LEGIBILITY NOTICE A major purpose of the Techni- cal Information Center is to provide the broadest dissemination possi- ble of information contained in DOE's Research and Development Reports to business, industry, the academic community, and federal, state and local governments. Although a small portion of this report is not reproducible, it is being made available to expedite the availability of information on the research discussed herein. BNL-52214 UC-410 (General Pbysics- DOE/OSTI-4500-R75)

BNL—52214

DE90 000868

This Was the Physics That Was:

The Years front P and C Violation to CP Violation

Gerald Feinberg Physics Department

BROOKHAVEN NATIONAL LABORATORY Associated Universities, Inc. Under Contract No. DE-AC02-76CH00016 with the Department of Energy DISCLAIMER This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, nor any of their contractors, subcontractors, or their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency, contractor or subcontractor thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency, contractor or subcontractor thereof.

Printed in the United States of America Available from National Technical Information Service U.S. Department of Commerce 5285 Port Royal Road Springfield, VA 22161 NTIS price codes: Printed Copy: A02; Microfiche Copy: A01 Introduction

As part of the joint Brookhaven National Laboratory-Columbia University celebration of the 1988 award to Leon Lederman, Mel Schwartz and Jack Steinberger, a special lecture was held at Brookhaven on February 17, 1989. Gary Feinberg of Columbia recreated for us the status and the style of at that time, and showed how it led to the discovery of the second , to new generations of - and to another kind of generation - a generation of fundamental investigations using the unique probe of high energy neutrino beams. In order to introduce such a talented speaker, we felt that we would need at least two very capable people. And who could do this better than Maurice Goldhaber of Brookhaven and T.D. Lee of Columbia? The whole event turned out so well, we decided to publish the lecture, and here it is. I hope you enjoy reading it as much as we enjoyed hearing it.

Nicholas P. Samios Introduction Maurice Goldhaber On behalf of Brookhaven National Laboratory I Dino Goulianos (now at Rockefeller) and Nari am delighted to welcome back the three men Misfry (now at Cornell). who recently reached a pinnacle of scientific acclaim by sharing the 1988 Nobel Prize for This prize is an honor for Brookhaven Physics: Leon Lederman. Mel Schwartz, and National Laboratory. Our laboratory was Jack Steinberger. At the time of their great dis- started soon after World War II with a new con- covery, which was made at the AGS in 1962. cept in mind. It would have facilities that were they were Professors of Physics at Columbia. not usually available at universities, and at They were cited by the Nobel Committee "For which university researchers would be wel the neutr.no-beam method and the demonstra- come. The guard at the gate was told to lei in tion of the doublet structure of the anyone with a good idea! One of the most through the discovery of the neutrino." important facilities built was the AGS. and the prize also honors all those people who were In the past, the Nobel Prize has often been responsible for its success. This magnificent given either for answering an important ques- machine, where dreams have often been turned tion or for opening up a new research field. But into reality, was built by many people. Let me in this case, it is noteworthy that both reasons single out just a half dozen for their special were cited. contributions: Ernest Courant, Stan Living- It has been a long journey from the old ston, and Hartland Snyder, who discovered the Brookhaven cafeteria to the table of the King of principle of alternating gradient focusing, and Sweden! One could conjecture, that one contri- Lee Haworth, Ken Green, and John Blewett, buting factor for the delay — though admit- who took the lead in building the AGS. tedly this is not a record-setting delay — lies in the nature of this discovery. It is unlike some In his blunt way, Isidor Rabi often asked me: other discoveries which also opened up new "Why don't those three get the Prize?" This fields of research and for which the Prize was occasion would have pleased him greatly, as given as promptly as possible. For instance, x acclaim both for them and for the two places rays, for which the first on which he left such an indelible imprint, the was given in 1901. and high temperature Physics Department at Columbia and Brook- superconductors, for which the Prize was given haven. It is sad that he did not live to enjoy our just a year before this one. were each discover- celebration. ies which led immediately to an explosion of The path to this discovery was partially research. The /z neutrino experiment, however. paved by the insight of someone I call the was a hard act to follow. There are few places second Pauli, or the Pauli of the second neu- where this type of research can be carried out. trino. Gary Feinberg, from whom we shall hear The experiments are difficult and time con- later. While Gary was a research associate at suming, and it took many years for reproduci- Brookhaven in 1958. he pointed out that it ble, significant new results to emerge with the would be difficult to understand why the beam method. Let me mention just did not decay readily into an and a y two important results: the discovery of neutral ray if there were only one kind of neutrino. He currents, and the determination of the Wein- had to use a subtler argument than the first berg angle which have helped make the electro- Pauli. who could rely on such hallowed conser- weak theory generally acceptable. vation laws as energy and statistics. At Brook- The three musketeers, as our laureates have haven and elsewhere the feeling was in the air often been dubbed, as well as those who collab- that "/j-ness" was conserved, and that the orated with them, are all here today. Their col- muon was trying to tell us something more laborators were our own Gordon Danby. Jean- general. Gary pursued his work with Steve Marc Gaillard, then an exchange-visitor at Weinberg, and then T.D. Lee and Frank Yang Columbia from France (now at Orsay and strengthened the argument for a second neu CERN). and two Columbia graduate students. trino further. Lee Haworth, as Director, had a steering whole world quickly learned of this magnificent committee for the Cosmotron consisting of discovery. A new conserved quantum number most high energy at Brookhaven was born, now called Flavor. with one outsider. Bob Serber. About the time when I became Chairman of the Physics But physicists cannot leave well enough Department, in May 1960, Lee Haworth also alone. There is a trend to question conserva- appointed me Chairman of a high energy advi- tion laws, or at least to test the limits of their sory committee, which would now have to con- validity. One of the important research fronti- sider experiments not only at the Cosmotron ers of today is the effort to find whether there but also at the AGS, which was then nearing are some unknown interactions which can completion. Rod Coo] became Secretary of the turn into , or muon Committee. A few original members of that into electron neutrinos. Brookhaven is in the Committee are here today: Ralph Shutt, Ronnie forefront of these searches and researches, and Rau and two university representatives. Bob if the three musketeers would like to join the Serber and Bob Adair. Our way of sifting prop- fun and make another proposal they can be osals for accelerator experiments became a assured that the present committee is kinder worldwide model as did many other Brook- and gentler; they even vote. But I should warn haven contributions, organization, al, techni- you that made a suggestion cal, and scientific. We now compete cheerfully long ago that we should put up a sign at the with our intellectual offspring. AGS: Only one Nobel Prize to a customer. I never got around to it, but perhaps Nick will do The Columbia proposal for the search for a so. second neutrino was approved by the advisory Let me close with a few remarks about our committee. Now let me tell you a story which laureates. After leaving Columbia they have the three musketeers may still remember. In had continued successes. the spring of 1961, while Trude and I were on a Nordita Lecture tour, they conveyed to me, Jack Steinberger, in his wisdom, has reached through Rod Cool, the urgency with which they the formal retirement age at CERN without ever wanted to do their experiment, suggesting that having been an administrator with a capital A. they use the only available beam, which led to He is still very active in research, and is com- the : if I would allow pleting a huge detector for the new CERN the bubble chamber to be taken out for six electron- collider, LEP. He calls his weeks, they would finish their experiment and detector Aleph. In this century, the nuclear and the bubble chamber could return. As much as I particle physicists have essentially used up the sympathized, I felt that 1 had to save them from Greek and Latin alphabets, partly due to their youthful enthusiasm. I was skeptical of Brookhaven discoveries. Perhaps Jack turned their promise to finish this important experi- to the Hebrew alphabet to give himself room for ment in six weeks, and insisted that they new particles which he may discover at LEP. would have to wait for their own beam so that they could do their work in whatever time it Mel Schwartz achieved something which took. I therefore assigned a beam for their sole may be as rare for physicists as a Nobel Prize: use. In the end, it took them eight months to He became his own boss, a successful entrepre- accumulate enough data to demonstrate neur. He runs a company which he has imagin- clearly, in a beautifully executed experiment, atively named "Digital Pathways", Inc., located that there was indeed a second neutrino. Dur- in Mountain View, California. This is one way ing that time, the AGS had met all the stringent to escape High Energy Advisory Committees. demands. Incidentally, Nick Samios reminds Leon Lederman became the Director of Fermi me that there was a bonus from two independ- Lab, and the place is humming. He has become ent beams: In the uninterrupted bubble such a well-known TV personality that my chamber run, he and his group found the as Secretary first spelled his name with two t's. well as the E *, so important for SU(3). Brook- What is worrisome is the prospect that he will haven had a well-attended press conference to soon vie for the title of "Funniest Director announce the arrival of the p. neutrino, and the Emeritus". Introduction T.D.Lee

