Sciences : histoire orale
Accueil du site > Individus > SCHWEBER Silvan Sam, 2001-08-27


SCHWEBER Silvan Sam, 2001-08-27

Enregistrer au format PDF

Silvan Sam Schweber

Pour citer l’entretien :

« Entretien avec Sam Silvan Schweber », par Arne Hessenbruch, 12 décembre 2000 Sciences : histoire orale, https://sho.spip.espci.fr/spip.php?article128.
— 

Entretien avec Sam Silvan Schweber, par Arne Hessenbruch, 27 aout 2001

Lieu : non précisé

Support : enregistrement non précisé

Transcription : Arne Hessenbruch

Edition en ligne : Sophie Jourdin

ARNE HESSENBRUCH (AH) : Would you tell me what kind of physicist you were, please ? In order to illuminate your perspective on the history of materials science and engineering.

SAM SCHWEBER (SS) : Let me say a little about my background. I started my studies at City College of New York in 1944 as a chemical engineer. I migrated to chemistry, and in my last year at City College I took various courses in the physics department, one in particular with Mark Zemansky. It became clear that I had greater interest in physics than in chemistry, so I went off to the University of Pennsylvania in ’47 to become a physicist. The kinds of courses that people took in those days were atomic physics, a required course offered by Harnwell, who later became the president of the University of Pennsylvania. He had written a book giving an overview of atomic physics. I had a course on quantum mechanics with Ufford, and also one on electricity and magnetism. Walter Elsasser was there, and I did a course on statistical mechanics with him. The person I got closest to was a man by the name of Herbert Jehle, who had an interesting background. He was originally German, a Quaker, who was trained by Schrödinger and obtained his PhD in the early ‘30s in Berlin. He knew Einstein very well and was very much interested in general relativity. Under his aegis I got interested in general relativity, I actually wrote a Master’s Thesis on variational principles in general relativity. He suggested that I transfer to Princeton, which I did in ’49. To complete the story of Jehle : being a Quaker, he got into trouble with Nazi Germany during the 1930s, and left Germany in 1938 to go to France. When the war broke out in 1939, he was promptly interned in Gurs as a German citizen. Eventually the Society of Friends got him out of the concentration camp in France and brought him to the United States, where he taught during the war at Harvard, and then moved to the University of Pennsylvania. He did lots of interesting things. He knew many people at the Institute of Advanced Study at Princeton, Einstein, Weyl, and many of the younger people in physics there, in particular Finkelstein and Wouthuyzen. . He also worked with Pauling on molecular mechanisms of identification in biology. He was an interesting and impressive man. I came to Princeton in September 49, right after the detonation of Joe 1, the first Soviet atomic bomb, at the height of the intensification of the Cold War and in the midst of debates over the hydrogen bomb. John Archibald Wheeler had disappeared and gone off to make hydrogen bombs, first at Los Alamos and later back in Princeton at what became the Forrestal Center. Eugene Wigner, whose research assistant I became, was there, as was David Bohm. The general tone was set by Wigner, who was really very much a phenomenological physicist. He was skeptical about making ultimate and general claims. During the war he had been very much involved in building piles, working with Fermi at the Chicago Met Lab. Actually, for a year before coming back to Princeton, he had been the head of Oak Ridge, trying to get people interested in nuclear power. When he came back to Princeton in ’49, after the advances in quantum field theory by Schwinger and Feynman, he did not get involved in these things. The kinds of topics he encouraged were illustrated by the theses his students wrote. There was a somewhat older student than I, Ed Jaynes, who had written a thesis on ferroelectricity. Another of his students, Ted Teichmann did a dissertation extending Wigner’s R-matrix theory, a generalized S-matrix approach to nuclear reactions. Probably the person at Princeton who influenced me most was David Bohm, whose Advanced Quantum Mechanics course I took during my first year at Princeton. He had just finished writing his Quantum Mechanics textbook. He taught a very beautiful course which took into account the recent advances in quantum electrodynamics. He was very interested in many-body theory at the time. He had had Eugene Gross as a PhD student the year before who had done a thesis on plasma physics. David Bohm had worked on the description of plasmas at the Berkeley RadLab during the War. He had come back knowing a great deal about plasmas. The question addressed was whether you could apply the insights of plasma physics to the description of the valence electron gas within a solid. Eugene Gross worked on the classical theory of plasmas. Another person who was finishing at the time was David Pines. He did the quantum mechanical formulation of plasma theory that eventually became the Bohm-Pines description of electrons in metals, a subject that Wigner was very much interested in, because back in the early 1930s he was one of the founders of solid-state physics with Seitz. Wigner had worked very hard on the treatment of electron correlations. The question was whether the plasma approach could give a simpler way of describing electron correlations — that is, whether a collective way of describing electrons in a metal was a more effective way of describing them than the quasi-single particle approach Wigner had used in the 1930s.

