“ABOVE THE SCEPTRED SWAY”: RETRIEVING THE QUALITY OF MERCY

Mercy is often thought to be a praiseworthy moral virtue. However, the quality of mercy has been thrown into doubt. Since to be merciful is to be lenient to a wrongdoer beyond the dictates of justice, it seems as if mercy cannot be a praiseworthy virtue. I argue that several recent attempts to reconcile mercy with justice fail, and that the subsequent endeavour to resurrect mercy as a praiseworthy moral virtue is doubtful. However, I also show how the praiseworthy quality of mercy can be retrieved, if we carefully reconstruct the moral case in its favour outside the law.


. Introduction
The status of mercy as a praiseworthy moral virtue is in doubt. 1 In the criminal law, it has been traditionally thought that the morally virtuous judge will, in some cases of punishment, temper justice with mercy. Mercy is supposed to be a praiseworthy moral virtue -it is good for a judge to be merciful. Justice, on the other hand, is supposed to be a central moral consideration -a judge is morally obligated to be just. In cases of punishment for wrongdoing, however, there is a prima facie conflict between mercy and justice. To be just is to exact a penalty in due proportion to the crime committed, whereas to be merciful is to exact a penalty less severe than justice dictates. Several attempts have been made to reconcile justice and mercy in the criminal law. 2 It may be, however, that mercy and justice are in fact irreconcilable in this context. In this paper I examine the case for mercy's status as a praiseworthy moral virtue, with the aim of determining in what context mercy can be regarded as genuinely praiseworthy. 3 The idea of justice in retribution is one of balance, or due proportion, between the severity of wrong done and the severity of punishment. Martha Nussbaum (1993, pp. 88-89) and A.T. Nuyen (1994, pp. 61-62) both refer to the ancient Greek conception of balance and equality in the "natural process" as being the basis of the concept of retributive justice. In Aristotle's view, "What the judge aims at doing is to make the parts equal by the penalty he imposes, whereby he takes from the aggressor any gain he may have secured" (NE 1132a 6-7). The wrongdoer has caused an imbalance in the natural order of things, and it is the obligation of the judge to restore the natural balance by inflicting an appropriate punishment on the wrongdoer; according to Aristotle, "What the judge does is to restore equality" (NE 1132a 17). In this way, the imbalance caused by the wrongdoer in his own favour is negated by the appropriate punishment. A straightforward contemporary understanding of justice might deem that the severity of punishment a just judge imposes on the wrongdoer answers to the severity of wrong done. The defining obligation of the judge in cases of wrongdoing and punishment is to be just, and in being just to inflict the severity of punishment that the wrongdoer deserves. The idea of mercy in retributive punishment, on the other hand, is one of leniency toward the wrongdoer going beyond due proportion. The merciful judge determines a just punishment, and instead of inflicting it decides to inflict a less severe penalty. Nussbaum refers to the Roman Stoic Seneca, who develops the concept of mercy as "that which turns its course away this [lenient] side of that which could be justly determined" (Nussbaum 1993, p. 102). At the core of the concept of mercy are the requirements that (1) the merciful agent is under no moral obligation to be lenient, and (2) that the wrongdoer has no moral claim to leniency. 4 The primary OED definition of mercy is consistent with these requirements, although it does not explicitly state that the merciful 2 See, for example, Brien 1998. 3 In my own view, the paradigm case of praiseworthy mercy emerges outside legal contexts, in cases of negligently self-inflicted suffering. I argue for this conclusion in section 4: "Mercy Outside the Law". 4 Based on the formulation of A.T. Nuyen (1994, p. 66). agent is under no obligation, but rather that the recipient is in his power: "1. a. Forbearance and compassion shown by one person to another who is in his power and who has no claim to receive kindness; kind and compassionate treatment in a case where severity is merited or expected." Thus, a further requirement of merciful agency suggests itself: (3) that the recipient of mercy is in the power of the merciful agent. 5 The institutional power of judge over wrongdoer in the criminal law is the paradigm case of this power relation in most discussions of mercy -as the use of the word 'leniency' in the prior two conditions hints-although it is by no means the only context in which mercy can be shown. The concept of mercy covers a far broader set of cases outside the law, as I shall discuss later. However, it is especially crucial to resist the temptation to generalise features belonging to the restricted institutional role of judges in criminal law cases to all merciful agency. Merciful agents who are not judges themselves, it ought to be noted, are not necessarily bound by the judge's institutional role or duty, which includes determining and then taking some corrective or punitive action towards wrongdoers. The power that the merciful agent holds is the power to relieve -or to refrain from inflicting-suffering, which is conceptually distinct from the power of determining and enforcing the punishment that wrongdoers should suffer. The dominant example of the criminal law judge in discussions of mercy tends to blur this important distinction. There is also a secondary concept of mercy, which applies to a power relation between persons in the absence of, or apart from, any moral considerations at all: "5. a. The clemency or forbearance of a conqueror or absolute lord, which it is in his power to extend or withhold as he thinks fit. . . ; b. at mercy. . . absolutely in the power of a victor or superior, at his disposal; liable to punishment or hurt at the hands of another. . . ." This secondary concept -I shall refer to it as mercy*-is easily confused with the original concept, although it is important to recognise some of the differences between the two. Under the secondary definition, the merciful* agent may well be morally obligated to treat his victim with compassion, and the recipient may well be morally entitled to lenient treatment. Indeed, a certain kind of qualified moral decency -well short of praiseworthiness-can be observed here: merciful* treatment often prevents or avoids immoral harm (i.e. the immoral harm that the merciful agent chose not to inflict). A captor who has it in his power to torture or even kill his captive, and yet who relents, shows mercy*, even when the captivity itself is immoral. Moreover, justice and mercy* within the criminal law are easily reconciled: to be both merciful* and just, the judge is simply required to be just, despite having the power to inflict a range of more severe -and unjustpunishments. This secondary concept of mercy*, however, isn't the concept that aspires to moral praiseworthiness, and nor is it in serious tension with the judge's obligation to treat wrongdoers justly.
When operating in the context of crime and punishment, the original concept of mercy is in prima facie conflict with the concept of justice. To formulate the conflict in terms of desert and obligations: (1) A person guilty of wrongdoing deserves a certain severity of punishment.
(2) To administer the deserved severity of punishment is to be just.
(3) To fail to administer the deserved severity of punishment is to be unjust.
(4) To administer a less severe punishment than is deserved, whilst under no obligation to do so, is to be merciful.
