logical matters

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

Contrary to what xorxes seems to say, McCawley does not end up
identifying Russell's descriptions with _le_.  He has clearly moved
on from Russell's case of exactly one thing with a property to a
definite description that picks out one thing with a property when
there are several things which actually have the property (Hilbert's
epsilon, roughly).  In the selecting, this is more like _le_ but is
already very different from Russell (FixGx is false if there is more
than one G).  However, the description McCawley talks about is
still singular and still veridical, unlike _le_.  To get non-
veridicality we have to go to Kaplan's "dthat,"  which McCawley
certainly did not mention in the old edition and not, I think, in the
new either.  As for plurality, well -- as I have said -- logicians do
do that.

And's description of the current view about situations shows that
Aristotle's system seems to have held up pretty well for another couple of
decades on top of its couple of millennia.  The terminology has been
tightened a bit -- or at least changed -- and perhaps some of the criteria
mainly used have changed (it is hard to tell from this brief description
but the cyclic/non-cyclic line of my day seems to have been replaced by
something about energy use).  Still, the pattern of four types remains,
with most of the contrasts intact, and -- I bet -- most situations still
assigned to the same type, even if for a different reason.  Happily,
Lojban decided to take the types, not the classification, as its base and
so can proceed without serious changes.
pc>|83