Essay On Reliabilism

Does reliabilism offer a viable solution to the problem of induction?

For the purpose of this paper I will refine the problem of induction to enumerative cases of induction. I shall explore whether reliabilism is a successful theory of knowledge, and propose that it is a viable solution to the problem of induction proposed by David Hume, but requires ad hoc amendments in attempt to satisfy the New Riddle of Induction put forth by Nelson Goodman.

The problem of induction, most notably attached to Hume, is the philosophical question of whether inductive reasoning leads to knowledge, in which Hume concluded that inductive inferences from observed instances to a general conclusion can never yield knowledge. Traditional philosophical epistemology assumes that knowledge requires certainty. Given such stipulations, inductive inference would have to guarantee complete certainty to warrant it being knowledge. Hence, because universal generalisations (all Fs are Gs) apply to indefinitely many instances, but we can only observe finitely many instances; no number of observations can entail that universal generalisations are necessarily true (there may possibly be a case where an F isn’t a G). Thus, it is logically possible that the premises (every rose I have ever seen is red) be true but the inferred conclusion (all roses are red) be false; accordingly, inductive inference is logically invalid and cannot yield knowledge.

Such problem, according to David Papineau, holds no grounds given the doctrines of reliabilism. Reliabilism is an externalist account of knowledge, which defines knowledge as true belief caused by a reliable process. Papineau maintains that reliabilism offers a viable solution to the problem of induction, but concedes...

... middle of paper ...

...regard certainty in the actual world. Subsequently all statements and claims, inductively inferred, are approximate, as opposed to being explicitly true.

Reliabilism puts forward a viable solution to the traditional problem of induction proposed by Hume, showing that despite enumerative induction being logically invalid, it can convincingly yield knowledge. Similarly, it can be shown the circularity involved in establishing inductive inference does not trivially guarantee its conclusion, unlike premise circularity. Nevertheless, Goodman’s New Riddle of Induction poses serious threat to what reliabilism can actually state as knowledge. If the reliabilist is willing to concede that inductive inferences are beliefs of less than full degree, they are faced with conceding that only deductive inferences and analytical truths yield certain knowledge in the actual world.