Printer Friendly

Why the deterrence argument for capital punishment fails.

Violence and the threat of violence are among the standard responses to situations of conflict. Often, when violence is used or threatened, an argument is provided in an effort to justify the use of violence in the given circumstance. Many such arguments, though certainly not all, point to the useful consequences of a given violent act or practice as the principal grounds for its justification. Such arguments are particularly prevalent in political thinking, wherein the practical efficacy of an act is one of the crucial issues for deciding whether the act should be done. Since violence is generally perceived to be at the very least prima facie wrong, some justification must be provided by the users of violence if they are to convince themselves and others that they are not guilty of moral wrongdoing when they do violence. It is the intent of these justifying arguments to establish that violence in the given circumstance is at the very least morally permissible, and sometimes even morally obligatory.

Given that justifying arguments are designed to provide a justification for doing something that is considered to be prima facie wrong, it makes sense to suppose that belief in the success of these arguments (on the part of both the user of violence and others) may contribute to the actual practice of violence. Thus, a careful scrutiny of such justifying arguments, particularly those that are convincing and are widely adhered to, is exceedingly important, given the objective of minimizing as much as possible the actual practice of violence.

Two circumstances in which there is widespread agreement that violence is justified are the circumstance of self-defense (in which one uses violence to prevent harm to oneself) and that of punishment (in which one uses violence on those who have performed some morally wrong, usually criminal, action). It is my intent in this paper to examine capital punishment (CP) as a paradigm example of violent action for which justifying arguments are provided. In the following analysis I will limit my focus to one field of justifying arguments, namely the arguments from deterrence. These are arguments which say that CP is justified by virtue of one of its consequences, namely the fact that it is uniquely effective in deterring persons from performing actions similar to the acts (usually murder) which are performed by those who would be subjected to CP.(1) Arguments of this sort are among the most widely used justifications for CP in political rhetoric and private thinking about CP, and therefore warrant close scrutiny.

In what follows I will provide a careful explication of deterrence-based arguments for CP; I will then show that, even if one grants that CP is a more effective deterrent than other forms of severe punishment, there are two serious problems with these deterrence arguments. The first of these is a problem simply with those arguments which support CP (although similar problems may exist with deterrence arguments supporting other forms of punishment). This problem, as we shall see, makes the deterrence arguments for CP inconclusive at best. The second problem, one which has so far escaped notice in the literature on the subject of deterrence (at least in the form presented here), may have more far-reaching implications, in that it is equally applicable to deterrence-based arguments for punishment taken as such, and not just for CP in particular. This second problem is, I think, a decisive one, and recommends the abandonment of deterrence arguments for CP, and perhaps for punishment in general.

It should be noted from the outset that those who wish to justify CP are almost universally compelled to accept that the kind of lethal violence they are advocating is at least prima facie wrong. The reason for this is clear: almost all supporters of CP wish to reserve CP for cases of murder, usually for cases of especially brutal murder or of multiple murders; and their primary reason for applying CP to cases of murder is their conviction that there is something wrong with using lethal violence against another human being. In other words, not only do those who wish to justify CP hold that lethal violence is prima facie wrong, but this conviction that lethal violence is wrong figures into their argument for CP. They use the fact that lethal violence is prima facie wrong as a premise in their argument for the justification of one particular kind of lethal violence. This fact involves the proponent of CP in an interesting dilemma; escaping this dilemma must be one of the central features of any argument defending CP.

The Deterrence Argument

Let us turn now to the deterrence argument (DA). There are many versions of DA, but they all share the common feature that they seek to justify CP on the grounds that it deters potential murderers from actually committing murder. Most recent debates concerning DA focus on the question of whether or not CP actually does deter more effectively than other forms of severe punishment (such as lengthy prison terms). These debates appeal both to statistical data and to common sense, and so far have been largely inconclusive.(2) This issue is an extremely important one in any analysis of DA, for if CP does not deter more effectively than other forms of severe punishment, then this particular justifying argument cannot hope to succeed. However significant this issue may be, I will not address it here. Rather, I will address a more theoretical claim. In particular, there seems to be a tacit assumption among many on both sides of the debate that if it could be shown that CP does serve as a more effective deterrent than do other forms of extreme punishment, this fact would automatically justify CP (this assumption may be more common in popular discourse than in philosophical essays, and is certainly more widespread among proponents of CP than among its opponents). In other words, it is assumed that there is some version of DA which is valid, and which is such that all of its premises, except the premise that CP actually serves as the most effective deterrent, are known to be true. It is my suggestion that such an assumption is false. It will be the effort of the following analysis to show that there are problems with DA aside from and in addition to the problem of the actual effectiveness of DA as a deterrent.

