Proposed label change 'entity' -> 'particular' in BFO-2020 OWL
73 views
Skip to first unread message
Alan Ruttenberg
Sep 19, 2022, 11:34:36 PM9/19/22
to bfo-d...@googlegroups.com
Entity in ISO 21838 is defined rather broadly.
anything perceivable or conceivable
That includes at least both particulars and universals, and probably also relations and other kinds of things that haven't been explored yet. In the OWL, entity is listed as a subclass of owl:Thing, and so it doesn't include universals/classes. Classes are not in the domain of discourse of OWL2 using the direct semantics, which is what we use.
Entity is also an odd class as probably the only thing that can be said of it is that definition. I can't think of any consequences of it being a superclass of anything. What is true of every entity? It's not like we could represent anything that isn't an entity, since if we represent it we've conceived of it.
So, really, the class that is called 'entity' in OWL is, at the moment, the class whose only members are particulars. I'd like to have the label reflect that. We could keep 'entity' as an alternative label with a note. 'entity' will still remain in the FOL version, for the moment, but it might be worth questioning whether it should even remain there.
I'm asking this about BFO-2020 because that's the version I maintain but I think it should be changed in earlier versions of BFO as well.
If I don't hear objections in the next few weeks I will make the change in BFO-2020
Alan
Bill Duncan
Sep 20, 2022, 4:20:58 PM9/20/22
to bfo-d...@googlegroups.com
DOLCE uses 'particulars' as its top-level class. So, this may be conducive for harmonizing BFO and DOLCE, or it may cause confusion. I'm not sure.
I've always thought the 'entity' class was in the OWL version of BFO was for engineering purposes. It allows you to have a sibling 'obsolete class' class. You can't do this if 'Thing' and not 'entity' is the root of the hierarchy. But, I may be wrong about this. Using the label 'particular' may (I think) provide a better distinction from owl:Thing.
Bill
--
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 on the web visit https://groups.google.com/d/msgid/bfo-discuss/CAFKQJ8myo-nsjDWG%3DtbFTNYzNULptysYEkLiatTJ7-yhQdMoRg%40mail.gmail.com.
Stefan Schulz
Sep 22, 2022, 4:43:12 AM9/22/22
to bfo-d...@googlegroups.com
I'm in favor of "particular". My problem with the word "entity" is that nobody could convincingly show me any non-entity up to now.
The word "entity" is also very ambiguous. It has a technical meaning in OWL, it is used in the NLP community as short form for "named entity" and designates either linguistic signs or their referents (which adds more confusion to the term).
Apart from that it is used in law and politics (particularly where there is a disagreement about their legal status of a political division, e.g. the "Republika Srpska Entity").
--
Stefan
To view this discussion on the web visit https://groups.google.com/d/msgid/bfo-discuss/CAM2Eo-cXPru74OPEsybLz%2BvMex40e6x0LzT8VFVWKq4akoD3%3DA%40mail.gmail.com.
Stefan Schulz
Univ.-Prof. Dr. med.
Univ.-Prof. Dr. med.
Institute for Medical lnformatics, Statistics and Documentation
Medical University of Graz, Auenbruggerplatz 2/5
A-8036 Graz, Austria
Tel: +43 316 385 16939 | +43 699 150 96 270
Medical University of Graz, Auenbruggerplatz 2/5
A-8036 Graz, Austria
Tel: +43 316 385 16939 | +43 699 150 96 270
Pierre Grenon
Oct 1, 2022, 7:33:49 AM10/1/22
to bfo-d...@googlegroups.com
Hi Alan,
Been mulling over this for too long.
Full disclosure: if it were a mere terminological preference, I wouldn't like it but I wouldn't care either. It seems to me it is more and that there is both a drift and a risk of confusion.
There are two questions that seem important to me:
a) does the move fit? or what does it entail?
b) what's the impact on domain ontologies? In particular, how is such a change communicated to communities?
I'm just going to address a) -- leaving aside lots, so apologies if there needs to be a bit of reading between the lines.
Entity does mean something very well established in KR and elsewhere, i.e., that which exists. In my understanding, that's been used and made abundantly clear by BFO documentation. The OWL implementation under OWL semantics only recognises particulars as entities because it elected to make them the only individuals in the domain under standard OWL semantics; it counts universals through their proxy classes, but not as things. That's an extensional thing though and also a peculiar way of shoehorning the distinction within OWL semantics.
