Re: On {lo} and existence

[Iain Alexander <ia@STRYX.DEMON.CO.UK>]

In message  <9502212349.aa01911@punt.demon.co.uk> pcliffje@crl.com writes:

>         pc:
>         Ahah! now I see (again, I think) the situation about _lo_ and _da
>         poi_.  _da_poi_ does not refer to anything, it is a quantifier
>         phrase, a syncategorematic, a piece of logical appratus, not a
>         content expression.  _lo_ phrases do refer to things, they are
>         descriptions on a par with _le_ descriptions and the like.

Once before I suggested that the term "description" had some meaning
to you, perhaps a technical one, that was unfamiliar to me, but I
failed to persuade you to explain.

As far as I'm concerned, a {lo}-description, unlike a {le}-description,
refers to something *in precisely the same sense* as an existentially
quantified variable does within the scope of its quantification.
In fact, I suspect it's for the same reason - they're both
(usually implicitly) quantified {su'o}.  The only difference is that
{da poi broda} uses up one of the {da}-series variables, whereas
{lo broda} effectively introduces a new variable {le broda}.

You seem to feel that a description {lo broda} refers to something
"specific", whereas a quantified variable {da poi broda} doesn't.

(Historical digression:  This list normally uses the term "specific"
to denote something _known to the speaker_.  I still haven't found
a good replacement for the more general sense of something precise,
unique (though not necessarily singular in number), which is the
sense I intend above.)

I disagree.  I think in both cases there is an (usually) implicit
prenex, potentially enclosing the entire discourse.  Looking at
the statement from outside the prenex, we of course view the
phrase (either one) as variable in meaning - we merely claim
that there exists a referent which makes the statement true.
But in the body of the statement, inside the prenex, we must
consider the phrase to identify a single (though not necessarily
singular:) referent.

BTW, if there were multiple instances of {lo broda} in an
utterance, would you take them to identify the same
referent each time?  I would imagine the situation to be
ambiguous, but with a likely presumption that they identify
different referents.

>         pc:
>         Whoa! if the quantification is _ro_da_poi_broda_ or just plain
>         _ro_broda_ then it does entail that there are broda; we set it up
>         that way.  The only "All elves" that does not is whatever has
>         become of _ro_da_kanoi_da_<elf>_ki_ and that is because of
>         _kanoi_, not _ro_ (and even it implies the rather wimpy _da_
>         _kanoi_ etc.).

This is news to me.  If this is
        {ro da zo'u:  ganai da <elf> gi da remna simsa}
        A(x):  elf(x) => humanoid(x)
I'd always assumed this to be interchangeable with
        ro da poi <elf> cu remna simsa
        ro da poi <elf> zo'u:  da remna simsa

--
Iain Alexander                    ia@stryx.demon.co.uk
                    I.Alexander@bra0125.wins.icl.co.uk