Encyclopedia of Educational Philosophy and Theory

2017 Edition
| Editors: Michael A. Peters


Reference work entry
DOI: https://doi.org/10.1007/978-981-287-588-4_362

The Justifications of Punishment

Analytic philosophy tends to separate the two main philosophical questions concerning punishment – those of meaning and justification. They tend to be separated and treated almost independently of one another. In this section we will follow this assumption. There are several traditional but opposed justifications of punishment, but, it will be argued*, none fit easily with the punishment of children (Marshall 1984). Here we will look at these traditional justifications, irrespective of whether or not they apply to children.

The retributive theory can be traced at least to the Old Testament adage of “an eye for an eye and a tooth for a tooth.” As the term “retributive” suggests, this involves a demand that punishment involves a paying back by the offender for his offense, and this notion of retribution is part of the justification of punishment. If the payment demanded by the Old Testament adage appears harsh and vindictive, the payment demanded by modern retributivists may not need to be as strong as that demanded by that adage. But, as the theory’s detractors often adduce, it is a “paying” back and is not concerned about the future behavior of the offender. In modern times, however, the recipient of any retribution has shifted from the victim to the State, though recent discussions on crime and punishment have centered on the plight of the victims of crime and how they personally are to be recompensed. Retribution then is backward looking, to the offense, and not forward looking and does little for the victim.

It is almost as if there is a moral accounts book, in which good and evil are entered on each side of the book in credit and debit pages. To keep the moral accounts balanced, the amount of evil brought into the world by a crime can only be alleviated by some credit of good, i.e., retribution or paying back by the offender. The moral accounts book must be kept balanced.

In less stern terms, we can see the retributive theory as holding that offenders need punishment and that justice is served on this account. Some retributivists however hold a stronger position that punishment is intrinsically good, i.e., that it is a good in itself and therefore in need of no further justification.

The deterrent theory (sometimes referred to as a utilitarian theory) is the major alternative to retribution as a justification for punishment. Its strengths are to be found in the differences which it has from retributive theories. While admitting that punishment is essentially an evil, the deterrent theory claims that it is justified because its purpose is to deter the offender, and others, from committing the same or similar offenses in the future. Thus it is forward looking and in opposition to the retributive theory which is backward looking. In its forward-looking aspect, it is concerned with the behavior of the offender and with changing that behavior in the future. This is deterrence’s strength in my view. At first sight then, this forward-looking aspect may have some attraction for a search for the justification of punishing children.

There are a number of problems associated with the deterrent theory (see, e.g., Acton 1963). First a deterrent theory of justification may also justify the punishment of innocent people so that others might be deterred. Here the objection might be that it is only the guilty who can be punished and that the innocent should not be punished. If this is part of the definition of punishment*, then it would appear that the infliction of pain should be described in some other way, as murder say, or as some other form of unlawful killing. Hence the “punishment” of the innocent cannot be justified as punishment because it is not punishment at all.

But perhaps it is the threat of punishment which deters and not merely punishment per se. First, there is a presupposition that people want to, or are likely to, or are disposed to do X. But for many people, the thought, desire, and any plan to do X just do not arise. X is not the sort of thing in which they engage. Second, in the notion of deterrence, there is an implied threat – if you do X, then Y will follow. Furthermore the deterrent theory is concerned not just with past offenders but with all people; for the punishment of one offender is meant to be a deterrent to others in that there is a threat for all others which accompanies the actual punishment of a particular offender. But why do we wish or need to threaten people in this manner, at all, especially for those to whom the thought of X-ing does not occur? Why do we treat the offender (who may be truly contrite and determined to change his/her ways) and the innocent others in this manner?

Sometimes punishment is said to be justified because it reforms people, but usually this putative justification of punishment receives short shrift in the philosophical literature; reform is reform and punishment is punishment. In other words the concepts of punishment and reform are claimed to be logically distinct. Reform implies a serious straying from the path and the need to be redirected, or to redirect oneself considerably. But at first sight this justification of punishment has its attractions.

But if it is meant to be a general justification for punishment, the way in which the legal institution of punishment works, especially in prisons, is hardly conducive to reform. Furthermore reform, as a justification, does not seem to work very well. In order to justify punishment because it reforms, there would need to be greater success with punishment as a reform than seems to be the case in actual fact.

Usually then reform receives short shrift as a justification for punishment. It would be nice if punishment did reform but it doesn’t. At best reform is a contingent bonus. Philosophically, reform is reform and is not identical with punishment.

Finally, A. C. Ewing (1929) talks about an educative function of punishment. He argues that punishment helps “din the words in,” i.e., that if the rational explanations of why something involves a crime or an intolerable state of affairs are accompanied by some form of punishment, then this may assist in the explanations being understood or accepted. But as Ewing’s examples seem to involve the young and their moral education, this account will be discussed elsewhere (consult other section on punishment).

The Legal Model of Punishment

In this section I discuss the legal model of punishment, sometimes known as the Flew-Benn-Hart model after three of its proponents. However there are conceptual problems in the Flew-Benn-Hart model which make its application to children problematic (Marshall 1972, 1975), and the traditional justifications of punishment do not seem to apply to children either (Marshall 1984). Why there are problems over the meaning of “punishment” and its justification in the case of children will be discussed in another section.

