[hist-analytic] Bootstrap

Roger Bishop Jones rbj at rbjones.com
Sun Jan 30 14:03:21 EST 2011

Well, JL, I wasn't intending to refer to Grice's "bootstrap" 
problem but it certainly is germane to the mid century 
disputations which ejected Carnap's programme from the 
philosophical centre ground (if it was ever there!).

Analyticity, and necessity were central concepts for Carnap, 
they coincided (except that analyticity was a property of 
sentences and necessity of propositions), and also coincided 
with the a priori.
The main planks in the dismissal of Carnap's programme were 
first Quine's attack on the analytic/synthetic dichotomy, and 
second Kripke's teasing apart of these three dichotomies 
(though Danny sounded as if he thought Kripke left 
metaphysical necessity and analyticity ("broad logical 
truth") together.)

I was really probing around the Kripkean end of things, 
whereas it seems to me that the "bootstrap" issue is more 
conspicuous in the debate about analyticity.
As far as that goes, Quine deliberately confuses the 
definition of the concept of Analyticity (which determines 
"broad logical truth" in terms of truth conditional 
semantics, and doesn't require a bootstrap) with the problem 
of defining (truth conditional?) semantics, which does get 
you into a bootstrap problem (at the same time as 
criticising Carnap for doing the same). 

I prefer to call this particular "bootstrap" problem "the 
problem of infinite regress in the foundations of semantics",
which is a real and interesting problem on which Quine says 
little of value in "Two Dogmas".
This problem is, that to define the semantics of a language 
you need to use a language, and in order for the definition 
to be meaningful, you must have a definition of the semantics 
of the language you use.
Hence, either infinite regress OR at some point you define the 
semantics of a language either directly or indirectly in 
terms of itself OR you slip into informal semantics, OR some 
combination of the above.

The fact is, that it can be done.  Not absolutely without 
ambiguity or with absolute precision, but with very high 
standards, and certainly (for formal language) much higher 
than one gets in natural languages.

As it happens my present interest is in the status of 
arguments which supposedly establish against Carnap that 
there is some notion of metaphysical necessity which goes 
beyond his conception of logical necessity, and more 
generally in the merits of the kinds of metaphysics which 
were pursued in the latter half of the twentieth century.
Plus ca change.

Roger Jones

More information about the hist-analytic mailing list