At Princeton physics and mathematics were rather close (Wigner was a professor in both departments). I became attracted to working with Arthur Wightman who was a young assistant professor at the time. This was particularly the case after David Bohm was not reappointed, and I started a thesis in (semi-)axiomatic quantum field theory with Arthur Wightman. Let me say something about the kind of training I received as a graduate student. Princeton was a little special in the following sense. The classes were small, as about 10-12 graduate students were admitted each year. The message conveyed to us was : "we invited you and think highly of you, and we want to give you every opportunity to realize your potentialities". There were no formal course requirements of any kind. Instead, there was a hurdle consisting of a very intense three-day written general examination which covered all of physics : classical mechanics, electricity and magnetism, optics, quantum mechanics, thermodynamics and statistical mechanics, special and general relativity. After that, there were two-three hours of oral examinations overseen by groups of faculty members. The onus was really on us to get acquainted with everything in physics. I had taken statistical mechanics with Wigner, and as mentioned earlier advanced quantum mechanics with David Bohm. I also took various courses in the Math Department - with Spencer, Schiffer and Feller, but in the absence of actual course requirements I just took what I considered useful. There was also the Institute ; after 1947 Oppenheimer was its head. There were joint theoretical seminars every Wednesday, and every theorist at the university would go out to the Institute. During the period I was in Princeton, from 1949 till 1952, most of the presentations were at the frontiers of quantum field theory. During the 50s the Institute was the finishing school for young theorists and all the bright young people trained in the US and overseas would come there. I remember people like Van Hove, Frank Yang, Bryce deWitt, Feldman, Karplus. During my second year at Princeton Murray Gell-Mann and Francis Low were there. I finished my thesis in the spring of 1952 and after I got my degree, I went up to Cornell in the fall of 1952. The atmosphere at Cornell was very different from that at Princeton. Freeman Dyson was there in 1952-53, and of course Hans Bethe. They were in the Laboratory of Nuclear Studies, which housed that segment of the department involved in high-energy physics. The Nuclear Lab had a separate building. There had been earlier tensions between solid state and nuclear/high-energy physicists that had resulted in a rift in the department. What was characteristic of my 1952-54 experience at Cornell was the intimate relationship between theory and experiment, something I had not felt at Princeton. I don’t mean to say that there was no active experimental work going on at Princeton. Dicke was there doing positronium experiments and various other things, in retrospect it seems to me that there were no fora for interaction in the way that existed at Cornell. As a post-doc at Cornell I attended a solid-state class with Overhauser, a young instructor at the time as well as lectures by Bethe on advanced quantum mechanics. Cornell gave me a different vision of what physics is about. One came to realize that high-energy physics was different from solid-state physics - both in approach and in status. It had been a little like that already before the war, then in terms of nuclear and solid-state physics, but this bifurcation became intensified after the war. The differentiation also manifested itself in the requirements on graduate students. What it took to become an experimental or a theoretical physicist differed at Cornell and at Princeton in one respect : during their first two years every graduate student at Cornell took the same classes : mechanics, electricity, magnetism, quantum mechanics, solid-state physics, nuclear physics, experimental physics. Experimental physics, a year-course, was required of every graduate student. Cornell also had a set of examinations, but these were less intense than Princeton’s. At Cornell, just as at Princeton, one was expected to know most of physics. The specialization was limited. Solid-state and high-energy physicists were all still members of the same department, and the curriculum required of all graduate students encompassed all of physics.