(5) However, to administer a less severe punishment than is deserved is to fail to administer the deserved severity of punishment, and is therefore to be unjust.
(6) Therefore, to be merciful is to be unjust.
Since a judge is obligated to administer justice -what the wrongdoer deserves for his wrongdoing-and since mercy is "conceptually exclusive of rights and obligations", it seems that mercy, in this context, is not a moral virtue at all (Nuyen 1994, p. 66). If a judge were to show mercy she would manifest injustice, and, since to administer justice is the defining moral obligation of the judge, she would be morally blameworthy rather than praiseworthy. Despite this apparently insurmountable problem, several recent writers have attempted to reconcile justice and mercy as virtues of judges in cases of crime and punishment. It will serve our purpose here briefly to revisit those arguments, so as to reconstruct the existing case against mercy. From this basis the case for mercy can then be examined outside the "criminal law paradigm" (Murphy and Hampton 1988) with the aim of determining whether and where mercy is a praiseworthy moral virtue.

. Mercy and the Criminal Law
H. Scott Hestevold (Hestevold 1983(Hestevold , 1985 proposes a solution to the conflict between justice and mercy using the concepts of "supererogation" and "disjunctive desert". According to Hestevold, mercy should be considered as a supererogatory moral virtue, one that is good to perform but not obligatory: mercy "involves the supererogatory tempering of deserved suffering", whereas, "justice involves the obligatory administering of deserved suffering for an agent's having performed a wrongful action" (Hestevold 1985, p. 281). The concept of "disjunctive desert" reconciles the two in cases of crime and punishment when the judge has the option of administering a disjunction of justly deserved punishments, but chooses to administer (one of) the less severe. Hestevold's solution seems to dissolve the prima facie conflict between justice and mercy: "Such a solution to the problem implies that the merciful person is one who acts justly, since he levies a penalty deserved; yet, he is also praiseworthy for his supererogatory choosing of one of the less severe penalties" (Hestevold 1985, p. 281).
Yet Hestevold's proposed solution fails on two fronts. Both responses to Hestevold highlight the fact that there ought to be a single deserved severity of punishment for a particular improper act. The first response points out that, in criminal punishment, it is the deserved severity of punishment that is at issue in the first place. A punishment of the wrongful agent is called for, and what is asked of the judge is what severity of punishment is appropriate or just for the wrongful act -not what range of variously severe punishments would be appropriate. Essentially, this response points out that there is no disjunction of appropriate yet varyingly severe punishments. There is no possible case where "the sufficient deserts may be of varying severity", since it is the deserved severity of punishment that is actually being decided in the first place (see, for example, Brien 1991). The second response criticises Hestevold's proposed solution from another direction. If the merciful penalty in Hestevold's disjunctive desert were held to be of the appropriate subjective severity -i.e. if it were held to be justly severe for that specific wrongdoer-then the alternative and more subjectively severe penalty in the disjunction of penalties would in fact be unjust, since it would be more subjectively severe than was justly deserved. To levy the more severe penalty, according to Nuyen, would be morally perverse (Nuyen 1994, p. 69). Furthermore, to levy the subjectively less severe penalty turns out to be obligatory rather than supererogatory, ruling out the question of mercy. Alwynne Smart, after first excluding some cases of justice that have been traditionally misnamed, proposes a different solution (Smart 1968). First, Smart identifies cases where the law is not sufficiently sophisticated to deal with significantly dissimilar cases and can therefore be unjustly harsh: "Where the law made no provision for this sort of difference a fair judge would exercise mercy; a judge who didn't would be regarded as unjust" (Smart 1968, p. 346). Furthermore, if the law cannot include in its deliberation on just penalties relevant facts about a particular case which make the improper act much less serious than it at first appears, then a merciful judge, it is commonly held, is one who takes these circumstances into account and levies a less severe penalty than the law prescribes (Smart 1968, p. 347). Smart points out, however, that even such special cases are in fact cases of justice, not mercy. These are cases of what I have called mercy*. A completely just law would need to "provide for every possible gradation of a crime", and since framing such a complicated and nuanced law would be practically impossible, "exceptions are provided for by the exercising of 'mercy' ", which, however, is really an application of justice (Smart 1968, p. 347). This kind of judgement is obviously a more sophisticated form of justice, a "better" justice, which renders the more general application of the law unjust. In these cases, genuine mercy is not being considered as an alternative to the just course of action, for "the judge who rigorously applies the law and declines to exercise 'mercy' is not being just as opposed to merciful, but is unjust" (Smart, 1968, p. 349).
Nevertheless, Smart does think there are cases of "genuine mercy". She identifies what a judge is doing in exercising genuine mercy: this involves acknowledging that an offence has been committed, deciding that a particular punishment is the appropriate or just one, and then deciding to impose a less severe penalty (Smart 1968, p. 350). In Smart's view, the appropriateness or otherwise of the exercise of mercy depends on the claims other moral duties have on the merciful judge. If the exercise of mercy "causes the suffering of an innocent party, is detrimental to the offender's welfare, harms the authority of the law, or where it is clear that the offender is not repentant or not likely to reform", or where "it involves unfair discrimination against others" (others with relevantly similar cases who are not shown mercy), mercy is inappropriate (Smart 1968, pp. 350-351). If, on the other hand, the exercise of mercy involves avoiding the suffering of innocent parties, or the supervision of the punishment would place intolerable burdens on a third party, or where it must be exercised to protect the authority and stability of the law, then it is appropriate (Smart 1968, pp. 353-354). Thus, "the answer to the original question 'when are we justified in being merciful?' must be: only when we are compelled to be by the claims that other obligations have on us" (Smart 1968, pp. 358-359).
Smart's answer, however, appears to violate one of mercy's conditions. In cases of "genuine mercy", the judge is obligated to show mercy to the wrongdoer. One of the main conditions for the exercise of mercy, however, is that the merciful judge is under no obligation to exercise mercy. It could be argued that there are two relevantly different types of obligation at play here, and that the no-obligation condition for mercy refers specifically and exclusively to the wrongdoer. This would mean that, although the judge is under no obligation to the wrongdoer to show mercy, there would be a legitimate sense in which a judge is obligated to show mercy anyway. On this view, the judge is under no obligation specifically to the wrongdoer, has been lenient due to a different type of obligation, and therefore has exercised mercy. However, the no-obligation condition unequivocally refers specifically to the judge. Mercy occurs where the judge is under no obligation, which means she has no obligation from any direction to be lenient. This is what Nuyen means when he says that "the merciful agent has complete discretion" (Nuyen 1994, p. 68). An appeal to the distinction between types of moral obligation to which putatively merciful judges are subject -if there is one to be madeseems misplaced. Smart's "genuine cases" of mercy turn out really to be cases of obligatory leniency. As Claudia Card observes, if on Smart's view mercy is only justified when we are obligated anyway, then it cannot be mercy (Card 1972, p. 184).
Card herself contends that we require an independent moral basis for mercy, and sets forth a principle based on the intuition that some offenders deserve punishment whilst others deserve mercy: the offender "deserves mercy" when otherwise he would be made to suffer "unusually more" than he deserves "in view of his basic character", or would come off worse "in this respect" when compared to those who exercise the right to have him punished (Card 1972, p. 184). However, Card's solution also seems to have severe problems. The second condition for the exercise of mercy has been violated. 6 Card's distinction between the offender's action and his "basic character", however, merits closer attention. Card implies that a certain type of character -a good person who has committed an out-of-character bad act-suffers more harshly under the same punishment than a bad character who has committed the same offence.
If this distinction is valid, then what a judge is doing in exercising mercy is recognising that a bad act does not always reflect a bad character, and tempering the severity of punishment accordingly. The judge, it seems, exacts a merciful penalty that is deserved by the basically good person, independently of or despite the bad act. Having two deserts, one for the act and one for "basic character", if the basic character is good where the act is bad, the merciful judge is free to respond in some way to the basic character's desert. However, this qualification seems to respond to a broader or richer conception of justice, rather than to the concept of mercy. 7 In any case, granting the distinction between character and act, Card's solution fails along the same lines as Hestevold's. Presumably, there is really only one severity of punishment for a crime, and since the more severe or "justly" harsh punishment would actually be perverse given the particular circumstances, the judge is obligated to administer the less severe punishment. This moral obligation, once again, disqualifies the case from being one of mercy (although it is, again, clearly a case for the exercise of mercy*).
A broader interpretation of the demands of justice in law lies at the heart of Martha Nussbaum's definition and defence of mercy. She links justice to mercy through the ancient Greek moral concept of epieikeia. Nussbaum claims that epieikeia unifies the concepts of equity, which is a superior form of justice, and mercy. This unification of justice and mercy through epieikeia involves, according to Nussbaum, "the ability to judge in such a way as to respond with sensitivity to all the particulars of a person and situation, and the 'inclination of the mind' toward leniency in punishing -equity and mercy". Nussbaum thus claims a link between situational appropriateness in just moral judgement and leniency, so that "Epieikeia, which originally designated the former, is therefore said to be accompanied by the latter: it is something mild and gentle, something contrasted to the rigid or harsh" (Nussbaum 1993, pp. 85-86).
The concept of epieikeia does indeed appear in Aristotle's Nicomachean Ethics, not as a separate concept, but as a superior form of justice. Aristotle introduces it by remarking on an apparent inconsistency: it seems strange that epieikeia is to be commended, yet that it is different from justice and that justice is also to be commended. That this could be the case in one and the same judgement certainly seems to be inconsistent. However, Aristotle explains that there is no real inconsistency, because "Equity [epieikeia], though a higher thing than one form of justice, is itself just and is not generically different from justice. Thus, so far as both are good, they coincide, though equity is to be preferred" (Aristotle NE 1137b 4-6). This solution is explained through a distinction Aristotle makes between the fallible generality of the law and the specific features of particular cases: What puzzles people is the fact that equity (epieikeia), though just, is not the justice of the law courts but a method of restoring the balance of justice when it has been tilted by the law. The need for such a rectification arises from the circumstance that law can do no more than generalise, and there are cases which cannot be settled by a general statement. So in matters where it is necessary to make a general statement, and yet that statement cannot exclude the possibility of error, the law takes no account of particular cases, though well aware that this is not a strictly correct proceeding. (Aristotle NE 1137b 7-10) Judges are apt to err in some cases -due to the poorly fitting generality of the law-and in these cases to strictly follow the law would be unjust. Equitable judges, by taking careful note of the details of particular cases, determine what is just when "the balance of justice [ . . . ] has been tilted by law", and act to correct the imbalance. Equity is thus essentially the same as justice -the defining moral obligation of the judge-and only arises as a separate consideration when the strict application of the law is unjust and requires correction. Equity is not, as Aristotle is at pains to make clear, a new or different concept, but is justice itself.
Aristotle's explanation of the relationship between epieikeia and strict legal "justice" does not, however, include the additional claim of Nussbaum's, that epieikeia is necessarily connected with leniency. Aristotle does oppose justice as equity to the rigidity or harshness of the law in the sense of its generality, but equitable judgment does not seem to imply for Aristotle a judgment that will always be lenient towards the wrongdoer; it only implies a genuinely just judgment, whichever way the balance has been tilted by the law. Reading Aristotle's formulation of epieikeia, it seems that reasons could be found in the particular case which might make the wrongdoer's crime actually seem worse than the general statement of the law allows, and thus that epieikeia would in some cases dictate a harsher penalty to restore the balance of justice.
Nussbaum acknowledges this point (Nussbaum 1993, p. 87). She argues, however, that the nature of retributive justice dictates that considerations of equity will never in fact demand a harsher penalty. Referring to the retributive idea of the ancient Greek philosopher Anaximander, Nussbaum traces the idea of retributive justice, or dikê, back to a harsh, universal ideal of "natural balance". Nussbaum claims that this harsh "retributive idea" derives from the moral and legal realm: "It is the idea that for encroachment and pain inflicted a compensating pain and encroachment must be performed" (Nussbaum 1993, p. 88). This historically "remarkably constant" retributive idea "is committed to a certain neglect of the particulars" -the neglect of who precisely can be identified as the wrongdoer, of whether the wrongful act was deliberate, of whether there were extenuating circumstances, and so on. The retributive idea often leads, according to Nussbaum, to the punishment of people -identified as wrongdoers because of some shared characteristic-who are in fact less deserving of punishment than the harsh generality of dikê will allow (Nussbaum 1993, pp. 89-90).
The true facts of the case are often ignored -there is a certain offence, which has caused a certain amount of "encroachment", and someone must receive the corresponding "encroachment" for justice to be done. No extenuating circumstances will be considered. In the case of Oedipus, "dikê says that parricide and incest have occurred here, and the balance must be righted" -retributive justice treats him "as a true or voluntary parricide would be treated, and crucial facts about him, about his good character, innocent motives and fine intentions, are neglected". Here, Nussbaum claims there is "substitution again, though of a more subtle sort, neglecting crucial elements of the person's individual identity" (Nussbaum 1993, p. 90). There is a clear need in moral theory, then, to improve the idea of justice to meet the moral demand for particularised judgement. Hence, according to Nussbaum, Aristotle's construction of a new and better conception of justice, epieikeia, to incorporate the insights of equity. And, according to Nussbaum's argument above, epieikeia will be more likely to be lenient towards the wrongdoer than dikê.
