3 Proceedings of the British Academy, 1956, pp. 109–32, republished in his Philosophical Papers (eds. J. O. Urmson and G. J. Warnock), 1961, pp. 153–180. Page references following are to the latter.
4 “Ifs and Cans,” Theoria, 1960 Pt. 2, pp. 85–100.
5 “Possibility and Choice”, Aristotelian Society Supplementary Volume, 1960, pp. 1–24.
6 “I Can”, The Philosophical Review, Jan. 1960.
9 There are perhaps purely sarcastic uses of the exemplified form of expression; but there are others which are not, and which are genuinely open to rebuttal.
10 Op. cit., pp. 13–14. See also Nowell-Smith, Ethics, p. 274.
11 In adopting this restriction I am simply making use of a familiar convention of ordinary language. For in replying to an accusation regarding something I am alleged to have done, I may insist either that I did it inadvertently, or that I didn't really do it—it was an unintended consequence of what I actually did.
13 Some purists may not approve of speaking of the contrapositive of a causal “if... then” (by contrast with a logical one), if only because the “contrapositive” is not itself a straightforward causal statement. I here simply follow the convenient loose usage of Nowell-Smith (Theoria, p. 88) and O'Connor (op. cit., p. 7). Austin does not use the term.
14 Austin heightens the illusion of oddity in the claim to infer the contrapositive by insisting, more strictly, on “If he would not have done otherwise, he had not chosen” (op. cit., p. 158). But if this is unintelligible, it is surely not any more so than a similarly strict inference from a similarly tensed version of Austin's avowedly causal example; for “If I had run I would have panted” yields “If I would not have panted I had not run”. I have therefore, in the text, eliminated this shared queerness in order to concentrate upon the question whether there is a difference of conditionality between the allegedly causal and non-causal examples.
17 O'Connor appears to slip, later in his paper, from this ‘only if interpretation into something more like the position I maintain myself below. In contrasting it with “I can if I choose”, he argues that “I shall if I choose” does not entail “I shall whether I choose or not” (p. 10) because choosing here is the last necessary condition to be satisfied. But on his previous showing, he should have interpreted “I shall if I choose” as “I shall only if I choose”; and the latter does not imply that choosing is the last necessary condition.
18 A similar point might be made about O'Connor's contention that “If we go by Oughterard we can get to Galway by dinner time” means “that I shall not get there by the desired time if I do not go by Oughterard” (op. cit., p. 6).
20 I say only ‘normally implies’ because, in spite of the support I later offer to the common assumption that there are no sufficient conditions for immanent choice of action, I should have to agree that we do speak of causes of action (and not just of movement) where this really means causes of choosing to act. I have argued elsewhere (see “Historical Causation and Human Free Will,” University of Toronto Quarterly, April 1960, pp. 357–69) that such causes do not complete sets of jointly sufficient conditions. The situation here, however, is the quite different one of choosing as cause of movement.
22 “He could have smashed that lob” (op. cit., p. 88). I do not use his example “He could have read Emma last night” because the reading is not a simple case of action as we are using the term here.
23 Op. cit., pp. 176–179. Austin juggles the two senses when he objects that ‘could have’ cannot mean ‘would have if because the latter, since it implies that the specified condition was unfulfilled, implies also that he couldn't have (p. 128).
25 Op. cit., pp. 9–10, 21.
29 See Taylor, op. cit., p. 87.
30 K. Lehrer (in “Ifs, Cans and Causes”, Analysis, June 1960, pp. 122–123) produces a similar puzzle, starting from a different point.