• penance;
  • punishment;
  • virtue;
  • sacrament;
  • communication;
  • Antony Duff;
  • Thomas Aquinas;
  • postsecular


  1. Top of page
  2. Abstract
  3. Liberal and Theological Limits to Communicative Punishment as Penance
  4. The State's Side of the Dialogue: Wrath, Censure, and Invitation
  5. The Offender's Side of the Dialogue 1: Intervention in Character and the Problem of the Abbot
  6. The Offender's Side of the Dialogue 2: Sacrament, Virtue and Sincerity, or The Problem the Priest
  7. References
  8. Biography

This essay suggests that while Antony Duff's model of criminal punishment as secular penance is pregnant with possibilities for theological reception and reflection, it proceeds by way of a number of separations that are brought into question by the penitential traditions of Christianity. The first three of these—between justice and mercy, censure and invitation, and state and victim, constrain the true communicative character of his account of punishment. The second set of oppositions, between sacrament and virtue, interior character and external action, and formal and moral reconciliation, subject the model of state punishment as secular penance to problematic liberal and libertarian constraints. A postsecular analogy, outlining a theology of the invitational nature of divine judgment, and drawing on Thomas Aquinas's account of penance as both sacrament and virtue, is proposed.

Liberal and Theological Limits to Communicative Punishment as Penance

  1. Top of page
  2. Abstract
  3. Liberal and Theological Limits to Communicative Punishment as Penance
  4. The State's Side of the Dialogue: Wrath, Censure, and Invitation
  5. The Offender's Side of the Dialogue 1: Intervention in Character and the Problem of the Abbot
  6. The Offender's Side of the Dialogue 2: Sacrament, Virtue and Sincerity, or The Problem the Priest
  7. References
  8. Biography

A number of contemporary philosophical justifications of criminal punishment, especially in the analytic tradition, make use of a sacramental analogy—namely criminal punishment is to be justified as a form of secular penance. The most prominent proponent of such a view is Antony Duff, upon whose work this essay will focus. Duff seeks to provide a justification for criminal punishment that upholds the practice as intrinsically valuable, and that avoids the sacrifice of the proportionality of sentence to crime that occurs in consequentialist accounts of deterrence, rehabilitation, or public protection. Duff's argument for retributive punishment proceeds by way of a theological analogy; where the term “secular” is held to chasten and refine the analogical extension in ways palatable to the norms of a liberal democratic polity. As a recent Festschrift puts it, “Duff is a particularly ‘humanist’ retributivist—while retributivists often see punishment as something done to offenders, Duff sees punishment as a kind of process, something that is done with offenders” (Reiff and Cruft 2011, 7). Thus punishment aims at the three stages of penance—repentance, reform (including reparation), and reconciliation (Duff 2001, 107–12). The state acts justly to impose only those forms of hard treatment which engage the offender in a communicative process in which she may be brought to suffer the pains of censure, repentance, and remorse. Unlike more basic forms of retributivism, there is no explicit justification here for a broader economy of deserved pains to rebalance some putative moral order. Duff has no interest in defending current penal regimes in late modern Western societies—which display, in practices like mass incarceration, a prevalent if rationalized vindictiveness and excess. Criminal punishment properly aims, then, to persuade the offender to make this censure her own, to recognize the need to reform herself, and to accept their punishment as a forceful means of apology and a move towards reconciliation both with the victim and with the community whose values have been violated (Duff 2003b, 300). Necessarily, this range of justifying aims need not all be met to establish the legitimacy of state punishment. The state, as the political authority of a fundamentally non-voluntary community, cannot easily be equated with the religious community—for one cannot easily leave one's state. Likewise, punishment cannot be conditional on the offender's confession of their wrongdoing or their recognition of the censure offered by the state (Duff 2003b, 302). Nonetheless, Duff maintains that an appropriately secular use of the penitential analogy models well the communicative nature of an adequately limited retributive and reformative punishment.

In this essay I briefly outline what I name a “postsecular” vision of punishment as penance. While the term “postsecular” is multifarious and contested within contemporary debate, I mean here to denote simply a greater sensitivity to the continued import of the theological claims and systems embodied in the penitential analogy. That is, for Duff, the term secular does not denote a strong sense of the exclusion or privatization of religious content. Likewise, my claim will only be that a penitential account of punishment remains founded on key Christian theological insights, and may draw more fully on these to address key objections and concerns arising from the proposal of a penitential model. In so doing it does not sacrifice its full public relevance (Bourne 2009). I take this to be a relatively minimal and uncontroversial usage of the term “postsecular.” Duff's extensive work on the justification of punishment explores how any such account flows from and has implications for a range of wider philosophical issues: including one's account of political community, the status and role of criminal law, and, most importantly for our present concerns, a moral and political anthropology. The practice of punishment carries with it a vision of the responsibilities and limits which shape a state's action upon the acts and attitudes of its citizens (Duff 2001, 35). There are thus two intertwined intended audiences in the essay. For theologians and scholars of religion, I suggest that a critical reception of Duff's highly suggestive approach is timely and important. For philosophers, theologians, and others who no longer approach public moral debate through a strongly secularizing approach to religious content, I suggest that ongoing discussions surrounding criminal punishment can fruitfully attend to the divergences and commonalities between Duff's liberal communitarianism and the moral and political anthropology underlying the Christian theology of penance.

Duff's model of secular penance has received a substantial amount of critical response within analytic philosophy of punishment, but there has been minimal reaction and evaluation from Christian theological ethicists. Significantly overstating the extent of explicit theological and religious reference, Duncan Forrester celebrates that Duff has “mined deep into Christian tradition” in his pursuit of a view of punishment as communicative, retributive, and reformative (Forrester 1997, 75). Andrew Skotnicki defends a retributive and broadly penitential theology of punishment with a few tangential citations of Duff (Skotnicki 2006a, 110; 2006b, 195; 2008, 142). Hans Boersma draws en passant on Brenda Baker's critique that the analogy of penance fails to reflect sufficiently the difference between the sacramental concern for the restoration of the offender and the larger interest for the community central to the practice of criminal justice (Boersma 2003, 77n31; Baker 1992). Timothy Gorringe notes, largely in passing, three factors congenial to my account here—he endorses Duff's communicative model (Gorringe 1996, 16), notes the expiatory assumption in his account of penance (Gorringe 1996, 244, 256), and raises some appropriate cautions regarding the rationalist presuppositions of his retributivism (Gorringe 2004, 84–85).

While the penitential analogy is helpful in regard to the communicative function of punitive practice, Duff's secular penance still assumes the characteristically liberal rejection of the avoidance of state interference in interior states or moral education. This latter refusal may strike some theological readers as odd. After all, is not penance precisely a means of moral and religious (re)education? In eschewing this, Duff does significantly more than simply delineate his account from those technically termed “moral education” theories of punishment (Morris 1981; Hampton 1984). He rejects the paternalistic tenor and intellectualist notion of wrongdoing he discerns in the term education. There is, at root here, a Kantian concern for the treatment of offenders as responsible rational agents. For Duff, moral reform is thus a broader category than moral education (Duff 2001, 89–92). The communicative enterprise of punishment is undermined when moral reform by persuasion is narrowed to a technical reorientation of moral awareness (Duff 2001, 212n13) Nonetheless, this moral reform is limited in Duff's account to an intermediate level—deeper than a reorientation of action, but not as extensive as an operation on a person's “character.” I show later that this account of the limited depth of moral reform is difficult to sustain. The concern to avoid totalitarian or otherwise manipulative excesses of state operations is salutary, though over-drawn—for not all relations, dependencies, and contingencies impacting interior states are inappropriately coercive. I suggest such dangers are not best guarded against on the basis of a typically attenuated rationalistic account of moral agency in which liberal society eschews any interference in character, properly understood.

In order to defend a limited engagement with character through the formation of a civic virtue of penance, I explore the intimate relation between sacrament and virtue found in the work of Thomas Aquinas. The model of postsecular penance I defend does not eschew state participation in moral formation, for such is unavoidable. There remain three limits to state punitive action—a limit to the grounds, a limit to the recipient, and a further limit consequent upon divergent moral anthropologies. The first limitation, of grounds, entails that such action is restricted in application only to that select range of criminalized public wrongs which may be reasonably expected to be amenable to some moral and material reparation. Likewise, the second limit requires that the offender may plausibly be regarded as capable of at least modest reform. While Duff clearly indicates the first of these limits, his account of the communicative function equivocates on the second and thus is unable to clearly reach what I shall show to be a necessary corollary of communicative punishment—the conclusion that such punishment must not make pain, beyond the inevitable pains of remorse, repentance and reparation, the justifying aim of punishment. The third limitation to state punishment pertains to the potential for excess in penitential punishment. Here the nature of the limit depends on an underlying moral anthropology. The secular analogy transposes penance into an alien anthropology in which many interior states are deemed immune from moral reparation. This risks rendering penance excessively retributive, exceptional, and episodic. By contrast, I suggest a postsecular theological limit. The original theological anthropologies of the penitential traditions focused far more on the ongoing formation of character and virtue. As such they hold out the vulnerable and fragmented hope of something more than a merely formal reconciliation, without thereby extending necessarily beyond the state's theological mandate to preserve order and sustain the space for the formation of civic virtue.

