Marcelino, (01)
I'll try to respond without going into too much detail for a
reasonable post. I believe that this is a false dichotomy. The
framing of this difference in program as "realist" vs. "cognitivist"
is significantly the result of arguments on the allegedly "realist"
side. But from the perspective on an epistemologist, the position
argued by that side is hardly representative of realism. Contemporary
epistemology, philosophy of science, and cognitive science all provide
rich resources toward understanding scientific and other realism,
informed by further understanding of the inherently partial and
purposeful nature of modeling and knowledge-seeking. Partial models,
influenced by the purposes for which those models are developed and
the nature of the confirmation/disconfirmation to which they are
exposed, are often locally necessary, based on both the processes by
which experts produce models and the processes by which scientists and
other reasoners use models. To acknowledge the appropriateness of
local, partial models -- particularly the kind of partiality by which
models differ in which aspects of reality they emphasize and capture
and which they deemphasize or entirely omit -- is not to say that
those models are not all models of reality. (02)
I personally believe that the most reasonable, best-supported, and
most explanatory way of understanding ontological, scientific, and
much other modeling is something (and perhaps some others) call
Contextual Realism. The most straighforward way to explain it is
something like this: (03)
* There is a real world.
* It is incredibly, fantastically complicated.
* One aspect of that complication is that the world looks very
different depending on where in it you are standing and what you are
doing.
* One's chance and ability to get things more *right* about the world
is much improved if one takes into account, or at least acknowledges,
the fact of this complex variation in perception of, cognition about,
and interaction with the world.
* It may be the case that a unifying model of reality, that takes into
account all of these contextual variations in perception, cognition,
and interaction while also incorporating the aspects of reality
partially captured in local models, is, in principle possible.
* It seems likely that any such unified model is itself incredibly
complex. Certainly, anything approaching it is a long way from what we
can manage now, even collectively.
* Knowledge advances in particular sciences (e.g., physics v.
chemistry) through use of local, partial models. Even were it possible
to force all chemists to work in the terms and models of physics,
there seems to be no advantage to doing so, and much disadvantage.
* Similarly, knowledge can and does advance in many areas, at many
levels of granularity, through us of local, partial models. Even if it
were possible to force abandonment of those local, partial models,
there seems little advantage to doing so, and there is much
disadvantage.
* Contextual Realism recommends acknowledging the locality and
partiality of the views of reality available and usable from within
different contexts (including different model-motivating activities)
* Contextual Realism also recommends that pursuit of understanding
that crosses such contexts, whether for the purpose of bringing
together knowledge from multiple contexts or for some other purpose,
is best supported not by attempting to erase the local models and the
complexities of reality and our interaction with it from which they
result, but by attempting to *capture* the variations and dependencies
as much as we can, where we encounter them. Doing so also has the
pragmatic benefit of enabling humans from outside the context of some
given model to begin to understand the assumptions, use, scope, and
limitations of that model, and thus to use or not use, and to
interpret it, accurately. (04)
Thus, it can be understood that an applied, computational ontology --
the kind of artifact with which we are concerned -- is a model of an
underlying reality. And as a model, It is itself a conceptualization.
It is also created for some motivating purpose, that is, to enable
understanding of , interaction with, reasoning about some aspects of
that reality. And as a model, it is inherently incomplete - created
in part via incomplete interaction with reality and contextualized
perception and cognition. (05)
It *might* be possible, in principle, to create a grand unified
ontological model that incorporates all of these partial models. But
(a) it is not certain that this is so, and (b) such a model would not
be usable to for any purpose for which we currently can use models.
Such a thing, if possible, would surely be too unwieldy to use for
local, specialized reasoning, by humans or machines. (06)
On the other hand, creating ontological models that reflect
acknowledgement of, and to some degree explicitly capture, their
contextual location not only serves the purposes that motivate
modeling in the first place, but also provides for the possibility of
using information about context to enable making connections between
local, contextual models. Who knows? The accumulation of information
about such conditions of variation *might* lead us slowly towards more
complete, less partial, understanding. It *does* give us models that
support reasoning and discovery within fields, in part by respecting
that localized, partial conceptualizations often reflect optimization
of models by experts and through experience. And, to the extent
context/provenance factors are captured, good local, contextual models
also enable sharing knowledge across contexts, by supporting the
incorporation of knowledge *about* context into interpretation. (07)
So, I say: it's a false dilemma. Ontologies can be models of reality
*and* conceptualizations, if both realism and conceptualization are
understood non-naively and with an acceptance of the extraordinary
complexity of reality and an appreciation of the ways in which real
contexts, including knowledge-seeking purposes, can motivate, drive,
and shape conceptualizations. (08)
Best,
Amanda (09)
On Wed, Mar 21, 2012 at 14:15, Marcelino Sente <zaratruta@xxxxxxxxx> wrote:
>
> In this paper, the author discuss some aspects raised by this topic and
>present a "cognitivist ontology".
>
>
>http://www.mind-consciousness-language.com/CarstensenTowardsCognitivistOntologies.pdf
>
> The paper present a contrast between two main views within the research
>program in ontologies:
> - Realist view: the ontology is build upon universals in reality rather than
>concepts. A good ontology is one which corresponds to reality as it exists
>beyond our concepts.
> - Conceptualist view: the ontology is a explicit specification of a
>conceptualization. A good ontology is one which captures our shared
>conceptualization.
>
> I would like to know the position of the members of this forum about this
>aspects.
>
> Thanks
>
>
>
> 2012/3/16 Marcelino Sente <zaratruta@xxxxxxxxx>
>>
>> Let consider this assertion (refered as A1):
>> “It should be emphasized that we are talking about a perceived world and
>not a metaphysical world without a knower” (Rosch 1978, p.29)
>>
>> How this assertion impacts on the research program of ontologies?
>>
>> Some initiatives regarding conceptual modeling, systems interoperability,
>and conceptual analysis have been using of theories coming from the domain of
>formal ontology. I say "formal ontology", in the sense of Husserl, as
>analogous to formal logic. Whilst formal logic deals with formal logical
>structures (e.g.,truth, validity, consistency) independently of their
>veracity, formal ontology deals with formal ontological structures (e.g.,
>theory of parthood, types and instantiation, identity, dependence, unity),
>i.e., with formal aspects of entities irrespective of their particular nature.
>Some (so called) foundational ontologies (as UFO - unified foundational
>ontology) embody several conceptions coming from the "formal ontology".
>So...What A1 say about the use of conceptions imported from "formal ontology"
>to the territory of semantic web, communication among computer and humans and
>systems interoperability?
>>
>> How can we view and compare the contributions related to the realist
>semantics and cognitive semantics, regarding our objetives expressed above
>(semantic web, communication among computer and humans and systems
>interoperability)? Does make sense think in terms of cognitive semantics in
>the reserach program of ontologies?
>>
>> Reference:
>> E. Rosch (1978) Principles of Categorization. in: E. Rosch and B. Lloyd
>(Eds.), Cognition and Categorization. pp. 27-48, Lawrence Erlbaum
>Associates, Hillsdale, New Jersey.
>
>
>
>
> _________________________________________________________________
> Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
> Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
> Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
> Shared Files: http://ontolog.cim3.net/file/
> Community Wiki: http://ontolog.cim3.net/wiki/
> To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
> (010)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J (011)
|