STANFORD UNIVERSITY
MEDICAL CENTER
PALO ALTO, CALIFORNIA

DEPARTMENT OF GENETICS DAvenport 1-100

School of Medicine

Cabies STANMED

November 2, 1966

Dr. H. J. Muller
Indiana University

Dear Jo:

I very much appreciated your letter, and your sending the draft type-
script of Zour Chicago talk. I will certainly read it carefully, and mind
your cautions, before writing about it.

I think I can sympathize with the frustrations of the discrepancy be-
tween your nervous and physical energies. You know best what compromises
you have to make. The paradox is that it must be just your friends and
admirers eager for some small contact with you who, in aggregate, must make
the largest demande: Anyhow, you know where I fall in outlook, and I surely
don't want to be any burden on you.

In the last five years or so, I have rather drastically changed my outlook
on science. The system is an efficient one for maintaining the essential
integrity and discipline of our approximations to reality, and I don't
see much about the research-system that I would want to change. But while
one is caught up in the process of discovery, one tends to overlook that
the only realities for a human being are in his communications with other
men. My column-writing is a very satisfying vehicle for matx reaching these
motives, and I am glad to have the chance to perfect the craftsmanship by
experience. Your general comments are very welcome, and so would be any more
concrete ones you might care to make at any time. You could not have made
a more flattering allusion than your reference to Haldane, whose flair for
commentary made an overt model for what I have been after too. Then, besides
yourself and perhaps Huxley, who else is usefully putting himself out this
way? Well, Medawar and lately Crick are dabbling too.

You made some brief reference to neurobiology and computer intelligence.
Shooter and Varon are doing some interesting things on protein-chemistry of
the brain and dispersed-neuron cultures and reaggregations, as the line of
this department in that direction. I have been dabbling myself (together
with Ed Feigenbaum of the Computer Science Dept.) trying to teach a computer
to think about organic chemistry. (My own first pass at genetics couldn't
find a good point of departure; chemasstty is farg easier.) As you might ex-
pect, the results are paradoxical-- the hard-looking things are easy, and
vice versa. But we do have some fairly clever programs by now. One has to
put as much effort into teaching them as any obstreperous undergraduates.

I would always welcome hearing from you.

Sincerely,

Joshua Lederberg