On Thu, 31 Jan, Harry Pollard <[EMAIL PROTECTED]> wrote:
>You are right to separate the two strands.
>
>However, my separation would be different.
>
>The "science" you speak of I think is mostly mathematics. Mathematics is
>a great tool, but is never better than its premises. And they are often
>highly suspect.
>
>These scientists have drawn around themselves a self consistent world
>which is not part of the real world. (As is said, Arthur, to an economist,
>reality is a special case!)
>
When mathematics is applied to the problem of the nature of the physical
world, it's called physics, and it works pretty well, within its domain.
When mathematics and sometimes, by extension, physics, are brought to
bear on problems in the real world, where dirt and warm bodies and
other inconvenient things get in the way of purely analytic solutions,
it's called engineering, and that is where economics rightly belongs.
Some decades ago, I took a course in celestial mechanics, which used
the beautifully elegant Newtonian formulations to develop a framework
for computing the positions and movements of bodies under gravity,
accounting for all the perterbations caused by the other gravitating
bodies in the neighbourhood. This magnificent triumf of mathematical
physics had taken about two hundred years to perfect, and incorporated
levels and levels of corrections and adjustments to allow the prediction
of orbital positions of objects at some considerable accuracy some
centuries into the future. When the prof had finished presenting this
aspect of the course, he said, more or less, "well, that's truly
wonderful, but now for the actual truth:". What good was this edifice?
Was it used in the computation of ephemeres? Well, no, the computing
engines could produce more accurate results more rapidly by using
a Ptolemaic model of the solar system, with simple circular orbits,
and simply heaping on more and more tiny epicycles to account for
the perterbations to arbitrary accuracy. Well then how about for
the computation of the paths of spacecraft, surely that would need
this sort of computation? Nope. It is much simpler and more dependable
to simply use the very first order approximation, and accurate
position sensing telemetry to feed back to the computing system
which could then correct the motion to the required accuracy just
as easily, and in fact more easily because the computations were
simpler. So in truth, the work honed to a high polish by 19th
century mathematical astronomers turns out to have about as much
utility as a steam powered abacus.
>[KH]
>*Economics also has to try and make sense of the oft-irrational
>side of human nature that erupts from time to time. And, like the other
>human sciences, economics can't be said to have made much headway so far.
>
>The latter is much more complex. What do we make of it?"
>
>The first thing is to disabuse ourselves of the notion that there is such
>a thing as "human nature" per se. Our "human nature" is really an
>accumulation of all sorts of different "drives", "instincts", "genetic
>propensities", call them what you will.*
>
>This is a real mess isn't it, Keith? A veritable anthill of humanity from
>which come the premises of mathematical economics leading inevitably to
>inadequate conclusions.
>
>However, the Classical Political Economists didn't hide behind
>mathematical jargon. They looked at people and particularly at persons.
>And they hypothesized the rules that would apply to all the different
>"drives", "instincts", "genetic propensities". And as you know they
>came up with the two Basic Assumptions of human nature that described
>the behavior of everyone - every single person. You must know them by
>heart, now.
>
>"Man's desires are unlimited."
>
>"Man seeks to satisfy his desires with the least exertion."
>
Which will make their computations as useless as any others which are
not based in systems engineering principles of what is actually
happening feeding back into the model to improve the accuracy of
its parameters. You see, as any sociological study of economists
will tell you, and has been discussed here before (where were you?)
economists more than any other group of people sorted by any
measure, regard people as venal, greedy, contemptible, robots,
"Homo economicus" I think Ray suggested for these imaginary
creatures, who defy all human virtues in order to act according
to the arbitrary dictates of the economists' dog-eat-dog fantasy
world. Real people, by contrast, sometimes actually treasure concepts
like fairness, compassion, and non-material goals. And each culture
possesses such individuals in different numbers, and values them
to differing degrees. Only a robust engineering structure can hope
to keep up with the vagueries of human nature well enough to
make a functioning economic model which takes this sort of
variable into account.
>There again, you'll recall that the single complicated human being is not
>analyzed in Classical Political Economy. Rather we look at his connection
>with the economic world, which is the way he exerts. The manifest
>indication of the person (no matter how complex) is found in the way he
>exerts.
>
> What other evidence have you?
>
>Once we have people somewhat pinned down, we have to look at the equally
>complicated world - so complicated that it is impossible to think about,
>which doesn't stop people trying.
You won't get people "somewhat pinned down" with any a priori
assumptions. You build your engineering structure to be able to
turn on a dime, and reflect the nature of people as you find them.
If the top-down (theory-first) economists had it right, economics
wouldn't be as lame as it is. That will continue to be the case,
as I've said often before, until economics is absorbed under
systems engineering, at which point the improvement in effectiveness
will develop so fast it'll make your head spin.
-Pete Vincent
