Initial questions about BFO

[Julius Hamilton]

Hello,

I'm very excited to become more active in the formal ontology community.

I expect to have a lot of questions in the near future.

I am working on a learner's project making a sort of enterprise knowledge graph. 

I am a low-level employee at Tesla, but my project is just my own private independent project.

I figured out how to import an ontology into Protege. I imported BFO from the IRI, and then I downloaded the merged Common Core Ontology from GitHub and loaded that in as well.

I am now reading the annotations of classes in the class hierarchy.

Here are some things I am noticing.

Many of the classes now have 2 rdfs:label annotations, one from CommonCoreOntologiesMerged, and one from bfo, with the same value. Is there a way to merge those two attributes into one?

The class 'entity' has an attribute 'skos:definition' from CCO. I guess it seems like CCO uses some pre-existing ontologies, like Dublin Commons and Simple Knowledge Organization System. Maybe the benefit of that is that it helps unify and integrate pre-existing ontologies within the BFO ecosystem?

skos:definition's value includes "(Elucidation)". It seems like this is communicating a "type" of definition. Shouldn't that be a formal attribute or class, though, rather than an informal annotation inside the string value of the skos:definition attribute?

rdfs:isDefinedBy just links to the IRI of the bfo:entity resource, but I guess this attribute is commonly used just to link a resource not to its exact IRI but a domain-specific ontology, perhaps, that controls its definition?

Then there are bfo:editor_note annotations. 

One is:

> BFO 2 Reference: In all areas of empirical inquiry we encounter general terms of two sorts. First are general terms which refer to universals or types:animaltuberculosissurgical procedurediseaseSecond, are general terms used to refer to groups of entities which instantiate a given universal but do not correspond to the extension of any subuniversal of that universal because there is nothing intrinsic to the entities in question by virtue of which they – and only they – are counted as belonging to the given group. Examples are: animal purchased by the Emperortuberculosis diagnosed on a Wednesdaysurgical procedure performed on a patient from Stockholmperson identified as candidate for clinical trial #2056-555person who is signatory of Form 656-PPVpainting by Leonardo da VinciSuch terms, which represent what are called ‘specializations’ in [81

This is really helpful information for understanding the concepts in the ontology, but the formatting seems unfinished. The examples of universals don't have spaces between them, and it looks like the text gets cut off. It refers to a citation that isn't provided.

The next editor note includes a link returning 404 not found, and it has a typo in "possibilites": 

> Entity doesn't have a closure axiom because the subclasses don't necessarily exhaust all possibilites. For example Werner Ceusters 'portions of reality' include 4 sorts, entities (as BFO construes them), universals, configurations, and relations. It is an open question as to whether entities as construed in BFO will at some point also include these other portions of reality. See, for example, 'How to track absolutely everything' at http://www.referent-tracking.com/_RTU/papers/CeustersICbookRevised.pdf

What is really helpful are the "examples of usage" attributes.

For the entity continuant, I feel like the definition is roughly clear enough to get an idea for how to use it. 

> A continuant is an entity that persists, endures, or continues to exist through time while maintaining its identity.

Part of me feels like the definition is allowing itself to use a lot of synonyms in order to convey certain shades of meaning, but ideally for a truly standardized ontology, there should be a more concise and exact definition. I feel like "maintains its identity through time" is a clear and important part of the definition of continuant. It might possibly give someone a semantic test they can do while trying to decide what type to assign something they want to model in an ontology.

One issue that occurs to me regarding the phrase "A continuant is an entity that persists, endures, or continues to exist through time" is that the human subject necessarily exists in time. In a way, it seems to me, there is nothing that is not temporal. I don't know if this gets at some deep philosophical debates in metaphysics and the foundations of ontology, but in my own attempts to develop a theory of ontology from first principles, I wonder if I have taken an extremely "idealist" perspective where it is the human perceiver and whatever notions they find a priori self-evident, in the Cartesian/Husserlian way, that is the starting place of the ontology. The two categories I have taken as primary are "entities" and "aboutnesses". Entities are specifically defined in a mental way: it is something your mind can direct attention to. We are not making any claims about their properties or constraints beyond that. For example, we are not claiming that an entity is a physical object in the world, or some highly theoretically developed notion of a "concept" (i.e., some theory of semantic networks, prototype theory, etc.). "Aboutnesses" are just predicates, it is the mental act of saying something *about* some entity you are in that moment regarding.

I went off on a tangent there, there is so much I am hoping to discuss with this community. All I wanted to say was, at least from my point of view, because it is the perceiving and self-perceiving human subject that is where we discover and establish all original categories of our ontology framework, "time" is on the topmost level of the ontology. Even if I think that "triangle" is somehow an "atemporal" entity, all we can really say is, it is something I am directing attention to *right now* in my mind, and so it necessarily exists in time. Maybe there is a whole other side of this that I need to think about though, like, why logical truths remain constant even when I am not thinking about them. I think this is what Husserl did and why he rejected psychologism in the philosophy of mathematics? 