However, Nussbaum believes that the insights that epieikeia affords lead the tradition not only to a more particularised and thus inevitably more lenient judgement, but also to a more sympathetic judgement -an attitude which is not apparent in Aristotle. On Nussbaum's view, because attention to the particulars of the case lead toward extenuation or mitigation far more frequently than otherwise, and never to aggravation, inevitably -and admirably-an attitude develops in the tradition that sympathy for the wrongdoer is an appropriate judgemental attitude. Nussbaum supports this claim by appealing to Aristotle's concept of suggnômê, or "judging with", in the Rhetoric. According to Nussbaum: "He links this ability with particular perception, and both of these with the ability to classify actions in accordance with the agent's motives and intentions ([Rhet.] 1374b2-10)." Aristotle's point, according to Nussbaum, is that, "One must [ . . . ] see things from that person's [the wrongdoer's] point of view, for only then will one begin to comprehend what obstacles that person faced as he or she acted" (Nussbaum 1993, p. 94).
Even if this is so, however, Nussbaum has not yet arrived at the point where justice -in the guise of equity-unifies with mercy. A sympathetic "judging with" is used to help in the search for a punishment of due proportion -justice-but not for the purpose of leniency. In fact, Nussbaum admits that she has so far only constructed a more nuanced case for justice (Nussbaum 1993, p. 96). Thus, one can endorse Nussbaum's account so far, yet still hold to the view that a genuine case of justified -just-mercy cannot occur in the criminal law. In fact, it is arguable that Nussbaum's discussion gives us more reasons, not fewer, to suppose that cases of alleged mercy are not mercy at all, but justice -justice as equity. It also reinforces the view that mercy itself may not be an important moral virtue at all, since justice already covers sympathy and particularity in judgement as far as they are morally warranted, under equity. Nussbaum admits that so far, from her discussion of ancient Greek thought on crime and punishment, "In effect, we are given a more precise classification of offences [ . . . ]. But once a particular offence is correctly classified, the offender is punished exactly in proportion to the actual offence." This at first may seem merciful, especially in contrast to "the archaic conception of justice", but it is in fact simply just. What we mean by mercy, then, must "involve a gentleness going beyond due proportion, even to the deliberate offender" (Nussbaum 1993, p. 97). Nussbaum seems to be saying that the inclination of the mind towards leniency, which results in gentleness going beyond due proportion, is an inevitable, and morally praiseworthy, extension of the sympathetic equitable ideal in the criminal law.
Nussbaum's argument for the unification of justice and mercy rests upon the idea of a "literary judge", one who closely reads each case as if reading a novel. Extending from the particularity and sympathy inherent in the equitable treatment of wrongdoers, Nussbaum claims that a "literary" moral attitude will develop toward a humanity that calls for a reconstruction of the judge's personal morality (Nussbaum 1993, p. 103). The merciful judge will respond to both the particularity of the case and the general plight of humanity, and will furthermore reject the retributive attitude itself (Nussbaum 1993, p. 103).
In making these claims about the retributive attitude and the role of the judge in punishing wrongdoing, however, Nussbaum ignores crucial facts about the nature of the law. Her claims rest, in my own view, upon a misconception about the institutional role a judge is in fact playing in cases of crime and punishment. The judge, qua institutional representative, has no moral grounds to act upon a "new attitude to the self" -this is precisely why the equitable judge is not free to go beyond "due proportion" in determining appropriate punishments. And this is why equity and mercy cannot, as Nussbaum wishes to argue, be unified in the law. Judges are morally restricted -moreover, are morally definedby the institutional role they inhabit, and that role is defined in light of the concept of justice. It is not open to the judge in a criminal case to represent her own attitude to herself, nor to extrapolate from that to an attitude of sympathetic leniency toward the wrongdoer, and thence to a punishment that is biased -as far as justice is concerned-in favour of the wrongdoer. The judge's overriding moral responsibility is to justice. In particular cases, where circumstances determine that what the law recommends is unjustly harsh, the defining obligation of the judge is still to justice itself, conceived as equity or epieikeia. The judge has the moral responsibility, in determining just punishment, to settle upon a penalty that resolves the imbalance caused by the wrongdoer's actions, and if necessary to correct any imbalance in the law, but this is not the same as to sympathetically tilt the balance in the wrongdoer's favour.
For a judge in these circumstances to use Nussbaum's favoured attitude -one of personal sympathy consciously going beyond equity-would arguably be both unprofessional and morally perverse. The judge is under no obligation to exercise mercy and is, moreover, obligated by the moral claim for appropriate balance in punishment -justice as equity-not to be merciful. Obviously, to do something you are morally obligated by the claims of justice not to do, cannot be considered morally virtuous. In my own view, the very idea of going "beyond due proportion" in favour of wrongdoers is as perverse and unjust as going "beyond due proportion" in the opposite direction. If the point of the criminal law as far as punishment is concerned is to find a sensitive and particularised balance, and if in Aristotle's epieikeia we have found such a balance, then Nussbaum's mercy -by creating a new imbalance-directly opposes justice. In other words, Nussbaum's proposal overcompensates for the harshness of dikê, and in overcompensating topples into fresh injustice. Her unification of equity and mercy fails because the searching particularity of equity finally finds a balance, only for overweening merciful agency to overturn that balance. So, although Nussbaum has found a situation in which a judge is not obligated to be lenient, and can therefore be said to be merciful, it turns out that the judge is in fact obligated by considerations of justice not to be lenient in the first place.
