On Fri, 24 Sep 2004, Syd Bauman wrote:
> Have you already considered using <w> and discarded the idea? If not,
> given your concerns about subtype= of <seg>, you may want to make an
> argument to use <w> instead. In addition to a type= attribute (which
> could be pressed into service for level_1 and level_3), it has both a
> lemma= and a function= attribute, which you might find useful.
This is an interesting point, and would be worth bearing in mind when
people come to address comparable issues in other languages. As far as
Sanskrit is concerned, I have deliberately shied away from <w> because
one of the requirements is to analyse compounds into their members,
and a tag whose basic meaning is "word" fits uncomfortably into such a
scheme. (Sanskrit compound members are bare nominal stems, not
Dr J. D. Smith * john.smith@...
Faculty of Oriental Studies * http://bombay.oriental.cam.ac.uk
Sidgwick Avenue * Tel. 01223 335140
Cambridge CB3 9DA * Fax 01223 335110