Historically, the treatment of particular and universal has been a little hand wavy in OWL implementations. My understanding of most of them is that they just weren't making a formal distinction. Usually, considerations around OWL's views on Class/Individuals and the assumption that universals ought to be classes are part of debates. Again, in my understanding, what I have seen are (OWL) implementations in which we have classes, some standing for universals, some not (approximated by the meta-concept of 'defined classes').
So this is potentially much more than a terminological change for even the OWL implementation. It seems to me that the question you raise points at the question of deciding between two alternative implementations that should be kept clean and well rationalised. The label change would run them together is my main objection.
Here's the tension: If we have this class of particulars, then why not have a class of universals etc. But then what of the traditional OWL implementation in which this distinction is of a higher order (meta level if you want)? To date, the main relevant advantage of having Entity in OWL is that it allows the possibility of introducing universals (and hence explicitly define the classes in relation to them), even though it does not introduce them in the reference implementation.
It'd be better to see either of the alternative implementations but not something in between. Either the OWL implementation does it's current "some classes stand for universals but there's no real way of telling which aside from using an annotation property" trick or it just bites the bullet and starts adding universals in the domain. OWL is not particularly elegant when you do this, I've done that in specific contexts, the neat thing is that it allows reference to universals and you can start doing things in OWL, however, arguably there is not much gain to it when the primary interest is the domain ontology.
Suppose you nevertheless do the label change proposed. As you point out, this is because of the extensional equivalence between Entity and Particular in the BFO-2020 OWL-- I agree with you that it extends because the underlying interpretation of the particular universal distinction in published OWL implementations has consistently been the same. Yet as you also articulate, it is also based on the current view. Suppose the view changes and moves towards one in which universals are added to the domain. When you add universals, will you reinstate an Entity class to subsume Universal and Particular? Why do it this way instead of adding -- say, as a fragment the distinction you want to have. E.g., make an additional module in which Particular subsumes the current children of Entity, place it under Entity, place Universal as a sibling and use instantiation, exemplification etc (not sure there is much etc though) and map the BFO-2020 universals to the corresponding classes.
In sum, the above may be a suggestion to have the two implementations, one is the 'as you have it currently' (with the informal interpretation of the universal-particular distinction that's been a convention historically with BFO OWL implementations) and the other as a consistent extension in which the distinction is formalised. I find the label change would conflate them and be more confusing than useful. I'm probably most reluctant to changes motivated by extensional arguments that may retroactively impact the informal theory and be hard to explain or navigate. I also find that keeping the historical line of implementations clean and consistently traceable (with their unactualised potential for introducing the distinction) is also simpler, cleaner and more understandable.
All the best,
Pierre
Been mulling over this for too long.
Full disclosure: if it were a mere terminological preference, I wouldn't like it but I wouldn't care either. It seems to me it is more and that there is both a drift and a risk of confusion.
There are two questions that seem important to me:
a) does the move fit? or what does it entail?
b) what's the impact on domain ontologies? In particular, how is such a change communicated to communities?
I'm just going to address a) -- leaving aside lots, so apologies if there needs to be a bit of reading between the lines.
Entity does mean something very well established in KR and elsewhere, i.e., that which exists. In my understanding, that's been used and made abundantly clear by BFO documentation. The OWL implementation under OWL semantics only recognises particulars as entities because it elected to make them the only individuals in the domain under standard OWL semantics; it counts universals through their proxy classes, but not as things. That's an extensional thing though and also a peculiar way of shoehorning the distinction within OWL semantics.
Historically, the treatment of particular and universal has been a little hand wavy in OWL implementations. My understanding of most of them is that they just weren't making a formal distinction. Usually, considerations around OWL's views on Class/Individuals and the assumption that universals ought to be classes are part of debates. Again, in my understanding, what I have seen are (OWL) implementations in which we have classes, some standing for universals, some not (approximated by the meta-concept of 'defined classes').
So this is potentially much more than a terminological change for even the OWL implementation. It seems to me that the question you raise points at the question of deciding between two alternative implementations that should be kept clean and well rationalised. The label change would run them together is my main objection.
Here's the tension: If we have this class of particulars, then why not have a class of universals etc. But then what of the traditional OWL implementation in which this distinction is of a higher order (meta level if you want)? To date, the main relevant advantage of having Entity in OWL is that it allows the possibility of introducing universals (and hence explicitly define the classes in relation to them), even though it does not introduce them in the reference implementation.
