Re: TECH: nested bridi anaphora

[Logical Language Group <lojbab@ACCESS.DIGEX.NET>]

mi'e .djan.
.i la lojbab. pu cusku di'e

> > (A series of nonsense gismu starting with each consonant sounds
> > interesting, too, as an analogue to the lerfu anaphora - but I don't see
> > proposing it unless we find usage demanding it.)

la .and. cusku di'e

> Would these simply work like lerfu anaphora but have more signal
> redundancy, or are there other advantages, in that the lerfu anaphora
> now become infinite in number. E.g. if the nonsense anaphor always
> corefers with the last word that starts with the same letter and
> hasn't already been assigned referred to with a different nonsense
> anaphor, and if a convention was adopted for assigning referents
> a nonsense anaphor, then you have easy & possibly quite mnemonic ways of
> referring to anything in the previous discourse.

The main difference would be that the lerfu anaphora are pro-sumti, whereas
these proposed anaphora would be pro-bridi.  Currently there is no analogue
to the lerfu anaphora in the pro-bridi system: you cannot use "me by."
because that means "pertaining to B" where B is the last sumti beginning
with "b".

The same effect can be achieved somewhat painfully with "broda xi by."
meaning "broda sub B"; "broda" is the pro-bridi analogue of "ko'a".

> What exactly did you have in mind? Anaphora is awkward in any language,
> so there is a good argument for providing as many varieties as possible
> & letting natural selection choose the best.

So we do.  See my paper in casper.cs.yale.edu:/pub/lojban/draft/refgrammar,
which explains Lojban anaphora and deixis words, viz. pro-sumti/pro-bridi.

--
John Cowan              sharing account <lojbab@access.digex.net> for now
                e'osai ko sarji la lojban.