Fram

a)

(<—)

J

i

 

——

October 16, 1963

Committee on Appointments & Promotions

Halsey Royden
Minsky Appointment

The documentation on Minsky seems to me to be very thorough. The
letters by cCarthy and Forsythe are detailed and, I believe, fairly
accurate summaries, You will note supporting Stanford letters From
Lederberg and Suppes, a dissenting letter from Gilbarg, and outside
letters from Miller, Perlis, Simon, and Wang. i have a few additional
comments for the Committee,

After consultation with the Greduate Dean and some of the users
of the Computer Center, our office has adopted the policy that the
Computer Science Division shouldhaye . two additional (half-time) men
in the near future, one in numerical and one in non-numerical aspects
of computing. It is planned to handle the numerical role at the
assistant professor level for the time being. i believe that the non-
numerical appointment should be strongly machine oriented. Of the
possible candidates for this position I would rate only Perlis above
Minsky, and we have already spent two years in an unsuccessful attempt
to induce him to Stanford.

Both the Provost and I have made inquiries in the East, and are
satisfied that the case presented for Minsky is fairly accurate, Most
impressive was my conversation with Wang at Harvard, Evidence for Wang's
enthusiasm is the fact that his division (Applied Science) at Harvard
has voted that Minsky be offered a Professorship. At the Columbia
Computer Center I was told, without my having mentioned Minsky, of
excellent progress on time sharing programs being done at MIT under
Minsky and at Stanford under McCarthy. Suppes is already planning a
considerable research program which will use these routines, and I am
confident that they will be valuable for other research in the social
sciences and perhaps also for "on-line" computing in the natural sciences.

+

It seems to me that Gilbarg's dissent is based largely on the
application of “centralist" standards. Such standards are widely held by
mathematicians, but I am not convinced that such standards should be used,
Moreover, it is my personal opinion that Minsky's paper on the unsolva-
bility of Post's tag problem is a first-class piece of work by any standards.

I think that the addition of Minsky to McCarthy will give us
considerable distinction in the non-numerical field and will considerably
enhance the usefulness of Computer Science to other areas of the University.