On Wed, Feb 22, 2006 at 05:43:33PM -0000, West, Matthew R SIPC-DFD/321
wrote:
> I'm not arguing that language is not important, just that it is not
> ontology. (01)
Well, I don't think anyone would disagree with *that*. (02)
> Mostly I hope we are talking about exchange between computers, so e.g.
> English language terms are not a critical element - not many databases
> read English, and it is between databases that most computers exchange
> information. (03)
I was not thinking about natural language at all, but the terms in
whatever formal representation language we are using to write our
ontologies. And my claim is that big-O Ontology is very much about
fixing the meanings of the terms in those languages. But look, to
assign meaning is to express a purported word/world connection, so the
endeavor is unavoidably connected to questions of existence -- I just
don't happen to think the answers to a lot of those questions matter all
that much to effective knowledge engineering. (04)
> So in ISO 15926 we distinguish between the thing of interest, and the
> term(s) that might be used to designate that thing. (05)
Well, distinguishing words from their referents is of course an
essential distinction to make. (06)
> Now whilst I accept that this amounts to giving a meaning to a term, (07)
Right, if by "this" you mean the association of a thing with a term. (08)
> this is a by-blow of ontologically identifying that there are terms
> and objects that they refer to as part of the ontology. (09)
You lost me there. (010)
> So it worries me if what we think we are really doing is defining
> terms. (011)
If by "defining" we mean not only the strict sense in which a defined
term is in principle dispensable but also *axiomatizing*, i.e., fixing
the meanings of primitive terms by means of sentences that express their
properties and logical connections, then I think that is pretty much
exactly what we are doing -- and I am nonplussed as to the grounds of
your worries. Ontology is about fixing meaning, but we don't do so
aimlessly or arbitrarily. We write axioms that reflect our intuitive
and/or scientific understanding of some domain and that embody
assumptions (e.g., that future objects are no less real -- and hence no
less available as values of our variables -- than objects in the
present) that might streamline the expression of that understanding. So
it's misleading to say that, in Ontology, "what ... we are really doing
is defining terms" as if defining/axiomatizing itself is the endgame.
The endgame is usefulness -- an ontology must serve to clarify our
understanding of its domain and enhance our ability to manage, reuse,
and share the information about the domain that we deem important. If
you also want to go on from there and draw philosophical conclusions to
the effect that this ontology or that gets at the True Nature of Things
or more clearly reveals What Really Exists, well, knock yerself out. As
a philosopher, I do think that there is (philosophically) interesting
and (philosophically) important hay to be made there. But as a
knowledge engineer, I don't much care. (012)
Chris Menzel (013)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/uos-convene/
To Post: mailto:uos-convene@xxxxxxxxxxxxxxxx
Community Portal: http://ontolog.cim3.net/
Shared Files: http://ontolog.cim3.net/file/work/UpperOntologySummit/uos-convene/
Community Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?UpperOntologySummit (014)
|