¹ What I call the ‘empiricist view’ roughly corresponds to Berlin's ‘negative concept’ of freedom; and what I call the ‘apriorist view’ has some resemblance to his ‘positive concept’, at least as he initially presents it (Berlin, Isaiah, Two Concepts of Liberty: an Inaugural Lecture (Oxford: Clarendon Press, 1958) pp. 16 f.)Google Scholar. However, he soon proceeds to what I regard as collectivised, romantic and illiberal perversions of this individualist and rationalist idea, which seems, in his treatment, almost to get swallowed up by its monstrous progeny.

² ‘Three Views Concerning Human Knowledge’, first published in Contemporary British Philosophy: Third Series, ed. Lewis, H. D. (London: Allen & Unwin, 1956)Google Scholar; reprinted in Popper, K. R., Conjectures and Refutations (London: Routledge & Kegan Paul, 1963) chap.Google Scholar

³ Actually, this view (at least as it was worked out by Hobbes) is not quite as straightforward and uncomplicated as it at first seems; see my Hobbes's System of Ideas, second edition (London: Hutchinson, 1973) chap. 7, especially pp. 95–6 and 128 f.Google Scholar

⁴ I have worded this cautiously in view of the notorious difficulties (culminating in Arrow's Impossibility Theorem) which attend the idea of the optimum collective satisfaction of individuals' preferences. I have reviewed these difficulties in my ‘Social Knowledge and the Public Interest’, Man and the Social Sciences, ed. Robson, W. A. (London: Allen & Unwin, 1972) PP. 185 f.Google Scholar

⁵ I presented this seeming counter-example to Hobbes's concept of liberty in 1965 (Hobbes's System of Ideas, first edition, §33 (2)).

R. S. Peters has drawn my attention to an article by Benn, S. I. and Weinstein, W. L., ‘Being Free to Act, and Being a Free Man’, Mind, LXXX, 318 (04, 1971)CrossRefGoogle Scholar with whose conclusions I am in full agreement. Thus they write: ‘There are cases, indeed, in which one clearly does not have it [the freedom of an autonomous chooser], cases of manipulated choice. The hypnotised subject and the brain-washed are “not their own masters”, not because the objective conditions of choice have been interfered with, but because the subjective conditions have. Though the subject may believe that he is choosing, the actions of other people may still sufficiently account for what he does; he may be as externally programmed as a computer’ (p. 210).

⁶ This phrase, which recurs rather frequently in my text, is adapted from Melden, A. I., Free Action (London: Routledge & Kegan Paul, 1961) p. 8.Google Scholar

⁷ In 1938 he wrote: ‘Libertarianism is certainly inconsistent with a rigidly determinist theory of the physical world. It is idle to pretend that there can be open possibilities for psychical decision, while at the same time holding that the physical events in which such decisions manifest themselves are determined in accordance with irrevocable law.’ Campbell, C. A., In Defence of Free Will (London: Allen & Unwin, 1967) p. 45.Google Scholar

⁸ ‘Physical determinism, we might say in retrospect, was a daydream of omniscience which seemed to become more real with every advance in physics until it became an apparently inescapable nightmare.’ Popper, Karl R., Objective Knowledge: An Evolutionary Approach (Oxford: Clarendon Press, 1972) p. 222.Google Scholar

⁹ See the reference in note 6 above.

¹⁰ Munn, Allan M., Free-Will and Determinism (London: MacGibbon & Kee, 1960) pp. 205 f.Google Scholar

¹¹ See, for example, Russell, Bertrand, ‘The Elements of Ethics’ (especially section IV) in Philosophical Essays (1910)Google Scholar and Schlick, Moritz, Problems of Ethics, trans. David Rysim (1939) chap. viiGoogle Scholar. For a recent defence of the reconcilability-thesis, see Grünbaum, Adolf, ‘Free Will and Laws of Human Behavior’, American Philosophical Quarterly, vol. 8, no. 4 (10 1971).Google Scholar

¹² I think that D. M. MacKay has introduced a novel consideration in his ingenious attempt to reconcile free choice with mechanistic determinism and scentific predictability; but I do not think that his attempt succeeds: see my ‘Freedom and Predictability: an Amendment to MacKay’, Brit. Jour. Phil. Science, vol. 22 (08, 1971)Google Scholar. (MacKay's reply in the same issue of the BJPS contains a bibliography.)

