Ted Honderich

reviewed by John Williams

    Ted Honderich's Punishment: The Supposed Justifications Revisited adds to his classic Punishment in addressing itself to an audience of more recent political times, notably the UK under New Labour and post-September 11 USA. It includes a new concluding chapter.

    Punishment involves a deliberate infliction of suffering or deprivation. Suffering and deprivation are in themselves bad things. So we need a justification for punishing an offender. What is it? Those who say 'Because she deserved it' look back to the past and appeal to desert. Those who say 'Because it will reduce crime' look forward to the consequences and are typically utilitarians -- those who think that an action is right just in case it brings more benefit to society in general than would be the case otherwise.  Honderich argues persuasively that neither answer will do. He also rejects other sorts of answers such as 'It sends a message' or 'It reminds her of what the right values are' or the reformist's 'It will make her (or others) a better person'. He is critical of answers that are mixtures of the preceding sorts of answers. He observes that any justification of punishment must account for the grim fact that a high proportion of people in prison are mentally ill. Yet consistently with this, he effectively debunks the idea that all offenders are ipso facto mentally ill and so should be treated, not punished.

    Honderich argues that our judgments about what justifies punishment in society are inseparable from our judgment of what the decent society is. For him, punishing an offender is justified just in case it is best judged a rational step towards keeping people from living bad lives, where a bad life is roughly defined in terms of benefits such as longevity, bodily well-being, freedom and power, respect and self-respect, relationships with others and culture, including  knowledge in place of ignorance. He accepts determinism -- that every event has a cause that makes that it inevitable that the event take place. Yet he argues that we may still sensibly take the attitude that actions are voluntary, so people can have moral credit for their actions and offenders may be morally blameworthy.

    Honderich starts by defining punishment as

"an authority's infliction of a penalty, something intended to cause distress or deprivation, on an offender or someone else found to have committed an offence, an action of the kind prohibited by the law"

where to be found to have offended is to be found to have

"actually broken a rule out of intention or negligence, somehow freely and responsibly, or broken certain rules without that, or have occupied such a position as employer with respect to a rule breaker in either of the preceding senses."

While quite a mouthful, this realistic definition avoids any reference to desert and so keeps the playing field level for various supposed justifications of punishment.

     Honderich then considers numerous backward-looking justifications of punishment -- those that appeal to what the offender has done, including, but not exhausting, those that appeal to desert, in other words retributive justifications. He argues convincingly that none of these will work. Either they are circular or they reduce to a claim for which no reason has been given, one we may reject. If we say that it is right to punish an offender because she deserves it, what answer can we give to question 'What makes her deserve it'?  Not 'Because it is right'.

     Some of the backward-looking justifications that Honderich rejects are as follows. Jurisprudents often say that punishment is deserved because it is linked to a worked-out system of penalties and offences. But a wicked system of penalties and offences, such as Nazi Germany's, would not make her deserve punishment.

    A retributivist might say that there are intrinsic goods, each corresponding to the offender's sufferings that are 'proportional' to the offence. But this does not capture what is substantial about retributivism. In any case why should we accept it? We could try saying that desert amounts to imposing a penalty of same kind as the offence, but that would rule out justifications and excuses of the offence.

    We might say that the offender deserves the penalty just in case the culpability of her behavior is equivalent to the distress of the penalty. But culpability depends on the extent of harm she has done and the degree to which she is responsible for it, so culpability is incommensurate with distress.

    Jurisprudents may say that an offender deserves the penalty, say five years in jail, when his offence, say rape, falls into a broad category all members of which attract the same penalty. But why does this category justify the penalty?  Not because jurisprudents have decided that five years is the right penalty for rape, on pain of circularity.

    We might claim that the offender establishes the principle that it is right to injure others and so should accept, if rational, that the principle be applied to herself. But even if she has a right to be injured, this does not mean that we have a duty to act on her right. A related idea is that the offender has consented to punishment in the sense that had she thought about it sensibly, she would have agreed to the social arrangements under which she is now being punished.  But why should we accept this hypothetical claim?

    Only slightly more promising is the idea that to offend is to consent to be used as a means to the prevention of offences. Honderich points out that the offender does not consent to his punishment but only to her loss of immunity to it.

     What about the idea that an offender (say someone who steals food) freely sheds himself of a burden (that of self-denial) that results in an unequal distribution of burdens (of self-denial)? Punishment corrects this unequal distribution and encourages others to stick to it. But it is unlikely that the distribution was ever equal in the first place. The rich have lighter burdens of self-denial than the poor.

      Honderich thinks that there is only one idea in the backward-looking justifications of punishment that has any substance. That is the idea that the offender deserves her punishment because

"The penalty satisfies grievances owed to the offence, and does neither less nor more than satisfy these grievances, and in doing so it is in accordance with a system that connects penalties with offences by way of grievances."