There are two principal versions of DA. One version focuses on the policy of CP, and attempts primarily to justify the existence of such a policy, justifying the actual practice of CP in particular instances only derivatively. Such argumetns take on the following general form:

1 Having a policy of CP in place for those who are convicted of committing murder has the effect of minimizing the actual commission of murders, either because the threat of violence deters potential murderers, or because the existence of such a policy instills a sense of abhorrence for the crime of murder in the general populace.

2 This effect could not be achieved (at least not as well) without having the policy of CP in place.

3 Given that having in place the policy of CP achieves this effect better than having in place some other policy, it is at least morally permissible to institute a policy of CP, even though there are prima facie moral considerations against instituting such a policy.

4 Therefore, instituting a policy of CP is morally permissible (1-3).

The other version of DA focuses on particular instances of CP, and seeks to justify these particular instances directly:

1 Performing CP on a convicted murderer has the effect of discouraging other potential murderers from committing murder in the future, either because they are afraid that they will also be subjected to CP, or because the act of CP instills an abhorrence for the crime of murder.

2 This effect could not be achieved (at least not as well) without performing CP on the convicted murderer.

3 Given that performing CP on a convicted murderer achieves this effect better than any other form of punishment, it is at least morally permissible to perform CP on convicted murderers, even though there are prima facie moral considerations against performing such actions.

4 Therefore, performing CP on a convicted murderer is morally permissible (1-3).

The premises which are the principal subjects of current debate are the first two premises in each version of the argument. I will focus on the third premise, since it is this premise which bridges the gap between the factual claim that CP is a better deterrent and the moral judgment that CP is permissible. Clearly, the third premise is not obviously true as it stands, and needs to be defended in some way. Thus, the success of both versions of DA will depend on the success of the subarguments for the third premise in each argument. If a satisfactory sub-argument cannot be provided, then DA will fail as a justification for CP.

For the sake of brevity, I will focus my attention on the arguments for the third premise in the second version of DA (the argument for the individual case of CP). My reason for this focus is simple. If the second version of DA fails, then the first version will succeed only if "having a policy of CP in place" is taken to involve the threat of CP but not its actual practice. For, if having in place a policy involves or presupposes that CP actually be practiced in particular circumstances, then the failure to justify the actual practice would likewise doom the policy. Thus, if the second version of DA fails, the first version could at most justify only the threat of CP, and not its actual practice (unless, of course, one combined the first version of DA with a non-deterrence justification for the actual practice, in which case we would not have a deterrence argument of the sort I want to look at here). The threat of CP, in order to be justified, would have to be an empty threat. It is unclear whether or not an empty threat could achieve the deterrent effect that is supposed to justify the policy of CP. That question aside, the failure of the second version of DA weakens the conclusion of the first version so much that it can hardly be viewed as a justification for CP in any ordinary sense. Thus, if the second version fails, CP cannot be justified on deterrence grounds alone. Henceforth, then, when I speak of DA, I will have in mind the second version.

The third premise of DA holds that CP is justified because it achieves a certain effect. In other words, it is by virtue of its consequences--in particular the consequence that murders are reduced--that CP is thought to achieve its sanction. Presumably, then, it is thought that CP is justified because its consequences are good, or at least better than the consequences of any alternative course of action. In other words, it seems that the foundation for the third premise of DA is the notion that the consequences of performing CP will be better than the consequences of any alternative action (given the truth of the preceding premises), and that it is primarily the consequences of an action which are to be considered in determining whether that act is right or wrong. In fact, it is precisely this sort of utilitarian defense of the third premise that one observes in the literature. Proponents of DA claim that there is a pair of alternatives which one must choose between when a murder has been committed, and the alternative with the better consequences involves carrying out CP on the convicted murderer. Ernest van den Haag, co-author of The Death Penalty: A Debate, outlines very well exactly what it is that proponents of DA see as the choice which must be made:

...we must choose either to 1) trade the certain death, by execution, of a convicted murderer for the probable survival of an indefinite number of murder victims whose future murder is less likely (whose survival is more likely)--if the convicted murderer's execution deters prospective murderers, as it might, or to 2) trade the certain survival of the convicted murderer for the probable loss of the lives of future murder victims more likely to be murdered because the convicted murderer's nonexecution might not deter prospective murderers, who could have been deterred by executing the convicted murderer.(4)

What the proponent of DA suggests is that the likely consequences of the first choice are better than the likely consequences of the second, and for this reason we are justified in choosing the first rather than the second, even though the first involves CP, which is an act of killing and which, in isolation, would be considered wrong.

Jeremy Bentham outlines this essentially utilitarian justification for punishment in the following passage:

General prevention ought to be the chief end of punishment as it is its real justification. If we could consider an offense which has been committed as an isolated fact, the like of which would never recur, punishment would be useless. It would only be adding one evil to another. But when we consider that an unpunished crime leaves the path of crime open, not only to the same delinquent but also to all those who may have the same motives and opportunity for entering upon it, we perceive that punishment inflicted on the individual becomes a source of security for all. That punishment which considered in itself appeared base and repugnant to all generous sentiments is elevated to the first rank of benefits when it is regarded not as an act of wrath or vengeance against a guilty or unfortunate individual who has given way to mischievous inclinations, but as an indispensable sacrifice to the common safety.(5)

What is said here is that punishment is justified when by carrying it out one promotes the common good by discouraging the future commission of crimes. What is said about punishment in general can and is extended by proponents of CP to the death penalty in particular: CP is justified when by carrying it out one promotes the common good by discouraging the future commission of murder.

We cac condense these points to form the following argument, which we will call the sub-argument SA:

1sa Assume that CP has the deterrent effect indicated in premises 1 and 2 of DA.

2sa Performing CP on a convicted murderer has certain bad consequences (namely, the death of the convict).

3sa Refraining from performing CP on a convicted murderer also has certain bad consequences (namely the lack of CP's deterrent effect)

4sa The consequences of performing CP and gaining its deterrent effect are better than the consequences of not performing CP and hence not gaining its deterrent effect.

5sa Whenever there is a choice between two courses of action, both of which have bad consequences, it is morally permissible to choose that act which has the best consequences (even though this entails certain bad consequences).

6sa Therefore, given the assumption that CP has the deterrent effect in question, it is morally permissible to perform CP on the convicted murderer despite the prima facie moral considerations against performing CP.

The conclusion to this argument, 6sa, is substantially the same as the third premise of DA. Thus, the preceding argument may be viewed as a sub-argument in defense of DA's third premise.

The Failure of the Argument

What can be said about this argument? It is my contention that there are difficulties with both premise 4sa and premise 5sa; the difficulties with 4sa render the argument (DA) inconclusive, while the problems with 5sa warrant the abandonment of DA altogether. Let us first look at the problems with 4sa. It is not entirely clear, even if the deterrent effect of CP is greater than the deterrent effect of life imprisonment or some other less severe punishment, that this difference constitutes a difference in the goodness of the consequences that outweighs the difference between the severity of CP and the severity of life imprisonment.

Let us put this another way. The evil consequences of CP are the death of the convict. The evil consequences of life imprisonment are the permanent incarceration of the convict. The evil of death presumably outweighs the evil of incarceration.(6) Thus, considered in isolation, the consequences of CP are worse than the consequences of life imprisonment. Even if we assume that CP's deterrent effect is greater than the deterrent effect of imprisonment, it is certainly possible that this difference is not enough to counterbalance the difference that exists between the severity of CP and life imprisonment taken in isolation. In order for 4sa to be true, then, it must be true not only that CP has a greater deterrent effect than life imprisonment, but that this difference is great enough to outweigh the greater evil that CP causes the convict.(7)

It is not clear how such a weighing of consequences could be accomplished at all. Given the inconclusiveness of the evidence concerning whether CP actually has a greater deterrent effect than life imprisonment, we can expect that the difference in deterrent effect is not very great even if there is a difference (I am presuming here that a great difference in deterrent effect would be less likely to lead to inconclusive evidence). Thus, in order for the proponent of DA to successfully defend the third premise of his argument, he is drawn into a morass of problems for which it does not seem any clear answers can be given.(8)

Let us suppose, however, for the sake of argument, that it could be shown that the consequences of performing CP, given that it has a unique deterrent effect, would be better than the consequences of any alternative course of action, despite the evil consequence of the convict's death. Does it follow that CP is morally permissible, given its deterrent effect?

