PUNISHMENT, THE PRINCIPLE OF HUMANITY, AND THE DEATH PENALTY by Ted Honderich This rather verdictive paper was given to an admirable conference on the morality of the death penalty in the Columbus School of Law at the Catholic University of America in Washington. It is the successor to a lot of old stuff of mine on the subject of punishment. It seemed to me to complement the orderly thoughts at the conference of Cardinal Avery Dulles and Hon. John Noonan, United States Court of Appeals. All of the thoughts, mine a little revised, are to appear in a volume edited by Professors John Grabowski and William Wagner. Mine also get expression in a revised edition in 2005 of my earlier punishment book, now very changed and under the title Punishment, The Supposed Justifications Revisited.    The most traditional of the supposed moral reasons for punishment by the state is retribution. What it is, traditionally stated, is that part or all of what makes punishment right is that it is deserved by the offender for his offence. Not a lot is added to this moral proposal by all the other language pressed into use in restating it, sometimes to make it sound less archaic or righteous. Nothing much is added, or subtracted, by speaking of a penalty as proportional, equivalent, commensurate, corresponding, owed to the offence -- or the retributive, fitting, merited or just penalty, or the penalty according to our right or the offender's right, or his loss of an immunity.1    That is not to say that the reason contained in the original or other statements of the reason for punishment is clear. Far from it. Nor is it one reason. It is remarkable, and a fact in need of some explanation, that lawyers and jurisprudents pass by the obscurity and ambiguity, indeed various obscurities and ambiguities, without any of the attention that they may give to the clauses of a will or a statute.    It is natural if not necessary to suppose that to say a punishment is proportional or whatever is to speak of some fact of relationship. More precisely, it is to advance some factual proposition, one that is true or false in the ordinary way. At least typically, it is such propositions that we use to defend our judgements as to the rightness of our practices and actions generally.    When punishment is said to be proportional, what exactly are the things that are in the relationship, and what exactly is the relationship -- be it factual or otherwise?    (1) Past actions or deeds are things that come to mind when imposing the death penalty is under consideration. That is, things like an offender's taking a life, where that is roughly a bodily movement owed to an intention. Actions or deeds in this ordinary sense, maybe the sense intended by some original proponents of the lex talionis or the rule of an eye for an eye, are not things that need much inquiry, say into the kind and degree of responsibility of the agent. As for the relationship between two life-takings, it is natural to say they are tokens or instances of the same physical type.    Forget about the problems inherent in the tokens-of-the-same-type relationship. No more needs to be said against this reason for punishment than that it cannot conceivably be right to take a life on the ground that its owner has taken one. Just for a start, kinds and degrees of responsibility on the part of the offender must be part of any arguable justification of punishment. Our police officers take lives in the line of duty. So does our airforce.    There is a reason for attending to this primitive and dim attempt to justify punishment. It seems to survive in feelings and judgements in this new millenium, most relevantly in our own ignorant societies -- American, English and like societies. Ideas lurk in minds, some of them nonsense. Ordinary defenders of punishment as retributive, and the death penalty in particular, when exasperated by problems and objections, have private or even public recourse to the simple-mindedness that somebody did something to somebody else and it is the same as what is about to be done to him.    (2) Contemplate the seemingly superior proposition that in talk of proportion the things that are proportionate or or not are the culpability of the offender in the offence and his dissatisfaction, distress or suffering in the penalty. An offender's culpability, in general terms, is a matter of his kind and degree of responsibility for it, along with the harm he does. And contemplate in particular with respect to the relation that there is a particular fact of proportionality -- a fact of quantity, an equality as in the case of an equation or things of the same size or weight.    This is not worth more attention than the lex talionis. It is clear that culpability is unquantifiable in any ordinary sense, if only because of its component of responsibility. Whatever else can be said, it is yet clearer that culpability and dissatisfaction are not quantifiable in the same units -- commensurable. So if the Retribution Theory's proposal is about serious quantification, real measuring or anything really like it, it is hopeless. A lot of ponderous talk and implication by judges has been vacuous. You can hear the implication still.    (3) Of course anyone can put numbers on categories of offences in terms of their culpability, and on punishments in terms of their severity, or, more likely, put offences and punishments into labelled categories in ordinal and cardinal scales. Neither ordinality or cardinality actually requires numbers. Then the categories of offences, 1st degree murder and so on, can be brought together with the punishments. A certain category of offences goes with a certain category of punishments.    