Nussbaum's argument fails to reconcile mercy with justice; rather, it favours mercy over justice. The ultimate failure of Nussbaum's case strongly suggests us that mercy and justice cannot be reconciled in the context of criminal law. The failure of Nussbaum's case also suggests, I think, that mercy in this context is not a praiseworthy moral virtue at all. Justice is the defining moral consideration in cases of wrongdoing and punishment, and there is no independent conceptual space for mercy to occupy as a virtue, especially given the practical versatility of Aristotle's concepts of epieikeia and suggnômê. In so far as the suffering of a wrongdoer is to be tempered, it is a virtue to pay sympathetic attention to the particular situation of the wrongdoer -to be equitable-but only to the extent to which it assists in determining a punishment in due proportion to the true degree of severity of the crime. As it is equity -the sophisticated alternative to the harshness of strict retribution-which fulfils this function in crime and punishment, any consideration that goes beyond equity is answering neither to the true severity of the crime nor to the defining moral obligation of the judge. Thus, mercy -which with Nussbaum's approval goes beyond equity in tempering the wrongdoer's suffering, causing fresh injustice-cannot be considered to be a praiseworthy moral virtue. Not only is mercy in crime and punishment a private response to a public matter, which is inappropriate due to the judge's defining moral role as arbiter of justice, it is also a morally bad response in that it confers inappropriate preferential treatment upon one who has already wronged. If to be just is one's defining moral obligation -if to be unjust defines what counts as morally wrongand to be merciful is to be unjust, then to be merciful is morally wrong. In this context justice and mercy are not to be reconciled, and the status of mercy as a praiseworthy moral virtue is in serious question.

. Mercy and Private Law
Is it possible, however, that mercy remains an important moral virtue as a disposition of persons (as in Nussbaum's case), but only as a personal response to a private case? Andrew Brien (1991) agrees with Nussbaum that the conflict between justice and mercy is a conflict between merciful and just dispositions internal to the agent -it is a private matter. An agent "may be disposed not only to acting mercifully but also be disposed to giving a wrongdoer his just deserts" (Brien 1991, p. 197). The merciful disposition, as we have seen from Nussbaum's account, is inappropriate in cases of crime and punishment when the agent inhabits an institutional role. However, when the case is a personal one -when the only victim of the wrongdoer's wrongdoing is the agent who has the role of punishing (or claiming restitution)-the choice of punishment then becomes a private moral response to a personal matter. Can we, then, construct a case of being merciful which doesn't turn out to be morally inappropriate?
Jeffrie Murphy (1988, p. 175) points out that the failure of mercy as a virtue so far is due to the paradigm in which it has been placed, "Thus far we have been operating within what might be called the 'criminal law paradigm' of mercy -thinking of mercy as a virtue that most typically would be manifested by a sentencing judge in a criminal case. [ . . . ] It is this paradigm [ . . . ] that is probably a failure" (Murphy 1988, p. 175). The failure of the "criminal law paradigm", according to Murphy, arises because, "A judge in a criminal case has an obligation to do justice -which means, at a minimum, an obligation to uphold the rule of law" (Murphy 1988, p. 175). Murphy goes on to outline a paradigm where mercy can be successful -the "private law paradigm". Distinct from possessing an obligation to the demands of justice and the need to "uphold the rule of law", the moral agent in the "private law paradigm", according to Murphy, "rather has [ . . . ] a right to impose harsh treatment" (Murphy 1988, p. 176). The moral agent possesses a right to exact a certain penalty, distinct from the obligation to justice demanded by the restrictive moral role of the judge in a criminal case. Thus, if a litigant in a civil suit chooses to show mercy, "he is simply waiving a right that he could in justice claim -not violating an obligation demanded by justice" (Murphy 1988, p. 176).
Murphy chooses Shakespeare's play The Merchant of Venice as his representation of mercy in the "private law paradigm". Murphy sees the resolution of a contractual dispute as the central focus of the play, where "Antonio has made a bad bargain with Shylock and, having defaulted, is contractually obligated to pay Shylock a pound of flesh" (Murphy 1988, p. 175). Portia acts as mercy's representative, asking Shylock in this case not to demand his full right, which he possesses in virtue of the contract between himself and Antonio. Shylock, because "he occupies a private role", will not violate any obligation if he chooses not to demand his pound of flesh from Antonio (Murphy 1988, p. 175). On the other hand, Antonio obviously does owe Shylock an obligation for defaulting. Shylock would not in this particular case be violating any defining moral obligation to justice -the obligation which has ruled out mercy in the criminal lawbut would instead be tempering his right as a victim of wrongdoing to exact a contractually stipulated penalty. On Murphy's argument, "the virtue of mercy is revealed when a person, out of compassion for the hard position of the person who owes him an obligation, waives the right that generates the obligation and frees the individual of the burden of that obligation" (Murphy 1988, p. 176).
This solution resonates with much of what has been argued by others above. And Murphy's solution appears to allow us the freedom, in the "private law paradigm", to praise mercy as a moral virtue, as we had originally wished. The arguments in favour of mercy rest on the intuition, brought out by Antonio and Shylock's contractual dispute in The Merchant of Venice, that "People who are always standing on their rights, indifferent to the impact that may have on others, are simply intolerable" (Murphy 1988, p. 176).
Have we arrived at a solution at last? Are we free to praise mercy as an important moral virtue again, now that we have placed it in a context outside the criminal law? A closer examination of the case in The Merchant of Venice -this time from a moral rather than a legal perspective-leads us, I think, to the rejection of Murphy's solution. To put it briefly: putting aside his private legal claim, how could Shylock be thought to hold anything like a moral right to extract his pound of flesh from Antonio? Is it because Antonio made a binding contract, in his ignorance, the conditions of which were to result in a grossly unfair outcome? It seems obvious that the conditions of the contract did not produce a moral obligation for Antonio to acquiesce literally to its terms; and nor did it produce a moral right on Shylock's side. This looks suspiciously like a case of mercy* rather than mercy. What we should really like to say is that the contract was grossly unfair in the first place, and that Antonio made a serious mistake in agreeing to its conditions. We would also like to say that, morally, Shylock did not possess any right to demand his pound of flesh. It turns out that what we want -morally speaking-is for Shylock to receive only what he can justly claim from Antonio; and we consider Shylock morally wrong for demanding full payment. In that case, it is in fact justice -in the guise of mercy*-for which Portia pleads, rather than the concept of mercy that we are interested in. If Shylock is reasonably lenient, we should say he is simply being just and merciful*, not merciful proper. And if Shylock insists on harshness, he is being unjust (as well as merciless*).
Although the strict, institutionally defined obligation to justice of the judge in a criminal case is absent, I would argue that the same moral considerations still hold. It is not the case that as soon as an institutional obligation drops out of the picture, we suddenly have no moral obligations as agents. It may be that Murphy mistakenly identifies having a private right with having no overarching or overriding moral obligations, or, more likely, fails to distinguish between mercy proper and mercy*. The concept of a private right might imply freedom from some moral obligations in some cases, but it seems clear that Murphy's private right is not one of these. It is, rather, a legal right which it would be unreasonable, or vicious, to use. Thus, what Murphy has found seems to be a paradigm in which to fulfil your moral obligations when you have a non-moral contract-based private legal right to be vicious is a virtue. To fulfil one's moral obligations -especially when there is no institutional framework dictating that one do so-is undoubtedly a good thing. However, when we are morally obliged to be lenient, genuine mercy is ruled out: Murphy's argument falls to the same objection as Smart's.