Very rarely do we encounter an experiment theory, developed by one — the super, which is so spectacular at the time of its execu- super, super physicist. tion and which later on becomes even more The oneness is good in dreams, but hard to important with time. When that happens we realize. So we may embark on a quest for two. know we have met a "true great" and we all feel In the thirties we had as strongly interacting privileged. I think the award of the Nobel Prize particles the and . And then to this discovery of two neutrinos, twenty-six there was another group,of the light particles, years after the discovery, is a clear tribute to electron and neutrino. Each group has two the greatness of the discovery. members: one charged and one neutral. This was pre-World War II. Then after World War H, There are perhaps three aspects which I may the muon came, so that made a family of three. emphasize. One. of course, is the discovery Then with cosmic radiation and the Cosmo- itself. First, one observed the ju-neutrino; and tron, came the A. And that stimulated the by that same observation, one also discovered Goldhaber model and the SU3 theory of Sakata. the e-neutrino and thereby resolved a paradox: which in turn then evolved to the SLT3 of the absence of the ^ — e + y decay. But there is Ohnuki, GeU-Mann and Ne'eman. So that a second element: this particular experiment became three. was the beginning of the high energy neutrino We then had two groups. Each had three program. It launched a brand new direction members. For the , proton, neutron, and changed the course of particle physics for or, in terms of more basic units, up, down and more than twenty-five years. It also added to strange . The other three are leptons: the massiveness of detectors. The neutrino neutrino, electron and muon. It was the discov- experiment was 10 tons. If you plot the weight ery of two neutrinos that made the leptons into of detectors versus time, I think you will find four. As you all know, four is not divisible by that the curve has a slope discontinuity in three and that made a fundamental impact It 1962. Since then it has risen from 10 tons to took more than a decade to realize that the the multi-kilotons of LEP detectors. only way to make two and three commensurate I think there is a third element for this dis- is to have six — three families with two covery and that is its impact on the quest for members each. That is the state that we are in the pattern of fundamental building blocks. today. What we all like is a quest for one — we live in It is a great pleasure for me to join with all of one world, and there is only one universe. We you in paying homage to one great laboratory, would like everything to be made of one funda- Brookhaven, to two elusive particles, the e- mental unit, maybe the super string, which is neutrino and mu-neutrino and to three superb governed by one theory, maybe the super, super physicists. This Was the Particle Physics That Was: The Years from P and C Violation to CP Violation

