terch:opaque

["John E. Clifford" <pcliffje@CRL.COM>]

After a couple of weeks with the theoretical logicians, it is a
pleasure to get back to practical McCawley.  Skipping over what
these worlds are, he gives some facts about the way they behave,
which we need to take into consideration.

First, the worlds for cognitive states (belief and its kin) are
different from the those for the volitional states (wanting,
intending, etc.), with dream worlds falling somewhere in
between.  Cognitive worlds are more nearly worlds, including
time spans (so other worlds in the temporal order) and other
people's belief worlds, indeed, even other possible-world
structures. Except as noted overtly, a belief world is assumed to
mirror the real world.  But it may differ in marked ways,
containing entities the real world does not, failing to contain
entities the real world does, and, of course, assigning to its
entities properties that they do not have in the real world.  A
belief world may even be inconsistent, a very big situation, then.
Finally, belief worlds endure largely unchanged through other
changes.

The upshot of this for opacity questions is that references to
beliefs really are opaque, that we cannot formally take an object
in such a context as being also in reality.  However, barring overt
signs to the contrary, we can make the move pragmatically, for
the assumption is, remember, that the belief world mirrors the
real unless marked.  But the marking may be late ("Oh, I thought
you know that he has this weird belief that I have a son, so, when
I told you that he believed my son was dropping out, I was just
bringing you up to day on his folie") so the formal argument is
invalid even for such clear cases as "my son" and proper names
(even self references can get skewed).

Because belief worlds endure, we can revisit them in the course
of a conversation and take into these revisits the anaphora from
previous visits, even after a long absence and even if the entry is
somewhat different ( that is, not "believe" again but some
operator that entails belief).

Volitional worlds are among the entries which rely on belief
worlds, so that , for example, he can want his daughter to marry
my (believed) son.  and once that volitional world is established,
it can be the basis for further ones, that he wants to name their
first child after Nelson Mandela, say.  But these volitonal worlds
are much smaller than belief worlds and do not endure in the
same way.  We tend to have to restart the whole volitional scape
to get back to them after we have left them.  But again they can
introduce totally new things (or lose others) relative to even the
belief world they start from and anaphora carries over as long as
the same (possibly developing) volitional scape is before us, even
as the superordinate state markers change.

The relation between dream worlds and the real (or even waking
belief) world is more complex.  One can, for example, dream that
one is someone else (real, believed real, or fictional) and even
that person is interacting with with one's (real/belief) self: "I
dream that I was Brigette Bardot (Okay, so I'm that old!) and was
kissing me" (not, note, "kissing myself," which, in this context,
would have BB kissing BB).  Dream worlds last only as long as
the dream and need not contain anything beyond the dream
material, but, so long as it is the same dream, we can return to the
anaphora from it fairly easily.

McCawley deals rather cavalierly (but no probably correctly and
certainly simply) with the problem of representation objects.
Reminding us that range of variables is the universe of discourse,
not reality, he holds that, when we are talking about art, the stock
material of art is always in the universe of discourse and, when a
particular work is at issue, its content is, of course, in that
universe.  So, the art work references are not opaque -- although
some care is probably needed to distinguish art work objects from
the real objects they seem to represent (and, in dramatic art, from
the real objects that present the representations).

I'm not sure whether any of this solves any Lojbanic problems, but it
does give us some useful data to consider in working for solutions.
pc>|83