It'd be better to see either of the alternative implementations but not something in between. Either the OWL implementation does it's current "some classes stand for universals but there's no real way of telling which aside from using an annotation property" trick or it just bites the bullet and starts adding universals in the domain. OWL is not particularly elegant when you do this, I've done that in specific contexts, the neat thing is that it allows reference to universals and you can start doing things in OWL, however, arguably there is not much gain to it when the primary interest is the domain ontology.
Suppose you nevertheless do the label change proposed. As you point out, this is because of the extensional equivalence between Entity and Particular in the BFO-2020 OWL-- I agree with you that it extends because the underlying interpretation of the particular universal distinction in published OWL implementations has consistently been the same. Yet as you also articulate, it is also based on the current view. Suppose the view changes and moves towards one in which universals are added to the domain. When you add universals, will you reinstate an Entity class to subsume Universal and Particular? Why do it this way instead of adding -- say, as a fragment the distinction you want to have. E.g., make an additional module in which Particular subsumes the current children of Entity, place it under Entity, place Universal as a sibling and use instantiation, exemplification etc (not sure there is much etc though) and map the BFO-2020 universals to the corresponding classes.
In sum, the above may be a suggestion to have the two implementations, one is the 'as you have it currently' (with the informal interpretation of the universal-particular distinction that's been a convention historically with BFO OWL implementations) and the other as a consistent extension in which the distinction is formalised. I find the label change would conflate them and be more confusing than useful. I'm probably most reluctant to changes motivated by extensional arguments that may retroactively impact the informal theory and be hard to explain or navigate. I also find that keeping the historical line of implementations clean and consistently traceable (with their unactualised potential for introducing the distinction) is also simpler, cleaner and more understandable.
All the best,
Pierre
--
Pierre Grenon
Oct 1, 2022, 7:38:25 AM10/1/22
to bfo-d...@googlegroups.com
True, DOLCE does, even though they don't have precisely the same underlying metaphysics for the distinction. It wouldn't be necessarily straightforward to link them.
Never really thought of the advantage you give, re. obsolete, helps me make sense of it though :) As i always cringed at that class. The rationale could be extended to adding to an OWL implementation other stuff that are convenient to have for KR and yet that the BFO theory rules outside of ontology. It might help making links from IAO too for the thorniest of them.
Cheers,
Pierre
To view this discussion on the web visit https://groups.google.com/d/msgid/bfo-discuss/CAM2Eo-cXPru74OPEsybLz%2BvMex40e6x0LzT8VFVWKq4akoD3%3DA%40mail.gmail.com.
Pierre Grenon
Oct 1, 2022, 7:43:34 AM10/1/22
to bfo-d...@googlegroups.com
Non-entities do not exist so I'm glad you never were convinced by anyone attempting to show you one :)
This being said, has anybody managed to show you a universal? They might challenge you to do so when using 'particular' and they will ask why we use this term.
This being said, has anybody managed to show you a universal? They might challenge you to do so when using 'particular' and they will ask why we use this term.
You're right that there is lots of ambiguity about words like these. Try talking about entities to someone in banking and they'll ask you in which jurisdiction it is regulated. We can understand the different technical senses involved here though.
Pierre
To view this discussion on the web visit https://groups.google.com/d/msgid/bfo-discuss/CAHM_EAteCf8TCqWBjYFYVQ1-HJ%2B9eS3aAczXvzEm_RTbhjG%2BTQ%40mail.gmail.com.
Thomas Beale
Oct 1, 2022, 9:21:07 AM10/1/22
to BFO Discuss
Ignoring the OWL-related technicalities, a practical point of view for us using BFO as a basis / guide for real information system is that BFO is purely a materialist ontology in the sense that even non-material things like (the text of) a play, mathematical equations etc are subordinate to their material bearers. Roles and Capabilities are similar. From a materialist perspective, this is fine, but from a social perspective, things like Roles look like first order entities, and we don't care about their material bearers (i.e. we just assume they are 'obviously' there). So 'GP', 'cardiologist', 'manager', 'company' etc are independent Entities (not dependent ones) in the social / legal / civic world, and thus most information systems. More importantly, the typical hierarchies one sees e.g. under classes like Party do not line up well materialist perspective of where their respective bearers lie (consider the bearers of Party / Person / Company / Group).