This talk is meant to evoke something of what of striking differences are immediately particle physics was like in the period leading apparent. up to the first high energy neutrino experi- 1. Experiments in particle physics in the late ments. I will therefore give the talk in the mode 1950s and early 1960s were done by what that was then universal, i.e.. I will not use would now seem like minigroups of physicists. transparencies. Figure 1 shows the number of authors of each The period that I will speak about is not liter- experimental physics paper published in Phys- ally that described in the subtitle, which would ical Review Letters in the second halves of begin in late 1956, but rather begins about a 1962 and 1988. The plots hardly overlap at all. year later, after the space-time structure of the The averages were six authors per paper in weak currents had been settled, on the one 1962, and forty-seven per paper in 1988. There hand by the theoretical work of Richard Feyn- is even one experimental particle physics paper man and Murray Gel) Mann, Robert Marshak in 1962 with a single author! (named Burt and George Sudarshan and John Sakurai, and Richter). The two-neutrino paper1, which had on the other hand by the experiments of Lee seven authors, was fairly typical, in this Grodzins, Andy Sunyar and Maurice Goldhaber respect, of papers published in 1962. In 1988. on the helicity of the neutrino emitted in beta several of the published papers have more decay. institutions involved than there were authors I want to avoid a common trap and so I will of most of the 1962 papers. Even the size of the not refer to the period in question as the typical experimental groups in 1962 repres- 'Golden Age' of particle physics. To paraphrase ented a major expansion compared to the what someone once said in another context, situation twenty-five years earlier, when it was the Golden Age of physics is when you were common for a single scientist to conceive an twenty-five years old. I was twenty-five at the idea for an experiment and bring it to realiza- beginning of the period I will discuss, but I am tion within a few days or weeks. This earlier content to leave it to historians to decide more situation is well described in a recent memoir objectively the merits of the accomplishments by Maurice Goldhaber. Thus within the work- of physicists in different time periods. I think ing lifetime of many of those present today, that when I am finished you will agree that the experimental particle physics has gone from an particle physicists who worked in the years individual effort, to an effort by a small group from 1958 to 1964 had a great deal to be proud of close collaborators, to a team effort by scores of. I was a postdoc at Brookhaven from 1957 to of individuals, to what, at the next generation 1959 and a faculty member at Columbia for the of accelerators, may require the combined rest of the period, two institutions where much efforts of whole countries of physicists. of the action 1 will describe took place, so I had 2. Particle physics was a much larger fraction a pretty good chance to observe what w^us going of the total effort in American physics in 1962 on. Of course, many of the trends that occurred than it has become lately. The domination by are much easier to observe through hindsight condensed physics had not yet taken than with direct vision. hold. On comparing the publications in Physi- I will begin with a few general remarks about cal Review Letters in a similar number of the sociology of particle physics then and now. issues in 1962 and 1988,1 found on the one Looking back on those days and comparing hand in the earlier year forty-nine papers were them with what has happened since, a number published on experimental particle physics and 25

1962 PR L 10

i i i i i i i i i i t i i i i i i i i i i i I i i i i i i i i i i i i i i

Figure 1. For each paper in experimental particle physics published in July-December of 1962 and 1988 in Letters, the number of authors is plotted on the ordinate. The scale for the 1962 papers is given on the right axis, that for the 1988 papers on the left. The abscissa is used only to order the papers by number of authors. an approximately equal number of papers were at least insofar as this can be measured by the published in experimental condensed matter number of papers published. In thirteen issues physics. On the other hand, in 1988, only eight of published in 1962. papers were published in experimental particle there were twenty-seven papers in theoretical physics and over a hundred were published in particle physics, and forty-nine papers in experimental condensed matter physics. The experimental particle physics. In 1988, in the priorities of American physicists have evidently same number of issues, there were thirty theo- changed considerably over this period. retical papers and eight experimental papers. The ratios of theory to experiment for 1988 are 3. A third sociological change that has taken about the same for papers in Physical Review place over this period involves the broadening D, Physics Letters B and B. If of participation in experimental particle phys- I add together all the current letter journals, ics. The forty-nine papers that I just mentioned then Jhe total number of experimental publica- as published in the second half of 1962 tions in 1988 is about the same as in 1962, but involved authors from twenty-five US institu- the number of theoretical papers is seven times tions, and seven foreign institutions, and the greater. I believe that this rapidly diverging majority of the authors were from just a few production of theoretical papers is an indica- places, such as Brookhaven, Lawrence Radia- tion of a disturbing trend in our field, about tion Laboratory, Columbia, and Cornell. For the which I will comment later. eight papers published in 1988, however, the authors came from thirty-five US institutions What We Knew in 1960 and twenty-six foreign institutions, and there was no trend towards dominance by any small I turn now to the actual content of particle group of institutions. physics in 1960. To give some idea of the state of knowledge in particle physics in 1960, Table 4. Finally, there has been a pronounced shift 1 lists the 'elementary particles' known at that from experiment to theory in particle physics, time. there are now as many, if we include color, as Table 1. The particles known and considered to there were known hadrons in 1960. The gauge be elementary in 1960. have had the most spectacular increase in number, rising to thirteen, if we include the Leptons electron e-neutrino yet unobserved . muon mu-neutrino ?? In 1960 as today, much of what we knew 7r-plus. 7r-zero. ir-minus K-plus. K-minus K-zero, K-zero could be summarized in terms of conservation bar laws, which came in two forms, those thought proton, neutron to be exact, and those thought to be at best lambda-zero approximate. Again, the lists have undergone sigma-plus, sigma-zero. sigma- some profound changes. Only one new exact minus conservation law has been discovered, that of xi-zero. xi-minus color, whose influence is muted by the fact that all forms of ordinary matter have zero for the graviton ?? conserved quantity. One conservation law. that Gaugons of CP, then thought to be exact, has definitively been moved to the approximate list. Several There were three known leptons, with suspi- others, including number and the cions of a fourth. Among the hadrons. there number of each of the types of leptons are were seven meSons. with hints of many more thought to be only approximate, but evidence that were about to be discovered. There were is lacking. It is also unknown whether the total eight baryons, and again indications that number is exactly conserved. many more were to be found. In fact, many of the vector mesons and excited baryons were As for the approximate conservation laws, we discovered in the two years after 1960. In addi- have found a few more, corresponding to the tion, there was one known gauge particle, the new hadronic quantum numbers of charm, photon, another firmly believed in, but unob- bottomness and presumably topness. These, served, the graviton, and suggestions of others, together with uppityness, downiness and such as the W particle. strangeness are all conserved in strong interac- tions. But our attitude towards the breakdown The passage of time has not been altogether of these conservation laws in weak interactions kind to what we thought we knew in 1960 has undergone a subtle change. In weak inter- (Table 2). actions, half of these conservation laws break down, because of the interaction of the quarks with the W , which causes transitions Table 2 The particles known and considered to within the quarks of one generation. But for be elementary in 1989. another aspect of the breakdown of approxi- mate conservation laws, that involving inter- Leptons electron e-neutrino generational transitions such as K muon mu-neutrino decay, we now believe that such decays are one tauon -neutrino manifestation of a mismatch between the Quarks up down states with definite and those charmed strange with definite weak interactions, a notion top?? bottom already introduced by 1960 as particle mixing. Gaugons photon The precise reason why this mismatch should graviton ?? occur remains elusive, and is presumably W Z related to the origin of itself. The 'armament' of theoretical particle physi- cists in 1960 was much more restricted than it The known leptons have survived as elemen- later came to be. We could do perturbative cal- tary particles, supplemented by their heavier culations in renormalizable field theories, cousins. The hadrons have been erased as ele- although with rare exceptions these were not mentary, and replaced by the quarks, of which carried out beyond one or two loops. While the 8

