¹ I mean this term to be used as widely as its use in common speech would suggest, its subject-matter as, in principle, all past thoughts, not just the rather individual meaning given to it by Professor Lovejoy and his pupils. The argument of the piece is that the histories of particular intellectual practices, of science, history, political theory, economics, theology, etc., are special instances of this single unitary category and that whatever autonomy they enjoy within it is simply a matter of literary convenience. In other words it is denied that a coherent account can be given of any of them which lends to them any sort of epistemological discreteness.

² This claim is clearly more plausible when made about the history of political theory than it is, for instance, about the history of philosophy. But it seems to me to be quite unmet by even such a helpful series of treatments as those edited by ProfessorPassmore, in Beiheft No. 5, ‘The Historiography of the History of Philosophy’, of the journal History and TheoryGoogle Scholar. For examples of the two different perspectives in the history of political theory in work of some distinction see on the one hand Ryan, Alan, ‘Locke and the Dictatorship of the Bourgeoisie’, Political Studies, vol. VIII, No. 2 (06, 1965), p. 219CrossRefGoogle Scholar and on the other, Skinner, Quentin, ‘Hobbes's Leviathan’Google Scholar (review article on Hood, F. G., The Divine Politics of Thomas Hobbes) The Historical Journal, vol. VII, No. 2 (1964), p. 333Google Scholar. For an example of the sort of difference which is likely to appear in full-length treatments from these different perspectives cf. Warrender, Howard, The Political Philosophy of Hobbes (Oxford: 1957)Google Scholar with the treatment of Hobbes in Macpherson, G. B., The Political Theory of Possessive Individualism (Oxford: 1962).Google Scholar

³ It seems to be the case that the interpretation of the famous passage in Hume, David's Treatise of Human NatureGoogle Scholar on deducing ‘ought’ statements from ‘is’ statements has been distorted in just this way. Cf. Treatise, Bk. III, 1, iGoogle Scholar, with, e.g. Hare, R. M., The Language of Morals (Oxford: 1952), p. 29Google Scholar. But this is controversial. Cf. Macintyre, A. C., ‘Hume on “Is” and “Ought”’, Philosophical Review, vol. LXVIII (10, 1959)Google Scholar with Atkinson, R. F., ‘Hume on “Is” and “Ought”. A Reply to Mr Macintyre’, Philosophical Review, vol. LXX (04, 1961)Google Scholar; Scott-Taggart, M. J., ‘Macintyre's Hume’, Philosophical Review, vol. LXX (04, 1961)Google Scholar. Later, Hunter, Geoffrey, ‘Hume on “Is” and “Ought”’, Philosophy, vol. XXXVII (04, 1962)Google Scholar. Flew, Antony, ‘On the Interpretation of Hume’, Philosophy, vol. XXXVIII (04, 1963)Google Scholar and Hunter, Geoffrey, ‘A Reply to Professor Flew’, Philosophy, vol. XXXVIII (04, 1963).Google Scholar

⁴ In practice, it does not always seem relevant in particular instances. The sense in which it is true is I hope made clear by the end of the paper.

⁵ This list is, of course, a caricature and intended as such. It is not even adequate as a preliminary typology of the sort of books there are. Notably it does not begin to give an account of the best or the worst of the books that are written. In the latter case this is hardly a vice. But it is important to emphasise, in order to avoid misunderstanding, the very remarkable quality of much of the work which has been done in these subjects by Cassirer, Koyré, Kemp Smith, Lovejoy and many others.

⁶ I have learnt most from the following, without fully agreeing with any of them: Collingwood, R. G., The Idea of History (Oxford pb. ed.: 1961)Google Scholar; Gardiner, Patrick, The Nature of Historical Explanation (Oxford: 1952)Google Scholar; Dray, William, Laws and Explanation in History (Oxford: 1957)Google Scholar; Gallie, W. B., Philosophy and the Historical Understanding (London: 1964)Google Scholar; Donagan, A., The Later Philosophy of R. G. Collingwood (Oxford: 1962)Google Scholar; Danto, A., Analytical Philosophy of History (Cambridge: 1965)Google Scholar; various of the articles edited by Gardiner, Patrick in Theories of History (Glencoe, Ill.: 1959)Google Scholar and the journal, History and Theory (ed. George Nadel); also from two striking works by practising historians, Kuhn, T. S., The Structure of Scientific Revolutions (Chicago: 1962)Google Scholar and Gombrich, E. H., Art and Illusion (London: 1959).Google Scholar

⁷ Most historical writing for better or worse does not consist largely of explanations. This lends an adventitious force to the position of the critics of ‘causal’ explanation. But if the stories are still to be true, some sort of concern for causality seems to be inexpugnable. The most elegant literary constructs in history do come I to grief on aesthetically trivial facts. Pragmatically the dispute is really over what I to do with the data, once gathered. The solution must surely be that a historian may organise them in any way which he can show to be conceptually coherent. In the particular instance which I am discussing in this paper the difficulty has been that the conceptual organisation chosen has often deformed the data. Different historians do (and there can be no reason why they should not) design their work as attempts at applied general sociology or at ‘stories which happen to I be true’. Professional disputes may, causally, arise over this difference of taste but they are conducted, by professional etiquette, as disputes over the truth of propositions about the data. In this, at least, professional etiquette seems unassailable.

