Message-Id: <m0r4D3d-00005bC@xiron.pc.helsinki.fi>
Date:         Sun, 6 Nov 1994 14:15:47 EST
Reply-To:     jorge@PHYAST.PITT.EDU
Sender:       Lojban list <LOJBAN%CUVMB.bitnet@FINHUTC.hut.fi>
From:         jorge@PHYAST.PITT.EDU
Subject:      Re: context in Lojban
To:           Veijo Vilva <veion@XIRON.PC.HELSINKI.FI>
Content-Length: 2373
Lines: 52

And:
> > > Even if we agree that the set of ro broda contains only one member, I
> > > still don't think this makes "lo broda" specific.
> >
> > For all purposes of truth values it does, I think.
>
> Not if +/-specific is a difference in how you go about evaluating
> the truth value: for +specific you find the in-mind referent, whereas
> for -specific you existentially quantify over (in this example)
> a one-member set.

Yes, but the result is the same, you always end up with the claim being
about the same referent, and the same things follow from the two cases.

We really agree, as you say. It doesn't have any effect on anything to
call {lo pa broda} specific or nonspecific.

I prefer to simplify things and say that every sumti with outer quantifier
{ro} is specific, because the "which one?" has been fully answered.

Any other outer quantifier makes the referent nonspecific, unless it is
equivalent to {ro}. For example {ro le re gerku} and {re le re gerku} have
the same referents, and I prefer to call them specific.

It doesn't really matter what we call "specific". What matters is that {le}
and {lo} have different outer quantifiers, and this results in them not
being interchangeable, even when the sumti is veridical. Since they are
not interchangeable, we don't choose one or the other depending on
veridicality, and therefore veridicality is not very relevant.

I don't remember any use of {le broda} where the referent isn't a broda,
except the example {le nanmu} that turned out to be a ninmu. But in spite
of that {le} is used more often than {lo}. Why? Because it can't be
substituted with {lo} retaining the same meaning.

> If I weren't married, but knew I was going to be, then I could talk
> about "my first wife", meaning "the first person that I'll be married
> to", without knowing who this person was going to be. This would be
> nonspecific.

I would call it specific, but it doesn't matter. We agree that it
has one and only one referent, and therefore commutes with negation.
It's logical properties are the same whether you quantify it with
{ro} or {su'o}.

> If you don't agree, then I think this shows we have a slightly but
> unproblematically different understanding of what specificity is.

Yes. It's just a matter of definition that doesn't change the fact that
what separates {le} and {lo} the most is not veridicality.

Jorge