It's true that the Guidance for Transcription Using Nemeth Code within UEB Contexts is not overly specific in regards to tables.
I certainly cannot say that there is a clear dividing line between transcribing the whole body of a table in Nemeth as opposed to using one or more pairs of Nemeth switch indicators for an individual entry or entries in a table.
I can attempt to outline a possible decision-making process for table transcription and Nemeth Code use, and I look forward to your thoughts on that. Here's that outline:
- Does every part of the table need to be in Nemeth Code? If yes, then do it.
- Does every part of the table's rows need to be in Nemeth Code? If yes, then insert switch indicators after the column separation line and after the last row.
- Does every entry (not including row headings) need to be in Nemeth Code? If yes, then do as above and treat the row headings as technical material.
- For any of the above, if the parts under consideration do not all need to be in Nemeth Code, but they are all either: a) material that needs to be in Nemeth Code or b) letter(s) and/or number(s) which could be read easily in either code, then use Nemeth Code as if all the parts under consideration did need to be in Nemeth Code.
- If one or more table entry is a phrase or something else that is out of place within switch indicators, then experiment with using one or more pairs of Nemeth switch indicators for an individual entry or entries in a table; this may be most reader-friendly in an alternate table format.
Does that help?
Your questions are excellent. I look forward to your response.