the TEI Council is seeking advice for the following issue:
We are lacking a mechanism to describe physically dispersed text-bearing objects (which formerly were parts of a single physical object) as one logical object.
Currently, the guidelines indicate the usage of <altIdentifier> for ‘scattered’ manuscripts  which seems not very convincing. A feature request was made for redefining the semantics of <msPart>  to include the case of scattered manuscripts. (At present, <msPart> is only defined to "contain information about an originally distinct manuscript or part of a manuscript, now forming part of a composite manuscript“ ).
Another proposed solution is to add a new dedicated element <msFrag> (or whatever name …) which would behave pretty much like <msPart> (in terms of content and context).
So, the choice is between 1) broadening the definition of <msPart> so that it can cover not only parts of a manuscript which were originally independent and later gathered together, but also parts of a manuscript which was originally whole but is now in fragments; and 2) creating a new element, <msFrag>, for the latter case.
Any comments appreciated!
Many thanks and
all the best