Description vs generation, Note to D'Agostino

[From Rick Marken (950913.0830)]

Hans Blom (950912f) --

Physics has succeeded because physics does not try to understand. It is
content with having theories and formulae that are descriptive and that make
accurate predictions.

Your notion of "understand" must be a heck of a lot differnt than mine.
I can't imagine what you could possibly mean by "physics does not try to
understand". The notion that physicists are content with description and
prediction alone is also off the map of my understanding.

Even the practical people who use physics every day are not just "content
with having theories and formulae that are descriptive and that make
accurate predictions". If description and prediction were enough, satellite
controllers would not be doing what they do all the time -- propagating
orbits. Instead of propogating orbits, all the satellite controllers would
need is the descriptive equation of the satellite's orbit at the time of
orbit insertion in order to predict where to look for the satellite in the
future -- say, on Sept 23, 1995 at 0128 zulu.

Unfortunately, the orbit of a satellite is not perfect; it degrades as
a result of unpredictable disturbances like solar wind, gravitational
irregularities due to the shape of the earth (especially for low fliers),
etc. In order to get the best prediction of where the satellite will be on
Sept 23, 1995 at 0128 zulu, the controllers measure the satellite's current
position and velocity (Keplerean elements) and use differential equations
(Newton's laws) to _generate _(propogate) the future positions of the
satellite, assuming no disturbances.

The satellite controllers are not using physics to _describe_ an orbit; there
is no orbit yet to describe; the controllers have only one data point (the
Keplearean or "2 element set"). The controllers are using the laws of physics
to _generate_ an orbit. Once the orbit has been generated, the controllers
can predict where the satellite will be on Sept 23, 1995 at 0128 zulu. The
accuracy of this prediction depends on distubances that occur between the
time of the propogation (say, today) and Sept 23, 1995 at 0128 zulu. But the
inaccuracy due to these disturbances is MUCH less than the inaccuracy that
would result if the prediction of the satellite's position on Sept 23, 1995
at 0128 zulu were based on the equation that _described_ the satellite's
orbit at the time of orbit insertion (which might have occurred years ago).

Brian D'Agostino --

My post to your compuserve address seems to have bounced so I am putting my
message to you on CSG-L. I hope you are still on the net.

I am wary of summarizing my theory, because the last time I did that in
Durango, it elicited "few questions and many polemical attacks," as I said
in my posting on CSGnet.

It seems to me that the net is a great forum for discussing ideas; and if you
really are interested in developing a PCT model you've got all the expert
right here in cyberspace;-) If, however, your application of PCT is already
satisfactory from your perspective then, I suppose, the net won't be that
useful.

If you let me know your postal address, I'd be happy to send you a
copy of the article, and you can then read as little or as much as you wish.

10459 Holman Ave
Los Angeles, CA 90024

I'd like to take a look at it; but I still think it would be most useful to
talk about it on the net.

Best

Rick