Belonging as I do to the Alice's Restaurant school of markup ("you can
get anything you want..."), our discussion of tei:choice has been
bothering me.

If tei:c is syntactic sugar for tei:seg[@type='character'], why can
tei:c not appear wherever tei:seg appears?

Or put otherwise: if tei:seg and tei:ab are generic models for other
elements that are largely refinements upon them, shouldn't the
refinements inherit all their ancestoral properties? It seems to me the
difference between tei:ab and tei:p is one of specificity regarding
offspring: tei:p is a special kind of tei:ab, not a different kind of
beast. It can have more narrow children, but it shouldn't be cut from
tei:ab's ancestors like some bastard cousin we don't want to talk about.

So what I'm wondering is if we shouldn't in P5 allow all syntactic sugar
to have the same rights as their more staid cousins? Once we accept that
tei:seg or tei:ab is an acceptable child of an element, shouldn't we
also admit that all cousins have the same ancestors? Virginia Woolf and
Lyton Strachey would be proud of us.

The main reason for asking this is that the non-sugary version of each
of the more specific elements is still allowed. I.e. I can encode
tei:seg[@type='character'] as a child of tei:choice, but not the sugary
form tei:c. All this does is promote inconsistent markup. Or choices
made for inconsequential reasons--i.e. if I know I'm going to have to
place tei:c in a tei:choice, I will encode using the long form
(tei:seg[@type='c'") rather than the short, since it causes me less

-dan "arlo" o'donnell
Daniel Paul O'Donnell
Associate Professor and Acting Chair
Director, Digital Medievalist Project
Department of English
University of Lethbridge
Lethbridge AB T1K 3M4

Vox +1 403 329-2377
Fax +1 403 382-7191