In Anglo-American philosophy of education, the analytic approach dominated for nearly two decades from the early 1960s. The account of punishment offered by philosopher of education R. S. Peters in Ethics and Education (1966) is analytic, follows the Flew-Benn-Hart model, and has been accepted, essentially, by a substantial number of analytic philosophers of education. It results however in a number of philosophical paradoxes for talk of the punishment of children. These are caused by taking from general philosophy analytic models of the meaning and justification of punishment in legal and adult cases and applying them to young people. Peters’ work is used to illustrate this general analytic position on punishment, because it is readily available, and has set an important general framework for discussion and debate on the meaning of “punishment” and its legitimization in philosophy of education.

Those philosophers who have been interested in punishment have tended to concentrate upon two major questions, concerned with meaning and justification. The following questions have almost come to dominate philosophical literature: “what is the meaning of ‘punishment’?” and “how is punishment to be justified?” The selection of papers edited by H. B. Acton in his important collection, The Philosophy of Punishment (1963), illustrates this point well. Our concern here is with the first question.

The Anglo-American philosophical literature generally concerns itself with a particular model of punishment. This (legal) model is presented, e.g., by H. L. A. Hart in Punishment and Responsibility (1968), as an answer to these questions of meaning and justification. As Hart’s work draws upon earlier work of Antony Flew and Stanley Benn, the model is sometimes referred to as the Flew-Benn-Hart model of punishment. And as it was essentially to be adopted by R. S. Peters in his enormously influential writings on philosophy of education (1966), it might also be referred to here as the Flew-Benn-Hart-Peters model of punishment.

Hart (1968) says that he is merely drawing upon “recent admirable work scattered through … philosophical journals.” That Hart specifically added the qualifier “English” to his list of journals need not be of too much concern. If this does represent a certain insularity or philosophical myopia, Hart was probably correct at that time that there was little need to go beyond this literature in English-speaking philosophy (see, in particular, the edited collection by Acton). However, it should be added that he is also writing from within an established legal tradition and with more than merely an analytic methodology as he relates his approach to that of Locke’s discussion of property.

Drawing upon Flew (1954) and Benn (1958), he says that he will define the standard case of the concept of punishment as containing five necessary elements or conditions for the correct application of the concept “punishment.” Hart argued that (1968, p. 4 f.):
  1. 1.

    It must involve pain or other consequences normally considered unpleasant.

  2. 2.

    It must be for an offense against legal rules.

  3. 3.

    It must be of an actual or supposed offender for his offense.

  4. 4.

    It must be intentionally administered by human beings other than the offender.

  5. 5.

    It must be imposed and administered by an authority constituted by a legal system against which the offense is committed.


Hart then drew a distinction between standard cases and substandard or secondary cases. For the accolade of standard to be applied, all five conditions listed above had to be met. Substandard cases were illustrated by the following cases, or possibilities: pain or consequences for breaching other than legal rules – here he gives as specific examples, the family and the school; by other than authoritative officials; and unpleasantness or pain imposed deliberately by authorities but upon non-offenders. Hence the paradox for educators over the meaning of “punishment” for how can we talk meaningfully, in more than a substandard sense of the term “punishment” and more than metaphorically, about the punishment of children? If we accept Hart’s position, then the use of “punishment” in relation to young people in home or school is a substandard case of the use of the concept.

This standard account is to be found, essentially, in R. S. Peters’ Ethics and Education (1966). Peters also adopts Hart’s fourfold division of meaningful questions about punishment. These are said to be (Hart 1968, p. 4) questions of definition, justification, and distribution, with the latter divided into questions of who should be punished, i.e., entitlement, and the form and severity. Peters, however, sees the first two questions only as being philosophical questions, with the remaining two being the province of jurists and administrators. In effect then, in Peter’s account, we have a philosophical division of labor, with the efforts of philosophers directed at the first two questions about punishment – definition and justification – and the relegation of the last two questions to the status of administrative, juridical, or, in Peters’ case, educational questions. If Peters’ particular account of punishment did not meet with universal approval (see, e.g., Wilson 1971; John Wilson 1977), nevertheless, the model sets the form of the debates that ensued in philosophy of education.


  1. Acton, H. B. (Ed.). (1963). The philosophy of punishment. London: Macmillan.Google Scholar
  2. Benn, S. F. (1958). An approach to the problems of punishment. Philosophy, 33, 325–341.CrossRefGoogle Scholar
  3. Ewing, A. C. (1929). The morality of punishment. London: Kegan Paul, Trench, Trubner.Google Scholar
  4. Flew, A. (1954). The justification of punishment. Philosophy, 29, 291–307.CrossRefGoogle Scholar
  5. Hart, H. L. A. (1968). Punishment and responsibility. Oxford, UK: Clarendon.Google Scholar
  6. Marshall, J. D. (1972). On why we don’t punish children. Educational Philosophy and Theory, 4(1), 57–68.CrossRefGoogle Scholar
  7. Marshall, J. (1975). Punishment and education. Educational Theory, 24, 148–157.CrossRefGoogle Scholar
  8. Marshall, J. (1984). Punishment and moral education. Journal of Moral Education, 13, 79–85.CrossRefGoogle Scholar
  9. Peters, R. S. (1966). Ethics and education. London: Allen & Unwin.Google Scholar
  10. Wilson, P. S. (1971). Interest and discipline in education. Henley: Routledge/Kegan Paul.Google Scholar
  11. Wilson, J. (1977). Philosophy and practical education. London: Routledge/Kegan Paul.Google Scholar

Copyright information

© Springer Science+Business Media Singapore 2017

Authors and Affiliations

  1. 1.The University of AucklandAucklandNew Zealand