Consequently, on a still-speculative information model of 'Entities' we ended up using Physical_entity for BFO:Entity and distinguishing it from Social_entity, which covers things considered to 'exist' in social / legal terms such as Parties and Assets (model here).
Particular therefore seems a good change...
Michael DeBellis
Oct 4, 2022, 12:25:39 PM10/4/22
to bfo-d...@googlegroups.com
a practical point of view for us using BFO as a basis / guide for real information system is that BFO is purely a materialist ontology in the sense that even non-material things like (the text of) a play, mathematical equations etc are subordinate to their material bearers.
If that is the case then I don't understand how BFO can be a good upper model for scientific disciplines. As Chomsky says, the concept of the "material world" stopped making sense with Newton's concept of gravitational fields: Noam Chomsky Lecture - "The machine, the ghost, and the limits of understanding: Newton's Contribution to the Theory of Mind" and even more so when we consider things like grammars (in computer science and linguistics), wave functions in quantum physics, games and strategies in game theory, and many more.
To take a specific example, in game theory a game and a strategy are first class objects that are studied abstractly even though they can represent behavior as diverse as competition and cooperation among multinational corporations or digger wasps competing for nests. The game that most people are familiar with is the Prisoner's Dilemma (PD). It can be studied abstractly (we can prove that the only strong Nash equilibrium for it is that both sides defect) but it can also represent various types of interactions between agents (e.g., see: https://en.wikipedia.org/wiki/Prisoner%27s_dilemma#Real-life_examples) but it isn't "subordinate" to any of these "material bearers". People like Von Neumann and Nash studied the pure mathematics of this and other games without being concerned about the specific ways they are represented in the physical world. And this doesn't apply just to mathematics. In the early days of game theory there were computer competitions for the best solutions to iterated prisoner's dilemmas that demonstrated (via empirical simulations) that the best strategy was "Cooperate on the first iteration and then play whatever your competitor played in the last iteration". Again, that had implications for the various material world processes that can be represented by the PD but the computer simulations that led to this result treated the games as just games with no reference to their "material bearers".
One other set of examples: Turing machines, Finite State Automata, Push Down automata, these are certainly "real" to any computer scientist but they aren't "subordinate to their material bearers". E.g., Turing's proof that the Entscheidungsproblem has no general solution is completely independent of whether we are talking about logic done in the head of a mathematician or as automated reasoning on a computer. Or algorithms such as Merge Sort, they are studied abstractly without knowing or caring if they are implemented in COBOL or Python.
One last example is the "It from Bit" approach to physics. I don't know what the status of Wheeler's theories are now but it seems to me that a scientific ontology can't axiomatically rule out approaches such as this.
I'm probably missing something, but it seems to me that If we say that BFO is a "purely a materialist ontology" then it is not an appropriate ontology for a significant amount of modern science.
Michael
--
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 on the web visit https://groups.google.com/d/msgid/bfo-discuss/1fcf35b4-77b8-440d-adc7-892839a0988dn%40googlegroups.com.
Ludger Jansen
Oct 5, 2022, 8:01:25 AM10/5/22
to bfo-d...@googlegroups.com
The "purely" was somewhat overstated. In many ways, BFO takes
material entities as basic and foundational, though.
For all other things it has to be stated how they relate to material entities. This can be tricky at times. As far as I know, it has yet been solved for fields and waves, but there are some attempts to spell this out.
There is quite some literature now on the question how social
entities relate to material ones. Because of this
interconnectedness, it is no good idea to open a wholly new and
disconected branch in the ontology for social stuff.
@Thomas Beale: The work-around you suggested has a problem with
this, I suppose.
Finally, idealizations like the Turing maschine or the Homo oeconomicus might be studied, but they do not exist. There are studies of the Turing maschine, there are lectures about the Turing maschine, but such a maschine has never been built and cannot be built (not enough matter around for the infinitely long coding tape). Analogously, there are reports about unicorns, pictures of unicorns, maybe even studies about unicorns -- but there are no unicorns. So the studies, lectures, reports, pictures go into the ontology, but the unicorns and the Turing machine don't.
My five pence on how BFO is useful (or should be used) as a top-level for scientific ontologies.
