'Group' , 'Organization', 'Object Aggregate' & 'Aggregate' in general within BFO

[petosa...@gmail.com]

The purpose of this post is:

1. to obtain clarification on definitions for the domain-specific terms 'group' and 'organization';
2. to determine how best to 'extend' these terms from BFO;
3. to obtain clarification on BFO's 'object aggregate'; and
4. to obtain clarification on applying 'aggregate' across all BFO categories.
   

The Oxford English Dictionary defines 'group' and 'organization' as follows:

group: a number of people that work together or share certain beliefs

 

organization: a group of people who form a business, club, etc. together in order to achieve a particular aim

 There are multiple OED definitions for each of these terms, but I chose to select these two. BFO neither defines (nor should it define) 'group' nor 'organization'.

The "Basic Formal Ontology 2.0 Draft Specification and User's Guide" [BFO Guide] offers an elucidation on 'object aggregate'.

b is an object aggregate means: b is a material entity consisting, at any time t at which b exists, exactly of a plurality of objects as member_parts at t.

The member_part_of relation is given this definition (Note: "<=" means "less than or equal to"):

b member_part_of c at t =Def. b is an object & there is at t a mutually exhaustive and pairwise disjoint partition of c into objects x(1), ..., x(n) (for some n > 1) with b = x(i) for some 1 <= i <= n.

Later and by example, the BFO Guide offers additional guidance on 'object aggregate'.

Different sorts of examples of object aggregates satisfy further conditions. For example, example (object aggregate) [an organization is an aggregate whose member parts have roles of specific types (for example in a jazz band, a chess club, a football team); a swarm of bees is an aggregate of members who are linked together through operations of their motor and visual systems]; and so on.

Object aggregates may be defined through physical attachment (e.g., the aggregate of atoms in a lump of granite); defined through physical containment (e.g., the aggregate of molecules of carbon dioxide in a sealed container; the aggregate of blood cells in your body). Object aggregates may be defined by fiat (e.g., the aggregate of members of an organization]; or defined via attributive delimitations (e.g., the patients in this hospital; the residents of Palo Alto; your collection of Meissen ceramic plates.

[76] provides a formal treatment of aggregates (there called 'collections') that is broadly consistent with the above except that it assumes that membership in a collection is fixed over time. However, as is true for many material entities, note (object aggregate) [object aggregates may gain and lose parts while remaining numerically identical (one and the same individual) over time. This holds both for aggregates whose membership is determined naturally (the aggregate of cells in your body) and aggregates determined by fiat (a baseball team, a congressional committee).]

[76] Thomas Bittner, Maureen Donnelly and Barry Smith, "Individuals, Universals, Collections: On the Foundational Relations of Ontology", in: A.C. Varzi and L. Vieu (eds.), Proceedings of the Third Conference on Formal Ontology in Information Systems (FOIS 2004), Amsterdam: IOS Press, 37-48.

Lastly, the BFO Guide states this about aggregates across BFO:

Object aggregate is to be treated as a close analogue of set of objects, in the mathematician's sense; thus an object aggregate has no parts other than the objects that are its members. In this document we concentrate on the use of 'aggregate' as it appears in the term 'object aggregate'. However, 'aggregate' should be understood as being generalizable to all continuant BFO categories. Thus for each BFO category X, the user of BFO as at his disposal also the category aggregate of X. [51].

[51] Kerry Trentelman, Alan Ruttenberg and Barry Smith, "An Axiomatisation of Basic Formal Ontology with Projection Functions", Advances in Ontologies, Proceedings of the Sixth Australian Ontology Workshop, Adelaide, 7 December 2010, Kerry Taylor, Thomas Meyer and Mehmet Orgun (eds.), 2010, Sydney: ACS, 71-80.

Questions:

1. Based upon the preceding OED 'group' and 'organization' definitions chosen for this post, would the following proposed definitions offer better precision? These definitions intend to define 'group' and 'organization' within the context of people and not generically of objects.
   
