Responsibility without Consciousness

Abstract

This paper addresses the relationship between responsibility and consciousness, in the light of both H.L.A. Hart’s and subsequent philosophical analysis. First, is consciousness necessary to responsibility-attribution? If so, how demanding a requirement is this? And does it make sense to pose these questions in the abstract? Second, when we move from the realm of moral argumentation to that of law, are there additional factors – institutional, functional, practical or otherwise – which alter the weight or implications of the argument? Third, conversely, what can moral philosophers learn from the way in which the issue of consciousness is dealt with in legal mechanisms of responsibility-attribution? The paper concludes that, in both law and morality, the conditions of responsibility are standards which are constructed and deployed within particular social practices and institutions for certain purposes. Hence the dominant issues underlying questions ostensibly about consciousness are evaluative questions about appropriate standards in particular contexts, in the light of the social functions and meaning of criminalisation; of a finding of responsibility in civil law; or of moral responsibility-attribution.

[A] hundred times a day persons are blamed outside the law courts for not being more careful, for being inattentive and not stopping to think; in particular cases, their mental or physical examination may show that they could not have done what they omitted to do. In such cases, they are not responsible; but if anyone is ever responsible for anything , there is no general reason why men should not be responsible for such omissions to think, or to consider the situation and its dangers before acting. ¹

I would like to begin by thanking John Gardner, his colleagues in the Oxford Law Faculty and the Tanner Trustees for the invitation to deliver this lecture in memory of HLA Hart. It is an inestimable privilege. Throughout my career—before I met him, as well as during the time when I had the good fortune to know him and even since his death—Herbert Hart has been a benevolent spirit in my life. As an indifferent, if not downright dissident, law student, it was his writing which first fired my imagination and bolstered my motivation; through the generous trust of his widow and his children I had the precious opportunity to try my hand at the genre of biography, and in doing so, to discover further depths to his integrity, intellect and imagination; and even as his work has become an object of critical reassessment, with my own work moving further towards the social sciences, it has remained a rich fund of inspiration, as well as a reminder of the importance of the effort to express one’s thinking clearly and concisely. In an increasingly competitive world, Hart’s life also stands as a reminder that it is possible to combine scholarly ambition, brilliance and seriousness with humility, a gentle manner, and personal and intellectual generosity to others. Indeed, one of the many joys of writing his biography was to receive letters from young scholars inspired in their turn by his life, and in particular by the way in which his intellectual courage allowed him to overcome the self-doubt which is often a characteristic of a truly creative mind. As I looked at the intimidatingly distinguished list of previous Hart lecturers, it was striking how wide-ranging has been his influence; as Joseph Raz remarked in his own Hart lecture six years ago, it is a mark of Hart’s quality, as well as of his open-minded intellectual curiosity, that his students, in the broadest sense of that term, have taken his ideas in so many different directions.

In this lecture, I follow the lead of at least four previous Hart lecturers ² in considering the factors which shape agency, personhood and responsibility in both legal and moral realms. In doing so, I return to the book which first inspired my interest in criminal law and in legal theory: Hart’s Punishment and Responsibility . ³ At the core of Hart’s argument in this book is a thought which goes back at least to Aristotle: the thought that our responsibility, in the sense of our answerability for our actions, is premised on two conditions: our cognitive capacity to grasp the relevant considerations bearing on our action and our volitional capacity to direct our own behaviour. This idea that human responsibility is founded in volitional and cognitive capacities runs through much of western philosophy up to and beyond Hart, but its apparent simplicity masks a huge fund of complications which have made ample work for philosophers, psychologists and legal theorists. In particular, the interminable debate about whether we have free will and, if so, in what sense, or, if not, whether it matters to our very notion of a responsible subject or to our practices of responsibility attribution has loomed so large in the history of moral philosophy that, as George Sher recently noted, ⁴ the cognitive condition for responsibility has remained comparatively neglected. Indeed, this series opened 20 years ago with Richard Wollheim’s psychoanalytic interpretation of the volitional conditions of responsibility, ⁵ which argued that the ‘pale criminality’ presented by unconscious impulses beyond our control should be put at the centre of any analysis of excuses. In Wollheim’s view, if many offenders act in the lure of a phantasy that they can purge pre-existing guilt by seeking punishment, guilt precedes and explains, rather than succeeding and being explained by criminal acts; and the law, in its assumption of voluntariness, misrepresents us to ourselves. The radical—and paradoxical—upshot of Wollheim’s argument was the troubling thought that, to the extent that crime is motivated by unconscious desires which render the criminal law’s very proscriptions a motivation for offending, the social institutions of criminalisation and punishment are counter-productive.

By contrast to the relative lack of attention to the cognitive conditions of responsibility in moral philosophy, in both criminal and civil law, the question of what we must know in order to be held accountable has been a very central one. To take just a few criminal law examples, the McNaghten test of insanity ⁶ depends on an inability to know the nature and quality of one’s act or, if one does know it, to know that it was wrong; and relevant mistakes of fact standardly operate to invalidate forms of mens rea such as intention or belief, or to invoke defences such as self-defence or duress. But even to allude to these very basic examples is already to see that the commonsensical proposition that we shouldn’t be held responsible for actions in relation to which we were ignorant of key features is much less straightforward than it at first appears. What are the relevant ‘features’ which we need to know in order to be responsible, and are they purely matters of fact or are they sometimes, as in the second limb of the McNaghten test, matters of law or other normative systems including morality? If the latter are included, why aren’t mistakes of law deemed as excusatory? ⁷ What if our ignorance could have been avoided by a reasonable effort on our part, or on the part of a person with standard capacities? And just how demanding is any requirement of knowledge, awareness or consciousness: do facts of which we have, as Antony Duff—another previous Hart lecturer ⁸ —has put it, latent knowledge count as known facts for the purposes of any knowledge or consciousness requirement? ⁹

