Aristotle's 'Genetic Account' and the Problem of Induction

Abstract

The problem of induction--the problem of how one can justify an inference from observations of some things of a type to a generalization about all (or most) things of the type--is one of the most important in logic and epistemology. In addition, one of the enduring problems of Aristotle scholarship is whether he dealt with the problem of induction, and if so, how. This problem is important in connection with _Posterior Analytics_ B.19, where Aristotle seems to provide an account, a piece of genetic epistemology, of how the principles of demonstration are acquired. The account seems to describe an inductive process. In my dissertation, I argue that Aristotle does have to face that problem of induction in the genetic account of APo. B.19, and has a putative solution to it. I argue that the putative solution to the problem is based on his doctrine of natural kinds.

I argue first that Aristotle in fact recognizes induction that consists of reasoning from particular propositions to a universal proposition. I then evaluate various readings of the genetic account of APo. B.19. I argue, in particular, that there is strong textual evidence against the claim that Aristotle thinks that principles of demonstration are secured by nous as intuition, so that nous is the solution to, or a way to avoid, the problem of induction.

What then is Aristotle's putative solution to the problem of induction? I argue, in essence, that for Aristotle the solution is a matter of retaining enough percepts of particulars in the subject kind of the conclusion (and maybe some outside of that kind) in memory for the universal form of the subject kind to become salient and clear in the rational soul. Thus, having the comprehension (nous) that all normal horses are quadrupeds is a matter of retaining enough percepts of horses, each having four legs, for the universal form of horse to "make a stand" (in Aristotle's words) in the mind of the inducer. In this way, the inducer comprehends that the four-legged-ness is part of that form, i.e. that it "belongs to" the form horse, i.e. that all normal horses are quadrupeds.

This gives us a picture of Aristotle as being aware of this problem of induction. But, because of his metaphysical commitment to natural kinds with universal forms that can be grasped through perception and abstraction, he, unlike Hume, does not think it is a major or unsolvable problem.