About Φ-stuff with respect to relevantly local samples

about Φ-stuff with respect to relevantly local samples of Φ-stuff specifically,
evaluate universally quantified beliefs about potable, transparent, sailable-on,
. . . etc., stuff with respect to relevantly local samples of stuff that is potable,
transparent, sailable-on, etc. Since, as it turns out, the stuff that satisfies
that description on Earth is of a different kind from the stuff that satisfies
that description on Earth2, beliefs about potable, transparent, . . . etc., stuff
get evaluated in different ways in the two places. And since, to run it into
the ground, the stuff that satisfies that description on Earth is H2O and the
stuff that satisfies that description on Earth2 is XYZ, it turns out that Earthwise
tokens of the phenomenological belief are true iff H2O is wet while
Earth2-wise tokens of that belief are true iff XYZ is wet. Which is, as the
patient reader may recall, just where we wanted to get to.
I have imposed the principle that universally quantified beliefs about
Φ-stuff should be evaluated with respect to "relevantly local" samples of
Φ-stuff, but I have not said what relevant localness comes to. If I had to make
a stab at it, I would guess that relevant localness is fundamentally an etiological
notion so that what the Principle of Reasonableness is telling us to do, in this
case, is to evaluate beliefs about Φ-stuff with respect to the kind of Φ-stuff
that gave rise to them. Since, it turns out, the stuff that gives rise to the
phenomenological belief on Earth is of a different kind from the stuff that
gives rise to it on Earth2, the phenomenological belief gets evaluated differently
in the two places. This, however, is very tentative, and I should want to keep
the issue of analyzing the notion of relevant localness clear of the issue whether
we are enjoined to evaluate universal beliefs with respect to relevantly local
phenomena. The latter question seems to me a good deal less murky than the
former.17
One more word about being reasonable. It is not only informal discourse
that demands circumspection in the evaluation of quantified variables. Consider
the quantifiers that bind variables in lawlike statements. The fact is that we
evaluate them too with respect to our "local" bits of the universe; roughly,
with respect to those regions of space-time for which isotropy can reasonably
be assumed.18 It would be worse than nit-picking, it would, in fact, be bloodyminded,
to object to the periodic table of elements, or to the germ theory of
disease, . . . etc., on the grounds that, for all we know, they do not hold prior
to the initial bang or on the other side of black holes. Nor is this exercise of
reasonableness merely "optional" (to use a term that Rorty coined for a
related issue). We could not say just how our nomologically bound variables
should be evaluated even if we wanted to, since we do not know with what
generality the laws of even our most basic sciences hold. Scientists are just like
us: they get to use bound variables even though they cannot in the usual case
produce a theory that will pick out the universe of discourse over which the
variables range.
Stopping Having come all this way, a brief retrospective may make
clear the structure of the argument. What there seems to be no way of doing
is to preserve simultaneously:
a. Putnam's intuitions about "water2"
b. the Gricean reduction of meanings to pro positional attitudes