SUO: Re: IFF Model Theory Ontology 1.0, Request For Comment
Hi Jon and Matthew,
Jim F. is out of town, so I'll give a preliminary response to your comment.
----- Original Message -----
From: "Jon Awbrey" <jawbrey@oakland.edu>
To: "Matthew West" <Matthew.R.West@is.shell.com>
Cc: "Robert E Kent" <rekent@ontologos.org>; "Jim Farrugia"
<jim@spatial.maine.edu>
Sent: Tuesday, May 14, 2002 6:42 AM
Subject: Re: IFF Model Theory Ontology 1.0, Request For Comment
[snip]
> roost, and any formal system that just assumes it knows of a
> hard and fast distinction between tokens and types, or atoms
> and predicates, or whatever, is not one that i would trust.
This comment seems to be directed mainly at the idea of a classification,
the central data structure of Information Flow, and only derivatively at the
idea of an IFF model, the central data structure of the IFF-MT, which is
represented as a the fibered product (pullback) of classifications and
hypergraphs.
In one sense, a classification is no more or less than a binary relation.
The difference is in attitude. In our approach to classifications we think
of one component as instances (or tokens) and the other component as types.
But this attitude is somewhat ephemeral, as witnessed by the fact that we
have an involution operation that transposes things -- type to instance and
instance to type. In a classification, types classifiy instances and
instances classify types.
It should also be pointed out that there is no requirement that the sets of
instances and types be disjoint (It is even useful sometimes to make them
identical). So there can be elements of a classification that function as
either an instance or a type in different circumstances. For example, "red"
is an instance of a "color" and the "color of your red Ferrari" is an
instance of the color "red."
In summary, in Information Flow an instance is not thought of as an
individual, but as an element in one pole of a polarity.
Robert E. Kent
rekent@ontologos.org