Character, Punishment and the Liberal Order

Character, Punishment, and the Liberal Order One of the chief features of the liberal political order (at least, in principle) is that it is restrained with regard to legal moralism. A liberal order does not seek to impose or enforce a particular conception of a well-led life or what persons should regard as good or worthwhile. Granted, there are several interpretations of liberalism but it is fair to say that, despite the diversity of liberal polities, each would regard enforcement of morality in a comprehensive manner as illiberal. However, sustaining a liberal order depends upon broad agreement on several fundamental moral matters so that there is a stable, widely endorsed framework of values and principles within which people can exercise extensive liberties. What is often overlooked in the discussion of the liberal order is that matters of moral psychology are just as important as matters of normative principle in the respect that, unless persons have the sorts of dispositions, attitudes, and sensibility to appreciate the liberal rule of law, it has little chance of succeeding and being sustained. This does not mean that the state has the responsibility of encouraging certain virtues and being directly concerned with persons’ characters. Yet, it is important to recognize the relations of mutual reinforcement (or weakening) between civil society and the liberal order. Unless civil society encourages certain sorts of dispositions, concerns, and abilities people will not be effectively able to sustain a liberal order or will lack an interest in doing so. The liberal rule of law makes possible the scope for a heterogeneous, dynamic, open civil society, and that sort of civil society can motivate persons to appreciate the liberal rule of law. There are important matters of moral psychology at the intersection of civil society and the liberal order despite the fact that an individual’s character is not a direct concern of the state in a liberal polity. Some fundamental matters of normative principle are not realizable without members of the society having certain attitudes and dispositions. For example, unless it is highly homogeneous ethnically, religiously, and in other ways a liberal polity depends upon citizens’ willingness to be restrained in the extent to which they attempt to impose their values on other citizens. Generally, a measure of forbearance, mutual respect, and toleration of disagreement on many matters are necessary on the part of participants in a liberal polity. Considerations of moral psychology have a significant role in explaining the success or failure of a liberal polity. In the remainder of the discussion I focus on some of the implications of this fact for criminal justice, and for criminal sanction in particular. The basic contours of the issues are as follows: (a) In the context of the institutions and practices of criminal justice in a liberal polity it is important to distinguish between causing suffering and worsening. Legal punishment can be justified on grounds of desert (supposing the offender is not exempt from the accountability attributable to a voluntary, rational agent). However, if criminal sanction regularly and predictably  worsens offenders insofar as it damages their agential capacities, rendering them less capable of reentering civil society, that damage would reflect failure to respect their moral standing as participants in a liberal order. (b) Granted, the varying sensibilities of individual offenders make this a complicated matter and there is no clear, straight line separating suffering from being damaged. But a liberal polity must face the task of trying to distinguish between the suffering justified by criminal sanction (which can still respect the person punished) and punishing in inhumane ways such that it is reasonable to expect those punished to be more or less disabled for participation in civil society. (c) The overall upshot is that there are respects in which matters having to do with character are important to a liberal polity even though persons’ characters are not properly a direct concern of a liberal state. The remainder of the discussion develops and defends these claims, with a view to highlighting ways in which considerations of character are explanatorily relevant to criminal justice in a liberal polity despite the stubborn difficulties and perplexities regarding how they should figure. I By a ‘liberal polity’ I mean a political order in which extensive individual rights and liberties are preserved and protected, and in which there is a significant role for elected representatives in government. There is no single canonical form of a liberal polity and, indeed, there are multiple reasonable conceptions of the liberal order. These range from minimal-state libertarian conceptions to egalitarian-oriented orders with a substantial role for state institutions, as in Rawls’s conception. In any case, there are several other plausible conceptions between those ‘bookends.’ They all share a concern for the standing of individuals and a concern to permit and preserve significant scope for voluntariness or self-determination or autonomy. This is not the place to explore the many different ways in which those are understood. It should be sufficient for present purposes just to indicate that they are members of a family of notions of special importance to a liberal political order. In addition to the multiple reasonable conceptions of a liberal order there are diverse grounds for endorsing such an order. It is not as though there is some single, rationally privileged ground for the liberal order or some unique method of arguing for it. Some individuals and groups may value the liberal order because they are committed to certain ends and interests and it enables them to lead the sorts of lives they wish to lead. Others may value it because they do not believe there are objective considerations of human good, and in the absence of such considerations, the state should not have the power and authority to impose or enforce a conception of how people should live. Still others may endorse liberalism because they do believe that there are objective considerations of human good but they believe it would be wrong for the state to have coercive power regarding them. And so forth. There is more to this than simply