Published online by Cambridge University Press: 14 March 2022
One of the recurrent temptations in theory of knowledge is to try to provide an inductive justification for employing the principles of inductive inference. The purpose of the present paper is to suggest that this approach is misguided by exposing the deficiencies of what appears to be the most carefully constructed attempt of this kind which has thus far been made. I refer to the attempt made by R. B. Braithwaite in his Scientific Explanation.
1 (Cambridge, England, 1955), chap. VIII. All page references in parentheses refer to this book.
2 Where the reasonableness of a belief is said to consist in the proposition believed being either “known directly to be true” or capable of valid inference from a proposition “known directly to be true” (p. 278).
3 In the sense that the conclusion can justifiably be added to the inferrer's body of reasonable beliefs (p. 280). Hereafter, I shall use the words “valid” and “validity” in this sense.
4 R. B. Braithwaite, “Probability and Induction,” British Philosophy in Mid-Century, C. A. Mace, ed. (New York, 1957), p. 149. (Italics his.)
5 Criteria I-V are certain possible sufficient criteria for the validity of an inference by a person B of the conclusion q from the premiss p in accordance with a principle of inference the effectiveness of whose use is asserted by the proposition r, where by “validity” is meant “justifiably carries a belief in q along with B's belief in p” (p. 280). They are as follows:
I. B believes p and believes r;
II. B believes p and reasonably believes r;
III. B believes p, and r is true;
IV. B believes p and believes r, and r is true;
V. B believes p and reasonably believes r, and r is true (p. 279).