Published online by Cambridge University Press: 01 January 2020
The proponents of the ‘vindication’ or ‘pragmatic justification’ of induction have attempted to show that induction will work if any method does. This in turn serves as grounds for their claim that we have everything to gain by using induction and nothing to lose. Hence, they conclude that it is rational to use induction. Their claim that induction will work if any mehtod does is based upon the following argument:
If nature is uniform, induction will work. If nature is not uniform and some other method works, the success of that method will then constitute a regularity to which induction can be successfully applied. In that case, induction will sanction the continued use of the successful method and hence will be successful as well.
1 Cf. Salmon, W. ‘Inductive Inference’ in Philosophy of Science, The Delaware Seminar, vol II, (lnterscience 1963),Google Scholar and Feigl, H. ‘De Principiis Non Disputandum …’ in Black, Max ed., Philosophical Analysis (Ithaca, N.Y.: Cornell University Press 1950)Google Scholar
2 Salmon, ‘Inductive Inference,’ 355
3 The following distinctions between levels of inference follows that of Skyrms, B. in his book Choice and Chance (Encino, Cal.: Dickenson Publishing Co. 1975) 44ff.Google Scholar
4 Skyrms, Loc. cit.
5 This criticism is entirely independent of the common one that it is difficult, on the vindicationist's account, to Justify the preference of one rule of inductive inference over another. The problem raised here defeats the vindicationist's argument even if a preference for one rule could be Justified.
6 Cf. Hume, D. A Treatise of Human Nature, ed. Selby-Bigge, L.A. part II, section VI (Oxford: Oxford University Press 1967)Google Scholar
7 Cf. Strawson, P.F. Introduction to Logical Theory, Chapter IX, (London: Methuen and Company, Ltd. 1952)Google Scholar
8 It is important to note that the argument presented in this essay need not deal with the traditional problem encountered by that position - i.e., which if an in· finite class of inductive ruled of inference to use. This is because the argument presented in this essay is addressed to the problem of induction at the general level; it is concerned with the issue of whether or not Hume's problem is a genuine problem. The question is therefore logically prior to the issue of which rules of inference we should use. However if the argument presented in this essay is successful, it would establish a basis upon which to decide which rules of inductive inference to use. If, that is, it is proven that it is rational to reason inductively in general, we can evaluate specific rules of inductive inference on the basis of their past performance or what would have been their performance if they were used. This approach would presumably avoid the difficulties encountered by the vindicationists, for they attempt to select the ‘correcr rule of inductive inference (the 'straight’ rule) on purely a priori grounds independent of the Justified status of induction. This seems to me to be doomed to failure, Just as any attempt to Justify the inductive method on purely a priori grounds would be doomed to failure, once Hume's argument is accepted. For difficulties with the vindicationists’ position, see Skyrms, B. ‘On Failing to Vindicate Induction,’ Philosophy of Science, 17 (1965) 253-68,CrossRefGoogle Scholar and Maxwell, G. ‘Induction and Empiricism: A BayesianFrequentist Alternative’ in Feigl, H. and Maxwell, G. eds. (Minneapolis: Min· nesota Studies in the Philosophy of Science, Vol. VI. University of Minnesota Press 1975)Google Scholar
9 For a discussion of this point, cf. Strawson, op. cit.
10 Cf. Popper, K.R. Conjectures and Refutations (London: Routledge & Kegan Paul 1963) 64Google Scholar
11 In this regard, cf. Salmon, W. The Foundations of Scientific Inference’ in Colodny, R. ed., Mind and Cosmos, (Pittsburgh: University of Pittsburgh Press 1965)Google Scholar