pointing out that in fact people offer diverse reasons for endorsing liberalism. That could be true even if many or even all of those reasons were inadequate, confused or of dubious relevance. However, one of the merits of the liberal polity is that it can accommodate a measure of disagreement over its own character and it can accommodate pluralism with regard to its justifying grounds. There are of course limits in regard to both matters but it is highly implausible—and perhaps illiberal—to insist that there is a single doctrine of liberal order in regard to its form or its justification. Thoughtful persons with diverse axiological and metaethical views could agree that there are good reasons in favor of liberalism while disagreeing over what those reasons are. That is not to say that it is a futile or fatuous enterprise to try to elaborate a principled case for liberalism. Liberalism is a political form in which the weight of good reasons has a genuine opportunity to actually matter, and much can be gained by encouraging a habit of offering good reasons for fundamental political commitments. The openness of a liberal polity can acculturate people to seek and to expect good reasons, and that can be one of its strengths. Nonetheless, it is probably unrealistically optimistic to expect a stable consensus on the grounds for liberalism and its specific form. At the same time, that does not imply that a liberal order is attainable and sustainable no matter what value-commitments and dispositions people have. The liberal order depends upon people having certain valuative commitments along with dispositions and attitudes reflecting them. Participants in a liberal political order need to recognize individual liberty as having fundamental value, they need to acknowledge that others each have like liberties, and they need to believe that the rule of law has among its primary concerns the protection of persons against being harmed and wronged (in ways that require careful specification). In addition to agreement on numerous fundamental matters people need to have effective dispositions to act in ways that realize those widely shared values. Even if a liberal polity is restrained with respect to imposing or enforcing morality it requires a widely endorsed framework of moral values and broadly based effective commitment to them. A liberal order cannot be sustained by accident.31 As Audi observes in his discussion of civic virtue; “an extensive agreement on moral practice is compatible with absence of agreement or even sharp disagreement in moral theory.”72 Motives for the willingness to respect a moral framework sufficient to uphold the liberal rule of law need not derive from some specific, ultimate commitment to political principles common across persons but, the willingness is vitally important. The relation between the liberal order and civil society is centrally important to the argument I will sketch out. By ‘civil society’ I mean all those contexts, associations, and activities in which persons and groups engage voluntarily, on the basis of their choices, interests, and desires. Participation in the various contexts and activities of civil society is not directed by the coercive power of the state.3 That participation may include many aspects of economic activity, diverse forms of religious life, various modes of education, cultural activities and leisure; all those departments of life that are not directly or comprehensively subject to imperatives originating from the state and its institutions. (That a young person should attend school or at least be educated to a certain level might be a requirement of the state but in a liberal political order there are options for educating young people and, to that extent, education is an activity of civil society.) When civil society is extensive, that is, when people’s lives are shaped in large part by voluntary association, engagement, and decision, and when the texture of the social world is largely spontaneous rather than commanded, individuals exercise voluntary, purposeful agency. This is so even if the spontaneity is thickly informed by tradition, orderliness, and considerable predictability. Civil society is not possible without liberty and the converse is true, as well. Civil society is a mode of social and economic life in which a diversity of values is pursued and there is openness to changing patterns of interaction, shifting interests, and patterns of association. There is extensive scope for individual and group purposefulness but there is no overall ‘plan’ requiring persons to have only specific roles and prospects. There are important relations of mutual reinforecment between the liberal polity and civil society. Participation in civil society under a liberal rule of law both requires and reinforces deliberative abilities, rational self-determination, and various modes of interaction and cooperation with other individuals and groups. The liberal order provides and preserves the legal framework for civil society to flourish, while participation in civil society can supply people with reasons to want to maintain the liberal order. The abilities to successfully lead a life in civil society need to be developed, encouraged, and trained by experience. The liberal order makes that possible. The experience of freedom of action made possible by liberty enables people to think of themselves and others as accountable, self-determining agents in more rather than fewer ways. Participation in civil society multiplies the occasions for persons to encounter each other as agents, as accountable, voluntary actors, acting for reasons.84 There is no assurance that this will create mutual respect and civility but there is greater potential for persons to regard and treat each other as rational agents when there are many opportunities and occasions for persons to be accountable agents. William Galston writes: A liberal polity guided (as I believe it should be) by a commitment to moral and political pluralism will be parsimonious in specifying binding public principles and cautious about employing such principles to intervene in the internal affairs of civil associations. It will, rather, pursue a policy of maximum feasible accommodation, limited only by the core requirements  of individual security and civic unity.5 Civil society is the complex context in which people