[Cf. the whole sequence: http://theunitedpersons.org/blog/posts-on-personal-ontology-1-0-alpha-on-bfo-list ]
On 6 December 2010 04:15, Alan Ruttenberg wrote:
>> 2. Alan, in a way it have obtained: I raised this issue as a routine
>> http://plato.stanford.edu/entries/thought-experiment 
>
> I don’t consider thought experiments to be evidence. They are good
> argument. In this case there are a number of substantive issues that
> we are far from understanding, so I don’t think the thought experiment
> is enough.
> 
ok
>> You say that AIs running on the internet would be generically
>> dependent continuants (not MaterialEntities) and all generically
>> dependent continuants have material bases, which are presumably
>> MaterialEntities, specifically, spatially dispersed sets of Objects.
>
> Yes.
ok
>> You say that I’m an independent continuant while I’m flesh and blood,
>> “uploading” would change me (still me) into a dependent continuant
>> with a material basis (presumably a set of Objects), and “downloading”
>> again would presumably change me back into an independent continuant.
>
> I don’t believe I said that.
> Rather: I don’t really know what constitutes the identity of a person.
> If it came to be known that this identity could be preserved when
> uploading, then we would know that it was a kind of generically
> dependent continuant. When in your body, the body would be the
> material entity that bears the concretization of that GDC.
> BFO doesn’t have such an entity (your sense of person/AI) so nothing
> would need to be changed in BFO should we learn this is possible.
> However if other ontologies had made a commitment that contradicted
> this they would need to be changed.
makes sense and consistent in itself: you say that BFO is independent (and you are agnostic, at least as long as possible) on a lot of issues. The concept of agency/personship is one of them.
>> But do I have the right to fork
>> http://en.wikipedia.org/wiki/Fork_(software_development) BFO at all?
>
> You do, but it would be considered bad practice to change any of the
> intended meanings of BFO terms, or to distribute it in any way that it
> might be confused by someone to be the real BFO.
>
> Better to create another ontology and import and use BFO terms where
> appropriate.
>
>> I feel so that the default copyright for ontologies is
>> http://creativecommons.org/licenses/by/3.0/ but I’d appreciate if you
>> would publish the copyright (e.g.
>> http://creativecommons.org/about/licenses/ ) of BFO.
>
> It is so said in the OWL file
> http://creativecommons.org/licenses/by/3.0
Copyright: thank you, I didn’t notice it.
Disambiguation: I understand your concern. I’ll do it bona fide.
>> On 29 November 2010 19:56, Alan Ruttenberg wrote:
>>>>[I think it would not be useful to regard
>>>> her [the AI] as something that has a material basis.
>>>
>>> It isn’t a matter of useful. It’s a matter of true.
>>
>> I’m sorry I was negligent.
>>
>> 1. Clarification: I think it would not be useful to regard an eventual
>> AI or the uploaded me as a dependent continuant with an independent
>> continuant material basis which is a scattered set of material
>> Objects. I think it would be useful to regard an AI and myself to be
>> independent continuants, invariant to eventual changes in our material
>> bases.
>
> That’s fine. Buy you are not using the terms as BFO intends them.
> Better to define new additional terms that define things the way think
> they should be. FWIW, I think it is possible that you don’t understand
> what it means to be a material basis in BFO.
ad1. see below
>> 2. Truth: I think what I say (modified by the Clarification above) is
>> (ontologically) true = invariant to (ontological) transformations (cf.
>> the end of my previous letter).
>
> Yes, but the truth I was referring to was deeper. There isn’t
> anything, according to BFO, that doesn’t depend on something material.
ad2. see below
>> 3. Usefulness: I think usefulness (and elegance etc.) is not
>> superfluous in ontological research either. Pragmatic approach to
>> ontology presupposes usefulness (and at least non-falseness). I think
>> there are useful ontologies just like there are useful theorems:
>> “[...]
>> – useful theorem: theorem that leads to many new ones [...]”
>
> It depends on what your definition of ontology is. BFO is trying to be
> an ontology in the sense of being a catalog of types of things that
> exist. It is proposed that building ontologies in this sense is
> already damned useful for it’s targeted purposes.
I think I understand what material basis means in BFO (but that thought of mine may be a part of the problem:) Let’s check it and see.
You say that “exists = (matter itself and/or) have material basis”.
‘Exists = (matter itself and/or) have material basis’ := ‘X=Mb’
Case I.
If you think X=Mb is an axiom of BFO 1.1, it’s ok. Then you can say that
- either you can slide “information” into the pigeon holes of the BFO 1.1 (e.g. an AI on the internet or the uploaded me or an imaginary magenta unicorn pegasus are generally dependent continuants with material basis),
- or “information” (which does not have material basis) is nonexistent (it would contradict to the axiom).
This is the scenario
- where you say that “BFO is trying to be an ontology in the sense of being a catalog of types of things that exist”, and you reject that meaning of “information” which may or may not have material basis (obviously along with other similar “things”, e.g. http://en.wikipedia.org/wiki/Luminiferous_aether ) presumably as “speculative metaphysics” (à la Kant, Heidegger), and
- where every continuant with a speculative mind (like me:), be it independent or dependent, using the fact that in BFO 1.1 owl:Thing =/= bfo:Entity, have the legitimate opportunity to reveal other categories under owl:Thing but outside the category bfo:Entity, e.g. a category Nonentity, or a category dolce-lite:Quale.
