[hist-analytic] AUNE: Analytic and A Priori

steve bayne baynesrb at yahoo.com
Wed Mar 18 12:13:41 EDT 2009

Avoiding certain technicalities, we can say that a sequence 

<a,b> satisfies a formula 'Fxy' just in case Fab. 

Yes, you are quite right about this, but it doesn't directly address the point: there 
is no "commuting" between languages in the hierarchy, nor (in Tarski) are there 
"language exit" rules. Tarksi is quite explicit on this, making clear in no uncertain 
terms that his system includes only four constants: negation, disjunction, 
quantification, and the universal quantifier. ("The Concept of Truth in Formalized 
Languages" in _Logic, Semantics, and Metamathematics_. Hackett, 1956. p. 168. 
Now, of course, you can actually "do" the sequences in some metalanguage, that is,
 make the correlations; but when you do you include something like 'a sequence 
<a,b>' and, so, introduce constants that don't occur in Tarski. So we must at some point 
give "rules" for correlating 'a' and some object, class, etc. The devil is in "certain 
technicalities," but this is where we want to go.

"Kripke used to show that some sentences true by virtue of meaning are 
contingent (e.g. "The standard meter is a meter long")..."

Kripke uses "meaning" very seldom in connection with this problem. He speaks 
of definitions, in particular stipulative ones. So when you use "meaning" I have to 
be apprised of your theory of meaning and how this relates to Kripke's reply to 
Wittgenstein. Read one way, Kripke's criticisms seem to assume Wittgenstein is
sort of foolish. Kripke is a far better logician than Wittgenstein was, but he was 
not foolish. I don't mean to suggest Kripke has a low opinion of Wittgenstein, either.
 Indeed he credits the latter with too much in my opinion in his book on the private
 language argument, which I think is a masterpiece! In addition the meaning of 
'contingent' must be explored. Before moving on let me give an example of my
concern over the use of 'meaning'. Take the definite description 'the man in the 
corner drinking a martini'. I don't think there is any better example of where we 
must take heed of Wittgenstein injunction to ask for use not meaning. The meaning 
of this expression on my view is invariant; but it may be used referentially or 
attributively (following Donnellan). The same can be said of 'the length of a stick 
one meter long'. 

On synthetic a priori, I'm undecided,  but I'm inclined to believe there are non-analytic 
a priori truths. Here I am motivated by Kant in particular, whose views I think are very 
advanced and better than many of his most articulate critics, such as Reichenbach. 
I haven't gotten to your criticisms in ETK which very possibly could lead me in the 
direction of empiricism, a direction with which I am most comfortable.

Best wishes

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://rbjones.com/pipermail/hist-analytic_rbjones.com/attachments/20090318/ee713c49/attachment-0002.html>

More information about the hist-analytic mailing list