We can get a better sense of the full array of these complications about just what counts as the relevant form of knowledge in relation to responsibility—and about just how we might determine whether it exists—by thinking about some examples. English literature abounds with them. I confess that, in planning this lecture, I rather hoped that Tom Stoppard’s new play, ¹⁰ which engages with the so-called ‘hard problem’ about the nature of consciousness, might provide some useful examples, given that Herbert Hart was a great admirer of Stoppard’s capacity to bring philosophical questions to life on the stage. Unfortunately from that point of view, Stoppard’s problem is different from—and very much harder than—the questions which I shall tackle in this lecture. But even bracketing that deeper ‘hard problem’, and taking human consciousness, and its significance, as given, just what form of knowledge or awareness is needed for an attribution of responsibility is in itself a problem worth contemplating. ¹¹ I shall begin this task by drawing some literary illustrations from a number of Hart’s other favourite authors.

The persistence of this puzzle about the relationship between responsibility and consciousness across the centuries is nicely illustrated by plentiful instances in Shakespeare’s plays. In Hamlet (1600–01), for example, Shakespeare teases his audience with conflicting evidence about just how much cognitive capacity his protagonist has, but also throws up more quotidian examples: when Hamlet strikes at what he thinks is a rat, or a dangerous intruder, behind the arras, killing Polonius, Shakespeare invites us to think about how far Hamlet’s ignorance should shape our interpretation of his fatal act. But the way in which writers pose—and resolve—the question changes significantly over time. Two hundred years after Shakespeare, Jane Austen presents us with the case of Emma Woodhouse (1815), who, though endowed with greater than average powers of observation and reasoning, with ample opportunities for exercising them and with the benefit of some sensible advice, persists in believing that Mr Elton is falling in love with her protégé Harriet Smith—in the teeth of plentiful evidence that he is far too ambitious to marry Harriet, and is in fact courting Emma herself. Austen’s view of this classic case of self-deception is clear: Emma’s mistake flows from a character flaw for which she must take responsibility. ¹² Another century on, by contrast, Henry James ( The Ambassadors 1903) seems to invite a more ambivalent reading of Lambert Strether’s persistence in ignoring the evidence, which is constantly before his eyes, that Chad Newsome, whom he has sent to Paris to rescue from European depravity for a life of service to American capitalism and provincial mores, is in fact not only in love with, but the lover of, the delicious and inscrutable Mme de Vionnet. Writing on the cusp of a modernist sensitivity to the ambiguity and plurality of the self—a plurality which might be seen as disrupting attributions of responsibility—James seems to invite us not to condemn Strether’s self-deception, but to accept his concurrent state of knowing, and not knowing, ¹³ as simply a fact of the human condition.

But perhaps the richest array of literary treatments of the relationship between responsibility and consciousness is thrown up by the Victorian era. This is hardly surprising, for this was a time in which a range of scientific developments—from Darwin’s theory of evolution and Spencer’s social Darwinism to the experiments of mind doctors, phrenologists and mesmerists—seemed to threaten not only the moral and theological doctrines on which understandings of human agency had been based, but even the notion of a stable, unified subject, present to itself, on which moral, legal and literary understandings of responsibility had rested in the modern period. Some of these literary cases are relatively straightforward: Franklin Blake, the hero of Wilkie Collins’s The Moonstone (1868), is clearly not responsible for a ‘theft’ committed while in a state of laudanum-induced sleepwalking; though Lydia Gwilt, the anti-heroine of his Armadale (1866), is most certainly held up as responsible notwithstanding her own liberal habit of self-medication. ¹⁴ In a different and yet more radical instance, Robert Louis Stevenson’s Strange Case of Dr Jekyll and Mr Hyde (1886) presents us with the question of how far responsibility stretches across a divided consciousness: is Dr Jekyll responsible for the murderous acts of his alter ego, Mr Hyde? ¹⁵ Nor do the uncertainties about the relationship between knowledge or awareness and responsibility end with cases of automatism and double consciousness: issues of memory, like what we might call fugue states, also preoccupied late 19th century novelists. For example, George Eliot’s Gwendolen Harleth of Daniel Deronda (1876), condemned to a wretched marriage by her own wilful misjudgement and desirous of her tyrranical husband’s death, is genuinely uncertain whether she has contributed to his death in a yachting accident because of an intervening period of unconsciousness and subsequent amnesia. Conversely, in her short story The Lifted Veil (1859), Eliot invites us to ponder just how much awareness is compatible with normal conditions of agency: the protagonist, afflicted with powers of clairvoyance and mind-reading, descends into a form of introspective neuroticism which seems more or less incompatible with agency. ¹⁶

