RE: [lojban] The Knights who forgot to say "ni!"

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

Rob:
> On Fri, Aug 31, 2001 at 06:04:28PM +0100, And Rosta wrote:
> > Rob:
> > > "whether D is a crook". Not every grammatical construct has to be
> "evaluated"
> > > to something shorter.
> >
> > In general, things *are* fully evaluated in Lojban. The times when you want
> > something that doesn't evaluate are special cases, that call for special
> > constructions, the ones that cause us so much perplexity.
>
> If you performed the kind of "evaluation" people are suggesting for {jei} all
> the time, then saying {la spat. gerku} would be instantly replaced by "true"
> and communicate nothing.

A bridi is one of those special constructions that is exempt from evaluation.
Indeed, it is *the( special construction par excellence. That is why when,
as in the Q-kau discussions, and the needing a box and seeking a unicorn
discussions, some sumti musn't be instantly-evaluated, we seek a solution
that involves embedding the sumti within a subordinate bridi, so as to
block evaluation. I'm not making this up, or laying down the law, or
speaking ex cathedra; I'm describing long-long-established Loglan/Lojban
semantic principles -- or so I believe.

--And.