    His objection to this is that it recommends a certain good, namely the satisfaction of people who have grievances, be gained at too great a cost, namely the denial of great goods to people. To this I might add my own objection: suppose that we execute a murderer when nobody has any grievance and the next day execute another, when there are plenty of grievances?  Must we admit that we are wrong today but not tomorrow?

     Turning to utilitarian justifications of punishment, Honderich shows how clear these are in comparison to backward-looking justifications. He provides a nice analysis of the factual claim that punishment prevents crime by deterring others that would otherwise have committed offences. In so doing he points out an advantage had by the utilitarian that is easily overlooked, that the utilitarian can explain the fact that punishment reduces offences, not by deterring would-be offenders who are caused to think twice, but by reinforcing an unreflective obedience to law. The empirical question of how well punishment deters would-be offenders is very difficult to answer. To that difficulty I add my own: in measuring deterrence, what can we know about those who have been deterred from committing an offence who spend their whole lives crime-free? Nothing of course, yet wouldn't we need to know this in order to assess the effect of deterrence?

     Honderich has a decisive objection to the utilitarian justification of punishment.  This is not that the utilitarian must treat people as means rather than ends, for Honderich argues deftly that this isn't so. Instead it is that the utilitarian is committed to wrongful victimizations. Against the 'stout and graceful philosopher Lord Quinton' who would object that it is impossible to punish the innocent on a definition of punishment like Honderich's, he points out that this difficulty is merely terminological (as Hart would say, a 'definitional stop'), since talk of victimization may replace that of punishment.   The utilitarian must say that a person must be punished (or penalized, when she is innocent) if the total resulting balance of satisfaction is greater than that which would result if she were not punished, regardless of how these satisfactions are distributed among a population. But we may imagine a possible case in which a penalty is imposed on an innocent person that causes her great distress, where the distress that would occur were she not penalized would be greater in total, yet so widely and thinly spread that very many people suffer trivial amounts of distress, such as not getting a cup of coffee in the morning. The penalty is wrong, or as the desert-theorist might say, unfair because undeserved. Even if such circumstances are unlikely to arise, the utilitarian is committed to doing something wrong in them, in virtue of a rule that he uses in all circumstances, including ordinary ones. That is enough to show that the utilitarian has the wrong moral compass, even in most ordinary circumstances. But it would be a mistake to think that this vindicates the desert theorist: a reason against victimization need not be a reason for punishment.

    Honderich seems right in making this the most serious objection to the utilitarian justification of punishment. But perhaps the utilitarian might reply by claiming that in the long run, avoiding such thinly spread inequalities of distributed satisfaction itself produces more total satisfaction than not avoiding them.

     Will we fare any better with the idea that what justifies punishing an offender is moral reformation? No. The victimization objection still bites since we may coherently imagine a wrongful victimization would improve the moral character of many others. Moreover we cannot justify punishment in terms of an improvement of the moral character of members of society unless we justify the claim that it really would be an improvement. Punishing offenders with the effect that many people change their moral character approved of by society would not be right if that society is itself wicked, as was Nazi Germany. So we still need an accurate conception of what the decent society would be. And Honderich argues that the offender who is reformed by punishment is reformed in the wrong way: the first step towards becoming a good person is not awareness that others condemn what she has done, but a sympathetic awareness of how breaking the law harms others.

    Honderich then effectively debunks the claim that all those who break the law do so as a consequence of mental illness and so should be treated rather than punished. Against the Freudian claim that offenders suffer from faculty relationships between id, ego and superego, he points out that there no evidence for this claim. Moreover the concepts it invokes have no precision. Eysenck's claim that children who are highly emotional and have a specific heredity are deterministically conditioned by the environment to become offenders ignores the fact that at least certain human decisions are different from a blink response. Against the sociological view that offenders offend because they have been formed by a certain social environment, it is difficult to see how this can be the complete explanation of criminality. In any case it would take enormous resources to treat all offenders -- or would-be offenders -- as if they were ill. Moreover, if the alternative to punishment is treatment, we would be justified in coercing offenders to undergo treatment, perhaps at great distress. We would have to give up argument and rational discussion with an offender to persuade her that what she did was wrong. We would also need the right view of what a well offender would be -- which again boils down to defending a view of the decent society.

