Quoting David Sewell <[log in to unmask]>:
> On Mon, 29 Jun 2009, Steve Tinney wrote:
> I think that there should be a non-container version of <supplied>
> Obviously there's a use case for this element. Are ther other members of
> the transcriptional element group
> for which the lack of a corresponding xxxSpan element is a problem?
> (Currently there are addSpan, damageSpan, and delSpan only.)
Interesting question. Within my own domain, no: we mark corrections
per-grapheme and don't use restore, though I could imagine that both of those
might be used in a spanning way.
> > <ptr> is not allowed inside a <w> element, but I don't see why.
> Is the need for this strong enough that it should be submitted as a
> Feature Request? I think that the guiding principle behind the content
> models of <w>/<m>/<c> was to keep the contents to elements representing
> linguistic phenomena so far as possible. But obviously <w> does contain
> a lot of elements outside that category.
I can't answer that fully because I am not sufficiently au fait with TEI
philosophy. Naively, my starting assumption was that a beginning and an end
should be represented as an anchor and a ptr, but that may be just me. For
sure, some generic beginning/end pair is necessary, but if it is acceptable to
use two anchors, then maybe ptr does not need to be allowed in w.