From what we have already, we know that the matchings cannot be made on the basis of actions or deeds simply understood, more or less physically, as in (1) above, or on the basis of a supposed real fact of measurable quantity, as in (2). But for the moment let us not ask for clarification of how the matchings are made. We know that such systems of punishment -- matching categories of offences and punishments -- do indeed exist. There are these structures existing in the statute and other law of our societies.    To come to the point, when it is said that a punishment is the deserved or the proportionate punishment, what may be meant is that it according to the system of offences and penalties that exists in the society. It is related to the offence in that way. That is a factual claim, true or false. Consider it by itself. That is, consider it entirely independently of whether the system is a somehow justified system. If you do this, which you must do if you are to have a plain factual proposition, something follows immediately.    It is that you do not get a reason for punishment at all in this traditional way. There have been horrendous penalty-systems, one of them in Nazi Germany. Almost all penalty-systems are open to some objection even in the eyes of their defenders. To put the point generally, it cannot be that the existence of a set of rules, with nothing said of their being good, bad or indifferent, justifies action according to them.    Is it possible to rescue something here by saying that if you stick to a penalty-system in punishing people -- whatever penalty-system -- you are treating similar cases similarly, and different cases differently? That is, you are proceeding in accordance with what was sometimes called formal justice? This comes to very little recommendation at best. I cannot put in for much mitigation of penalty by pointing out that I did not discriminate against the woman I tried to rape -- I try to rape all women of just that sort in that kind of situation. Nor can a judge defend himself morally by saying only that he is treating similar cases similarly.    No doubt a reference to a penalty system can enter into other reasons for punishment. Indeed such a reference can be written into any of the reasons considered below. There is no need to do so, however, and there is the danger of more confusion.    (4) Here is a sample of less traditional thinking about retribution that also tries to come up with a fact of relationship and succeeds. Somebody you know gets mugged and the mugger is put in jail for some time for it. Think about the two experiences. Compare the prospect of being mugged yourself in the same way or being put in jail yourself for the same time. You could prefer the first to the second, it is said, or the second to the first, or be indifferent between the two prospects. What we come to here is the idea that there is a reason for a punishment's being right in the fact that an ordinary and reasonable man would be indifferent between suffering the offence and suffering the penalty. (Goldman)    This is as bad as what has gone before. Certainly it cannot be that whenever you think of two experiences between which an ordinary and reasonable man would be indifferent, and somebody has actually had an instance of the first one, that is a reason for somebody else having an instance of the second one. There may be indifference, for example, between giving a tip to an obnoxious waiter and being delayed a little in traffic on the way home. Nor does it help if you make the rule, for whatever other reason, that he who has the second experience is the person who caused the first experience.    Indeed I mainly mention this attempt to make sense of retribution in order to illustrate that defenders of it are forced to go to extraordinary lengths to try to defend it. Does it have some real core that they wish to keep dark?    (5) Let us give up looking for hard facts of proportion, desert or whatever that can enter into the justification of punishment. If we do so, we may think first of what is most plain and salient about the utterance by someone that a child-killer deserves what he is getting. This is that that it is right that he is getting it. Whatever one's feelings about child-killers, it is clear that if that is what the desert-claim comes to, it is no argument for what he is getting, but a circularity, a begging of the question. We are given no reason at all for the rightness of his punishment.    Nothing is more common in reflection on the justification of punishment than such circularity, somehow obscured or bundled up into a lot of stuff. Return to the matter of a penalty-system, noted in (3). What is implicitly offered by someone who defends a punishment by its being according to such a system is indeed very likely that it is according to a system that is right or whatever. In effect that is to beg the question, not advance discussion at all. There are quite a few instances of this absurdity, some involving the ambiguity of talk of justice or rights or the like. Lawyers when they turn to moral reflection are often curiously oblivious to the circularity.    (6) That is not to say that all non-factual reasons for punishment beg the question in this way. It is possible to begin with or rest on a moral rather than a factual premise, without circularity. You can define punishment in a way that is possible if unconventional, as behaviour narrowly conceived -- behaviour conceived in terms of only some of its effects. It is no part of the punishment of putting somebody in prison, conceived in this way, that what you do causes dissatisfaction, distress or suffering. The punishment is the putting of them into prison.       