Re: [lojban] About plural 'ro'

[Jorge Llambías <jjllambias@gmail.com>]

On Wed, Apr 21, 2010 at 10:21 AM, John E Clifford <kali9putra@yahoo.com> wrote:
> Well, I don't like the the 'ganai... gi ..' format since that would allow for empty sumti, which are undesirable (if not incomprehensible) for a variety of reasons.

It seems to me that the two issues are independent of one another.
Empty sumti are indeed weird objects, but they are weird in any
context, not only when used in "me <sumti>". The weirdness of:

    ganai da me <sumti> gi ...

in a context where <sumti> doesn't have any referents comes from the
fact that <sumti>, which is supposed to be a referring term, doesn't
have any referents, not from the fact that  "da me <sumti>", "da
is/are among the referents of <sumti>" will then have to be false, and
therefore "ganai da me <sumti> gi ..." true. If faced with such a
sumti, the listener will respond with na'i or ki'a, i.e. "I refuse to
enter the universe of discourse you are leading me to", or "please
clarify because I don't get where you are leading me to", or perhaps
even "ja'o", "ok, I'll play along, let's enter the realm where this
apparently referentless sumti does have a referent".

But anyway, all this is leading us astray. Let's focus on sumti with
clear referents. Given a sumti with (one or more) referents, my
contention is that:

        ro <sumti> cu broda
=      ro me <sumti> cu broda
=      ro da poi ke'a me <sumti> zo'u da broda
=      ro da zo'u ganai da me <sumti> gi da broda

and that those equivalences should hold whether ro/da are defined as
plural or singular.

If ro/da are defined as singular, all four expressions say that each
referent of <sumti> is broda (we probably agree about this).

If ro/da are defined as plural, then my contention is that all four
expressions say that any one or more of the referents of <sumti>
broda. You disagree with this contention. You say that even with ro/da
defined as plural, we need to know something else about the referents
of <sumti> before we can decipher (some of) those four expressions. We
need to know _how_ those referents came to be referents of <sumti>
before we can say what those expressions mean. That's what I find odd.
(Forget about empty sumti. We are under the assumption that <sumti>
has one or more referents.)

> And yes, there is no general rule for 'ro<sumti>' nor can I think of a reason to expect one.

The obvious reason is simplicity. It seems to me that the onus is on
you to explain why we want <sumti> to bring anything more than its
referents into the question.

> Sumti refer to things in a variety of ways and quantifiers naturally take these differences into account.

Why should sumti refer to things in a variety of ways? The obvious
starting point is that they refer to things period. It is the job of
the quantifier, not of the sumti, to specify how those referents
relate to other referents.

> I we are going to have plural quantification, then that has to be the fundamental form and others derive from it.

Yes. But it seems to me that what you are introducing is not just
plural quantification, but plural quantification plus some other
information contained in sumti besides its referents.

> The most obvious way to deal with the problems that appear to arise from this is to adapt the rules for quantifiers to what is quantified over (given that we are now in fact quantifying over things that no actual logic quantifies over and so we are winging it).

I don't understand this point. If "ro" is singular, it quantifies over
the set of referents. If "ro" is plural, it also quantifies over the
same set of referents (but in the way that plural quantifiers do it).
The things over which they quantify are the same things in both cases.

> If this gives undesirable results, then perhaps we need further rules about transitivity, though these get increasingly hard to formulate.  Alternatively, we can do away with plural quantification. which leads to problems,
>  given that terms have plural referents and instantiate bound variables, so that, then, variables don't cover their instances.

If we use singular quantification, it is the referents of the terms
that instantiate bound variables, but only one referent at a time.
With plural quantification too, it is the referents of the terms that
instantiate bound variables except in this case more than one referent
can instantiate at the same time.

> Of course, we can also drop plural reference and go back to singularity and get plurality explicitly when needed by the distinction between, say, 'lo' and 'loi' (well, not plurality exactly, rather the interesting correlate of it, collectivity).

We could do lots of things. Personally, I don't like the loi mess.

> Quantifiers continue to work differently depending on what the term applied to is -- 'ro lo broda' is presumably partitive, 'ro loi broda' multiplicative, and so on.

How can "ro lo broda" be partitive without plural reference? What is
the single referent of "lo broda" in that case?

mu'o mi'e xorxes

-- 
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to lojban@googlegroups.com.
To unsubscribe from this group, send email to lojban+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.