I noticed working with interpellation questions that pre199495, questions submitted in writing that are included in protocols are often annotated with seg and u elements, though they are not representing speech events. It's probably part of a larger issue, not only related to these debates:
[ ] Identify parts of the protocols that are not representative of speech events (IP questions, tables of contents, budget tables, etc...)
[ ] decide how to annotate: note with type attr, other types of elements, sth else?
[ ] re-label non-speech event text from seg/u to whatever other annotation scheme we decide on
I noticed working with interpellation questions that pre199495, questions submitted in writing that are included in protocols are often annotated with seg and u elements, though they are not representing speech events. It's probably part of a larger issue, not only related to these debates: