On the Justification of Induction

On the Justification of Induction

Hans Reichenbach
The Journal of Philosophy, Vol. 37, No. 4. (Feb. 15, 1940), pp. 97-103.

This paper is an answer to the paper of Isabel P Creed.
The question of justification can be raised based on a previously chosen aim (A) or objective, it concerns the question of appropriateness of a certain means (M) or procedure.

Different kinds of justifications arise depending on what we know about the attainability of the aim, the author defines 3 cases as follows:

we know something about the objective possibility of reaching A by applying the means M
we know that by applying m we certainly will reach A
we know the probability p that A will occur if M is applied
we know at least that P > 0
we know that although p=0, A is possible
we don’t know whether or not application of the means M will lead to A
we know that by applying the M we shall never reach A

M is not justifiable in case three, Reichenbach says that all the others are justifiable but
Miss Creed disagrees with cases 1d and 2, instead she proposes a new case (b) which she thinks justifiable in terms of belief, Reichenbach answers by criticizing the use of the belief concept, how can a belief lead to a justification of induction? Instead he states that this case is just his case 2. Here the author also criticizes David Hume and his defense of inductive belief as a habit, in relation to Miss Creed’s paper title.

Then he continues explaining the terms objective possibility and epistemic possibility The first gives 2 options: something is possible or impossible, the latter gives the chance of more interpretations, being the one that views epistemic “possibility “the same as epistemic “not impossible “including the case of indeterminacy into the term “possible”, the one that would be use to demonstrate the possibility of success in a justification, making the inductive procedure possible.
Justification of the M is conditional in so far as it refers to a certain A. If somebody does not want the aim A an application of the means M by him it would not be justified. There are two types of condition c and c’, being c the less sufficient, just wanting the aim in general and c’ wanting it in face of the more or less problematic chance of success.

The next part is a critic to a kind of analysis proposed by Miss Creed, Reichenbach disagrees with the necessity of this analysis for a justification. The mathematical expectation method is introduced, it determines whether or not a bet is acceptable, where a would be the pleasure and m displeasure and p the probability of A if M is applied, the mathematical formula p*a –m , gives an acceptable value if it’s greater than 0, this condition is extended to cases in which a and m are emotional values and if there is a possibility of measuring them, it would be the perfect justification of induction for Miss Creed, however if we do not know p it would be impossible to determine this value, and it is irrelevant because it has a subjective element, and even if such calculation could be given, it would not necessary mean an obligation to act (in the case of a bet), in other cases the positive value is easily asserted making this calculation unnecessary Reichenbach continues by saying “at this point we have left the logical field and are concerned with the psychological motives of our actions. “


The problem of the justification of induction includes both the question of the decision to attempt predictions and the question of the choice of the best means of making them.
If the decision on the attempt is justified when the aim is not proved to be unattainable, and if it is proved that the means are the best we have, the problem of justification is solved.