Distinguishing roles from externally grounded qualities

[Niklas Beckmann]

Dear colleagues,

as far as I can tell, there is no reason to assume that BFO qualities can
not be externally grounded (e. g. in a social context).

The CCO type "Financial Value" for example seems to me to be a quality that
is externally (socially) grounded. It is defined as „A quality that inheres
in an independent continuant to the degree that that independent continuant
can serve as a medium of exchange in an economic system at a particular
time.“

But here I think it might make more sense to say that a „Financial Value“ is
only manifested in some process of exchange and/or of value appraisal.

I tend to accept the distinguishability of realizable and non-realizable
dependent continuants in general.

But it seems to me that distinguishing socially grounded qualities and roles
is especially difficult because externally grounded qualities almost always
can be said to require a process to be manifested/realized.

I also noticed that I haven't seen any quantitatively determinable roles yet
(or dispositions, for that matter), although I see no ad hoc reasons why
roles could not be quantitatively determinable like qualities.

Any insights would be appreciated. Thank you.

[Barry Smith]

On Thu, Sep 24, 2020 at 2:30 PM Niklas Beckmann <s3ni...@uni-bonn.de> wrote:

Dear colleagues,

as far as I can tell, there is no reason to assume that BFO qualities can
not be externally grounded (e. g. in a social context).

The CCO type "Financial Value" for example seems to me to be a quality that
is externally (socially) grounded. It is defined as „A quality that inheres
in an independent continuant to the degree that that independent continuant
can serve as a medium of exchange in an economic system at a particular
time.“

But here I think it might make more sense to say that a „Financial Value“ is
only manifested in some process of exchange and/or of value appraisal.

I tend to accept the distinguishability of realizable and non-realizable
dependent continuants in general.

But it seems to me that distinguishing socially grounded qualities and roles
is especially difficult because externally grounded qualities almost always
can be said to require a process to be manifested/realized.

BFO is neutral as to the contrast between externally and internally grounded qualities. Financial value, for the reasons you give, seems not to be a good example of the former. A better sort of example might be the historically determined qualities, for example of being been made in America, or cleansed in the waters of the Ganges, or having shaken hands with Elvis, or having been registered to vote. (Note that the latter does not imply that one is registered to vote; nor that one is eligible to vote.)

 

I also noticed that I haven't seen any quantitatively determinable roles yet
(or dispositions, for that matter), although I see no ad hoc reasons why
roles could not be quantitatively determinable like qualities.

Any insights would be appreciated. Thank you.

I think the ball is in your court to provide interesting and useful examples under either heading

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 view this discussion on the web visit https://groups.google.com/d/msgid/bfo-discuss/ximss-20880574%40be4.uni-bonn.de.

[Niklas Beckmann]

Thank you for the very helpful reply.

Am I right in thinking that historically determined qualities could in theory all be modeled as ontologically posterior "defined classes" like below?

cleansed in the waters of the Ganges(x) = def. quality(x) and for some r, s, t (Person(r) and Process of cleansing in the waters of Ganges(s) and Cleansee role(t) and r has_role t and t realized_in s [at t] and x inherits_in r [at t+1])

The reason I brought up determinable roles is that I at one point wondered whether one needs externally grounded qualities to represent externally grounded quantities.

As far as I can see, a mass of 1 kg can be represented as an instance of the type "mass of 1 kg". "mass of 1 kg" is a determinate quality and subtype of "mass" which is a determinable quality. Alternatively one could say that an instance of "mass" is measured by some ICE which has some value.

By the same token, if we indeed take "Financial Value" to be a role, it seems to me that an instance of it could be measured by an ICE as well and there could als be a determinate subtype "Financial Value of 1 USD“.

Kind regards

[Niklas Beckmann]

PS: The determinable role should of course better be called something like "Financially valued entity role“ and the determinate role „Financially valued at 1 USD entity role“.

[Alan Ruttenberg]

On Thu, Sep 24, 2020 at 8:03 PM Barry Smith <ifo...@gmail.com> wrote:

On Thu, Sep 24, 2020 at 2:30 PM Niklas Beckmann <s3ni...@uni-bonn.de> wrote:

Dear colleagues,

as far as I can tell, there is no reason to assume that BFO qualities can
not be externally grounded (e. g. in a social context).

