Re: [lojban] Re: "la" rule

[Bob LeChevalier <lojbab@lojban.org>]

Mark E. Shoulson wrote:
> Bob LeChevalier wrote:
>
> > For all that Mark thinks that the problem will be lasting, it should 
> > be trivial for the tools - lexer/parser/glosser etc, to include such 
> > things as a check for "la" and related strings in what otherwise 
> > parses as a cmene token, and flag it with a clear error message.  In 
> > the absence of  skilled editors, the classic standard of evaluating 
> > Lojban text for publication has been to run it through the parser.
>
> "You must have faith in the future, young one.  Soon, everyone will 
> always have computers to check their grammar."

No.  Many will have computers to help them LEARN their grammar, for 
those features of grammar which are necessary for computer 
understanding, but which don't seem to be a problem for humans.  And 
this one is a very trivial check for a computer, if not for humans (who 
aren't so hot even at proofreading for spelling errors in their native 
language, or we wouldn't have spell-checkers).

> I should put "won't this be solved by better technology?" as one of the 
> responses on the wiki page.

The la rule isn't required with standard (spaces included) orthography, 
and without a speaking community.  When we have a skilled speaking 
community, we can find out whether la in names in the spoken language 
causes problems for humans - after all, humans can parse English and 
other human languages successfully, and they aren't audiovisually 
isomorphic.

We know that la in names would cause problems for a computer with no AI 
capability at all.  So it may take a computer to be stupid enough while 
rule-bound to recognize that it is misunderstanding something that 
humans have no trouble understanding.

lojbab