On Tue, Oct 12, 2010 at 3:22 PM, Robert LeChevalier <
lojbab@lojban.org> wrote:
>
> If the
> language is defined to extend into multiple units called "text", the
> boundaries between "text"s need to be unambiguous, and only a parser that
> accurate reflects those boundaries is "correct".
the generator of the texts. It can generate infinitely many different
texts. The idea that it can only generate one text just doesn't make
any sense. I don't even know what you mean by it.
> Right now, there are no formal rules governing multiple texts and their
> possible interactions.
And that's just as it should be. There are also no rules in Lojban
about how many bananas you can eat every day, or at what time you
should take a shower, so why should there be any rules about how many
texts you can produce, or when? It has nothing to do with the formal
grammar. All the formal grammar does is tell you how to generate a
correct Lojban text, not when to generate it, who you give it to,
under what circumstances, and so on.
> Perhaps there should be, but I am inclined to think
> we should wait till the BYFY finishes the simpler "single text" problem
> we've been stuck on for years before making the job harder.
Indeed. And we should avoid legislating on things that don't really
have anything to do with what a grammar is supposed to do.