You are now in a position to avoid saying in effect that punishment is right because punishment is right. What you say, rather, is that punishment in your sense is right because it results in what is right, which is the dissatisfaction, distress or suffering of the guilty. (Davis)    Is this not another ndication of struggle in extremity, in no way helped by invoking the supposedly self-evident nature of moral truth? One evident weakness of it is that the moral premise, about the suffering of the guilty, does seem to call out for support. Is there a fact that supports it? Well, we know already that such a thing will be hard to find. If you do insist on the line that the suffering of the guilty is an intrinsic good, that there just is a moral truth that can be somehow discerned, something that doesn't depend on something else, you face an obvious difficulty. You immediately run up against someone else's so-called moral truth, perhaps the moral truth that it is wrong to make anybody suffer, however guilty, if absolutely no good effect at all comes of it.    So much for six traditional and not so traditional attempts to make sense of the the supposed reason for punishment by the state that it is deserved, proportionate or whatever. A philosopher as clear-headed as any who considered the subject drew a certain conclusion from such reflections. (Mackie) It was that retributivism cannot be made sense of, that the various ideas of retributivism are in one way or another hopeless. He said it was a paradox that they persist in our common thinking.    That retribution cannot be made sense of has seemed to me exceedingly unlikely. It can hardly be that there is an emptiness, a hole, where more or less the whole history of mankind has found an argument and acted in accordance with it. The situation cannot be that there is no argument contained in retributivism. Rather, almost certainly, the situation is that we have found the argument embarrassing and turned our eyes away from it. There is self-deception here, supported by what may be the necessary self-pride of our lawyers and the like.    (7) If you and I meet, and you offer what I take to be an insult about my intelligence as exhibited in this paper you are reading, I will have some desire, however small and fleeting, for something like your discomfiture. I will also have to deal with a desire, certainly different, if you kill my son or rape my daughter. It is a fact of human nature, whatever qualifications you can put on it, that lesser and greater injuries give rise to grievance-desires, desires for exactly some degree of dissatisfaction, distress or suffering on the part of the person who has done the injuring. The distinctive thing about these desires, not only on the part of the victim, is that they are exactly desires for another's distress or the like. They are, as we sometimes hurry to say in one way or another, low desires.    So retributive talk about punishment does in fact cite a perfectly factual fact. To say that part of what makes punishment right is that it is deserved or proportional or whatever is to rest on the argument that it satisfies certain desires -- grievance-desires brought into existence by the offender's offence. Or, more precisely, it is to rest on the argument that the particular penalty corresponds to or is proportional to the offence in the clear sense that that penalty will do neither more than less than satisfy the existing grievance. (Honderich 1984/1989, Ch. 2, Postscript; 1988, Ch. 10)    It is notable that the truth of this account of the retribution reason for punishment, which account might be clarified in response to scepticisms of one kind and another, has become more obvious in some societies over recent decades. The English press, at its lower end, has never been decorous. What is now true of the lower end of it is that it reports on the families of victims of offences in a certain way. What they want, it says, is satisfaction. They are quoted as using the word. No doubt there is an American counterpart to this more open sense about retribution.    One other thought on this particular retributive reason for the death penalty and other penalties, as against the previous six high-minded reasons, is a radical one but maybe right.    If you take retributivism seriously in one of the high-minded ways, starting with the six ways above, you suppose that there is a reason for a practice, including the practice of killing people, that satisfies no desires. That is implicit in all these enterprises, as a little reflection will indicate. You take the view, if you are clear-headed, that what we have in the stuff about retribution is not the satisfaction of any desire. But reasons of the kind with which we are concerned, certainly including moral reasons, are reasons for action. And surely nothing can be a reason for action that does not promise the satisfaction of some desire? (Honderich 1996)    So the supposed high-minded reasons, whatever else is said against them, are not reasons at all. There is no reason given by citing, just, a supposedly quantifiable relation between culpability and distress or the like. There is no actual reason given in any of the other supposed reasons (1) to (6) if they are taken in the intended and official way. If they have worked as reasons, which indeed they have, this must be because they have had another content. It is given in (7).    