In 1954 I went as a research associate to Carnegie Tech in Pittsburgh. The atmosphere there was similar to the one at Cornell. There were some very good people there (the senior theorist was Gian Carlo Wick ; also in the department were Walter Kohn who had migrated from doing quantum field theory with Schwinger to doing many-body solid-state physics , and Julius Ashkin who had switched from being a theoretical physicist and became in charge of the Carnegie Tech cyclotron). I had to leave Pittsburgh after one semester and go back to New York, because my wife had fallen ill. I never returned to Carnegie Tech but accepted in the spring of 1955 a job at the recently established Brandeis - its first class had graduated in 1952. At that time Brandeis was essentially an undergraduate college, but it had a commitment to build a graduate school. In 1957 a graduate program in physics was started. The emphasis was initially heavily on theory because it is cheaper to hire and maintain theorists than experimentalists. In 1958 we started a summer school in theoretical physics that became well known and helped put Brandeis on the map. It was quite useful and ran through the late ’60s. Partly by virtue of the emphasis on theory, the people we hired were predominantly engaged in field theory and general relativity. In 1957 Eugene Gross, who had been a student of David Bohm’s, arrived, doing many-body physics. By 1960-61 it was recognized that we should have an in-house experimentalist program and experimentalists doing solid-state physics (Steve Berko) and atomic beams (Edgar Lipworth) were hired. The examinations to be admitted to candidacy for a PhD in physics at Brandeis were no different from the ones given at MIT or anywhere else in the country. Graduate students were expected to know what mesons and their properties were, even if they were to continue in solid-state physics. The training of graduate students was such that they did not have to make a decision whether they wanted to become experimentalists or theorists, and in what field, until the end of their second year of graduate studies. That continued until the mid-1960s. Entering graduate students mostly would want to become theorists, but if the faculty felt they were not quite strong enough, they would become experimentalists. It is certainly the case that by the early 1960s, one had a sense of sub- disciplines existing, that is that solid-state physics was a different branch from the rest of physics ; and that the same was true for high-energy physics. There were some overlap and mutual interests, for example in atomic physics where precision measurements involved quantum electrodynamics. Brandeis became well known in the late 1960s for doing experiments that determined the Lamb shift in positronium — Berko, Cantor and Mills did that experiment. Positron sources for the exploration of properties of the solid state is what Brandeis became famous for in experimental physics, and that was mostly due to Steve Berko. He had done positron physics before coming to Brandeis and really put positrons on the map as probes in solid-state physics. Until the 1970s support for all the research activities in the department came from two or three places : initially it was mostly the Office of Naval Research, then the National Science Foundation, and then the Air Force. There were no strings attached. You submitted an application and it was looked upon on its merits. The intent of the ONR and the Air Force was to have people trained, to have a pool of scientists available to meet national security needs and national educational and economic interests.

During the 1960s the size of the physics faculty kept on increasing, as did the number of students, and as did the funding by the national agencies. From the mid-’50s on, computers were housed in the physics department and computer science was taught by the department. Max Chretien, who had been a high-energy physicist, had slowly gravitated more and more toward what we now call computer science. By the late 50s, early 60s, the physics department offered several courses in software and in hardware. Computing activities took up an ever-greater role, partly because the high-energy people required it. But it was also simply a way for the department to grow.

