Grice's characterisation of "extensionalism"

Speranza has given us an expansion of the characterisation of "extensionalism" by Grice in "Reply to Richards".

I can't say this has completely sorted out things for me but I did come away with some useful clues as to what Grice was getting at.

Before discussing that, I recapitulate some context.
I am a substitute Carnap discussing with Speranza pp Grice some topics on which they might have been difficult to reconcile.  Grice's Bete's Noirs are a source of examples, and "minimalism" and "extensionalism" are acute examples, for Carnap is certainly minimalistic and extensionalist in some sense, and Grice seems adamantly opposed to some such doctrines.

I have already mentioned one general way in which Grice and Carnap might be talking past each other on this matter, by distinguishing between "dogmatic" and "pragmatic" minimalism.  A key feature of minimalism for Grice seems to the the denial by the minimalist to others of the use of the kinds of entities which he has discarded.  For example a nominalist may deny the existence of abstract entities and hence refuse to accept any ideas or theories tainted by them.
However, for Carnap, as evidenced by his "principle of tolerance" and also by his "Empiricism Semantics and Ontology", there is no tendency to proscription, he is a minimalist who lacks what seems in Grice to be the principal failing of minimalists.

Turning more specifically to "extensionalism" we can ask, while acknowledging that Carnap is some kind of extensionalist, whether he is the kind of extensionalist which Grice has in mind, and whether any of the criticisms which Grice levels at extensionalism applies to the kind of extensionalist which Carnap exemplifies.

For this purpose a clear definition of "extensionalism" would be nice, but we will have to settle for Grice's rather oblique characterisation and the further refinements we might infer from the detail of his critique.

In response to my failing to grasp Grice's characterisation, Speranza points out (inter alia!) that "red" is a property (a universal) while "Paul Grice" is a proper name. I think the story might run on, that universals have meanings and so there is something definite which they attribute but that names do not, they simply refer.

This leads us to a story about why this characterisation fails for Carnap, for Carnap does not consider names to be devoid of meaning, not at least in the formal languages with which he is principally concerned.
A typical Carnapian method of defining such a language would be to define the syntax of the sentences and then give L-rules and P-rules, in which L-rules determine which sentences are analytic and may be understood as defining the meaning of the language (and the constant names in it). (P-rules express some synthetic scientific theory).  Later the meanings of constants are given in "meanings postulates".
All this means is that a Carnapian might easily fail to understand Grice's characterisation because he does not think of proper names as meaningless.

This is not entirely an eccentricity of Carnap, for his two distinguished predecessors Frege and Russell approached logicism on the basis that anything derivable in a logical system from definitions given in that system is a logical truth.  That certainly doesn't work for a notion of logical truth which corresponds with analyticity, unless the defined constants have a meaning given in their definitions.  Of course, they didn't talk about analyticity, but they probably were thinking broadly along the lines of Hume's "relations between ideas" which is essentially the same idea, and Frege thought of logical formulae as if they were mathematical expressions which evaluate to truth values (which leads to Wittgenstein's characterisation of logical truths as tautologies, and thence to the semantics methods in Carnap's "Meaning and Necessity").

Further. what Grice seems to have in mind here is that an extensionalist will insist on universals such as "red" being understood as extensions.
Red is the set of red objects, so that what we might consider the essential notion of redness is lost, and the explanation of redness is denuded of most of its content.

So the second point here is that Carnap is not that kind of extensionalist, at all.

Quine perhaps, the disagreement between Quine and Carnap on modal logic (which one might think forced on Quine by his stance on analyticity, given that for Carnap analyticity and necessity are interdefinable) points us in that direction. Quine objected to "intensions", but Carnap, though in some sense an extensionalist (to which I may come later), embraced them.

From here I propose first to give an account of the kind of extensionalist which Carnap might have been, and then to consider whether any of Grice's criticisms hold good of that kind of extensionalist, and if not whether Grice might possibly be reconciled to that kind of extensionalism (and more broadly, to at least some of Carnap's other minimalisms), in a later post.


RBJ