encounter others with different values, concerns, and attitudes, and in which acknowledgment and toleration of such differences are required. Participation in a dynamic, open civil society can encourage the development of humane concern and regard for others. While that is not assured, the prospects for a liberal order are dim if civil society does not supply regular opportunities for orderly disagreement, rule-governed voluntary activity of many types, and contexts in which demands upon mutual understanding, collaboration, planning, and forbearance are met. That is an important respect in which a liberal order depends upon civic culture even if, at the same time, the basic order of the polity is what makes that civic culture possible. Civil society can encourage a common sense that preservation of the scope for different values and aims is vitally important. In that way civil society not only requires but can also attain certain shared values and presuppositions regarding the social world. II With this background we are in a position to consider the significance of considerations of character with regard to some important issues of criminal justice. I shall sketch out the main contours of a view of why character is both explanatorily and normatively significant with respect to criminal sanction in a liberal polity. Character is not a direct concern of the state, yet criminal sanction’s impact upon character can create serious moral and social difficulties for a liberal polity. Institutionalized practices that are known to more or less disable people as self- determining agents, rendering them much less capable of reintegrating into civil society motivate important more questions about the use of state power. My thesis is that long-term incarceration—at least in the U.S.—is morally suspect on those grounds. The damage done to prisoners’ characters exceeds what justified criminal sanction can permit. By character I mean relatively stable dispositions and patterns of sensibility, practical reasoning, choice, valuing, and reacting shaped by features of natural temperament and voluntary activity. The disposition and patterns figure in explanations of persons’ actions and they can also have important places in agents’ self-conceptions, aspirations, and regard for themselves. Elements of character are grounded in features of natural temperament but are also shaped by agents’ choices and actions—by voluntariness. Thus, one might have a volatile temper by nature but strive to control and moderate it; both of those attributes are elements of character. An agent might aspire to be more courageous, fairer, and more resolute but struggle against tendencies to be timid, selfish, and imprudent. The aspiration and the effort to realize it are elements of character, as are the natural tendencies against which the person struggles. Some features of mature character are voluntary in a significant respect despite the role of natural tendencies and despite the fact that certain states of character are often not acquired intentionally.6 The actions that shape dispositions partly constitutive of character are often voluntary even if the agent does not deliberately aim at being the sort of person for whom it is second nature to act from those dispositions. For example, someone’s lies are typically voluntary and the person who lies often may become a liar, a dishonest person, though he or she never intended to become a liar in that character-grounded sense. Someone who repeatedly makes excuses for failure to keep commitments or to follow through on assurances given may become both self-deceived and unreliable in ways that are entrenched and enduring. That was never part of what was intended. Yet, voluntariness had a role in bringing that about. Thus, while coming to have specific states of character is not achieved directly by decision or volition, it is achieved in part by patterns of decision and patterns of voluntary action. This is independent of whether the individual takes responsibility for the actions that shape states of character. Human beings have capacities for acting for reasons, for acknowledging (for better or worse) a multiplicity of distinct types of practical necessity, what we might call distinct types of ‘have to’ to which one might respond. This is a significant aspect of human voluntariness. Character is both shaped by, and reflected in the ways in which persons recognize and respond to what they take to be different types of ‘have to.’ There is the ‘have to’ of desire, the ‘have to’ of legal requirement, the ‘have to’ of religious commitment, the ‘have to’ of moral requirement, the ‘have to’ of etiquette, and so on. Even though ‘I have to see that new film; I go to every film starring so- and-so’ expresses a reason based upon desire, the agent may regard that reason as the basis for something he or she ‘has to’ do at least in the respect that failure or being prevented would not be a matter of indifference. We might think that this is a pretty weak basis for practical necessity but different people weight things in quite different ways. Moreover, what one regards as what one ‘has to’ do, what one thinks is practically necessary, can also reflect a person’s aspiration to have a certain type of character. Someone might think that he has to strive to be more cooperative, less selfish, and more honest. Thinking this may be motivated by more than just a very pragmatic prudence. This person might have come to a realization that some of his states of character are objectionable and that he can and should make an effort to acquire different ones. That is one way it can matter to us what we take to be practically necessary, what are the sorts of ‘have to,’ to which we respond and why. That is also a respect in some people do indeed take a measure of responsibility for what they are like. Granted it is not clear that, with respect to character, we can judge with confidence the extent to which someone is, so to speak, self-made. The answer to the question of how each person came to be like that, for whatever specific dispositions, propensities, and susceptibilities the person has, is likely to be enormously complex. A further complication is that the significance of a type of conduct can be different for the same person at different times. As a young man one may regard the ‘have to’ of satisfying the requirements