Path: utzoo!mnetor!uunet!lll-winken!lll-lcc!ames!ucbcad!ucbvax!sdcsvax!ucsdhub!hp-sdd!hplabs!hplabsz!taylor
From: cik@l.cc.purdue.edu (Herman Rubin)
Newsgroups: comp.society
Subject: Re: University Education and Industry Needs
Message-ID: <1278@hplabsz.HPL.HP.COM>
Date: 29 Dec 87 00:26:45 GMT
Sender: taylor@hplabsz.HPL.HP.COM
Organization: Purdue University Statistics Department
Lines: 52
Summary: the view of a professor
Approved: taylor@hplabs

Kurt Guntheroth asks:

> Why should industry give more money and resign all control?  

I believe that many firms realize that, in the long run, it is necessary for
basic research to be done.  They also need good applied research.  I do not
agree with many (I do not know if they are in the majority) university
faculty who object to professors doing applied, or even applicable, research.
In fact, I would like to strongly encourage industry researchers to teach in
the universities! 

However, I do not believe that it is reasonable to expect a good researcher
to perform well if the research is directed, whether by the firm or the
government.  That does not mean that some of the time should not be spent
on directed research.  Even a fair researcher is likely to find it
necessary for sanity to work on several problems at the same time.  (One
time I got the solution to a clearly formulated problem by discussing an
apparently unrelated result on a different problem with some colleagues at
lunch; the "experts" I had discussed the problem with were also familiar
with an intermediate problem which gave the actual solution and did not
recognize it.)  Some of the more intelligent corporations realize this, and
act accordingly.

Another reason to encourage people with research ability to work on problems
which are "pure" is that we should recognize the state of our ignorance.
Many outstanding mathematicians, some of them interested in applications
and some not, have stated that certain fields of mathematics would never
have practical applications.  How wrong they were!  It is frequently _not_
possible to do the necessary basic research when the applied problem
arises.  I think that if we imposed a 5% tax on medical treatment, the
proceeds to go to those promising researchers whose ideas are considered
"far out", that the rewards would more than justify the expenditures.  One 
editorial which appeared in _Science_ many years ago pointed out that if the
government had regulated research in Polio that we could very well have
efficient respirators, but no vaccines.

We must also recognize that basic research is inherently inefficient, and
necessarily so.  _Sometimes_ we know how to approach a problem.  If one
reads the scientific literature, the impression is often percieved that the
authors went directly from the problem to the experimentation or the
analysis presented.  Do not believe it!  I am not sure that theoretical
research is even 1% efficient.  However, the 1% useful time can not be
separated from the 99% "wasted" time.  If we know how to treat a problem,
it is not research but development.

I hope I have answered the author of the previous article, and made the point
that industry should support university research and get involved in the
educational system, but not in such a way as to stifle it.  The government
is, possibly inadvertently, already doing a good job of stifling, as 
predicted by Szilard before 1950.  I hope the industrialists show better
judgment.

Herman Rubin