The recent discussion on the tdwg-content email list has clarified my understanding about the community consensus on the nature of an Occurrence. In my previous view that each Occurrence had a single token that provided evidence for the Occurrence, I felt that it was necessary to have a term in the Darwin Core type vocabulary for each type of token that could serve as evidence for an Occurrence and which could serve as a controlled value for dwc:basisOfRecord. On the basis of this understanding, I proposed adding DigitalStillImage as a dwc:type.
However, given the apparent consensus that the documenting token is a separate entity from the Occurrence and that there may be multiple tokens associated with a single Occurrence, I no longer see any need for adding DigitalStillImage to serve as a controlled value for basisOfRecord, since still images that are tokens will be a separate resource from Occurrence and can simply be rdf:type'd as dcterms:StillImage . (In fact, I no longer believe that basisOfRecord itself serves any useful purpose - I'm still waiting for someone to show me a case where it is necessary for it to exist as a term.)
I'm not actually withdrawing my request, because if there is actually a purpose for basisOfRecord, then DigitalStillImage should be a value for it. People who want to use basisOfRecord will need it. But given my current understanding of an Occurrence, I would vote against my own proposal (if I were voting, which I'm not). Anyway, unless there is a desire for further discussion of basisOfRecord, I feel like the discussion on tdwg-content of adding DigitalStillImage to the DwC type vocabularly has progressed far enough for the TAG to move on to the voting stage.
Steve