A priori

2. Innate concepts, certainty and the a priori

Many philosophers deny that having a priori knowledge requires having innate concepts, concepts that do not derive from, or depend for their being understood on, sensory experience. (Some theorists, in the tradition of Platonism, hold that mathematical concepts, among others, are innate.) Propositions, one might suppose, consist of concepts, perhaps analogously to the way in which sentences consist of terms. Propositions knowable a priori, according to the philosophers in question, need not consist of innate concepts. The notion of a priori knowledge depends on a notion of a priori warrant, not on a notion of a non-empirical origin of the concepts constituting the known proposition. A notion involving special conditions for the justification of a believed proposition is not automatically a notion involving special conditions for either the origin or one’s understanding of the belief in question.

The notion of a priori knowledge, construed as a notion of non-empirically grounded knowledge, is not the same as a notion of epistemic certainty. Philosophers have understood ‘epistemic certainty’ in various ways: for instance, as epistemically indubitable belief or as self-evident belief. A belief is epistemically indubitable if and only if it would not be epistemically justifiable to doubt that belief under any circumstance. It is not obvious that a priori warrant for a proposition requires epistemic indubitability of this proposition. A proposition’s being warranted a priori for someone seemingly allows for an expansion of their relevant evidence, whereby a proposition justified on the original evidence ceases to be justified on the expanded evidence. (One might, for example, come to appreciate further implications of a proposition that was justified a priori.) A priori justification for a proposition apparently can be subject to ‘epistemic defeat’ given a change in a priori evidence (see Certainty; Doubt).

Philosophical talk of ‘self-evidence’ is often unclear. On a literal construal, a self-evident proposition is justified but does not depend on anything else for its justification. The problem in linking a priori warrant to such self-evidence is that a priori warrant is compatible with inferential warrant, wherein a proposition owes its warrant to inferential relations with other propositions, as might a theorem in a mathematical system. (It is a separate issue whether all a priori warrant might be inferential.) The notion of a priori knowledge should thus be explained independently of a literal construal of self-evidence. Other construals of self-evidence will contribute here only if they elucidate a notion of non-empirical warrant that differs from notions of necessary truth, analyticity and certainty as epistemic indubitability.