* I wish to thank Kai Nielsen for reading an early draft of this essay and for many helpful suggestions and comments.

1 Anthony, Quinton “On Punishment.” Analysis, Vol. 14, No. 6 (June, 1954)Google Scholar.

2 Cf. Antony, Flew “The Justification of Punishment,” Philosophy. Vol. 29. October, 1954Google Scholar.

3 The first five derive from Antony Flew, op. cit. The sixth is suggested by Benn, S. I. and Peters, R. S. Social Principles and the Democratic State (London: Allen & Unwin, 1959), p. 174Google Scholar.

4 It may be argued that the whole notion of necessary and sufficient conditions for use of a word is misguided especially since we are told with respect to “punishment” that it is both vague and “open-textured.” (Cf. Flew, op. cit.) However, though this reminder may serve to rule out many glib counter case objections to the proposed criteria, we should be careful not to allow it to paralyze all critical evaluation through anticipation of the retort “but that is not the standard case,” or “but that is a secondary or extended use.” First, we would have to be given a good reason why such uses are non-standard and/or secondary apart from the fact that they fail to meet the proposed criteria. And second, it is not clear why secondary and extended uses are not here philosophically relevant.

5 An interesting question is why it is not necessary that there be actual unpleasantness for an act to qualify as punishment. My suggestion is that we regard as punishment, both in the law and without, only those penalties which, as a general rule, fulfil certain requirements among which is the involvement of evil, unpleasantness or deprivation of good. Both in legal and in less formal systems of rules one must, in the interests of practicability, address himself to what is generally found most effective in furthering the ends of legislation or rule making. Specifically with respect to penal law, one cannot countenance individual or idiosyncratic dislikes and reactions to particular punishments. The exigencies of legal enforcement require that penalties be attached to crimes in such a way that only the unpleasantness which is usually found to be associated with a given penalty be reckoned. Account cannot be taken of the possibility that a particular offender will find the penalty for his offense desirable for there is no theoretical limit to such possibilities. What is crucial is that penalties for offenses be of a kind which are found unpleasant given the usual reactions of men—the only criterion which affords some predictability from the standpoint of either formal or informal legislation. (The assumption is, of course, that offenders who come before the law will not deviate so much from the usual in their likes and dislikes that the penalties which the law has attached to various offenses will become pointless.)

6 Antony Flew, op. cit. p. 293.

7 Benn, S. I. and Peters, R. S. The Principles of Political Thought: Social Foundation of the Democratic State (New York: Free Press of Glencoe, 1959), Chapter 8, p. 211–12Google Scholar. Originally appearing under the title Social Principles and the Democratic State (London: George Allen & Unwin, 1959).

8 Antony Flew, op. cit., p. 293. Emphasis added.

9 The suggestion that punishment is tied to the acknowledgment or at the least to presumption of guilt rather than guilt per se is elaborated by Kurt, Baier (“Is Punishment Retributive?” Analysis, Vol. 16, No. 2 (December 1955)Google Scholar).

This suggestion is an improvement over the traditional view (that “punishment” is a felicitous description only of suffering inflicted on the known guilty) both in its recognition of the latitude of our use of “punishment” to cover cases of presumed but not necessarily established guilt, and the understanding this betokens for the type of criticism to which utilitarian ethics seems open.

10 S. I. Benn and R. S. Peters, op. cit., p. 212.

11 Cf. English Defense Regulation (18b).

12 Benn, S. I. “An Approach to the Problems of Punishment.” Philosophy. XXXIII (October 1958). 325–41CrossRefGoogle Scholar.

13 Oddly enough, both the suggestion of the defense detention case as one which fails to be punishment and the proposal of intended as opposed to incidental hardship as a criterion of punishment are put forth by Benn and Peters. However, they fail to see any connection between the two.

14 McCloskey, H. J. “The Complexity of the Concept of Punishment,” Philosophy, October 1952, pp. 321–322.Google Scholar

15 I shall ignore the contention that talk of collective punishments is, for our purposes, not illuminating since “collective punishment” is somehow metaphorical or secondary. (Cf. Flew, op. cit., p. 295.)

It is not clear to me why we should regard collective punishments as metaphorical or secondary even given the very loose definition of metaphor (offered by Ogden and Richards) that “Whenever a term is . . . taken outside the universe of discourse for which it has been defined, it becomes a metaphor and may be in need of fresh definition.” For even Flew admits that “. . . the original unit on which punishment was inflicted was not the individual but the family, tribe, village, clan or some other group.”

Cf. Ogden, C. K. and Richards, I. A.. The Meaning of Meaning (New York: Harcourt, Brace and World, Inc., 1923), p. 111, and Flew, op cit., p. 295Google Scholar.