¹³ ‘The ordinary common-sense man has always had a belief in his personal free-will… Up to the seventeenth century [the validity of his belief] was not really denied. Though Christian theology might talk about an omniscient and omnipotent supreme being, it was a highly abstract theory. At the practical level the churches acted as if man did possess free-will… The situation changed with the development during the seventeenth, eighteenth, and nineteenth centuries of physical science. Here for the first time in the history of man's thought appeared a rigorously deterministic theory, extending its range of application in all directions without encountering any apparently insurmountable obstacle; and by the middle of the nineteenth century there was every good reason to believe it would be eventually extended to a detailed description of man himself. What to do with man's belief in free-will when considered against the background of a universal deterministic physics? The continuing success of physics resulted in the abandonment of any direct attacks upon its determinism…’ Allan M. Munn, ibid., p. 213.

¹⁴ See pp. 311–14 of the paper by Grünbaum referred to in note 11 for a recent statement of this dismissive view.

¹⁵ Treatise, Bk. I, Part III, Sect. XIV (ed. Selby-Brigge, p. 171)Google Scholar. This claim of Hume's has been criticised by Popper, , Objective Knowledge, pp. 227 f.Google Scholar Hume's view was reiterated by Schlick: ‘the alternative “either determination or chance” is a logical one, there is no escape from it, no third possibility’ (in Feigl, and Sellars, , Reading in Analysis (New York, 1949) p. 532).Google Scholar

¹⁶ Although Russell was for the reconcilability-thesis in 1910 (see note 11 above), he had stressed the ‘alien and inhuman’ aspect of the physical determinism of contemporary science in ‘A Free Man's Worship’ (1903): ‘Man is the product of causes which had no prevision of the end they were achieving;… his origin, his growth, his hopes and fears, his loves and his beliefs, are but the outcome of accidental collocations of atoms’ (Mysticism and Logic, p. 47).Google Scholar

¹⁷ This is an adaptation of Alfred Landé's argument to the effect that determinism leads to a theological conspiracy-theory of random sequences. (See, for example, his ‘Determinism versus Continuity in Modern Science’, Mind, vol. 67 (04, 1958)Google Scholar and his New Foundations of Quantum Mechanics (Cambridge, University Press, 1965) chap, ii.)Google Scholar

¹⁸ Joseph Priestley, for instance, would have endorsed this implication unhesitatingly. He was a determinist and also a Christian; and for him a decisive objection to libertarianism was that it would deprive God of His ability to foresee what would happen in His own creation. (The Doctrine of Philosophical Necessity Illustrated, 1777Google Scholar, sect. III, ‘Of the Argument for Necessity from the Divine Prescience’; reprinted in Priestley's Writings on Philosophy, Science and Politics, ed. Passmore, John A. (London: Collier-Macmillan, 1965) p. 63.)Google Scholar

¹⁹ As Hume put it: ‘Let no one, therefore, put an invidious construction on my words, by saying simply, that I assert the necessity of human actions, and place them on the same footing with the operations of senseless matter. I do not ascribe to the will that unintelligible necessity, which is suppos'd to lie in matter. But I ascribe to matter, that intelligible quality, call it necessity or not, which the most rigorous orthodoxy does or must allow to belong to the will.’ (Treatise, Bk. II, Part III, Sect. II, ed. Selby-Bigge, p. 410.)Google Scholar

²⁰ Ethics, Bk. V, prop, xxxii.

²¹ ‘It is evident from what I have said, that we are in many ways driven about by external causes, and that like waves of the sea driven by contrary winds we toss to and fro unwitting of the issue and of our fate.’ (Ethics, Bk. III, prop, lix, note.)

²² Groundwork of the Metaphysic of Morals, second edition, p. 115Google Scholar (pp. 123–4 in Paton, H. J.'s translation, The Moral Law, London: Hutchinson).Google Scholar

²³ A new version of it has recently been advocated by Stephan Körner: see his Abstraction in Science and Morals, Eddington, Memorial Lecture (Cambridge: University Press, 1971).Google Scholar

²⁴ Ibid., pp. 112–13 (pp. 122–3).

²⁵ Popper, Karl R., ‘Indeterminism is Not Enough’, Encounter (04, 1973) p. 20.Google Scholar

²⁶ Farrer, Austin, The Freedom of the Will (London: Black, 1958) pp. 2–4.Google Scholar

²⁷ He seems to have been a would-be interactionist inhibited by the old belief that something non-physical cannot interact with something physical. He rejected psychophysical parallelism and epiphenomenalism, insisting ‘that consciousness should do some real work, and not exhaust her efficacy in the mere business of being conscious’ (ibid., p. 99). But confronted by the full-fledged interactionist, dualist, neo-Cartesian position presented by Sir John Eccles in The Neurophysiological Basis of Mind he exclaimed: ‘We will have nothing to do with the fantastic suggestion, that what the supersensitive “reactors” in the cortex react to, is the initiative of a virtually disembodied soul’ (p. 87). He took refuge in a kind of identity-theory: ‘We say that the reactions of the reactors are, in cases of free will, themselves the work of the soul… incarnate in … these very reactors’ (p. 92).

