reference, grice, and irony

this post might be a bit obscure and probably a lot shorter than it should be, but here goes anyway.

quick review about the nature of refernce give to specific words (as i understand it):

descriptivists and bundle theorists (russell, searle, maybe frege) think that proper names don't really refer, but are in fact abbreviations of descriptions. so a name like "stephen colbert" refers to {something that is a person & something that hosts a show right after jon stewart's show & something that is the subject of the website colbertnation.com & something that is running for president in south carolina &... etc}. this leads to somewhat deterministic view of names.

direct reference theorists (kripke, braun) say that proper names designate their objects and nothing else. so "sam clemens" in this context only works as a name that refers to the lump of atoms (or whatever people are) that was linguistically baptized "sam clemens" by his parents at his time of birth. but this runs into the problem of explaining how "mark twain" is "sam clemens" while explaining at the same time how this equation isn't trivial.

now comes the grice part: his approach to language is coming at it from a completely different angle. the idea is that language works because there are certain intentions in the speaker to create to the listener both a perceived meaning and the awareness of this intention of the speaker by the listener (by performing an utterance).

all righty... now to the irony part. while discussing grice, it came to me that certain expressions, or performance utterances follow grice's rules for non-natural meaning, that

1. person A intends person B to believe that x by uttering 'u'
2. A intends B to recognize A's intentions of B believing that x by uttering 'u'
3. B's belief that x is directly caused by A's utterance of 'u'

yet in certain situations, we still won't want to say that 'u' means x. and this is the type of situation that one calls a the village idiot, w, "brilliant." when we call w an idiot, we're attributing to him the quality of being an idiot. now how can we explain the fact that we understand the intention of the speaker while the words she's uttering are exactly the opposite of what she means?

if you adopt a direct reference theorist's ideas about names then the utterance "w is brilliant" is interpreted as the heterogenous set of:

<[the individual designated by a name, or w], the property of being brilliant>.

the direct theory of reference has little to say about metaphors, as far as i know (although i'll admit i haven't read davidson on metaphor just yet), so it seems to me that a theory of direct reference would have to provide a separate account of metaphors altogether.

however, if you adopt a descriptivist account of reference, then irony seems more explicable. consider just what "w is brilliant" might look like to them:

<[the descriptions that define w, including him not being brilliant], the property of being brilliant>

of course, in this view there would be an inconsistency in the properties being assigned to w and thus the ironic effect is captured, as that inconsistency.

i think this might be a point for the descriptivists :D