There remains, of course, the question of the acceptability of premise 5sa, which states that whenever there is a choice between two courses of action, both of which have bad consequences, it is morally permissible to choose that act which has the best consequences. The evaluation of this premise, of course, depends to a large extent on the evaluation of consequentialist moral theory as such. I cannot in this limited context address all the issues related to such an evaluation.

It would not be inappropriate, however, to consider certain specific issues in this broader context that have a direct bearing on the morality of CP. In particular, I wish to argue that given the crucial importance of premise 5sa to the effectiveness of DA--a premise which grounds the moral justification of an act exclusively in the consequences of that act--the actual guilt or innocence of the potential CP victim (with respect to the crime in question, presumably murder) turns out not to be a necessary condition for the justification of CP. I will show in what follows that there is at least one realistic circumstance in which committing CP on an innocent person would be justified, given 5sa. And it follows from this that if we accept 5sa as a legitimate premise, we will be forced to admit the possibility not only of the justified execution of the guilty, but also of the justified execution of the innocent.(9)

An examination of certain features of DA will show why this is so. In essence, the premises of DA can be classified into two categories: those premises that attempt to show that CP is uniquely capable of achieving a certain effect (namely the effect of minimizing the total number of murders committed in the community); and those premises that attempt to show that, because this effect is achieved, CP is morally permissible. The former premises are designed to establish what I call the minor premise of the two-premise version of DA. The latter are designed to establish the major premise. Premise 5sa is the crucial premise in this latter category. DA could, in essence, be reduced to the following two-premise argument:

1 Implementing CP for those guilty of murder achieves the deterrent effect, while not implementing CP for those guilty of murder compromises that effect. (Minor Premise)

2 CP would be justified if through its commission, and only through its commission, the deterrent effect were achieved. (Major Premise)

3 Therefore, CP is justified.

This formulation of DA shows clearly the two main steps in a justification for DA framed on deterre4nce grounds. What DA must show is, first, that by performing CP in certain circumstances, and only by performing CP, the deterrent effect in question is achieved (the minor premise), and secondly, that it is morally permissible to perform CP given that it achieves this effect (the major premise). Premises 4sa and 5sa serve to support this latter position. What we see in DA is that it is the achievement of the deterrent effect that is crucial for outweighing the prima facie moral considerations against killing. The actual guilt of the intended CP victim is relevant only insofar as practicing CP on those who are not judged to be guilty will not serve the deterrent effect. Executing those not judged to be guilty will promote no deterrent effect whatever, and will rather have the evil of the killing itself compounded by the unease such an act would instill in the community.

The point to note, however, is the following: if we assume that the major premise is true, then, were there some circumstance which grounded the minor premise other than the circumstance in which the intended CP victim is in fact guilty of murder--some circumstance in which implementing CP achieves the justifying effect (and not implementing CP compromises that effect)--then CP would be justified in such a circumstance. Put another way: were it possible for CP to achieve its deterrent effect in a framework that did not include the guilt of the intended victim, CP would still be justified in such circumstances, if we assume that the major premise above is true. And this is a possibility we must accept if we accept as true premise 5sa, since premise 5sa is the central element of the major premise in the above two-premise version of DA. In short, because the actual guilt of the intended victim figures only in the minor premise of DA, and not in the major premise, it remains an open question as to whether or not there are circumstances such that the intended victim of CP is in fact not guilty, in which CP is still justified according to DA.(10)

This question can, I think, be answered; the answer is that there are cases in which DA would justify inflicting CP on an innocent person. These cases are made possible by the fact that our system for judging the guilt or innocence of accused criminals is fallible. While mistaken judgments can be reduced through careful and conservative procedures, the fallibility itself cannot be totally eliminated. And if we accept 5sa, we must accept the conclusion that the infliction of CP on those who have been mistakenly judged to be guilty, but who are in fact innocent, is not merely an excusable tragedy, but is justified and is morally laudable.

