(1) Some have argued that its solution bears directly on anthropological practice (see, e.g., Winch, P., Understanding a Primitive Society, American Philosophical Quarterly, I (1964), 307–324Google Scholar, where Evans-Pritchard's account of witchcraft among the Azande is held to be partly vitiated by his supposedly mistaken answer to it). I agree with this position, but in this paper I do not seek to substantiate it.

(2) Leach, E., Political Systems of Highland Burma (London 1954), pp. 13–14.Google Scholar

(3) Ibid. p. 182.

(4) Firth, R., Essays on Social Organisation and Values (London 1964), p. 237.Google Scholar

(5) Ibid. pp. 238–239.

(6) See Beattie, J., Other Cultures (London 1964)Google Scholar, chs. V and XII, and Id., Ritual and Social Change, Man: The Journal of the Royal Anthropological Institute, I (1966), 60–74.Google Scholar

(7) Beattie, J., loc. cit. (1966), p. 68.Google Scholar Thus, magic is “the acting out of a situation, the expression of a desire in symbolic terms; it is not the application of empirically acquired knowledge about the properties of natural substances” (Beattie, , op. cit. (1964), p. 206).Google ScholarCf. Parsons, T., The Structure of Social Action (New York and London 1937), p. 431 (2nd edition 1949)Google Scholar: “Ritual actions are not […] either simply irrational, or pseudo rational, based on prescientific erroneous knowledge, but are of a different character altogether and as such not to be measured by the standards of intrinsic rationality at all” (cited in Beattie, , loc. cit. 1966).Google Scholar Parsons wrongly attributes this position to Durkheim: as I shall show, Durkheim did not see religion as merely symbolic.

(8) Beattie, (1966), p. 72.Google Scholar For Beattie magic and religion “both imply ritual, symbolic ideas and activities rather than practical, ‘scientific’ ones […]” (Id. (1964), p. 212). For an example of the procedures Beattie advocates, v. Turner, V., “Symbols in Ndembu Ritual” in Gluckman, M. (ed.) Closed Systems and Open Minds (Edinburgh 1964), pp. 20–51.Google Scholar

(9) Beattie appeals to the authority of Suzanne Langer (Beattie, Ritual and Social Change, loc. cit. p. 66), but I am unsure how far his allegiance to her views goes. I do not know whether he would wish to argue, as she does, that rationality and even logic can be ascribed to expressive symbolism and whether he would subscribe together general view that “[r]ationality is the essence of mind and symbolic transformation its elementary process. It is a fundamental error, therefore, to recognise it only in the phenomenon of systematic, explicit reasoning. That is a mature and precarious product. Rationality, however, is embodied in every mental act […]” (Id., Philosophy in a new Key (Harvard 1942, p. 99Google Scholar; 3rd edition 1963). Miss Langer's is in any case a special sense of ‘rationality’. As I hope to show, the fundamental meaning of rationality is essentially linked to the phenomenon of systematic, explicit reasoning.

(10) Lore, Maori Medical, Journal of Polynesian Society, XIII (1904), p. 219Google Scholar, cited in Lévy-Bruhl, L., Les fonctions mentales dans les sociétés inférieures (Paris 1910), p. 69 (2nd edition 1912).Google Scholar

(11) C. G. and Seligman, B. Z., Pagan Tribes of the Nilotic Sudan (London 1932), p. 25Google Scholar, cited in Evans-Pritchard, E. E., Lévy-Bruhl's Theory of Primitive Mentality, Bulletin of the Faculty of Arts, II (1934), 1–36.Google Scholar

(12) Evans-Pritchard, E. E., The Intellectualist (English) Interpretation of Magic, Bulletin of the Faculty of Arts, I (1933), 282–311.Google ScholarCf. also Id., Theories of Primitive Religion (Oxford, 1965), Ch. II.Google Scholar

(13) Cf. Leach, E., Frazer and Malinowski, Encounter, XXV (1965), 24–36Google Scholar: “For Frazer, all ritual is based in fallacy, either an erroneous belief in the magical powers of men or an equally erroneous belief in the imaginary powers of imaginary deities” (p. 29).

(14) Horton, R., African Traditional Thought and Western Science, Africa, XXXVII (1967), 50–71 and 155–187.CrossRefGoogle ScholarCf. also Id., Destiny and the Unconscious in West Africa, Africa, XXXI (1961), 110–116Google Scholar; The Kalahari World View: an Outline and Interpretation, Ibid. XXXII (1962), 197–220; Ritual Man in Africa, Ibid. XXXIV (1964), 85–104. (For a symbolist critique of Horton, v. Beattie, Ritual and Social Change, loc. cit.). For other ‘neo-Frazerian’ writings, v. Goody, J., Religion and Ritual: the Definitional Problem, British Journal of Sociology, XII (1961), 142–164CrossRefGoogle Scholar; Jarvie, I. C., The Revolution in Anthropology (London 1964)Google Scholar; Jarvie, I. C. and Agassi, J., The Rationality of Magic, British Journal of Sociology, XVIII (1967), 55–74.CrossRefGoogle Scholar

(15) Horton, , loc. cit. pp. 50–71.Google Scholar

(16) Ibid. p. 53.

