Re: XML for linguists?

Charles:
> I looked at XML again after reading all this discussion;
> now it looks to me like "just" a syntax for a metalanguage,
> without any actual content. It merely imposes a tree-structure
> using what amount to circumfixes with attribute tags.
> The real value-added part is the DTD (?) which we ain't got.
>
> I still don't quite see how to apply this to my old idea of
> an interchange format for constructed languages. Somewhere
> there may be a definition of language as that which is
> its own metalanguage, or some other Godel-like constraint
> that makes it impossible to accomplish.
The major hurdle is establishing the content of the markup
scheme - the semantics, or the lexicon, as it were. SGML/XML just
provides the syntax. There are lots of descriptive and
theoretical linguistic frameworks that have developed a
metalanguage for representing language. The snag for a
"standards"-oriented project such as yours or the TEI is
there is no worthwhile consensus to be achieved on what
the 'lexicon/semantics' of the common metalanguage should
be.

--And.