Banner
    1988 - The Second Discovery Of The Top Quark
    By Tommaso Dorigo | November 7th 2013 07:28 AM | 9 comments | Print | E-mail | Track Comments
    About Tommaso

    I am an experimental particle physicist working with the CMS experiment at CERN. In my spare time I play chess, abuse the piano, and aim my dobson...

    View Tommaso's Profile

    These days I am trying to reconstruct some stories from my old experiment, CDF. The CDF experiment was conceived in 1979 and constructed in the early eighties at the Fermi laboratories in Batavia, near Chicago. CDF took the first proton-antiproton collisions in 1985, and it collected data in1987-88, 1992-96, and 2001-2011, thus becoming the longest-lasting particle physics experiment in the history of science.

    In 1994 CDF announced the first evidence of the top quark, measuring its mass at 174 GeV, a value which still holds its ground the increase by one order of magnitude of the precision of today's measurements. But few know that the top quark had been discovered much earlier... By Rubbia in the mid eighties, and then by other CDF researchers who were, for good or for bad, not trusted enough.

    I thought the following excerpt, which tells an anecdote of one of these early sightings, would be of interest to some readers here...

    ---

    In December 1988 a one-day workshop was organized in the Ramsey auditorium, the conference room at the basement floor of the Hirise, the main building of the Fermi National Accelerator Laboratory. The workshop was the first of a series of meetings that would take place in the course of the following few years, and it was specifically devoted to focused discussions on the top quark search, which was being performed independently by several groups of CDF physicists. The Argonne National Laboratory partecipated to the CDF experiment with a strong group of researchers; they were very active in the search for the top quark in "lepton plus jets" events.

    I should explain that back then it was not events containing top-antitop quark pairs that were being sought, but rather ones containing a single top quark. In fact, in the first CDF searches the top quark was believed to have a mass not much larger than the fake 40 GeV signal just published, and soon retracted, by Rubbia's UA1 collaboration. A top quark lighter than the W boson could still, of course, be produced in pairs, but it would be much smarter to search for single top quarks produced in the decay of a W boson, since the latter process has a larger production rate.

    One was thus looking for a W boson decaying into a top quark and a bottom quark. The bottom quark would produce a low-energy jet of hadrons; the top would itself decay by turning into a lepton, a neutrino, and a second bottom quark which would yield another hadronic jet. These were thus "lepton plus jets" events of much smaller energy than those later called "W+jets" which would be used for the top discovery six years later.
     
    The Argonne group gave a very detailed presentation on lepton plus jet events. The researchers who presented the work were Jimmy Proudfoot and Barry Wicklund, both experienced and careful physicists. They had looked at many things: jet activity, momentum of tracks, lepton isolation, missing energy. The presentation spurred a lively debate on the details of their event selection, and on their definition of the variables used to identify clean electron and muon events. The conclusion of the Argonne researchers was that it looked like there was no clear signal of a top quark in the data.

    After the Argonne presentation Pekka Sinervo got on stage. Pekka worked for the CDF group of Pennsylvania State University. He also had looked for the top quark, and he also was reporting a negative result. His talk received wide interest, criticism, suggestions from the audience. It was all in the right spirit of a workshop where the best minds of the experiment would collaborate to get the most out of the data they had worked so hard to produce.

    On the other hand, the picture could also have felt a bit depressing: the talks were giving the message that the top quark was not there to be found, and furthermore, one got the feeling that a well-defined strategy for the top search was missing. Indeed, back then it was not even clear to most CDF researchers that the main background to top production was constituted by events featuring a W boson together with hadronic jets produced by QCD radiation.

    Finally, the time came for the talk by Kuni Kondo. Prof. Kondo was a Japanese physicist who led a sizable group of researchers from the University of Tsukuba. In his late fifties, he was lean, not tall, with black hair combed straight above an incipient baldness; he usually dressed in black or grey suits. He was a charming and very polite person, who spoke with a soft tone of voice and smiled a lot. It looked like nothing could ever upset him.

    Kondo had devised a very complex, deep method to discriminate top quark events from the background, based on an analysis approach he had dubbed "dynamical likelihood" which would become a sophisticated standard only a decade later, but which was taken with quite a bit of scepticism at the time; in private, quite a few of his American and Italian colleagues would even make silly jokes on it. The method consisted in constructing probability distributions for the observed kinematics of the events, which could then be used to derive the likelihood that the events were more signal-like or background-like.

    It is ironic to think that nowadays all the most precise measurements of the mass of the top quark rely on the method called "matrix element", which is nothing but Kondo's original idea recast in the context of a measurement of the mass rather than the discrimination of a top signal. Kondo was way ahead of his time, and like most pioneers in science he did not have an easy life getting his work appreciated and accepted, in a situation dominated by a conservative mainstream.

    It is by now four in the afternoon, and Kondo finally gives a full status report of his analysis. His presentation is thorough and yet almost unintelligible by a good half of his listeners; his analysis includes highly unorthodox and yet brilliant tricks, like taking a jet from one event and mixing it in with other jets in a different event to study the behaviour of some of his selection variables for background events. His colleagues listen in an atmosphere of disbelief mixed with awe. Despite the complexity of the material and the possibility to object on a hundred of details, no questions are asked. As Kondo reaches the end of his talk, he concludes with a tone of voice just a milli-decibel higher than the rest of his speech:

             "And therefore", a pause, and then  "I thinkwe have discovered the top quark".

