Thank you all for your thoughts.
> - For once we have to distinguish between "editorial" interventions that
> took place inside the source and those carried out by the modern editor.
I was indeed assuming that when the definition of restore mentions 'restoration
of text to an earlier state by cancellation of an editorial or authorial marking or
instruction' that word *editorial* would refer to a modern editor. I now realize
that is not necessarily the case.
> I suppose that Peter was talking about an authorial deletion which a
> modern editor would like to restore, right?
> if you want to include that information you could also look into the @status
> attribute of <del>. It’s scope is a bit different, but in another project I used
> status="faulty" to mark an intervention that the editor did not agree with.
In our case, there's nothing wrong with the deletion as such. It's just that the
material is too interesting and important to be hidden away in the apparatus.
> e.g. XSLT. Accordingly I'd rather use a xml:id attribute for the relevant
> <del> element and display it by XSLT whereas the other <del> elements
> are left out.
As Torsten also mentions, I think the decision to display the text should be
recorded in the XML. For the moment, my conclusion is that rend="display"
or something like that is really the best solution.