[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: More about questions and the like (was:What I have for dinner...")
- Subject: Re: More about questions and the like (was:What I have for dinner...")
- From: Pycyn@aol.com
- Date: Sun, 12 Dec 1999 20:39:42 EST
One names-as-predicates. It is odd semantically: names usually (certainly in
English) don't have a sense, just a referent -- unlike predicates, which have
both -- and the referent is an object, not a set -- as it is for predicates.
On problem 2, using the nap approach makes all sentences about Pegasus true,
which is as objectionable in context as the non-denoting names approach that
makes them all false (or undefined). Pegasus was the winged horse but
Pegasus was not a unicorn. With the xu'a approach, this sort can take place
without odd readings of names and as part of a general rule about intensional
operators, which we will need anyhow.
As for problem one, if John doesn't know Paul as Paul, he probably does not
know that he has the property is-Paul either and so not that something both
is Paul and went to the party, i.e., that Paul went to the party on the one
reading. On the other hand, it probably does cover the other reading, that
there is something which is-Paul and John knows that it went to the party.
But this does not require the odd predicate is-Paul (rather reads it as "=
Paul") to work. Indeed, that is one general solution for these cases, treat
"Paul" or whatever as an external "quantifier" to work in: in Lojban, set
some variable to "Paul" in the prefix: "For x = Paul, John knows that x went
to the party." The problem is with the name inside the the intenly waysional
context, not outside. The trick is always to disambiguate in the less likely
way -- when the thing involved is real. So the real trick is to know when
that is.
pc