for Incompatibilism’, Stanford Encyclopedia of Philosophy, <http://plato.stanford.edu/entries/incompatibilism-arguments/> (2011); McKenna, ‘Compatibilism’, Stanford Encyclopedia of Philosophy, <http://plato.stanford.edu/entries/compatibilism/> (2009); T. O'Connor, ‘Free Will’, Stanford Encyclopedia of Philosophy, <http://plato.stanford.edu/entries/freewill/> (2010); Kane, R. H., ‘Introduction: The Contours of Contemporary Free Will Debates’, The Oxford Handbook of Free Will, 2nd edn., ed. Kane, R. (Oxford, 2011), pp. 3–38 CrossRefGoogle Scholar; G. Strawson, ‘Free Will’, Routledge Encyclopedia of Philosophy, <http://www.rep.routledge.com/article/V014SECT4> (2011); Timpe, K., Free Will: Sourcehood and its Alternatives (London, 2013)Google Scholar; R. Clarke and J. Capes, ‘Incompatibilist (Nondeterministic) Theories of Free Will’, Stanford Encyclopedia of Philosophy, <http://plato.stanford.edu/entries/incompatibilism-theories/> (2014)., and, in particular, Levy and McKenna, ‘Recent Work’.

⁵⁵ This is not to say, of course, that there can be no challenges to criminal justice through undermining moral responsibility. In fact there are strong arguments in the recent literature on moral luck to that effect. Zimmerman, Immorality; Levy, N., Hard Luck (Oxford, 2011)CrossRefGoogle Scholar. These are not, however, relevant to a neuroscience challenge.

⁵⁶ Baumeister, R. F., Bratslavsky, E., Muraven, M. and Tice, D. M., ‘Ego Depletion: Is the Active Self a Limited Resource?’, Journal of Personality and Social Psychology 74 (1998), pp. 1252 CrossRefGoogle Scholar–65; Hagger, M. S., Wood, C., Stiff, C. and Chatzisarantis, N. L., ‘Ego Depletion And The Strength Model Of Self-Control: A Meta-Analysis’, Psychol Bulletin 136 (2010), pp. 495–525 CrossRefGoogle ScholarPubMed.

⁵⁷ Evans, J. S. B. T., ‘In Two Minds: Dual-Process Accounts of Reasoning’, Trends in Cognitive Sciences 7 (2003), pp. 454 CrossRefGoogle Scholar–9; Haidt, J., ‘The Emotional Dog and its Rational Tail: A Social Intuitionist Approach to Moral Judgment’, Psychological Review 108 (2001), pp. 814 CrossRefGoogle ScholarPubMed–34; Schnall, S., Haidt, J., Clore, G. L. and Jordan, A. H., ‘Disgust as Embodied Moral Judgment’, Personality and Social Psychology Bulletin 34 (2008), pp. 1096–1109 CrossRefGoogle ScholarPubMed.

⁵⁸ One should generally be wary, I believe, of claims about where the burden of proof lies. There is more than a grain of truth in Peter van Inwagen's observation that: ‘[w]hen a philosopher says, “The burden of the proof lies on you”, he means, “You must deduce your conclusion from the truths of immediate sensory experience by means of an argument that is formally valid according to the rules of elementary logic; I on the other hand may employ any dialectical tactic I find expedient”’ ( Van Inwagen, P., An Essay on Free Will (Oxford, 1983), p. 18 Google Scholar). If the reader shares my wariness towards such claims I beg her patience only for a short while. I shall argue for placing the burden where I think it lies, not assume it.

⁵⁹ Note that cases where the relevant authorities fail to apply this standard support rather than undermine the intuition. Whatever else may be wrong with such failures of oversight – corruption, incompetence, deception – a central part of the explanation of why they are wrong would seem to be the fact that they fail to ensure that the procedure or product generates a benefit that justifies imposing the health risks and economic costs of its use.

⁶⁰ Blackstone, W., Commentaries on the Laws of England in Four Volumes (Oxford, 1765)Google Scholar, vol.4, p. 352.

⁶¹ Singer, P., ‘Famine, Affluence, and Morality’, Philosophy & Public Affairs 1 (1972), pp. 229 Google Scholar–43.

⁶² This lack of a principled approach to uncertainty is, it seems to me, a theoretical embarrassment to deontology, which has only recently begun to attract the attention it deserves. Lazar, Cf. S., ‘Anton's Game: Deontological Decision Theory for an Iterated Decision Problem’, Utilitas (2016), pp. 1–22 Google Scholar.

⁶³ Vilhauer, B., ‘Free Will and Reasonable Doubt’, American Philosophical Quarterly 46 (2009), pp. 131 Google Scholar–40; Vilhauer, B., ‘Free Will and the Asymmetrical Justifiability of Holding Morally Responsible’, Philosophical Quarterly 65 (2015), pp. 772 CrossRefGoogle Scholar–89.

⁶⁴ Duus-Otterström, Cf. G., ‘Betting against Compatibilism’, Res Publica 16 (2010), pp. 383 CrossRefGoogle Scholar–96.

⁶⁵ In 2013, the most recent year for which reliable data are available, an estimated 4,239,619 persons suffered imprisonment as a sanction in criminal justice systems worldwide. United Nations Office of Drugs and Crime, ‘Crime and Criminal Justice Statistics’, <https://data.unodc.org/> (2016).

⁶⁶ Note that the fact that a given justification is dubious, or that adhering to it requires that one expend certain efforts, does not itself give one reason to believe a different justification. The trilemma cannot, that is, be an argument for e.g. consequentialist justifications of criminal justice. Denying this is tantamount to assuming that criminal justice must be justified, and hence that if one justification does not work, this is evidence that another justification must be true. As such, taking the horn of the trilemma where one provides a different justification is not so much a way of tackling the neuroscience challenge as it is of avoiding it.

⁶⁷ I have presented drafts of other papers that eventually metamorphosed into this article as well as earlier versions of it at a MANCEPT 2012 workshop on bioethics, a 2013 Oxford Uehiro Moral Philosophy seminar, a Roskilde University Criminal Justice Research seminar, an Aarhus University ELPP seminar, the 2015 meeting of the Nordic Network for Political Theory, and a 2016 Oxford Applied Moral Philosophy Seminar. I am grateful to audiences at these events for many useful comments and suggestions. I owe particular thanks to Martin Marchman Andersen, David Birks, Jon Rostgaard Boiesen, Tom Douglas, Katrien Devolder, Alexandre Erler, Nir Eyal, Rune Klingenberg Hansen, Fredrik Hjorthen, Robert Huseby, Katrine Krause Jensen, Kasper Lippert-Rasmussen, Carsten Fogh Nielsen, Rebecca Roache, Raffaele Rodogno, Johan Roduit, Jonathan Pugh, Jesper Ryberg, Fatima Sabir, Julian Savulescu, Jørn Sønderholm, Jens Damgaard Thaysen, and above all to Roger Crisp, Søren Flinch Midtgaard, Sune Lægaard and Thomas Søbirk Petersen for extensive and extremely helpful feedback. Finally the article owes a heavy debt in spirit to the work of Michael Zimmermann, whose clarity, rigour and originality of thinking on the topics of moral responsibility and criminal justice inspired much of its writing, even as it fails to meet the exemplary standard he sets.