Maybe I am angling towards some hardcore stance that temporality is a necessary condition of existence, there is no such thing as non-temporal existence; I don't know. I haven't studied a lot of metaphysics.

Anyway, coming back to this definition:

> A continuant is an entity that persists, endures, or continues to exist through time

I read a little bit of Barry Smith et al.'s book about BFO, and it seems he emphasizes the importance of non-circular definitions. Part of me feels like the sequence of words "persists, endures, or continues to exist through time" is kind of vague and maybe circular; it doesn't immediately give one a sense of, "Oh, *that*." It simply says, "A continuant is something that continues in time." The word "continue" already implies time in the background, so we might simplify it to "A continuant is something that continues." One might say, "Are dead people then not continuants? They used to continue in time, but they are not continuing in time any longer."

I also feel like if "occurrent" is disjoint with "continuant", it should be easy to understand what it is simply by negating the definition of "continuant". For example, perhaps it could be defined as "a thing which does not have fixed identity across time". But if time is already a given, maybe both of these definitions should focus more on the nature of *identity* than that of *temporal character*. I wonder if that would make the notion of an occurrent much clearer. Because I also find the definition of an occurrent hard to understand:

> An occurrent is an entity that unfolds itself in time

This sounds like very metaphorical, non-technical language. A continuant "continues" in time, but an occurrent "unfolds itself" in time. Those sound like they could be the same thing: they have a temporal dimension. They occupy time. "Continue" could imply "is static" and "unfold" could imply "changes, is dynamic", but this doesn't match the way these terms are being used: a continuant can be a stateful entity. Like a person, a continuant can change through time, but it maintains fixed identity. I am starting to wonder if the difference between a continuant and an occurrent is more about the relationship between universals and particulars? We (perhaps wrongly) think of classes/types as atemporal and even unchanging - i.e., the Platonic ideal of a chair or cup or person is taken to be fixed, but why? In Plato's time, perhaps he thought forms simply had an immanent definition; in our time, maybe we can say our brains have learned a fixed schema for a certain object, so they appear superficially to be "static". When a class is instantiated, it seems like the instance gets transported to the realm of time: there, its accidental features can change (i.e., humans can grow old, get their hair cut), but one thing maintaining the instance's "fixed identity" is that it is tethered to its atemporal class "up above" in the timeless plane. I wonder if what makes occurrents occurrents is that they do not even have this class vs. instance character, which is why the nature of the way they possess "identity" is different? 

Sorry if this is amateur rambling, but I really do hope to develop some of these thoughts.

Hopefully I will have some more thoughts soon about a clearer way to try to define what an "occurrent" is, I feel like I will have some ideas in the coming week. 

I would also like to note that "occurrent" does not have the helpful "examples of usage" that "continuant" does.

Thank you,

Julius Hamilton

[Werner Ceusters]

Julius,

I am not sure what rendering of BFO you are using, but the latest version is here: https://github.com/BFO-ontology/BFO-2020.

Rather than to just look at the OWL rendering you must look at the axioms in first order logic. You will see that both instances of occurrent and instances of continuant exist-at a time. The difference is that instances of occurrents - while existing at different times - do not change. Definitions and axioms always need to be considered together to avoid confusions and surprises. See f.i. https://ebooks.iospress.nl/doi/10.3233/FAIA250486

The link you posted (http://www.referent-tracking.com/_RTU/papers/CeustersICbookRevised.pdf) doesn't work. But anyhow, the most recent work in matter of Referent Tracking is here: https://osf.io/preprints/osf/q8hts_v1

Aboutness in the context of the BFO is discussed here: https://www.referent-tracking.com/RTU/files/ICBO2015Proceedings-Smith-Ceusters/1.0/ICBO2015Proceedings-Smith-Ceusters.pdf

W

[Barry Smith]

To Julius, regarding your

"I don't know if this gets at some deep philosophical debates in metaphysics and the foundations of ontology, but in my own attempts to develop a theory of ontology from first principles, I wonder if I have taken an extremely "idealist" perspective where it is the human perceiver and whatever notions they find a priori self-evident, in the Cartesian/Husserlian way, that is the starting place of the ontology."

This "idealism" is a common philosophical position. I agree that Husserl's account of the a priori is a good starting point for ontology -- indeed Husserl was the first to use the term 'formal ontology' in his Logical Investigations, written when Husserl was still a realist. However it is important not to draw any conclusions from the thesis that "it is the human perceiver and whatever notions they find a priori self-evident that is the starting place of the ontology." To see why you should read this.

Barry

--
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 view this discussion visit https://groups.google.com/d/msgid/bfo-discuss/274896ab-0200-4da7-aef6-fa885fe98383n%40googlegroups.com.