[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[lojban] Re: Is Lojban a CFG? (was Re: [lojban-beginners] Re: Enumerating in Lojban)
Probably because the alternative can be very unintuitive.
Our impulse with handling unintuitive strings differs, then; my
impulse is that they be noted to be unintuitive, and therefore bad
style in communication involving humans, throwing up a nice big red
flag saying "humans, this probably doesn't mean what you think it
means", and leave it formally in the language; I'd rather have
statements that are unintuitive than a language that is so
tantalizingly close, but not quite, context-free. It'd only apply to
those areas that are not currently in lojban, so it doesn't seem to me
that anything would be lost. I've yet to meet a programmer, at least,
who when faced with uncertainty about which way an expression will
parse, doesn't just insert parens to force it to be the way they want;
that's what I had envisioned elidable terminators doing; providing a
quick and straightforward method of forcing a string to parse in the
way you meant, not the way it would be without the terminator.
Wouldn't left-grouping give you:
(le (nu (le broda) brode) brodi)
You are in all probability correct; I was assuming the existence of
something to force the fragment "le broda brode" into being "le (broda
brode)", because that was the behavior I remember being told of when
the terminator was elided. That's what I get for posting at 0300...
and why I check homeworks after sleep if I'm doing work late/early.
It's also, of course, not neccessarily the best grouping to use; I was
just pulling up an example that would disambiguate all of the
terminator-less strings, not make them behave in sensible manner to
humans.
So basically, what I'm thinking, is let the formal language keep those
statements, let the formal parsers parse them, and avoid using them
because they're unintuitive. That also would be the behavior I would
prefer for something serving the function of jbofi'e - it'd
misunderstand, which would point me to the need to change the string
to force it to parse the way I meant, rather than just saying "That's
not lojban."
Not saying that humans should use these statements, just that a
decider for membership in lojban should accept them. Adjusting the
definition in such a manner puts that decider into one of the most
studied and understood classes of deciders in existence, rather than
an ad-hoc or less well-explored decider.
-Jonathan
To unsubscribe from this list, send mail to lojban-list-request@lojban.org
with the subject unsubscribe, or go to http://www.lojban.org/lsg2/, or if
you're really stuck, send mail to secretary@lojban.org for help.