Issue 13 in bfo: has_material_basis_in

[b...@googlecode.com]

Status: New
Owner: ----
Labels: Type-Defect Priority-Medium

New issue 13 by albertgoldfain: has_material_basis_in
http://code.google.com/p/bfo/issues/detail?id=13

We often want to establish a relation between a disposition and the
material entity that confers that disposition. This manifests itself in
OGMS as a relation between a disease and the disorder(s) that confer the
disease.

After some email discussions with Barry Smith, Lindsay Cowell, Bjoern
Peters, and Alan Ruttenberg, I believe something like this might do the
trick:

has_material_basis_in
domain: bfo:disposition
range: bfo:material entity
definition (*): d has_material_basis_in m iff
(1) d is_a bfo:disposition
(2) m is_a bfo:material entity
(3) i is_a bfo:independent continuant and i has_disposition d because i
has_part m

(* Note: d, m, and i are instances per RO policy)

A synonym could be 'conferred_by'. Another
proposal, 'has_physical_basis_in' I don't like as much because
the 'physical' doesn't seem to make you think of the proper bfo type in the
range (i.e., material entities).

[Bill Hogan]

I'd like to emphasize the need for this relation for work on the
ontology of diseases.

I think the proposed relation and its definition are quite good and
worthy of adoption now, with ongoing refinement done as needed.

One minor change I'd suggest though is that if d, m, and i are
instances, then it should read "instance_of" instead of "is_a":

has_material_basis_in
domain: bfo:disposition
range: bfo:material entity
definition (*): d has_material_basis_in m iff

(1) d instance_of bfo:disposition
(2) m instance_of bfo:material entity
(3) i instance_of bfo:independent continuant and i has_disposition d
because i has_part m

Bill

> --
> You received this message because you are subscribed to the Google Groups
> "BFO Discuss" group.
> To post to this group, send email to bfo-d...@googlegroups.com.
> To unsubscribe from this group, send email to
> bfo-discuss...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/bfo-discuss?hl=en.
>
>

[Ludger Jansen]

We think that the proposal still confuses the basis of a disposition (which
is a dependent continuant as the disposition itself) and the parts of its
bearer.

For in many cases it seems doubtful that some material part of the bearer
is
the base of the disposition.

In standard examples from chemistry it is rather some structure of lower
level properties of the whole bearer of the disposition. If we take the
molecular structure of NaCl as material base of the water solubility, it is
clearly that particular structure which confers the disposition, not the
mere fact that the substance has Na+-Ions as parts.

Of course, there are cases where only a part of an entity confers the
disposition to the whole. The ability of my body to digest food ist based on
the respective ability of some body parts, viz. my digestive system. But the
digestive system as whole has the disposition, not only in virtue of its
parts, but of their acting together in a properly functional way etc. And
only this is usually meant in the philosophical debate with "base" of a
disposition. If a whole x has a disposition in virtue of a part p, we are
still faced with the question what confers the disposition on that part p.

So our suggestion would be to take as base m in general not a material
entitiy and a separate part of the bearer, but some quality/property or
stucture of properties of the bearer (or of some part of the bearer).

In line with this approach, we suggest to use the relation name "has_basis":

d has_basis q iff

(1) d instance_of bfo:disposition

(2) m instance_of bfo:quality

(3) i instance_of bfo:independent continuant

and i has_disposition d because i has_quality q

The "because" should probably be spelled out along the line that every
independent continuant that has the quality q is also bearer of the
disposition d, as follows:

(Q is_a Quality & D is_a Disposition & Q base_of D) -->
For all q: (q instance_of Q -->
There is a b and a d such that:
q inheres_in b &
d instance_of D &
d inheres_in b).

While equal bases bring with them instances of the same type of
dispositions, the inverse does not hold. As a rule, dispositions in the
biomedical domain can be constituted by different types of base qualities.
Our general formulation leaves leeway for several candidates for bases. If a
patient lacks a certain enzyme one could either take that absence as the
basis of the disposition for the resulting patho-physiological processes.

Or, to avoid "negative entities" like absences, one could say that the whole
pathological (micro-)structure of an organ is the base of the respective
disposition for a pathological process. Such a pathological structure can be
further analysed into its constituents and their respective dispositions.
Thus we get something like a "cascade" of dispositions: The disposition to
produce insufficient amounts of the enzyme leads to the bodily state
characterized by lack of the enzyme and that state is in turn the base for
the disposition for pathological processes resulting from that state.