(17) Ibid. p. 58.

(18) Ibid.

(19) Ibid. pp. 155–156.

(20) Durkheim, E., Les formes élémentaires de la vie religieuse (Paris 1912), pp. 339–341.Google Scholar

(21) V. Les Carnets de Lucien Lévy-Bruhl (Paris 1949)Google Scholar, passim; where it is made explicit and partially resolved.

(22) It is worth noting that Durkheim differed crucially from Lévy-Bruhl, emphasising the continuities rather than the differences between primitive and modern scientific thought: v. Durkheim, E., Les formes élémentaires de la vie religieuse, op. cit. pp. 336–342Google Scholar, and Review of Lévy-Bruhl, L., Les fonctions mentales dans les sociétés inférieuresGoogle Scholar, and Durkheim, E., Les formes élémentaires de la vie religieuse, in Année sociologique, XII (1913), 33–37.Google Scholar

(23) L. Lévy-Bruhl, A Letter to Evans-Pritchard, E. E., British Journal of Sociology, III (1952), 117–123.Google Scholar

(24) Lévy-Bruhl, L., Les fonctions mentales dans les sociétés infirieures (Paris 1910), p. 30.Google Scholar

(25) Ibid. pp. 30–31.

(26) He eventually abandoned it: v. Les Cornets de L. Lévy-Bruhl, op. cit. pp. 47–51, 60–62, 69–70, 129–135, etc.Google Scholar

(27) Lévy-Bruhl, L., La mentalité primitive (Herbert Spencer Lecture) (Oxford 1931), p. 21.Google Scholar

(28) E. E. Evans-Pritchard, Lévy-Bruhl's Theory of Primitive Mentality, loc. cit. Lévy-Bruhl's general endorsement of this article is to be found in Lévy-Bruhl, “A Letter to E. E. Evans-Pritchard”, loc. cit.

(29) Evans-Pritchard, E. E., Lévy-Bruhl's Theory…, loc. cit.Google Scholar

(30) Ibid.

(31) Ibid.

(32) Lévy-Bruhl, , Les fonctions mentales…, op. cit. p. 77.Google Scholar

(33) This position he did not abandon: v. Les Cornets de L. Lévy-Bruhl, op. cit. (e.g. pp. 193–198)Google Scholar, where it is strongly reaffirmed.

(34) Evans-Pritchard, E. E., “Lévy-Bruhl's Theory of Primitive Mentality”, loc. cit.Google Scholar

(35) Ibid.

(36) Lévy-Bruhl, L., “A Letter to E. E. Evans-Pritchard”, loc. cit. p. 121.Google Scholar

(37) Evans-Pritchard, E. E., “Lévy-Bruhl's Theory…”, loc. cit.Google Scholar

(38) Ibid.Cf. Les Carnets…, op. cit. p. 61Google Scholar, where he recalls that he had begun from the hypothesis that societies with different structures had different logics. The theory of the ‘prelogical’ was a modified version of this hypothesis, which he only finally abandoned much later, when he came to Understand that “the logical structure of the mind is the same in all known human societies” (Ibid. p. 62).

(39) Lévy-Bruhl's final position was as follows: “there is no primitive mentality which is distinguished from the other by two characteristic features (being mystical and prelogical). There is one mystical mentality that is more marked and more easily observable among ‘primitives’ than in our societies, but present in every human mind”. (Les Cornets…, p. 131.)Google Scholar

(40) Winch, P., Understanding a Primitive Society, American Philosophical Quarterly, I (1964), 307–324.Google Scholar

(41) Gellner, E., Concepts and Society, Transactions of the Fifth World Congress of Sociology (Washington 1963), I (1962), 153–183Google Scholar; and MacIntyre, A., “Is Understandhold ing Religion compatible with Believing?” in Hick, J. (ed.), Faith and the Philosophers (London 1964).Google Scholar

(42) Evans-Pwtchard, E. E., Witchcraft, Oracles and Magic among the Azande (Oxford 1937), p. 63.Google Scholar

(43) Winch, P., Understanding a Primitive Society, loc. cit. p. 308.Google Scholar

(44) Ibid. p. 309.

(45) Ibid.

(46) Witchcraft…, op. cit. pp. 24–25.Google Scholar

(47) Understanding a Primitive Society, loc. cit. p. 315.Google Scholar

(48) Ibid.

(49) The philosophical basis for this position is to be found in Winch, P., The Idea of a Social Science and its Relation to Philosophy (London 1958). Cf. in particular the following passage: “[…] criteria of logic are not a direct gift of God, but arise out of, and are only intelligible in the context of, ways of living and modes of social life. It follows that one cannot apply criteria of logic to modes of social life as such. For instance, science is one such mode and religion is another; and each has criteria of intelligibility peculiar to itself. So within science or religion, actions can be logical or illogical: in science, for example, it would be illogical to refuse to be bound by the results of a properly carried-out experiment; in religion it would be illogical to suppose that one could pit one's own strength against God's, and so on[…]” (pp. 100–101).Google Scholar

(50) Winch, P., Understanding…, loc. cit. p. 317.Google Scholar

(51) MacIntyre, A., “A Mistake about Causality in Social Science”, in Laslett, P. and Runciman, W. G. (eds.), Philosophy, Politics and Society, Second Series (Oxford 1962), p. 61Google Scholar. This formulation suffers from its emphasis on the location of these norms rather than on their nature.

