Reading Questions for Salmon’s “Rational Prediction”

Due Date: 11/1

What is the point of the balloon example? Why does the balloon move forward?

1. Salmon summarizes Popper’s falsificationist anti-inductivism as follows: failed predictive tests give rational (because deductive) grounds for rejecting generalizations (theories); successful predictive tests (testing which does not result in failed prediction) do not confirm or inductively support generalizations; rather they build a record of past corroboration for the generalizations involved; this record of corroboration is seen by Popper to have no implications for the future performance of the generalizations; conclusions about future events (predictions) result deductively from two sorts of premises: (i) statements of past and present events, and (ii) generalizations (theories). Salmon then presents the “problem of rational prediction”—a problem for Popper’s anti-inductivism. What is this problem, and how/why does it seriously challenge Popper’s whole view?

Salmon concedes, on p. 436, that the problem of rational prediction may not be a problem for theoretician merely interested in generating and testing an explanatory theory. Why does he concede this?

2. Salmon turns to practical prediction, and considers Watkins’ response to the problem. What is Watkins’ response, and why is Salmon unsatisfied with it?

3. Salmon then considers Popper’s claim that “corroboration statements have no predictive content although they motivate and justify our preference for some theory over another”. The issue here, for Salmon, is what Popper could possibly mean by ‘justify’ in this claim. Popper cannot me ‘inductively justify’, for he rejects that notion. How does Popper (in the quotations on pp. 440-41) claim that choosing the more corroborated theory is rational? Why is Salmon still not satisfied that Popper has solved the problem of rational prediction without smuggling in some form of inductive inference?

4. Salmon suggests an explanation (not necessarily a justification) for Popper’s insistence that corroboration provides a rational basis for practical prediction. What is this suggested explanation, and why does it make Popper appear much like a traditional inductivist?

The main thrust of Salmon’s argument in this paper is as follows: (i) prediction serves at least three purposes (deductive test, satisfying intellectual curiosity, as part of practical decision making); while Popper’s anti-inductivism may be consistent with the first of these, (ii) the second two seem to require that the generalizations involved in drawing predictive conclusions be rationally supported; (iii) either they are not rationally supported on Popper’s view, or they are inductively supported, undermining Popper’s anti-inductivism; i.e., every sense in which Popper tries to show prediction can be rational either fails, or imports induction. Is it possible that Salmon, in relying on (iii), is pressing a false dilemma? I.e., is it possible that (whether or not Popper has made the case himself) reliance on well corroborated theory could be rational in some sense not depending on induction?