• [proposed] group: An ad hoc assemblage of people for which the assemblage has no prescribed purpose.
• [proposed] organization:A prescribed and contractually obligated assemblage of people for which the assemblage has one or more prescribed purposes for achieving one or more prescribed goals.
  
  
2. Since the proposed 'group' and 'organization' definitions suggest aggregations of people, would these two mutually exclusive universals extend from 'object aggregate'?
   
   
3. The BFO Guide is unclear on the nature of the 'member parts' of an 'object aggregate'. While the 'member parts' are required to share roles of specific types, it does not specify if the objects serving the same-typed role(s) must themselves be of the same type. Take, for example, a (traditional) String Quartet as a specialization of 'object aggregate'. The member parts must serve the following roles: 'Violionist I', 'Violinist II', 'Violist', and 'Cellist', where each of these roles specialize the 'Stringed Instrumentalist' role. Traditionally, people serve in these roles. However, given advances in music technology, there is no reason why the 'objects' serving these roles cannot be, for instance, a mixture of 'people' and 'computer music software systems'.
   
   Does BFO restrict 'object' types within the context of an 'object aggregate', or can an 'object aggregate' have member parts that are a mixture of 'object' types serving the same role(s)?
   
   
4. The BFO Guide states that already "it is clear that BFO or its conformant domain-ontologies will in due course need to recognize also other sub-universals of 'material entity', in addition to 'object', 'object aggregate' and 'fiat object part' - for instance, 'aggregate of fiat object parts' [29, 82]. Thus the treatment of 'material entity' in BFO 2.0 should not be associated with any closure axiom pertaining to the three distinguished categories, in other words it should not be associated with any claim to exhaustivity."
   
   Recall this excerpt:
   
   

   "Object aggregate is to be treated as a close analogue of 'set of objects', in the mathematician's sense; thus an object aggregate has no parts other than the objects that are its members. In this document we concentrate on the use of 'aggregate' as it appears in the term 'object aggregate'. However, 'aggregate' should be understood as being generalizable to all continuant BFO categories. Thus for each BFO category X, the user of BFO has at his disposal also the category aggregate of X. [51]."
   
   While 'object aggregate' defines class membership for 'object', the BFO Guide constrains membership to 'objects' with the same-typed 'role(s)'. The 'object aggregate' set-theoretic construct is ontologically constrained by the member part role type(s), as it is ontologically constrained through its specialization of 'material entity'. However, this raises an interesting point. If 'object', 'object aggregate' and 'fiat object part' do not exhaustively close 'material entity' and if a BFO user has "category aggregate of X" at his disposal, then would it not follow that no part of BFO exhaustively closes? If we consider 'aggregate' merely as a set-theoretic construct, then, for example, cannot there be an aggregate of 'Entity' at the top-most level? I suppose one cannot impose an aggregate of, for instance, 'Material Entity' and 'Process' (category X does not resolve to the same immediate parent) but can define an aggregate of 'Material Entity' and 'Immaterial Entity' (category X resolves to the 'Independent Continuant' immediate parent).
   

[Barry Smith]

Antony,

My thinking on this matter goes as follows. We have a very poor understanding of how to represent goals, aims, objectives. 

But we can define group, organization, etc., along the following lines:

group = an object aggregate whose member-parts are persons who act together

organization = a group in which some member-parts have specifically defined roles

in many cases these roles will be reciprocal, thus an organization may have a CEO and a CFO and their respective roles are defined in terms of each other (compare: the husband and wife roles)

BS

--
You received this message because you are subscribed to the Google Groups "BFO Discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to bfo-discuss...@googlegroups.com.
To post to this group, send email to bfo-d...@googlegroups.com.
Visit this group at http://groups.google.com/group/bfo-discuss.
For more options, visit https://groups.google.com/d/optout.