(52) Winch, , Understanding…, loc. cit. p. 318.Google Scholar

(53) Evans-Pritchard, E. E., Nuer Religion (Oxford 1956), p. 131.Google Scholar

(54) Ibid.

(55) Ibid. pp. 131–132.

(56) Ibid. p. 318. Emphasis mine. Professor Gellner's comment on this approach is that it “absolves too many people of the charge of systematically illogical or false or self-deceptive thought”. Moreover (Gellner, E.loc. cit. p. 171):Google Scholar

The trouble with such all-embracing logical charity is, for one thing, that it is unwittingly quite a priori: it may delude anthropologists into thinking that they have found that no society upholds absurd orself-contradictory beliefs, whilst fact the principle employed has ensured in advance of any inquiry that nothing may count as prelogical, inconsistent or categorically absurd though it may be. And this, apart from anything else, would blind one to at least one socially significent phenomenon: the social role of absurdity.

(57) I think Max Weber is largely responsible for this. His use of these terms is irredeemably opaque and shifting.

(58) Philosophers have disputed over the question of whether ‘belief’ involves reference to a state of mind. I agree with those who argue that it does not; thus I would offer a dispositional account of ‘acceptance’. As will be evident, I take it that belief is by definition propositional. As to the philosophical status of propositions, this does not affect the argument.

(59) This is the sense of rationality stressed by Professor Hare. V. Hare, R., Freedom and Reason (Oxford 1963).Google Scholar

(60) Cf. Horton, R., African Traditional Thought and Western Science, Africa, XXXVII (1967), 50–71, and 155–187, especially pp. 167–169CrossRefGoogle Scholar. For numerous examples of this, v. Evans-Pritchard, E. E., Witch-craft…, op. cit.Google Scholar

(61) See, e.g., Jarvie, I. C. and Agassi, J., loc. cit.Google Scholar

(62) Cf., e.g., Parsons, , op. cit. pp. 19 and 698–699.Google Scholar

(63) Cf., e.g., Homans, G. C., Social Behaviour: Its Elementary Forms, (London 1961), p. 80Google Scholar for senses (g) and (10). It is perhaps worth adding here that I do not find Mr. Jonathan Bennett's stipulative definition of rationality germane to the present discussion (“whatever it is that humans possess which marks them off, in respect of intellectual capacity, sharply and importantly from all other known species”, in Bennett, J., Rationality (London 1964), p. 5).Google Scholar

(64) I take “criterion of rationality” to mean a rule specifying what would count as a reason for believing something (or acting). I assume that it is only by determining the relevant criteria of rationality that the question “Why did X believe p?” can be answered (though, of course, one may need to look for other explanatory factors. I merely claim that one must first look here).

(65) Cf. Strawson, P., Individuals (London 1959)CrossRefGoogle Scholar, and Hampshire, S., Thought and Action (London 1959), Ch. I.Google Scholar

(66) Winch, P., The Idea of a Social science, op. cit. p. 15.Google Scholar

(67) Winch, P., Understanding…, loc. cit. p. 308.Google Scholar

(68) I owe this argument to Martin Hollis. I have profited greatly from his two unpublished papers, “Winchcraft and Witchcraft” and “Reason and Ritual” (the latter to appear in Philosophy).

(69) Winch, P., The Idea…, op. cit. p. 126.Google Scholar

(70) Winch, P., Understanding…, loc. cit. p. 318.Google Scholar

(71) Though, as Horton shows, they may be unnecessary (African Traditional Thought…, loc. cit. p. 58).

(72) Cf. Les Cornets de L. Lévy-Bruhl, op. cit. pp. 80–82 and 193–195.Google Scholar

(73) Ibid. p. 194.

(74) Beattie and Firth see the sense of this argument but do not accept its conclusions (see quotations in text above and Beattie, J., Other Cultures (London 1964), pp. 206–207).Google Scholar

(75) Cf. Goody, J., Religion and Ritual: the Definitional Problem, British Journal of Sociology, XII (1961), 142–164, especially pp. 156–157 and 161CrossRefGoogle Scholar. As Evans-Pritchard (somewhat unfairly) says: “It was Durkheim and not the savage who made society into a god” (Nuer Religion, op. cit. p. 313).Google Scholar

(76) Cf. Evans-Pritchard, , Witchcraft, Oracles and Magic among the Azande (Oxford 1937), pp. 475–478Google Scholar, where twenty two reasons are given why the Azande “do not perceive the futility of their magic”.

* I am most grateful to Martin Hollis, John Beattie, Rodney Needham, Jean Floud, John Torrance and Vernon Bogdanor, among others, for their very kind and helpful criticisms of an earlier draft of this article.