[hist-analytic] Carnap's Bootstrap, Pulled Up By Grice

Jlsperanza at aol.com Jlsperanza at aol.com
Sun Feb 28 17:24:56 EST 2010



In a message dated 2/28/2010 4:42:35  P.M. Eastern Standard Time, 
rbj at rbjones.com writes:
Carnap argued in Meaning  and Necessity that an extensional 
metalanguage would suffice for the  intensional object 
languages which he deals with in that book, but I'm not  
aware of him having offered a more general argument.

---  

I would suggest we consider, somewhat seriously, the half page Grice  
devotes to Bootstrap, in "Reply to Richards" google.books.
 
He never claimed to have _proved_ it, but surely neo-Grice can do that, 'at 
 one fell swoop' as he'd say.
 
The thing is _bound_ to please neo-Carnap (or not entirely displease  him!)
 
-----
 
historical note.

When getting tired of people ab-using 'meta-', I took the time to  consult 
the OED, and R. Hall may have been my benign demon, we find the first  use 
to go to Lord Russell, which is VERY respectable indeed. He coined the  
thing, and he coined the uglier thing, the
 
   object-language
 
My litmus test is: you know a philosopher is serious when he does use  
'object-language'. By this Russell may have meant, language _as_ object or  
universe of discourse, or domain, of a meta-language (alla Boole, Jevons, etc.), 
 rather than anything fancier -- i.e. that it is the language of _objects_!
 
So back to Carnap and Grice.
 
What "Bootstrap" states is that the meta-language can NEVER be so much  
richer than the object-language, because you may not, tomorrow, be able to pull 
 up by your own bootstraps, and we don't want that.
 
How this connects with Carnap:
 
For Carnap:
 
  FL-1  Anything goes. Anything-Goism. Intensional  isomorphisms,
            anything  is pretty much tolerated. Surely his was not
             zero-tolerance.
 
 MFL-1 i.e the meta-language for FL-1
          This HAS to be  'extensional'. None of your fancy
          intentional  isomorphisms here, or, even 'meaning postulates'
          _intentionally_  conceived.
 
----
 
So that is totally consistent with Grice.

What Grice may have found offense in would be:
 
 
someone postulating:
 
 
object-language: whatever.
 
meta-language: whatever; intentionsional;
          intensional  isomorphisms, etc.
 
 
-----------
 
There are more applied aspects to this.

E.g. Grice's theory of meaning! After all, it is the received (by  
Bennett), if wrong, view that we have some sort of historical development  or crisis 
here:
 
 
         Carnap
  Meaning and Necessity
             1947
                          . 
                            .                      Grice, "Meaning"
                            .                           1948
                   (You Know Who
                Thinking This Dogmatic)
 
                                  . 
                                   .  .  .  .  . Grice/Strawson
                                          coming to defend
                                          the underdog(ma)
                                                   1956
 
I.e. Bennett sees Grice/Strawson as _needing_ a theory of 'meaning' to be  
able to defend the under-dog of the 'meaning' postulate. But Grice had said 
'...  means ...' is INTENTIONAL. So:
 
   object-language:    x means y
 
 metalanguage:         A  intends B
 
And 'intend' is NOT extensional!
 
But this would not be a problem for Grice his qualm is with a  
Meta-Language which is UNABLE to be reduced to the Object-Language. And  "intentional" 
talk _is_ soluble in (more) "intentional" talk.
 
 
------ So much for the object-language/meta-language distinction, but more  
later, I hope.
 
J. L. Speranza



More information about the hist-analytic mailing list