renormalization group had been introduced such a gauge theory, given mass by means some years earlier, little use had been made of unknown, with the physical vector mesons it to break out of perturbation theory. whose existence was being suggested by The set of known renormalizable theories electron- scattering experiments. These was not large, including only QED, non- suggestions eventually seemed to be borne out derivative couplings of spinless mesons to by the discovery of the p meson, but as we now , and 04 self-coupled meson theories. know, the connection of the p meson to gauged While non-Abelian gauge theories had been isospin turned out to be an illusion. What described by Chen-Ning Yang and Robert Mills seemed like an even wilder possibility was sug- some years earlier, little use had been made of gested in unpublished work by Feza Gursey them, because it was not known how to give and myself in 1959. and probably by others as the quanta finite mass, and this was deemed well. This was a unification of the ideas of Yang necessary in order for the gauge theories to and with those of Yang and Mills. describe anything physical. For the same rea- The known spin-zero mesons were all taken to son, it was not yet known that such theories be composites of baryons and antibaryons, could be made renormalizable. Indeed, various bound by forces generated by the exchange of proofs had been published to the effect that the Yang-Mills isospin gauge particles. they were not. Nevertheless, there were some Although this idea contains the germ of the attempts, notably by and by ultimately successful color SU(3) theory of had- Sheldon Glashow to use non-Abelian gauge rons, too many of the crucial ingredients were theories to unify weak and electromagnetic missing for it to have been taken seriously, interactions, culminating in Glashow's 1961 even by Gursey and myself. Both the notion of article2 in which the SU{2) X U( 1} gauge group confinement, which allows us to have quarks was first used. and color and yet not see them, and the notion of asymptotic freedom, which allows us to do The basic ideas that ultimately made the calculations with confined quarks, remained possible were not yet clearly ideas for the future to bring forth. recognized in 1960. Spontaneous symmetry breaking and degenerate vacua had been dis- cussed already several years earlier by Werner The Discoveries of the Early 1960s Heisenberg in the context of his non-linear The period from 1960 to 1965 was an theory of everything', but had met with a frigid extremely fruitful one for particle physics, reception from the reigning priests of quantum although at the time it did not always seem to field theory. 's effort to find an be that way. A series of experimental discover- analogy in particle physics to the symmetry ies, and a somewhat unrelated series of theo- breaking that takes place in the BCS theory of retical ideas laid the groundwork for much of superconductivity seemed to run afoul of the what has come to be known as the standard Goldstone theorem. Schwinger had raised the model. Apart from some of the work that led up possibility that this theorem might not be rele- to the two-neutrino experiment, I will describe vant to gauge theories, but ' specific these discoveries only very briefly, as each of realization of that possibility was yet several them by itself could well be the subject of a talk years away. like this one. Although the notion of higher symmetries of strong interactions had been discussed for sev- 1. The Filling in of the Particle eral years before 1960, there was not yet any Spectrum. clear indication that these ideas were relevant Perhaps the most striking discovery of the to physics. The large mass difference between early 1960s was the recognition that there strange and non-strange particles seemed to were hundreds of hadronic particle states make useless theories that placed both types beyond those known previously. These new into a single multiplet. The only clearly valid states did not carry new internal quantum symmetry for strong interactions was isospin. numbers; instead they were similar to pre- Already, there had been an attempt by Sakurai viously known particles, but with greater mass, to 'gauge' isospin and to identify the quanta of and sometimes with higher spin. Although some of these, such as the p meson had been main reasons why the discovery of CP violation anticipated, most of them were unexpected, came as such a surprise. One is that after the and certainly nobody expected new particles to discovery of P and C violations, physicists appear in such numbers. became comfortable with the proposal by , and by Lee and Yang that the 'true" An obvious consequence of the discovery of mirror symmetry was CP invariance. To become the many new hadrons was that they made stripped of that remainder of an earlier time possible a precise determination of the correct when space-time symmetries were all exact approximate symmetry of the strong interac- seemed to plunge us into uncharted territory, tions. While there were a number of candidate where there was no real guide to how we symmetries that described the previously should proceed. The other reason for the sur- known quasi-stable hadrons. only what we now prise was that the extent of CP violation, as call flavor SU(3) survived the test of classifying measured by the ratio the excited baryons and mesons. This line of reasoning culminated in 1964, when Nick R = KL - 2TT / Ks - 2TT Samios and his coworkers at Brookhaven 3 was so small. Indeed, several years before the found the first ii particle , with the properties Fitch-Cronin result, an experiment by Leon predicted for it by SU(3). Lederman and coworkers had put an upper But in spite of this great success. I think that limit on the ratio R which was not much larger the first main effect of the discovery of the mul- than the actual value found in 1964. Again, the tiplicity of hadrons was psychological. It put an experience with , where the observed vio- end to the idea that the old familiar particles lation was maximal, had convinced us that such as the nucleons and the were some- arbitrary small coefficients did not appear in how the basic constituents of matter, and physical phenomena. opened the way for the present notion that the forms of matter that are best known to us may I think that it was the combination of these be only incidental aspects of a much richer two reasons that initially made many theorists subatomic universe. Of course, that is an idea grope for alternative explanations to CP viola- which may not have yet reached its final frui- tion for the decay KL — 2TT. None of these tion, as we are probably many orders of magni- turned out to be right, and after a few years, tude away in energy from probing the final everyone was convinced that CP violation was a secrets of matter. With the removal of the idea fact. that pions and nucleons were somehow funda- The reason for CP violation is another mat- mental, the way was opened to look for some- ter. At the time of its discovery there were var- thing that was more fundamental, and in a few ious hypotheses put forward about the origin years Gell-Mann and George Zweig responded of CP violation. Most of these were soon dis- with quarks and aces. proven experimentally. One of the survivors was Lincoln Wolfenstein's 'superweak' interac- 2. The Discovery ofCP Violation tion model, which attributes CP violation to a It is just about twenty-five years since Val new interaction that is smaller than the weak Fitch. Jim Cronin and their coworkers sur- interaction. This model, which essentially prised the world of particle physicists by their attributes CP violation to mass mixing between discovery4 that the long-lived neutral K-meson K° and K°, predicts that the branching ratios of decays into two pions with a branching ratio of KL and Ks into different In modes should be about 10 3. This experiment, like the two- equal. neutrino experiment, was carried out at Brook- It was recognized by Makato Kobayashi and haven, and again like the two-neutrino experi- Toshihide Maskawa5 that another viable model ment involved what by present standards of CP violation existed if there were three would be considered a minigroup, containing generations of quark doublets. In the KM only four investigators. I think that it is fair to model, CP violation is attributed to the occur- say that this discovery was unexpected, rence of a complex phase in the matrix that although it is conceivable that some theorist describes the mixing of the different types of had written a paper predicting it. There are two quarks. Such a phase can be eliminated for two 10 quark generations, but not for three or more. amounts of material as the neutrino detector, The KM model generates CP violation both using a figure of ten tons. Even with this large through induced K°-K° mass mixing and detector, relatively low event rates are to be through an an additional term of direct CP vio- expected, and Schwartz calculates one event lation. The latter term seems to have been per hour for each 10 tons of detector. observed recently in experiments at CERN and It is very interesting to compare these Fermilab. which show a difference between the numbers with the actual numbers in the two- 2TT° to 7T~TT branching ratios of K and K . L s neutrino experiment carried out two years It should be emphasized that even though it later1. The experiment was done at the Alter- appears that what we know of CP violation is nating Gradient Synchrotron, which then pro- well parametrized by the KM quark mass mix- duced about 3 X 10' > /sec, at a some- ing matrix, this does not mean that we have yet what higher energy than Schwartz had understood the fundamental origin of that vio- considered. The detector was a ten ton spark lation. In fact, we do not yet really understand chamber. The use of a spark chamber was not how any aspect of the quark mass matrix specifically envisioned by Schwartz, as I believe arises. Gaining that understanding is the main that none had been successfully operated in intellectual argument that has been given for 1960. The running time for the experiment the construction of the Superconducting seems to have been about three hundred Super Collider. For example if. as many people hours, and produced about one hundred expect, CP violation arises through some spon- events, of which about one-half were used to taneous symmetry breaking among an draw conclusions. Thus the results of the first extended set of Higgs mesons, then there is high-energy neutrino experiment conformed every reason to expect that at energies much very closely to the suggested parameters des- higher than any we have yet explored, the patt- cribed by Schwartz. ern of CP violation is much more complicated In view of that agreement, I was quite sur- than anything that can be described by a sin- prised when I recently reread Mei's paper, and gle complex number in the quark mass matrix. found that after giving the numbers mentioned In particular, as pointed out by Scott Dodelson, for a proposed experiment, he wrote the follow- most of the models that have been put forward ing: "This estimate places the experiment out- to explain baryogenesis in the early universe side the capabilities of existing machines by require many more complex parameters in the one or two orders of magnitude."6 He then goes Higgs interactions than appear in the KM on to discuss the possibility of doing it at sev- matrix, so that we cannot use what is known of eral proposed machines, most of which were CP violation at low energies to calculate the never built. He does not specifically mention baryon asymmetry. the AGS. which was then a few months away from operation, but I imagine that he was 3. The Discovery of the Muon Neutrino including it among "existing machines.' Why a. The New Experimental Possibilities then was he so pessimistic about an experi- Several distinct streams flowed together to ment that was successfully done within about form the experiment that discovered the muon two years? neutrino. By far the most important of these I should let Mel answer that question, but let streams was the recognition by Mel Schwartz me offer one bit of speculation. I have already that experiments with high energy neutrino mentioned the gradual change in the sociology beams were possible. In his first paper6, pub- of experimental particle physics over two lished in Physical Review Letters in March of generations. Part of this change was the widen- 1960. he clearly laid out the directions of ing of physicists' time horizons, from consider- future experiments. These included the use of ing only experiments that could be done in a high energy beams as the source for the few days, to the present situation, in which neutrinos, so that a 3 GeV proton beam of I012 physicists blithely plan on experiments that protons/sec would generate a flux of about will be done ten years from now. by graduate 5000 high energy neutrinos/cm2/sec at the students presently still in high school, on target. Schwartz discusses the need for large machines that are still to be funded. Mel's 11

