For clarification, in cases like #1, we ARE allowing one-to-many relationships between "composite" Individuals and Identifications as long as those Identifications represent differences of opinion about the common taxon, or refinements to a lower taxonomic level (e.g. I'm not capable or don't have time to determine the lowest taxonomic level common to all of the biological individuals in the jar, but later I am able to find that out). What we AREN'T allowing is for subsets of the composite "Individual" that belong to different lower level taxa to be identified to those taxa without first separating them into different Individuals. I think I'm stating the principle that Rich laid out correctly.
Yes, exactly. But see also my reply to Dusty about the "Erebia youngi or Erebia lafontainei" example. I don't think this breaks the rule, because it's still two competing and mutually exclusive assertions of taxonomic identity -- just that happen to have been made by the same person at the same time. Aloha, Rich