Punishment and Retribution

Leo Zaibert

• Publisher: Routledge
• Year: 2016
• ISBN:
  9781317073239
• Edition: 1
• Topic: Derecho
• Subtopic: Sentencias

Discussions of punishment typically assume that punishment is criminal punishment carried out by the State. Punishment is, however, a richer phenomenon and it occurs in many contexts. This book contains a general account of punishment which overcomes the difficulties of competing accounts. Recognizing punishment's manifoldness is valuable not merely in contributing to conceptual clarity, but in that this recognition sheds light on the complicated problem of punishment's justification. Insofar as they narrowly presuppose that punishment is criminal punishment, most apparent solutions to the tension between consequentialism and retributivism are rather unenlightening if we attempt to apply them in other contexts. Moreover, this presupposition has given rise to an unwieldy variety of accounts of retributivism which are less helpful in contexts other than criminal punishment. Treating punishment comprehensibly helps us to better understand how it differs from similar phenomena, and to carry on the discussion of its justification fruitfully.

Chapter One

Theories and Justifications

The debate regarding the justification of punishment, that is, the debate between retributivism and consequentialism, once appeared straightforward. I do not mean to suggest that the choices the debate forced upon us were ever easy (they have never been); my suggestion is rather that the distinction between the opposing alternatives was, more or less, conceptually straightforward. Traditional consequentialist justifications of punishment asserted, roughly, that punishment is justifiable only by the (good) consequences that follow from it, whereas retributivist justifications of punishment asserted, roughly, that punishment is justifiable only by its being deserved. Thus, to ascertain whether a justification of punishment was retributivist or consequentialist used to be relatively easy: for example, Kant and Hegel were, without a doubt, retributivists; similarly, Bentham and Sidgwick were, without a doubt, consequentialists. But recently all sorts of mixed justifications of punishment have sprung up, supposedly coherently combining retributive and the consequentialist rationales.

Many authors refer to these mixed justifications as “mixed theories”. Flew’s admonition, to which I have referred in the introduction, has not been heeded: most people continue to refer to theses related to justifications of punishment as “theories” of punishment. While in a sense this is a minor terminological point, there is another sense in which it is important to emphasize a noteworthy difference between “theories” and “justifications” of punishment. A theory of punishment seeks to tell us what punishment is, what the necessary and sufficient conditions for something to be punishment are, how punishment relates to and how it differs from related phenomena, and similar questions. A justification of punishment, on the other hand, seeks to tell us when it is morally (or politically, or in any other normative way) legitimate to inflict punishment. The famous mixed “theories” of punishment do not even try to answer theoretical sorts of questions; instead, they seek to reconcile two opposing ways of justifying punishment (retributivism and consequentialism), that is, two different sets of reasons why punishment should be inflicted. For these reasons, I favor referring to “mixed theories” of punishment as “mixed justifications” of punishment.

Early in this book, I shall be interested both in putting forth a theory of punishment, and then, later, in discussing the problem of the justification of punishment. Yet, before presenting the details of my own theory of punishment, I would like to devote some attention to showing why mixed justifications of punishment fail. Their failure, I will argue, is inseparable from a problem in the definition of punishment that they presuppose. My initial foray into the discussion of the (mixed) justifications of punishment is geared mainly at showing this single specific point: the failure of the mixed justifications has a lot to do with a problem regarding the definition of punishment. My plan, then, is to discuss first the problems with the mixed justifications, then to present my own definition of punishment, and finally to come back to the thorny problem of the justification of punishment, but only after taking seriously the manifoldness of punishment.

The relationship between a theory of punishment and the justification of punishment is subtler than it might seem on first approximation. An important thesis I shall defend in this book is that we are well advised to try and keep the definitional and the justificatory enterprises as separate as possible. There are two main reasons why I preface my discussion of the definition of punishment with a consideration of the failure of the mixed justifications of punishment. First, since both retributive and consequentialist rationales are persuasive, it might be thought that a justification of punishment which combined these two would be all the more persuasive – the “best of both worlds” sort of scenario. But none of the existing mixed justifications succeeds in this syncretic fusion, and it is important to show this at the outset. Since later in the book I shall come back to the problem of the justification of punishment, it serves me well to show at once why I discard as viable options precisely those positions which seem, to many authors, to be so poignantly attractive. Second, the talk of definition itself has fallen, for different reasons, into such disrepute, that I would like to convince even those who are suspicious of definitions, that here we have a case in which a good definition would go a long way towards helping to solve a difficult practical problem. A good definition of punishment shall, at least, direct us down the right path in trying to understand what justifies it.

The attractiveness of mixed justifications of punishment is so great that contemporary punishment theorists doubt that there remain any authors who could be described as embracing only retributivism or only consequentialism in anything like the standard, traditional articulations of those views I have just sketched. Ted Honderich for example, as he begins a chapter in Punishment: The Supposed Justifications (entitled “Compromises”), expresses the fact that, in his opinion “there no longer are defenders of the traditional retribution theory, or at least the version that we are obliged rather than permitted to punish offenders because they deserve it”. Honderich further claims that “the traditional deterrence view is also in decline, for different reasons, if not so abandoned as the view that punishment is justified by reformative effects [both consequentialist justifications of punishment]”.1 Similarly, H. L. A. Hart begins his Prolegomenon to the Principles of Punishment with the following assessment of the current state of punishment theory:

General interest in the topic of punishment has never been greater than at present and I doubt if the public discussion of it has ever been more confused. The interest and the confusion are both in part due to relatively modern scepticism about two elements which have figured as essential parts of the traditionally opposed “theories” of punishment. On the one hand, the old Benthamite confidence in fear of the penalties threatened by the law as a powerful deterrent, has waned with the growing realization that the part played by calculation of any sort in anti-social behavior has been exaggerated. On the other hand a cloud of doubt has settled over the keystone of “retributive” theory.2

While the years following the publication of Honderich’s and Hart’s books have witnessed a veritable retributivist revival,3 and there are famous retributivists – most notably Michael Moore – who seem to make no concessions to consequentialism, and who assert that desert obliges us to punish offenders, most of the recent retributivists are, as Honderich and Hart point out, retributivists only in some newer sense. Something similar happens with contemporary consequentialists.