The State's Side of the Dialogue: Wrath, Censure, and Invitation

  1. Top of page
  2. Abstract
  3. Liberal and Theological Limits to Communicative Punishment as Penance
  4. The State's Side of the Dialogue: Wrath, Censure, and Invitation
  5. The Offender's Side of the Dialogue 1: Intervention in Character and the Problem of the Abbot
  6. The Offender's Side of the Dialogue 2: Sacrament, Virtue and Sincerity, or The Problem the Priest
  7. References
  8. Biography

Duff presents the process of criminal trial as an attempt to “engage the defendant in a rational dialogue about the justice of the charge which she faces, and to persuade her—if that charge is proved against her—to accept and make her own the condemnation which her conviction expresses” (Duff 1986, 233). The process of trial and punishment aims not simply to encourage a prudential decision for obedience to law, but to elicit from the offender an appreciation of the wrongness of his or her action. Punishment is communicative, rather than simply expressive insofar as it represents a rational and reciprocal act in which the offender is engaged as an active participant (Duff 2001, 79–80). The communication of censure is accompanied by a communication of that society's demand for repentance and reform. Indeed, unlike expressivist justifications, whether secular (Feinberg 1970; Primoratz 1989) or theological (O'Donovan 1998; 2005, 101–24), it also presents the offender with a means of response.

The reciprocity and reconciliatory intent of this communicative model is appealing. The secular act of judgment and censure retains an analogical link with the biblical account of divine wrath and judgment; both are essentially communicative and invitational. God's wrath (orgē) is not primarily an emotional disposition or condition (affectus) but an act (effectus)—the opposition to sin known as judgment. Judgment aims to elicit a response of repentance, remorse and moral change.1 Thus wrath and reconciliation are far from antonymic. Furthermore, wrath and judgment are then not opposed to God's reconciling mercy, but find in that mercy their operative condition (Romans 9:22–23, Ephesians 2:3–10; see also Barth 1957a, 386–406; Travis 2008, 53–73). The outworking of wrath is not a pain excessive or extrinsic to the act, but the permissive and reluctant outworking of the full awful consequences of the wrongdoing itself. The censure of divine judgment aims to evoke what R. C. Moberly called the “wistful desire for renewed fellowship” (Moberly 1909, 65; Fiddes 1989, 104–5). Thus divine judgment is the enacting of the opposition to sin that invariably entails an invitation to obedience. The invitation is not secondary or distinct to the act of judgment.

Here emerges the first substantive divergence of the secular and postsecular accounts. Both seek to hold together censure and invitation. This becomes difficult in Duff's account because he retains a common retributivist separation of justice from mercy. In contemporary analytic defenses of penitential retributivism this characteristically takes the shape of a construal of mercy as exceptional and intrusive on the domain of justice (Duff 2007, 2011; Garvey 2004, 2007; Murphy and Hampton 1988). Historically, the antonymic separation of justice from mercy emerges most strongly alongside the medieval development of Gregorian canon law. There, mercy is tamed, becoming at best a defense of equity in punishment, and a rationale for modest and occasional clemency. Mercy was held to temper the excesses of the demand for justice; mercy seasoned justice to render it more palatable. It was, however, subordinate to the divinely maintained balance between transgression and retribution (Slaughter 2010, 437). The increasingly rationalist juridification of justice thus took leave of the earlier properly mysterious coextensive relation of justice and mercy (Berman 1983, 179). The shift in the relation between the two was far from uniform. Thomas Aquinas advanced a simple retributivist demand for punishment deserved for damage to the order of human law (ST I-II, Q. 87, A. 1, 6). Yet he also provided an account of the non-competitive compatibility, if not simple harmony, of justice (iustitia) and mercy (misericordia) which had the (unfortunately underdeveloped) potential to at least chasten the triumph of retributive justice-as-fairness over the restorative justice-as-reconciliation. In Aquinas the deserved punishment for violation of human law is distinct from the pain of remorse which arises from the sin violating reason itself, and from the violation of divine law, in which punishment is, or may be, inflicted by God. The latter was understood in a clearly expiatory and propitiatory defense of redemptive suffering (ST III, Q. 49, A. 3–4); on the basis of which it is hardly surprising to hear him argue for something strikingly close to Duff's picture of punishment:

The stain of sin cannot be removed from man [sic] unless his will accept the order of Divine justice, unless either of his own accord he take upon himself the punishment of his past sin, or bear patiently the punishment which God inflicts on him; and in both ways punishment avails for satisfaction. Now when punishment is satisfactory, it loses somewhat of the nature of punishment: for the nature of punishment is to be against the will; and although satisfactory punishment, absolutely speaking, is against the will, nevertheless in this particular case and for this particular purpose, it is voluntary. … We must, therefore, say that, when the stain of sin has been removed, there may remain a debt of punishment, not indeed of punishment simply, but of satisfactory punishment. (ST I-II, Q. 87, A. 6)

It is notable that this sense of the transformation of punishment in becoming “satisfactory” is here predicated on the continued pertinence and power of divine mercy. Unlike the simplistic construal of mercy as a supererogatory form of punitive leniency, Aquinas insists on the more basic foundation of mercy in charity, such that the community may aid the offender's recovery of virtue (ST II-II, Q. 25, A. 6, ad. 2).2

It is necessary to note two areas of contrast between Aquinas's view and that of Duff. In the first instance, Duff's view is clearly more appealing in that it repudiates the most basic form of retributivism which may be called the justification of deserved pain; whereby hard treatment is justified on the basis of an uncritical baptism of the allegedly universal intuition that wrongdoers deserve to suffer. The pain of remorse is central to Duff's argument that restorative justice still enacts an essentially retributive restoration (Duff 1998a, 51–52; 2002; 2003a; see also Daly 2000). Though somewhat ill-defined in most retributive accounts, remorse is clearly somewhat more than pangs of regret. It is, one might suggest, regret driven to self-censure. In remorse I do more than merely regret the harms done (though I may come to appreciate these in a deeper empathy with the victim); I also begin the process of disavowal of the desires, passions, intentions, or choices from which the wrongdoing flowed. Importantly, it is in seeking to induce remorse that practices like mediation and community service retain, for Duff, a retributive justification (Duff 1992; 2001, 92–99; 2003a). The pain of remorse is the recognition of guilt. This is to be distinguished from the anguish generated in the realization that such guilt warrants my own hard treatment. This latter claim is found, for example, in Michael Moore's notion of “feeling guilty unto death” for committing some horrific offense (Moore 1997, 145). Here Duff rightly regards Moore's essentially intuitionist account of remorse as reducible wholesale to what has here been named deserved pain (Moore 1997, 104–52; Duff 2001, 24–25). The justifiable and necessary pain of remorse is far closer, one might suppose, to the “godly grief” induced by St. Paul's lost letter of rebuke to the Corinthians that moved the recalcitrant community to repentance (2 Corinthians 7:8–13). This, as Aquinas points out, is the root of the Latin contritio—to crush. Thus “because remission of sin requires that a man completely relinquishes the effects of sin, through which he had a certain continuity and solidity in the sensitive part of his soul, this act by which sin is forgiven is … called contrition” (In IV Sent d.17.q.2.a.1. Qa.1; quoted in Luijten 2003, 56).

Where, by contrast, Aquinas retains the upper hand over Duff is in retaining a strong connection between justice and mercy. Here Aquinas may have been an ally for Karl Barth (contrary to Barth's own reading). For both, because divine misericordia is (to adapt a Barthian trope) the operative condition of divine judgment, this judgment functions as “the veil of the divine mercy” (Barth 1957b, 214). On this basis Barth then rightly insists that: “The primal, basic decision of God with regard to man is His mercy, the engagement of His heart, and therefore His most intimate and intensive involvement with the latter's existence and condition. It reveals that even God's judgment is sustained and surrounded by God's mercy, even His severity by His kindness, even His wrath by His love” (Barth 1957b, 211; see also 393). Likewise, in Aquinas the relation of iustitia and misericordia is not one of contradiction, for mercy exceeds and perfects the demands of justice (ST I, Q. 21, A. 3). Unlike mercy, divine anger is never deemed to be an attributa dei. No balancing act is required to steady the competing demands of anger and mercy (De Grijs 1998, 36).

By contrast, for Duff, mercy is “an intrusion into, rather than an aspect of, the criminal law's normative structure” (Duff 2007a, 362; see also 2011, 478). This is not to say that he thinks such leniency inappropriate or unjustifiable. Rather, it is the exceptional interposing of justifiable concerns for elements of the human condition of the offender (their socio-economic situation, their family obligations, etc.) which are not admissible within the consideration of the criminal act itself at trial. In so arguing, Duff elides “criminal justice” with “retributive justice” (Tasioulas 2011, 41). Thus, in offering countervailing reasons for the mitigation of an offender's sentence, mercy is found to run counter to justice itself. Important as they may be to the practice of sentencing, the grounding events of mitigation (not all of which may be adequately described as merciful) do not impact the punishment deserved. Duff claims that, once we eliminate from our consideration the excessive harshness of current penal practice, there are no grounds for conceding a role for mercy in attenuating genuine desert (Duff 2007a, 387). In such circumstances any punishment which seeks, as mercy must, to exact a penalty more lenient than deserved, must thereby compromise the effective communication of censure and a demand for fully public repentance. Duff thus construes the prior repentance of an offender as a mere private contrition, which does not warrant mitigation (Duff 2001, 118–21; 2007, 384–66; 2011, 485–87). This would be the case even if an offender had acted on his or her remorse by seeking out and apologizing to the victim, and offering some privately agreed upon reparation. Only undertaking the modes of public repentance and reparation stipulated by the trial and sentencing of the offender may constitute genuine repentance. On this basis any prior repentance by an offender would not merit leniency, but would permit the already repentant offender to regard their punishment as a genuine and necessary penance from the very start (Duff 2001, 121).