By the early 1960s, we started having discussions at senior faculty meetings whether or not to hire a solid-state physicist rather then the brightest and ablest theorist. The demand of sub-disciplines had become assertive. That is the way the department is now run. There are sub-groups, semi-autonomous, quasi-departments. Discussions in the early 1960s still had a common background in questions such as : "what is physics really about ?" We had metaphysical discussions about what counted as more fundamental. By the late 70s and the early 80s, this kind of question had disappeared. It was now about where support can come from. This determined in part who would be hired. Another crucial questions became : "Where would graduate students be able to find employment, and what kind of training should they obtain keeping that in mind ?" It also became clear in the post-Vietnam 70s that only the very best theorists would find employment. And there was then a gradual shift of the balance of the department towards the experimental side and more students interested in experimental physics were admitted. Robert Meyer, an experimentalist who studies liquid crystals, came to Brandeis and started a program in that field. Gradually his group has become fairly large and one of his best graduate students has joined the faculty. Interest in using viral particles to test phase transitions and that sort of thing has become a part of the experimental activities of the department. There is also at Brandeis a research institute in biochemistry, molecular biology and biophysics called the Rosenstiel Center, that functions by having its research staff placed in various departments. Alfred Redfield, an outstanding experimentalist working in NMR, was appointed to both the physics and biochemistry departments. Over the last 15 years the focus of experimental physics has shifted toward biophysics, partly reflecting the fact that the greatest amount of support will be forthcoming in that area, and that graduate students obtaining PhDs in this area can go in many directions thereafter. The newest additions to the department in theory have been condensed matter physicists. There is a woman by the name of Bulbul Chakaborty who is a condensed-matter physicist, doing computational physics. I would say that she is now the leading theorist in the department doing that sort of thing. Two years ago we hired a young condensed-matter physicist by the name of Jane Kondev who has made a big difference. These are people well grounded in modern condensed-matter theoretical physics, and their interests are much wider than those of the high-energy physicists. They have the feeling that theirs is the way of the future ; that young physicists ought to be trained in this field when considering their employment upon graduation. What causes tension within the department is the fact even though there are on the faculty several outstanding high-energy theorists - people like Stanley Deser, Marc Grisaru, Howard Schnitzer - it has nonetheless become more and more difficult over the last 15 years to attract good graduate students. The very best go to Harvard, CalTech or such places. In a nutshell, that is what has happened in the Brandeis physics department. There has been a rethinking on what physics is about and what the mission of the physics department is.

AH : Is it a question of status ?

SS : Also.

AH : The way you described it was in terms of numbers of students. But are you also saying that solid-state physics used to be low status and that this has changed ?

SS : Yes, there is no question about that. I think that Brandeis, a secular university but founded, established and supported primarily by the Jewish community, felt that it had a certain mission. Clearly the conception of the university was such that "pure" contributions had a higher priority than applied ones, and in the sciences this meant theoretical physics and mathematics. Furthermore, as I said it is much cheaper to do theory and by virtue of the initial economic context at Brandeis, it was much easier to hire theorists rather than experimentalists. The initial bias was towards theoretical physics, and since until the mid-1970s high-energy physics was considered more "fundamental" than solid-state physics or what later became condensed-matter physics the theorists were either general relativists or field theorists. It was within the physics community at large that the question of status got rectified. It has to do with what solid state physicists did in superconductivity - using field theoretical techniques, and in particular recognizing the role and meaning of spontaneously broken symmetries in the description of the dynamics of such systems — that similar ideas were developed and then grafted onto quantum field theory in general, and a rethinking of the status of condensed matter took place. Phil Anderson was very influential in this both by virtue of his technical contributions, and because of his article in Science "More is different." There were great theoretical advances in the late 60s and early 70s, in particular, the re-adoption of quantum field theory as the seminal representation of microscopic phenomena, because gauge theories conjoined with the Higgs mechanism for symmetry breaking seemed to be the right way to describe both the strong and the electroweak interactions. This after ’t Hooft showed that one can renormalize gauge theories even if they have broken symmetries This led to what is now called the standard model. But ironically what happened at the same time was a recognition that what had been thought of as renormalization is really something very different, namely, an effective way of putting all the high-energy effects into certain observable parameters at low energy. For the most part you will not see anything of the effects of the high-energy processes beyond the cut-off energy you have introduced to make the theory free of divergences. It implied that even though renormalizable theories still had a special status, renormalization qua renormalization was not quite so important. More fundamentally, it implied that most theories that we know of, be they quantum electrodynamics, electroweak theory, or quantum chromodynamics, are only effective theories. They have a limited range of validity, the range being essentially the masses of the particles, which you need to introduce into the theory in order to explain the data, and which therefore reflect the context (the available energies) in which you are doing physics. People began to recognize that to some extent the theories that condensed-matter physicists used to describe superconductivity, liquid helium, or some such system, was as foundational for its domain of validity as the standard model was for the 0-200GeV range of subnuclear entities - the quarks, gluons and leptons.

AH : But that is an argument that will not be accepted by an older generation of physicists. I mean, this is the kind of thing that Paul Forman argues against, right ?

