Ontological categories vs. phenomenological categories
At 10:11 AM -0400 11/7/2000, John F. Sowa wrote:
>The important distinction that Peirce was trying to make is the
>difference between ontological categories (what a thing is) and
>phenomenological categories (how we perceive, interpret, or
>think about something).
>For the SUO, we will need both kinds of categories, since we have to
>represent the words that people use for talking about the world, and
>they freely intermix both kinds of terms in the same sentences.
>
>There is no contradiction between the two kinds of categories, since they
>describe different aspects: how a thing is and how we think about it.
>We need both ways of description, but we do have to distinguish them.
This time I agree with you, John. Although it may be difficult (or
impossible) to really pin down what a thing is at what you call the
"ontological" level (since, as Chris Welty noticed, we are
superimposing our judgements on the reality anyway), however it is
very important in practice to distinguish between between these two
perspectives: in simple terms, the former may be called "physical",
the latter "cognitive", or "conceptual". I am reluctant to use
"ontological" for the former since I believe we can use the tools of
formal ontology to describe both: you may have an ontology of reality
and an ontology of commonsense [I agree with Bill Burkett that
commonsense is "real", too, but in a distinguished sense].
Reality and commonsense represent two different (although related)
subject matters for ontology. I agree very much on the fact that, for
our purposes, we need to take both into account, and to explore the
relationships between them while keeping them distinct. I believe
that a good deal of the hot debate existing in philosophy (and
computer science) about ontological issues is due to the fact that
one perspective is privileged to the other, or the two perspectives
are confused.
Take for instance the classical example of the vase and the clay it
is made of. From a physical perspective, the vase is just a piece of
clay with a certain shape. From a cognitive perspective, that shape
is extremely relevant, so relevant that we use a specific name for
pieces of clay exhibiting that shape. As a result of this cognitive
process, we tend to ascribe different identity criteria to the vase
and the clay. If we take this intuition seriously, we have to admit
that, at a certain time, there are two different (co-located)
entities there, the vase and the clay. The former ceases to exist if
the vase is destroyed, the latter does not.
Of course this vision creates a lot of philosophical and technical
problems. An easy solution is to stick to the physical perspective,
maintaining that we only have a piece of clay there, and being
(shaped as) a vase is just a property like being red, being full of
water, and so on. A radical variant of this solution is to think
that, after all, there is no such an enduring thing like a piece of
clay, but rather a series of "stages", each one different from the
other, corresponding to a particular four-dimensional distribution of
some basic physical variables (mass, charge, spin...).
Both these moves are perfectly plausible, and - I would say -
correct. The problem is that they refuse to "describe" the cognitive
perspective, but rather try to reformulate it in more basic terms. I
understand that this is the so-called "revisionary" approach to
ontology, as opposite to the "descriptive metaphysics" approach.
Philosophers adotping these views tend to say that the puzzle related
to the difference between the vase and the clay is only apparent, and
it is an artifact of our language, culture, and so on. Certainly -
they say - this is not an ontological problem, but rather a
linguistic (semantic) problem: the goal of ontology is not to
describe the way a certain term like "vase" is used in natural
language. Therefore, the linguistic tests that may justify the
cognitive position are not relevant.
My position, in accordance (I believe) to the above comments by John
Sowa, is that for the SUO purposes we can't rely *just* on the
physical perspective. In particular, we *must* account for the way a
term like "vase" is used in natural language differently from "clay".
This means that we must develop an ontology of commonsense as well as
an ontology of "physical world". The interesting aspect, I believe,
is that we can adopt the same formal tools to develop both. In
particular, the basic notions of so-called "formal ontology"
(parthood, unity, identity, dependence...) can be effectively used
exactly because they are "formal", independent on a particular
domain. Physical world and commonsense are just two different (but
related) domains.
As a final note, we may wonder how to distinguish between an entity
belonging to the physical ontology and one belonging to the
commonsense ontology. Implicitly, this was the problem of Chris
Welty, since he observed that everything is ultimately perceived by
us... My practical solution here is the following: if the identity
criteria ascribed (within a certain conceptualization) to a certain
entity x are assumed to be universal (independent on particular
cognitive judgements, and so on), then x belongs to the physical
ontology, otherwise it belongs to the commonsense ontology. So
"electron" or "clay" would be physical, while "vase" or "mountain"
would be commonsense.
-- Nicola
---------------------------------
Nicola Guarino
National Research Council phone: +39 O49 8295751
LADSEB-CNR fax: +39 O49 8295763
Corso Stati Uniti, 4 email: Nicola.Guarino@ladseb.pd.cnr.it
I-35127 Padova
Italy
http://www.ladseb.pd.cnr.it/infor/ontology/ontology.html
(***updated 12/7/2000***)