RE: [lojban] Why is there so much irregularity in cmavo/gismu?

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

> From: Bob LeChevalier (lojbab) [mailto:lojbab@lojban.org]
> At 02:08 PM 11/15/01 +0000, And Rosta wrote:
> >If it makes sense to ask a yes/no question about a specific part of
> >the sentence then it also makes sense to affirm or negate a
> >specific part of the sentence. Just as xo behaves like a PA and
> >ma behaves like a KOhA, so xu should behave like a JAhA.
> 
> No, that is an argument for all the JAhA to behave like a UI, except that 
> then we would lose the capability for logical negation.

That's why xu/ja'a/na should be in JAhA rather than UI.
 
> My answer is that the strict question which a JAhA answers is not "xu", but 
> ja'axu or naxu, but we interpret sentence scope xu to be the same as 
> ja'axu.

Not a bad answer, but in {du'u xu kau broda} the {xu kau} is to be
replaced with {ja'a} or {na}, and by the usual rule of replacing
like with like, that would mean that {xu} should be in JAhA.

--And.