A fully public form of hard treatment is held to be “intrinsically appropriate” to the penitential model of punishment (Duff 1998a, 51–52). Unfortunately Duff nowhere satisfactorily explains how or why hard treatment necessarily “constitutes a forceful and public apology” (Duff 2001, 119) in a way that brings about, or at least makes possible, reconciliation with the political community. Reconciliation requires the initiative and gracious invitation of the offended parties,3 and thus also a reasonable level of attentiveness to the response of the offender. It is thus somewhat surprising to find Duff rejecting the mitigation of sentence on the basis of the prior repentance of an offender. Communication with no real interest in reception and response is really only expressivism. Thus mercy is not merely the compassionate leniency of sentencing springing from norms and values outside of justice. Rather, on the account of judgment and censure established above, the drive for reconciliation and the costly enterprise of seeking for it are merciful from the first. There is, then, some basis for recognizing the prior repentance of an offender within the norms of justice. The assumption of the sincerity of such repentance must be the defeasible presumption of the criminal trial. Recognition of prior repentance would not vitiate the need for punishment entirely, because, Duff rightly insists that repentance requires the kind of public performance which can only be sustained within the imperfect practices established by a community for its reception of such messages. These are found in certain forms of hard treatment in which the pains of remorse and repentance are undergone. They thus retain a retributive element. Nonetheless, in order to avoid the total subsumption of the wrong done to the direct victim under the broader category of the wrong done to the community as a whole, it is necessary for a communicative vision of justice to both make space for (state-funded) mediatory processes, and to have sufficient recognition of the ad hoc equivalents to these that may have been undertaken independent of state apparatus at the initiative of either repentant offender, or amenable victim. To such an extent the separation of civil mediation from criminal prosecution would be blurred in practice, but not, I suspect, in a way that would undermine the integrity of either process. The inclusion of mercy as both an enabling condition of justice, and a mitigation of retribution, within the communicative enterprise is thus consistent with the diverse range of potential responses open to the offender. “The purpose of reformative punishment … is not so much to reform the offender, as to persuade her to reform herself: to “correct” her as a rational and responsible moral agent” (Duff 1996, 50). It seems disingenuous, then, to discard some of these responses in privileging the retributive over the reconciliatory.

It is the responsive constituents of the dialogue that represent the strongest elements of the theological analogy; for secular penance desires three connected aspects of its penitent:

  • 1) 
    communication of repentance through the acceptance of punishment as a penance (not merely as a cost of criminal activity);
  • 2) 
    a form of public apology through the willed undertaking of hard treatment of an appropriate magnitude;
  • 3) 
    the making of reparation, where possible.

Importantly, for Duff, the success of all aspects of this process is not required for the legitimacy of punitive action. Rather the expression of censure immediately justifies the imposition of hard treatment. It is sufficient, for Duff, that the offender be addressed “as someone who could respond appropriately” (Duff 2009, 91; see also 2011, 121–25). This is not the strict or traditional variant of retributivism in which punishment is held to be appropriate solely as a “backward-looking response to a past crime” (Duff 1996, 46). Necessarily, it cannot be held to fail in its basic retributive justifying aim even if the offender makes no satisfactory response. What justifies the practice is the attempt, not its success. However, if the event of censure is valid only if it contains within itself the invitation to penitence, it is not at all clear why this should remain a dominantly retributive form of punishment, if by that we continue to mean, as Duff does, a form of deserved “hard treatment.” To the extent that a communicative theory of punishment aims to reinstate the invitational aspect of censure, it is to be commended. However, Duff's form of retributivism undercuts something of the communicative justification in its construal of mercy as intrusive upon justice; and a focus on the justificatory sufficiency of censure regardless of response, which all too easily bleeds into a more basic account of retributive suffering as deserved pain.

The Offender's Side of the Dialogue 1: Intervention in Character and the Problem of the Abbot

  1. Top of page
  2. Abstract
  3. Liberal and Theological Limits to Communicative Punishment as Penance
  4. The State's Side of the Dialogue: Wrath, Censure, and Invitation
  5. The Offender's Side of the Dialogue 1: Intervention in Character and the Problem of the Abbot
  6. The Offender's Side of the Dialogue 2: Sacrament, Virtue and Sincerity, or The Problem the Priest
  7. References
  8. Biography

In the final two sections I consider two important aspects of Duff's account of the offender's side of the dialogue. Many of Duff's critics accuse him of advocating “compulsory attitudinizing” or an illiberal intervention in interior states and character (Von Hirsch 1993, 83; Von Hirsch and Ashworth 2005, 94). The fear is of the state overstepping its liberal limits and straying into an excessively moralizing form of paternalism. The concern with illegitimate intervention is associated with the applicability to the state of the religious role of the monastic abbot. Arising from the same issues of moral and political anthropology, my final section considers a second objection, raised far less commonly, regarding the sincerity of the offender's putative acts of penance. This essay has already discussed the problem of the already repentant offender. In discussing the nature of sincerity, I now turn to consider the obverse scenario—to what extent may we regard the punishment of an avowedly unrepentant offender as sufficiently penitential? Here I reflect on the limitations of the role of the priest. In both cases I show how fuller attention to Christian penitential traditions, and in particular the thought of Thomas Aquinas, may provide an alternative solution that is preferable for both a theological ethics of punishment and broader postsecular political discourse: the state is neither abbot nor priest, and in respecting those theological limits, it may legitimately regard itself as pursuing justified penitential punishment.

Duff defends himself against the connotation of an imposed script for the offender's role in the dialogue by insisting that while participation is required this respects the standard liberal refusal to intervene in attitudes or feelings (Duff 2001, 110). The gentle communitarianism of Duff's approach is susceptible to libertarian objections that a state that establishes the appropriate mode of apology and reparation and then compels the offender to follow this route, fails to respect the liberty of the offender (Brownlee 2011, 55). Likewise von Hirsch and Ashworth argue that Duff's claim that the offender is not forced to mean what they are required to enact is scarcely sufficient because “what is degrading about compulsory attitudinizing is precisely being compelled to express what one does not mean” (Von Hirsch and Ashworth 2005, 94). Von Hirsch objects that Duff's account places the state in the role of a monastic abbot, charged with a level of personal scrutiny unfitting for relations between state and citizen (Von Hirsch 1993, 73–74).

Both this critique, and to a lesser extent Duff himself, assume here a liberal moral limit prohibiting state intervention in the attitudes, dispositions, and character of its citizens. This limit is predicated on an account of interiority as self-contained and inviolable that owes far more to seventeenth century conceptions of the self than to the Christian tradition. Despite some ambiguity and even playfulness in Augustine's Confessions on the function and audience of confession, there is a marked difference between this liberal construal and the interiority of the confessor in pre-modern Christian penitential theology; namely that the latter is a socially constituted interiority flowing into public liturgical expression. Augustine's confession is not simply a silent and private transaction between confessor and God—this is a confession intended to be overheard (Taylor 2009, 35–38). In contrast the sixteenth-century development of a private confessional box feeds more directly into John Locke's conception of the self as a private dark room shut away from the prying gaze of others (Cary 2000; Bernasconi 1988, 85).

More than his liberal and libertarian critics, Duff escapes the worst excesses of this anthropology in affirming that “the distinction is thus not between two sharply separate realms of ‘inner’ and ‘outer,’ or of ‘character’ and ‘action’; it is rather between deeper and shallower, or between more or less substantial, interpretations of the person and her agency” (Duff 2002, 174). Despite this affirmation, and others like it (Duff 1993; 2001, 68, 127–28; 2007b, 115–21) this conclusion is not borne out in the diverse and imprecise accounts of character in Duff's discussion, in which character and action are again separated (for example, Duff 2002, 178). This is most problematic, for our current purposes, when Duff eschews character as a legitimate object of punitive action. He does so, unfortunately, through a vacillation in his understanding of the nature of character. At times, for Duff, “character” names the “whole moral being” of a person (Duff 2003b, 303) thereby conflating modern “personality” with the more precise notion of character, and thus including religious belief, sexual orientation, marital behavior, and “taste in food, music, literature or sport; his good or bad table manners” (Duff 2011, 479). This involves a dangerous rhetorical slippage—for it bundles in a variety of preferences irrelevant to the task of criminal judgment (such as musical taste) or inadmissible in such determinations (racial prejudice), and on this basis rejects intervention in “character” as a whole.

