RE: SUO: 2000-7-26 example - nature of organisation
John,
I think I agree with you, I think I do.
We get traction (to use Mike Ushold's term) - the rubber hits the road -
when we look at "the structural relations that relate the parts". Without
this, studying parts makes no sense. But this does not make trying to
understand the general structure that parts can have useless, does it - I
presume you were not suggesting that. If there is some general organizing
structure, this is useful to know.
Regards,
Chris
-----Original Message-----
From: sowa@bestweb.net [mailto:sowa@bestweb.net]
Sent: 31 August 2001 18:07
To: West, Matthew R SITI-GREA-UK
Cc: 'mail@ChrisPartridge.net'; standard-upper-ontology@ieee.org;
phayes@ai.uwf.edu; guarino@ladseb.pd.cnr.it
Subject: Re: SUO: 2000-7-26 example - nature of organisation
Chris and Matthew,
The following summary of Peter Simons' work highlights
what I believe is a crucial insight about mereology: it is
insufficient in itself to serve as a foundation for ontology.
MW> My reading of Peter Simons book "Parts" is that there are two
> > types of mereological object.
> >
> > 1. Pure mereological sums, where the whole is precisely
> > the sum of the parts. These are important theoretically,
> > but in practice are generally uninteresting.
> >
> > 2. Mereological sums where the whole is something more
> > than the sum of the parts. In this case there is at
> > least one structuring relation that collects the
> > parts together.
> >
> > The range of structuring relations varies enormously, and
> > results in objects whose parts are more and less tightly bound to
> > each other.
The moral that I derived from reading Simons' book is that the
title is misleading. It suggests that the notion of "part" and
mereology as a theory of parts are sufficient to serve as a
foundation for ontology. But the critical point of the book
comes in the later chapters, where he says that something more
is needed. That little bit more is what's really important.
As Matthew's summary indicates, pure mereology, like pure
parthood, is not very interesting in itself. What are really
fundamental to all the most important applications of mereology
are the structural relations that relate the parts.
That is the point I have been making again and again: if you really
want to get to the heart of ontology, writing axioms for parts is
a red herring. What you should really be studying are the kinds
of structural relations that link the parts.
Conclusion: Mereology is useful as an alternative to set theory
for combining physical entities. But the really important subject
is the theory of the combining relations -- and there are many such
relations and kinds of relations. Parthood is only a side effect.
John Sowa