WikiDiscuss

Posted by pycyn on Tue 09 of Nov, 2004 01:58 GMT posts: 2388

wrote:

>
> pc:
> > I do, of course, agree with what I
> > take it you are trying to say, that {na} has
> some
> > whole bridi as its scope (in some sense)
> while
> > {na'e} has at most a selbri.
>
> Good.
>
> > What I disagree
> > with is the inference from that to the claim
> that
> > they have radically different functions.
>
> I'll settle for plain different functions.

Good. But they are both negations and negations
related (apparently) in a number of ways, which
the grammar — as youread it — only partly
shows.
>
> > {na(ku)}
> > means either that some sentence is false or
> that
> > its complement is true and that latter is
> > essentially the role of {na'e} and {to'e},
> with
> > differing restrictions on what the complement
> > involves in addition.
>
> Yes, but {naku} can negate quantified and
> connected
> sentences, and {na'e} can't. {na'e} can negate
> just part of the selbri — {na'e broda brode}
> is
> {(na'e broda) brode} — and {na} can't. {na}
> and {na'e}
> may be the same when: {na} negates a bare
> sentence
> (by that I mean that no quantifier or
> connective
> operates on the sentence before {na}) and
> {na'e} negates
> a complete selbri. At that point they may
> touch.

As noted earlier, these seem accidental factors
of the grammar. There is nothing inherent in
{na'e} to prevent these other moves other than
the peculiarity of its grammar (relative to that
of {na}). One could with some effort and
ingenuity achieve the results of those
modifications in very periphrastic ways. So they
are sayable, I think.

>
> > I note that {na'e ku} fror example is
> > also unlisted,
>
> Right, because {na'e} is not a tag. A tag
> attaches
> to a selbri, whereas {na'e} attaches to a
> brivla,
> i.e. a selbri component.
>
> > which raises an interesting
> > question whether bringing the two negations
> into
> > a single class would clarify matters a bit;
> it
> > appears that it would interfere with no usage
> but
> > generate some new ones, covering cases that
> now
> > are apparently somewhat obscure.
>
> Maybe for LoCCan III. Probably not a
> possibility for
> Lojban. But it might be interesting to see if
> it can be
> done from the point of view of the syntax. I'm
> not sure
> either way.
As noted, Lojban appears to have the capability
that this envisions, it just can't do it tidily.
The result of that seems to be that some aspects
of even {na} go underused or reported — the role
of the placement of {naku} when quantifiers and
connectives are not involved, for example: you
seem to think there is none (and consequently
that descriptions are constants) for example.