Best
Ludger
To view this discussion on the web visit https://groups.google.com/d/msgid/bfo-discuss/CALGFikfR8je3gT%2B8drJtbtyEH4PSZX0xTLfGL5vVsXXYiUkP-Q%40mail.gmail.com.
-- Prof. Dr. Ludger Jansen Institut für Philosophie Universität Rostock D-18051 Rostock NEW BLOG https://biomimetics.hypotheses.org
Michael DeBellis
Oct 6, 2022, 3:31:17 PM10/6/22
to bfo-d...@googlegroups.com
If you are going to have an ontology that you claim is a good upper model for all science it seems rather perplexing to say things like "we haven't figured out wave-particle duality or the social sciences yet". So it's an upper model for all of science except Physics (the foundational and most mature science) and the social sciences. All that's left is chemistry and biology.
In fact, if you exclude things like game theory you aren't just excluding the social sciences but evolutionary biology as well. Game theory is very important in modeling behavior and adaptations, that's why I mentioned digger wasps in my first post. That's an example from John Maynard Smith's seminal book: Evolution and the Theory of Games. And in that book Smith also discusses various games (like Prisoner's Dilemma and Hawk-Dove) in the abstract before showing how they can apply to specific behaviors like Digger Wasps.
How exactly do you define what is real? The best definition I've seen is in a video from a physicist on YouTube named Sabine Hossenfelder. She defines real (in a scientific sense) to mean any concept required for a scientific theory. So by her definition imaginary numbers are real because you need them to solve problems in electrical engineering.
As I said in my first post, science moved past naive materialism with Newton. If you stick to naive materialism and exclude all the things I mentioned I am somewhat amazed you can claim BFO is a vocabulary for all of science. And not just science. What if my ontology wants to represent individuals like Jay Gatsby (the fictional hero in F. Scott Fitzgerald's The Great Gatsby)? Is that also excluded? Or even Julius Caesar? He had a material body at one point but not any more.
I don't mean to be giving anyone a hard time, I'm legitimately trying to understand BFO and this confuses me.
Michael
To view this discussion on the web visit https://groups.google.com/d/msgid/bfo-discuss/d31b2cb5-7b48-b82b-1262-a4c25840ceaa%40uni-rostock.de.
Ludger Jansen
Oct 7, 2022, 9:24:09 AM10/7/22
to bfo-d...@googlegroups.com
Am 06.10.2022 um 21:31 schrieb Michael DeBellis:
> How exactly do you define what is real? The best definition I've seen
> is in a video from a physicist on YouTube named Sabine Hossenfelder.
> She defines real (in a scientific sense) to mean any concept required
> for a scientific theory. So by her definition imaginary numbers are
> real because you need them to solve problems in electrical engineering.
There are plenty of scientific theories. Most of them are wrong. Do they
> How exactly do you define what is real? The best definition I've seen
> is in a video from a physicist on YouTube named Sabine Hossenfelder.
> She defines real (in a scientific sense) to mean any concept required
> for a scientific theory. So by her definition imaginary numbers are
> real because you need them to solve problems in electrical engineering.
also count as benchmarks of reality?
There is a scientific theory that requires the concept of phlogiston. Is
phlogiston real?
>
> As I said in my first post, science moved past naive materialism with
> Newton. If you stick to naive materialism and exclude all the things I
> mentioned I am somewhat amazed you can claim BFO is a vocabulary for
> all of science. And not just science. What if my ontology wants to
> represent individuals like Jay Gatsby (the fictional hero in F. Scott
> Fitzgerald's The Great Gatsby)? Is that also excluded? Or even Julius
> Caesar? He had a material body at one point but not any more.
Jay Gatsby is an instance of FictionalCharacter to which membership to
the kind HomoSapiens is ascribed (or presupposed).
Or something similar. For a more detailed suggestion cf.
http://ceur-ws.org/Vol-2518/paper-WODHSA4.pdf
Best
Ludger
Thomas Beale
Oct 10, 2022, 7:02:46 AM10/10/22
to BFO Discuss
On Wednesday, 5 October 2022 at 13:01:25 UTC+1 Ludger Jansen wrote:
There is quite some literature now on the question how social entities relate to material ones. Because of this interconnectedness, it is no good idea to open a wholly new and disconnected branch in the ontology for social stuff.
@Thomas Beale: The work-around you suggested has a problem with this, I suppose.