In recent years, three prominent philosophers have made important interventions on the very particular question which is my concern here: whether knowledge and awareness are necessary conditions for attributions of responsibility. In his 2009 monograph Who Knew? Responsibility without Awareness , George Sher ¹⁷ mounted a spirited and meticulous assault on the ‘consciousness thesis’: in other words, the conventional wisdom that responsibility must be premised on knowledge or awareness. In Sher’s view, the consciousness thesis confuses conditions relevant to our first person and third person perspectives. ¹⁸ While, from a first person point of view, I can’t be expected to respond to things I am not aware of, the detached, third person perspective from which we make judgments is retrospective and properly takes into account a wider range of factors. In place of the consciousness thesis, Sher argues that there are two different paths to justified ascriptions of moral responsibility: one proceeding via consciousness, the other proceeding via ‘origination’. In other words, where an agent has evidence that an act is wrong, and fails to respond to that evidence, she is responsible wherever her behaviour both falls below an acceptable standard and is agent-specific in that it stands in a causal relationship with her constitutive psychology: her settled attitudes, character traits and so on. In making this argument, Sher provides a picture of responsibility which fits well with widely shared intuitions about agents’ responsibility for instances of inattention, forgetfulness or indifference which lie just beyond the boundaries of consciousness; and. Sher also dismisses as unrealistic—I think with good reason—the thought that these cases can be always brought into conformity with the consciousness thesis by locating some previous time at which the agent knowingly opted for action which presented a risk of later departure from the requisite standard of conduct. ¹⁹

At about the same time, Joseph Raz was working on two papers which form the final part of his 2011 monograph From Normativity to Responsibility , ²⁰ one of which originated as the 2010 Hart lecture. ²¹ In that lecture, Raz tackled one important part of the terrain encompassed by the question of how responsibility relates to consciousness, setting himself the task of vindicating—in accordance with Hart’s view, quoted at the head of this lecture—the idea that we are genuinely responsible for negligence proceeding from misjudgements attendant on forgetfulness, slips of the mind and inattention. Like Sher, Raz defends a dual, alternative track to responsibility, albeit one unified by the fact that, on each side of the track, the acts for which we are responsible relate to our way of being in the world as rational agents: both our conscious reasoning processes and our effects on the world bear some meaningful relation to us—indeed, to who we are. According to Raz, we are responsible where we either consciously act in a particular way, in the exercise of our rational agency, or act without awareness of some relevant consideration but within what he calls our domain of secure competence—the sphere in which we generally expect ourselves to be acting as rational agents who are responsive to reasons—and fail to reach the standard which rational agency would have secured. In making this argument, Raz produces an account of responsibility which resonates with the notion of outcome responsibility famously defended by another of Hart’s closest associates, Tony Honoré ²² —though Raz narrows the scope of outcome responsibility by juxtaposing the insight that the outcomes we cause in some sense become part of our selves, ²³ with an insistence that the conditions of responsibility include either rational deliberation or some form of failure of judgment or attention within what the agent generally regards as the domain of her secure competence for rational judgment and action. (Even read as an account of a certain kind of social practice and understanding, notwithstanding its focus on rational agency, note that Raz’s account might, counter-intuitively, be read as consistent with Wollheim’s: agents might, of course, be deluded about the efficacy of their rational functioning, misled about the existence of, or about what falls within, a domain of secure competence—hence opening up the possibility that our settled social practices of responsibility attribution may operate on the basis of ontological assumptions radically at odds with the facts of the human situation.)

In making these arguments, Sher and Raz were reaching, albeit by somewhat different routes, a similar position to that meticulously staked out in 1990 by Antony Duff. ²⁴ Criticising the dualist approach typical of contemporary legal philosophy, which has increasingly tended to separate act from accompanying mental state or mens rea , Duff argued that criminal responsibility—which he understands as a species of moral responsibility—is rather an interpretation of the attitudes manifested in the defendant’s actions. While acting knowingly or intentionally expresses the strongest form of identification between an agent and her actions, actions which disclose an attitude of practical indifference—for example, where an agent’s latent knowledge might readily have been called upon to avoid a mistake—also justify a finding of responsibility, and indeed shape the very meaning of the agent’s conduct. This is not, in Duff’s view, to accept that responsibility is imposed ‘objectively’, irrespective of the agent’s own capacities; rather, practical indifference describes the agent’s own (culpable) attitude. Conversely, John Gardner’s 1998 positon on the ‘gist’ of excuses, which argues that to excuse an agent’s conduct is to make an assessment that she has conformed to established standards of character, would exclude many instances of inattention or otherwise unconscious failures of judgment from the boundaries of excuse. ²⁵

Most recently, in his 2014 monograph Consciousness and Moral Responsibility , ²⁶ Neil Levy has mounted a robust defence of the thesis that consciousness of the morally significant factors bearing on action is indeed a precondition for moral responsibility. Of particular interest is the way in which Levy constructs his case. Rather than relying, as does Sher, on the consistency of his thesis with prevailing intuitions, or elaborating, as does Raz, from an abstract account of rational agency, Levy grounds his argument firmly in empirical findings emerging from cognitive psychology and neuroscience. He accepts that the (normative) philosophical questions about responsibility cannot be settled by the empirics, but insists—ironically, rather in the style of Wollheim—that some effort should be made to test philosophical ideas of responsibility against what we know about human psychology. His argument accordingly builds a picture of agency informed by experimental and neuroscientific evidence; one in which consciousness plays a special role in integrating the causal factors and impulses being produced and processed by our complex brains—impulses and processes which he argues to be very much more diverse, and less logically ordered, than those presented in Freud’s theory of the unconscious. ²⁷ Levy therefore takes the position that consciousness is not simply something relevant to our first person perspective. Rather, consciousness has a distinctive role: in allowing for general domain integration between many functionally separate modules within the brain, ²⁸ it delivers the highest possible degree of a subject’s responsiveness to reasons and hence counts as a condition of agency. While some degree of responsiveness is consistent with incomplete awareness—one telling example is that of the improvising jazz musician ²⁹ —only consciousness enables the sort of deliberation and responsiveness which is truly compatible with responsibility.