paper was one of the first, if not the first, to My own involvement with neutrinoless muon expand that horizon to periods of several years. decays began in a quite indirect way. I spent With that change in the time horizon, it the year 1956-1957 as a Postdoctoral NSF Fel- became much more difficult to estimate future low at the Institute for Advanced Study in Prin- possibilities, and it was perhaps prudence on ceton, where I shared an office with Pasha Mel"s part to be cautious rather than overly Kabir. One day. while waiting at the train sta optimistic about what could be done. Evidently, tion, we met Robert Oppenheimer. then the his collaboration with Leon Lederman. Jack Institute's Director. (Pasha remembers the con- Steinberger, Gordon Danby, Dino Goulianos, versation as instead being in Oppenheimers Jean-Marc Gaillard and Nariman Mistry was office). He mentioned to us a recent report able to squeeze the extra one or two orders of about a bubble chamber experiment at Berke- magnitude out of an "existing machine." (Mel ley by Luis Alvarez and coworkers. which des- tells me that the reason for his written pessim- cribed events in which a negative muon ism about the immediate prospects for high stopped, and a short time later was ejected at energy neutrino experiments was that the high speed. Oppenheimer thought that this design intensity of the AGS was only 109 prot- observation might be interpreted as implying ons/sec. Thus it was the fact that the actual that there was both a heavy and a light muon. intensity of the AGS beam was much higher and that Alvarez was observing a radiative than anticipated that allowed neutrino experi- transition between them. He proposed that ments to be done there.) Kabir and I should calculate the expected tran- sition rate. Actually, Alvarez had discovered not b. The Test of a New Conservation Law a heavy muon but rather muon catalysis of fusion. Nevertheless. Kabir and I set about The second stream that led to the two-neutrino looking into radiative transitions between two experiment was the recognition of a need for a of different mass. leptonic conservation law in addition to that of total lepton number. This need arose from the Our effort led to a result that surprised us fact that neutrinoless conversions of muons very much. Instead of coming out with a finite into electrons were not observed to occur. For decay rate, we ended up with zero! What we did example, an experiment7 by Steinberger and for the calculation was something that later his student Subramanian Lokanathan in 1955 came to be known as mass mixing. That is. we limited the branching ratio for the decay took an interaction of the form: fi — e + y) to less than 2X10 5. The idea that the non-occurrence of neutri- noless /a — e transitions could be understood This interaction, together with the usual by the assignment of different quantum electromagnetic interaction, generates two numbers to muon and electron had been put Feynman diagrams for the decay FH — FL + y . forward by Hormoz Mahmoud and Emil Kono- But a simple calculation shows that the contri- pinski8 in 1953, in the context of a single four bution of these two graphs cancels, giving no component Dirac neutrino. That proposal was decay amplitude. One does no better by taking later taken up by Schwinger9 and by Kazuhiko a derivative interaction Hamiltonian and intro- Nishijima10 in 1957. Nishijima also indicated ducing electromagnetism by the minimal inter- that this assignment would forbid the neutri- action rule nos from IT —fi decay from producing electrons. H, = H.C., What was missing from these early sugges- nt tions of the need for a new conservation law a procedure that came to be known as current was any indication of how significant a prob- mixing. lem the absence of neutrinoless muon decay For this case there are three Feynman was. After all, one could readily have taken the graphs, which again combine to give zero decay attitude that it was not a conservation law that amplitude. A few years later1', Steve Weinberg, forbade such decays, but rather the absence of Pasha and I proved a general theorem that a specific interaction that would allow it to such mixing could always be transformed away occur. and that it gave no transitions, so long as all 12