²⁸ In my contribution, ‘The Unity of Popper's Thought’, submitted back in 1968 to the forthcoming volume, The Philosophy of Karl Popper (The Library of Living Philosophers, ed. Schilpp, P. A.).Google Scholar

I wanted to avoid duplication of that paper in this one. But the delayed publication of the former (though it has appeared in a German translation, by Albert, Gretl, in Grundprobleme der Grossen Philosophen, ed. Speck, Josef (Göttingen: Vandenhoeck & Ruprecht, 1972)Google Scholar and my desire to make the present paper reasonably self-contained have resulted in some overlapping. [Added in proof: The Philosophy of Karl Popper was published in 1974.]

²⁹ Popper, Karl R., ‘Quantum Mechanics without “The Observer”’, chap. 1 in Quantum Mechanics and Reality, ed. Bunge, Mario, volume ii of Studies in the Foundations, Methodology, and Philosophy of Science (Berlin and New York: Springer Verlag, 1967)Google Scholar, and ‘The Propensity Interpretation of Probability’, Brit. Jour. Phil. Science, vol. 10 (05, 1959)Google Scholar.

Popper himself does not seem to have exploited his propensity theory in his criticism (see note 15 above) of Hume's necessity'chance dichotomy.

³⁰ von Mises, Richard, Probability, Statistics and Truth, Second English edition by Geiringer, Hilda (London: Allen & Unwin) pp. 11–12.Google Scholar

³¹ My debts to Popper here are indicated in §4.1 of the paper referred to in note 28.

³² Gilbert, William, De Magnete, Book V, chap. xii.Google Scholar

³³ Bohm, David, Causality and Chance in Modern Physics (London: Routledge, 1957) p. 45.CrossRefGoogle Scholar Or as Einstein put it: ‘For this theory [the electrodynamics of Faraday and Maxwell] and its confirmation by Hertz's experiments showed that there are electromagnetic phenomena which by their very nature are detached from every ponderable matter — namely the waves in empty space which consist of electromagnetic “fields”.’ (‘Autobiographical Notes’, in Albert Einstein: Philosopher-Scientist, ed. Schilpp, P. A., The Library of Living Philosophers, vol. 1, p. 25.)Google Scholar

³⁴ See the section on ‘The Construction, Possession and Utilisation of Theories’ in Ryle, Gilbert, The Concept of Mind (London: Hutchinson, 1949) pp. 286 fGoogle Scholar.

In Spinoza's case, we know from the Short Treatise on Godf Man, and His Well-Being, trans, and ed. Wolf, A. (1910) reprinted (New York: Russell & Russell, 1963)Google Scholar; and from the correspondence with Oldenburg in The Correspondence of Spinoza, trans, and ed. Wolf, A. (London: Allen & Unwin, 1928)Google Scholar that there had been earlier versions of his system which differed significantly from the version in the Ethics.

³⁵ See Lakatos, I., ‘Proofs and Refutations’, Brit. Jour. Phil. Science, vol. 14 (1963–1964).Google Scholar

³⁶ See note 2 above.

³⁷ On the ‘world 3’ status of problems, see Popper, Karl R., Objective KnowledgeGoogle Scholar, chap. 4. Russell's problem of the class of all classes that are not members of themselves will serve as an example. Russell did not manufacture this problem, he found it; and when he found it, he at first seriously underestimated it, imputing it to ‘some trivial error in my reasoning’, Russell, Bertrand, My Philosophical Development (London: Allen & Unwin, 1959. P. 76).Google Scholar

³⁸ Op. cit. (see note 18 above), section II (p. 62). Priestley's standard case was that of a child who likes apples and dislikes peaches being offered an apple or a peach. But even here the distinction is not entirely obliterated: an enterprising child might respond inventively to the situation by taking the peach in the hope of swapping it for two apples later.

³⁹ This and the next two paragraphs were added in response to a suggestion by Alan Musgrave.