[petosa...@gmail.com]

 > group = an object aggregate whose member-parts are persons who act together

This implies a trans-ontological, specific dependence relation between 'person member parts' and 'process', whereby the person member parts participate in some process at some point in time (either temporally contiguous or temporally discontiguous). Acting together is to be concurrently and/or co-dependently engaged in some act (i.e., 'process'). Is this relation mandatory for defining a 'group'? Cannot a group simply be an ad-hoc collection of spatially co-located people? For example, a group instance could be an 'object-aggregate' particular of people who are demarcated by some fiat boundary at the intersection of 57th Street and 3rd Avenue in New York, NY 10021 on Thu, 20 Aug 2015 @ 1:00 PM (EDT). This group is defined by 'attributive deliminations'. As the group awaits a traffic light change, it can gain or lose parts over time as people step into or out of the group's fiat boundary.

> organization = a group in which some member-parts have specifically defined roles

If one considers 'organization' to specialize the more general 'group', then the same points apply here. Your proposed definitions present this generalization-specialization relationship.

By comparison, here again are the proposed definitions of 'group' and 'organization'.

[BS] group: an object aggregate whose member-parts are persons who act together
[AP] group: an ad hoc assemblage of people for which the assemblage has no prescribed purpose.

[BS] organization: a group in which some member-parts have specifically defined roles
[AP] organization: a prescribed and contractually obligated assemblage of people for which the assemblage has one or more prescribed purposes for achieving one or more prescribed goals.

In your proposed definitions, the differential is having specifically defined roles. With my proposed definitions, the differentia are (a) a qualification on the assemblage of people and (b) a qualification on purpose. This takes me back to a question I asked a long time ago on the Aristotelian definition structure (i.e., A is a B that C, where 'C' presents the differentia). With tangible objects, which physically occupy three-dimensional space, the differentia can be scientifically classified, for instance, down to the molecular level, and verdicality can hold at other 'ontologically zoomed', yet orthogonal, views of reality. This proves to be tricky with intangible objects. The 'group' and 'organization' terms do not present physical entities. What specific criteria must be applied for their differentia?

[Barry Smith]

On Thu, Aug 20, 2015 at 1:26 PM, <petosa...@gmail.com> wrote:

 > group = an object aggregate whose member-parts are persons who act together

This implies a trans-ontological, specific dependence relation between 'person member parts' and 'process', whereby the person member parts participate in some process at some point in time (either temporally contiguous or temporally discontiguous). Acting together is to be concurrently and/or co-dependently engaged in some act (i.e., 'process'). Is this relation mandatory for defining a 'group'? Cannot a group simply be an ad-hoc collection of spatially co-located people? For example, a group instance could be an 'object-aggregate' particular of people who are demarcated by some fiat boundary at the intersection of 57th Street and 3rd Avenue in New York, NY 10021 on Thu, 20 Aug 2015 @ 1:00 PM (EDT). This group is defined by 'attributive deliminations'. As the group awaits a traffic light change, it can gain or lose parts over time as people step into or out of the group's fiat boundary.

​But would not such a case be simply an object aggregate of human beings?

> organization = a group in which some member-parts have specifically defined roles
If one considers 'organization' to specialize the more general 'group', then the same points apply here. Your proposed definitions present this generalization-specialization relationship.

By comparison, here again are the proposed definitions of 'group' and 'organization'.

[BS] group: an object aggregate whose member-parts are persons who act together
[AP] group: an ad hoc assemblage of people for which the assemblage has no prescribed purpose.

​I think that an organization is a special kind of group, and a group is a special kind of object aggregate. I fear that your account of 'group' will have difficulties, e.g. when a group assembles for no prescribed purpose but then decides in an ad hoc way to become an organization​

 

[BS] organization: a group in which some member-parts have specifically defined roles
[AP] organization: a prescribed and contractually obligated assemblage of people for which the assemblage has one or more prescribed purposes for achieving one or more prescribed goals.

​do you have a definition of 'purpose', 'goal'? 

and are not these terms 'contactually obligated', etc. logically more difficult to understand than the term 'organization'?

