SUO: Re: Enhancing Data Interoperability with Ontologies...
Hamish,
I just wanted to comment about models:
JS> I agree with you completely on this issue. I wasn't
> burying the ontology issue. I was just deploring my
> own role in helping to popularize the word "ontology".
> I think the word "model" would be less confusing.
HH> OK. Though where I am "model" would be an outright
> distaster, and I think it might be more generally
> problematic too. When I say "model" to a colleague,
> they think immediately and specifically of a numerical
> simulation model of an environmental process. Or usually
> of some particular implementation of that, or worse a piece
> of software which wraps that implementation in a whole raft
> of cute GUI and data management toolery, or or ... Model,
> in other words, is a massively overloaded word.
I agree that it's massively overloaded, but I also believe
that there's a common core of meaning to all those senses.
I like to identify models in Tarski's sense with models
in the engineering sense (and I came across a paper by
Petri, of Petri-net fame, in which he said that he also
believes there's an underlying affinity between them).
One advantage of the word "model" is that I can then
quote George Box's comment about engineering models:
All models are wrong, but some are useful.
And I say the same about models in Tarski's sense.
That, of course, ruffles the feathers of some logicians
who deserve as much ruffling as I can give them.
But that gets to my reason for preferring the word "model"
to the word "ontology": every engineer and scientist
recognizes that models are tentative things that are
intended to be tested and more often than not rejected.
But people who use the word "ontology" tend to think
that an ontology must be cast in bronze and preserved
for all eternity.
I don't mind using the word "ontology" if it is recognized
as one language game (in Wittgenstein's sense) among many.
But I panic when people glorify their pet viewpoints and
set up an Inquisition to eliminate the heretics.
John