In the rest of this lecture, I draw on these important contributions to examine the following questions:

First: is consciousness indeed necessary to responsibility attribution and, if so, how demanding is this requirement—what level of awareness or knowledge is needed? And does it make sense to pose this question in the abstract? Here, I will argue that, once we take into account the place of evaluation in all three accounts, Levy’s account can in fact be seen to be closer to Sher’s and Raz’s positions than he is inclined to accept. In particular, I will suggest that Levy’s application of standards of reason or rationality, to determine what the agent is either taken to have known or justly held accountable for having failed to appreciate or attend to, in fact displaces any purely empirical question of consciousness, moving from fact to evaluation. Taking the philosophical debate on its own terms, this militates in favour of wider approaches, such as those of Duff, Sher and Raz. Yet it also calls into question whether the issue of responsibility’s relationship to consciousness can be resolved purely in terms of philosophical intuition or in the abstract: if we must draw the relevant lines through evaluative judgments, the contexts in which, and purposes for which, those judgments are made become important.

Secondly, therefore, I will ask whether, when we move from the realm of moral argumentation to that of the law, there are additional factors—institutional, functional, practical or otherwise—which alter the weight or implications of the arguments. In raising this question, I am attending closely to Hart’s—to my mind unanswerable—insistence that attributions of responsibility in legal contexts should not be assumed to follow exactly the same logic as attributions of responsibility in moral discourse and reasoning. ³⁰ Moreover, we need to consider whether the conditions of responsibility vary across different domains of law or whether—as Raz might be taken to imply—the conditions of responsibility are invariant, with merely the rules pertaining to liability varying in terms of whether responsibility is a condition precedent. How far, in short, are these debates in moral philosophy relevant to the law?

Thirdly, and conversely, can moral philosophers learn from the way in which the issue of consciousness is dealt with in legal mechanisms of responsibility attribution? Contrary to Sher’s view that the legal debates are of limited relevance to the moral debates because of the technical and functional features which make law a special case, ³¹ and notwithstanding the law’s special practical and institutional features, I shall argue that on one familiar and plausible view of morality some very similar issues arise, implying that lawyers’ easier view of the socially constructed nature of responsibility may turn out to be an advantage rather than a limitation. So, while not venturing to go quite as far as endorsing Arthur Goodhart’s amusing claim in the 1950s that ‘analytic philosophy is old hat to lawyers’ (!), ³² I will suggest—echoing another previous Hart lecturer, Tony Honoré, who argued in 1993 for ‘the dependence of morality on law’ ³³ —that the learning here should not be all in one direction. While law and morality solve the relationship between responsibility and knowledge or awareness in different ways, because of their different methods, institutionalisation and orientation as normative systems, there is reason to think that moral philosophers may have a great deal to learn from the law. Moreover, legal philosophers’ increasing tendency over the last 40 years to look to moral philosophy for their inspiration in theorising responsibility has had some unfortunate consequences which Hart himself would have regretted has, as we shall see in the final section of this paper, some consequences which Hart himself. The conditions of responsibility, I conclude, in both law and morality, are standards which are constructed and deployed within particular social practices and institutions for certain purposes. Hence, evaluative questions about appropriate standards in particular contexts, in the light of the social functions and meaning of criminalisation, of a finding of responsibility in civil law or of moral responsibility attribution, are the dominant issues underlying questions ostensibly about consciousness. In short, what really matters here is producing the most plausible interpretations of responsibility attribution, understood as a complex set of social practices ranging across a variety of settings in which moral concerns or questions of criminal or civil law are at stake.

To clarify: in what follows I will be talking only about agents who satisfy basic conditions of cognitive and volitional capacities such that they are candidates for being held responsible for their actions. ³⁴ Hence, I will not discuss the particular case of radical mental incapacity, which arguably exempts subjects from the realm of responsibility rather than providing them with excuses which preclude or mitigate that judgment of responsibility on any particular occasion. ³⁵ In addition, I leave aside the question of how the arguments developed in this lecture would apply to the attribution of responsibility to collectivities, such as corporations or states, or to non-human subjects, such as animals. These questions are of undoubted importance, and I hope—again, following the lead of previous Hart lecturers—to pursue them further in future work. ³⁶

Moral Responsibility without Consciousness?

As a lawyer rather than a philosopher, I am not qualified definitively to adjudicate between the respective merits of Sher’s, Raz’s and Levy’s accounts of the relationship between responsibility and consciousness. In this part of my lecture, rather, I want to question whether Levy’s argument—or, at least, its upshot for ascriptions of responsibility—is ultimately as different from Raz’s and Sher’s positions as might appear. Certainly, all three of these philosophers share some important views: they all assume that a coherent account of the conditions of moral responsibility can be given (though Levy explicitly leaves it open as to whether these conditions are ever in fact satisfied); ³⁷ they all accept that action proceeding from conscious deliberation implies responsibility; and, with varying degrees of explicitness, they all accept a considerable degree of complexity in the causal neurological events that eventuate in action and render problematic the fiction of a wholly unified self. ³⁸ In other words, agency is something which human beings achieve or assemble by subjecting the signals and impulses coming from their brains, and from their environment via their brains, to some form of rationalising process which itself forms part of a complex social process of mutual interpretation. While that reasoning process is central for Levy, just as rational functioning is for Raz, even for Sher there is a strong assumption of some degree of consistency over time and stability in the subject’s constitutive psychology—the self as an ‘enduring causal structure’ ³⁹ —which turns on the relationship between settled character traits, attitudes and actions. All three authors have more to say about reasoning and rationality than about the emotions which we know to shape human cognition and deliberation—though all give accounts which would be capable of giving some place to emotions as revealing reasons for action. While Raz and Sher deny that knowledge is a necessary condition for responsibility, they do not deny that conscious action is one route to a justified finding of responsibility; indeed, I see no reason to think that either Sher or Raz would have much objection to Levy’s picture of consciousness as a ‘global work space’ ⁴⁰ which broadcasts messages from the brain which are ‘online’ and available to the agent to draw upon in rational deliberation, drawing conflicting messages into some sort of order of coherence or priority in deciding how to act. And, finally, all three philosophers proceed on the basis that there is some timeless conceptual truth about responsibility which is capable of being revealed by meticulous philosophical analysis.