 

In your proposed definitions, the differential is having specifically defined roles. With my proposed definitions, the differentia are (a) a qualification on the assemblage of people and (b) a qualification on purpose. This takes me back to a question I asked a long time ago on the Aristotelian definition structure (i.e., A is a B that C, where 'C' presents the differentia). With tangible objects, which physically occupy three-dimensional space, the differentia can be scientifically classified, for instance, down to the molecular level, and verdicality can hold at other 'ontologically zoomed', yet orthogonal, views of reality. This proves to be tricky with intangible objects. The 'group' and 'organization' terms do not present physical entities. What specific criteria must be applied for their differentia?

​I would not like to specify this in advance. Certainly some organizations were formed to meet a prescribed purpose, but organizations may change their purposes over time, for instance​

 

​BS​

 

[Joe Behling]

Barry,

What are your thoughts on the relation between organization as you have defined here (organization = a group in which some member-parts have specifically defined roles) and a corresponding legal entity?

For example, if "Apple Computer Organization" as an instance of "organization" then I assume there will be a distinct corresponding "Apple Computer Legal Entity" that is responsible for paying taxes, paying employees, etc... 

Where should "Apple Computer Legal Entity" be classified (i have seen other discussion threads leaning toward generically dependent continuant)?

What is the relation between the "Apple Computer Legal Entity" and "Apple Computer Organization"?

-joe

[Alan Ruttenberg]

Here is one analysis:

Even if the phrase seems to designate an entity "Apple Computer Legal Entity" it doesn't mean that there is a corresponding entity in the world - it may be "a manner of speaking". I will argue that there is no distinct legal entity. Rather that in any situation in which one would invoke the concept (word deliberately chosen) we are referring instead to the conditions under which some organization member or members role(s) would become (perhaps in part) realizable.

For example when the "legal entity" is billed for rent on a building, the rent is paid as realization of a fiduciary role of someone in accounting, a member of the organization.

The only argument I can see for there being a distinct legal entity would be if it is the case that

1) The organization O can lose all members but obligations are legally taken up by an outside organization G, such as a government entity and 

2) G is empowered (has a role) to reconstitute A by appointing new members, who are now considered to be members of A

The argument in this case is that there is an entity that can persist independently of there being any member. Still, even in that case there is the m matter of the gap - how do the roles and responsibilities survive the gap if there is no one to bear them.

OTOH, if, legally, it would be considered that rather than reconstituting A, A's obligations (roles) are ceded to a new entity G or A', then I would argue against the existence of the legal entity, as it would seem to coincident (another way of speaking about) the organization.

My gut says that the latter situation obtains - that the roles are 'transferred' to G or A'.  Note that this isn't an uncomplicated issue. Even within an organization we need to understand whether 

a) a fiduciary role, such as one held by someone in accounting is some sort of generic dependent. This is something that we don't yet represent in BFO (roles are specific dependents). 

b) whether the role is borne by the (singular) organization, but realized by one or another part of the organization. This can currently be represented in BFO. An argument for this might be that in an emergency (e.g. the CFO is incapacitated) the role can be realized by some other member of the organization. 

c) roles are borne by individuals but when, for instance, a new CFO is appointed there is a new role instance of the same type as the previous CFO role. The creation of such a role would be in the realization of some other role (hiring committee role). This is something can arguably be represented in current BFO. If an argument it would be over the nature of the reference to the CFO role universal. There are currently very few relations that relate particulars to universals (instance_of would be one).     

My thinking tends towards (b) or (c), in part because there is a better story explaining the how of change in the organization. (a) simply says that it can happen but says nothing about why or how.

If there is someone who is versed in the law perhaps they could comment on the legal issues.

Alan   

--

[Barry Smith]

I do not see the need to define two entities, Apple the organization and Apple the legal entity.

Apple the organization has responsibilities -- the proper treatment of which is still being worked on -- in something like the way in which you have responsibilities.

BS

On Thu, Oct 29, 2015 at 9:31 PM, Joe Behling <behli...@gmail.com> wrote:

Barry,

What are your thoughts on the relation between organization as you have defined here (organization = a group in which some member-parts have specifically defined roles) and a corresponding legal entity?

For example, if "Apple Computer Organization" as an instance of "organization" then I assume there will be a distinct corresponding "Apple Computer Legal Entity" that is responsible for paying taxes, paying employees, etc... 