Against the argument here, it could be said that I am in some sense mistakenly identifying morality with justice. It seems, on my view, that if justice is competing with mercy as a disposition in private matters, then justice will always triumph in moral agency, because it has a greater moral claim on private, as well as public, morality. Brien argues against this type of intuition -on Brien's view, "justice and mercy are two amongst a number of act-options. The task of the moral agent is to determine which option, on this occasion, is morally appropriate" (Brien 1991, p. 198). However, it is still difficult to see how genuine mercy could manifest itself in the private law paradigm as a morally praiseworthy virtue against the claims of justice. It has been seen that in Shylock's case in The Merchant of Venice he would, if he were to exact an appropriate penalty, be manifesting justice. If he were instead to manifest genuine mercy -to go beyond justice in Antonio's favour-then it really is difficult to see how mercy is morally praiseworthy. One would think that a close and sympathetic examination of the particulars of Antonio's case on Shylock's behalf -which is no more than justice requireswould decide the most fitting penalty, morally speaking, for him to impose. To say that "justice always trumps mercy" seems reasonable on my view, especially when the pervasive, subtle, and ameliorative qualities of justice as equity are fully recognised. Aristotle quotes approvingly the proverb that "All virtue is summed up in dealing justly" (Aristotle NE 1130a 7).
However, cases could be considered where the just penalty has been determined, and claimants, holding a moral right to redress, decide instead to exact a penalty less than they have the equity-based moral right to inflict. Since it would not be a case of injustice towards the offender, would not be a case of any moral obligation to be lenient, and would prevent or alleviate suffering, it would seem that being merciful would be genuinely praiseworthy. Yet immediately the question arises whether, morally speaking, it is a good thing for one person to give up just claims on another -especially when their welfare has already been taken into account (due to the dictates of equity). A case of this kind would arguably amount to giving the wrongdoer a free advantage, much as is the case with Nussbaum's argument, except that in this case the agent seems prima facie to be free of any obligation not to do so. That is unless, of course, agents have some basis for moral obligations to themselves in private law cases. A case where one person holds a moral right to exact a penalty decided by due proportion and equality -an equitable penaltybut decides instead to exact a penalty which is less severe to his own disadvantage, although it may be genuinely merciful, seems to instantiate a certain kind of potentially self-destructive vice.
Can it ever be unequivocally praiseworthy to be lenient beyond what is fair to oneself in private law cases, in order to be merciful? It could be argued that if a moral agent can in a certain sense "easily afford" to show mercy to another who has wronged her, then mercy could be seen to be a virtuous moral response despite the apparently self-inflicted unfairness it entails. Consider the case where a wealthy creditor waives a debt owed by a struggling borrower, although he is well within his legal rights to claim payment. According to Aristotle, who also argues that people cannot literally do themselves injustice, "This kind of renunciation is held to be a characteristic of the gentleman, whose instinct is to take less than he is entitled to" (Aristotle NE 1138a 1). This seems to be the essence of what is meant in many private law cases of mercy. When we have the discretion to impose the repayment of a debt or not, but, cognisant of the fact that enforcing our right would have bad consequences for the debtor (if e.g. it would nearly bankrupt or otherwise disproportionately disadvantage him) and, easily able to absorb the loss, we instead refrain, then it seems that we have a paradigm case of genuine and unequivocally praiseworthy merciful action. In the first place, it seems genuinely to be "up to us" as creditors whether to pursue the private case or whether to drop it, and secondly our debtor has no claim to leniency. This kind of case represents a more subtle objection. In brief, the claim is that in at least some cases of morality in private law there is scope for genuine discretion: one does oneself no serious injustice by not enforcing one's morally legitimate claims, and one prevents bad -possibly disastrous-consequences for one's debtor. The unequivocal praiseworthiness of mercy seems to lie in this: that acting mercifully costs little and prevents or alleviates significant suffering. This view has strong intuitive appeal. Aristotle's praise of those who are willingly prepared to take less than their fair share brings to mind some of our more intuitively attractive ideals: self-sufficiency, altruism, benevolence, compassion. However, I think that this view makes a twofold mistake: in overstating the scope of moral discretion, it underestimates the moral scope of equity. That we as creditors are wealthy and can easily afford to absorb losses is of some significance to determinations of equity in the case at hand; as is the fact that the debtor stands to suffer disproportionately. Indeed, "disproportion" is precisely what equitable agency aims to uncover. As in the criminal law paradigm, equitable judgement, when properly executed, takes into account any and all particular circumstances relevant to the proportionate or fair resolution of private cases. Acknowledgement of the fact that we can easily afford to absorb the loss derives its moral weight from the intuition that to demand payment would be, in this case, disproportionate -i.e. inequitable, unfair. And the unequivocal praiseworthiness of our conduct, I would argue, derives from the fact -once again-that despite our having the power (via a legal right) to inflict disproportionately harsh conditions in our own favour, we equitably refrain. The case only serves to uncover further considerations relevant to equity, not separate conditions of supererogatory moral discretion. Morally, the kinds of considerations mentioned are at bottom a matter of equity. Moral discretion in favour of a fair resolution -given that fairness equals equity in the particular case-is a chimera, and without the inadmissible intuitive appeal to fairness, this kind of objection loses much of its force. The case is one of equity again, this time in a deeper disguise.
Mercy requires both genuine moral discretion -not the discretion of mercy*-and disproportionate leniency.
Does my argument show that in the private law there is never any room for genuinely discretionary cases of mercy? On my view, it is not a matter of there being no conceptual space for mercy to inhabit. I think that there are genuine cases of mercy; perhaps mercy is even common. My point, however, is that when we reflect carefully on mercy's moral foundation in the context of the law, criminal or private, what we find, morally speaking -beside equity's broad scope-is a surprisingly dubious notion. Being merciful includes going beyond due proportion in lenient treatment of another, whilst being under no obligation to do so. Given that the concept of equity will reliably "mop up" all genuine considerations on both sides relevant to fairness, mercy's moral weight seems to derive solely from a certain kind of consequentialistic consideration. An act of mercy in the private law paradigm, on one hand, prevents or alleviates proportionately determined (i.e. fair) but subjectively bad consequences for its recipient. On the other hand, an act of mercy requires of its doer that she relinquish her fair claim to redress in order to prevent or relieve fair -but, for the recipient, subjectively bad-consequences. When mercy's consequentialistic good for the recipient is weighed against the costs to the merciful agent, and moreover to the concept of justice as fairness -assuming that these values are not somehow irretrievably incommensurable-it is difficult to see how mercy, as Portia has it, 'blesseth him that gives and him that takes'.