I agree with the spirit behind Honderich’s and Hart’s remarks, that is, I believe that the boundary between retributivism and consequentialism has become blurred. It is of course interesting that only very recently do we find the first systematic, self-conscious attempts to coherently combine the retributive and the consequentialist rationales: the sophisticated “mixed justifications of punishment” are creatures of the twentieth century – and mostly creatures conceived by analytic philosophers inebriated with Oxford’s ordinary language philosophy and its concomitant fascination with logic. Yet, Honderich and Hart, along with most punishment theorists, restrict their investigation, for all practical purposes, to one single manifestation of punishment: criminal punishment carried out by the State, and I will argue that this is problematic.

I am not the first person to object to these mixed justifications of punishment, of course. After all, as one would expect, both partisan defenders of retributivism and partisan defenders of consequentialism would be naturally opposed to the mixed justifications in that such justifications, in their opinion, smuggle unacceptable consequentialist or retributivist elements which they are simply unwilling to accept. In other words, for a hard-core, single-minded retributivist the mixed justification might concede too much to consequentialism, and for a hard-core, single-minded consequentialist the concessions to retributivism might be similarly unacceptable. Since I am neither a partisan retributivist nor a partisan consequentialist, the sort of objection that I wish to level against the mixed justifications of punishment is different from the typical objections: I will here claim that one important and overlooked reason why the mixed justifications fail is that they presuppose an inconveniently narrow definition of punishment. Those endorsing mixed justifications of punishment, following a general trend, overwhelmingly assume that punishment is criminal punishment carried out by the State. But I will argue that any candidate definition of punishment must do justice to the fact that there are instances of punishment which occur outside of State institutions.

One would wish that at least one of the mixed justifications would be successful, not only given the inherent importance of the problem of punishment, but also given the obvious appeal of both retributivist and consequentialist rationales. But if we pay attention to punishment which occurs in these other much-ignored (non-State) contexts, we will clearly see the deficiencies of the mixed justifications. I will argue that independently of whether the mixed justifications of punishment may have attained some success within the context of criminal punishment carried out by the State, they are utterly unsuccessful in shedding any light whatsoever over the problem of punishment in other contexts. And it is simply not true that the tension between retributivism and consequentialism obtains only within the context of the State; in fact, this tension is probably as old as humanity itself, and surely the State is not that old.

Whenever an ordinary person deliberates about whether or not to punish another person, she struggles with the appeal of retributive rationales and with the appeal of consequentialist rationales, and she will look for ways of having the best of both worlds. And unless this person is a sovereign, or an agent of a sovereign, deliberating about an instance of State punishment, the mixed justifications would have very little to tell her. Thus, the first two sections of this chapter are devoted to the discussion of the mixed justifications of punishment and to their failure. In the last two sections of the chapter I present my own account of punishment, emphasizing how it differs from any existing account, and how it accommodates non-State punishment.

The Rise and Fall of the Mixed Justifications of Punishment

In somewhat of a contemporary rendition of the simultaneous discovery of infinitesimal calculus by Leibniz and Newton, Anthony Quinton in On Punishment and John Rawls in Two Concepts of Rules simultaneously “discovered” a way of reconciling consequentialism (which they both misleadingly called “utilitarianism”)4 and retributivism. True, as Rawls claims, there are some differences between the two articles, but they nevertheless remain strikingly similar.5 In fact, most mixed justifications of punishment (not only these two) conspicuously exhibit a pair of problematic maneuvers:

(1) An appeal to logic as the solution to the tension between retributivism and consequentialism.

(2) An appeal to institutions as the solution to the tension between retributivism and consequentialism.

In his famous article, Quinton set out to resolve the “prevailing antinomy about the philosophical justification of punishment”, that is the antinomy confronting “the two great theories[:] retributiv[ism] and utilitarian[ism]”.6 Quinton’s solution to the antinomy is, in spite of its simplicity, far-reaching:

retributivism, properly understood, is not a moral but a logical doctrine, and […] it does not provide a moral justification of the infliction of punishment but an elucidation of the use of the word.7

This passage contains, in embryo, the kernel of not only Rawls’ mixed justifications of punishment, but also Hart’s – which I will discuss below. I shall dub this “the two-question strategy”.

Quinton’s solution to the antinomy consists, then, in claiming that retributivism and utilitarianism are responses to two different questions. Retributivism is relevant in connection to the question “when (logically) can we punish?”, and “utilitarianism” is relevant in connection to the question “when (morally) may we or ought we punish?”8 The essential strategic move that allows Quinton to postulate such a radical thesis is a peculiar definition of retributivism, according to which all that retributivism amounts to is the claim “that punishment is only justified by guilt”.9 This is a form of what has been termed minimalist retributivism, insofar as it reduces retributivism to a rather humble thesis.

The logicist approach to punishment and its justifications exhibited by Quinton (and as we shall see also by Rawls and Hart, amongst many other mixed “theorists”) is further overhauled by a strong emphasis upon the institutional aspects that frame crim...