⁸ This does not, of course, means that such a novel form of comprehension could never come our way, just that it would be novel, i.e. we cannot know what it would be like until we know what it is like. See very helpfully Taylor, Charles, The Explanation of Behaviour (London: 1964), pp. 45–48 esp.Google Scholar

⁹ This claim is ambiguous. It does not hold for those propositions the truth of falsity of which depends solely upon the speaker's sincerity in asserting them; reports of intentions, more dubiously promises. For a superb account of the I problems raised by these see Austin, J. L., How to do things with words (Oxford: 1962)Google Scholar. Issues of sincerity do affect the truth status of propositions in works of the highest intellectual complexity (indeed, this insight has been made the key to an entire method of interpretation by Professor Leo Strauss and his distinguished group of pupils from the University of Chicago), but it is clear that the truth status of any proposition of any descriptive complexity cannot rest purely on the sincerity of its proponent.

¹⁰ There is an important conventional sense in which one can understand what I anyone says without knowing whether it is true or false. But consider, for example, the project of writing a history of science without beliefs as to the truth or falsity of any scientific proposition. Conversely, if Aristarchus thought that the earth moved around the sun, we can understand the notion, as expressed in these terms, without much difficulty. But we do not thereby know, or at least may not know (i.e. do not know) what Aristarchus meant unless we know the ontological and physical contexts at the very least which gave definition to his claim. Rudely, what we know is that Aristarchus anticipated one of our more firmly established beliefs. But this is self-celebratory gibberish, not history. It is a poor attempt at understanding Aristarchus.

¹¹ Here, as elsewhere, this phrase is used for exemplary purposes. I have no wish to foreclose on any form of attained causal explanations of behaviour, but I do not wish, particularly in the face of Taylor, Charles's The Explanation of Behaviour (London: 1964)Google Scholar, to assert that the explanations must be ultimately reducible to statements in a ‘physical-object’ data language—whether a peripheralist analysis of behaviour or a centralist analysis in terms of neuro-physiology or bio-chemistry.

¹³ Danto, Arthur C., Analytical Philosophy of History (Cambridge: 1965)Google Scholar. The whole book is devoted to expounding the importance of the difference between contemporary-specific and future-specific descriptions of events, say, for example, ours and theirs, to the understanding of historical analysis. The Yeats poem is very deftly quoted at page 151. To rephrase Danto's point, the data-language of history changes throughout history. The future constantly changes the set of true descriptive statements which could in principle be made about the past. No contemporary description of an event can take this particular sort of account of the future which it ‘engenders’.

¹³ Op. cit. (note (1) above), p. 219Google Scholar. Such an account (sc. an analysis of Locke's concept of property, taken from the Second Treatise alone) ‘may perhaps be in danger of refutation by the historian as an account of what Locke intended. It is in less, even no, danger of contradiction from such a quarter as an account of what Locke said. And in case this is thought too small a claim, let me point out that we usually hold people to what they say, rather than to what they suppose to follow from what they meant to say.’ I should like to emphasise that the very able article in question does not in practice suffer at all from ill consequences deriving from this, to me, misconceived methodological doctrine.

¹⁴ This again is a wild over-simplification. I have deliberately begged the most intractable question about psychological explanation (what the form of an adequate causal explanation of a piece of human behaviour would be) by talking of the more behavioural ‘activity’ rather than the more intellectualist ‘act’. I quite accept that understanding an act is never just a matter of subsuming a piece of behaviour under a set of causal laws, but I should certainly want to claim that part of doing so is frequently just such an operation. But cf. Macintyre, Alasdair, ‘A Mistake about Causality in Social Science’ in Laslett, Peter and Runciman, W. G., Philosophy, Politics and Society (2nd Series) (Oxford: 1962)Google Scholar and convergent arguments in Kenny, Anthony, Action, Emotion and Will (London: 1963)Google Scholar; Taylor, Charles, The Explanation of Behaviour (London: 1964)Google Scholar. Also Winch, Peter, The Idea of a Social Science (London: 1958).Google Scholar

¹⁵ I.e. it may not have been what he meant, cf. notes (13) and (9) above. What a man meant to say may differ from what he succeeded in saying in numerous ways. He may, for instance, as in many of the cases considered by Sigmund Freud in the Psychopathology of Everyday Life, speak a word different from the word which he intended to speak; or he may speak a word in a foreign language thinking that it has a meaning which is other than the meaning which it does have, a common occurrence; or he may use a word in his own language persistently while misunderstanding its meaning (whether by mistaking it for another with a similar sound, a malapropism; or by simple mis-identification). All of there seem peripheral. It is hard to imagine a man who never said what he meant to say (not, of course, one who never said what he meant (where meant = felt like saying). There is nothing conceptually difficult in the notion of a consistent hypocrite. Indeed, if by chance one were confronted by someone who never said what they meant to say, one could only interpret their behaviour as the result of severe and peculiar brain damage. There are, of course, numerous instances where men say things which are not consistent with other things which they say or feel and one could, under some circumstances, describe these states of affairs as instances of men not meaning what they say. But this is a very derivative usage and surely cannot be construed as meaning that they did not intend to convey what they did convey but rather that they did not realise the implications of what they, intentionally, said and would not have said it if they had realised these. The suggestion in note (13) above depends upon there being a general distinction between what men succeed in saying and what they intend to say. The distinction which does exist between these two scarcely seems of the right type. If one is interested in trying to understand an argument, the least one wculd normally attempt to do is to establish what the protagonist meant.