of gang membership as an important imperative, and later on regard it as spurious and a serious error. Other sorts of ‘have to’ may retain their genuineness for the agent, as in the virtuous agent’s commitment to honesty. How one regards what one ‘has to’ do is not necessarily fixed by objective considerations, even if there are objective action-guiding considerations. Suppose there are objective moral reasons. Even if a soundly virtuous agent would regard them as decisive, as ‘silencing’ other considerations (as McDowell put it) whether those considerations are effectively practically necessary (‘silencing’) depends in part on the character of the agent.7 The incontinent agent and the vicious agent do not see that they have to act on those reasons. How action-guiding considerations, whether objective or not, figure in an agent’s deliberations and actions can depend substantially upon the agent’s character. McDowell writes, “the dictates of virtue, if properly appreciated, are not weighed with other reasons at all, not even on a scale which always tips on their side.”8 [26] And, speaking of the relation between the temperate person’s proneness to physical pleasure and his perception of what are reasons for acting in certain ways he says …his clear perception of the requirement [of temperance] insulates the prospective enjoyment—of which, for a satisfying conception of the virtue, we should want him to have a vivid appreciation—from engaging his inclinations at all. Here and now, it does not count for him as any reason for acting in that way.9 [27] Contrasting virtue and continence, McDowell writes “The lack of struggle [on the part of the virtuous agent] is ensured by keeping the attention firmly fixed on what Aristotle calls “the noble”; not by a weighing of attractions which leads to the conclusion that on balance the virtuous course is more desirable.10 [27] Considerations of character often have an important role in interpreting and explaining people’s behavior, whether the accounts refer to virtue, vice, continence or dispositions that are mixed. Consider the kinds of questions and answers that give us insight into people’s decisions and actions. For instance, “why did he suddenly decide that he should cease his ‘off the books’ business activity? He’s been getting away with it for years.” Or, “what led her to throw herself into philanthropic work with so much energy and focus; that’s not like her?” Or, “what has motivated her to be so much more considerate and willing to listen than she had been?” Explanatory answers to these questions are likely to refer to facts about agents’ characters even if specific circumstances or events have a prominent role in the explanations. The answers will often include reference to how the agents see themselves and their situations, and how they regard their own states of character, whether or not that regard seems accurate. Perhaps a particular situation triggered the person’s change in attitude or a reconsideration of how to act or to react. Still, what the situation counts for, how the facts are regarded by that person at that time, depends upon how it resonates with features of character and with their self- conception. Reactions and decisions are often mediated, as it were, by how the agents see their own characters in the respect that a person’s self-regard and self-evaluation can make a difference to their decisions and attitudes. Sometimes it is important to people that they should be able to see themselves in certain ways— even if that involves falsehood or dishonesty; still, their self-respect may be bound up with seeing themselves as having a certain kind of character. How one thinks about one’s character, and also one’s lack of thought about it, can make an important difference to one’s self-conception. This is especially clear in agents with genuine aspirations to acquire and strengthen virtues or, perhaps, to free themselves of vices. We often evaluate ourselves in terms of what sorts of durable and prevailing dispositions we have, and self-knowledge can depend substantially on honest awareness of one’s character. Many people are complacent about their vices or even take perverse pride in them. Someone might tell us, “listen, I’m a liar. OK; just get used to it; at least I’m not dishonest about that” or, “Look, that’s just the sort of person I am. Don’t expect me to do you any favors or to show any gratitude; I’m in this for myself and myself alone”. Our understanding of our characters, and what in fact our characters actually are, runs through all of this fabric of self-regard and reflection, for better or worse. It is quite plain that how we regard others, and the ways in which we are inclined to react and respond to them is often substantially shaped by what we believe about their characters. Given what we know of them and given our notions of their characters, we are sometimes surprised by people. In many cases this is not a reason to think their characters have changed but rather, that we have come to know more about them. Character, Aristotle argued, tends to become fixed, becomes second nature in the mature individual. Aristotle argued that because of the role of voluntariness in human action a person is responsible for even those actions reflecting fixed, unchangeable character.11 Voluntariness—if not explicit intention—figured centrally in acquiring those states of character, even though natural temperament, too, has a role. Moreover, in his view, there is a respect in which fixity of character is an ideal, as in the soundly virtuous person; the one for whom non-virtuous considerations are silenced. At the opposite end, fixity in vice is possible, and the vicious agent may have no realistic prospect for revising his or her character. Even if this view overstates the fixity of character we can see that what it says about virtue expresses a plausible ideal and what it says about vice expresses a plausible conception of ethical incorrigibility. Moreover, the Aristotelian view includes the notion that while we know that what we do voluntarily can shape dispositions that then figure in character, we do not know with certainty the degree to which this or that action influences character. The person may insist, “ I am only going to do this sort of thing a couple of times; it’s not like I’m becoming the sort of person who always does it.” But