SS : Well, I would put it differently - and it is not meant pejoratively. Most physicists have become like chemists, in that they are exploring novelty in the world. They are creating new things, or new systems with new properties. And the power of the theory is such that they can actually pinpoint the elements that will give you the kind of qualities or properties that you want. There is still a subset of condensed-matter physicists who see their job as testing foundational issues - for example, "Under what circumstances is describing certain processes as a Bose-Einstein condensation appropriate ? How does this depend on the number of particles involved ?" - However, the foundational theories involved are not questioned. Since the 1970s more and more condensed-matter physicists think that conceiving of nuclei in terms of quarks is irrelevant. An atomic physicist doing experiments that determine the Lamb shift in various atomic states to some amazing accuracy, measurements that indicate structure at a smaller scale, does nothing of great consequence for chemistry, for most phenomena in chemistry do not depend on effects to that order of precision.

There is a further observation. Since the late 1970s the impact of computers has been very great. It is a striking fact about both experimental and theoretical physics that much computer modeling takes place. People are now trained quite narrowly in modeling on the computer, so that computational physics has become a third branch of physics. There are so few things that one can do analytically by virtue of non-linearity that one has to do it on the computer. The results are really quite impressive. The rise of the role in the training of students in computer modeling is really very striking.

AH : Does that have a bearing on the solid-state vs. high-energy physics issue ?

SS : In theoretical high-energy, the impact of computers has been on how to deal with the complexity of perturbative calculations when you start to have 2 to 3 hundred diagrams to evaluate. You have programs where you draw the relevant Feynman diagram and the computer will grind out the finite radiative corrections at given energies or whatever else you are calculating. When you are concerned with real experiments and you have to compute radiative corrections and cross-sections, you people rely very much on computers, the calculations having become so complex. The use of computers among string theorists is much more having access to the web so that you can download the latest pre-prints available from Los Alamos. They log in every morning to learn what new preprint is available. From a sociological point of view the web - and the Los Alamos archive - has democratized high-energy theory : a researcher in Pakistan or Ghana has the information as readily available as the one at SLAC. Also there is a general recognition that physicists need to know a great deal about computing, if only because at the end of their graduate studies they could land a job on Wall Street if they are adept and innovative in modeling on the computer. I think you are right in saying that the older generation still feels that high-energy and elementary particle physics is more "fundamental" than condensed matter physics. I think it is quite painful for my colleagues in high energy theory at Brandeis to consider the possibility that once they retire the university will likely not replace them with someone in their field. They have invested their lives in this field. In high-energy physics and cosmology, where practitioners - still - set their own internal agendas, work gives meaning to lives in a way that is - probably - not the case in condensed-matter physics. For a condensed-matter physicist, solving a difficult and important problem is an impressive thing, and brings about many rewards both personal and communal, including the possibility of a Nobel Prize. The same is true for the high-energy physicists but in addition there is the feeling of being some kind of secular priest, for they believe that they are reading and writing the book of Nature, and formulating something ultimate about the world.

AH : That is certainly the way I perceived it as an undergraduate of physics in Freiburg, Germany, starting in 1978. Pure physics was supposed to be independent of the demands of society. It had nothing to do with the demands, say, of the nuclear industry or the military. That seemed to be the credo, whether it was correct or not. Solid-state physics was very small, and it was openly commercial. The solid-state physics institute actually paid students a salary to write their theses. It attracted some students but it repelled even more. Does that ring a bell ?

SS : In the US, the autonomy of physics departments was greater. There was a system of governmental funding until the end of the Cold War. One did not have to go to industry to get support for your students. In chemistry, funding for research and for students has always come in part from industry. In that field, a faculty member, as the principal investigator, would get a grant or a contract from a particular chemical or pharmaceutical company to carry out specified investigations of great value to that company and in this way obtain the funds to carry out the research and support his postdoctoral fellows and graduate students. In physics, it would not be unheard of during the 60s and 70s for a graduate student to go to Lincoln Lab or Oak Ridge to do his PhD thesis there, but it would have been very unusual to obtain a PhD for work done in an industrial setting.

