Tarski's Truth ...

Thu, 16 Dec 1993 12:43:00 PST

Graber writes:
>J. McCreery presents, I think, a straw-man version of positivism and the
>role therein of the semantic conception of truth.

While I agree with the sense and thrust of what Graber writes, there are some
knotty problems that are being glossed over, I believe. Let me recast
Graber's argument in a slightly different format. Mathematicians use truth
in the sense that theorems derived in an axiomatic system are true given the
axioms. Most mathematicians are happy to take the axioms as a priori and the
question of whether or not the axioms are true when given a real world
interpretation is largely irrelevant. E.g., a variety of geometries are
definable using different axioms regarding infinitely extended lines. At
most one of these geometries captures the reality of the universe, but
mathmaticians are nonetheless interested in all of these geometries and do
not worry about which one might be the best model for the universe
(physicists worry about such questions, though). In effect, mathematicians
recognize that the question of whether or not the axioms are "true" is of a
different nature than knowing of sucha and such a theorem is true given the
axioms. Now the truth that Graber is talkins about is the truth of the
axioms when given a real world interpretation. And here is where the
difficulty arises. Our only means for empirical testing relies on
categorizations we make of real world sensory input--categorizations made
unconsciously by our brains, or categorizations we more-or-less consciously
construct. Thus, I say: the sun is shining, and you go and look out the
window and say, yes, that's true. In doing so we are using implicit
catetories and agreed upon understanding of these categories. That is, we
have an implicit mapping that links the assertion "the sun is shining" with
certain sensory perceptions we agree should be "observed" if that statement
is true. But how do we know that the categorizations made by our brain are,
in fact, isomorphic with objective reality? We know that the latter
assertaion is not always valid; e.g. optical illusions, so how do we know
when our cognitive categories are isomorphic with objective reality? This
seems to be the point at which the post-modernists leap in with a vengeance:
if all depends upon our categorizations, and categorizations are (in some
sense) arbitrary, then in the absence of a criterion for deciding which
categorizations are isomorphic with objective reality we might as well
consider all categorizations equally valid, hence science has no more claim
on producing "knowledge" than, e.g., right-wing christian fundamentalism.

I think the error lies in assuming that in the absence of an absolute
criterions for knowing which categorizations are isomorphic with objective
reality and which are not the only viable alternative is the nihilistic one
that all are equally valid. It seems more reasonable to assert that while we
cannot make absolute distinctions, we can rank categorizations into more
plausiable and less plausible. Indeed, from an evolutionary viewpoint if
what our brain produced were merely arbitrary categorizations (ie. noise),
then the brain would never have evolved as it did, for by simply producing
noise it would reduce fitness. The brain evolved only by virtue of the fact
that it could construct reasonably accurate categorizations and models of
external reality that enabled our ancestral forms to deal correctly more
often with the external world than would otherwise be the case.

The methodology of science can be seen as one of assessing categorizations
as to their validity indepenent of particualr individuals at particular times
in particular situations. Thus, the Voyager spaceship that is escaping from
our solar system carries a disc that includes knowledge it is presumed ANY
"intelligent" life form must also know about by virtue of that intelligent
life form also being part of our universe and making observations about our

If we think of assessing the "truthfulness" of a statement as a process
of subjecting it to the test of whether or not accepting the validity of
the statement is contingent upon particular persons at particular times in
particular contexts, then I think Graber is on solid ground in saying that
science is about "truth".

D. Read