To see that this is so, consider the following: suppose it is the general opinion in a community that a certain person is guilty of multiple murders. After a careful consideration of the evidence, the person has been judged guilty in a court of law. Let us suppose, however, that the individual in question is in fact innocent. Because it is believed that the person is guilty of murder, executing this person will presumably achieve the deterrent effect (it is our assumption here, as it has been all along, that the other controversial premises in DA are true, namely that CP deters better than all other forms of severe punishment). Furthermore, not executing this person would compromise the deterrent effect since the populace would perceive this as the non-execution of a murderer. Thus, executing this person would be justified, if we accept the major premise of DA as true, even though the person is innocent. For the purposes of DA, then, the public judgment that a person is guilty is sufficient to justify CP (assuming, of course, that this judgment is made after careful and conservative procedures--the lack of such procedures preceding an execution would instill enough fear and insecurity in a population to presumably outweigh the value of the deterrent effect), whether or not the person is in fact guilty. Put another way, the actual guilt of the potential CP victim is not necessary for justifying CP.

And this is a troubling consequence. In effect, this consequence suggests that it is justified to kill an innocent person in certain circumstances. But the moral impermissibility of killing the innocent is one of the most central tenets of most moral theories. While the killing of innocent persons might be excused due to some sort of ignorance, it seems very strange indeed to suggest that killing an innocent person might be justified due to some sort of ignorance. But that is precisely what is being suggested here: if we accept the major premise of DA as true, we must accept that the mistaken judgment that a person is guilty does not merely excuse the act of killing an innocent, but in fact justifies it. This consequence of DA, that its principal normative premise would not only excuse CP when performed as a result of a mistaken judgment of guilt, but would in fact justify CP in such a case, is one that has to my knowledge so far gone unnoticed. That ignorance and mistakes can justify actions that would not otherwise be justified seems an untenable consequence of accepting the major premise of the two-premise version of DA. Since the major premise relies principally on premise 5sa of the sub-argument, we see that accepting 5sa leads to untenable consequences.(11)


Of course, someone might want to hold that it really is morally permissible to kill innocent persons so long as one has come to a judgment of guilt after careful and conservative examination of evidence. Someone might claim that, while the actual recorded cases of innocent persons being killed due to judicial error are indeed tragic, they are not examples of acts that violate moral requirements. Since the possibility of such errors is always present in even the most careful of judicial systems, one must tolerate erroneous infliction of punishment or abandon the practice of inflicting punishments altogether. It might be argued that since a system of punishment is necessary for social cohesion; it is justified so long as the system is designed to minimize as far as possible the punishment of the innocent. However, even in such a system innocents will unavoidably be punished. Therefore, the punishment of the innocent is justified in such a system.

There are several problems with this line of thinking. First, this line of thinking confuses the justification of a system of punishment with the justification of a particular infliction of punishment. It may be that a system can be justified, and yet when the human agents of that system make mistakes, the particular infliction of punishment is unjustified but excusable (so long as the mistake does not result from negligence). In fact, this seems to be a natural way of talking about punishment. My objection to DA is that it does not allow us to talk this way.

Secondly, I think this line of thinking overlooks the possibility of excusing erroneous inflictions of punishment without justifying them. It may indeed be true that erroneous inflictions of punishment are unavoidable, no matter how careful we are, and that we must therefore be willing to tolerate such erroneous infliction if we are to have a system of punishment at all. But maintaining that errors justify inflictions of punishment on innocent persons is not necessary in order to "tolerate" such inflictions. We can excuse such inflictions without justifying them: we can free the court system of culpability (at least if we can be convinced that the court system's failure is not due to any carelessness in the judging procedure) and yet insist that to punish the innocent is always wrong. In fact, it seems that the moral impermissibility of punishing the innocent is precisely what impels the courts to be careful in the judging process. To say that mistaken acts of punishment are morally justified would cripple the moral imperative to constantly improve the accuracy and reliability of the judicial system.

