Re: Theory and experiment

[Trond Andresen <Trond.Andresen@xxxxxxxxxxx>]

Jim D. says

> Trond writes:
> "The
> definite proof of the superiority of a theory in engineering, physics,
> chemistry is the lab demonstration, showing that results are better
> (in some agreed-upon sense) than that given by alternative theories.
> Either in explaining a phenomenon,
> or in solving a problem.
>       ^^^^^^^^^^^^^^^^^^   <- my emphasis, added now, Trond

> "The main problem with economics isn't that better theories are (in
> principle) impossible to find.  It is that only one type of experiment
> on a national* economic level is allowed...."
>
> Though I agreed with most or all of what Trond said after that,
> I think that we shouldn't put too much emphasis on lab experiments.
> As Stephen J. Gould makes clear, sciences like palentology can't
> do lab experiments either. (He terms palentology, astronomy, and
> the like "historical sciences," and should include economics, too.)
> What one can do is "postdiction": one can test a theory by looking
> for new historical evidence to refute or confirm the hypotheses
> generated by the theory.

The fundamental difference between paleontology and economics is that
economics additionally is a science for _doing_ and _steering_, not only
for observing and understanding what has happened until now. You must
have a climate where different national and other economics projects
are allowed to be tried out without obstruction or sabotage, if the
"doing"-part of economics science shall develop soundly.


Trond


-----------------------------------------------
| Trond Andresen (Trond.Andresen@xxxxxxxxxxx) |
| Department of Engineering Cybernetics       |
| The Norwegian Institute of Technology	      |
| N-7034 Trondheim, NORWAY		      |
|				              |
| phone (work)	+47 73 59 43 58               |
| fax	(work)	+47 73 59 43 99               |
| private phone +47 73 53 08 23  	      |
-----------------------------------------------