16 Flew, op. cit., p. 293.

17 Unfortunately, talk about what distinguishes one sense from another is, from what I can find in the literature, somewhat lackinn in linguistics. To he sure, there is much discussion concerning criteria of non-literal and metaphorical uses and of what distinguishes these from literal or ‘straight’ uses. There is also, unquestionably, much excellent work dealing with word usage from the standpoint of a primative/ derivative classification, i.e., from the standpoint of historical priority. (Cf. for example, Michael Breal's Semantics: Studies in the Science of Meaning, translated by Mrs. Henry Ceest (New York: Dover Publications, Inc., 1964), esp. chapters X, XI, XII, XIV and XV.)

Generally speaking, however, there seems little concern with those features of a term in virtue of which it is classified as secondary in a sense which is indicative neither of historical dependence nor of non-literal meaning. (At least one exception is the work which has sometimes emerged as the by-produce of philosophers’ discussions concerning the analysis of certain concepts. J. C. B. Gosling is a good case in point. While analysing the related concepts of desire, pleasure, and wanting. Gosling comes up with some interesting and often overlooked points about the various distinctions that can be marked of between uses of a term. Of especial relevance here is his insight that a variety ofterms is employed to distinguish senses of a term as primary and secondary, among which are central/ peripheral, basic/derivative, original/developed, straight/extended, standard/parasitic. Although the different sets here are sometimes used interchangeably, not all these contrasts mark off the same distinctions: Some are etymological and have to do with which words, as a matter of linguistic fact, were the first to be employed. Some are non-contingent, i.e., where one is applicable the other sense is presupposed. And some may reflect that conceptual connection of priority whereby if the one term is applicable then there must in fact be cases where the other term is applicable but not vice-versa. The conclusion is thus that classifications of terms as “secondary” are not univocal. Each such classification must be examined independently.

Cf. Gosling, J. C. B. Pleasure and Desire (The Case for Hedonism Reviewed) (Oxford: Clarendon Press, 1969), pp. 114–126CrossRefGoogle Scholar.

18 Thus, in appealing to other languages to assess whether there harbors in our own a secondary or non-literal sense of a term it is well to heed Breal's remarks that

Semantics: Studies in the Science of Meaning, op. cit., pp. 131–132.

19 Cf. Robert, A. Samek “Punishment: A Postscript to Two Prolegomena” in Philosophy, Volume XLI (1966), p. 220Google Scholar.

20 I am thankful to Sidney Morgenbesser for helpful discussion here.

21 From three lectures given at Harvard University in the William James Lecture Series 1966’ also Cf. Grice, H. P. “The Causal Theory of Perception,” Proceedings of the Aristotelian Society, Supplementary Vol 35 (1961)Google Scholar and “Meaning,” The Philosophical Review, July 1957 (Volume LXVI, No. 3), pp. 377–388.

22 I am grateful to John King-Farlow for this example. His comments, questions and examples have been very suggestive and helpful to me.

23 Antony Flew, op. cit.. p. 295.

In fairness, it should be noted that not all writers on punishment who focus on legal or quasi-legal contexts do so ‘under cover.’ Hart, especially, makes quite clear his restriction of whal he calls standard cases of punishment to punishment under a legal system. Cf. Hart, H. L. A. “Prolegomenon to the Principles of Punishment,” Proceedings of the Aristotelian Society. Vol. 60 (1959-60}Google Scholar.

(To be sure, the attraction of the view that the concept of punishment entails those features present primarily in its use in legal contexts is due to the fact that we tend to consider punishment chiefly in the context of legal systems. But we should not be misled here. For it is too easy to think of cases of punishment which do not conform to this model. One writer asks us to think of such informal punishments as kicking and punching, treating with coldness, coolness, aloofness, cutting dead, not speaking to, showing up, sending to bed without supper, tor such varying ‘offenses’ as rudeness, jumping the queue, presu-ptousness, conceit, cowardice, bad manners, bed-wetting, being a nuisance, etc. It will not do to dismiss these cases as fringe cases, for the question arises why we treat them as cases of punishment at all. It is surely not because they resemble cases of punishment under a legal system.) (Samek, op. cit., p. 220.)

24 Cf. Samek, ibid., p. 221.

25 J. D. Mabbott, “Punishment,” op. cit., p. 42.

26 Flew, op. cit., p. 299-300.

27 An example of this is found in Kurt Baier's claim that the necessary preliminary conditions of the ‘game’ of punishment are rule-making, penalization, finding guilty of a breach of a rule and pronouncing sentence, conditions which are marks of a legal or quasi-legal system of punishment and not of the concept of punishment. Cf. K. Baier, Is Punishment Retributive?, op. cit.

28 Ibid., p. 28.

It is interesting that Baier is able to see the error of conflating the defeat of a system of punishment with a change in the meaning of the concept when he himself offers as criteria for the concept of punishment criteria more appropriate to punishment as a system.