well now; isn't that interesting

I happened to stumble across this fascinating little article while I was drooling over lovingly restored /5s, and liveaboards in Juneau with Force 10 2 burner propane stoves and roller furling headsails, and jobs for philosophers in places like Prague and Glasgow.

... sigh ...

So, anyways, I'm strangely drawn to the idea of applied ontology, but I think the cog sci folks are getting it wrong. Here - read this:

An ontology research pipeline

Mike Uschold
Boeing Phantom Works, Seattle, WA, USA

(I snipped the first four sections because, well, they don't make much sense. You can go read them on your own at this website, if you really want: http://iospress.metapress.com/media/eafy7jvwvmdjwb0mhm2u/contributions/0/g/a/p/0gap235ukd5e9mge.pdf)

5. Summary and conclusion
I look to Applied Ontology to support the development of a research pipeline that disseminates
academic research results into industry and government. I call attention to various important issues that need to be addressed. First, we must clearly articulate the value propositions for using ontologies. What are the business cases? Under what circumstances do they apply? Benefits need to be demonstrated in a convincing way. This requires progress in evaluating specific techniques and tools used in ontology-based solutions. Where possible, evaluation should be done using a scientific methodology, and in all cases the range of applicability of approaches should be made clear. Limitations as well as strengths need to be highlighted. Ontologies need to be engineered for particular purposes to ensure successful deployment, yet they also should be re-usable as much as possible. Tools and infrastructure to support
the creation, use, and maintenance of ontologies must be created and evaluated. Finally, I call attention to the challenge of living with ambiguity and exploring the limits of what kinds of semantic processing can be automated.

[Applied Ontology 1 (2005) 13–16 13; IOS Press]

Can you see why I think this is a little screwy?

Problem 1: From academia to government? Since when have philosophers (especially ontologists) performed a prescriptive role in civil discourse? Anyone? Why would ontologists be especially well-equipped to tell folks in business and government what sort of ontology to adopt? Why should anyone listen?

Problem 2: Since when do I need to articulate the "business cases" of an ontology as a criteria for determining their value? Is he really saying that the best ontology is the one that gives me the ability to describe the most stuff in the way that gets me the most items of value? Is cash value a concept applicable to an ontology? That just seems flat-out wrong - I can't arbitrarily assign that sailboat or this motorcycle an ontological status of "mine" (or owned-by-me, if you'd rather), because it simply isn't true.

Problem 3: Evaluation should be done using "a scientific methodology," apparently. Doesn't seems to matter which one. No questions about the sort of ontological assumptions inherent in any of the methodologies one has at one's disposal.

Problem 4: "Ontologies need to be engineered for particular purposes to ensure successful deployment, yet they also should be re-usable as much as possible. Tools and infrastructure to support the creation, use, and maintenance of ontologies must be created and evaluated." Say what?

So ontologies are like parachutes? Or old german sports cars? Or bombs?

Ontologies don't seem inuitively to be the sort of things which one engineers, or deploys, or sends to the shop for maintenance with some sort of conceptual socket wrench - it's a strangely mechanical way to speak about an organic process. (Organic meaning "arising in a non-silicon brain/mind" as opposed to mechanical meaning "arising in a constructed apparatus." An arbitrary distinction, true... but bear with me.)

As far as an infrastructure to support my ontology, that would have to be an epistemology and a metaphysics... now there are some under-studied fields of applied philosophy.

I applied ontology gives you a bunch of engineers and cog sci folks sitting around the lab grumbling about how to get their computer programs to talk to one another, then what would a room full of applied epistemologists look like?

A library? a Mensa meeting?

I don't think you could get two applied metaphysicians into a room together, but if you could, it would probably look like something out of a Beckett play.