Why care about lexical semantics?

I *love* formal semantics.  I love it because I love formal systems generally, especially set theory, and I love the structure of languages.  However, I can't see much value in it (besides the sheer fun of transcriptions, better than Xbox for sure). 

 

Philosophers should be interested in arguments.  An argument is a kind of assertion.  An assertion is a kind of speech act.  Thus arguments are a kind of speech act. 

 

But what is asserted has a tenuous tie to the lexical semantics of an utterance which is *used* to assert a certain proposition.  Depending on the context of utterance, there are almost no conceivable limits on what proposition can be asserted by a given utterance.  That is, we can *use* utterences in very creative ways to *assert* a wide variety of things.  

 

Note that grammatical categories don't always map nicely onto speech acts.  For example you can use a declarative sentence to issue a command: "There's the door!"  It can also go the other way around, you can use an imperative sentence to make an assertion: "Get that idea out of your head!"  You can also use an interrogative sentence to make an assertion: "Do you really think you can fool me?"

 

Then also consider how in different contexts of utterance the same sentence-type can have so many different "meanings" (to use the term in a quite natural way).  "It's hot" can express a sober assertion; an imperative "Don't touch it!"; a conditional "If you don't talk, I'll burn your arm!" or even ask a question given the intonation. 

 

I've heard one philosopher of language suggest that utterances express a plethora of propositions and context picks out which one we have asserted.  Expression is taken to be fixed by lexical semantics and context narrows the scope of what is asserted.  The expression relation is somewhat mysterious to me, but even leaving that aside, I think that on this view there would be so many propositions expressed that it would reduce the contribution of lexical semantics nearly to nil. 

 

Pragmatics and context is where the action is, it seems to me.

PS - #5 on Kent Back's Top Ten list of misconceptions about implicature is that "For what is implicated to be figured out, what is said must be determined first."  I think that helps what I'm saying here, though I think many would accuse me of being subject to #8.  That might be so, but I'm not so much concerned with the specific notion of implicature that Grice talked about as the general phenomenon of relying on extra-linguistic facts to use utterances to do all kinds of things not very tightly connected to lexical meaning.  

Bach's full piece is here.