🚀 go-pugleaf

In article <1992Nov17.221542.17555@husc3.harvard.edu>, Michael Zeleny (mz) writes:

mz> Even granting your alleged ability [to find contexts to separate
mz> synonyms], it is obvious that it does nothing to exclude the
mz> context-dependent theories of meaning.  

The problem here is that the phenomenon becomes purely theoretical.
You _can't_ put two phrases or whatever into the same contex for the
same reason you can't put two physical objects into the same 4-space
location, not enough room.

mz> The best NL examples are furnished by good translations.

Give me one example. Even phrases in languages I know nothing of are
not interchangable unless they are phonetically _very_ similar, which
is a bit of a degenerate case for synonymy. If I understand, even
vaguely, one of the languages, then all kinds of contextual effects
are going to ruin any equivalence you wish to set up.

mz> Following Montague, I believe that deicticaly disambiguated natural
mz> languages *are* formal languages.

This is either empty or patently false, either way it doesn't reflect
the Montague as I have seen. I think you need a different verb in that
opaquely embedded phrase :-).

--
rjc@cogsci.ed.ac.uk				_O_
						 |<