Cite the 'untruncated' form and gloss that.

Asterios Chardalias where would you cite it? In a note?

Since it is spontaneous speech and the framework I use is Interactional Linguistics and Conversation Analysis, the "turn at talk" needs to be written in the transcript as it is pronounced in the video.

So I was thinking of going with:

25 C; [ͦtsahah::=avvocà.°]

tshahah lawyer.VOC

tshahaha::counselor

26 F; nzomma viè da sé

So.ADV come.IND.PRS PREP.ABL PRO.3SG.IMPRS

Well, it goes without saying

And then add a note with the whole words.

Wondering whether this is a good idea...

I've been taught to work with the Leipzig Glossing Rules. That means glosses align vertically for ease of identification. You can have as many lines per item as needed for your analysis. I have worked with spoken Greek dialectal data (which I wanted to present to English-speaking audiences). I had 6 lines per item: 1. Dialect transcribed in Greek 2. Standard Greek equvialent transcribed in Greek 3. Standard Greek equivalent transliterated with roman characters 4. Word for word English translation 5. Morphological analysis 6. English Translation

I'm not sure how that would feed back into a Conversational Analysis piece that also has its own coding conventions. I'd imagine going with a 6-line gloss per turn will be way too much.

But then again it is more of a matter of what you want to actually claim. Is there morphological variation away from the standard language in your data? Or is it that the dialect only differs phonologically? Remember a morphological analysis deals in morphemes. So the dialect may present different morphs (allomorphs), but not different morphemes. If that's the case, glossing from the standard equivalent is fine. If not then you'could go with a phonemic gloss first and then morphologically gloss that (essentially treating the dialect as a 'different' language).