The difference between them, of course, lies in Levy’s explicit rejection of Sher’s view that we may be responsible for acts which simply originate in and express our constitutive psychology or character, independent of awareness, and in the inconsistency of Levy’s thesis with Raz’s somewhat different view that we may be responsible for acts which fall within what would normally be our domain of secure competence for rational action but where some failure of knowledge or judgment has disrupted our usual competence. ⁴¹ But how radical is this difference in fact? To get a clear view of this, it is useful to consider a couple of examples. One is the case of practical indifference or wilful blindness: the case of the rape defendant who, socialised in an extreme instance of a culture in which women are taken to be generally sexually available though likely to deny it (a world, sadly, not as different from our own as we might wish), has had sex with a woman who did not consent, but in whose consent he genuinely and positively believed; or as to whose consent he simply failed to advert. The second is a recent case of journalists charged with misconduct in a public office after making payments to prison and immigration detention centre officials in return for information about life inside the prison/detention centre, under circumstances in which they were unaware that the offence existed, or that prison and detention centre staff would count as public officials. ⁴²

How would Levy deal with these sorts of cases? For him, the key question would be whether there is sufficient consciousness of the morally relevant factors—those which bear on the act’s moral significance—to justify the view that agency is engaged and hence to underwrite an ascription of responsibility. But where, and how, does Levy draw the line between knowledge and unconsciousness in these borderline cases? His approach is to insist that what is needed is greater than dispositional potential to be aware, albeit falling short of full-scale ‘occurrent’ awareness: the bottom line, he suggests, is ‘access consciousness’ ⁴³ —in other words, we are conscious of things which are not merely ‘online’, but readily personally available to us, even if we have not brought them to the surface of our minds. ⁴⁴ On this basis, it may seem obvious that the journalists failed to meet the relevant standard. But even here, in judging whether the ‘access consciousness’ condition is satisfied, it is difficult to imagine that we would not be influenced by considerations such as what it would be reasonable to expect a journalist to inform herself of in relation to legal norms bearing on her professional activities or role, or by judgments about whether the existing social and professional cues should have put the journalists on notice that they were doing something wrong.

In the case of the rape defendant, the idea that we can draw a simple empirical line between sufficient consciousness and its lack is equally problematic. Levy might be inclined to say that, if prevailing and widely diffused conventions depart from the defendant’s own conventions, this is a case of non-responsible ignorance: yet, once again, one wonders how we would draw the line other than in terms of some sense of what agents can reasonably be expected to know about prevailing social norms. What of the practical indifference example? Here we have a form of attitude or outlook, but one which falls short of conscious knowledge or awareness. Is this sufficient? What seems obvious, once again, is that the answer to this question cannot be supplied by empirical evidence.

Notwithstanding metaphors such as broadcasting, or phrases such as the requirement that relevant factors be ‘occurrent’ or ‘personally available for report’, ⁴⁵ at least as soon as we accept, as Levy does, that responsibility reaches beyond actions in relation to the morally significant features of which we have occurrent awareness, where we draw the line between what is taken to be within and without agents’ knowledge cannot be simply a question of psychological fact. ⁴⁶ Rather, it is strongly shaped by prevailing conventions and assumptions about what an agent in their position ought to have been, or could reasonably have been expected to be, aware of. This is perhaps particularly clear in the case where the morally relevant factor is a norm rather than a factual circumstance, though in principle the two cases raise the same issue. ⁴⁷ In short, Levy can’t get what he wants here, because there simply isn’t any empirically grounded, evaluatively neutral way of judging whether an agent is ‘sufficiently’ conscious of the morally relevant features of her situation to be held responsible for what she has done. The question of whether, short of actual awareness, the relevant information was ‘readily available for report’ cannot be a purely empirical question. And if it is right that the sufficiency of consciousness is inevitably in part defined in terms of evaluation or interpretation, it follows that the difference between Levy’s position on the one hand and Raz’s, Sher’s or Duff’s on the other is less radical than it first appears. For Levy’s position is no less evaluative than Sher’s reliance on a finding that the agent failed to meet an ‘acceptable’ standard or Raz’s insistence that the second track of responsibility attribution, that relating to non-aware conduct within the domain of secure competence, should be confined to ‘failures’ of rational judgment. Note, finally, that this has a broader upshot: for the evaluative presuppositions which inform where we draw the boundaries of responsibility attribution are founded not merely in philosophical intuition, but are shaped by historically situated and practical concerns about when it makes sense to hold someone responsible. And this, I will argue, is where moral philosophers have something to learn from law. ⁴⁸