we do not have fine-grained control over the way in which our actions and reactions influence states of character. That is what lies behind the— often very apt—advice, “you don’t want to make a habit of that,’ especially when it appears that the person might indeed want to do that, whatever it is, habitually. The natural temperament from which we start is not up to us but there is typically a measure of plasticity in it, and that is why decisions and actions can make a difference to what we are like even though we do not have fully executive control over the difference made by this or that occasion of acting. Nonetheless, one’s voluntary actions, reactions, and the ability to consider them and to assess them underwrite responsibility for character whether or not it is taken. Also, change in character, even when it is brought about with the agent’s effort, is not brought about directly by volition or decision. We can make considered judgments of what to do and why, and that is what enables us to have some control over what dispositions become constitutive of mature character despite the fact that we do not have accurate knowledge of just how much our actions influence our dispositions. We are sufficiently aware of the voluntariness involved to know that what we do makes a difference to what we are like. III What has this discussion of character to do with criminal sanction in the liberal polity? The main elements of the issue I wish to formulate are these: (a) In a liberal polity an individual’s character is not a direct concern of the state. It is not a proper project of the state to arrange the social and political world in such a way as to support or enforce any particular conception of individual perfection. (b) Despite that, character, as a matter of moral psychology, is relevant to the liberal polity. For one thing, in a liberal political order people have extensive freedom to lead lives of the kinds they wish to lead. They are not required to have certain virtues and the state does not have authority to determine their careers, interests, and guiding concerns. There is extensive political freedom to try to develop the sort of character one wishes to have. (There are important limiting factors and impediments. But that is always the case and they are not the result of institutionalization by the liberal state.) The form of civil society under a liberal rule of law permits wider rather than narrower scope for persons to lead the sorts of lives and become the sorts of people they wish. Participation in the numerous contexts of civil society can encourage the development of realized capacities for self-determination, cooperation, association, and interaction of many kinds as well as enabling persons to develop prudence and an appreciation of accountability. That is one of main ways in which open, dynamic, heterogeneous civil society can lead to recognition of the value of a liberal order and an interest in preserving it. There are respects in which the liberal polity depends upon  persons’ characters at the same time that it is restrained with respect to directly shaping them. (c) In general, criminal justice is concerned with conduct rather than character though considerations concerning character sometimes have a role in sentencing (in the U.S. and some other countries). Having a direct concern with character would put the state in a position to exercise coercive power in strongly illiberal ways, policing people’s souls, as it were, and perhaps also requiring them to change in specific ways. The law regards persons as voluntary agents, capable of reasoning about how to act and capable of acting on the basis of reasons. Still, the law in a liberal polity cannot require a person to be virtuous and cannot require a person to have and act upon specific motives. (d) Even when punishment is regarded as having a correctional aim whether a prisoner reconnects with correct values through an effort of ethical self- correction or simply decides that crime is not worth it and resolves to be law-abiding (or makes no attempt at correction, ethically, prudentially or otherwise) that is up to that prisoner, it is not an imperative of the state. (e) While the imposition of suffering via by legal punishment is justifiable, the damaging of prisoners, in the sense of disabling their agential capacities, is not justifiable. If, in regular and predictable ways, punishment encourages dispositions and sensibility antithetical to civil society or impedes them in to a significant degree, that is not justifiable. Unfortunately, punishment does just that, on a large scale, at least in the U.S. and almost certainly in other countries, as well. That motivates important questions about whether current penal practice is morally justified, and there are reasons to conclude that it is not. Before developing these points a bit more fully I want to comment on one important matter concerning the relation between liberalism and retributivism. I believe this merits comment because, in my own view, retributivist considerations should have a central place in the justification of punishment in a liberal polity.12 George Fletcher suggests that there are inherent tensions between liberalism and retribution. He says that the liberal is skeptical about retribution as a rationale for criminal punishment. If the state cannot posit fundamental values for the society as [sic] whole, then, by like token, the state cannot arrogate to itself judgments about ultimate justice for criminals. It is not the job of the liberal state to do God’s work on earth. Retribution—the pursuit of “ultimate justice”—is not, therefore, a proper aim of a liberal criminal law.13 (“Political Theory and Criminal Law,” Criminal Justice Ethics, Volume 25, No. 1, Winter/Spring 2006, p. 23) However, I do not think that retribution need be interpreted as “ultimate justice.” Rather, it is an important element in the justification of criminal sanction in a liberal polity because of how retributivism registers the significance of desert and the standing of persons as voluntary, responsible agents, capable of understanding the values and principles reflected in criminal law. I see retributivism not as an attempt at ultimate justice but as recognizing standing in an important sense. As Antony Duff writes: …we can best understand the authority and claims of criminal law in a liberal