Furthermore, it seems that to allow a mistake, even an honest one, to justify an act that would normally be considered wrong is to radically misunderstand the significance of moral justification. Honest mistakes, understandable ignorance, and impotence are precisely the sorts of conditions which serve to exculpate wrongful acts, to excuse them rather than to justify them. Once we allow mistakes to justify such acts we are blurring the distinction between justification and excuse, between the condemnation of acts and the condemnation of agents.(12) Ignorance of a person's innocence is a characteristic of the agent or agents inflicting punishment, not a feature of the act of punishment itself. Thus, such ignorance should at most exculpate the agent (withdraw condemnation from the agent); it should not justify the act (withdraw condemnation from the act).(13) Since a consequence of accepting premise 5sa is that we would have to justify CP when it is erroneously inflicted, and not merely excuse it, we can legitimately conclude that 5sa is false on the grounds that it leads to untenable consequences. Because DA depends on 5sa, DA should be rejected as well.


(1)Arguments from deterrence, as I am understanding them here, are to be distinguished from those arguments which incorporate elements of both deterrence reasoning and retributive reasoning. W.E. Cooper and John King-Farlow, for example, argue (following Rawls) that while the policy of CP should be justified on grounds of deterrence, particular acts of CP should be justified on retributive grounds. It seems that in this sort of argument, retributive considerations have a kind of moral priority: it is possible to justify the policy of CP on grounds of its deterrent effect only because it is permissible to kill murderers for retributive reasons. Thus, this kind of argument is ultimately not a deterrence argument at all, and will not concern me here. See Cooper & King-Farlow, A Case for Capital Punishment, 20 J. Soc. PHIL. 64 (1989). See also Rawls, Two Concepts of Rules, 64 PHIL. REV. 3 (1955).

(2)Isaac Erlich has tried to show with statistical evidence that CP does in fact serve as a superior deterrent. See Erlich, The Deterrent Effect of Capital Punishment: A Question of Life and Death, 65 AM. ECON. REV. 397 (1975). Erlich's conclusions are belied, however, not only by other studies that arrive at contrary conclusions (see, for example, Glasser & Zeigler, Use of the Death Penalty vs. Outrage at Murder, 20 CRIME & DELINQ. 333 [1974]) but also by strong challenges to the validity of his statistical argument. See especially Passell & Taylor, The Deterrence Controversy: A Reconstruction of the Time Series Evidence, in CAPITAL PUNISHMENT (H. Bedau & C. Pierce eds. 1976).

Michael Davis argues that common sense by itself clearly supports the superior deterrent power of CP over other forms of punishment. See Davis, Death, Deterrence, and the Method of Common Sense, 7 Soc. THEORY & PRACTICE 145 (1981). Jeffrey Reiman has contested this position in his essay Justice, Civilization, and the Death Penalty: Answering Van Den Haag, 14 PHIL. & PUB. AFF. 115 (1985). Even if Davis is correct, however, this fact does not establish the moral permissibility of CP, as I show in this paper. Davis himself admits this point.

(3)For a useful discussion of the various ways in which CP's deterrent effect is thought to be achieved, see the report of the British Royal Commission on Capital Punishment, 1949-1953.


(5)Bentham, Principles of Penal Law, in 1 THE WORKS OF JEREMY BENTHAM 396 (J. Bowring ed. 1962).

(6)This premise has been contested by, among others, J.S. Mill and Jacques Barzun. Both contend that death is preferable to life imprisonment. See Barzun, In Favor of Capital Punishment, 31 AM. SCHOLAR 181 (1962). While this is an interesting position, it seems a difficult one to defend. Those who do defend such a position appeal to the suffering associated with imprisonment, and note that the suffering associated with a death sentence is much shorter. It seems unreasonable, however, to consider only the amount of suffering which is involved in the two forms of punishment. For what makes CP so feared is not the suffering it produces, but the fact that it ends life. The cessation of life itself, no matter how little suffering accompanies it, seems a very severe thing indeed. See Bedau, Capital Punishment, in MATTERS OF LIFE AND DEATH 163-64 (T. Regan ed. 1980), for an argument defending the greater severity of CP over life imprisonment.

(7)David Conway makes this point in Capital Punishment and Deterrence: Some Considerations in Dialogue Form, 3 PHIL.&PUB. AFFAIRS 431 (1985).

