process histories and associated dependent continuants
bpeters
entity branch.
Several terms we have describe objects in terms of past processes that
they have participated in. There are implicit modifications to the
object by this process. These changes are hard to define without
referring to the process itself. Examples are: 'irradiated', 'heat-
killed', 'digest'.
I pick 'irradiated' in the following. Irradiation is a protocol
application in which the input material is exposed to radiation. We
are having trouble saying what the output material is. We know what it
_was_, but the irradiation process has now introduced random changes
to the physical structure. We solved this by defining defining
'irradiated material' as the output_of an irradiation process.
So far so good. However, a persistent request is to use the term
'irradiated' by itself. To do that, we are struggling to say what
would it would be in BFO?
Our approach so far: 'irradiated' would be a specifically dependent
continuant. It seems closest to role or quality, but we don't believe
they would fit completely. it is not realizable (otherwise, how does
it inhere in an entity before the realization? It is not a quality
(otherwise we should be able to define it without referring to what
the entity was.
Would 'irradiated' require a different kind of specifically dependent
continuant? Something differently named, but similar to 'state'? What
would be the relations between 'irradiated', the process of
irradiation, and the irradiated material?
As with the last post, any help would be greatly appreciated. The more
specific it is to the application of BFO (in contrast to discussing
its fundamental design choices), the better.
Thanks!
bjoern
Barry Smith
>Another OBI help request, this time on behalf of the (bio)material
>entity branch.
>
>Several terms we have describe objects in terms of past processes that
>they have participated in. There are implicit modifications to the
>object by this process. These changes are hard to define without
>referring to the process itself. Examples are: 'irradiated', 'heat-
>killed', 'digest'.
>
>I pick 'irradiated' in the following. Irradiation is a protocol
>application in which the input material is exposed to radiation. We
>are having trouble saying what the output material is. We know what it
>_was_, but the irradiation process has now introduced random changes
>to the physical structure. We solved this by defining defining
>'irradiated material' as the output_of an irradiation process.
>
>So far so good. However, a persistent request is to use the term
>'irradiated' by itself. To do that, we are struggling to say what
>would it would be in BFO?
PATO has occasionally included adjectives rather than nouns to
represent qualities. I think this is unfortunate, but something we
will probably have to live with where no noun presents itself.
We have, then, a quality, irradiated. It is a dependent continuant.
It is a quality entities have only if they have undergone processes
of irradiation, and consists in the presence those characteristic
qualitative features which such processes create.
>Our approach so far: 'irradiated' would be a specifically dependent
>continuant. It seems closest to role or quality, but we don't believe
>they would fit completely. it is not realizable (otherwise, how does
>it inhere in an entity before the realization? It is not a quality
>(otherwise we should be able to define it without referring to what
>the entity was.
I think it is a quality. (I am pretty sure that PATO already has
other quality terms with this feature.)
>Would 'irradiated' require a different kind of specifically dependent
>continuant? Something differently named, but similar to 'state'? What
>would be the relations between 'irradiated', the process of
>irradiation, and the irradiated material?
We could think about creating a special subclass of qualities --
those whose existence necessarily involves a prior process.
BS
Alan Ruttenberg
On Jun 14, 2008, at 11:46 AM, Barry Smith wrote:
> PATO has occasionally included adjectives rather than nouns to
> represent qualities. I think this is unfortunate, but something we
> will probably have to live with where no noun presents itself.
> We have, then, a quality, irradiated. It is a dependent continuant.
> It is a quality entities have only if they have undergone processes
> of irradiation, and consists in the presence those characteristic
> qualitative features which such processes create.
>
>
>> Our approach so far: 'irradiated' would be a specifically dependent
>> continuant. It seems closest to role or quality, but we don't believe
>> they would fit completely. it is not realizable (otherwise, how does
>> it inhere in an entity before the realization? It is not a quality
>> (otherwise we should be able to define it without referring to what
>> the entity was.
>
> I think it is a quality. (I am pretty sure that PATO already has
> other quality terms with this feature.)
The problem is that processes don't always leave a trace of
themselves in the structure of the object. Moreover, the qualities
that they do leave, when they do leave a trace, are not always
uniquely generated by the process.
As an example of a process that doesn't leave a trace, but does get
noted as history, consider dephosphorylated protein.
Dephosphorylated protein is the same as protein before it was
phosphorylated, i.e undecorated protein. One can't tell the
difference, other than looking at the history, between protein that
was never phosphorylated and protein that was phosphorylated and then
dephosphorylated.
As an example of the latter, consider protein being "precipitate".
The quality in this case is insolubility. However, being precipitate
only means that there was a material transformation that occurred
that caused the quality to be acquired. However, the same quality
might be acquired in different ways, in particular in other than in a
situation where there abundant purified protein in solution - the
usual situation in which would call such insoluble protein precipitate.
-Alan
Barry Smith
A better proposal would be: to create a special subclass of qualities
-- those whose existence necessarily involves a prior
engineered/artificial process (a process that is the product of
deliberate human intention) (since of course every quality in the
natural world involves some prior processes in its history).
Use of the corresponding qualities would then go hand with interest
in artifactually created substances (deliberately created toxins,
drugs ...), where history matters. Only if the it matters that the
dephosphorylated protein is artifactual would we need a corresponding
quality term.
(Artifacts are, by definition, entities whose history matters.)
BS
Alan Ruttenberg
> A better proposal would be: to create a special subclass of
> qualities -- those whose existence necessarily involves a prior
> engineered/artificial process (a process that is the product of
> deliberate human intention) (since of course every quality in the
> natural world involves some prior processes in its history).
>
> Use of the corresponding qualities would then go hand with interest
> in artifactually created substances (deliberately created toxins,
> drugs ...), where history matters. Only if the it matters that the
> dephosphorylated protein is artifactual would we need a
> corresponding quality term.
> (Artifacts are, by definition, entities whose history matters.)
> BS
There's a couple of things here. On the question of whether they are
a subclass of qualities, I have to say that this doesn't feel quite
right. They seem to me to be more like roles, except that they point
backwards in time rather than forwards. If the are about "things that
matter", then they are mind co-dependent in a way similar to roles,
but not similar to qualities.
On the issue of artifacts, I'm not sure how to understand your
comment about them being "by definition, entities whose history
matters". Their history matters as long as it matters, but there are
other things whose history matters, for example that a part used to
be part of an organism.
I think we will do an experiment in OBI and make these a dependent
continuant class that is sibling to role and quality and see how it
plays out. Once we have some more examples it may be easier to review
them and see if they form a coherent group, and whether, upon review
they belong as a subclass of one of the other dependents.
Best,
Alan
jennifer fostel
process (in the example from Bjoern "irradiated") could be considered
a transient change, for example "fed" which produces satiated animals
only for a period of time.