What then, of the restriction of the criminal law to a “depth” above that of the offender's character? Again one finds equivocation this time in relation to the affective dimension of wrongdoing—for he claims that penance includes attitudes, dispositions, and feelings (Duff 2003b, 301) and appears to commend the Aristotelian-inspired account of the essential cognitive and evaluative dimension of the emotions (Duff 2002a,2002b, 161–62) but nonetheless holds this distinct from character or “the innermost citadels of [our] souls” (Duff 2003b, 303). Yet he then also rejects inquiry into the sincerity of remorse; and claims that the public wrongfulness of criminal actions is not generally constituted even in part by attitudes or motives (Duff 2002a,2002b, 171–72). Only if one first accepts that “character” names that private self in some way “deeper” than one's public acts, can one then maintain that the concern of the criminal law solely with public wrongs (Duff and Marshall 2010; Duff 2007, 140–46) requires a limit of punishment to a level of abstraction away from character (Duff 2002a,2002b, 178–79). Duff clearly and rightly defends a form of modest legal moralism (Duff 2001, 67; 2012), but seeks to hold this separate from the processes of the habituation of virtue and the shaping of character.

Duff, drawing on Aristotle, describes character as a settled or stable disposition (Duff 1993, 364). However, whereas the Aristotelian account of a stable disposition is qualified by a strong teleological account of moral formation as habituation, Duff's account of stable character is somewhat static—it lacks teleology and malleability. In such invocations of settled character Duff endorses Jeffrey Murphy's distinction of “character retributivism” and “grievance retributivism” (Murphy 1999; Duff 2003b, 303). Murphy and Duff object to the latter on the basis of the danger of an excessive even vindictive notion of desert if retributivism is based on permanent fault in a person's moral core. Further Murphy argues that character retributivism is to be regarded as incompatible with Christianity insofar as it violates theological and epistemic limits regarding the place and expertise of human creatures to make such judgments (Murphy 2003, 270–71). Here Murphy is quite right to raise, but somewhat simplistically over-draws, the theological chastening of state punishment inherent in its recognition as a solely penultimate judgment. He is right to insist on the impermissibility of any final declaration or judgment on a person's moral core. But only if the interior self is entirely inviolable and in possession of an immutable and natural character can a more proximate and modest reorientation be dismissed as a legitimate goal of penitential punishment.

While Duff does not share the simpler libertarian account of inviolable interiority per se, his refusal of intentional censure of character leads, against his intentions, to an excessive separation of exterior action and interior virtue. A far more satisfactory account takes its cues from the claim, articulated most explicitly by Aquinas, that penance is both a sacrament and a virtue. The Christian penitential traditions are thus unlike those elements of Kant which contemporary retributivists most emphasize, which regard moral sanctification and the substantial improvement of character as beyond historical realization (Kant 1998, 68).4 In denying the malleability and contingency of character, Murphy's “character retributivism” assumes that judgment of character is permanent and wholesale—an imputation of an inner wickedness or deep and culpable depravity (Murphy 1999). The enacted judgment of punishment thereby signifies a stigmatization of the morally weak character, which distances “them” from the rest of us. The corollary of this vision of character would be the kind of tyrannical demand for immediate and incontrovertible confirmation of the completion of penitence found, for example, in Montaigne's On Repenting.

As for me, I can desire to be entirely different, I can condemn my universal form and grieve at it and beg God to form me again entirely and to pardon my natural frailty. But it seems to me that that should not be called repenting any more than my grieving at not being an angel or Cato. … Before I call it repentance it must touch me everywhere, grip my bowels and make them yearn—as deeply and as universally as God does see me. (Montaigne 2004, 916–17)

In contrast with this, Christian eschatology imagines a restoration, rather than a replacement, of our nature. Punishment as postsecular penance constitutes but one intense episode in the ongoing narrative of character formation. Without this broader focus, the diachronic element of virtue or vice is separated from the teleological. Such a separation is impossible for the Christian penitential traditions in which the life of discipleship is understood around two distinct but inseparable metaphors—of both a “dialogue” and a “journey” (Hauerwas 1985, xxix–xxxii; Meilaender 1987, 34–38). The dialogic element, which I have here outlined in reference to wrath and invitation, finds an impressive secular analogue in Duff's communicative theory (though with the divergence on mercy noted above). However, the “journey” element appears constrained within the short period of the criminal sentence, and is largely lacking in the case of the unrepentant offender (to whom we shall return).

How then are we to answer the problem of the abbot? Much of the concern about paternalist intervention evaporates when one adopts the view that character is both malleable and contingent, and, importantly, that both offense and punishment entail an unavoidable impact on character. In distinction to the wide variety of deserved pains defended in many retributive justifications of punishment, I suggest that the pains of remorse, repentance, and reparation are intrinsic to moral repair, precisely because they impact the offender's character. Importantly, the repudiation of wrong and the reconciliation of wrongdoer to community are not peculiar to process of criminal punishment. Rather social life must always entail holding some cares, commitments, or values in common. Public censure is best conceived as a means of communicating to the offender that they have violated values held dear by that society; that society cares about bodily integrity, private property, or some other public good. Censure rightly adopts an imperative mood. In so doing the community or the state invariably “puts pressure” on the offender—not merely on their actions, but on their affections, desires, and intentions—that is, their character. Such pressure is inevitable, though open to abuse. As Stanley Hauerwas rightly claims, we “necessarily impinge on the ‘freedom’ of others. … There is no morality that does not require others to suffer for our commitments. But there is nothing wrong with asking others to share and sacrifice for what we believe to be worthy. A more appropriate concern is whether what we commit ourselves to is worthy or not” (Hauerwas 1983, 9).

In contrast to Duff's account, the limit preventing the state becoming an abbot is to be conceived not so much through imagery of the depth of intervention, as of its methods and extent. In terms of method, state punitive action is normatively constrained to only those forms of hard treatment that may elicit the pains of remorse and repentance, and enact, however incompletely, some form of reparation.5 That is to say, it is constrained by the norms of effective moral communication—including respect, the absence of manipulative strategies, and freedom of thought and speech. These latter (both liberal and Christian) relative goods are not compromised by censure or an invitation to reform of character. The moral objection to state manipulation or coercion of character is not that it delves into the forbidden territory of private intentions, but that it is counterproductive to the goods sought through communicative punishment. The practices of restorative justice (mediation, dialogue, and a variety of forms of symbolic reconciliation and material reparation) are intrinsically appropriate to these goods. Indeed, there is no “imposed script” in such practices, if we mean by this the arbitrary setting of “lines” by the state. For the notion of “intrinsic appropriateness” claims that the state is imposing the action, but not arbitrarily determining the form. Indeed in many practices of restorative justice the form is not set, beyond the creation of a communicative setting in which reconciliation may be appropriately improvised. The guard against excessive paternalism arises through the limitation of the extent of such intervention both by the tests of communicative adequacy and by the limitation of prosecutable crimes to significant public wrongs. As such, state punishment aims not to shape the whole of the offender's character, but to engage only in the formation of civic or public virtues.6 That is to say, it seeks to address those wrongs and harms impacting the goods of political society (ST I-II, Q. 96, A. 2–3). The reorientation of the will in penance is thus the habituation of a disposition of thought and action to the public good.

The Offender's Side of the Dialogue 2: Sacrament, Virtue and Sincerity, or The Problem the Priest

  1. Top of page
  2. Abstract
  3. Liberal and Theological Limits to Communicative Punishment as Penance
  4. The State's Side of the Dialogue: Wrath, Censure, and Invitation
  5. The Offender's Side of the Dialogue 1: Intervention in Character and the Problem of the Abbot
  6. The Offender's Side of the Dialogue 2: Sacrament, Virtue and Sincerity, or The Problem the Priest
  7. References
  8. Biography

The problem of the priest follows from the problem of the abbot, precisely as the declaration of absolution follows on from the cure of souls. The latter problem identifies concerns of excessively paternalist intervention. Regardless of whether one draws the limits of state paternalism in respect of the putative “depth,” “extent,” or “method” of such intervention, a further challenge then arises for a communicative account of punishment—how are we to judge the effectiveness and sufficiency of the state's penal communication with the offender? Does the success and sufficiency of penal communication, and the implied declaration of absolution, depend on the sincere contrition of the offender? A positive answer to this question would have two undermining consequences—first, it would render the justification of state punishment contingent upon the variable outcome of punishment; and second, it raises the epistemological question of how one may satisfy oneself of the genuine repentance of the offender. Duff's primary answer is to place the strongest justificatory burden on the expression of censure. So long as the state addresses the offender as if they were someone who may respond appropriately, it has met the justifying aim of punishment (Duff 2009, 91; see also Duff 2001, 121–25). We might call this the test of communicative adequacy. Such criteria seem to be an appropriate, indeed unavoidable, approach to the issue. Duff then moves from this to a wider set of claims regarding the sincerity of the offender and their recognition and reception of punishment as a form of penance. In the course of these reflections, however, Duff moves to a liberal formalism which regards the sincerity of the offender's reception of penance as not just indemonstrable, but immaterial to the justification of punishment. Thus the implied “absolution” declared by the state (its partial mirroring of the priest's role) is a function solely of the formal observance of a ritual, and not related to the attempted formation of civic virtue. Therefore, a postsecular account of the sincerity of the offender's remorse and repentance will have to be drawn differently. Must one regard the sincerity of the offender's penance as either immaterial to or fully constitutive of the justification of punishment as penance? I suggest not.

Duff equivocates regarding the capacity of retributive hard treatment to demonstrate sincere repentance. He rebuts the simplistic desire for an assurance of sincerity through hard treatment.