(8)Of course, someone might argue that the death of the convict is not evil at all because the convict deserves death. If, however, the convict deserves death, then killing the convict is not in need of justification on deterrence grounds, and the argument at issue does not arise. It might also be argued, however, that the death of the convict, while an evil, is a much lesser evil than the death of an innocent. If such is the case, then even a very slight difference in deterrent effect between CP and life imprisonment might be enough to justify CP by virtue of this effect. Someone who wants to argue along these lines, however, must defend the view that some human lives (innocent lives) are significantly more valuable than other human lives (guilty lives). The debate over the relative value of human lives, however, is one of the most contentious in value theory; it is unclear how an assessment of comparative human value should be made. Conway notes as much in his article (id. at 441-42).

(9)The argument I present here should not be confused with the sort of argument made by H.J. McCloskey and E.F. Carritt, to the effect that utilitarian punishment theory would in some circumstances require that one frame and punish an innocent person as a scapegoat. Rawls, in Two Concepts of Rules, supra note 1, has argued strongly that utilitarian considerations alone would rule out any penal practice that left room for such tactics. The argument I present here, however, suggests that, if we accept DA, innocent persons can justifiably be killed even within the context of a penal practice that never knowingly punishes an innocent and makes every effort to punish only the guilty. Thus, Rawls's criticism of Carritt's and McCloskey's view, even if successful, does not undermine the argument raised here. See E.F. CARRITT, ETHICAL AND POLITICAL THINKING 65 (1947); and McCloskey, A Note on Utilitarian Punishment, 72 MIND 599 (1963).

(10)It is easy to see how the two-premise argument outlined here could be modified to represent the general form of a deterrence-based argument for punishment as such. Thus, what I say here with respect to DA will apply equally well to deterrence arguments for punishment in general (of the sort espoused by Bentham).

(11)Again, what can be said for CP can be said for punishment generally. Recently, Rex Martin has presented an essentially consequentialist justification for punishment (he argues that punishment is justified when, and on the grounds that, it minimizes the violation of rights in the state, i.e., when certain consequences obtain), and has tried to show that the punishment of innocent persons could not be justified within such a framework. What he argues is that it would be justified to punish only those who are adjudged violators of rights, and only if the judging procedure is reasonably accurate. What Martin fails to discuss is the fact that any judging procedure is fallible, which means that on some occasions adjudged violators of rights will in fact be innocent of any such violation. Within Martin's system of justification, punishing those innocents will not merely be excused, but will be justified. Given the unacceptability of such a consequence, Martin's framework for the justification of punishment is seen to be flawed. See Martin, Justifying Punishment and the Problem of the Innocent, 20 J. SOC. PHIL. 49 (1989).

(12)It may be the case that the blurring of this distinction is common in the law. In fact, I believe that excuses are not given their proper place in the court system, perhaps because of the strong contextualism of excuses, and the concomitant difficulty of codifying excuses. J.L. Austin makes a similar point in A Plea for Excuses, 57 PROC. ARIST. SOC'Y 1 (1957). The difficulty of making room for excuses in the law is exemplified by the cases of battered women who kill their husbands in their sleep. Those who want to withhold punishment from these women often do so by trying to justify their actions on grounds of self-defense, even though alternatives to killing the husband were in fact available and had not been tried. See, e.g., Stell, The Legitimation of Female Violence, in JUSTICE, LAW, AND VIOLENCE 241 (J.Brady & N. Garver eds. 1991). In such cases, it seems much more natural to excuse the women than it does to justify their actions.

(13)I am implicitly drawing a sharp distinction here between moral justification and epistemic justification. One may be epistemically justified in believing that an act is morally justified, and yet be mistaken in this belief. It is precisely in cases where the agent has such epistemic justification that I believe an act can be excused (but not morally justified). The agents of the court who execute an innocent person may be epistemically justified in believing that the person is guilty, and may therefore be excused for killing the person; but such epistemic justification does not by itself provide moral justification of the act.
COPYRIGHT 1993 Institute for Criminal Justice Ethics
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Reitan, Eric
Publication:Criminal Justice Ethics
Date:Jan 1, 1993
Previous Article:Shame, forgiveness, and juvenile justice.
Next Article:Editor's introduction.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters