[tei-council] notes on TC Critical apparatus

[Syd Bauman]

AC> - the attribute @with has as data type 'data.pointer' when within
AC> the class att.textCritical or att.rdgPart, but has data type
AC> 'data.code' when it is used as attribute of <witDetail>. I can
AC> see a possible reason for this, but I thought I pointed it out if
AC> we are aiming to get rid of these inconsistencies as much as
AC> possible.

LB> A palpable error. It should be data.pointer everywhere: I have
LB> changed witDetail accordingly.

I'm not sure I agree. As currently defined (over my objection),
data.code and data.pointer are almost semantically identical. But if
I can spare any vestige of my preferred distinction between the two
(a data.pointer has no semantics of its own, and thus really must
point to something, whether in the current document or elsewhere; a
data.code may make use of its value to provide the semantics, so it
may point to something, ostensibly in the current document perhaps by
inclusion), then I'd say they should all be data.code. After all, if
a <witness> can be empty, why require that it be there?


LB> Though I begin to think witDetail is a really stupid idea.

I've never been really fond of it myself.