the other interactions of the two species of fer- dered in the IVB theory. It was known from mions were symmetric. work in the 1950s that a theory with massive The conclusion that we drew in 1957 was vector bosons, interacting with and that to generate electromagnetic transitions leptons. was not renormalizable because of the between two fermions, something more than high energy behavior of the longitudinal terms simple mixing was required. It would be neces- in the propagator. Nevertheless. I sary to involve some internal structure of the felt that at least some estimate of the rate for fermions, such as that induced by the higher M — e + y could be obtained by calculating the order effects of weak interactions. But at that Feynman graph and so I set about doing it in time the Fermi theory was generally accepted, early 1958, when I was a postdoc in the Brook- and while that theory would generate a decay haven Physics Department. amplitude for n — e + y , it would appear only 2 The calculation was reasonably straightfor- in order GF , which we then thought would ward, although it was complicated by the fact make it insignificant. that the answer was logarithmically divergent, The situation changed considerably within a and some gauge invariant method was needed year. The propcsal of a universal V-A weak to produce a definite answer. In doing the cal- interaction, by Feynman and Gell-Mann, Mar- culation, I was greatly helped by advice from shak and Sudarshan, and Sakurai, and the the late Ralph Behrends. also at Brookhaven rapid indication that their theory was consist- then, who was one of the few people who had ent with experiment, reopened a possibility calculated radiative corrections to weak inter- that had been suggested some years earlier, actions. Since dimensional regularization had that the known weak interactions are mediated not yet been invented, I used Pauli-Villars regu- by the exchange of a vector boson, which soon larization. Another complication was that we came to be known as the W particle. While this did not really know what the proper electro- intermediate vector boson (IVB) hypothesis was magnetic theory was for the charged vector consistent with all known observations, it was boson. The logarithmic divergence that I men- not seen to have any immediate consequences tioned occurs for the minimally coupled theory, that could distinguish it from a theory in corresponding to a boson with a magnetic which the was a sum of pro- moment of one magneton. I remarked in my 12 ducts of various currents. However, one such paper that if the boson was given a moment consequence soon suggested itself to me. of two magnetons, the answer came out to be finite. I believe that this is the value that cor- responds to the Glashow-Weinberg-Salam the- ory, so that if that theory had been available at the time, there would have been a unique pre- diction for the fi — e + y rate. The result of both this finite calculation12, o l_ w and of the calculation with any reasonable value of the cutoff gave a branching ratio for p—e + y compared to ordinary muon decay of about 104, which was considerably above the existing limit of Lokanathan and Steinberger. Subsequent searches for ii — e + y over the Figure 2. A Feynman diagram contributing to next few years pushed the limit down by sev- the decay n — e + y in the IVB theory. eral orders of magnitude. This showed a clear In the IVB theory, the decay n — e + y should contradiction with the [VB theory under the assumptions that had been made. occur to first order in GF, through the pictured Feynman graph (Figure 2), This would mean My own view at the time was that this result that the rate for fi — e + y should not be sup- showed there was something wrong with the pressed compared to that of ordinary muon IVB theory. In fact in a letter that I wrote to my decay by any powers of GF. However, this was a then current girl friend I paraphrased a remark loop graph, the first such that had been consi- made by Joe Louis about someone he was 13