Cf. R�hl/Jansen: "Representing Dispositions"
(OBML workshop Manheim, Sept 9-10, 2010)
http://www.onto-med.de/obml/ws2010/obml2010report.pdf

Best greetings from Rostock,
Johannes R�hl
Ludger Jansen

[Cristian Cocos]

The "base/basis" that Ludger seems to be looking for has a consecrated name in contemporary science: is called a "state," only BFO doesn't recognize states--not as far as I know (though I may be a bit behind in my BFO approfondisments). At one point I think a state was supposed to be represented as something akin to "process boundary," though I'm not sure that that proposal ever led to anything. (To be entirely honest, science is (programatically) ambiguous as to the range of structural properties that are constitutive of the very system as opposed to constitutive of the *state* of the system.) How about "structural quality" as a closer approximation to "state"?

C

Cf. Röhl/Jansen: "Representing Dispositions"

(OBML workshop Manheim, Sept 9-10, 2010)
http://www.onto-med.de/obml/ws2010/obml2010report.pdf

Best greetings from Rostock,

Johannes Röhl

--
"People think that I must be a very strange person. This is not correct. I have the heart of a small boy. It is in a glass jar on my desk." -- Stephen King

[b...@googlecode.com]

Updates:
Labels: -Type-Defect Type-BFO2-design

Comment #1 on issue 13 by alanrut...@gmail.com: has_material_basis_in
http://code.google.com/p/bfo/issues/detail?id=13

(No comment was entered for this change.)

[b...@googlecode.com]

Updates:
Labels: -Type-BFO2-design Type-BFO2-Reference

Comment #2 on issue 13 by alanrut...@gmail.com: has_material_basis_in

[Barry Smith]

Does Christian have a definition of 'state' ?
BS

[Stefan Schulz]

I would say a state is a process, in a context which ignores the
internal structure of this process.
(patient in a semiconscious state, in a hypothyroid state, in a manic
state, in a septic state etc.)

-
Stefan





2012/5/8 Barry Smith <phis...@buffalo.edu>:

Stefan SCHULZ  (Univ.-Prof. Dr. med.)

Institut für Medizinische Informatik,
Statistik und Dokumentation
Medizinische Universität Graz
Auenbruggerplatz 2/V
8036 Graz (Austria)

http://www.medunigraz.at/imi
http://g.co/maps/aqedt

+43 (0)316 385 16939
+43 (0)316 385 13201

http://purl.org/steschu
mailto:stefan...@medunigraz.at
Skype: stschulz

[  home: Afritschgasse 32/3
[  8020 Graz (Austria)
[  mobile:  +43 (0)699 150 96270
[  http://g.co/maps/m8rau

[Cristian Cocos]

On 5/8/2012 08:46 AM, Barry Smith wrote:

Does Christian have a definition of 'state' ?
BS

A satisfactory definition of "state" that does not essentially leverage the term "set" in some form or other is still beyond my reach, which is (among others) why I did not follow up on my earlier promise to deliver a draft of a BFO-compliant definition. (One of the other factors that dampened my enthusiasm was that it looks to me that pursuing a system-state outlook might lead to some consequences that might be harder to swallow by the some of the more conservative BFO advocates.) At any rate, I am tossing below a couple of random (and far from carefully curated) thoughts that emerge from the use of the term in Physics as I know it, should anyone want to pursue it:

Processes are sequences of states (...whence one might infer that states are t=constant process sections? Would that be something akin to a SNAPshot?)
A state is the totality of (values of) the system's state parameters, meaning those parameters that fully determine the temporal evolution of the system. (Note that I attempted to avoid any epistemic intrusions, as this is usually present in the literature as "the knowledge of which allows predicting the evolution of the system at any given time in the future.")
The mathematical underpinning of this outlook is obviously the idea of Cauchy boundary conditions, as system evolution is usually given my means of differential equations.
"(State) parameter" is what initially drove me to propose a view on states as bundles of qualities of a certain continuant.

[Ludger Jansen]

I think that the word "state" is ambigous in an ontologically relevant
way. It can be used to refer to a continuant as well as to an occurrent.
Compare the following sentences:

(1) Continuant meaning: "This state endures since 1 May."
(2) Occurrent meaning: "During a continuous change, an object is in a
certain state only instantaneously."

Of these two, I think, the latter is more basic, as in the first
sentence "this state endures" can be rendered as "Object #22 is in state
thus-and-thus since 1 May".
My spontaneous candidates for an object's being in a certain state are:

a- this object being bearer of certain qualities or quantities
b- this object being the bearer, together with other things, of certain
relations
c- this object's parts being the bearer of certain relations

Christian's conception of system states seems to be a variant of c.
These things seem to be compliant with BFO, only that "state" allows for
being vage on exactly which of a,b,c and their subtypes we mean.

