Group of Agents and subclasses--three recommended changes + one extra observation

[michaelrabenberg]

Here's the definition:

Group of Agents. An Object Aggregate that has only Agents as parts.

Recommendation 1: Change "as parts" to "as member parts."

Rationale: There's no reason to think the only continuant parts of a Group of Agents are Agents. Furthermore, the actual subclassing axioms on Group of Agents use member-part, not continuant-part.

Here are two subclasses of Group of Agents:

Group of Organizations. A Group of Agents that has only Organizations as parts.

Group of Persons. A Group of Agents that has only Persons as parts.

Furthermore, the subclassing axioms for these two classes use has-continuant-part, not has-member-part.

Recommendation 2: Change "as parts" in textual definitions to "as member parts."

Recommendation 3: Revise subclassing axioms for these classes, so they use has-member-part, not has-continuant-part.

Rationale for both should be obvious from above.

One extra observation: Organization is a subclass of Group of Agents, and so also of object aggregate, so the textual definition of Group of Organizations entails that an object aggregate can have object aggregates for member parts. I believe this was a disputed matter some time ago. Tagging @alanruttenberg because I think he was a party to that dispute.