As for the worth of the real retributive reason for punishment, that is plain enough. To say the least, it would go against the fact and practice of ordinary morality clear-headedly to take a man's life only for the satisfaction of others. This is not to say that such satisfaction is inhuman and to count for nothing. That is more confused high-mindedness, of a different kind. There is something to be said even for an execution if, say, it allows some peace of mind to the widow of a man who has been murdered. It is of course unthinkable that there is enough to be said for the execution if only this can be said for it.    The tendency of all that has been said so far2 is of course towards reasons for punishment of a consequentialist kind -- reasons that have to do with the consequences or effects of the practice. Most familiarly, such reasons have to do with protection and deterrence. Punishment has the recommendation that it deters possible offenders, including the man being punished, from further offences. Far from satisfying no desires, punishment has the recommendation of satisfying the most respectable of future desires, the desires not to be robbed, raped, tortured, killed -- and, we can add, desires not to have your life and the lives of your children destroyed by respectable business corporations and the like.    Some are keen to say that punishment, and the death penalty in particular, has no deterrent effect. What is more, they seem to have facts to prove it. It would be agreeable to join them, in a way, but this is not easy. If I am right, what this line of thought come to is that the incidence of offences or some offences stays much the same in the absence of a particular kind of punishment. So, it is said, there is no deterrence.    This line of thought seems to overlook a certain proposition. It is not the familiar proposition that our practice of punishment has an effect on those who are actively or really contemplating an offence. The overlooked fact is not that such people are consiously frightened out of an offence, or consciously judge that the gamble is not a good one, that the offence would not be rational. Rather, the proposition is that punishment is essential over time to the existence of much larger numbers of people who do not actively or really contemplate any offence of seriousness.    It would be remarkable if so large a fact as punishment, a fact that has been large in the whole histories of our societies, had no significant effect on conduct. It is my own view that it has contributed as much as anything else to the fact that serious offences do not really come to the minds of very many people as possibilities. It has contributed as much as anything else to unreflective law-abiding behaviour. It continues to make that contribution. You can have the thought that the past death penalty in a particular American state contributes to such behaviour after the death penalty has been abolished. Also the thought that the existing death penalty in another state contributes to such behaviour.    This paper must be particularly unsatisfactory in not considering further the question of the consequences of our practices of punishment, preventive as distinct from only deterrent effects. The paper does rest, however, on the assumption that our practices of punishment do have consequences of importance, that they do affect behaviour and could do so to a greater extent. They could affect it differently from the way they do.    To think of reasons for punishment of a consequentialist kind is to think, maybe first of all, of Utilitarianism, which continues to bury its philosophical undertakers. Jeremy Bentham did indeed set out to justify punishment by the Principle of Utility and nothing else. To be less ambitious, do we get one moral reason for punishment from the principle? It is best understood as being the principle that the right action or practice or the like is the one that when compared to alternatives is likely, according to the best available knowledge and judgement, to produce a greater total of satisfaction.    It was persistently objected that this principle would in certain circumstances justify the punishment of the innocent -- or at any rate the victimization of the innocent. To that objection it was possible to reply that we already have in our bodies of law something close to such victimization, some of it as a result of rules, principles and statutes of strict liability. Is this victimization not tolerable? But a different and perfectly conclusive objection to the Utilitarian reason for punishment was that certain victimizations it would sanction would have none of the recommendation of existing victimizations in the law.    To think of the Utilitarian reason for punishment is to be led to or reminded of something important. There is nothing significant, nothing worth attention, that can be called the consequentialist or the preventive reason for punishment. Rather, there are different reasons that recommend punishment by different consequences it is supposed to have. Utilitarianism recommends it by total satisfaction, misdescribed as the greatest happiness of the greatest number.    There have been religious, racist, Fascist, Communist and other reasons for punishment. To think of these is to have a first ground for another conclusion that cannot be avoided. It is that the question of the justification of punishment does not exist on its own, and never did, but is absolutely necessarily bound up with political philosophy and the worth of particular societies. If it was possible in the past, it would now be unworthy of philosophy or any serious inquiry to pretend otherwise.    What consequences make or would make our systems of punishment right?    