Dick:
> And:
> And to Jasper:
> >I'm eager to see the evidence. As you could probably guess, in the
> >face of such evidence I would conclude that that stuff is in fact
> >expressed in (phonologically invisible) syntax. The more substantive
> >issue, though, is whether er and ee can be defined in purely
> >semantic terms (as Agent and Theme can).
> ## Is anyone suggesting that semantic notions have to be defined in
> *purely* semantic terms? I'm happy to see them (like everything else in
> language and cognition) as reflecting a convergence of many different
kinds
> of patterning; so in this case, I think the concepts of semantics reflect
> links to syntactic and lexical concepts as well as real-world links. In
> other words, as I said earlier, I think semantics = thinking for speaking.
Semantic notions that aren't defined in purely semantic terms, I would
call 'syntax'. That is, for me something is syntactic if it can't be
defined in purely semantic terms.
--And.
|