[YG Conlang Archives] > [engelang group] > messages [Date Index] [Thread Index] >
* Tuesday, 2012-09-11 at 23:12 -0300 - Jorge Llambías <jjllambias@hidden.email>:
> On Tue, Sep 11, 2012 at 7:38 PM, Martin Bays <mbays@hidden.email> wrote:
> > * Tuesday, 2012-09-11 at 11:15 +0100 - And Rosta <and.rosta@hidden.email>:
> >>
> >> The current situation is the one the truth-conditions of the l-
> >> formula is evaluated against. I think distinguishing situations from
> >> UoD is a helpful move.
> >
> > Maybe. But in situational semantics (disclaimer: I don't actually know
> > anything about situational semantics), claims are still meant to be
> > evaluated against whole worlds.
>
> I don't know anything about it either, but the second sentence from
> http://plato.stanford.edu/entries/situations-semantics/
> says pretty much the opposite:
>
> "In situation semantics, linguistic expressions are evaluated with
> respect to partial, rather than complete, worlds."
Hmm. I was going off Kratzer's '89 paper, where she seems to use
situations purely as calculational devices - declarative sentences are
interpreted as claims that the sentence holds in the world, but the
interpretation process sometimes involves considering situations. In
fact she specifically ensures that sentences are "persistent" - true in
a situation iff true in the situation's parent world.
But yes, some things in the encyclopaedia article (see section 3 in
particular) are doing it differently, and more in line with what
I thought And might be doing.
So quantifiers would (unlike in Kratzer) be evaluated in the obvious way
in a situation, with no mention of the parent world?
I'm not clear on how to handle intensions, though. e.g. would it make
sense to say something like
"la mlta [a might have many kittens]" ,
or would the assumption that there's a unique cat bleed over into
other worlds? Would it make a difference if "might" were replaced with
"will"? I'm assuming that "does" is right out.
> > In what sense is that equivalence to be read? If we interpret both
> > sentences within the same situation, and that situation satisfies
> > the presupposition, then of course they get the same truth value.
> > Is that all that's meant?
>
> Yes, that's about it.
>
> > Or do you mean something stronger, that when speaking I can
> > substitute one for the other and expect to convey the same
> > information? I don't see how to get the latter.
>
> How is it stronger? When speaking, I am describing some situation, and
> therefore I can substitute one for the other and expect to convey the
> same information.
I'm not sure anymore what I was worried about. So the presupposition
breaks out of scope, and we have
na la mlta xkra --> "ATTENTION: evaluate the following in a situation
containing a unique cat! It's false that the
unique cat is black" .
OK.
Martin
Attachment:
binSAch0eS4qa.bin
Description: Digital signature