[hist-analytic] Kripke on the A priori and A posteriori

Roger Bishop Jones rbj at rbjones.com
Mon Nov 9 16:54:52 EST 2009

Since Steve is banging this drum at the moment I thought I would throw in my 

There are two fundamental defects in Kripke's conception of the a priori/a 
posteriori distinction which render it unsatisfactory and cast doubt on the 
validity of his supposed refutations of predecessors.
Of course I am more interested in rebutting the suggestion that Kripke refuted 
Carnap than Kant, but the problems may be relevant to both.

The first is that Kripke takes the ways we come to believe a proposition as 
crucial to its epistemic status, whereas in my opinion they is quite 
irrelevant.  To get a classification of propositions one must consider not how 
a belief originates, but what we are prepared to accept as a justification of 
the claim that the proposition in question is true.  Not only should the 
manner of discovery be disregarded, but even the justification offered by the 
believer. The only question is what kind of justification would be acceptable 
for that kind of proposition.

The second problem is that in determining whether a justification offered for a 
proposition is a posteriori or a priori one must carefully separate out 
matters which relate properly to the justification of the proposition from 
those which relate to the determination of which proposition is expressed by 
some sentence under consideration.
The connection between a sentence and the proposition it expresses will in 
general be contingent, and empirical evidence may therefore be necessary to 
explicate the meaning of the sentence.  This should be disregarded in 
determining the epistemic status of the proposition expressed by the sentence, 
otherwise all propositions will be a posteriori.

This latter issue is of particular importance in Kripke's philosophy because 
of his views about rigid designation, the effect of which is to make the full 
meaning of a proposition obscure.   I would argue that we do not know fully 
the meaning of a sentence containing a rigid designator until we know which 
object it designates, and therefore that any empirical observation needed to 
establish what objects are so designated should be disregarded when 
determining the epistemic status of the proposition expressed by a sentence 
involving the designators.

Presumably Kripke would disagree with me on this point about the semantics of 
rigid designators, but if he does so then his theories become irrelevant to 
the refutation of Carnap on these matters, since Carnap's conception of 
semantics is a truth conditional account which entails that the intension of a 
rigid designator must include its referent. Kripke therefore argues at cross 
purposes with Carnap and his criticisms miss their mark.

Roger Jones

More information about the hist-analytic mailing list