[hist-analytic] Carnap and Grice on "logical"

Roger Bishop Jones rbj at rbjones.com
Sun Mar 7 16:28:01 EST 2010

On Friday 05 Mar 2010 17:26, Jlsperanza at aol.com wrote:

> Thanks v. much for the explanation. I guess I won't be
>  using 'logical constant' for a while. 

I hope you don't take me to be prescribing how you should 
use that term.
It was my intention to clarify my own usage and connect it 
with some of the other usage with which I am familiar, and 
in particular with what was said by some important players 
in the run-up to Carnap's (purely verbal) shift from using 
L- to A- concepts.

The more general usage which yours exemplifies is just as 
well represented (if not better), but since you questioned 
my observation about first order logic, I sought to connect 
it with the usage of mathematical logicians in that context. 

>  Grice does speak
>  of 'device' in various contexts:
> They seem to be "logical devices" but he just says
> "formal devices" in WoW:ii, first page.
> and he has both the turth-functors (monadic, -; dyadic:
>  &,v, ->) and  the three quantifiers ((x), (Ex) and
>  (ix)). This gives the list as comprising, let's see
>  _seven_ formal devices. He does not mean to be complete,
>  because his  point is about "some of the formal
>  devices". And he is not into the mathematic,  as you
>  say, first-order predicate-calculus (we agree there
>  that's the stuff of  mathematics) BUT of what he calls,
>  vaguely, 'philosophical logic' (as opposed to 
>  'philosophy of logic'). I think Grice and Strawson were
>  in this informal  campaign of highering the status of
>  what they were doing, from "philosophy of  logic" --
>  where they would be philosophers doing logic -- to
>  "philosophical  logic" where they would be logicians
>  philosophising.

Not convinced that that is the correct delineation.
"Philosophy of logic" surely must be philosophy, whereas 
"philosophical logic" must be a kind of logic; doesn't that 
jump out from the grammar??

>  Matter of style --  Similarly, he saw
>  himself as a philosophical psychologist, rather than as
>  a  philosopher of mind.

That's interesting.  An expansion would be too.
But don't we get a similar problem.
Surely Grice should be a philosopher of psychology, and a 
philosophical psychologist someone approaching the similar 
problems from the opposite direction?

> I agree with you that 'logical' should be given a higher
>  status and that it cannot be dependent on the choice of
>  a linguistic framework like that. And so I  see very
>  well Carnap's point in changing from "L" to "A".

I think you may have mistaken me there. 
I think also that I put myself badly.
I did intend firstly to opine that Carnap's change was a 
purely verbal concession and did not reflect his agreement 
with Quine and Tarski on the use of the term "Logical".

Second when talking about whether Logical Truth is language 
dependent, we have to bear in mind that it must be in one 
sense, that it depends upon the semantics of the language 
which sentences of the language express logical truths (just 
as in the case of analyticity).
However, in the case of analytic, once one has the truth 
conditions, one needs no further information to determine 
analyticity.  But if logical truth is taken to be narrower, 
then you do need something extra to determine logical truth.
You have to have the truth conditional semantics split into 
two parts just for this purpose, and this is quite 
> It seems authors who have spoken of the 'logical
>  constants' have confused the things.

I don't know that I would say that myself.
I think it is legitimate to use "logical constant" just to 
refer to certain features of language, and the distinction 
which is captured by a narrower usage may not be relevant to 
what they are saying.

It is arguably not relevant to the description of logical 
truth in these narrow senses, since whether the sentential 
constructors of first order logic are constants or not is 
immaterial to this point, one needs their semantics to be 
taken into account (in determining logical truth).

On the other hand, though legitimate, it is a hostage to 
fortune because of the narrower usages, and another term 
might be less likely to result in confusion.


More information about the hist-analytic mailing list