Re: SUO: Re: Industry takeover
----- Original Message -----
From: "Robert E. Kent" <rekent@ontologos.org>
To: "John F. Sowa" <sowa@bestweb.net>
Cc: "SUO" <standard-upper-ontology@ieee.org>
Sent: Saturday, May 03, 2003 12:06 PM
Subject: Re: SUO: Re: Industry takeover
[snip]
> IFF-OOv2, which is due out this fall, is geared towards a better
> representation of populated ontologies (ontologies with instance data).
> These are represented as IFF logics (an IFF logic is an IFF theory and an
> IFF model-theoretic structure that shared a common IFF type language).
OK. Perhaps I was a bit too fast here. Actually, in the IFF the notion of a
logic breaks up into two notions. See the current IFF logic namespace
document at
http://suo.ieee.org/IFF/metalevel/lower/namespace/logic/version20021205.pdf.
The above notion is called a prologic. An IFF *prologic* is an IFF theory
plus an IFF model(-theoretic structure) that share a common IFF type
language. This FOL notion corresponds loosely to the Information Flow (IF)
notion of a (not necessarily normal) local logic. See chapter 12 of
"Information Flow" by Barwise and Seligman. Here the model may not satisfy
all of the axioms in the theory. The restricted notion is called a logic. An
IFF *logic* is an IFF prologic where the model _does_ satisfy all of the
axioms in the theory. This FOL notion corresponds closely to the IF notion
of a normal local logic.
Every logic is (embeds as) a prologic, and every prologic can be restricted
to a logic by dropping non-compliant tuples from the model. This two way
passage is a reflection called "restriction". For more on the category
theory of this reflection see in general the architectural diagram on page
36 and in more detail the restriction reflection discussion and
axiomatization starting on page 47 of the metatheory document at
http://suo.ieee.org/IFF/metalevel/lower/metatheory/ontology/version20021205.
pdf.
Robert E. Kent
rekent@ontologos.org