Example 3 - include forms not listed?

[pdurusau]

Now understanding &c. as etc. to indicate an incomplete listing, do you want to encode other forms of #Bryght(e)# that don't appear in the vocabulary? Reasoning that we can find occurrence and auto-generate pointers more easily than Tolkien and make the listing complete. Then of course to distinguish between Tolkien's list of forms versus a more complete one.

[jtauber]

I think the right way to do that would be to set up a separate abstract lexicon, lemmatise Sisam's texts linking to that lexicon, and linking Tolkien's entries to that lexicon.

My lemma lattice (originally developed for NT Greek Lexicons) is highly suitable for this use case too where you want to both be able to reference a particular spelling and the lexeme as a whole.

The citations in Tolkien can help bootstrap the lemmatisation but ultimately would not be the primary assertion of the lemmatisation.

Of course, there is a lot that can be done with both Sisam's texts and Tolkien's glossary even before any of this is done.

[pdurusau]

Separate abstract lexicon works but I assume you still want to account for the refs that Tolkien has to Sisam's texts. Yes? Which within <orth> would be <oRef>, using the @target attribute to point to an occurrence mentioned by Tolkien. Assuming you want to accept each reference to a text number + line number or just text number as a string. (Not encoding the text number separate from the line number.)