[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
xorxes:
> la djan cusku di'e
>
> > > But if you divide {pa nanmu} in many parts,
> > > you don't end up with {so'i nanmu}, no matter how big the
> > > pa nanmu was
> >
> >Ah, but you can, if "pa nanmu" means "a regiment of men"
> >(which is justified on the same basis as saying "pa birje" can mean
> >"a vat of beer"). Gricean considerations will block this in most
> >cases, but not all
>
> If this is so, then we're teaching the word wrong, because nobody
> is using it like that. The gi'uste should be modified to read
> "x1 is a quantity of men"
>
> But that's just a disagreement about the meaning of {nanmu}
> If we agree that {nanmu} is about quantities of men rather
> than plain men, then we can define a lujvo {naurka'u}, "x1 is
> a man-quantum", for the normal meaning of English "man". Then
> {pa naurka'u} cannot be divided into {so'i naurka'u}
This is a good reply, and I think we must conclude that "pa nanmu"
does mean "one man"; that is not just merely the most probable
interpretation.
I think the best thing is to see Lojban as working like English
(and yes, I *have* been passionately in love with English for
over 30 years, but I don't think this makes me blinkered). In
English, any noun can be used as a mass noun or a count noun.
But you can mostly predict from the sense of the noun whether
it is more commonly count or mass:
"a dog" is more common than "dog"
"water" is more common that "a water"
"apple" and "an apple" and "brick" and "a brick" are about
equally common.
Where Lojban ought to depart from English, though, is in the
arbitary-seeming restrictions on the interpretation of some
normally-mass nouns used as count nouns. So "a rice" can't
mean a grain of rice, say.
In summary, then, the sense of the Lojban predicate *can*
specify predicate-specific criteria for intrinsic boundaries,
as with "nanmu", or it can leave such criteria unspecified,
as with "djacu". If you use a Substance-gadri with "nanmu"
then the interpretation must 'erase' the intrinsic boundaries
from the meaning. If you use an Individuals-gadri with
"djacu" then the interpretation must add intrinisc boundaries
to the meaning.
All I would add to this is that IMO Collective-gadri require
criteria for intrinsic boundaries to be added or left unerased.
> >In short, I think of the notion of collectives as orthogonal to
> >the individual/substance distinction. A collective may constitute
> >an individual or a substance
>
> I can agree with that. But gadri are not supposed to be
> predications on the referents of the term, but rather different
> ways of referring to the same objects. For the same set of
> referents {le'i broda} I can make two claims {le broda cu brode
> ije lei broda cu brodi}. The objects described are the same, the
> gadri only indicates how they enter into the relationship
> {le ze plise cu xunre ije lei ze plise cu se culno le tansi}
> "each of the seven apples is red, and the seven apples as a whole
> fill the bowl". I describe the same objects with different gadri
> because they enter each relationship differently
>
> The individual/substance distinction, on the other hand, seems to
> be used to describe different objects, so it is a distinction to
> be made by using different selbri. There is no "apple stuff" that
> corresponds directly to the set of seven apples, so if we were
> to have a substance gadri it would not be related to the set in
> the way other gadri work
What you say here is reasonable. But the individual/substance
distinction applies to every predicate, in the sense that every
predicate can have an individuals-version and a substance-
version. Or rather, every *sumti place* with have two versions.
So if every sumti place has two versions, it is impractical
to insist of as solution as ad hoc as lujvo creation. So if
we were to reject making the distinction in gadri, we would
need a way of doing it other than using lujvo.
--And.