Where should "Apple Computer Legal Entity" be classified (i have seen other discussion threads leaning toward generically dependent continuant)?

What is the relation between the "Apple Computer Legal Entity" and "Apple Computer Organization"?

-joe

--

[Asiyah Yu Lin]

RE: Alan's comment:

" Even if the phrase seems to designate an entity "Apple Computer Legal Entity" it doesn't mean that there is a corresponding entity in the world - it may be "a manner of speaking".  "

What if the "Manner of Speaking" is the thing that people wants to represent.

Does this mean that we should not consider BFO in this case?

I think, more importantly, how can we use BFO's concept and relation to code the 'Manner of Speaking' kind of things out there.

I encountered some similar situations as well.

Best,

Asiyah

[Barry Smith]

For those cases where we really are dealing with mere façons de parler BFO uses the strategy of defined classes; this is discussed briefly in the BFO book and we will publish a more detailed documentation in due course. 

Thus there is no entity (for instance no quality) which things have in virtue of having bathed in the Ganges; but people might want to use language in a way which seems to refer to such a quality. For this we can create a defined class

x is-at-t someone who bathed in the Ganges =def. x is a person at t and there is some process p and x participated in p at some time earlier than t and p is a bathed in the Ganges process

(or something like that)
BS

---------- Forwarded message ----------
From: Asiyah Yu Lin <lini...@gmail.com>
Date: Fri, Oct 30, 2015 at 9:01 AM
Subject: Re: [bfo-discuss] 'Group' , 'Organization', 'Object Aggregate' & 'Aggregate' in general within BFO
To: "bfo-d...@googlegroups.com" <bfo-d...@googlegroups.com>

RE: Alan's comment:
" Even if the phrase seems to designate an entity "Apple Computer Legal Entity" it doesn't mean that there is a corresponding entity in the world - it may be "a manner of speaking".  "

What if the "Manner of Speaking" is the thing that people wants to represent.

Does this mean that we should not consider BFO in this case?

I think, more importantly, how can we use BFO's concept and relation to code the 'Manner of Speaking' kind of things out there.

I encountered some similar situations as well.

Best,

Asiyah

On Fri, Oct 30, 2015 at 8:40 AM, Barry Smith <phis...@buffalo.edu> wrote:

[Asiyah Yu Lin]

Thanks, Barry.

Just want to point out that, in the reality, we then create shortcut relations for this kinds of long way description..., since it is difficult for someone who doesn't know much about BFO to understand.

[Alan Ruttenberg]

Asiyah, people _want_ to represent all sorts of things. An example would be fictional people who are people, but fictional. BFO takes a stance on such issues: not all sentence of the same form represent the same sorts of entities. A translation of "Hamlet died" will necessarily be very different than a sentence (in a scientific paper) that one of the mice died. In the former case there is no instance of organism, no instance of dying. In the latter there is.

Alan

[Asiyah Yu Lin]

Thanks, Alan.

I understand your point.

I am trying to figure out a way to use BFO to solve some issue at my work, and at the mean time not introduce too much barriers for adopting BFO.

Maybe we shall communicate off line if I am allowed to present you those "real-world" problems at my work.

Best,

Asiyah

[Joe Behling]

Barry, your advised approach makes sense.

So then, i would be interested in thoughts on how to model a "Non-Profit Organization"

The legal definition of a non-profit organization entails:

1. organizational structure: corporation, trust, or unincorporated association

2. activities: set of do's and don'ts with respect to kinds of activities, i.e. BFO Processes

3. stated purpose: charitable, educational, religious, scientific, literary, fostering national or international sports competition, preventing cruelty to children or animals, and testing for public safety.​

https://www.irs.gov/pub/irs-pdf/p4220.pdf​

Point 3 "stated purpose" seems to be the most problematic.  Given the assertion from Barry above: "responsibilities -- the proper treatment of which is still being worked on", perhaps there is not a clear path for modeling many real-world organization types in BFO.