At 15:11 +0300 1-04-2005, martin wrote:
Hi Aldo,
Aldo Gangemi wrote:
Hi all, just a few comments.
IMO, categories (as used in ontology engineering) are not in
psychology or in nature, but they must be good enough in order to
construct an efficient ontology for a large variety of tasks.
OK. So are they constructed, and then from what? BTW "psychology" I meant
structures in our mental states. I'd argue that some are "built-in" in
our mind and some constructed. At least I read such arguments about the
"boot-strapping" of a new-born's mind?
I'm a cognitive scientist, and believe that some invariants exist
which underlie our rational processes ... the problem is how to
objectify such invariants in a way that practical ontological
categories can be derived from them. An interesting investigation we
are making is on "cognitive schemata" proposed by people like Lakoff,
Talmy, Langacker, etc. For example Rob Roy at MIT is trying to use
them in his research in robotics. As far as ontology is concerned, a
complex machinery is needed to formalize such schemata, and deploy
them appropriately in ontological engineering.
Another route is designing psychological experiments on the use of
different sets of categories, but the few studies are quite
unsatisfactory until now.
The reason why we endorse more or less the same set of categories
(object, event, quality, space region, situation, context, agent,
etc.) probably deals with the impressive cultural history behind
them: philosophy, mathematics, AI, linguistics, social sciences, etc.
have contributed so much work, and (specially Western) natural
languages are so much organized around similar distinctions (again,
probably due to the work of scientists, lexicographers, etc.), that
we are practically obliged to think that way. But once we stress the
distinctions by combining them and critizing, comparing, etc., we
often realize that different settings are just different
representations of a common rationale ... for which we do not have
the vocabulary (or even the cognitive ability) to describe (from a
different vantage point).
Philosophers usually do a different work, which consists in
challenging every distinction: their work should not be taken as
definitive. Of course, when biology will tell us more on the nature
of rationality, we'll measure the biological compliance of an
ontology.
Concerning end/perd (object/event), I've just had a long discussion
with Pat Hayes on the W3C SWBPD OEP list. We arrived at a good
compromise: endurantism and perdurantism (in their better,
state-of-art logical formulation) result to be notational variants,
with different pros and cons from the readability, efficiency, etc.
viewpoints. For those interested, I can forward parts of the
discussion.
Yes, please forward some of the discussion.
See separate email
--A
Concerning logical basis of categories, I think there is none:
logic uses its own categories which are detached from any practical
application of them (they are "neutral"). Then there is no reason
to choose one category or the other, if logic is the rationale. But
if we want useful notions that drive our modelling projects,
categories become relevant.
I like that.
Best,
Martin
Best
--Aldo
At 13:07 +0300 1-04-2005, martin wrote:
Hi Edmund,
I agree with you completely. My point was however the opposite:
Practitioners distinguish clearly between things and events, but
the philosophers seem not to have found so far a clear understanding
of what the difference between the two is. So some propose not to
distinguish between both, whereas in the CRM we declare that there is
no common instance of Persitent Item and Temporal Entity. If two concepts
are not well-distinguished, one would however expect to find
instances in the
"grey zones" in between. So either they don't exist, or we have
not encountered
them in the practical scope of the CRM, or the generalization of the
psychological difference between things and events is based on
other essential
properties than those discussed by the respective philosophers.
So, either the distinction is real, and our psychology supports it due to
innate biological experience, or we are deceived by our
psychology, and there is no
difference, or, the nature of the difference is different from the one
discussed (e.g. "being wholy present at each point in time).
I would not exclude the case that certain very high abstraction levels
have no logical justification. I am no philosopher, but I have the
impression
that Aristotle and Feyerabend support this idea, which does not mean that
this case is already at this level of abstraction.
I like this discussion for methodological reasons, because the AI and
Semantic Web community attempts to declare everything logically, which
poses the question of where the limits of logical explanation are.
Cheers,
Martin
LEE, Edmund wrote:
Hello folks,
Anyhow it seems, that practitioners do not have any problem
with the perdurant-endurant distinction, independent of how it puzzles
logic-driven thought. To which degree should an ontology be
non-logic?
I've been fascinated by this on an intellectual level, but I think
Martin has it exactly right. If practice does not require us to make
this distinction, then our tool for conceptualising the world (the CRM)
need not be concerned.
Another idea could be, that nature gave us an intuitive understanding
of
the difference, but not the means to describe it...
I'm reminded of theories (of which I know little) that suggest that we
(i.e. humans) only put as much effort into perception as will be useful
for the task in hand. A car driver will be familiar with the experience
of driving a very familiar route and not being able to recall a few
minute later what the traffic conditions were at a particular junction
although it was crossed without a problem. Compare this to driving in an
unfamiliar city, where a much greater level of perception is used.
Not really on topic, but I thought I'd chip in
Best wishes
ed
--
--------------------------------------------------------------
Dr. Martin Doerr | Vox:+30(2810)391625 |
Principle Researcher | Fax:+30(2810)391638 |
| Email: [email protected] |
|
Center for Cultural Informatics |
Information Systems Laboratory |
Institute of Computer Science |
Foundation for Research and Technology - Hellas (FORTH) |
|
Vassilika Vouton,P.O.Box1385,GR71110 Heraklion,Crete,Greece |
|
Web-site: http://www.ics.forth.gr/isl |
--------------------------------------------------------------
--
--------------------------------------------------------------
Dr. Martin Doerr | Vox:+30(2810)391625 |
Principle Researcher | Fax:+30(2810)391638 |
| Email: [email protected] |
|
Center for Cultural Informatics |
Information Systems Laboratory |
Institute of Computer Science |
Foundation for Research and Technology - Hellas (FORTH) |
|
Vassilika Vouton,P.O.Box1385,GR71110 Heraklion,Crete,Greece |
|
Web-site: http://www.ics.forth.gr/isl |
--------------------------------------------------------------
--
Aldo Gangemi
Research Scientist
Laboratory for Applied Ontology
Institute for Cognitive Sciences and Technology
National Research Council (ISTC-CNR)
Via Nomentana 56, 00161, Roma, Italy
Tel: +390644161535
Fax: +390644161513
[email protected]
*******************
!!! please don't use the old [email protected]
address, because it is under spam attack