going to box. Billy Conn, who was noted for his A strong additional argument for this view fancy footwork, and I said about the W bosons. was given by Lee and Yang, who pointed out "They can run, but they can't hide." that at sufficiently high energies the Fermi the- 1 did realize that there was another way out ory would violate unitarity. Whether or not W which could save the IVB theory. If the neutrino bosons existed, some means would have to be emitted by the muon was different than that found to cure this disease. Whatever that cure which could be absorbed by the electron, then was. it would surely introduce some structure of course the graph I calculated would give into the weak interaction matrix elements, and zero. This difference in neutrinos would allow such a structure would, in the absence of a for the possibility of assigning different quan- selection rule, allow the ju — e + y decay to take tum numbers to electron and muon. While this place, at a rate that would presumably not be procedure amounted to introducing a second too different from the prediction of the IVB the- two-component neutrino. I phrased that possi- ory. There seemed no way out from the conclu- bility somewhat differently in a footnote to my sion that a conservation law to forbid the decay paper. Taking advantage of the fact that there was required, and given the other experimental were two unused* components of the original data about muons and electrons, this was pos- Dirac neutrino. I suggested that the muon sible only if there was more than one type of interacted with two components orthogonal to neutrino. The question then became how to those with which the electron interacted. It was detect that other neutrino. actually known from work by Incidentally, in the modern age. when all glo- and others that this was mathematically equi- bal conservation laws have come to be thought valent to introducing a second two component of as only approximate, there has been a revival neutrino. But at the time, it seemed to me and of interest in searching for n — e + y and sim- to others that the language using a single four ilar decays that change the muon number. The component neutrino was superior. In retro- present limit for the branching ratio is some- spect that was not the case, since thinking of thing like 10 )0 and there is an experiment in the interactions as involving a single four com- progress at Los Alamos to look for ju — e + y at ponent neutrino cannot be extended beyond the level of 10 13. There is little firm theoretical two generations, while the number of two com- guidance about the rate one should expecL ponent neutrinos could be extended indefi- Various effects, such as neutrino mixing would nitely, as we now have done to three. generate i± — e + y, but with really tiny branch- ing ratios. In my opinion the theoretically most My calculation of the rate for ju — e + y was promising vehicle leading to an observable rate published in June of 1958. It was in fact the for n — e + y involves introducing and break- last paper to be published as a letter in Physi- ing some sort of leptonic flavor gauge group, cal Review. Simon Pasternack. then the editor, but even that idea has not led to any sharp pre- gave me a choice between that and having it in dictions for n -• e + 7. the first issue of Physical Review Letters, which came out in July 1958. Since I was It is also worth mentioning that it still is not uncertain of the future prospects of Physical known whether the muonic conservation law is Review Letters. I chose to stay with tradition. additive as usually assumed, or multiplicative, The view that there were probably two distinct as suggested by Weinberg and myself, and oth- neutrinos gained support over the years ers almost thirty years ago. When we made the directly following the n — e -*- y calculation. latter suggestion, we did not have any plausible Builders of group theoretical models of the model that led to a multiplicative conservation weak interactions, such as Glashow in his law. However, there now is such a model, in 1961 paper on SU(2) X U( 1), found two neutri- which the neutrinos are taken to be Majorana nos more convenient than one. Searches for particles with mass, in that model, transitions other neutrinoless transitions between muon between muonium and antimuonium are 2 and electron, such as ^ — 3e or ^ + nucleus — e allowed with a rate proportional to GF , whe- + nucleus also gave negative results, streng- reas n — e + y is not. unless additional neu- thening the case that a conservation law was at trino mixing is introduced, so that there is at work. least an approximate multiplicative conserva- 14

