Re: na`e

[HACKER G N <c9709244@ALINGA.NEWCASTLE.EDU.AU>]

On Sun, 5 Oct 1997, JORGE JOAQUIN LLAMBIAS wrote:

> >> Quantifiable pro-bridi are an abomination on the language.
> >
> >Ha, ha, ha! Okay. Would you like to explain that a little more? :)
>
> They are essentially unworkable as soon as you add a little of
> complexity. You can use, for example, {su'o mlatu} in a prenex,
> to mean "at least one cat". But you can't use {su'o bu'a} to mean
>  "at least one bu'a", because bu'a has a special rule for how it
> works in the prenex. (The way I understand it, this contradicts
> the claim of syntactic unambiguity. To keep that claim true bu'a
> should be in a selmaho of its own, i.e. the parser should identify
> it as a different thing than a normal selbri.)

Well, I agree, and it sounds like that aspect of the language isn't well
thought-out. Possibly that's something to think of for a slight
re-baseline after the five-year design freeze. As far as quantifying
selbri goes, what about talking about "nu bu'a", "su'u bu'a", or something
like that?

>As for our example, does
> it really work with cei? Is {su'o bu'a cei na vreta} "some <X> which is
> not {vreta}", or does it mean "some <X> which is {na vreta}"?
> I would have said the last one, but in any case, whichever it
> is, how do we say the other?

I don't like that rule with "cei" either. It all sounds really ad hoc. I
really think some selbri equivalent of "poi" would be preferable, or else
"poi" with an abstraction sumti of the "su'u bu'a", etc. variety.

Geoff