. Mercy Outside the Law
It seems reasonably clear that mercy is not an unequivocally praiseworthy moral virtue within the context of the law, whether criminal or private. The moral obligation of the judge to justice in the case of crime and punishment denies moral praiseworthiness to genuine mercy -it represents disproportionate leniency. In the private law paradigm, mercy is of similarly dubious moral value: when cases of mercy* have been excised, and the broad scope of equity is fully acknowledged, genuine mercy inhabits a dubious moral space. I think it is clear in the end that mercy is not a praiseworthy moral virtue in legal contexts, criminal or private. Is there, then, any case for mercy as an unequivocally praiseworthy moral virtue at all?
Elements of the discussion above point to the possibility of rescuing mercy as an important moral virtue in personal or private lives, outside any legal context. John Kleinig, criticising Smart's article, makes the salient point that "Mrs. Smart has been misled by the fact that questions of mercy often arise within the context of wrongdoing and punishment into thinking that this is the only context in which they arise" (Kleinig 1971, p. 341). It has often been overlooked "that mercy can be shown in a wider context than that of wrongdoing and punishment" (Kleinig 1969, p. 341). Kleinig considers mercy in a broader context of moral agency, "that of treating with benevolence those who are in need [ . . . ] and who have no claim to benevolent treatment. Whether the claim is one of desert or entitlement does not seem to matter" (Kleinig 1969, p. 341). Kleinig claims that mercy can be shown "to the widow, the fatherless and the destitute", and that "To show mercy to the widow, the fatherless, and the destitute is not necessarily to treat them other than they would be if treated justly" (Kleinig 1969, p. 341). In treating the unfortunate with benevolence, a moral agent is doing something morally praiseworthy -alleviating suffering-and is also doing something she is under no moral obligation to do. However, Kleinig's analysis is not quite satisfactory -the idea of mercy I endorse considers moral agents merciful and morally praiseworthy if they treat with benevolence those who really are in a certain sense undeserving of our benevolence-especially those who have in some sense negligently brought suffering upon themselves. I think this condition captures more accurately the concept we intuitively grasp, when we think of the receiver having "no moral claim" to mercy. The widow, the fatherless, and the destitute do not figure in my proposed cases of mercy, but are receivers of a different species of benevolence, in the light of which they do have "a moral claim" on our benevolent treatment. If we are in a position to offer succour to suffering innocents, then it seems that we have what Kantians call a "wide" obligation to do so: they are prime targets for our duty of benevolence. The distinction outlined here has the advantage of retaining one of the original characteristics of genuine mercy -that its recipient has no moral claim to benevolent treatment-whilst also leaving room for mercy's distinctive moral praiseworthiness to emerge.
In my view, the paradigm case of praiseworthy mercy emerges in cases of negligently self-inflicted suffering. That people are liable to bring unnecessary suffering upon themselves is a fact of life. When a person has negligently brought suffering upon himself, and there is no other victim of his actions -i.e. he has not done harm to anyone else-any obligation to justice, whether institutional or private, drops out of the picture. There is no sense in which any other moral agent holds either a moral right or responsibility to administer justice; and in at least one sense, justice has already been done. The agent suffers the natural ('just') -and subjectively badconsequences of his own negligence. The balance and due proportion of justice proper, however, only applies when someone has "tipped the balance" in favour of himself, which is clearly not the case here. Thus, the moral obligation to do justice does not apply to any putatively merciful moral agent who comes upon the scene. (Nor, on the other hand, is there any obligation not to be merciful.) The first condition for showing mercy -that the merciful agent is under no moral obligation to be merciful-is thus satisfied. Secondly, the person who has negligently inflicted suffering upon himself, since he "brought it on himself", has not the kind of moral claim on the merciful agent's aid that the innocent sufferers had. Their moral claim arose because they suffered innocently and thus undeservedly, whereas the self-inflicted sufferer suffers deservedly through his own negligence. The merciful moral agent in this kind of case is not under any moral obligation, and the recipient has no moral claim to the merciful agent's benevolent treatment.
This kind of case seems to escape the objections to mercy raised in the preceding discussion. However, two questions still remain: how, exactly, does being merciful in this context make mercy a praiseworthy moral virtue? Also, when is being merciful in this context appropriate, and when not? One possible solution to these problems lies in Nussbaum's previous analysis. I agree with Nussbaum that to determine when mercy is appropriate and morally praiseworthy is predominantly "a literary task" of reading the particulars of the case. In the private context, the merciful moral agent is free not only to "read" the details of a life, but is free to act -i.e. is not obligated not to act-on feelings of sympathy "going beyond due proportion" engendered by the details of the story. The fact that in the case of the negligent sufferer there is no other victim of his moral wrongdoing, and yet that a wrong has been done and someone is suffering, frees the autonomous moral agent from other overriding moral obligations. One possibility is that literary perception -and in particular the form of the novel-can allow the merciful moral agent to judge on the basis of good character, and to act mercifully on the basis of that judgement. Mercy would then be appropriate when the merciful moral agent finds reasons, in reading the particular case, for believing that the negligent agent is basically a good person (i.e. is fundamentally of good character). Mercy is then morally praiseworthy when it alleviates the self-inflicted suffering of someone who is found -subsequent to a careful reading of the case-to be, at bottom, a good person who has made a bad choice (or series of choices).