The point here is not just an evidential one—that undertaking penance gives others stronger evidence of sincere repentance than merely verbal confession and apology could give. … It is rather that undertaking a penance, giving this outward and materially burdensome expression to the painfully burdensome recognition of one's own wrongdoing, is a way of taking the matter seriously; it part constitutes the repentant sinner's earnest repentance. (Duff 2003b, 299)

Unlike in personal relationships, “in less intimate contexts (such as our dealings with our fellow citizens) there is more room for purely formal apologies whose sincerity is not an issue. What matters is that the wrongdoer apologizes. We do not inquire into her sincerity” (Duff 2001, 110). This may be seen to encourage precisely the duplicity or bad conscience Duff would otherwise rather avoid. He argues that, at the time of sentencing the offender should have the opportunity to affirm or deny the apologetic character of their punishment (Duff 2001, 111). Up to this point, Duff unarguably protects the freedom of the offender. But then he also claims that, regardless of the offender's explicit affirmation or denial of the act's apologetic connotation, society should treat the offender as if he had “paid his debt” because his or her act remains the act most consistent with genuine apology (Duff 1999, 84; Duff 2001, 111). This refusal to give due attention to the sincerity and extent of the offender's remorse and repentance treats as one a number of groups of offenders with radically different stances—those who are genuinely fully repentant prior to punishment (discussed above) (Duff 2001, 118–21); the conscientious offender breaking an unjust law (whom the law should still punish, but in acknowledging the values the offender defends the wrongfulness is to be portrayed not as malum in se but a malum prohibitum) (Duff 2001, 122); and the unrepentant offender who may or may not come to be repentant at a later date, subsequent to the punitive action. We shall return to this last problematic example shortly.

Any satisfactory answer to the problem of sincerity must acknowledge the epistemic limit inherent in human action—it is not possible to fully reassure oneself of the sincerity of another. In the face of this epistemic limit how can the penitential metaphor function? In a religious context the apparent voluntariness of submission to penance is assumed to demonstrate sincerity to a pragmatically sufficient degree. This clearly cannot be the case in necessarily coercive state action. One theorist who addresses the sincerity problem with a strongly liberal limit is Christopher Bennett. Bennett pursues a similar argument to Duff, albeit framed in terms of the more generic “apology” rather than the language of “penance” (Bennett 2006, 2008). For Bennett “ritual” connotes an activity undertaken in formal conformity with legal or communal requirements, but for which no judgment regarding sincerity is possible or desirable. Bennett thus distinguishes between a formal and a moral reconciliation; arguing that only the former is appropriate for state punishment. He goes further than Duff in claiming that formal reconciliation is achievable and sufficient not only in cases of uncertainty over sincerity, but even in the face of explicit repudiation of sincerity. This formal observance retains both the expression of censure and the vindication of the victim. It retains an essentially expressive function—to the offender that they ought to sincerely apologize, and to the victim that the community “stands behind” them (Bennett 2006, 134–35). Nonetheless, Bennett stops short of endorsing Duff's communicative model, arguing that the expression of censure and blame is a sufficient legitimating aim (Bennett 2008, 188–89). The problem for what we might call Bennett's expressivist formalism lies, however, in the protection of the integrity of the offender's interior state through an insistence on the adequacy of mere formal observance of apology. There is an equivocation in the limitation placed on the state's interest in the attitudes of its citizens. “It is permissible for the State to declare a certain attitude appropriate (by requiring that the offender perform those actions that would express that attitude) but impermissible for it to put pressure on the offender to espouse that attitude if he does not hold it” (Bennett 2006, 135–36). Does the declaration, especially when expressed in coerced action, necessarily entail a form of pressure on the offender's disposition? The equivocation consists in a conflation of an epistemic limit—that we cannot publicly assure ourselves of the sincerity of another person's apology, with a moral limit—that we should not inquire into interior states, for that is beyond the remit of the state in a liberal order.

Insofar as the expressivist rationale for punishment does not focus on the potential response of the recipient of punishment, a formalist agnosticism regarding sincerity is not out of place. However, it is certainly counter-intuitive to claim that an offender adopting an apologetic posture, all the while vocally disavowing his sincerity, meets the test of symbolic adequacy. Likewise an offender free to adopt an inner attitude of insincerity, but constrained from public expression of this, would hardly meet the ideal liberal condition of an un-coerced interiority. External action and interior attitude are not so easily separable. There are, of course, moral and pragmatic limits to the ascription of sincerity in punishment, but even those values held most dear in liberal societies cannot be neatly separated from internal dispositions. The limitation of state or community enacted punishment to criminal public wrongs does not thereby prevent it from minimal expectation of practical assent to basic values. This is perhaps most clearly the case in perfectionist forms of liberalism (Raz, Galston, etc.), for which the defense of central liberal values of liberty, autonomy, and toleration are dominant. Even the non-perfectionist Rawls, of course, did not think political liberalism could or should strive to sustain the kind of schizoid and hamstrung “neutrality” that refuses even any state action that may influence the value preferences of its citizens. So even on a strongly liberal account, it is necessary to demonstrate if and how formal completion of a public sentence may be taken to be an adequate demonstration of assent to central public goods and values.7

The primary differences between expressivist (Bennett) and communicative (Duff) formalism is that the latter must give due account to the nature of the hoped for communicative reception and response of the offender, namely their reception of hard treatment as a penance. In exploring this we shall find that there are two moments of the process construed differently in the secular and postsecular accounts of punishment as penance; adopting the sequence of the penitential rite, we may say these are the moment of public contrition and the declaration of absolution by the priest. In this consideration let us consider an example of an unrepentant offender. Peter is convicted of a minor assault—he had punched in the face a man harassing his (Peter's) girlfriend in a bar. He acknowledges his criminal guilt but remains convinced that his victim “got what he deserved.” Peter is sentenced to a community sentence comprising of 100 hours of “community payback” clearing up graffiti and repainting public buildings, a court mandated attendance at a brief series of anger management therapy, and the invitation to victim-offender mediation. Peter refuses the invitation to mediation, attends but does not fully participate in the anger management therapy, and continues to regard the “community payback” as both an unjust imposition and an unfortunate cost of his justified action. To that degree he may be described as an unrepentant offender, who refuses to receive his punishment as if it were a penance. For Bennett, Peter is still regarded as formally reconciled with the community and victim. Duff's position is worth quoting at some length here:

The [unrepentant] offender has been subjected to what would constitute an appropriately reparative apology if he undertook it for himself. His fellow citizens should therefore now treat him as if he had apologized. This is not a matter of pretending what they know to be false—of pretending that he has apologized when they know he has not. Instead, it is another aspect of treating him as someone who can redeem himself—as someone who can, and to whom we owe it to hope that he will, refrain from crime in future. He might not have paid the apologetic debt that he owed, if his punishment was simply inflicted. But something like that debt has been exacted from him, and those who exacted it should now treat him as if the debt has been paid. (Duff 2001, 123–24)

Duff correctly insists on the sufficiency of a sentence broadly proportionate to the public wrong, and thus denies both the possibility of extending the sentence of the unrepentant offender, or of regarding them as “not worth” bothering with, on the basis of some assumed incapacity to change. Any postsecular or theological reception of Duff should concur with these caveats.

However, two notable shifts occur in the second half of this quoted passage to which such a reception of Duff should be attentive. First the introduction of a language of payment of debt fits better with more traditionally retributive, even expiatory accounts of hard treatment as the converse response of an economy of pains. There are both theological and pragmatic reasons for avoiding such language wherever possible. Most substantively though, I note the inclusion of a note of hope. Duff rightly claims that his form of retributivism includes a forward-looking component absent from non-communicative accounts. Hope names the forward-facing desire that, despite current defiance, the unrepentant offender may still, in the longer run, be shaped positively by the punitive action they have undergone. It is thus a theological virtue with an important secular analogue. Duff is certainly right to insist that punishment can be justified as penance “as a proper attempt to pursue [its] ends even when we have good reason to believe that it will fail to achieve them” (Duff 2001, 129). However, the failure of the communicative function of punishment within the discrete bounds of the criminal sentence does not need the language of debt to establish the virtue of hope, any more than the Christian theology of atonement needs an economy of penal pain to establish its eschatological hope. The economic and the communicative are two distinct rationales for establishing the sufficiency of the sentence of the unrepentant offender. The civil community cannot look upon the coerced actions of the unrepentant offender “as if” they were an adequate public enactment of remorse or the beginning of repentance, when the symbolic adequacy of such actions requires the affirmation, if not the demonstration, of sincerity. It is not clear why a coerced action that is formally similar to repentance and remorse, should be held a sufficient payment. The fiction of a formal equivalence of debt and payment lacks demonstration. By contrast the language of hope regards the proportionate criminal sentence not as sufficient payment, but as all that can be properly said. This is a communicative, not an economic limit. That is, the sentence sufficiently communicates censure, the need for reform, and the desire for reconciliation. The state can justly do no more. As in my dealings both with family and friends and with strangers, I am constrained by the vulnerability inherent in communication. I am required not just to express censure, but also to remain open to dialogue—including the full range of possible responses.