polity by understanding it as a law for citizens: a law to which citizens subject each other and themselves, under which they call public wrongdoers to account for what they have done, and are themselves ready to answer to their fellows.14 Duff’s view involves more thickly communicative and republican elements than the view I am defending. For him, a criminal wrong is to be understood as a “breach of the values which underpin our life together, and of the social and moral bonds which make that life possible…”.15[Dagger 16] Moreover, Duff maintains that punishment has, as a proper aim, to motivate, through suffering, an effort at ethical self-correction and reconnection with sound values. However, there is an important point of contact between his view and my own in the way they regard the retributive element of punishment as responsiveness to the standing of persons in a liberal polity. One of the challenges for a liberal polity is that, in certain respects, the success of a state as a liberal polity depends upon members having dispositions and attitudes it would be wrong for the state to seek to encourage or strengthen in a direct, coercive manner. That is true even in regard to offenders found guilty of criminal conduct. Criminal sanction should not be applied to make people virtuous, even if the hope is that it will motivate those punished to not only conform to the law but also acknowledge the values informing it. But the framework of values of a liberal polity is needed to preserve extensive freedom and to protect persons from harm, not to enforce specific aretaic ideals or specific conceptions of a well-led life. It prohibits certain types of conduct antithetical to a liberal order, and that is a much more modest aim than seeking to cultivate citizens with specific conceptions of good and of virtue. The law is concerned primarily with conduct, with violations (and in some cases, attempts) not with virtue and vice. A liberal polity faces the difficult task of finding the balance between focusing on legal liability rather than a deeper, fuller notion of moral blameworthiness, and yet, employing a notion of desert that is adequate for making the appropriate discriminations regarding legal responsibility. It is very difficult to see how that can be done without inclusion of features of a person’s character. As Murphy notes, “Thus a preference for a mens rea system over a strict liability system is, I think, not only consistent with liberalism but required by it.”16(Ibid., p. 101) There are numerous difficulties with the notion of mens rea, and it might be more suitable to the analytical clean-room conditions of conceptual analysis than the complicated, world of actual agents, acting in particular circumstances, with motives and intentions it is difficult to ascertain with exactness and certainty. Nevertheless, features of character are often relevant and illuminating in regard to judging blameworthiness in ways responsive to the law’s attempt to recognize different degrees and types of voluntary action. The law cannot require that someone acquire a certain kind of character but sometimes knowledge of character is crucial to legal judgments. IV Though a liberal polity can punish, and thereby inflict suffering, there is a conflict between a liberal order and punishment that is known to disable or destroy persons’ agential capacities, and that is the present focus. Punishment addresses offenders as rational agents, and does not just express the society’s disgust, anger, and power.17 There is no guarantee that the imposition of criminal sanction will be received accordingly, offenders recognizing the justification of being punished. Nonetheless, punishment that damages persons regularly risks its moral justification. What could be the point of punishment that causes those punished to be less able to be self-determining, independent agents, less able to function in civil society—to be worsened in that way? It is quite clear that the conditions of incarceration often have a role in disabling prisoners for constructive, prudent, accountable participation in civil society. A significant complicating factor is that, in many cases, those persons were already disabled in the relevant sense prior to being incarcerated. That raises questions about whether offenders have had fair opportunity to comply with the law and whether their circumstances constitute excusing conditions to some extent. As Reiman notes, “If the obligation to obey the law is the fair return for benefits from others’ cooperative efforts, it follows that the duty to obey the laws is conditioned on the justice of the society that those laws govern.”18 (“The Moral Ambivalence of Crime in an Unjust Society,” p. 7) “How great that reduction is will depend on how badly victimized they are, on how great or small a chance there is to remedy this victimization legally, and so on—a set of factors too complicated to put into a simple formula.”19 (p. 12) Even if we do not explain punishment as a response to a breach of rules of fair play or of taking unfair advantage, we can see that significant unfairness in the distribution of benefits and burdens can at least suggest doubts about whether some offenders have diminished responsibility. The question of whether some offenders’ abilities to conform to the law is diminished by the unfairness they have suffered is not our primary focus here. Apart from the question about opportunity to comply with the law we need to confront the question whether current penal practice unjustly disables many prisoners for participation in civil society. Punishment may be justified but the conditions of punishment and their effect on prisoners’ characters might raise doubts about their justification. The points made earlier concerning relations between the liberal political order and civil society and how they can be mutually reinforcing are relevant here. The scope for civil society preserved by a liberal rule of law makes it possible for people to shape a dynamic, complex, textured social world by purposeful, rule-governed activities and practices, which can support a shared sense that such a social world is worth preserving. That is how civil society can support the liberal order. Sustaining healthy civil society requires certain shared values and presuppositions regarding the social world. Coercive removal from civil