BookReview: Three Degrees Above Zero
by Peter L. Bernstein, Scribner, September 1, 1984, 978-0684181707

Peter Bernstein takes us on a fascinating tour of Bell Labs, the research institution. He covers the numerous discoveries, the management philosophy, and above all, the people. The people made Bell Labs the premier research institution in its heydays.

[p31] [Ronald Graham:] If you really knew what you were trying to do. that would often be the biggest part of the battle. There does not seem to be any obvious way of knowing how some development here will impact on something over there. You just hope you have good people who are excited and that they can communicate."

[p108] [Addison White:] Incidentally, as a comparison, I think that our country [p109] has been conducting its affairs in an absurd way on cancer research. much of which is being done without a guiding idea. We are trying to throw hundreds of millions of dollars at the problem without enough ideas. which is nonsense. It would have been the same sort of thing if we had tried to move more rapidly in the case of the transistor."

[p113] [Addison White:] When it came to the transistor I had no real ideas of my own. and that. I think. is a good thing for a technical manager. because it means that he does not compete professionally with the men who are reponing to him. 1 think Bill Shockley's fundamental problem as a manager was that -he was competing. Perhaps because of my personal history. I have always been utterly skeptical that any school of business administration can teach people how to manage. My Own concern has always been. how does one motivate these extremely able people? Pan of the answer is, of course. to hope that one way or another you can arrange it so that they will be doing work that is just a little beyond their capacity. That's the way people grow. I think that when it came to the responsibility I had for the materials side of the transistor. it worked. A good deal of this early transistor work was really development as opposed to physics research. When it comes to research, about all one can do is to recruit very able people in fields relevant to the Laboratories' mission and turn them loose, as is done in a university.

[p140] [Phil Anderson:] "It is hard to realize that in most other fields of science this close collaboration is not accepted. The machines are too big. the specialties too sharply defined. and the whole sociology is hostile to inviting a theorist to think about raw data and experimental arrangements or an experimentalist to propose a theoretical model himself. The idea is. 'A gentleman doesn't interpret another gentleman's data.' I feel strongly that nature is hard enough to understand without inserting artificial barriers. So many theorists refuse to look at experimental facts. on the grounds of some kind of false idea of 'fairness' or of playing science like a game with set rules. Or they insist on using false and sometimes meaningless standards of mathematical rigor to test a theory intended to explain data amassed on fallible instruments or from samples of unknown purity. There is a rigor of physical thinking that is necessary but entirely different from these mathematical games.

"Collaborations are, at least now. easy here at Bell. because the presumption and the atmosphere are one of openness. That is encouraged by the fact that our managers manage. not in the sense of telling people what to do but of knowing what people are doing. After the war Peter Debye. who was a consultant here. ran what were known as the Debye sessions. In those. everyone talked about their work in progress in front of Debye and in front of the directors. There was no way not to be open about one's working progress.

[p146] [Phil Anderson:] There are a few very rare examples where, if you are very lucky, you can start with the material and really calculate everything from first principles. I have done that kind of work. My thesis was sort of like that. But normally you can't do that. A real material contains all kinds of junk. There are too many electrons, and so on, so one has to abstract something relevant from the material. The first time I did that was with the ferroelectrics. I did not realize then that I could do it; I just did it. I am told that I have a mind that thinks quantum mechanically. I have always felt that quantum mechanics is a lot easier than classical mechanics. That feeling goes right back to my orals at Harvard-which I almost didn't pass-where they asked me to explain how the top works in classical mechanics, which I couldn't do. I can do at least the simple versions of the top in my head quantum mechanically.

[p146] [Phil Anderson:] Chemists, for example; are notorious empiricists, and they are capable of living indefinitely with anomalies that stare them [p147] right in the face. People are perfectly capable of ignoring them, or they might puzzle about them for a few years, and if nobody solves them everyone forgets about them. In this sense, they just disappear and get encapsulated. I have seen that encapsulation of anomalies occur again and again in science. They stay there until someone notices them."

Via Rob 8/25/2007