1. From Morality to Law?

Let me now turn to my second question: that of how, if at all, these debates about moral responsibility bear on the criminal or civil law. At first sight, it seems clear that both civil and criminal law support the plausibility of Sher’s and Raz’s cases for the extension of not just liability to compensation or punishment but, more substantially, of genuine attributions of responsibility to cases of negligence, inattention and misjudgement which do not feature anything approaching Levy’s sense of conscious awareness. Even leaving aside the case of strict liability, the negligence standard in a wide range of criminal offences and the application of reasonableness requirements within a range of defences, notably defences involving mistakes, testify to legal responsibility attribution reaching beyond consciousness. The civil law goes yet further in this direction: Lord Atkin’s famous ‘neighbour principle’ establishing a duty of care to those foreseeably affected by one’s actions in the law of negligence ⁴⁹ provides a striking example of the core intuition at work here. Moreover, there is strong reason to think that the rationale guiding these extensions of responsibility attribution beyond consciousness lies—as Honoré noted 27 years ago ⁵⁰ —in a deep sense that, under certain conditions, agents who cause particular harmful outcomes unconsciously nonetheless stand in some significant relation to them in terms of how they stand in the world, their identity, who they are as agents in the role which they occupied at the relevant time and hence for what we expect them to regret. ⁵¹ The due diligence defence which is applied to many cases of strict (though not absolute) liability would, for example, support the idea that these are cases of genuine agent responsibility rather than mere cases of liability. And, pace Sher, the ‘partially subjective’ interpretation of negligence liability akin to the one advanced by Hart—asking not simply what a reasonable person would have known, but what it would be reasonable for a person in this particular role and situation ⁵² to have known—seems to offer an appealing vision of how negligence liability fits within the criminal law. Note that it is one which does not depend on what Sher rightly sees as a tenuous claim that responsibility is based on the agent’s knowledge at an earlier point in time (a time which is impossible to identify in a significant set of cases); ⁵³ and that it fits well with Duff’s account of practical indifference in identifying the action as genuinely proceeding from agency.

Moreover, the dual track to responsibility fits well with the clear coexistence of different patterns and, indeed, principles of responsibility attribution in criminal law. As I have argued elsewhere, English criminal law in practice discloses at least three different patterns of responsibility attribution, with the balance between them shifting over the course of modern legal history. ⁵⁴ Today, the dominant view as reflected in most appellate decisions and leading textbooks is that criminal responsibility—what most clearly justifies the holding of an agent to account in criminal law—requires proof that a defendant’s core volitional and cognitive capacities were engaged at the time of the allegedly criminal conduct. In one view, what is at issue here is the thought that criminal conduct in this sense is genuinely chosen, and this has manifested as a strong preference in some quarters for the instantiation of ‘subjective’ conditions of criminal responsibility such as intention, knowledge, foresight or belief. ⁵⁵ But it is incontrovertible that criminal law in fact quite often treads a different path to a finding of responsibility—one which focuses rather on the quality of character which the relevant conduct discloses, either in the sense of global traits of character or in terms of the attitude to the norms of criminal law specifically which are expressed by the conduct in question. ⁵⁶ This approach to criminal responsibility dominated, I have argued, in the 18th century, then declined in the late 19th and early 20th centuries, but is having a revival in the early 21 ^st century. ⁵⁷ Finally, criminal liability is sometimes based on responsibility simply for causing particular outcomes or posing particular risks—with the application of due diligence defences to many of these instances suggesting, pace Raz, that in legal terms they are understood as cases of genuine responsibility attribution as opposed to mere liability. Responsibility in legal contexts is a product of the law, and not a given to which the law must respond or which it must reflect. Hence the salience of particular puzzles about the conditions of responsibility will alter in tandem with particular legal constructions of responsibility, which in turn relate to the context in which those constructions have been formed. This is a matter to which I will return in addressing my final question.

One way of reading this variegated terrain of responsibility—attribution in criminal law would be to regard it as strongly confirming, in the legal context, the applicability of Duff’s, Raz’s or Sher’s notion of a dual or multiple track to responsibility, each with somewhat different relationships to consciousness or awareness. Note, moreover, that we can reconcile this approach with the account of criminal responsibility most elegantly articulated by, and strongly associated with, Herbert Hart: the thought that responsibility consists not so much in choice as in the finding that an agent had a fair opportunity to conform their conduct to the demands of criminal law. For we might take this ‘fair opportunity’ account to indicate not so much a dual track alongside responsibility with consciousness, but rather the existence of a more general category (in the style of Raz’s notion of justified attributions of responsibility being premised on conditions which express our way of being in the world) in which conscious choice is just one way of meeting the fair opportunity condition, and hence of satisfying the basic condition of responsibility: that our conduct has in some adequate sense expressed ourselves as agents. Note, too, that the agency arguments of Sher and Raz also provide insights into where the lines here should be drawn so as to avoid excesses of imputation on the basis of status or imputed ‘dangerousness’ which come close to Levy’s cases of predictiveness but not agency. ⁵⁸ So while Levy is right to suggest that there is a strongly felt distinction here, it cannot be the case that where we draw the line between conscious and non-conscious action is a self-executing empirical matter. Evidently, and as Herbert Hart always insisted, two important questions arise in relation to the law: first, what practices of responsibility attribution does the law disclose? And secondly, can these practices be justified in terms of the moral or otherwise evaluative precepts which we believe ought to be brought to bear on the law’s practices? I would, moreover, add a further question in which Hart was, perhaps regrettably, less interested: that of the institutional and other conditions of possibility for the law’s meeting of the precepts of our moral or political ideals. ⁵⁹ Each of these questions is in my view key to our understanding of responsibility in legal contexts.

There is, of course, a great deal more to be said about why the law might be in need of a dual or multiple track theory of responsibility. I shall set out just three ways in which the legal context for responsibility attribution is special, and is distinctive, moreover, in ways which might make a difference to how we would want to conceptualise responsibility. These are, first, the practical role of responsibility attribution in the law; second, the institutional context for legal judgments; and third, the broader regulatory and expressive roles of the law in particular social contexts. In what follows, I will confine my attention to criminal law, for reasons of both my own expertise and of time. But I would like to note at the outset that what I have to say about responsibility attribution in criminal law itself implies that the conditions of responsibility might be expected to vary—both between criminal and civil law, and between different areas of each.