I don't really have a strong opinion or even a decent working theory of how to deal properly with constructed realities in ontology. We could go a 'possible worlds' route or some other modal logic approach. I do know that a basic challenge is that much of what we find in constructed social realities - even mundane things like how entities are classified in government databases (e.g. taxation) - are based on accidental / contingent properties (from a material point of view). E.g. entities with names, addresses and legal standing. It is easy to imagine an alternate history in which (say) corporations were treated entirely differently from a legal point of view, or in which geographical addressing was done entirely differently (indeed it is, in Japan compared to US).
What this means practically is that the material reality of such entities will usually either have no significance (e.g. the tax database isn't going to contain an organisational and physical model of offices of each company), or else will have a torturous relationship with the social reality. Social entities are usually constructed on the basis of named entities and relationships between them, along with all kinds of contact info, classifications and so on. Theoretically, BFO:Relational quality can be used for relationships, but the BFO material bearers / owners of the relationships don't fall out cleanly with respect to the social / legal bearers / owners.
BFO is really very good for biomedical reality, and that's a primary interest for us. What an ultrasound machine sees can be well described by biomedical ontology; but the ownership, leasing, maintenance, etc details, not so well.
If we wanted to use it to represent social and legal reality, I think we have to take an approach that builds an IS-A hierarchy of the entities recognised in such spheres. Clearly, there will be challenges of cultural and societal specificity to account for, but some close-to-universal things can be found. In the end, a multiplicity of such ontologies may be unavoidable.
- thomas
Michael DeBellis
Oct 12, 2022, 12:31:42 PM10/12/22
to bfo-d...@googlegroups.com
> As I said in my first post, science moved past naive materialism with
> Newton. If you stick to naive materialism and exclude all the things I
> mentioned I am somewhat amazed you can claim BFO is a vocabulary for
> all of science. And not just science. What if my ontology wants to
> represent individuals like Jay Gatsby (the fictional hero in F. Scott
> Fitzgerald's The Great Gatsby)? Is that also excluded? Or even Julius
> Caesar? He had a material body at one point but not any more.
Julius Caesar is an instance of HomoSapiens.
Jay Gatsby is an instance of FictionalCharacter to which membership to
the kind HomoSapiens is ascribed (or presupposed).
Sorry for the late reply. Thanks for clarifying that and for being patient with my stupid questions. And sorry to keep asking questions but I'm still a bit confused because earlier in this thread someone said:
a practical point of view for us using BFO as a basis / guide for real information system is that BFO is purely a materialist ontology in the sense that even non-material things like (the text of) a play, mathematical equations etc are subordinate to their material bearers.
So if you allow individuals like JayGatsby and classes like FictionalCharacter then I don't see how you can say "BFO is purely a materialist ontology in the sense that even non-material things like (the text of) a play... are subordinate to their material bearers" Unless you are saying that FictionalCharacters are really dependent on the material books that they are written in. That doesn't seem consistent with what we typically mean by a fictional character. When I say "Jay Gatsby was partly based on Fitzgerald's love for his wife Zelda" how is Jay Gatsby "subordinate to their material bearers"? Clearly when I refer to Gatsby I don't have in mind any specific copy of the book right? Rather I have in mind the fictional character that I assume is brought to the minds of most literate Americans by that name. So what does it mean to say that he is subordinate to "material bearers"?
Michael
--
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 on the web visit https://groups.google.com/d/msgid/bfo-discuss/4c0aead5-de4b-0337-bc0e-ba0983ad4140%40uni-rostock.de.
Michael DeBellis
Oct 12, 2022, 12:38:36 PM10/12/22
to bfo-d...@googlegroups.com
BFO is really very good for biomedical reality, and that's a primary interest for us. What an ultrasound machine sees can be well described by biomedical ontology; but the ownership, leasing, maintenance, etc details, not so well.
That's interesting because from what I've read that isn't the way I've seen it described. I always thought it was assumed to be an Upper Model for everything. Which is something I'm very skeptical of because I also think having one Upper Model for all of science, including the social sciences and sciences where common sense materialism breaks down such as physics, is highly questionable. So am I just misunderstanding BFO and really it's just an Upper Model meant (or primarily meant) for the Biomedical domain?
--
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 on the web visit https://groups.google.com/d/msgid/bfo-discuss/aee5cbe2-e033-47af-a3db-6cc6af9d46fdn%40googlegroups.com.
Reply all
Reply to author
Forward
0 new messages