Asserted subclass relation between Group of Persons and Group of Agents may be incorrect

[gregfowlerphd]

Group of Persons is currently treated as a subclass of Group of Agents. However, it would seem that not all Persons--i.e., members of the species Homo sapiens--are Agents. Consider newborns or brain-dead Persons. Hence, the subclass assertion here seems incorrect.

[swartik]

@gregfowlerphd: Personally, I've never had a problem with the subclass assertion. When I've needed to use class Group of Persons, I've always been interested only in Persons that are capable of being Agents. I think that was the intent of the class.

However, by that logic Ethnic Group and Populace shouldn't be subclasses of Group of Persons. Individuals of these classes are likely to include members that lack agency, such as your examples. Of course, lack of agency is temporal. Most newborn persons will have agency in the future. Most brain-dead persons had agency in the past.

In brief, you can make arguments both for and against. Ultimately it comes down to this: what do you want class https://www.commoncoreontologies.org/ont00000914 to express?

[gregfowlerphd]

@swartik: I basically agree with your points. I'd simply say that if we want Group of Persons to express Group of Persons who are Agents, that ought to be reflected in the definition, which it currently is not.

[BrendaBraitling]

Ok this discussion made me look some stuff up. I am still new to ontological thinking, but very experienced in standardization and assuring systems work well and as they are intended to be used.. It is possible for us to choose how to use things, but it would be helpful to be unambiguous in order to provide for standardized use of this tool by others...

I happened upon this PROV to BFO/CCO mapping project , noting figure 5 of this document which maps PROV to BFO/CCO - Person, Agent, Group of Agents and Organization. I remain confused about all the versions of BFO and CCO so I do not know what has changed over time and which versions are used here.

This author sees CCO Person as a BFO Object which is downstream from organism, while it sees CCO Agent as a Material Entity - not an object. I also like the author's helpful distinction when mapping the PROV:SoftwareAgent .

An Agent and a Person seem to be very different as this author interprets them. Being different seems to solve some messy problems about how real persons exist while agents take actions... and one person can be many agents at various times or have no agency at all - as mentioned above.

As an information system engineer and solution architect, I deal with Human - Person - Agent entities regularly and find them quite distinct. I realize that ontological thinking is specialized - so here are some use cases for human, person, agent that I deal with...

• I translate signals and meaning across the human-computer interface for every system solution. Human usability and ergonomic factors...
• I deal with Personally Identifiable Information (PII) - so to me Person is an instance of a Human - Organism - Object.
• I handle "need-to-know" data security - so actions impacting any part of the information system - software, hardware, organizational role or specific account - need to be aligned to data that is carefully restricted.

So I wonder:

• How can a Person (object) or Group of Persons (object aggregate), as instances of Human - organism Objects be in two branches of the hierarchy? an object as a human person and a material entity as an agent in another?
• The article seems to say that material entity makes the agent more flexible than object when used in applications. How are they different? Is this important?
• Objects can be aggregated so we can have a group of persons, group of humans, group of apples... But can a material entity be an aggregated object if it is not an object?
• If an organization can be one or more agents, then why do we need groups of agents at all? I see organizations as formal or informal - according to the myriad textbooks for organizational behavior... This seems to handle all interrelated organizational structures (even the annoying matrix organization - where Jack serves two teams doing two jobs with two bosses.) Is this the difference between having agent as a material entity vs an object?

Appreciate any words of wisdom as I am still trying to make sense of ontological thinking,,,

[mark-jensen]

@gregfowlerphd the axiomatization and def "A group of agents" require it , adding "persons that are agents" is redundant.

What it means for something to have the "capacity to realize [an agent capability] in a planned act" is somewhat glossy still. BUT, we recently did make iterative improvements #186 #520