⁴⁰ For the notion of ‘completeness’ or ‘connectedness’ see, for example, Sen, Amartya K., Collective Choice and Social Welfare (Edinburgh: Oliver & Boyd, 1970) p. 3.Google Scholar

⁴¹ In my ‘Imperfect Rationality’, Explanation in the Behavioural Sciences, ed. Borger, & Cioffi, (Cambridge University Press, 1970) esp. pp. 197 and 202 f.Google Scholar; and my ‘Self-Interest and Morality’, Practical Reason, ed. Körner, Stephan (Oxford: Blackwell, 1974) pp. 67–77.Google Scholar

⁴² Additional, that is, to the arguments in Popper, Karl R., The Poverty of Historicism (London: Routledge, 1957).Google Scholar

⁴³ Latsis, Spiro J., ‘Situational Determinism in Economies’, Brit. Jour. Phil. Science, vol. 23 (08, 1972) p. 211 and passim.Google Scholar

⁴⁴ Of course, trying to escape is not the only way in which a prisoner can try to preserve his autonomy, I have an unbounded admiration for those political prisoners of totalitarian regimes — Alex Weissberg, the pseudonymous F. Beck and W. Godin, Edith Bone, Eugenia Ginzburg, Paul Ignotus, George Paloczi-Horvath, Bela Szasz, and others — who, though they sometimes came near to breaking-point, continued basically to defy the system of lies and terror in which they were caught, hoarded their experiences like misers anxious to let no bit of truth slip away, and — eventually — turned the tables by telling the world, in sharp detail and without bitterness, the truth as they knew it.

⁴⁵ Lorenz, Konrad, King Solomon's Ring, second edition (London: Methuen, 1961) pp. 108–10.Google Scholar

⁴⁶ Hume, David, Enquiries, Appendix ii ad fin. (ed. Selby-Bigge, p. 302).Google Scholar

⁴⁷ My analogy is inspired by Dostoyevsky's short story, The Friend of the Family.

⁴⁸ An unusual and alarming example was reported by Newton (in a letter to Locke). Newton had found that, on going into a dark room after looking at the sun in a looking glass, he could revive an image of the sun at will. But after doing this a good many times he found that he could not make it go away again for several days; and ‘for some months after, the spectrum of the sun began to return as often as I began to meditate upon the phenomenon, even though I lay in bed at midnight with my curtains drawn’. (Quoted in Cranston, Maurice, John Locke (London: Longmans, 1957) pp. 346–7.)Google Scholar

⁴⁹ Waismann, F., How I See Philosophy, ed. Harré, R. (London: Macmillan, 1968) pp. 196–8.CrossRefGoogle Scholar

⁵⁰ In terms of Popper's schema ‘P ₁ → TT → EE → P ₂ (where P₁ is an initial problem, TT is a tentative theory, EE is error-elimination, and P ₂ is the resulting problem), I have been emphasising the relation of XT to P ₂ at the expense of its relation to EE.

⁵¹ The idea of a decision-scheme is explained in my ‘Imperfect Rationality’ (see note 41 above), pp. 206 f.

⁵² Quoted in Mill, John Stuart, Collected Works, vol. x, p. 186.Google Scholar Attention is drawn to this pasage by H. B. Acton on p. xiii of the work mentioned in the next note.

⁵⁸ Mill, J. S., Utilitarianism, chap. 2 (p. 9Google Scholar in John Stuart Mill, Utilitarianism, Liberty, Representative Government, ed. Acton, H. B. (London: Dent, 1972)).Google Scholar

⁵⁴ Autobiography, ed. Laski, H. J. (Oxford University Press) p. 143.Google Scholar See note 16 above for a parallel development in Russell's estimate of determinism.

⁵⁵ Logic, Book VI, chap. ii. R. P. Anschulz, after quoting the passage referred to in the previous note, comments: ‘Nor in spite of much painful pondering did Mill ever succeed in producing a satisfactory solution of the problem. On his premisses, indeed, it is an insoluble problem.’ (The Philosophy of J. S. Mill (Oxford: Clarendon Press, 1953) p. 172.)Google Scholar I agree with this and also with his subsequent remark: ‘In the upshot Mill contrives to retain both these views of human nature [the naturalistic or scientific view and the romantic or self-formative view] by expounding them in different parts of his philosophy.’ (Ibid., p. 173.)

⁵⁶ Autobiography, p. 177.Google Scholar

⁵⁷ There is a moving account of Freud's last days, as he was dying from cancer of the jaw, in Ernest Jones, Sigmund Freud, vol. iii, pp. 261 f. During his long illness Freud had never taken more than an occasional aspirin, preferring ‘to think in torment than not to be able to think clearly’. He took an interest in outside events to the end. When things got too bad (‘It is only torture now and it has no longer any sense’) he was given, by prior arrangement, a dose of morphia (‘Tell Anna’) and died peacefully.