    The audience remains silent. The convener is a tall, lean guy with a sharp nose and a penetrating stare; he looks like an English gentleman from a XIXth century novel, especially thanks to his considerable aplomb. He is not impressed, and that much does show.

             "Thank you very much Kuni. Is there any question ?", one, two, three, four, "...No questions. Okay, thanks again Kuni. The next speaker is...".

    In retrospect the convener's attitude and lack of consideration toward an esteemed colleague and a visitor from another country, who had brought to the experiment lots of resources and had contributed significantly to the detector construction, sounds at least rude and unjustified. Still, back then CDF was not a place where people would exchange courtesies and compliments (it never was, in truth): there everybody had to work hard and the only way to earn the respect of colleagues was through the good physics output of one's analysis results. If your analysis methods were not considered publishable or your results were thought fallacious, you would be considered a potential threat to the good name of the experiment, and you would suffer little short than boycott. But the way Kondo was treated was all flowers in comparison to what other physicists would experience, along the way to the top discovery.

    Comments

    Thanks for some details beyond what is in the book "The Evidence for the Top Quark" by Kent W. Staley (Cambridge 2004) about Kondo's claim that the 1988-9 candidate event "discovered the top quark"
    and
    the collaboration bureaucracy that blocked his effort to publish in PRL.

    Since Kondo was from Tsukuba, could he have published his analysis in a Japanese journal such as Progress of Theoretical Physics

    which would have been similar to what Sliwa, Goldstein, and Dalitz did by publishing their similar analysis of the same event in the European journal Physics Letters B ?

    Would Kondo have suffered the same consequences as Sliwa et al suffered had he done so ?

    Would the Japanese journal have shown independence as did the European journal
    or would it have been as controlled by Fermilab as the American journal PRL ?

    Would the collaboration have treated Tsukuba as it treated the home university of Sliwa and Goldstein - that is, telling the university that if they did not shut up about their analysis, then Fermilab would bar anyone from Tufts from ever participating in their collaboration ?

    Note that (afaik) Fermilab did not present a similar ultimatum to Dalitz's home university, Oxford.

    Tony

    dorigo
    Hello Tony,

    interesting questions. I think there was a difference between Kondo and Sliwa. Kondo did respect the CDF rules to the letter. Had he not done this, there is a chance that the CDF management would have had fewer ways to handle the situation because Kondo was Japan and Japan could not be shown the door.
    As for Dalitz, he was not a CDF member so CDF could really not do much to him.

    Cheers,
    T.
    So in retrospect what was the big flaw in Kuni's analysis? I think the last time I checked the top was more heavy than the W boson.

    dorigo
    Hi Matt,

    I don't actually know ! But Kuni was a bit, let's say, too enthusiastic about his tentative signals, and this sometimes led him to overstate the significance of his finds. I think in his analysis he had just spotted some weird W decays with jet activity.

    Cheers,
    T.
    I am not sure about all the details of the 1988 event that Kondo analyzed
    but
    here is a URL for a histogram of a later candidate event, from the 1997 UC Berkeley PhD thesis of Erich Ward Varnes
    http://www.tony5m17h.net/12814euM.gif
    In this event, there were 3 jets instead of the usual 2 for dilepton events.
    As you can see there are two peaks.
    The solid line peak around 170 GeV is if all 3 jets are included.
    The dashed line peak around 130 GeV is if only 2 jets are included
    and one of the 3 is excluded.

    My guess is that the Kondo 1988 event might have had 3 jets
    and that Kondo cut one of the jets so that the event would look more like a typical dilepton event.

    That would allow the 1988 event to have really been at 170 GeV
    but as analyzed by Kondo to have looked like 130 GeV.

    Tony

    Vocabulary tip: Don't say "fake signal" because it implies willful trickery, fabrication by researchers. Instead say "spurious signal" to describe a signal that is thought in good faith to exist but does not hold up to closer scrutiny..

    dorigo
    Thanks! Very appreciated, will keep it in mind !
    T.
    Hi Tommaso,
    In the papers on the actual observation of single top quark production in 2009 by DZero and CDF, the multivariate analysis methods prototyped by Kuni Kondo and others were completely trashed by the first set of PRL referees, even though they had been applied and published many times in the intervening two decades. Old attitudes die hard! From discussions with the CDF analysts, I think that both papers were reviewed by the same people, and that the second referee saw the report of the first one before sending in her/his one. I have not seen the comments on CDF’s paper but the ones on DZero’s were shocking in their misunderstanding of what had been done, and I gather those on CDF’s paper were similar in nature. Fortunately we were able to address all the issues and get the papers published five months later, and the two discovery papers now have nearly 300 citations each. Similar analysis techniques were applied in the Higgs boson discovery last summer without all the fuss over their validity. Thank you for this report on the second “discovery” of the top quark, it made interesting reading.
    Sincerely,
    Ann Heinson

    dorigo
    Hello Ann,

    thank you for the interesting note! I did not know the story of those two papers.
    Cheers,
    T.