¹⁶ It is more common for someone from an alien culture to misunderstand what a person has said than it is for people in any culture not to say what they mean.

¹⁷ Cf. Durkheim, Emile, Socialism and Saint-Simon (New York, pb. ed.: 1958), p. 41.Google Scholar

¹⁸ I.e. say it yelpingly; not yelps are propositions. Just that the truth-status (whatever that may be) of the proposition ‘God is Love’ is no different where it is gasped out by the dying martyr in the blood-stained arena from when it is enunciated with the plummy self-assurance of a well-fed agnostic in a role which is religious only as an inconvenient historical hangover.

¹⁹ See Austin, J. L., How to do things with words (Oxford: 1962), passim.Google Scholar

²⁰ For an example of the acute interpretative difficulties which this fact raises, see the remarkable reconstruction by ProfessorRyle, , Plato's Progress Cambridge: 1966).Google Scholar

²¹ This does not mean that the ideas of stupid people can be explained causally while the ideas of those who share our own incomparable advantages elude such crude determination–though there is a faint and horrible grain of truth in an extreme version of such a view. Cf. Popper, Karl's famous argument (set out in brief in the preface to The Poverty of Historicism, pp. ix–xi and refs. there (London: 1960, 2nd ed.)Google Scholar. But the Popper argument does not apply to making causal statements about past ideas—it is the logical oddity of predicting new ideas on which it insists. It is a purely contingent (though highly intractable) fact that in the case of the Republic the sort of data which survive go no distance whatever towards providing an account of the sufficient conditions for the writing of the book.

²² Cf. Kamenka, Eugene, ‘Marxism and the History of Philosophy’, in Beiheft 5. History and Theory, pp. 83–104.Google Scholar

²³ Passmore, John, ‘The Idea of a History of Philosophy’, in Beiheft 5, History and Theory, p. 13.Google Scholar

²⁴ See, briefly, Runciman, W. G., Social Science and Political Theory (Cambridge: 1963) cap. 6Google Scholar. For a penetrating account of the sources and deficiencies of the notion as employed by Malinowski see Leach, E. R., ‘The Epistemological Background to Malinowski's Empiricism’, in Man and Culture: An Evaluation of the Work of Bronislaw Malinowski (ed. Firth, Raymond, London: 1957).Google Scholar

²⁵ Kenny, Anthony, Action, Emotion and Will (London: 1963), pp. 28–51.Google Scholar

²⁶ There are, of course, dangers in learning to talk precisely about fictions instead of trying to talk about the confusions of the world. Cf. in another area Mills, C. Wright, The Sociological Imagination (New York: 1959)Google Scholar. But it is still important in all innocence to advocate the attempt to combine both.

²⁷ Winch, Peter, The Idea of a Social Science (London: 1958), passim.Google Scholar

²⁸ As a part of the causal story, this can be very considerable indeed. Cf. on Plato, Adkins, Arthur W. H., Merit and Responsibility (Oxford: 1950).Google Scholar

²⁹ Not that one would not employ philosophical notions at any point in the attempt to explain and assess them; only, that most of the operation of understanding them (even after the story of how they come to be there is told completely) has nothing to do with philosophy.

³⁰ Mutatis mutandis, this would apply to the history of any specialised form of reflection. Each such special extrapolation is derived from an unitary matrix, the causal story of past human thought, the set of necessary and sufficient conditions for the set of past human thoughts.

³¹ See Laslett, Peter (ed.), Locke, John, Two Treatises of Government (Cambridge: 1960), pp. 67–76Google Scholar. I am currently attempting to set out the character which the book did bear as it was written, in a full-length study, ‘The Political Philosophy of John Locke’.

³² Proc. Arist. Soc. supplementary vol. XXII, ‘Things and Persons’ (quoted by Passmore, John, ‘The Dreariness of Aesthetics’, in Elton, William (ed.), Aesthetics and Language (Oxford: 1959), p. 40.Google Scholar

³³ This paper arises out of several years of discussion of the subject with Mr Peter Laslett and, especially, MrSkinner, Quentin (see his article ‘The Limits of Historical Explanations’, Philosophy, vol. XLI (07, 1966)Google Scholar, for a partly analogous, partly contrasting view). I am very grateful to them both. Dr M. I. Finley, Dr R. M. Young and Mrs Joanna Ryan very kindly read it through and helped me to clarify a number of points. Where it remains opaque, it does so through no fault of theirs, but merely as a result of my own obstinacy.