There are a few things going on in the OP, some of which I'm myself trying to disentangle by writing about it. This...:
(3) Interpretation comes prior to truth value assignment — fdrake
...comes closest to what's going on, but the sense of 'interpretation' here is very particular. True, we must 'interpret' a claim before deciding if it is true or not, but quite specifically, we must also 'decide' if
this is what the truth claim
applies to; we must decide if
this is what
counts as what the truth claim is about. So: out of the blue, you say "houses turn into flowers"; I imagine my first thought is: you're not talking about roses and such - that's not what 'counts as' as flower in your locution; no, you're being metaphorical, you're being a bit annoyingly enigmatic about what you're talking about, but I'm sure it'll be cleared up if I inquire further (I suspect you're saying something like: an MMA fighter can't become a professional boxer). And that's alright.
But no, you
really mean that houses turn into flowers. After a moment of shock, assuming you're not joshing me, I realize I no longer know what
counts as a house; nor a flower. The world in which these terms took on their significance has been totally upended for me. Note that something has shifted massively between the first and second 'receptions' of the claim 'houses turn into flowers'. The 'metaphorical reception' 'fits' into the world I know: I still know, despite the metaphorical use, what here counts as a flower and house. The literal reception throws that all out of what: what counts any more as a house or a flower? I'm no longer sure, the grammar of my concepts needs to be revised; what kind of thing(s) I say about houses and flowers needs to be revised.
(I wrote, in a draft: The
kinds of things I say about houses and flowers must change entirely. Not facts, but the 'relations between facts' (and these are not to be found 'in' the facts; they are found in our grammar, in how concepts relate to other concepts) must change. The change occurs 'in the space of reasons',
qua Sellars).
Anyway, what this brings out is that there must then be
currently a set of kinds of things that I or 'we' say about houses and flowers - there must be, if this is what must undergo revision upon the revelation that houses turn into flowers. One of the points, I guess, is that
this is really hard to see. It's only at point of 'grammatical crisis', we might say, that our sedimented grammar shows itself up
as sedimented. A bit like Heidegger's broken hammer.
But there are other points to be made too: about how the intelligibility of language (of what we say) is derivative or premised upon the world, the life in which it is used; in 'this' world, houses are not the kind of thing that (can) turn into flowers. And the very intelligibility of our speaking of (what we call) houses and flowers is derived from this (more or less) stable fact. And while we can, as you said, simply stipulate (as with metaphor) what we mean by such a phrase that does not turn upon this fact, we would not be speaking about the same thing as someone who
is speaking about red bricks and peonies.
--
Another way to make the same point: one might make a distinction (which Cavell kinda does) between words which are 'only' their meanings (stipulations, metaphor), and words which take their significance from the world in which they are embedded - 'lived' meanings, as it were. Cavell speaks of words which 'have nothing
but their meanings' (which are 'merely'/only conventional), and contrasts this with words that have a relation to the world, which take their intelligibility from how things are in the world (like the fact that houses are not the kind of thing which turn into flowers!). Both 'kinds' of words are of course meaningful - one cannot
deny that metaphor and so on are meaningful; but the danger is in
confusing the two, in treating the one like the other.