[tei-council] s vs seg, ticket 578

[Lou Burnard]

The reason we have both <s> and <seg> is that an eminent corpus linguist
(now sadly deceased) opined very strongly that there should be a TEI
element which enabled users to divide a text  into smaller units (as is
commonly done in many corpora) which did not nest and which tessellated
the text completely. That element is <s>. It was pointed out at the time
that a more general kind of segment which could self nest and which was
not required to tesselate the entire text would also be very useful.
That element is <seg>. I don't understand why this distinction , which
is pretty clearly stated in the Guidelines (see eg 
http://www.tei-c.org/release/doc/tei-p5-doc/en/html/AI.html#AILCW) , 
seems to have become
problematic all of a sudden. Do people think the distinction is not
useful? Do we want to abolish one or other of these elements? (no point
in keeping both if they are to be used in the same way)? Do we want to
swop the names over? Obviously <s> is a special case of <seg>,
so we could remove it, but that seems a bit unkind.