Rich Cooper: (01)
> how do we account for personal ontologies in a semantic system? (02)
I have a useful suggestion about this. Please bear with me as I lead up
to it. (03)
Communication of any kind is founded on rhetoric. Rhetoric is prior to
all systems of classification or model. For the transmission of an idea
from one awareness to another, the really basic prerequisite is that the
awarenesses both employ similar techniques for identifying things.
(Yes, a technique for identifying things may be a model, but it may be a
model that's broken, and its brokenness may not be a bug, but rather a
feature. Curiously, the model can be different in the hearer than in
the speaker, and it/they can still do its/their job(s) of supporting
communication. Rhetorics are allowed to be slippery, but we expect more
consistency from "models".) (04)
Techniques for subject identification are more a matter of rhetoric than
logic. However, we tend to think some logical system of classification
is necessary for identification, and therefore for communication. And
it is true that rhetorics always seem to imply logical classification
schemes. The problem is that when we try to discover the logical
foundations of rhetorics, things fall apart in a way that's reminiscent
of the Babel myth, or perhaps the mythic image of a snake eating its own
tail. At the end of the day, communication doesn't submit to such
analysis; it refuses to be modeled in any way that's satisfactory to us
model-makers. (05)
We can more successfully strengthen communication (and, therefore,
strengthen our ability to share systematic/logical tools like
ontologies) by focusing, first, not on the "logic of rhetoric", but
rather on rhetoric itself, with no prior demand or expectation that
rhetoric be logical or even modelable, in the normal sense of modeling. (06)
Recognition of the priority of rhetoric is embodied in some (but
certainly not the majority) of work on Topic Maps. The approach is
evident in Part 5 (the "Topic Maps Reference Model") of ISO 13250. Part
5 demands disclosure of the rhetoric(s) of subject identification that
is/are used in any given topic map. Such a rhetoric may or may not have
logical features. It may or may not involve one or more ontologies. It
may or may not involve a database used for reference. No supporting
tools are excluded, although one may certainly question, in any given
situation, the usefulness of, for example, oracular or stochastic tools
for subject identification. Part 5 only demands an answer to question: (07)
How shall the recipient of the topic map be expected to
distinguish whether two things under discussion in the
topic map are different, or the same? (08)
This demand does not require any particular answer, nor does it require
that the answer reveal internal consistency, or predictability, or any
of the other goals that modelers normally have in mind. (09)
When we recognize the priority of rhetoric over logic, Rich Cooper's
question, "How do we account for personal ontologies in a semantic
system?" becomes pragmatically answerable. We can stop requiring that
the identities of subjects under discussion be disclosed in any
particular way, or even in any way that provides any particular kind of
satisfaction. (010)
Having shed our demand for any particular kind of power over the
material under discussion, we are then in a position to devise abstract
rhetorics that allow multiple rhetorics, and multiple inconsistent
logical systems founded on them, to coexist and "play nicely together",
even in a single message. We can perhaps even exploit the powers
provided by one rhetoric in the contexts of others. (011)
Obviously, it's a matter of opinion whether a given subject of
conversation is really the same even when it's under discussion in two
different rhetorics, but, hey, that was always the case. Every attempt
at communication is an attempt to share an opinion. The implicit claim
of Part 5 is that it may be practical to share opinions in ways that
encompass multiple rhetorics. That we can, in the words of Rich Cooper,
"account for [multiple] personal ontologies in semantic systems". (012)
(If your reaction to the above is, "Huh?", or "So what?", you can take
comfort in the fact that you're in very good company, and I thank you
for reading this. If your reaction is to ask, "What are the
implications?" or "What do such disclosures look like?", then maybe we
should talk.) (013)
Rich Cooper wrote:
>
> So it looks like the consensus among those in this discussion is:
>
>
>
> An ontology is a collection of
>
>
>
> classes, each with possibly unique property values;
>
>
>
> a few constant instances (e.g., equilateral triangle = special
> instance of generalized triangle, etc);
>
> and
>
> logical relationships among the classes and instances.
>
>
>
> And nothing else. If that satisfies everyone, then any operational
> system would require more than just an ontology. It would also
> require that information nobody seems to want to call ontological,
> like the specific employees in the employee table.
>
>
>
> If we accept this definition among the group of us, an ontology with a
> database to back it would be about the simplest semantic system I can
> imagine being useful. The database would store the instance data
> beyond the ontology, but the ontology would define the classes,
> properties and relationships among the entities.
>
>
>
> But then how do we account for the diverse viewpoints going into the
> system from multiple users? We all agree that each user has a unique
> ontology of her personal world. We know that subjectivity gets
> squeezed into the tightest databases with the strictest controls.
>
>
>
> So how do we account for personal ontologies in a semantic system?
>
>
> (014)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (015)
|