The CCO type "Financial Value" for example seems to me to be a quality that
is externally (socially) grounded. It is defined as „A quality that inheres
in an independent continuant to the degree that that independent continuant
can serve as a medium of exchange in an economic system at a particular
time.“

But here I think it might make more sense to say that a „Financial Value“ is
only manifested in some process of exchange and/or of value appraisal.

I tend to accept the distinguishability of realizable and non-realizable
dependent continuants in general.

But it seems to me that distinguishing socially grounded qualities and roles
is especially difficult because externally grounded qualities almost always
can be said to require a process to be manifested/realized.

BFO is neutral as to the contrast between externally and internally grounded qualities. Financial value, for the reasons you give, seems not to be a good example of the former. A better sort of example might be the historically determined qualities, for example of being been made in America, or cleansed in the waters of the Ganges, or having shaken hands with Elvis, or having been registered to vote. (Note that the latter does not imply that one is registered to vote; nor that one is eligible to vote.)

That doesn't ring quite true to me. We've many times discriminated as not qualities things which manifest cambridge change. An "externally grounded quality" such as financial value is exactly the sort of thing that can change without there being anything about the bearer changing, vice price of gold.

It is the case that financial value can *also* change because the bearer changes, such as when the value a fruit goes down if it gets bruised or old.

So, before even the question of quality version realizable comes up, it seems that the dependencies of something like financial value should be accounted for. BFO has relational qualities which is a starting point. I have argued in the past that most dependents have relational cases. But, in the case of say dispositions, the other dependency can land up, in current BFO, being masked by being wrapped up in the condition/trigger for the disposition, something we don't have an explicit representation for.

So let's say that financial value depends on both an entity that is perceived as having the value, and the entity which sets that value. I see some different cases. It can be the case that there is an interested buyer and an object that has value to that buyer, and in that case the two dependencies are clear. In other cases it is diffuse, a result of the "market", or the even more diffuse "society". One view is that those things are aggregates/populations at least, so there is a potential target for the dependency. Even in the absence of categories for the relational cases, we can still record the dependencies with the s-depends relation.

To the question of whether "financial value" is quality or realizable, I'm with you that it falls most clearly in the realizable bucket. That said, I think it's hard for practicing ontologists, in many cases, to make the distinction. Take the classic example of "quality" color. There's no color if there isn't a process in which light shines on the object, sensors register light and some process classifies the measured spectrum of the reflected light. So is color a realizable? I'd say so. The quality is reflectance, which is a property of the material.

What to do about this? I'm honestly not sure. I try to make the distinction and think I can make it reasonably well, but often enough my judgement on these things does not make it into an ontology. I think, at least in some quarters, that the distinction between qualities and realizables is considered more trouble than it's worth. So in practice maybe things land in the realizable bucket when it's considered that the manifestation is relevant in the domain. The unfortunate thing about that is that a judgment of relevance isn't objectively reproducible, which harms interoperability, which depends on systems in which judgements are reproducible.

As an aside, I've had conversations about exactly these issues with Barry as he's worked toward a definition of capability.

 

I also noticed that I haven't seen any quantitatively determinable roles yet
(or dispositions, for that matter), although I see no ad hoc reasons why
roles could not be quantitatively determinable like qualities.

Any insights would be appreciated. Thank you.

I think most of the dependent types have cases which are determinable/determinate. In biology, for example, molecular functions are graded. Dispositions like malleability are also graded. But again, BFO doesn't have explicit representation of determinable vs determinate. So determinable/determinate is pretty much just talk. What we *have* said is that there are qualities, such as temperature or mass, which persist - one per material entity - but change what subtype they instantiate. Conceptually that makes me think about the structure of the space of types. But we have no mechanism to represent such structure, to say things like: the space of types of temperature is a one dimensional manifold. As such there is a total order of the subtypes of temperature, and the ordering relation is hotter than (or colder than). This kind of thinking has pushed me towards arguing that there are relation "universals", not just relations that as we have now that are not in the realms of entities.

We've thought of, but not followed through on, adding to BFO something like Dolce's quality regions, and I think there's consensus that this would be a good addition to BFO. We just haven't gotten to it. So for now all such things are pushed to the information/measurement side of things.

 

I think the ball is in your court to provide interesting and useful examples under either heading

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 view this discussion on the web visit https://groups.google.com/d/msgid/bfo-discuss/ximss-20880574%40be4.uni-bonn.de.

--
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/CAN_82STmkOCFZQHE8gYGmXMSnGMPM4skwwohpn6uLLHRvMMntw%40mail.gmail.com.