* Monday, 2011-08-22 at 12:28 -0300 - Jorge Llambías <jjllambias@gmail.com>:
> On Mon, Aug 22, 2011 at 7:23 AM, Martin Bays <mbays@sdf.org> wrote:
> >
> > We seem to want that a simple sumti, like {lo broda} or {ko'a}, should have
> > interpretation a Whole, which I'll denote [lo broda] resp. [ko'a].
>
> If we talk in terms of plural reference, then there need be no Wholes.
> Presumably both metalanguages are reducible to one another, but I'm
> not sure if the reduction is always trivial. I think
> Whole-metalanguage goes with purely singular reference, and plural
> reference should not be mixed with Whole-metalanguage.
Agreed. In this subthread, I was trying out the mereological "Wholes"
approach John Clifford appeared to be suggesting, but precisely running
into the issue that for quantification to work as we want it to, we seem
to have to introduce global atoms - and then it indeed reduces to the
"plural reference of individuals" approach.
I'm thinking here in terms of "global atoms" and "global individuals",
such that translation between the two approaches is just given by
identifying atoms with individuals.
You seem to want contextually dependent atoms; do you correspondingly
want contextually dependent individuals? (see also below)
> At least part of the reason we often don't seem to get very far with
> these discussions is that everybody uses their own choice of
> metalanguage and the confusion created by the mix often clouds any
> points of contention.
>
> If we use Wholes, then (at least in my understanding of Lojban) "ro"
> is not the universal quantifier over Wholes. I suggest using "ro'oi"
> for the universal quantifier in that case, which corresponds to the
> plural universal quantifier in terms of plural reference.
>
> "ro" quantifies over atoms in Whole-metalanguage, and it is the
> singular universal quantifier in terms of plural reference.
Yes, I think we're speaking the same metametalanguage.
> This is what causes the weird mix we have of plural reference and
> singular quantification, (or alternatively reference to Wholes with
> quantification over atoms) but no other option seems to do what we
> most often want to do.
Yes.
For whatever record, I feel I should point out that I'm not actually yet
convinced that introducing plural reference (/Wholes) was ever a good
idea, i.e. an improvement on having always-distributive predicates with
groups-as-individuals to substitute for collective predication. Do you
have any killer argument for that?
> > How, without invoking absolute atoms, can you give a meaning to
> > {ro ko'a broda} based only on the Whole [ko'a] and on the meaning of
> > {broda}?
>
> I agree that we end up invoking atoms, but maybe we mean different
> things by "absolute atoms". In some (many) contexts the atoms will be
> people, in other contexts they may be human scale "dacti", and so on.
> And these atoms can change from sentence to sentence, and perhaps even
> from reference to reference in the same sentence. What we probably
> don't need is context independent absolute atoms. Once we have settled
> on an interpretation, then we do have absolute atoms for that
> interpretation, but there's no guarantee that the next sentence won't
> bring up the need to reassess what the "absolute" atoms are.
I don't see that this presents a problem for the "absolute atom"
approach.
It does if the 'part of' (aka 'among') relation is considered to have
anything to do with the world model, e.g. a limb being part of its owner
or a drop being part of an ocean, so perhaps the mereological
terminology is misleading. I'll switch to plural reference, where the
question is whether we can have a context-independent notion of
individual. (To be clear: here I am understanding 'plural reference' and
'individual' to mean that any term is interpreted as having
a referent-set which is a set of individuals.)
Now prenu and dacti and nu broda and so on are all individuals. So {lo
broda} can have any of them as referents, as long as they broda; which
of them it does is of course context-dependent.
So in other words: I don't see why the reassessment you mention can't
just be of what the referents of this next instance of {lo broda} should
be, rather than of what counts as an individual.
> > I also don't see how to formalise your atomising poi.
>
> I think my "poi brodi" can only be suggestive: by introducing a
> predicate with a strong affinity for certain type of atoms (like
> "prenu"), you help the listener pick the right interpretation. But I
> think you can't ever force the listener into an interpretation, you
> can only guide them there. If "prenu" has an interpretation where it
> is non-distributive over certain atoms, adding "poi prenu" does not
> exclude the possibility that the reference is to those atoms (or to
> Wholes that have those atoms as its parts).
>
> "lo sruri be lo dinju", without other contextual clues, could refer to
> people, rocks, sidewalks, air, groups of people, who knows what else.
> "lo sruri be lo dinju be'o poi prenu" could be people or groups of
> people, but it's much more likely just people because if the atoms
> were meant to be groups of people you could have said more clearly "lo
> sruri be lo dinju be'o poi gunma be lo prenu".
Again, can't this just be treated as giving information to specify which
set of (absolute) individuals is the intended referent-set of {lo
sruri...} (equivalently: which of the absolute atoms are among the
intended Whole)?
Martin
Attachment:
pgpsPO5xDGDOd.pgp
Description: PGP signature