* Wednesday, 2011-10-19 at 20:28 +0100 - And Rosta <and.rosta@gmail.com>:
> Martin Bays, On 19/10/2011 05:30:
> > * Wednesday, 2011-10-19 at 03:56 +0100 - And Rosta<and.rosta@gmail.com>:
> >> Is there consensus on what fractional quantifiers should mean?
> >
> > Not to my knowledge.
> >
> >> I find it hard to think of an valid argument for piro being distinct from ro.
> >
> > There seems to be at least some consensus that {ro} is a singular
> > quantifier. {piPA} has tended to be used for other things.
> >
> > If {pi za'u} is to be a plural existential quantifier, which it would be
> > very useful for it to be, then it seems we're obliged to have {pi ro
> > ko'a} == {ko'a} (just a null-op), and have {pi ro broda} being, for
> > distributive broda, the plurality formed from the extension of broda.
> > For non-distributive broda, it's less clear.
>
> Ah, I see. So for "pi mu plise" there are three candidate meanings:
> [i] "half an apple" (Pierre's),
> [ii] "half of appledom" (my stab at glossing yours),
(sounds right)
> and [iii] "one in every two apples" (what I had vaguely thought it
> meant before this conversation).
(ii) and (iii) are likely to have similar effects; the main difference
is that (iii) has {pi mu} being a singular quantifier.
With (i), {pi mu} doesn't seem to be a quantifier at all.
It's true that having {pi so so so broda} have wholly different effect
from {pa broda}, as (ii) and (iii) would imply, is a little ugly...
For reference: the current text in the definition of {loi} on the gadri
section page appears to more or less accord with (ii) - although only
after translating it from the 'mongrel' system to pure plural semantics,
such that {loi broda} has a single plural referent rather than multiple
referents each of which is a "group"...
(At some point, we do need to write up this pure plural semantics - which
I believe at least xorxes, John Clifford and I were in agreement on, and
which seems to be necessary to make real sense of xor{lo}. But this may
be premature if, as I'm currently hoping, JC's idea that plural
semantics also solves these kinds/types issues works out.)
> >>>>> So maybe {loi} should actually be defined like that. {loi cinfo} means
> >>>>> precisely the same thing as "the lions".
> >>>>
> >>>> I think "the lions" would mean {lei cinfo}, actually, but that's
> >>>> a point about English, and doesn't contradict your underlying point.
> >>>
> >>> Just making a veridiciality distinction? Or specificity too?
> >>
> >> I don't know how sclerotic my thinking is, but I'm thinking "the
> >> lions" is {lo co'e voi cinfo} (or maybe also your {loi co'e voi
> >> cinfo}) and "le broda" is "lo co'e voi broda" (and "lei broda" "lei
> >> co'e voi broda").
> >
> > So just adding non-veridiciality?
>
> adding nonveridicality with voi, and specificity with co'e.
How does {co'e} give specificity, sorry?
Martin
Attachment:
pgpT9prNy1a4j.pgp
Description: PGP signature