Hi @abentley. I interpreted @cboelling to mean that we should treat these as three different types of digital object, each with their own representation and data standards (metadata, etc.), although with some well-defined relationships. This is in contrast to the idea (implied in your previous message) that these all belong to the same class of object. It may be that it was your earlier message that was misleading. Perhaps you simply meant that we needed to support hierarchies of sets of specimens all as nested collections, including a “super-collection” that represents all the materials held by an institution, a “collection” that represents a management unit that typically has a recognisable CollectionCode, and an “infra-collection” that represents a potentially unnamed set of specimens of interest in some context.