Yes, you can legally say just <g type="arabesque"/> -- though obviously
it might be better to say <g type="arabesque">Ӓ</g> (where Ӓ
produces something that looks like an arabesque for you) and still
better to say <g type="arabesque" ref="#abcd"/> (where #abcd points to
full documentation of the intended mark, including mappings to Unicode
or other systems, descriptions, sample graphic renditions etc etc). But
-- for once-- the TEI also supports the Simple Way too!
Gabriel BODARD wrote:
> I've been meaning to look at this question for a while. Does this mean
> that what we are currently doing in P4 with <mark/> we can now do in P5
> using <g>, without abusing the element in any way?
> That is to say: the only information we record about the "mark" on the
> stone is (i) that it is there, on a line, in the same context as an
> alphabetic character could be; and (ii) a short description--usually a
> single word--of the mark that can be displayed or, I suppose, searched
> in some way (e.g. "scroll", "leaf", "arabesque", "dolphin"). We do *not*
> include any kind of link to a standard table, a unicode codepoint, a
> URL, a graphic, or any other external information. If that's legal, then
> it looks like our problem is solved once we upgrade to P5...
> Lou Burnard a écrit :
>> This seems strikingly similar to the <g> element introduced in TEI P5.
>> Tom Elliott wrote:
>>> I wonder if it's relevant to this discussion how epidoc is trying to
>>> handle "a mark in the original text that is not readily handled by
>>> conventional type or fonts (e.g., does not correspond to a Unicode
>>> code point) but is not really an illustration."
>>> An empty extension element with a "type" attribute, e.g.:
>>> <mark type="claudian_ps" />
>>> Tom Elliott, Ph.D.
>>> Director, Pleiades Project
>>> Ancient World Mapping Center
>>> University of North Carolina at Chapel Hill