The central question in the problem of sincerity relates to the priestly role in two moments of the process of the conversion of what Aquinas would have called vindictive justice into penance. In the case of vindictive justice, compensation or satisfaction takes place in accordance with the judgment of the judge. By contrast penance aims not merely for formal balance and equality of justice but the fuller reconciliation of “friendship.” The difference is found in that vindictive justice is essentially the enforcement of the judge's decision. I might suggest that it establishes only a mythic status quo in an economy of pains,8 whereas, by contrast, penance atones for an offense according to both the will of the sinner and the judgment of God. The will to contrition is present in penance but absent in vindictive justice (ST III, Q. 90, A. 2).

Repentance and the conversion of vindictive justice to penance

It is all too common in formalist accounts to understand public repentance as a singular moment or discrete process in which the offender moves from pride to sorrow, and thus becomes fully penitent. The case of Peter is but one example of a more complex picture. It is suggestive that a distinction between perfect “contrition” and imperfect “attrition” is important for the medieval and early modern understanding of those “imperfectly disposed” to the grace of absolution (Kelly 2008, 259–60). The conversion of attrition (the imperfect, fearful, and unformed sorrow for sin often without any attendant determination for future avoidance of wrongdoing) to full contrition is no inevitable or linear progression. It occurs, insists Aquinas, only with the infusion of the theological virtue of caritas—charity. In post-Tridentine formulations (though not in Aquinas himself) this is not least because attrition names not just an imperfect form of contrition, but a sorrow founded on fear rather than the reception of grace. Aquinas's contrast is less stark; higher forms of attrition are still an orientation of imperfect love of God and the good. Here the analogy of penance would lead us to the view that state action may aim directly only at an imperfect attrition. It may seek to achieve this only in encouraging and hoping for, but not enforcing the conversion of attrition into contrition. Famously, for Aquinas, even punitive acts undertaken in fear may tend toward the development of virtue (ST I-II, Q. 95, A. 1). A tendency toward the formation of virtue is not a description of an inevitable process, but rather indicates its rationale and potential beyond backward-looking forms of retributivism.

This more gradual and variegated picture of repentance connects with our refusal to separate the formal practice from its relation to character. For Duff, “penance” names a formal rite or ritual; but for Aquinas it is both and inseparably a sacrament and a virtue. In relating the rite and the reality of penance Aquinas distinguishes between the sign only and in itself (sacramentum tantum); the action entailing both the thing itself and the sacrament in which it is found (res et sacramentum) which in this case is the confluence of the operation of contrition and repentance within the offender's heart and character with the penitential acts which are a necessary, but not sufficient, means of this transformation; and finally the gracious operation of forgiveness—the thing only and in itself (res tantum) (ST III, Q. 84, A. 1, ad. 3). By contrast the formalist reconciliation of secular penance seeks to sustain the sacramentum tantum without recourse to the two other elements. While some later in the tradition sought similarly to argue that attrition was in itself sufficient as a basis for undergoing the sacramentum tantum, Aquinas holds that contrition is both required for the effective res et sacramentum, and may be infused by grace during the sacramentum tantum. There can be no unwilling reception of the res tantum (ST III, Q. 84, A. 3, ad. 5). The formalist performance of state penance enacts a clericalist excess—by regarding the sacramentum tantum as sufficient it does not remain practically agnostic about the occurrence of the res tantum. Rather it prematurely declares absolution on the basis of the sacramentum tantum alone.

Absolution and the limited role of the priest

For Aquinas, the sign (signum) may also be a cause (causa). The declarative act of the priest's absolution mirrors the constitutive act of the penitent's turn to the good. This mirroring is superfluous in the formalist view—rendering the priest's role internally and discretely sufficient, regardless of the offender's response. In the sacrament the operation of grace is such that the very process of undergoing the sacramentum tantum may be found causa of the res tantum. Sacraments are said then to effect that which they signify through the very act of signifying. This can only be maintained in the case of some prior acceptance of self-censure (contrition or higher attrition) on the part of the penitent. This, one assumes, must mean that the state cannot and must not enforce penitential rites on those unwilling to accept them, but may encourage those whose motivations for undergoing the rite are a typical combination of higher attrition and self-interest, in the hope of performative sacramental causation of true contrition.

Aquinas distinguishes between two functions of sacraments—they are both expressive and sanctifying. These are not two discrete elements—as if the sacrament were a mere accidental representation of the virtue. For God gives sacraments as the sensible, or one might say material and practical, mode through which the spiritual reality is made real and manifest. For Aquinas the relation of sacrament to virtue is largely cyclical rather than linear—the sacrament of penance is a cause of grace, but as virtue penance is an effect of grace. The virtue is completed and perfected in the sacrament. Thus there is, as Duff insists an “intrinsic appropriateness” (Duff 1998a, 51–52) to penitential forms of punishment, but it resides in its connection with virtue, not hard treatment. Thus, Peter's punishment is sufficient and proportionate not in that he has “paid his debt”; but because it entails the kinds of actions which communicate to him the need to repent, and provide him with avenues for the formation of civic virtue. The state can say no more. It is not just that the state has fulfilled its own responsibility to communicate, and can now rest content regardless of the response. Rather, given the unification of form and matter in the sacrament, it has exhausted its own capacities, and must now look to grace in hope. The state, as no priest, lacks the promise of the power of the keys, and thus can only assemble the ingredients of penance, and hope.

We have seen that Duff asserts that the community should regard the unrepentant offender's undertaking of punishment as sufficient, on the grounds that the sincerity of the action is not required for the state to be warranted in its declaration of the satisfaction of the demands of justice. Here the separation of sacrament from virtue claims too much for the capacity of the state to achieve justice and declare civic absolution. This stance of expectant waiting for grace is notably enacted when the priest often, though not always, included the pronunciation of absolution prior to the completion of the penance prescribed.9 The root of this may be found in medieval debates concerning the necessity of the priest's officiation in pronouncing absolution. Aquinas favors the option that insists that the ministrations of the priest are indispensable—they both reflect and affect absolution—thus lending his considerable support to the late twelfth-century shift from the deprecatory “Christ absolves you” to the indicative “ego te absolvo.” The deprecatory form, dominant up to that point, denies for the sacrament the full declarative and constitutive complexion of the act. What Aquinas sought to sustain in insisting on the indicative formulation is that, while absolution can be obstructed by an indisposition of the will, the hindrance affects only the arrival of the effect in the recipient, and not the saving power of the absolution itself (Luijten 2003, 168–71).10

The relation of rite to disposition is pivotal here. The indicative formulation of absolution can only accrue a highly clericalist connotation of a discrete constitutive power in possession of the priest himself once the role of priest is construed as a controller or dispenser of grace, rather than a servant of it. This is guarded against in penitential liturgy by the insistence that it is by the power of Christ alone that absolution is offered, thus the deprecatory request for God to absolve occurs in many formulations directly prior to the indicative pronunciation “ego te absolvo.” The virtue of penance is a lifelong calling to sanctification, enabled and achieved in the power of Christ through the operation of the Holy Spirit. The rite of penance serves, enables, and affects the virtue. The priest's pronunciation of absolution is not discretely constitutive of pardon, or a final declaration of a verdict of virtuous character. The Thomist account concludes that the words “I absolve thee” mean “I grant thee the sacrament of absolution” (ST III, Q. 84, A. 3, ad. 5). That is, the declaration of the priest, and by partial extension the state, does not fully constitute the final absolution of the offenders; but sends them out in the hope of continued formation in the virtue of penance.

Importantly for the applicability of the penitential analogy to state operations, Aquinas argues that even the ungodly (for example, Ninevites) and those that follow the “Old Law” [sic] (Jews) may exhibit the virtue of penance (ST III, Q. 84, A. 7, ad. 1) but notes that in so doing they display faith and hope in the pardon of God. This, for Aquinas, is an ontological description, not necessarily a psychological awareness. The offender is held within the remit of hope in God's sanctifying work. Both the offender's turning from wrong and the priest's declaration of absolution are integral to the sacrament, through which God turns the heart of the offender to the truest good. The theological virtue of faith precedes the virtue of penance in that the former is an orientation toward God and the good, the latter being the love of God actively directed against sin. In this way one can argue that the sacrament of penance presupposes the virtue. In that it does not require perfect contrition prior to reception of the sacrament the analogy with state punishment may sustain application to those whose remorse and repentance are incomplete or nascent.

