[hist-analytic] Aune and the a/s distinction
baynesrb at yahoo.com
Sun Feb 8 10:34:06 EST 2009
I want to thank Speranza, Danny, Roger, and Rogerio for their astute
comments. i will be looking at them more carefully. I've been engaged
in finishing the book I'm doing on Anscombe, which is about complete.
I am considering not submitting this thing to the academic presses.
I'm not sure I like the "cut," the demands, the groveling. I quit groveling
a few decades ago and have no desire to return to the "game."
Now some VERY half-baked comments on the a/s distinction.
"Quine thought that, for philosophical purpose, a conception of analytic
truth as truth by virtue of meaning was too vague to be taken seriously."
- Bruce Aune
If we reject the idea of truth in virtue of meaning, we might ask: "Truth, then,
in virtue of what?" Powerful arguments (Davidson) have been adduced for
being suspicious of facts, but if not facts what? Suppose we take a
pragmatist view a fact is what makes a theory "work." Few, if any, pragmatists
would accept this simplistic formulation, but despairing of his loss of facts,
the philosopher will care to consider what distinguishes "I ate my breakfast
before my dinner" or "I laughed at 6:00am" where their contradictories are
"false." The pragmatist answer would be what? So truth in virtue of meaning
is one way of viewing analyticity, it is NOT what Kant had in mind, and I think
Kant understood the role of the a/s distinction better than Quine, as evidenced
by the use he made of synthetic a priori, a category of *propositions*.
Elsewhere Aune has indicated that as for what makes the a/s important we
can say this:
"To show the error of a wide range of claims by epistemological rationalists."
I have deferred comment because I need to take a close look at an ETK,
Aune's book on Hist-Analytic, before making a serious reply; but, for now,
consider this: I, share, Aune's empiricist proclivities; still I remain skeptical
about the a/s distinction.
Central to the issue is whether there are facts; if there are no facts then
the truths of algebra and the truths of physics are truths in the same sense
of 'truth' differing only in some other property requiring the introduction of
meaning, perhaps. My, immediate, concern is the place of probability in
philosophy of science. If we look at the mathematical theory of probability,
then I think we can establish a pretty clean connection between the propositions
of science and two valued logic. But if there is no way of making this connection,
"truth," in science may best be regarded outside the semantic web the
Tarskians have weaved. This, I think, was pretty much Reichenbach's view
and I am inclined to share it. What makes an analytic proposition true? We
don't want to say that it is analytic because it has such and such resistance to
revision, etc. and then go on to say that it resists revision because it is analytic.
If there are no logical "facts" then the truths of logic are expressed as propositions which are "formal" truths lacking factual content. This in some quarters is orthodoxy.
If there is no "factual"/"formal" divide then there is, ultimately, no distinction between
the truths of science and the truths of mathematics. I find this unacceptable, but
only because I don't regard the distinction as epistemological; nor do I believe that
the most productive consequences of such a distinction are of interest to a belief
that a conceptual framework is linguistic.The subject matter of science is the world;
the subject matter of logic is not; logic lacks a subject matter. This is by no means obvious or clear. Is a thought a part of the world? Does ontology go beyond "atoms
and the void"? These unresolved issues figure into all this and I don't have an opinion I would "go to the mats" to decide.
When Carnap speaks of the "logic of the language of science" in the Logical
Syntax of Language, the language of science must be understood as the
"physical language." If, as Aune points out, Carnap agreed that there is no
useful a/s distinction in natural language we can say, at least, one of two things.
We can say the a/s distinction has nothing to do with language, and that it
concerns concepts, alone (Kant) or we can say that the notion of meaning in a canonical language is not the same as in natural language. What then is the
place of meaning in canonical languages? It is tied to reference, ala, Tarski.
Once again the language of science is tied to truth in a formalized language,
and where that language is science, and "facts," are no longer at issue, the
position is untenable. Analyticity becomes a matter of decision in the construction
of a "viable" canonical language. Not unrelated here is that in a three valued logic applied to probabilities, where probability is understood on the frequency theory,
these laws of logic simply do not apply. If the world is best described on such a frequency theory then this is not a matter of decision or "interest." Whether such
a logic is employed is determined in large measure by the way the world is, not "meaning."
It would be interesting to know what Quine/Carnap would have regarded as a "philosophically useful distinction." I think, for example, that Kant's treatment of
the synthetic a priori marvelously relates ethics and mathematics! But I don't
think either would be inclined to consider this application a "phiosophically
useful distinction." A look at their ethical theory may explain why.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the hist-analytic