That's fascinating -- and, at the same time, I'm glad I don't have such
complicated issues to contend with! OK, then, now that I see what you're
really doing, I agree that treating the speaker-repeats as a variety of
<fw> makes as much sense as anything else.
On Sun, 28 Oct 2001, Ingo Mittendorf wrote:
> Just a quick reply.
> The original is a manuscript. 'Speaker' headings are repeated after the
> page-break everywhere (once there's a mistake with the wrong speaker turning
> up, never mind).
> Unfortunately, the spelling in the text (it's in Welsh) is rather variable
> and unpredictable. We do not and cannot regularize the spelling (it's not
> done with medieval and early modern Welsh texts). This makes omitting the
> speaker-repeats a problematic issue.
> Page-breaks we need to document as we need them as a reference system (word
> X occurs on p. 19 ...).
> So, all these exits are blocked.