society by imprisonment does not have to be a de-civilizing disaster, but it often is. It certainly seems to be for a great many incarcerated offenders in the United States. It is caused by isolation, routine violence, and the lack of meaningful activities, experiences, and interactions. I am not arguing that imprisonment should be a resource-rich environment of education, skill training, counseling, enjoyment of the arts, and overall bildungsroman. It is punishment, which, when deserved, causes suffering justifiably. The sting of punishment is important to its communication of denunciation. Still, a distinction between justified punishment and damaging or worsening persons is needed in order to be properly responsive to the respect owed to persons, even when criminal sanction is merited. Some persons may be worsened by conditions that are not inhumane. That is unfortunate but does not call punishment into question. Given individual sensibility, even decent conditions of imprisonment may be a disaster for some persons. Conditions might lead to depression and loss of self-respect, while for someone else they amount to a moral ‘wake-up call,’ motivating that person to reconsider the values and motives that led to criminality. A society’s laws, practices, policies, and institutions cannot reasonably be expected to be fully responsive to the particularities of individuals. That punishment is not calibrated with great precision for each person sentenced is not, as such, a legitimate ground for a claim that the law is unfair though, clearly, particular applications can be unfair. Some offenders show strong signs of being incorrigible. Some appear to be genuinely and deeply committed to wrong or perverse values. Others may have dispositions so entrenched that, even if they are capable of genuine regret, they may not be able to change. Their bad habits may be very firmly entrenched even though they attain some perspective on their badness.1020 Still, there are both epistemic and moral reasons against simply ‘writing off’ a person as morally hopeless despite the fact that, in some cases, confidence in such a judgment may seem well placed. The epistemic reason is that we do not know for certain if we are correct. People can surprise us—in the best and worst ways. The moral reason is that we should guard against regarding others as morally so far beneath us that we might cease to regard them as having any standing whatsoever in the moral world. That reason applies even when confronted with a horrible, unrepentant moral monster. What is monstrous is that this is a person, not some sort of non-person. The moral caution may slow what might otherwise be a rush to judgment. Also, in slowing us down in that way it helps us remember that we cannot be sure how much of our own decency and virtue is to our credit and how much is a matter of luck.1121 It is morally instructive to be reminded of how terrible human beings can become, and how fragile our own decency or goodness might be. I am not suggesting that we should expect that there is a ‘better self’ in each person, and if only we are patient and attentive, we can help draw out that better self. Maybe some people’s vices are so entrenched that there is precious little capacity for any better self to emerge. Maybe they can acknowledge their wrongs but are incapable of revising their dispositions. However, even if severe punishment is deserved it is not the case that further worsening the person is acceptable, and the conditions of many prisoners’ lives behind bars could hardly be expected to have a result other than worsening them. This suggests that in the highly complicated intersection of ethical, empirical, and explanatory considerations concerning punishment in a liberal polity there is a role for considerations of character. The focus on character is a helpful way to identify and explicate a serious issue regarding the moral justifiability of practices of criminal sanction. Far too frequently practices of criminal sanction disable persons for returning to civil society. They are worsened by punishment in more than the incidental senses having to do with lost job opportunities, damage to family relationships and friendships, incurring debt, and so forth. Understandably, much of the public discussion of criminal law and sentencing reflects public opinion regarding what is needed to make society safer, to rid communities of criminal elements, and to ‘put away’ repeat offenders and even many first-time offenders ‘for a long time,’ in prison, ‘where they belong.’ However, there is much less attention to the effects of imprisonment upon prisoners. Of course, imprisonment is punishment; I am not suggesting that it should be a pleasant diversion from life outside of prison. Nonetheless, it would be prudent for society to take an interest in whether, as a result of imprisonment, offenders are rendered much less able to reintegrate into civil society upon being released into it. It is not in anyone’s interest that persons being released should be more disposed to commit crimes or should be so diminished in their capacities as rational agents that they have slim chances of successfully reintegrating. The difficulty is not that the state should be in the business of reforming offenders, and yet it does a very bad job of it. Even if rehabilitation or perfectionist ends are not among prison’s primary aims it is not a matter of indifference how punishment harms those punished. It is important to try to distinguish between causing suffering and worsening or damaging. Plato argued that we worsen a criminal by not punishing, because the sting of punishment can motivate the criminal to reconsider his values and his reasons for action. The reform-of-the-soul-for-the-criminal’s- own-good conception of punishment exceeds the aims proper to a liberal polity though, to be sure, a repentant response would be welcome. If serious damage to character is among the predictable effects of the state’s policies and practices, that is reason to regard those policies and practices as unjustified. This is not a quasi-utilitarian argument that we should seek from criminal sanction the socially most consequences we are able to obtain. It is the recognition that even retributively justified punishment, which does not aim primarily at improvement of the punished, treats the criminal in a morally dubious