What difference does this complex account of the conversion of vindictive justice to penance make to an offender, like Peter, who repudiates the penitential significance of their sentence? There are three elements of the theological basis of a communicative account of punishment as penance that I suggest may tentatively point to a fuller, postsecular, justification of punishment. Firstly, Peter is not declared to be absolved by a fictive “as if” that runs contrary to his self-description. We do not entertain a myth of equivalent pains. Instead, because Peter has been addressed appropriately, in the space of waiting for his reception of the message, there may be grounds even here to recognize some form of nascent attrition. Some of those who currently disavow remorse and repentance nonetheless offer grounds of mitigation or extenuation for their own culpability. In appealing to extenuation (provocation, the “heat of the moment,” etc.) Peter might still concede the contingency of his action; perhaps its root in a somewhat irascible character, etc. That may be, but is not necessarily, a step on the route to greater penitential self-reflection. The inclusion of such unrepentant offenders in the postsecular penitential model would be sustained only on the basis of a hope for future realizations of guilt and the burden of remorse and reparation. This may well not come to fruition during the discrete and bounded sentence permitted in legal sanction. Secondly, our account of postsecular penance places a stronger emphasis on the forward-looking component that Duff rightly seeks to articulate. This hope is found in the regard in which the community is said to hold the ex-offender: not so much “You have paid something like your apologetic debt and may thus be reconciled with us,” more “You have undergone an enacted invitation to and symbol of the remorse and repentance that are appropriate to your crime. We now hope you will come to acknowledge these as such in your future actions.” Thirdly, and finally, the more explicit articulation of hope avoids the twin dangers of despair and presumption (ST II-II, Q. 20–21) which Aquinas names as the sins against hope. These are set against the gift of filial fear (ST II-II, Q. 19) in a way that may sustain a significant role for the pains of remorse and repentance while (against Aquinas himself) avoiding the retributive excess of deserved pain. Filial fear, unlike servile fear, eschews evil in the hope of sustaining a proper relation with God. Servile fear acts solely out of fear of the consequences and pain of punishment. In seeking formal compliance devoid of interest in character secular penance blurs this boundary and runs the risk of falling into despair or presumption. In its more nuanced forms, as with Duff, secular penance may steer a course to avoid despair, for it holds out the expectation of a reform of behavior, if not of character. However, it is not merely the emptiness of the formal performance of the unrepentant offender that may give us pause, but also its presumptuousness (ST II-II, Q. 21, A. 2). For presumption, says Aquinas, is an inordinate hope conformed to a false intellect claiming that God grants forgiveness to those who persevere in their sin. Formalist solutions to the problem of the priest represent, ironically, a diminution of hope for they hope too little in the immensity of the goodness of God. Consequently they hope too little in the gracious scope for reform of character. Importantly, though, Aquinas goes on to say that presumption is less grave than despair because “it is more proper to God to have mercy and to spare, than to punish: for the former becomes God in Himself, the latter becomes Him by reason of our sins” (ST II-II, Q. 21, A. 2).

I have identified a number of aspects of Duff's account of punishment as penance which should be well received by theologians and scholars of religion working in the area of crime and punishment: the concern for significant reform of current penal practice, the pressing engagement with restorative justice (which rightly enjoys great popularity in Christian theologies of punishment), the articulation of a strong communicative model of punishment, and, of course, the pregnant use of the model of penance. However, I have also noted some difficulties in Duff's account—often related to the more dynamic relation between rite and character in Christian thought. If, as I have claimed, character is continually subject to formation and negotiation then neither censure nor penance are exceptional to social life. If this is the case then the enabling of penitential punishment forms one key part of a broader responsibility of the state to create conditions conducive to repentance in both punitive and non-punitive settings (see also Tasioulas 2007). Thus, in contrast to formalistic limits to penance, I suggest that a stronger account of the relation of sacrament to virtue, rite to disposition, would sufficiently answer the problems of abbot and priest, while also avoiding the attenuated account of agency toward which Duff's view of penance strays. Placing the penitential analogy back into fuller contact with the theological account of judgment, justice, and mercy from which it emerged, will also enhance the answer to the question of how one addresses the repentant offender, and thus how one constructs the normative basis of punishment. While far more remains to be said in this important set of debates, both within theological circles and in wider postsecular discourse, I have identified, in only a suggestive a fashion, elements of a penitential moral anthropology that, in drawing far more on the rich and nuanced penitential traditions of Christianity, may justly be denominated a view of punishment as postsecular penance.

  1. 1

    There is no space here to discuss the important issue of whether such wrath is thus fundamentally personal (though not merely affectus) as, for example, both Karl Barth 1957a (370–71) and Stephen Travis 2008 (54–55) claim, or whether it arises inevitably, even mechanistically, as part of the process of a morally ordered universe—as is maintained by C. H. Dodd 1932, John Howard Yoder 2002 (318), and A. T. Hanson 2010.

  2. 2

    It may be objected here that Aquinas then goes on to outline the grounds for the withdrawal of such merciful charity in cases where an “incurable wickedness” justifies a capital sentence. However, the justification for such punishment is clearly distinct from the broader compatibility of justice with mercy. One need not follow Aquinas's regrettable verdict of the incurability of very great wickedness, nor of the need to address such through the death penalty—both of which militate against the virtue of hope that, as we shall see below, sustains a truly just practice of punishment.

  3. 3

    This begs a range of questions regarding offenses where reparation appears impossible, or at least, severely limited. The all too obvious case of murder may permit some minor “reparations” to the family of a victim, but not to the victim themselves. I cannot enter into these questions here, and need not do so, for the problem of such a conceptualization of reparation besets both secular and postsecular penance, among many other varieties of the justification of punishment. It will have to suffice to note that the justification of reparative justice does not seek to escape the tragic, time-bound “predicament of irreversibility” (Arendt 1958, 237).

  4. 4

    This is, of course, a highly selective and inadequate reading of Kant's account of the potency and nature of moral change. For an insightful critical appraisal of Kant on this issue, see Reno 2002.

  5. 5

    This does not preclude a restricted use of imprisonment on grounds of public protection against violence or sexual predation. This would still, of course, represent a massive downscaling of current imprisonment rates in the majority of liberal societies. It would also still be normatively bound by the requirements of penal communication.

  6. 6

    The term civic virtues has multiple meanings. I use it here in reference to the older Thomist account of a disposition of the will to the common good. This is distinct from the common usage of the term to denote public spiritedness. It is thus far more than being willing to join the PTA.

  7. 7

    There is no space here, of course, to enter into the vexed questions of the nature and extent of the values presumed in such public symbolic actions. For now I simply note that the wall between formal and moral reconciliation will not hold absolutely.

  8. 8

    On the mythic nature of this practice see Ricoeur's famous “Interpretation of the Myth of Punishment” (Ricoeur 1974, 354–80).

  9. 9

    I am grateful to an anonymous reviewer for highlighting this practice.

  10. 10

    The latter history of the indicative form tended to locate in the priesthood a constitutive power which had, against its original intentions, the effect of privatizing, clericalizing, and absolutizing the “power of the keys” and severing the sacrament from its essential Christoformity—which would remain pivotal for the kind of penitential anthropology defended here.


