Marking alteration with del@rend

[adunning]

Could the schema provide a value of ‘alteration’, ‘alt’, or something similar to mark a letter that has been modified (e.g. an o changed to a b), sometimes without technically having been erased? A couple of examples:

gladius → gladiis [Technically the letter here has not been added at all, but the inline value would presumably make most sense? One could also, I suppose, make a case for inRas, but the point is that only part of the letter has been erased to change the 'u' into an 'i'.]

gladi<subst><del rend="alteration">u</del><add place="inline">i</add></subst>s

ualeuimus → ualebimus

uale<subst><del rend="alteration">u</del><add place="overstrike">b</add></subst>imus

[stenskjaer]

My first thought was that in my view this mostly belongs in the diplomatic schema. But then, what happens if an editor wants to give this sort of information in a critical context? We could of course include it in the critical, but it would violate the principle that <add> and <del> wrap whole words, not just the changed characters.

I wonder whether we would add an attribute to the <subst> element rather than the individual sub-elements. This would decrease the granularity with which you can describe the change, but increase the generality.

The problem I have often thought of with the @rend is that it approaches infinity in the range of possible descriptions that would be necessary. That makes it difficult to schematize.

I am open to discuss any of this. And I definitely see ways of including a solution that could encode such information, and where it would be useful and work wel.

[adunning]

I'd certainly be interested in any other approaches: my aim here is to implement what the TEI and EpiDoc guidelines seem to be assuming while making it possible to render a documentary edition in a readable fashion. There are certainly circumstances in which it's useful to indicate modification within a critical apparatus, since the alternative is adding another siglum for a corrector, and that is both less efficient (at least in TEI) and more error-prone.

(Perhaps I am misunderstanding your second sentence, but I should note for the record that there is nothing saying that <add> and <del> must wrap an entire word: the guidelines at http://www.tei-c.org/release/doc/tei-p5-doc/en/html/PH.html include several examples showing the substitution of individual characters.)