To take the first question: responsibility plays several distinctive practical roles within criminal law. It operates as a key condition in legitimating criminal law as a decisive instance of the state’s power to punish . In a society which is highly individualistic and which is attached, at least in its public self-conception, to broadly liberal values, this legitimation role for the conditions of responsibility is particularly urgent—and particularly problematic. The conditions of responsibility also have a role in coordinating the operation of criminal law: identifying what forms of knowledge and judgment must be brought to bear in the court room in particular and in legal argumentation in general. And the conditions of criminal responsibility must serve the distinctive ends of the criminal law in coordinating social behaviour. ⁶⁰ Note that this very basic insight into the role of the conditions of responsibility in criminal law immediately implies that the conditions of criminal responsibility are inevitably shaped by what is regarded as the broader social role of criminal law itself—whether regulatory or expressive, or both. This implies, in turn, that there is a close link between the conditions of responsibility and the very standards and coordinating norms which criminal law seeks to establish—as, indeed, is demonstrated by the inescapably evaluative aspect of responsibility attribution to which I have already drawn attention. This is a point which I will pursue in more detail below.

Secondly, any attempt to analyse the conditions of responsibility in criminal law must take into account the fact that criminal law as a social practice takes a particular institutional form . As compared with other forms of regulation, modern law operates through general rules and addresses its subjects as self-directing and rational. Yet attributions of criminal responsibility have to be delivered by particular institutional processes, in accordance with particular protocols of proof, executed by a range of legal and, increasingly, professional actors; and the institutional infrastructure necessary to the realisation of this modern form of legalism has been assembled over several centuries, and takes different forms in different legal systems. Concretely, the shape of the relevant institutions and actors—the trial, the legal profession, policing and prosecution—have changed over time, driven by both shifting views of the role of criminal law and, more importantly, changing configurations of interests in the broader social and political economy. The key point here is that the institutional features of law are among the most important conditions of existence of particular ideas of responsibility, for protocols of proof and practices of legal judgment have to go forward within particular institutional structures. Hence, for example, the standard 18th century felony trial, the average length of which has been calculated to have been around 20 minutes, ⁶¹ was simply not equipped with many of the features—legal representation, law reporting, a system of appeals for testing points of law—which would allow responsibility understood as a conscious mental state to be an object of proof. Rather, it operated on the basis of assumptions about the qualities of character exhibited by the defendant’s conduct, judged in the light of his or her reputation in a still relatively settled community. In this system, in other words, local knowledge informed and underpinned the legitimacy of the criminal process in interpreting what any given defendant could be taken to know.

With the coming of a more anonymous, urbanised, individualised world, criminal law from the late 18th to the early 20th century gradually assembled the complex institutional, ideational and professional infrastructure which allows different conceptions of criminal responsibility based on assessment not only of character or outcome but also of engaged capacity to be invoked, and subject to test in the trial process (a development which, however, inevitably lent to the length and complexity of trials—arguably causing the invention of pragmatic mechanisms, notably plea bargaining, which are inclined to invoke predictive character- or status-based short cuts to capacity responsibility). Moreover, with the (ironic) silencing of the defendant as witness in the emerging 19th century trial process, just as a psychological conceptions of mens rea were emerging, there was a need to develop protocols of proof which squared this institutional circle given the lack of any direct access to evidence of aware mental states. Hence the presumption that a person intended the natural consequences of his or her actions remained central to legal doctrine until the early 20th century. ⁶²

Thus, and thirdly, as Sher notes, the regulatory functions and ambitions of law make a difference to what criteria of responsibility are appropriate . ⁶³ As I have argued in relation to character responsibility, ⁶⁴ one way of looking at this is that the least the criminal law can do is to respond to us as we genuinely are; but, conversely, its regulatory role is such that we can ask no more of law than that it respond to us as we are and in relation to actions which genuinely express our character, in the restricted sense of our attitudes, as agents, towards the relevant criminal law standards, as manifested in our conduct. ⁶⁵ And sometimes criminal law requires us to ‘take responsibility’ for the unforeseen but harmful consequences of particular, often profit-making or otherwise advantageous, roles which we have taken on. ⁶⁶ Hence, far from Sher being right that the legal literature fails to capture various distinctions important to the moral debate, rather it raises a whole host of further interesting philosophical and practical questions. And this brings me to the final question of my lecture.

2. From Law to Morality?

It may seem obvious to you—as it has to many philosophers—that law is a very different creature from morality: that its special social functions and institutional features are such that moral philosophers have little to learn from legal philosophers, let alone from lawyers’ debates about the conditions for responsibility attribution. All three philosophers on whose views I have mainly focused, for example, base their arguments about the relationship between responsibility and consciousness on an (unstated) assumption that the conditions for moral responsibility just are —whether that are is an empirical or a metaphysical one—a certain way. For them, the assumption is that there is ‘one right answer’ within the fabric of the universe to the question of how responsibility relates to consciousness. ⁶⁷ Thus, even when conceding, to quote John Gardner, that ‘the law … faces many constraints, but also many choices, in settling how morality is to be institutionalised’, ⁶⁸ the assumption is that some given notion of ‘morality’ provides the fundamental source of the choices available. Herbert Hart was famously sceptical about ‘one right answer’ claims in the context of adjudication; ⁶⁹ and, though I am taking a liberty in drawing such a sweeping analogy, I like to think that he would have been similarly sceptical about the idea of a single right answer about responsibility. For not all moral philosophers—let alone all of those non-philosophers who have thought seriously about morality—think about this issue in such a metaphysical way. Many—indeed, Hart himself was drawn to this view, though he is well known to have been reluctant to take a definitive position on it—would be inclined to see morality itself as a fundamentally social practice of practical coordination, mutual interpretation and meaning-making. And if we view morality in this way—a view, by the way, which does not exclude the existence or importance of critical debate about moral precepts—law turns out not to be such an entirely different creature, and indeed looks like an obvious fund of reflection for our practical understanding of moral concepts. ⁷⁰

