Re: (1) loi; (2) le v. la

[ucleaar <ucleaar@ucl.ac.uk>]

> >Perhaps I am biased by English, since, so far as I am aware, everything
> >in English is +veridical. Consequently I may lack the appropriate
> >intuitions about -veridical.
>
> English is certainly NOT veridical.  Among other things, we use metonymy
> heavily, and metonymy is inconsistent with veridicality.  "The White House
> announced a new policy last night."  Houses do not announce.  And is the
> policy really "new"?   Then there is the classic JCB example of ""le"
> "That man is really a woman".  "That man" cannot be veridical if the statement
> is true.

I take it that veridicality is a relationship between a sentence and
the proposition (or fragment thereof) the grammar allows us to derive
from the sentence.  Given the word "milk", we derive a reference to
something that is milk.

>From "The White House announced a new policy" the grammar derives a
proposition which is something like "There is a house and there is
a policy and the policy is new, and the house announced the policy".
>From "That man is really a woman" we derive something like "there
is a man that is a woman". Thus both sentences are veridical: the
grammar cannot derive from these sentences the propositions
'the banana announced a banana' or 'that banana is really a woman'.
This is why English is veridical.

Both the grammar-derived propositions expressed by the sentences you
give are (in all likelihood) false.

>From here we leave language in the strict sense, and move into
more general areas of cognition and communication. From the propositions
the grammar derives from the sentences, we can derive further
propositions, e.g. by processes of figurative thought. Thus we can
take the 'literal' grammar-determined propositions expressed by
your two sentences, and from these propositions use those general
pragmatic processes to derive the propositions 'The president
announced a policy he said was new' and 'What appears to be a man
is really a woman'. This is pragmatics, and of course is very
interesting, especially to linguists, but it is certainly beyond
the bounds of grammar.

Perhaps I have misunderstood what veridicality is. Perhaps it
acts as metalinguistic pragmatic constraint denying figurative
derivations. A weird notion, but interesting. I imagine it would
be hard to square with fuzzy categories.
In this case English would be nonveridical, though a few words, like
"literally", would serve as markers of veridicality.

-----
And