Re: Coextensive properties?

From: Pierre-Normand Houle (houlepn.nospam_at_attglobal.net)
Date: 10/13/04 

Date: Wed, 13 Oct 2004 12:53:45 -0400

"Neil W Rickert" <rickert+nn@cs.niu.edu> wrote in message news:ckjkfp$3ap$1@usenet.cso.niu.edu...
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> "Pierre-Normand Houle" <houlepn.nospam@attglobal.net> writes:
>
> >Quine might say (but I would have to look further into this) that the class
> >"red" which is useful in science (as a starting point) is the class of objects
> >x that have in fact prompted, and will in fact prompt, users to
> >ascent to the proposition "x is red" when they are (have been) visually
> >presented with them.
>
> The trouble with this, is that a scientist would have to poll the
> populace to see whether something is in class red. He could no
> longer rely on his instruments.

Yes, this, or maybe he would have to "over-rely" on his instrument.
If the class of red things is the class of things that are, in fact, positively
identified as such by the instrument, then the scientist cannot make
sense of the instrument malfunctioning. An instrument functions
properly if it works as it is *supposed* to do. An irreductively intensional
notion.

What would Quine say? Here is a conjecture : A faulty instrument is not
one that fails to track a real property but one relative to which we define
an operational concept (class) that, in fact, contributes badly to future
prediction generated using our whole system of (operational) concepts
and theory. Quine's holism saves him but there are further problems
with this answer. One is that it merely presupposes that intentional
notions have weak predictive values. The best explanation of an
operational concept being useless might be it's being tied to a
broken or ill concieved instrument! (The return of intentions with
a vengeance.)