Page 33 - Encyclopedia of Philosophy of Language

SECTION II
Language, Metaphysics, and Ontology
A Priori
E. J. Lowe
The notion of the 'a priori' has its primary application in the field of epistemology, where it is standardly used to characterize a species of prepositional knowledge (knowledge that p, where p is a proposition) and, derivatively, a class of propositions or truths, namely, those that are knowable a priori (though strictly this way of classifying propositions should be relativized to a type of knowing subject, the usual presumption being that human subjects are in question). In a related usage, certain concepts are sometimes classified as a priori, namely, those that figure as substantive con- stituents of a priori truths.
1. APrioriKnowledge
Knowledge is said to be a priori (literally: prior to experience) when it is knowledge which does not depend for its authority upon the evidence of experi- ence. This is not the same as saying that it is knowledge which is acquired independently of experience, whether because it is innate knowledge or because it is knowledge learned without the substantive con- tribution of experience (for instance, knowledge learned through the exercise of pure reason). How knowledge is acquired and how knowledge claims are justified are quite distinct, albeit related, matters. The converse of a priori knowledge is 'a posteriori' knowl- edge. The history of the a priori/a posteriori dis- tinction may ultimately be traced back to Aristotle, but modern usage owes much to the influence of Immanuel Kant. Until the twentieth century, the dis- tinction was viewed with increasing skepticism by epistemologists, but interest in and respect for it have been revived.
2. APrioriandInnateness
Although the notion of a priori knowledge and the notion of innate knowledge are quite distinct, his- torically philosophers have tended to run them to- gether by confusing questions of justification with questions of acquisition. Mathematical knowledge is usually held up as the paradigm of a priori knowledge and certainly it is true that mathematical knowledge
claims, unlike the claims of physical science, do not normally depend for their justification or con- firmation upon observational or experimental evi- dence (an exception being claims based on the results of electronic computation rather than on direct math- ematical proof). Few would think it appropriate to test the truth of the arithmetical proposition 7+ 5=12 empirically, by repeatedly conjoining and counting sets of seven and five objects. (Unusually amongst major philosophers, however, John Stuart Mill did believe that mathematics rested ultimately upon induction from experience.) But even accepting the a priori status of mathematical knowledge, it is quite another matter to hold (as Plato did) that math- ematical knowledge is innate (though this in turn is not to deny that experience may be needed to 'trigger' such latent knowledge, as happens in Plato's account of the slave boy in the Meno). Conversely, con- temporary linguists like Noam Chomsky and philo- sophical psychologists like Jerry Fodor, who notoriously hold that much of our knowledge of lan- guage is innate, do not therefore wish to claim for (say) principles of universal grammar the same epis- temological status as mathematical truths as far as their justification is concerned: linguistics, unlike mathematics, is an empirical science, answerable to observational evidence.
3. A Priori and Analyticity
'Empiricist' philosophers have traditionally held that all a priori knowledge is of necessary, analytic prop- ositions, and conversely that all a posteriori knowl- edge is of contingent, synthetic propositions. David Hume, for instance, is standardly interpreted as adopting this view. Kant, however, famously held that we have some a priori knowledge of certain very gen- eral synthetic propositions (such as the proposition that every event has a cause), though he too believed that all a priori knowledge could only be of necessary truths. However, in his highly influential onslaught upon the 'dogmas of empiricism,' W. V. O. Quine was to argue that none of these distinctions could be
11