"A given reading is associated with the set of witnesses attesting it by listing the witnesses in the wit attribute on the <rdg>, <lem>, or <rdgGrp> element."
Comment: Perhaps a clue could be given as to how @wit can be used on <rdgGrp>? Can a pseudo-witness occur here?
Good point. The problem seems to be that, for convenience, rdg and rdgGrp are members of the same attribute class, from which they inherit a bunch of attributes including @wit. I am not sure what you mean by a pseudo witness -- presumably you are thinking of the "[unattested]" hypothetical lemma -- but in any case, it seems odd to supply a default witness for a group of readings, given that each member must be from at least one different witness. Not clear whether to fix this anomaly by requiring that @wit be explicitly defined for each element that needs it and removed from the class, or by taking rdgGrp out of the class.
Agreed to remove @wit from att.textCritical; define new class att.witnessed (vel sim) with members rdg and lem
Implemented at rev 10243