The connection between this genuine form of mercy and moral virtue is exemplified in Joseph Conrad's novel Lord Jim (Conrad 1957). An outline of this novel brings out what is meant by genuine mercy, the connection between literary perception and genuine mercy, and the reasons why mercy can be a truly praiseworthy moral virtue. The novel revolves around the life story of a young sailor, Jim, as told by Marlow, an older sailor who follows Jim's story and intervenes mercifully on his behalf. Jim has appeared at an inquiry into the circumstances surrounding his crew's abandonment of their ship in the middle of the ocean with several hundred passengers aboard. The ship did not sink and no passengers were killed or injured -they were rescued and towed to port. Despite there being no victims of the actions of the crew, abandoning ship is possibly the most dishonourable, cowardly act any sailor can imagine. The rest of Jim's crew escape the inquiry, and Jim is left to face the consequences on his own. He has had the opportunity to escape, but has refused. Marlow attends the inquiry, and in his narrative comments: "I became positive in my mind that the inquiry was a severe punishment to that Jim, and that his facing it -practically of his own free will-was a redeeming feature in his abominable case" (Conrad 1957, p. 57). Marlow is convinced that Jim feels the disgrace of being the public face of such a shameful incident, but that despite this self-inflicted suffering -the result of his 'abominable' act-Jim acts virtuously. This action of Jim's sparks Marlow's sympathy, and causes him to take a particular interest in Jim's plight -to find out more about Jim's history and character. He has found sympathy for Jim, and reinforcing this sympathy is Marlow's familiarity with Jim's background (Conrad 1957, p. 52).
Marlow offers Jim lodgings during the inquiry, with the aim of coming to understand just what made Jim do what he did -to discover how Jim could have failed so abominably, yet could still appear honourable. Jim recounts to Marlow the details of his disgrace -how he came to abandon his ship in the middle of the ocean to save himself. Hearing Jim tell his tale, Marlow is moved by the difficulties Jim faced, and by the suffering he inflicted on himself in choosing to act as he did. As Jim describes the moment when the life-boat drifted from the ship, Marlow is deeply moved by Jim's regret and anguish: "I believe that, in this first moment, his heart was wrung with all the suffering, that his soul knew the accumulated savour of all the fear, all the horror, all the despair of eight hundred human beings pounced upon in the night by a sudden and violent death . . . " (Conrad 1957, p. 108). Marlow can see that Jim is fundamentally a good person who has made a terrible mistake, a mistake which he immediately felt and regretted. After the official inquiry concludes -finding Jim guilty of misconduct and cancelling his sailor's certificate-Marlow finds Jim, and arranges for his employment as a water-clerk in a distant port. Marlow's first act of mercy results from his sympathy for Jim's plight and his close and particular knowledge of the affair -gleaned from both the official inquiry and Jim himself. Marlow well knows that it would be easier just to walk away from the whole situation, and certainly that it is not his own particular responsibility to act with benevolence towards Jim.
Eventually, Marlow realises that at the heart of Jim's case is the need to redeem himself -to clear his stained conscience. Jim is a good person who wishes only to do what is right and honourable. Marlow offers Jim the opportunity to redeem himself by arranging for Jim a new start in a remote part of the world, where the shame of his actions can be both forgotten and annulled. Marlow, with the help of a trader friend, sends Jim to a remote trading-post far from civilisation. It is exactly the act of benevolence Jim requires: "He left his earthly failings behind him and that sort of reputation he had, and there was a totally new set of conditions for his imaginative faculties to work upon. Entirely new, entirely remarkable. And he got hold of them in a remarkable way" (Conrad 1957, p. 191). Jim makes a great success of running the trading-post, and becomes a leader of the community. It is Marlow's act of benevolence which allows Jim to redeem himself and thus relieve himself of the burden of his self-inflicted suffering. Marlow has paid close attention to Jim's particular life-story, has found reasons to be merciful, and has acted on those reasons. The result is that Jim's suffering is alleviated -a good person, who has no moral claim, has been relieved of suffering by the benevolent actions of another who was under no obligation in the first place to act at all. Thus, an example of a genuine case of appropriate and morally praiseworthy mercy can be found in Conrad's Lord Jim. But what of cases where mercy is inappropriate, and morally blameworthy, in this context? It turns out that such a case is found in the conclusion to Lord Jim. Ironically, it is Jim himself -now admired leader of his own remote community-who misjudges a situation and is inappropriately merciful. A desperate band of starving men arrive, led by Brown. Brown is a pirate trying to escape from the authorities, and his history is a brutal one. Marlow, our narrator, recounts Brown's past deeds, of which Jim is unaware. Brown is known both for "the arrogant temper of his misdeeds and a vehement scorn for mankind at large and for his victims in particular" (Conrad 1957, p. 297). According to Marlow, such is Brown's viciousness of character that, "he would bring to the shooting or maiming of some quiet, unoffending stranger a savage and vengeful earnestness fit to terrify the most reckless of desperadoes" (Conrad 1957, p. 298). Brown's starved state and desperation is self-inflicted -his parlous state has been brought about not just through negligence, but also more extreme vice-yet Jim does not inquire in any depth into the particular circumstances of Brown's life. When Brown arrives with his crew, desperate and starving, Jim allows him food and a means of escape from his pursuers. It is obvious that Brown is not basically a good person, and that if Jim had taken the trouble to inquire into the particulars of Brown's character and brutal past, he would have reacted quite differently to Brown's plight. Jim tells his followers that, although there is no obligation to help Brown on their part, "He believed that it would be best to let these whites and their followers go with their lives. It would be a small gift" (Conrad 1957, p. 330). Mercy, in this case -though still genuine-is not a praiseworthy virtue but a blameworthy vice. Jim's judgement is badly wrong: Brown brutally murders several innocent members of Jim's community, before escaping to the open sea. Jim's standing in the community is ruined because of his poor judgement, and this leads directly to his own demise. Jim saw only the self-inflicted suffering of Brown -he had no interest in Brown (sympathetic or otherwise), nor any knowledge of Brown's previous actions, life or character. Jim had not sought to find out the particulars of Brown's case. Nevertheless, Jim had shown Brown mercy -a case where to be merciful was both inappropriate and morally blameworthy.

. Conclusion
Ultimately, the virtue of mercy is a private matter. Determining when mercy is appropriate and morally praiseworthy is -as Nussbaum argues-a literary task. But mercy is not a praiseworthy moral virtue in a legal context, be it criminal or private. What is praiseworthy turns out each time to be equity misnamed as mercy. Mercy is a praiseworthy moral virtue in private lives, typically where people negligently bring suffering upon themselves, and other people -subsequent to reading the particulars of the case-find good reasons for benevolent action. Acting on a close and sympathetic analysis of the particular case is morally praiseworthy because it alleviates the suffering of a good person. In this kind of case, mercy is the subject of neither moral claim nor obligation, yet is nonetheless a praiseworthy moral virtue. However, there is no reason why this conception of mercy as a praiseworthy moral virtue must be limited entirely to an individual's response. It seems plausible that this model of mercy can be extended to a wider social sphere, to groups in society who take the care to inquire carefully into the circumstances of people's suffering, and who find there compelling reasons to be merciful. If this were the case, it might well be said that mercy "droppeth as the gentle rain from heaven".