February 1996: A Media Storm Hits CDF
    By Tommaso Dorigo | February 4th 2014 05:13 AM | 1 comment | 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
    On January 25th 1996 the CDF collaboration submitted for publication to Physical Review Letters  the result of their measurement of the rate of jet production as a function of jet transverse energy, performed on 20 inverse picobarns of data collected by the experiment in the 1992-93 run of the Tevatron collider. That measurement deviated at its high-energy tail from predictions of quantum chromodynamics, suggesting that the underlying model -and most likely, the parton distribution functions (PDF) which describe the probability of finding partons with given fractions of their parent momentum- was at fault. An alternative explanation, quite intriguing and groundbreaking, was that the excess was due to the presence of a inner structure in the quarks: preons, as hypothesized twenty years before by Pati and Salam.

    The graph below, which summarizes the 1996 CDF result, is rather complex, so let me guide through it. In the top panel you can see a main distribution and an inset. First let us look at the inset: it shows the CDF measurement of the rate of jet production as a function of the transverse energy of the jet, as a steeply falling histogram. Overlaid with the data points is the theoretical prediction, obtained by next-to-leading order QCD calculation complemented with a MRSD0' model of the parton distribution functions in the proton. You might observe that the agreement is spectacular over several orders of magnitude; and yet there is a small departure of the data from the curve at the right end of the spectrum.

    The main graph shows the fractional difference between the CDF data and the theoretical model, in percentage of corrected rate (technically, a fractional difference in the differential cross section). The data show excesses of as much as 120% from the model at large jet ET. Other PDF models provide similar conclusions; they are shown as curves with various choices of dashing.

    At the bottom there is a systematic error band, which shows how systematics sources of uncertainty may affect the rate measurement; note that however those errors are completely correlated across the x axis.

    At the time of submission of the paper the particle physics community had already digested the CDF result. The public debates on whether to call agreement or discrepancy the comparison of the CDF spectrum to its DZERO omologue had been losing strength, as both collaborations were more interested in improving the statistical power of their measurements by using the four-times larger dataset acquired in 1994-95, which had recently become available to analyzers.

    The QCD group in CDF was also in the process of producing a precise measurement of the angular distribution of the jet pairs in those same events which populated the highest-ET bins of the inclusive jet ET distribution: the angular information offered a powerful test of the nature of the excess, and its study would soon reveal whether the hypothesis of preons was to be swept off the table or considered more seriously.

    Further, the CTEQ group had already published a preprint where they indicated that the likely cause of the CDF anomaly was the underestimate that all existing PDF models made of the gluon PDF at large momentum fractions. It looked like the paper would land peacefully on PRL and be soon forgotten, but things were going to be much different.

    In the course of January 1996 the Science magazine conducted several interviews and collected material for a publication of the story. The Science reporter was James Glanz, a physicist with a Ph.D. in astrophysics obtained at Princeton: this was close to the best possible scenario that CDF could hope for, as the article was guaranteed to contain no sensationalistic claims nor incorrect or deceiving statements. And indeed it did not. However, Science had a policy of "leaking out" to local newspapers the most interesting articles they were about to publish, just before the magazine got distributed. And there was the rub: would local newspapers keep the same high standards of the flawless text produced by Glanz ? Not at all, of course.