As an answer to your question : The older generation of physicists might deplore the fact that string theorists lack knowledge of general physics, because their field of activity is so close to mathematics. It is not that they look down upon it, but it is felt that as long as string theory is without empirical relevance it is not really physics. Since the 1980s there has been a great deal of exchange of techniques and knowledge between quantum field theorists and condensed-matter physicists, partly because experimental techniques have become so good that one can devise one-dimensional and two-dimensional systems. The exploration of two-dimensional field theories is thus of great interest to both sets of practitioners. They exchange insights about conformal field theories. But both are concerned with the constraints and relevance of experimental data. String theory thus far has little to say about experimental findings.

AH : Do young people who enter into high-energy physics nowadays also have these - shall we call it lofty aspirations ?

SS : I really don’t know. There is some split. One type are the theorists who try to figure out what it means if the experimentally measured value of the m meson anomalous magnetic moment no longer agrees with the standard model calculations. Does it imply super-symmetry ? String theorists are not likely going to do that. They count states in black holes, so the division is more where you stand in relation to actual experiments.

AH : I don’t understand the historical development of solid-state into condensed-matter physics. Could you explain that ?

SS : Solid-state physics was principally concerned with solids, primarily metals and semi-conductors. Gradually it became clear, partly due to field-theoretical methods, that what you learn about Fermi systems is equally applicable to electrons in solids and gases and liquids composed of He3 at low temperature.

AH : So it was an expansion ? When did this take place ?

SS : In the late 1950s and early 1960s. This is when field theoretic methods became powerful tools in the treatment of many body systems. This was also the time when Bardeen, Cooper and Schrieffer solved the superconductivity problem and when experiments on He3 were being done. You ought to ask Paul Martin or Henry Ehrenreich about this. There was a period in the middle 1950s to early 1970s when these field theoretical methods become standard components of the solid state physicist’s toolkit. And since the field theoretic Green’s function methods used allowed the consideration of the properties of the system as a function of temperature, phase transitions became an essential part of the field. Thus all phases - liquid, gases, solids - became the concern of the field.

Interestingly, there is a parallel development in nuclear physics. After the establishment of the Standard Model, nuclear physicists recognized that they had fallen upon hard times trying to understand nuclei in terms of nucleons, that the challenge was to understand how to go from quarks and gluons to mesons and nucleons. The field theoretical methods were used to understand the properties of nuclear matters, and diffused to such esoteric topics as neutron stars, and the collapse of supernovae. Incidentally, the development in that field in the United States can only be understood if it is remembered that the physicists who became high -energy physicists had been at Los Alamos during World War II and the most distinguished of them populated PSAC (the President’s Scientific Advisory Committee) and setting the priorities for governmental support of the various fields of science. This began to change already in the late 1960s.

AH : Already ?

SS : Look at PSAC : In the 1960s there were almost no chemists on it. The support of science before and after Vietnam was different for many reasons. First of all there was the 1969 amendment of senator Mansfield about funding by the department of defense. Vietnam also marks end of the expansion of physics. The number of annual PhDs in the field began to decline.

AH : Where did Materials Science and Engineering enter into your world ? I imagine that when the interdisciplinary laboratory came into being at MIT in 1960, you must have noticed it. Did it look like physics to you ?

SS : We noticed it. The experimental approach resembled work in physics departments. But it was also like chemistry. My colleague and friend, Eugene Gross, had an interest in the statistical mechanics of polymers. I heard from him about polymers and polymerization. I went to the chemistry department colloquia. In particular, I remember a lecture in the mid-1970s by a polymer chemist from PennState in which he talked about designing polymers with specific electric and optical properties. The use of quantum mechanics to understand not elementary but complex systems and to make it the basis of a "molecular engineering" intrigued me. Later I discovered that Slater had such a vision already in the late 1920s after the advent of quantum mechanics.

In 1973 we had another crisis because of the Arab-Israeli war and the oil crisis. This led to extensive investigations of solar cells - Henry Ehrenreich was head of the committee that was in charge of these activities. He went around to various departments asking what kind of materials are required to obtain such and such an efficiency. By that time materials science conceived in this "applied" way was recognized as a valid field of inquiry.

AH : So it became known to you as a problem, but not as a discipline ?

SS : Right. We did not see these activities as belonging in a department called Materials Science. No, I associated it more with polymer chemistry and with semiconductor physics. I think it came into greater prominence through these solar cells investigations, in these designs for capturing solar power.

AH : We are talking about the mid-1970s now ?