If I understand Stefan correctly, he would interpret
(3) Patient #12 is in a septic state
as
(3*) Patient #12 is participant of septic processes.
Though this resembles b, it is distinct from this. For here we have

d- an object standing in a certain formal relation to another entitiy.

Maybe d adequately captures medical talk. But probably we do not need d
for accounting, e.g., for the basis of dispositions (which was the
context of Christian's remark that has been quoted by Barry). For the
basis of a disposition is normally seen as a continuant entitiy, not as
the participation in an occurrent entitiy.

Best
Ludger

>>>> (Q is_a Quality& D is_a Disposition& Q base_of D) -->

>>>> Cf. R�hl/Jansen: "Representing Dispositions"

>>>> (OBML workshop Manheim, Sept 9-10, 2010)
>>>> http://www.onto-med.de/obml/ws2010/obml2010report.pdf
>>>>
>>>> Best greetings from Rostock,

>>>> Johannes R�hl

PD Dr. Ludger Jansen
Institut f�r Philosophie
Universit�t Rostock
18051 Rostock

[Bill Hogan]

"Medical talk" encompasses all of a-d, I would say.

Examples:

a. Febrile state
b. Carrier state (w.r.t. certain infectious diseases)
c. (admittedly having trouble coming up with example here)
d. Sleep state (participating in sleep process)

>>>>> Cf. Röhl/Jansen: "Representing Dispositions"

>>>>> (OBML workshop Manheim, Sept 9-10, 2010)
>>>>> http://www.onto-med.de/obml/ws2010/obml2010report.pdf
>>>>>
>>>>> Best greetings from Rostock,

>>>>> Johannes Röhl

> Institut für Philosophie
> Universität Rostock
> 18051 Rostock

[Albert Goldfain]

The proposal is to have a *material* basis for a disposition (hence the relation name 'has material basis'), not a processual basis. 

If you wanted to accommodate the processual reading of 'state', you might have a relation like 'has processual basis', although I think that might confuse things since it is one more level of indirectness in terms of dependence. 

Consider: A pregnant woman bears the disposition to vomit during the first trimester.
The realization of this disposition is clearly vomiting.  What might the basis for the disposition be:

"state of pregnancy" is ambiguous
(1) 'having a part containing a zygote/fetus'  is not ambiguous (material basis that confers the disposition...simplifying a lot, I admit)
(2) 'participating in a pregnancy process' is not ambiguous (processual basis???)

In any event, both (1) and (2) are more precise than "state of pregnancy" and the scope of the proposed relation is just BFO material entities.

AG

[Chris Mungall]

On May 10, 2012, at 6:41 AM, Albert Goldfain wrote:

The proposal is to have a *material* basis for a disposition (hence the relation name 'has material basis'), not a processual basis. 

If you wanted to accommodate the processual reading of 'state', you might have a relation like 'has processual basis', although I think that might confuse things since it is one more level of indirectness in terms of dependence. 

Consider: A pregnant woman bears the disposition to vomit during the first trimester.
The realization of this disposition is clearly vomiting.  What might the basis for the disposition be:

"state of pregnancy" is ambiguous
(1) 'having a part containing a zygote/fetus'  is not ambiguous (material basis that confers the disposition...simplifying a lot, I admit)
(2) 'participating in a pregnancy process' is not ambiguous (processual basis???)

In any event, both (1) and (2) are more precise than "state of pregnancy" and the scope of the proposed relation is just BFO material entities.

In the case of (1), the range of the relation isn't a BFO material entity, is it?

The material basis is not the embryo or the uterus or the mother (all MEs); it is the having-a-part-containing-an-embryo (let's go with your simplification for now), which is a different thing from any of these 3 MEs.

What kind of thing is the having-a-part-containing-an-embryo? It seems like a state/quality type thing. I don't think anyone would object if "pregnant" were to be added to PATO. 

Perhaps states could be modeled as qualities. This requires a certain amount of liberalism in what constitutes a qualities - can we just take any collection of relationships about an entity and reify them as a quality? I suspect that this would be met with objection by BFO purists. But I'm not sure how to distinguish "true qualities" from arbitrary reifications (or if there is a need to). 

[Cristian Cocos]

Ludger, I just think that you're being too delicate when calling the
first sentence (#1 below) as "ontologically relevant" in any way. I
happen to think that it's just sloppy language, though there's a way to
explain how that came to be (basically an abstraction artifact, as
Stefan seems to be suggesting)--which I will attempt to do as soon as I
get some respite.

C