manner and also defeats its own purposes in an aggravated way by worsening criminals. Incarceration may have incapacitated the offender for the period of imprisonment but it often results in agents even less able to lead prudent, responsible, law abiding lives. And that is avoidable; it is not as though all incarceration inevitably or unavoidably damages and diminishes people. It could be objected that what I said earlier about the voluntariness of character pulls in the opposite direction from the argument just given, and that if prisoners come to have worse characters as a result of incarceration that is largely attributable to them as agents more or less responsible for their characters. It is fair to raise that point but the argument was not that individuals are responsible for their characters no matter what or that character is voluntary in all circumstances. It is true that people respond differently to the conditions of imprisonment; that was the point of the remarks made above regarding the diversity of sensibility. Still, there is little doubt that, in many cases, the conditions of incarceration are such that it would not be reasonable to expect persons to be improved by them or to manage not to be worsened by them, especially in light of the fact that so many sentences are very long. Often, they are conditions in which those who are either already disposed to, or willing to engage in violence, intimidation, blackmail, theft, the degradation of others, and other strongly antisocial, uncivil practices are best able to protect themselves and their interests. In many cases, there is little in prison life supplying incentive to adopt norms and values more consistent with civil society. This view does not involve an equivocal use of the notion of liberalism, employing it in a more libertarian sense some of the time and a more republican sense at other times. As I have said, it is not part of this view that punishment should aim at repentant rehabilitation of the criminal and restoration of him as a public-spirited citizen eager to do his civic duty. The main point is to highlight the relevance of character in a context in which it is easily overlooked. No equivocation concerning liberalism is involved in defending a view in which the state is more rather than less restrained in regard to legal moralism and in which retributive considerations are central to the justification of punishment on the one hand, and holding that certain kinds of damage done to prisoners’ characters is morally unjustified on the other. There is dispute over whether disenfranchisement and other legal impediments faced by former prisoners are justified; there is much less discussion of whether diminution of one’s capacities as an agent is justified. While there is hardly a straightforward measure of how much a person’s agential capacities have been diminished it should not be inscrutable whether conditions of imprisonment have an influence on dispositions, attitudes, and propensities, i.e., on character. We should distinguish the political issue of whether or not liberalism can be entirely neutral with regard to different conceptions of the good from the question of the extent to which character-based moral blameworthiness is properly a concern of criminal law and criminal justice. Even in a state striving for neutrality agreement on several values is necessary as a framework for law. Moreover, character-based considerations can be crucial to determinations of blameworthiness and a variety of important legal distinctions. And we should distinguish those issue from the question of whether it is acceptable that current forms of legal punishment (a) encourage dispositions and sensibility antithetical to civil society, and (b) impede or inhibit the dispositions and sensibility needed for constructive participation in civil society. Even if punishment is justified primarily on retributive grounds (which is not to say that it has no non-retributive aims or that it is necessarily harsh or severe) it is wrong for it to damage or destroy offenders’ agential capacities by the corruption of character. End Notes 1. ? 2. Robert Audi, 3. My discussion of civil society is influenced by Edward Shils. He wrote: 4. ? 5. William Galston, Liberal Pluralism, New York: Cambridge University Press, 2002, p. 20. 6. I develop this view in some detail in Choosing Character: Responsibility for Virtue and Vice, Ithaca: Cornell University Press, 2001, esp. Chs. 1 and 2. The view owes a great deal to Aristotle, esp. Bks. II and III of Nicomachean Ethics. 7. McDowell, “Are Moral Requirements Hypothetical Imperatives?” 8. Ibid., p. 9. Ibid., p. 10. Ibid., p. 11. Aristotle, in Bk. III of Nicomachean Ethics, argues that the fixity of character in an adult is not a reason to regard the agent’s actions as less voluntary than those of someone whose character is not yet established as ‘second nature.’ Indeed, an action is, in some respects, more fully one’s own for reflecting fixed character. 12. I argue for a central role for retributivist considerations in the justification of punishment in “Luck and Retribution,” Philosophy, 1999, pp. 13. George Fletcher, “Political Theory and Criminal Law,” Criminal Justice Ethics, Volume 25, No. 1, Winter/Spring 2006, p. 23. 14. Duff 15. Duff 16. Jeffrie G. Murphy 17. Antony Duff has argued extensively for punishment having a significant communicative dimension. This is part of what is required for it to be justified in a community of rational agents capable of understanding (and endorsing) the values reflected in the law. Otherwise, law is just a coercive imposition. 18. Jeffrey Reiman, “The Moral Ambivalence of Crime in an Unjust Society,” p. 7. In “A Paternalistic Theory of Punishment,” Herbert Morris argued that the following two conditions must be met for punishment to be justified: The first is that the norms addressed to persons are generally just and  and that the society is to some substantial extent one in which those who are liable to punishment have roughly equal opportunities to conform to those just norms. The second condition is equally import- ant. The theory presupposes that there is a general commitment among persons to whom the norms apply to the values underlying them. (‘A Paternalistic Theory of Punishment,’ in Punishment and Rehabilitation, ed. Jeffrie G. Murphy, p. 165.) 19. Reiman., p. 12. 20. ? 21.