Full Idea
The view that the meaning of language of thought expressions is based on their conceptual role (derived from Wittgenstein's idea of meaning as use), is most plausible for the logical connectives like "and", but implausible for, say, "animal".
Gist of Idea
The meaning of "and" may be its use, but not of "animal"
Source
Georges Rey (Contemporary Philosophy of Mind [1997], 9.1.2)
Book Reference
Rey,Georges: 'Contemporary Philosophy of Mind' [Blackwell 1997], p.239
A Reaction
It was the logical connectives that got LW started on this track. If it doesn't work for 'animal' then does that mean we need two different theories?