    Non-utilitarians must or do hold that an offender may be punished justly only if she committed her offence voluntarily. How can they square this with causal determinism? Honderich accepts determinism. He holds that quantum physics does not threaten determinism, very roughly the claim that all events have causes that make their effects inevitable. Yet he argues that we may keep the attitude that actions are voluntary.  He maintains that we have two notions of freedom. One is freedom as origination: your decision is free (so you are responsible for it) just in case it originates in you as an uncaused cause, namely as an act of will. The other is freedom as voluntariness: your decision is free just in case it accords with your desires. Incompatibilism holds that freedom is voluntariness plus origination, so since determinism rules out origination, determinism rules out human freedom. Compatibilism holds that freedom is voluntariness, so determinism is consistent with freedom, because your decision, say to steal, may be inevitable given a preceding cause, such as a brain-event, yet still be in accord with your desires. Honderich holds that both incompatibilism and compatibilism are mistaken. Both assume that we have one fixed notion of freedom. In fact we have two. When we accept determinism and think of freedom as origination, our attitude is dismay, because our judgment that you were wrong to steal is threatened by the fact that you could not have done other than steal, so you were not responsible for stealing and thus not wrong. When we accept determinism and think of freedom as voluntariness, our attitude is intransigence: we stubbornly refuse to compromise on our judgment that you were wrong to steal. Therefore we can keep the attitude that actions are voluntary.

    I found this segment of the book the most technical. Perhaps I missed something, but I find this position obscure. Surely the important question is not how we conceive of freedom, but which conception is correct. Since freedom as origination is ruled out by determinism, the alternative seems to be the claim that you freely chose to steal only if you could have done otherwise had you wanted to -- despite the fact that what you in fact wanted was causally determined. If this isn't compatibilism, what is?

     Honderich then considers mixed theories, which roughly reduce to the claim that punishment is justified because it is or may be deserved by the offender and also prevents future offences. He shows, against Hart, that the fact that we can ask what justifies the practice of justification separately from asking what justifies particular punishments of particular offenders, does not mean that we must answer the first just in terms of prevention of crime and the second just in terms of desert. That won't work, given his objections above. Against Nozick's proposal that punishing the offender is justified if it reconnects him with the correct values, he points out that the

"aim is to defend our societies as they are, or to defend societies to which we might extrapolate from our societies. But we do not know, in any enlightening way, what is supposed to make these societies worth defending."

So the communication theory is incomplete.

      Honderich ends his book with an impassioned argument for his own justification of punishment. For me, this was when the book came alive, especially in comparison with the rather dry discussion of determinism. His position is not politically liberal, or communitarian, or conservative (to which he objects on the grounds that it assumes, with no moral justification, that we should all be self-interested). Rather, it is based on the Principle of Humanity: we must take what are best judged rational steps towards keeping people from living bad lives, where a bad life is roughly defined in terms of benefits such as longevity, bodily well-being, freedom and power, respect and self-respect, relationships with others and culture, including  knowledge in place of ignorance.

    Punishment is justified just in case in accords with this principle. Honderich shows -- or rather, forcefully reminds us -- that our own societies are indecent. The best-off hundredth of the population of the UK has about 25% of total wealth. In American the best-off tenth has had over 70%. The political power of the top tenths is thousands of times greater than that in the bottom tenths, partly because the richer have advantages they can buy, including lawyers. This translates into children who will remain sick or ignorant, gross unemployment, race discrimination in jobs and bad deaths for the old. It also explains socio-economic disparities in terms of crime; those with greater freedom and power have less risky ways of satisfying their desires than resorting to rape or assaulting someone who persistently harasses them. Our societies, which are indecent, are largely formed by our punitive systems. So our punitive systems are wrong. They are not rational means of keeping people out of bad lives. Instead they strengthen a society that mainly only improves already good lives, by preserving an unfair distribution of benefits. For example, punishment for offences against private property serves the end of those with power that seek to profit themselves.

    Also important is what our punitive systems do not punish, such as those that profitize what was public property, corporations like Bhopal that offend against public health, or a  prime minister who lies to his nation about the necessity of war. Honderich concludes that our penal systems serve a society whose good lives rest on many more bad ones.

    He ends by claiming that we must reform the nature of our societies in any rational way we can. We need more disrespect for the law and disdain for merely hierarchic democracy. We can turn to mass civil disobedience, organize boycotts, withdraw investments and refuse to pay taxes.

    Whether you agree with Honderich's conclusions, there is no doubt that this book would be invaluable for anyone who wants to start thinking seriously about what justifies punishment, not only because it surveys a high proportion of the classical literature but because it connects theories in broad yet subtle ways. It adroitly anticipates the reader's objections. It combines impressive breath with meticulous dissection of ideas. Its conclusion is refreshingly iconoclastic. It contains wit and a healthy contempt for politicians.

    The definition of punishment from which Honderich starts is admirably circumspect and qualified. I just wished that the argumentation wasn't so often of the same character.  Meaner and leaner would have helped me more in identifying inferences to sub-conclusions. Yet patience paid dividends: I got a lot more out of it on a second reading.

    Honderich says in the last chapter that there is another book that should be written about how actual systems of punishments in various countries fall short of the Principle of Humanity. I hope he writes it.

This review by John Williams, Professor in Philosophy, Singapore University, comes from Metapsychology, and appeared on 21 November 2006. For a review of the original edition of the book by Professor Jonathan Glover, go to Justifying Punishment.

HOME to T.H. website front page
HOME to Det & Free website