COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	\input buslet[1,ra]
C00005 ENDMK
C⊗;
\input buslet[1,ra]
\jmclet
\vskip 10pt
\address 
Editorial Correspondence
Science
1333 H Street NW
Washington, D.C. 20005

\body
To the Editor:

	Gerard Piel, {\it Science} 5 September, makes many proposals,
and I'd like to report on the discussion of one of them during lunch at the
Stanford Faculty Club --- namely the proposal that part of the money
for support of science be allocated to universities as institutions
for internal distribution.  Many disadvantages were mentioned and no
advantages.

	One said, ``When I was a dean I was always glad that I
didn't have to decide the relative merits of the research of
my colleagues''.  Another said that he thought that people in
a field all over the country were more in a position to evaluate
research proposals than other people in the same university in
different fields.  Indeed when we need to evaluate someone's research
for the purposes of making a tenure decision, we rely mainly
on opinions from outside the university.  Another said that
he was happy to get his research support by mail-order, since
he didn't consider himself competent at internal university politics.

	My own field, artificial intelligence, would have been
delayed many years if it had been necessary to reach a consensus
among the faculty or deans of any university that it should be
supported.  Let me conjecture that the greater promptness of
Americans in developing new fields of science compared to other
countries is due precisely to the fact that young researchers
don't have to persuade the other professors in their own university
to give up some of their own plans in order that the newcomers
can get started.

\closing
Sincerely,

John McCarthy
Computer Science Department
Stanford University
Stanford, CA 94305

\end