  1. Top of page
  2. Abstract
  3. Liberal and Theological Limits to Communicative Punishment as Penance
  4. The State's Side of the Dialogue: Wrath, Censure, and Invitation
  5. The Offender's Side of the Dialogue 1: Intervention in Character and the Problem of the Abbot
  6. The Offender's Side of the Dialogue 2: Sacrament, Virtue and Sincerity, or The Problem the Priest
  7. References
  8. Biography
  • Arendt, Hannah 1958 The Human Condition. Chicago: University of Chicago Press.
  • Aquinas, Thomas 1981 Summa Theologiae. 12561272. Translated by The Fathers of the English Dominican Province. Westminster, Md.: Christian Classics.
  • Barth, Karl 1957a Church Dogmatics II.1: The Doctrine of God. Translated and edited by G. W. Bromiley and T. F. Torrance. Edinburgh: T&T Clark.
  • Barth, Karl 1957b Church Dogmatics II.2: The Doctrine of God. Translated and edited by G. W. Bromiley and T. F. Torrance. Edinburgh: T&T Clark.
  • Baker, Brenda M. 1992Penance as a Model for Punishment.” Social Theory and Practice 18.3 (Fall): 311331.
  • Bennett, Christopher 2006Taking the Sincerity Out of Saying Sorry: Restorative Justice as Ritual.” Journal of Applied Philosophy 23.2 (May): 127143.
  • Bennett, Christopher 2008 The Apology Ritual: A Philosophical Theory of Punishment. Cambridge: Cambridge University Press.
  • Berman, Harold J. 1983 Law and Revolution: The Formation of the Western Legal Tradition, Volume I. Cambridge, Mass.: Harvard University Press.
  • Bernasconi, Robert 1988The Infinite Task of Confession: A Contribution to the History of Ethics.” Acta Institutiones Philosophiae et Aestheticae 6: 7592.
  • Boersma, Hans 2003Liturgical Hospitality: Theological Reflections on Sharing in Grace.” Journal for Christian Theological Research 8 (January): 6788.
  • Bourne, Richard 2009 Seek the Peace of the City: Christian Political Criticism as Public, Realist, and Transformative. Eugene, Oreg.: Cascade.
  • Brownlee, Kimberly 2011The Offender's Part in the Dialogue.” See Cruft, Kramer, and Reiff, 2011, 5467.
  • Cary, Phillip 2000 Augustine's Invention of the Inner Self: The Legacy of a Christian Platonist. Oxford: Oxford University Press.
  • Cruft, Rowan , Matthew H. Kramer , and Mark R. Reiff , eds. 2011 Crime, Punishment, and Responsibility: The Jurisprudence of Antony Duff. Oxford: Oxford University Press.
  • Daly, Kathleen 2000Revisiting the Relationship between Restorative and Retributive Justice.” In Restorative Justice: Philosophy to Practice, edited by Heather Strang and John Braithwaite , 3354. Aldershot: Ashgate.
  • Dodd, Charles H. 1932 The Epistle of Paul to the Romans. London: Hodder & Stoughton.
  • Duff, R. Antony 1986 Trials and Punishment. Cambridge: Cambridge University Press.
  • Duff, R. Antony 1992Alternatives to Punishment—or Alternative Punishments?” In Retributivism and its Critics, edited by W. Cragg , 4368. Stuttgart: Franz Steiner.
  • Duff, R. Antony 1993Choice, Character and Criminal Liability.” Law and Philosophy 12.4: 345383.
  • Duff, R. Antony 1996Penal Communications: Recent Work in the Philosophy of Punishment.” Crime and Justice: A Review of Research 20:197.
  • Duff, R. Antony 1998Desert and Penance.” In Principled Sentencing: Readings on Theory and Policy, 2nd ed., edited by Andrew Von Hirsch and Andrew Ashworth , 161167. Oxford: Hart Publishing.
  • Duff, R. Antony 1999Response to von Hirsch.” In Punishment and Political Theory, edited by Matt Matravers , 8387. Oxford: Hart Publishing.
  • Duff, R. Antony 2001 Punishment, Communication and Community. Oxford: Oxford University Press.
  • Duff, R. Antony 2002aRestorative Punishment and Punitive Restoration.” In Restorative Justice and the Law, edited by Lode Walgrave , 82100. Cullompton: Willan Publishing.
  • Duff, R. Antony 2002bVirtue, Vice, and Criminal Liability: Do We Want an Aristotelian Criminal Law?Buffalo Criminal Law Review 6.1 (April): 147184.
  • Duff, R. Antony 2003aRestoration and Retribution.” In Restorative Justice and Criminal Justice: Competing or Reconcilable Paradigms, edited by A. Von Hirsch , J. Roberts , A. E. Bottoms , K. Roach and M. Schiff , 4359. Portland, Oreg.: Hart Publishing.
  • Duff, R. Antony 2003bPenance, Punishment and the Limits of Community.” Punishment and Society 5 (July): 295312.
  • Duff, R. Antony 2003cThe Limits of Virtue Jurisprudence.” Metaphilosophy 34.1–2 (January): 214224.
  • Duff, R. Antony 2007aThe Intrusion of Mercy.” Ohio State Journal of Criminal Law 4.2 (Spring): 361387.
  • Duff, R. Antony 2007b Answering for Crime: Responsibility and Liability in the Criminal Law. Portland, Oreg.: Hart Publishing.
  • Duff, R. Antony 2009Can We Punish the Perpetrators of Atrocities?” In The Religious in Responses to Mass Atrocity: Interdisciplinary Perspectives, edited by Thomas Brudholm and Thomas Cushman , 79104. Cambridge: Cambridge University Press.
  • Duff, R. Antony 2011Mercy.” In The Oxford Handbook of Philosophy of Criminal Law, edited by John Deigh , and David Dolinko , 467492. Oxford: Oxford University Press
  • Duff, R. Antony 2012Towards a Modest Legal Moralism.” Criminal Law and Philosophy Vol. 8. Minnesota Legal Studies Research Paper No. 12–28.
  • Duff, R. Antony and Sandra E. Marshall 2010Public and Private Wrongs.” In Essays in Criminal Law in Honour of Sir Gerald Gordon, edited by James Chalmers , Fiona Leverick , and Lindsay Farmer , 7085. Edinburgh: Edinburgh University Press.
  • Feinberg, Joel 1970The Expressive Function of Punishment.” In Doing and Deserving: Essays in the Theory of Responsibility, 95118. Princeton, N.J.: Princeton University Press.
  • Fiddes, Paul S. 1989 Past Event and Present Salvation: The Christian Idea of Atonement. London: Darton, Longman & Todd.
  • Forrester, Duncan B. 1997 Christian Justice and Public Policy. Cambridge: Cambridge University Press.
  • Garvey, Stephen P. 2004Is it Wrong to Commute Death Row? Retribution, Atonement and Mercy.” North Carolina Law Review 82.4 (May): 13191343.
  • Garvey, Stephen P. 2007Questions of Mercy.” Ohio State Journal of Criminal Law 4.2 (Spring): 321327.
  • Gorringe, Timothy 1996 God's Just Vengeance: Crime, Violence and the Rhetoric of Salvation. Cambridge: Cambridge University Press.
  • Gorringe, Timothy 2004 Crime. London: SPCK.
  • De Grijs F. J. A. 1998Thomas Aquinas on Ira as a Divine Metaphor.” In Tibi Soli Peccavi: Thomas Aquinas on Guilt and Forgiveness, edited by Henk J. M. Schoot , 1946. Utrecht: Peeters.
  • Hampton, Jean 1984The Moral Education Theory of Punishment.” Philosophy and Public Affairs 13.3 (Summer): 208238.
  • Hanson, Anthony T. 2010 The Wrath of the Lamb. Eugene, Oreg.: Wipf & Stock.
  • Hauerwas, Stanley 1983 The Peaceable Kingdom. Notre Dame, Ind.: Notre Dame University Press.
  • Hauerwas, Stanley 1985 Character and the Christian Life: A Study in Theological Ethics. San Antonio, Tx.: Trinity University Press.
  • Kant, Immanuel 1998 Religion Within the Boundaries of Mere Reason. 1793. In Religion Within the Boundaries of Mere Reason and Other Writings, edited and translated by Allen Wood and George Di Giovanni, 31192. Cambridge: Cambridge University Press.
  • Kelly, Henry A. 2008Penitential Theology and Law at the Turn of the Fifteenth Century.” In A New History of Penance, edited by Abigail Firey , 239318. Leiden: Brill.
  • Luijten, Eric 2003 Sacramental Forgiveness as a Gift from God: Thomas Aquinas on the Sacrament of Penance. Nijmegen: Peeters.
  • Meilaender, Gilbert 1987 The Limits of Love: Some Theological Explorations. University Park, Penn.: Pennsylvania State University Press.
  • Moberly, Robert C. 1909 Atonement and Personality. London: Murray.
  • Morris, Herbert 1981A Paternalistic Theory of Punishment.” American Philosophical Quarterly 18.4 (October): 263271.
  • Montaigne, Michel de 2004On Repenting.” In The Complete Essays, translated by Michael A. Screech, 907921. London: Penguin.
  • Moore, Michael A. 1997 Placing Blame: A Theory of the Criminal Law. Oxford: Oxford University Press.
  • Murphy, Jeffrie G. and Jean Hampton 1988 Forgiveness and Mercy. Cambridge: Cambridge University Press.
  • Murphy, Jeffrie G. 1999Moral Epistemology, the Retributive Emotions, and the ‘Clumsy Moral Philosophy’ of Jesus Christ.” In The Passions of Law, edited by Susan A. Bandes , 149167. New York: New York University Press.
  • Murphy, Jeffrie G. 2003Christianity and Criminal Punishment.” Punishment and Society 5.3 (July): 261277.
  • O'Donovan, Oliver 1998The Death Penalty in Evangelium Vitae.” In Ecumenical Ventures in Ethics: Protestants Engage Pope John Paul II's Moral Encyclicals, edited by Reinhard. Hütter and Theodor Dieter , 216236. Grand Rapids, Mich.: Eerdmans.
  • O'Donovan, Oliver 2005 The Ways of Judgment. Grand Rapids, Mich.: Eerdmans.
  • Primoratz, Igor 1989Punishment as Language.” Philosophy 64.248 (April): 187205.
  • Reiff, Mark R. and Rowan Cruft 2011Antony Duff and the Philosophy of Punishment.” See Cruft, Kramer, and Reiff 2011, 324.
  • Reno, Russell R. 2002 Redemptive Change: Atonement and the Christian Cure of the Soul. Harrisburg, Penn.: Trinity Press.
  • Ricoeur, Paul 1974 Conflict of Interpretation: Essays in Hermeneutics. Evanston, Ill.: Northwestern University Press.
  • Skotnicki, Andrew 2006aGod's Prisoners: Penal Confinement and the Creation of Purgatory.” Modern Theology 22.1 (January): 85110.
  • Skotnicki, Andrew 2006bHow is Justice Restored?Studies in Christian Ethics 19.2 (August): 187204.
  • Skotnicki, Andrew 2008 Criminal Justice and the Catholic Church. Lanham, Md.: Rowman & Littlefield.
  • Slaughter, Marty 2010Sublime Mercy.” Law, Culture and the Humanities 6.3 (October): 436454.
  • Tasioulas, John 2007Repentance and the Liberal State.” Ohio State Journal of Criminal Law 4.2 (Spring): 487521.
  • Tasioulas, John 2011Where Is the Love? The Topography of Mercy.” See Cruft, Kramer, and Reiff 2011, 3753.
  • Taylor, Chloe 2009 The Culture of Confession from Augustine to Foucault: A Genealogy of the ‘Confessing Animal’. London: Taylor & Francis.
  • Travis, Stephen H. 2008 Christ and the Judgment of God: The Limits of Divine Retribution in New Testament Thought. Peabody, Mass.: Hendrickson.
  • Von Hirsch, Andrew 1993 Censure and Sanctions. Oxford: Oxford University Press.
  • Von Hirsch, Andrew and Andrew Ashworth 2005 Proportionate Sentencing: Exploring the Principles. Oxford: Oxford University Press.
  • Yoder, John Howard 2002 Preface to Theology: Christology and Theological Method. Grand Rapids, Mich.: Brazos.