RE: sets, masses, &c. (was: RE: [lojban] Re: [jboske] RE: Anything but tautol...

["And Rosta" <a.rosta@ntlworld.com>]

Jorge:
> >There are, though, still some problems. The first is that the
> >category derived from subtracting the individuating properties
> >from another category (as with mass nouns derived from
> >counts), is not necessarily equivalent to material; one
> >can massify immaterial things (e.g. misfortunes : misfortune).
> 
> All right, but articles in Lojban are not used to change
> categories. They provide different ways of referring to the
> same category. We might need something other than {marji}
> for categorical deindividuation, but I think it would
> still have to be a brivla (or perhaps some other modifier,
> something in NAhE?). {loi} only provides referential
> deindividuation.

I agree, sort of. It is not a job for gadri, but nor is it
a job for brivla. That is, we need brivla meaning "is a unit
of" and "is a mass of", but the second argument of these
predicates would better be equivalent to "loi ka broda"
than to "lo(i) broda".

It's not really a fixable problem at this late stage in
Lojban's development: any solution would be too cumbersome
to be worth using.

--And.