Should we say the maintenance of the liberal society? I doubt that we have that option, since there is no elucidated and determinate thing that is liberalism. The ramified theory of justice of John Rawls is by far the best attempt to set out a conception of the liberal society, but to my mind it does not come near to succeeding. (1990, 2002)    It propounds a principle having to do with traditional liberties, and leaves them vague. This principle is to have precedence over a principle of allowable socio-economic inequalities. If the principle about liberties is vague, what is the upshot of this precedence? This is certainly not the only uncertainty about this liberalism. To my mind it amounts to good intentions never carried into clear prescriptions. It is different from the grim selfishness of libertarianism (Nozick; Honderich 2002), but that is no sufficient recommendation.    Among other consequentialist reasons advanced for punishment is the reason that it reforms or socializes offenders, or serves what is called their spiritual good. It is hard to find time to give this attention. For a start it now seems plain that punishment more often does the opposite. Let me instead move towards my own consequentialism, and a proposition or two about the only defensible reason for punishment by the state.    Consider the question of fundamental human goods, which is to say our fundamental categories of human desire and satisfaction. There is everything to be said for resisting the temptation to retreat into generality here -- into some general principle or notion of value. Those who resist the temptation, including some Catholic proponents of natural law (Finnis), have a chance of staying in touch with life, and with the natural fact and practice of morality.    Almost all of us, indeed all of us save for ordinary suicides and martyrs old and new, want a life of decent length, say 75 years. We want the same thing for those close to us, children first. A life that lasts long enough -- enough of living in an elementary sense -- is both an intrinsic good and also a means to other great goods.    The second of these goods is a quality of life that is bound up with certain material means. We want more than what makes for elementary existence. We want what can reduce or save us from pain, and more than rudimentary shelter and food and drink. We all want what it is easier to disdain if you have them, which is consumer-goods that satisfy more than manufactured wants.    To specify the other great goods still more quickly, but not because of their lesser importance, one is freedom, power and safety. It comes in different kinds and settings. Individuals struggle against personal oppression or coercion. Whole peoples struggle for freedom in a place to which their culture and and recent history attaches them. The outstanding instance at this time is the moral struggle of the Palestinians against what it is proper to call the rape of that people, a rape by Israel aided and abetted by the United States.    All of us, individuals and groups, also want respect and self-respect. Our lives are made poorer things, maybe  things of little worth, by condescension, racism, self-disdain and the like. All of us want, fifthly, the great goods of relationship, both to individuals and as members of groups. None of us wants a life made thin or barren by isolation. All of us, sixthly, desire the goods of culture, these encompassing education, enrichment of feelings, religion, diversion, art, and more.    Reflection on these great goods and on being deprived of them issues in a conception of bad lives. If there is ordinary truth somehow or somewhere in the natural fact and practice of morality, from which worked-out moralities derive, it does not produce by intuitive or deductive means, or by any other indisputable means, a decision as to what lives are bad. Coming to a conception is indeed more a matter of decision than discovery.    By my lights, bad lives are those that are cut short, typically by starvation or disease, and also lives that are dragged down or ruined in any of four other ways. A life is bad if it is long enough but without material well-being -- maybe the means of dealing with river blindness. It is bad too if it is without freedom, power and safety. It is bad if it is without the human standing of respect and self-respect, and it is bad if it is long enough but has little of all the other great goods in it.    There is a principle to put in place of other moral principles and such ideologies as liberalism. It is a principle of humanity, well-being, equality, fellow-feeling or generosity. It has a good history. It was, you can say, the principle of Colonel Rainborough, who is remembered for what he said in the Putney Debates at the time of the English Civil War in the 17th Century. `Really I think that the poorest he that is in England hath a life to live as the greatest he....' Something like the same note was struck, whatever notes he struck at other times, by another remembered speaker. He spoke of `a new nation, conceived in liberty, and dedicated to the proposition that all men are created equal'. Abraham Lincoln, 1863.    To my mind this principle is the best that can be done to make consistent the elements of our natural morality. Where else is one to start, by the way, in setting out to find a worked-out general judgement of what is right? Those who are aided by religion remain dependent on this human foundation. Those to whom religion must be a matter of a kind of hope rather than belief are less distracted from this foundation.    