SUO: Re: ontology as science
o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o
John,
Again, I have no argument with any of this --
I am, as the phrase goes, way ahead of you --
my so-called career is one of being kicked
around many different blocks many different
times, and a person does not cycle through
majors of math, phys, psy, phil, comp sci,
communications, rad-lib arts, sys engine,
and then back again through the whole wash
and rinse cycle 2 or 3 more times without
picking up a number of pov's from which
to select, if nothing else.
I have been working on the nitty-gritty, pointy headed data structure
plus algorithmic side of this for some time now, and my questions are
directed to specifics. I personally abandoned the container-metaphor
approaches to context back in 1985-86, the last time they were going
through one of their recursed hey-days. The sign relational tack
is a better idea -- it does not send you spiraling outward into
an ever-expanding contextless context of boxes within boxes --
so that's the angle that I have been taking for 15+ years.
I should think that you would be copacetic with this.
I know most of the ways that theorem provers, human or machine,
can crash, and I know what has to be done to pick up the bits.
When a few other inference engineers catch up on this, then
maybe we can start having some "experienced" discussions.
Jon Awbrey
o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o
John F. Sowa wrote:
>
> Jon,
>
> I believe that a good ontology should include the basics
> of all the major science subjects, as taught at the junior
> level in universities.
>
> Unfortunately, those subjects are not consistent with one another
> or even with themselves. Every subbranch of any of the so-called
> "hard" sciences makes approximations that are inconsistent with
> every other. And they even make further approximations for
> every problem to be addressed.
>
> If you look at an EE textbook, you will see nice little
> square pulses running down the wires. But if you put an
> oscilloscope on those wires, you will see squiggly blobs
> bristling with spikes and slopover.
>
> In fluid mechanics, the Navier-Stokes equations are so difficult to handle,
> that every application uses a different approximation. My favorite example
> is the engineer who "proved" that nothing could move through air faster
> than the speed of sound. It turned out that he started with equations
> that were based on the assumption that all speeds were much less than
> the speed of sound.
>
> Bottom line: You need modules or microtheories to deal
> with the basics of any branch of science or engineering.
> And a monolithic ontology will be hopelessly inadequate
> to deal with any real problem of science or engineering.
>
> John
o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o