On Sat, Oct 9, 2010 at 3:19 PM, Robert LeChevalier <
lojbab@lojban.org> wrote:
>
> I'm not sure that I am understanding what the issue is in this discussion.
> "Text" is mostly meaningful in Lojban grammar as a term referring to
> parsible chunks of Lojban. Your conversation between A and B is a valid
> single "text", simply because it can parse.
when stuck one after the other (in fact they often do), but that's
just a lucky coincidence. The resulting third text does not
necessarily have any related meaning with the two original texts. If A
says something grammatical, and then B says something grammatical in
response, it may happen that if you join the two texts together you
end up with a grammatical text, but that new text may have a meaning
totally unrelated to A and B's conversation. For example:
A: do klama ma
B: lo zarci
The text "do klama ma lo zarci" happens to be grammatical, and happens
to be what you get when you join A's text with B's text, but it has
nothing to do with the conversation between A and B.
> ("Text" is still ambiguous,
> because that which lies within a parenthetical or a quote is also
> grammatically a "text", but my default usage is to refer to the largest or
> highest-level chunk
I think we are using the same sense of "text": the chunk of input that
the parser is meant to parse in one go. Quoted and parenthetical texts
are themselves part of another text, but they could also be seen as
independent in some sense, so I don't think that's a problem.
> But in case it is relevant, I have not seen any mention of fa'o, the
> reserved cmavo that explicitly indicates the end of a piece of text (in the
> sense of the largest parsible unit), but which I believe is found in no
> formal grammar and is almost never used.
It is in the PEG grammar, but it's always elidable, unless you want to
follow it with non-parsable nonsense that will be ignored.
> It was specifically conceived for
> situations where one knows that what one is saying cannot parse as a
> continuation of what has gone previously, but has been superfluous in Lojban
> parsers which were designed to inherently assume a single text.
Why would something that you put at the end of the text have that
effect? Are you saying that the intention was that B's answer to "do
klama ma" should be "lo zarci fa'o" so that "lo zarci" is not taken as
a continuation of "do klama ma"? That seems like an odd way to go
about it.
> The most likely "real" use of fa'o to me has been when one gets into some
> kind of nested parenthethetical and isn't sure how many and what kind of
> terminators are needed to get out to the highest level.
Just SU it! :)
> Perhaps no one talks about "fa'o" because as designed, it practically cannot
> be talked about in Lojban, since its use has absolute metalinguistic force.
> "zo fa'o" MAY be the only possible override (and then only if we define it
> as such). Any actual unquoted use otherwise inherently breaks off the text
> in which it is used without being part of that text.
I believe in PEG you can use it also after ZEI, inside LOhU ... LEhU,
and also of course inside ZOI (even probably as the delimiter word). I
don't remember whether BU and/or SI kill it.
mu'o mi'e xorxes
--