tion law. (Actually there are up to three multi- One important aspect of the influence of the plicative conservation Jaws in that model, one Lee-Yang paper was that it was an indication of for each lepton generation with a Majorana the close ties between theory and experiment neutrino.) The search for muonium- in that era of particle physics. At least at antimuonium transitions is now approaching Columbia and Brookhaven. and. I am sure, also the level where an effect might be expected if at the other institutions where particle physics the law is multiplicative, and it will be interest- research was carried out. there was an intense ing to see what these searches give. interaction between theorists and experimen- ters, which was beneficial to both groups and c. What Else Can Be Learned From to the progress of physics. The fact that in Neutrino Experiments physics theorists and experimenters have The final stream that combined to generate something constructive to say to one another the two-neutrino experiment was the recogni- seems altogether obvious to us, but it is a point tion that there were many open theoretical that can be entirely missed by people in other questions about weak interactions that might branches of science. be answered through such experiments. The Some researchers in other sciences have cri- existence of two neutrinos was only one of ticized particle physics because experimenters these questions. A good summary of the situa often do experiments suggested to them by the- tion in 1960 was given in a paper by Lee and 13 orists. Even a little reflection should make Yang , published in the Physical Review Let- them recognize that this situation is actually a ters directly following Schwartz's paper on the beneficial consequence of the fact that particle possibility of high energy neutrino experi- physicists have some theoretical understand- ments. Some of the questions discussed in the ing. If such collaboration does not occur in paper by Lee and Yang were the following: other fields, it may indicate a lack of any effec- tive theories. 1. Are there neutral lepton currents? 2. Do the leptons interact at the same Unfortunately, it seems to me that the space-time point in the weak current? intense collaboration between theorists and 3. What are the isospin properties of the experimenters in particle physics has waned, weak hadronic currents? and is in some danger of disappearing. One 4 What happens to weak interactions at indication of this is the change that I menti- very high energies? oned earlier in the ratio of theoretical to experi- 5. Are the weak interactions mediated by mental papers. Much of the work being done in an intermediate vector boson? theoretical physics today is unmotivated by any desire to make contact with experiment, While several of these questions had been either past or future. I believe that a good deal previously posed in various contexts, the sug- of the blame for this lies with theorists. There gestion that they could all be investigated by a has come to be a kind of pecking order among single experimental technique played a power- us. in which those who retain an interest in ful role in stimulating experimentalists to what experimenters are doing are disparag- develop that technique. All of these questions ingly referred to as 'phenomenologists'. and are were in fact investigated through neutrino considered to be at a lower intellectual level experiments over the following decade or two, than those who pursue mathematical models and all but the last of them were answered for their own sake. While creating new models through those experiments. The nondiscovery is an important tool for realizing the physics of of the W-boson by neutrino experiments was a the next century, and it is surely impossible to consequence of its high mass, then unexpected tell which of today's mathematical models will by anyone other than Schwinger and Glashow. turn out to be relevant to experiment, 1 believe The rest of us, misled by the fact that no parti- that there is presently an overemphasis on cles known in 1960 had rest energy more than such abstract modeling, as opposed to con- 2 Gev. expected the W mass to be in that range fronting the phenomena that have been or can as well, in which case it would have been easily be observed. This overemphasis is harmful to found in the later neutrino experiments. both aspects of physics. To paraphrase a paraphrase. " Theory without experiment is we don't really understand leptonic (or quark) lame, experiment without theory is blind." flavor, the way we think about it is on out- It may also be that the great success of the growth of the two-neutrino experiment. Both standard model has removed much of the sti- the pairing of charged with neutral leptons, mulus to theorists to be concerned with experi- and the symmetry of quark doublets and lep- ments, as the experiments now being done ton doublets are an outgrowth of the discovery appear to some to be unlikely to lead to any real of the muon neutrino. novelties. In any case, such behavior, like social But I think that even more important than fashions, often run in cycles, and it may well be the specific discovery of the muon neutrino that. Congress willing, the next generation of was the demonstration that high energy neu- experiments will produce enough new physics trino experiments could be done. This demon- to lure some theorists back from their stration led to a tremendously influential series mathematical reveries. of neutrino experiments at Brookhaven, CERN and Fermilab. Neutrino experiments, together The Legacy of the Two-Neutrino with electron- scattering, dominated Experiment high energy physics in the late 1960s, and The two-neutrino experiment was performed in together with electron-positron colliders domi- 1962 and the rest is history. In trying to evalu- nated it in the 1970s. Only in the 1980s has ate its legacy. I am reminded of Christopher the focus shifted to hadron colliders. A list of Wren, the architect who designed St. Paul's the major discoveries for which experiments cathedral and who was also one of the discover- with neutrino beams played the dominant role ers of the conservation of energy. On the include the following: entrance to the cathedral appear the words "Si monumentum requiris, circumspice". or 1. The muon neutrino roughly translated, "If you would see his monu- 2. Hadronic and leptonic neutral currents ment look around." Evaluating the legacy of the 3. Charm two-neutrino experiment is a bit like seeing the 4. The quark structure of hadrons work of Wren. Much of our picture of the inter- All in all. there has been quite a legacy from a nal symmetry of the weak and electromagnetic coffee-hour conversation on the eighth floor of interactions stems from that discovery. While Pupin almost thirty years ago. 17

References

1. Observation of high-energy neutrino 7. Search for improbable meson decays, reactions and the existence of two S. Lokanathan and J. Steinberger, kinds of neutrinos, G. Danby, J-M Phys. Rev. 98,240(A), (1955) Gaillard, K. Goulianos, L.M. Lederman, 8. The universal Fermi interactions, H.M. N. Mistry, M. Schwartz and J. Mahmoud and E. J. Konopinski. Phys. Steinberger, Phys. Rev. Lett. 9,36, Rev.92, 1045.(1953) (1962) 2. Partial-symmetries of weak 9. A theory of the fundamental interactions, S. L. Glashow )NucL interactions, J. Schwinger, Ann. Phys. Phys., 22, 579, (1961) (NY) 2,407, (1957) 3. Observation of a with 10. Vanishing of the neutrino rest mass, strangeness minus three, V.G. Bames K. Nishijima, Phys. Rev. 108,907. et al. Phys. Rev. Lett, 12, 204, (1964) (1957) 4. Evidence for the 2 n decay of the K° 11. Transformation of muons into meson, J. H. Christenson, J.W. Cronin, electrons, G. Feinberg, P. Kabir and S. V.L. Fitch and R Turlay, Phys. Rev. Weinberg, Phys. Rev. Lett, 3,527, Lett 13. 138,(1964) (1959) 5. CP-violation in the renormalizable theory of weak interactions, M. 12. Decays of the fi meson in the Kobayashi and T. Maskawa, Prog. intermediate meson theory, G. Theo. Phys. 49,652, (1973) Feinberg, Phys. Rev., 110, 1482, (1958) 6. Feasibility of using high-energy 13. Theoretical discussions on possible neutrinos to study the weak high-energy neutrino experiments, interactions, M. Schwartz. Phys. Rev. T.D. Lee and C.N. Yang, Phys. Rev. Lett, 4,306, (1960) Lett, 4, 307, (1960)

U.S. GOVERNMENT PRINTING OFFICE: 1989—612-038-00002