The principle of humanity, in the past called the principle of equality by me, has the goal of saving people from bad lives. It does not have a goal of bringing people into relative standings, say equality, with respect to the great goods, although to act on it must have some egalitarian side-effect. If it is also connected in another way or two with equalities, it is true to the very greatest of moral imperatives, the ending of deprivation, distress and suffering.    To state it a little more explicitly, it is the principle that the right thing to do, or the right practice or institution or society to have, is the one that according to the best judgement and information is the rational one -- the most effective and economical one -- with respect to the end or goal of saving people from bad lives as defined.  (Honderich 2002b) That is a good start on setting out the principle but no more. What is needed, mainly, for a fuller understanding of it, is something on the mentioned rationality, on the effectiveness and economy of of the actions, practices, institutions and societies. They will be in accordance with a number of policies.    The first is to help those with bad lives and to prevent bad lives in ways that do not significantly affect the the well-being of the better-off. The most obvious of these is the transfer of means to the great goods from the better-off -- means that they can spare without significant effect on themselves. Certainly these exist.The second policy is to help the badly-off by real means that do affect the well-being of the better-off, without making them badly-off.    The third policy is a realistic one about incentive-rewards. It does not suppose for a moment that the incentive-rewards in the enterprise of humanity are those now demanded in whatever end by the better-off. The fourth policy, implicit in the previous ones but best made explicit, is a prohibition, if not an absolute one in every respect, on wounding, attack, killing, torture, sexual attack and violation, threat, intimidation and other violence and near violence.    You will anticipate my conclusions. The first is that the only thing that can morally justify punishment by any state is the principle of humanity. It supplies the only reason there can be. It is not part of some bundle of principles whose upshot is uncertain. It has the required or essential recommendation of any morality that gives what a morality is supposed to give -- determinate guidance, an insurance against self-serving self-deception and the like. That recommendation is a single principle, quite distinct from a congeries of principles, commitments, values and what-not out of which you can get what you want.    So, to repeat, the only possible institution of punishment that can be right is one that is given over to the aim of getting people out of bad lives and keeping them from falling into bad lives. That there are few if any such institutions of punishment in existence is no reason to doubt the conclusion. It is the curse of moral thinking now, particularly in the English-speaking world, to respect reality, to suppose that there must be silliness in conclusions that go against the good sense of the Rotary Club and local mores, not to mention the institutions of media, religion and government. Real moral thinking, in which we can all engage, has nothing to do with all that -- a truth, incidentally, that has turned up throughout the history of religion.    It is also my conclusion, certainly, that our existing institutions of punishment, more or less from the ground up, are without moral justification and are wrong. Those mild epithets reveal something that is in need of further attention and resistance. It is the new extent to which the language of moral denunciation has become the property of those who use it only to denounce some terrorism, usually not including state-terrorism, and some well-chosen tyrannies, and like things that do not serve our self-interest. There are other things that call for new language, new use of the resources of language.    Let me be a little more explicit about our punishment systems.The American and British ones, to concentrate on them, have in them an element of retributivism that involves such confusions as noticed above in (1) to (6) and also the grim desire for the distress of others as an end. To this is added a much larger consequentialist element. The systems have the goal, about which no conspiracy theory is needed, of defending and advancing a way of life, one that includes hierarchic democracy and a kind of capitalism. (Honderich 1995, 2002b) That way of life, from the proper viewpoint of the principle of humanity, as you can begin by saying, is indefensible.    We here encounter, incidentally, an attempted reason for or justification of punishment related to the one in terms of liberalism. But this attempted justification eschews theory, doctrine and high talk of rights, and takes punishment to be OK when it serves the end of maintaining an existing society, such as the American or the British one. You use ostensive definition -- point at the society to get the justification.    The United States leads a number of societies that are in fact morally vicious. That is a proper name to put on a society where the means to the great goods are shared out in a certain way. Those goods, you will remember, are a decent length of life, a material quality of life, freedom and power, respect and self-respect, and the goods of human relationship and of culture. In the United States, the best-off tenth of population has at least 30.5% of the fundamental means to these goods. The worst-off tenth has at most 1.8%. The effect in terms of life-expectancies is great.    