How different is law from morality? Of course, the law is concretised in institutional arrangements to a far greater degree than morality; and partly as a result, as Hart observed, its content is susceptible to deliberate change to a qualitatively different degree from morality. ⁷¹ But the similarities are as striking as the differences, with the regulative ideals ⁷² and social role of morality in securing peaceful coexistence and social order finding close analogues in law, as is confirmed in very different fields of research, such as social anthropology and evolutionary psychology. Moreover, such a view of morality can help us to see that, within a very broad understanding of responsibility conceived in terms of how people are interpreted as being related to and answerable for their actions, particular conceptions of responsibility, as well as of the conditions which have to be met for attributions of responsibility, shift in moral as well as legal practice over space and time.

If we allow for the fact that cultural forms like literature reflect the moral and practical concerns of their day, our literary examples should be able to give us some further insights here. In Hamlet , we see the broad questions about responsibility, and about the impact of both madness and mistakes on how we should interpret and evaluate a person’s actions, already in place. But, as between Jane Austen’s confident focus on Emma’s ⁷³ responsibility for mistakes deriving from failings of character, the Victorians’ preoccupation with the mental and psychological conditions of responsibility, and Henry James’s displacement of the question of responsibility, we see significant changes in the understanding of selfhood or agency, of what it means and how much it matters, to hold people to account for their actions. While I can only touch on this question here, it is also significant that issues about responsibility for mistakes and misjudgements take interestingly different forms in 18th and 19th century literature. The early to mid-18th century writers are mainly preoccupied with their characters’ adequate fulfilment of particular social roles grounded within a hierarchical and status-oriented view of the world: responsibility attribution is grounded in assessment of external factors manifest in conduct and comportment. By contrast, the more individualistic world of the 19th century generates an outpouring of literary reflection on the internal, mental conditions of responsibility, as writers ponder the changing conditions of social life in an increasingly democratised setting. As I argued in Women, Crime and Character , ⁷⁴ the increasing importance of the question of responsibility, and the rising influence of psychological understandings of responsibility, in the 19th century derived not merely from the rise of the medical and psychological sciences, but from a declining confidence in the world-views—theological, moral and, indeed, political—which had hitherto underpinned the regulative practices of social morality.

In recent years, criminal law theory has been strongly in the grip of what we might call a form of moral-philosophical imperialism, and has been marked not only by the claim that criminal responsibility is a form of moral responsibility, but also by the assumption that the philosophical analysis of legal responsibility is therefore a second class relation of philosophical analysis of moral responsibility. ⁷⁵ This, I want to argue—and I hope Hart would have agreed—has been unfortunate, diverting attention from the insights to be gained from a careful interpretation of the relationship between principles of responsibility attribution and the expressive and regulative social practices in which they are embedded—a relationship powerfully evoked by Hart’s close colleague Peter Strawson in his classic essay ‘Freedom and Resentment’. ⁷⁶ Note, too, that a commitment to interpreting responsibility as a social construct embedded in different normative orders unravels one of the persisting meta-problems of moral philosophy: its tendency to rely simply on intuitions. As in the case of the debate between Levy and Sher, moral debate has often seemed suspended in an unappealing dilemma between reliance on intuitions on the one hand and reliance on scientific claims incapable of generating the relevant judgments on the other. In a view of morality, like law, as a social practice itself informed by evaluative commitments, this dilemma is avoided, for it becomes clear that the issue is simply and irreducibly one of interpretation of actually existing social practices. Perhaps, then, we should read Sher, Raz and Levy—even if counter to their own intentions—as discussing moral responsibility attribution as a social practice with a particular role or function. Such a reading would make their work a fruitful source for thinking about the analogies, and disanalogies, between responsibility in law and morality.

3. In Conclusion

I hope that I have said enough to convince you not only that the question of how legal and moral conceptions of responsibility relate to consciousness is one which merits our attention, but also that practices of responsibility attribution, whether in law or morality, are shaped by the institutional frameworks, social aspirations and practical goals within which they are embedded. This implies that there is no order of priority or importance as between law and morality; rather, both provide fascinating fields within which a properly context-dependent analysis of practices of responsibility attribution can go forward. I should conclude by admitting that the title of course massively oversimplifies the subject matter of the lecture, for there is no one relationship between responsibility and consciousness, or even between responsibility and knowledge or awareness, but rather a number of relationships relative to the regulative systems within which ideas of responsibility are socially produced, and within which they remain embedded.

I am very grateful for the insight and generosity of Antony Duff, Lindsay Farmer, Victoria McGeer and Hanna Pickard, who commented on a draft of this lecture; of Mike Redmayne and David Soskice, with whom I discussed its argument; and of the audience at the lecture, delivered on 19 May 2015, who posed some excellent questions. Responsibility for any carelessness, misjudgement or ignorance—as well as one or two wilful decisions not to follow advice!—rest with me …