- This approach is too “postmodern” to be an ontology in one (the “continental”) meaning (“[Philosophy] the branch of metaphysics that deals with the nature of being” [= owl:Thing, including a rigorous and exhaustive organization of it that is hierarchical and contains all the relevant entities and their relations]),
- but it’s ok in the (“analytic”) other (“[Logic] the set of entities presupposed by a theory” / “[computer science] a rigorous and exhaustive organization of some knowledge domain that is usually hierarchical and contains all the relevant entities and their relations”) http://www.thefreedictionary.com/ontology
Case II.
But if
- you think that X=Mb is a deep (basic) truth (of ontology in philosophical meaning) mapped into an axiom of BFO, i.e. BFO is a basic ontology in the philosophical meaning above, and
- tertium non datur,
then I think X=Mb is simply false.
(I don’t know whether BFO has this – type II – ambition at all:
- on one hand, it is so close to it that it would be a sin:) to miss the opportunity,
- on the other, in BFO 1.1, owl:Thing =/= bfo:Entity as if BFO would not be confident enough to be type II and would ensure place for the diversity of ontologies.)
Why false? Suppose for a while that I completely accept the solution of BFO 1.1 for AI-s running on the internet (i.e, they are generically dependent continuants, not MaterialEntities, and all generically dependent continuants have material bases, which are MaterialEntities, specifically, spatially dispersed sets of Objects).
However, this would not solve
and so would not mean that X=Mb.
If you say that matter is the nesessary building block for owl:Thing, you make the mistake in logic I made at first (“I may like information to be the “infrastucture” of matter but you are right: for an ontology this would be a superfluous presumption and a logical fallacy of this kind: http://www.philosophyexperiments.com/mary/Default.aspx “).
So in Case II we do the right thing if we abandon our demand for the preeminence of information and of matter and in general of any specific owl:Thing.
What the universe (more exactly: owl:Thing) is made of at the “fundamental level” is an open empirical (rather than a speculative / a thought experiment kind of) question yet (cf. e.g. http://arstechnica.com/science/news/2010/12/lhc-spots-no-black-holes-eliminates-some-versions-of-string-theory.ars ): I don’t think so that BFO in Case II should decide on it in advance. BFO in Case II should be explicitely independent on all “speculative mathaphysical” answers.
Solution
In Case I: There is no problem to solve.
Your axioms/categories are yours, mine are mine: there is no further need to argue (you don’t argue fiercely e.g. on Dewey numbers:) and no need to change anything on BFO 1.1. The only (absolutely legitimate!) task left is to keep BFO away form revolution and let it evolve into other disciplines.
In Case II: Back to the future!
Fortunately, even if I am right in what I am saying above, BFO itself(!) has a solution for the problem of Case II: your internal solution in Case II is BFO 1.0 ( http://www.ifomis.org/bfo/1.0 ) The structure of the previous version of BFO is “agnostic”: independent on what owl:Thing is supposed to made of.
Of course, BFO 1.1 is an improvement on 1.0: Disposotion, Function and Role are bundled into a category in 1.0, and bundling Object, FiatObjectPart and ObjectAggregate into a new category in 1.1 is a parallel idea. BFO may retain the new structure in the next version of BFO as well.
The problem is, as the forensic investigations show:), that this stuctural change was accompanied by the silent introduction of X=Mb by calling the new category ‘MaterialEntity’. This is why BFO 1.1 is not just a harmless improvement of 1.0. From the point of view of Case II, 1.1 is 1.0 constrained by the false X=Mb.
I think (the Platonic idea of) BFO is deeper than 1.1, and 1.0 is closer to it.
Suggestions
I’m a just a user of BFO, not even a heavy one, one among many thousands. You should put my suggestions in their right place.
For me it is very important to clarify: how would look like the BFO I would be happy to work with, without the need to fork it?
1. It would be modular: a core and auxiliary axioms (or alternative auxiliary axiom systems). BFO is independent on a lot of issues. The concept of agency/personship is one of them. For me, agency/personship is ontologically important, but you are right: this doesn’t imply that the core of BFO should contain personship. However, it implies that BFO should not preclude the ontological concept of personship.
BFO should be modular: its core
- should remain indenpendent on theories in ontology (like the ontological construction of the concept of agency and personship) and
- should become independent on open empirical questions (like the fundamental constitution of owl:Things).
2. It would have a Type II core.
The core of BFO seems to me the right place for the philosophical/continental meaning of ontology (Case II above). This is why
2.1. I would consider removing the the root category bfo:Entity from BFO. In philosophical meaning of ontology bfo:Entity is not different to owl:Thing, it would be an inappropriate postmodern modesty:) not to erease it.
2.2. I would change the label of the BFO 1.1 category ‘MaterialEntity’ into another label without reference to matter and to any other specific hypothetical constitution of owl:Thing.
(2.3. I would revisit the examples and the wording of the definitions of BFO 1.1 based on the arguments above.)
3. It would have different Type I auxiliary axioms.
3.A. With the BFO core + X=Mb as an auxiliary axiom, we could essentially reconstruct BFO 1.1.
3.B. With the BFO core + an auxiliary axiom alternative to X=Mb, I would be able to describe what I think to be ontologically important.
Results
In this https://groups.google.com/group/bfo-discuss/browse_thread/thread/433bfa9718cae15 conversation BFO 1.0 has revealed itself as more profound than BFO 1.1. Minor changes in BFO 1.0 could result in a future BFO core independent on material or any other specific hypothetical constitution of owl:Thing. BFO core + auxiliary axiom X=Mb would reconstruct BFO 1.1. BFO core + alternative auxiliary axioms would ensure conceptual space for alternative ontologies without the need to change or fork the BFO core, and without any interference with the reconstructed BFO 1.1.
[Original: https://groups.google.com/group/bfo-discuss/browse_thread/thread/433bfa9718cae15?hl=en ]
