appendix: empiricism without inductivism
One of the main points of Popper’s philosophy is his suggestion that empiricism is distinct from inductivism and therefore that one can be an empiricist without accepting induction.
This point seems to me not to have had the notice it deserves.Consider the following inference:
First premise: We learn from experience.
Second premise: Whatever we learn from experience,
we learn it by using induction.
Conclusion: We learn from experience by induction.
Evidently the conclusion follows from the two premises. Hume showed that the conclusion is false if by ‘learn’ we mean ‘gain theoretical knowledge’. Let us accept without debate this interpretation and Hume’s criticism. As an inference with a false conclusion is either invalid or contains at least one false premise, the following approaches are open to us. (1) We may render the inference invalid. We may do so, with Hume, by reinterpreting ‘learn’ in the first premise only, or in both premises only, but not in the conclusion, to mean ‘acquire habits.’ (2) We may deny the truth of the first premise and say that our theoretical knowledge is not derived from experience but is a priori valid. This is the Kantian approach. (3) We may accept the first premise and deny the second premise. This is the Popperian approach. Other approaches are possible, but they were never seriously entertained. The Humean and the Kantian approaches are now in almost universal disfavor.
The Popperian approach raises the question of how we learn from experience if not by induction. Popper’s answer to this question may, if it is acceptable, strenthen his rejection of the second premise. In any case, it will show that the second premise can be denied without inconsistency. However we view Popper’s answer to this question, we must acknowledge that we have no argument for accepting the second premise.
The only argument in its favor (namely that there is no conceivable empirical method other than the inductive method) was superseded by Popper’s proposal of his own methodology. This should be admitted even if Popper’s methodology is rejected.In defense of the second premise, one may argue that ‘we learn from experience’ is synonymous with ‘we learn by the use of inductive method.’ But then the phrase ‘inductive method’ can be omitted altogether and instead of using it one can characterize methods which, according to Hume’s criticism, are not workable. These are the methods of generalizing from observations, and of rendering some hypotheses more probable than others with the aid of observations. And the question remains, what alternative methods there can be, which may be said to explain the fact - if it is a fact - that we gain theoretical knowledge from experience. Popper’s methodology - his view that we refute hypotheses with the aid of observation - may or may not be acceptable. But it has not been shown to be erroneous, while all existing alternatives to it have.
Those who nonetheless reject Popper’s view and seek alternatives should provide empiricist but not inductivist views; i.e., ones which are not hit by Hume’s arguments.