That is a start, but only a start, on an account of the moral viciousness of our societies. The punishment systems in societies like those of the United States and Britain also support our contribution, mainly by omission, but also by commission, to facts in much of the rest of the world, notably certain African countries. To take but a sample that is easily calculated, we have a responsibility for a loss in living-time, a loss of time for the having of the other great goods, of 20 million years. (Honderich 2002b)    Little has been said by me of just the death penalty. This is not so much of an embarrassment as it might be. What has been said here of punishment in general applies particularly to the death penalty. Even if the death penalty raises special problems, there is no possibility of discussing the morality of it without being engaged mainly with the general reasons given for punishment.    Also, one can contribute to a specific subject, e.g. the death penalty, by trying to correct an assumption about the size or importance of it. Our great moral problem about punishment is not the death penalty, but the problem to which the whole raft of other penalties contributes more. That is the contribution of our punishment to worlds of bad lives. Those who die are not flies.    But, with respect to the death penalty itself, I myself come by way of the principle of humanity to more or less the conclusion, as it now is, of the Catholic Church. The death penalty ought not to be imposed, except in barely conceivable circumstances of necessity. That is a whit less than an absolute prohibition. It would of course be a kind of absurdity to have an overriding principle or principles of morality, which necessarily are of some consequentialist kind, and absolutely to prohibit what could in some conceivable circumstance demonstrably serve that end.    Maybe it is the case, however, and indeed maybe it is the Catholic case, that in practical terms, that with respect to what is foreseeable, we ought as societies never to kill. Or at any rate we ought never to do so if our societies are in the range of ordinary historical circumstances. Necessities are different in extraordinary cicumstances. I again have in mind the plight of the Palestinians, under the terrible necessity of defending their existence as a people against a moral crime of commission rather than omission. Their terrorism, it is possible to feel, is a moral right. Notes 1. It would have been good to bring some new philosophy about punishment to the conference from which this book derives. I could not do that, and what follows here is a summary and stiffening-up of the main contents of the third and fourth editions of Punishment, The Supposed Justifications and also some subsequent writings, including A Theory of Determinism, Ch. 10. The stiffening-up has to do with particular arguments but in the main with a view of the moral indecency of our societies, as conveyed in another book, After the Terror. As for the conference, let me record that it was possible for an atheist and socialist to feel a member of a certain moral community in which many Catholic thinkers are also members. 2. This discussion of retributivism must leave out things that seem to me nearby or indeed bound up with it. One is the idea that there is a reason for punishment in the fact that the offender somehow consents to it (Nino). Another is the more tempting idea that punishment is an act of communication (Duff). There is also the older idea that punishment is an annulment (Hegel). They are given more or less consideration in Honderich 1984a/1989. Another large matter in need of attention is the presupposition of ideas of retributivism having to do with freedom as against determinism, far too rarely considered by any advocates of punishment. (Honderich 2001, Ch. 19; 2002a) BACK References Bentham, Jeremy 1970 , An Introduction to the Principles of  Morals and Legislation, ed. J. H. Burns and H. L. A. Hart. London, Athlone Press. Davis, Lawrence H. 1972. `They Deserve to Suffer', Analysis. Duff, R. A. 1986. Trials and Punishments Cambridge, Cambridge University Press. ------ and N. E. Simmonds, eds., 1984. Philosophy and the Criminal Law. Franz Steiner. Finnis, John 1980. Natural Law and Natrual Rights. Oxford, Oxford University Press. Goldman, Alan H. 1979. 'The Paradox of Punishment', Philosophy and Public Affairs. Hegel, G. W. F. 1942. Philosophy of Right, trans. T. M. Knox. Oxford, Oxford University Press. Honderich, Ted 1984a/1989. Punishment, The Supposed Justifications. London, Penguin/Cambridge, Polity. Revised as Punishment, The Supposed Justifications Revisited, Pluto, 2005. -----------1984b. `Culpability and Mystery', in Duff and Simmonds. -----------1988. A Theory of Determinism: The Mind, Neuroscience and Life-Hopes. Oxford & New York, Oxford University Press. -----------1990. Conservatism. London, Hamish Hamilton, Penguin; Boulder & San Francisco, Westview. -----------1995. `Hierarchic Democracy and the Necessity of Mass Civil Disobedience'. Conway Lecture. London: South Place Ethical Society. Also at http://www.homepages.ucl.ac.uk/~uctytho -----------1996. `Consequentialism, Moralities of Concern, and Selfishness', Philosophy. -----------2001. Philosopher: A Kind of Life. London & New York, Routledge. -----------2002a. How Free Are You?, 2nd edition. Oxford & New York, Oxford University Press. -----------2002b. After the Terror. Edinburgh, Edinburgh University Press, distributed in the U.S. by Columbia University Press. Mackie, J. L. 1982. `Morality and the Retributive Emotions', Criminal Justice Ethics. Reprinted in Mackie 1985 --------- 1985. Persons and Values. Oxford, Oxford University Press. Nino, C.S. 1983. `A Consensual Theory of Punishment'. Philosophy and Public Affairs. Nozick, Robert, 1974. Anarchy, State and Utopia. New York, Basic Books. BACK