[lojban-beginners] Re: Enumerating in Lojban

["Jonathan Gibbons" <jonored@gmail.com>]

Ah, I think I see the trick. So, instead of using a grouping rule like
everywhere else, with the elidable terminators they decided to just
say that anything caught being ambiguous because of a lack of a
terminator was to be thrown out of the language? such that rather than
"le xekri ckafi" being disambiguated by a grouping rule like "tanru
group tighter than descriptions", it's ungrammatical, because it could
either be "(le xekri) ckafi" or "le (xekri ckafi)"? it really does
seem more sensible to leave the string in the language and apply a
grouping/precedence rule to get at the meaning, but yes, that behavior
does seem non-context-free. A bit silly, but non-context-free. If, on
the other hand, you use precedence and grouping, which is to say, the
same sorts of mechanisms that have been neccessary even to get at what
is meant by "1+2*3" rather than just say that yes, that is a valid
mathematical formula, it seems reasonably doable to me; and "le xekri
ckafi" has a meaning. It might not be what was intended, but it has an
unambiguous meaning. I seem to recall quite a bit of talk of grouping
in the reference grammar, though. CFGs are nice for describing
languages, which is to say, sets of strings, and precedence and
grouping rules are nice for picking particular derivations of a string
to associate meaning with.

-Jonathan