SS : Yes.

AH : And Brandeis does not have a Materials Science department, but much work done there could now well be done in a department with that name.

SS : Correct. Let me put it this way : I have some interest in the limits of computation. How far can you reconstruct the world if I tell you the foundational theory and the entities make it up ? Where does computational complexity put a stop to your efforts ? Every once in a while I go to Harvard to listen to material scientists talking about the design of materials by modeling on computers. What is striking to me is how uncritically they accept the foundational theory as formulated in their computer codes. They just model.

AH : Complexity is in a way also pure vs. applied. If you have the true foundational theory, then in principle you can calculate everything from it.

SS : Yes, until you are stopped by the complexity of the calculation.

AH : Right, and my point is that thinking about the world in this way is tantamount to accepting the notions of pure and applied science.

SS : To overcome the computational complexity you have to cleverly make assumptions simplifying the calculation. Organic chemists designing complicated biological molecules make up little models with hook-like forces and it works very well. The challenge to the theorist is how to justify such assumptions.

AH : I am trying to fit what you have told me to the demise of the pure vs. applied model. For instance in materials science, many people will tell you that the notions are outdated or that they are simultaneously doing both.

SS : I would argue that quantum mechanics differs from everything else in that it gives a highly accurate and stable foundational theory on the atomic level that can’t be tampered with. Heisenberg called it a closed theory. You also know the basic entities populating that realm, i.e. electrons and nuclei. There is a confluence of theory and ontologies. In electroweak theory you can compute the anomalous magnetic moment of the electron to an unheard precision : of one part in 10 to the 10th. You never need that accuracy in materials science and what I am trying to say is that the foundational theory is secure. That is the reason that pure and applied no longer exists. Nobody doubts the accuracy, veracity, efficacy, and efficiency of the foundational theory. There is nothing these people do that probes the foundational theory.

AH : I imagine that an imaginary particle physicist in the 1950s would have said that one ought to find the foundational theory beyond quantum mechanics.

SS : Even though people really felt that the Schrödinger description of atoms is a limiting version of the more foundational theory, you did not have the means to show this. The 1960s provided the means to show that with the re-conceptualized notions of what renormalization means. Not that it is easy, but the foundational theory has been secured. More than that : any foundational theory beyond quantum mechanics must recover all the successes of the standard model : for example, the ability to account for the Lamb shift in hydrogen to the known accuracy, to account for the magnetic moment of the leptons to the known accuracy,...

AH : It seems to me that the meaning of pure has shifted. When materials scientists tell me that they do pure and applied at the same time, they refer to calculating local densities of states and energy levels of an atom moving in a groove of a surface or some such thing as pure. But from the perspective of the old high-energy physicist that is not pure.

SS : Okay. I would phrase it differently : I would call it pure if the intellectual agenda is set by the practitioners without external considerations. Of course, with this definition of purity even high-energy physics - even though the practitioners set their own intellectual agenda - is not pure because it is supported by the government, etc. The same is true for people investigating the onset of turbulence. Clearly there are industrial applications, but that community also sets its own agenda and rewards research very much as high-energy does. It is a question of autonomy. Purity simply reflects the state of autonomy.

AH : So the shift in the meaning over time is interesting.

SS : It is difficult to phrase this - I can make it ad hominem. Your activity as an historian of science gives meaning to your life. You are trying to find out something ; it is personal and there is an intellectual commitment going beyond earning a living. I am not sure that a person at DuPont trying to figure out a better plastic gives much meaning beyond that of monetary reward and job security. There is of course the personal gratification if the challenge is met and the problem solved - and something useful is created - but I believe that success does not answer the quasi-metaphysical questions. So purity has something to do with going beyond what pertains to the economic and sociological sphere. I don’t know how to make it sharper than that. Something about extradisciplinary rewards.

AH : The importance of which has declined ?

SS : In every field. Not just physics or history of science.

Fin de l’enregistrement


haut de page

accueil du site


Haut de page | Accueil du site | Table générale

Contrat Creative Commons

Sciences : histoire orale de https://sho.spip.espci.fr/ est mis à disposition selon les termes de la licence Creative Commons Paternité - Pas d’Utilisation Commerciale - Pas de Modification 3.0 non transcrit.