    On the February 9th 1996 issue of Science, James Glanz titles open-mindedly: "Collisions Hint That Quarks Might Not Be Indivisible". His article is quite good and balanced, and it explains well the physics without making unsupported claims or imprecise statements. The article also contains several quotes from CDF members, starting from the spokespersons:
     "This is just the sort of effect you would see" says CDF-co-spokesperson William Carithers, "if quarks were not fundamental particles but had some sort of internal structure".
    And then Giorgio Bellettini:
    "As these events began to accumulate [...] "a fierce fight" broke out within the collaboration over how to gauge the small chance that systematic experimental errors could explain the results. The researchers made exhaustive tests of the possibility that a "conspiracy" of random or systematic errors might be fooling them [...]. Finally, he says, the collaboration reached a consensus that the excess had to be real."
    The article closes with a quote from Steve Geer, the convener of the QCD group, who actually points out what would only later be accepted as the correct solution:
    "Steve Geer, a CDF team member at Fermilab, describes the most dramatic possibility: "It might mean that, just as in Rutherford's atom, there's a hard center" lurking inside the quarks, as some speculative theories suggest. But Geer points out that several other explanations might account for the measurements. The more mundane possibility, he says, has to do with how momentum is parceled out among the components of a speeding proton. The hardest collisions occur when two quarks that happen to carry a high fraction of each proton's momentum meet head-on. But the massless gluons can carry momentum as well. So if, say, QCD underestimates how often gluons carry a high fraction of the momentum, then the quarks they encounter could suffer an unexpected number of violent collisions, and "we could end up with more energetic jets than expected".
    Frankly, the Science article does not appear to be damaging to the CDF collaboration: no strong claims, and just a report of work in progress. But in 1996 an influential minority of the CDF members were uneasy with any sort of media coverage of their work; a few of them were quite frightened that the lack of control over what could be perceived from the outside might damage their reputation, and worsen their chances of getting money and grants from the funding agencies. By and large, that paranoid feeling had been originated by an article by Malcolm Browne which appeared on the New York Times on January 5th 1993, just months before the Superconducting Supercollider project was canceled by the US congress.

    The article was titled "315 Physicists Report Failure in Search for Supersymmetry", and it was built around the rather unremarkable news that a paper had been sent by CDF to Physical Review Letters to report lower mass limits in a search for squarks and gluinos. Browne had insisted on reporting the point of view of critics of "big science": just to let you taste the flavour of the piece here are a few clips:
    "[...] despite this arsenal of brains and technological brawn assembled at the Fermilab accelerator laboratory, the participants have failed to find their quarry, a disagreeable reminder that as science gets harder, even Herculean efforts do not guarantee success." [...] "Some regard such failures as proof that high-energy physics, one of the biggest avenues of big science, is fast approaching a dead end." [...] "Failed experiments of such a grandiose scale offer easy marks to critics who contend that "big science" produces too meager a crop of knowledge for what it costs." [...] "all these projects soak up more human talent and more public money than society should permit, skeptics argue."
    To his credit, Browne concluded the article with a beautiful quote by Wolfgang Panofsky, the director-emeritus of the SLAC laboratories:
    "If certain answers crucial to man's understanding of nature can be obtained only by large effort, is that sufficient reason for not seeking such answers?".
    But that was not nearly enough. The damage of similar press, real or hypothetical, was quite present in the eyes of many inside CDF, and their feelings would be fortified by the cancelation of the SSC which followed a few months later.

    Of all the quotes in Glanz's article, the one by Bellettini was judged badly. It spoke of fights within the collaboration, and of a "real effect"; and it came from a spokesperson, which was an aggrievating clause. Worse still, as the Science paper hit the road, Bellettini was also featured in a National Broadcast Radio interview, in which he explained the matter in his usually colorful, enthusiastic way. Giorgio has always been a true enthusiast of particle physics research: his verve and his passion have in truth always fascinated his students, bringing them to share his love of particle physics. I personally regard this as a great gift in a scientist and a university professor, and I actually admire him for that; I cannot understand those who criticize him for his positive, energetic, inspiring attitude. In the radio interview, Giorgio gave a remarkable answer when asked what he believed was the source of the CDF anomaly:
    "Sometimes when I go to bed I wake up and think it's new physics, and sometimes I wake up and think it is just QCD".
    As mentioned above, several newspapers throughout the US managed to organize quick articles on the CDF anomaly as they got notified by Science. Many of those pieces appeared on February 8th. On the New York Times it was Malcolm Browne again, titling "Tiniest Nuclear Building Block May Not Be the Quark". Browne had got a hold of Brenna Flaugher, one of the authors of the study on which the paper was based, and quoted her:
    ""What's different about these results," Dr. Flaugher said, "is that there are many more high-energy transverse jets than theory predicts. The highest energy transverse jets were 120 percent more frequent than theory would explain," she said."
    Brenna could not have possibly said "high-energy transverse jet" in place of the correct "high-transverse-energy jets" -no CDF members would have done that; but the quote was otherwise technically accurate. Another source identified by Browne as  "a spokesperson", and which most in CDF would bet was Bellettini, was also quoted as saying:
    "We spent more than a year working on the data to be reasonably sure we weren't making some big error," a spokesman said, "and now we're confident at least that something unexpected has happened."
    Again, not a terrible misrepresentation of facts, but the mention of "unexpected" was already enough to disturb the most sensitive collaborators. On one thing the writer fully showed he had a clear vision, as he concluded his piece with the following forecast:
    "the C.D.F. paper is expected to undergo intensive scrutiny and criticism in the months ahead."
    Curt Suplee on the same day wrote for the Washington Post and the Los Angeles Times a piece titled "Quark as Basic Particle May Be in Dispute". The quotes contained in his article did not have names attached, and the tone was possibilistic:

    "It is possible that quarks--the smallest known constituents of protons and neutrons--are not fundamental, indivisible particles, but may be made up of yet smaller entities of unknown nature, said researchers from the Collider Detector at Fermilab" [...] "Until alternative explanations or possible errors have been ruled out, the 450-member team reports in a paper submitted to the nation's leading physics journal, Physical Review Letters, "any claim about the presence or absence of new physics is not defensible.""
    In this case what was most disturbing to CDF collaborators was the rather imprecise description of the findings of the experiment: one read that
    "The team turned up nearly 1,200 observations during a year of experiments in which colliding particles were deflected or "scattered" in ways that apparently cannot be reconciled with the predictions of current particle-physics theory. But the odd trajectories and energy levels observed in the collisions might make sense if quarks were composed of tiny sub-units that could send matter flying off in unexpected ways.".
    The above description was frankly sub-par, even for a national newspaper more accustomed to deal with politics and Football than with particle collisions. It is quite surprising to notice that its author is a renowned and appreciated science writer, who has won during his career several awards for outreach activities. "Odd trajectories" and "matter flying off in unexpected ways" are rather awkward misrepresentations. They offer a good example of what I consider going below an acceptable threshold in the process of simplifying a complex physics concept . Also, the mention of "1200 observations" must have been the result of a complete misunderstanding of the way the experiment collected its data; one which I cannot even trace back to a possible source. 1200 runs ? 1200 events above 200 GeV ? We will never know, but surely Suplee's article had its share of  responsibility for the discontent among CDF members.

    The Chicago Tribune also published on February 8th an article on the CDF anomaly, titled "Fermilab's `Preons': Bad Math Or A Profound Find? As Basic Matter, Quarks Superseded Atoms Now A Flash Of Energy May Shake Up Physics". The author Ronald Kotulak wrote it after interviewing Rob Harris, who was careful enough to downplay the practical significance of the CDF find. He clearly explained to Kotulak that the theory was probably wrong, as he is quoted to say:
    "If one were a betting person, one would tend to bet that it's more likely to be an error in the calculations than it is to be a new quark substructure".
    That sentence was apparently the origin of the mention of "bad math" in the title. Alas, no matter how hard you try to give a correct, careful interview, there are always ways to be misquoted in the resulting newspaper article. Because of the Chicago Tribune article Rob got a number of unpleasant comments from colleagues who thought that too much emphasis had been put on new physics; but it really was not his fault.

    Indeed, to many the fault was in the way the whole interaction with the press had been managed by the spokespersons. Bellettini in particular became the target of the criticism; however, rather than being expressed openly in the appropriate venues, which would have been the Executive Board meeting or the associated mailing list, it remained at the level of corridor rumor, a sort of stomach rumble for the collaboration. As such, it was even worse than a public trial, as Bellettini never got a chance to defend his actions in front of his collaborators.

    The only surfacing critique of Bellettini's action was a letter which Avi Yagil posted on February 9th to the CDFNEWS forum, a mailing list where CDF members mainly exchanged news about papers in preparation, meeting times, or newborn babies; but also one which, unlike the executive board mailing list, reached all CDF members. In his posting Avi summarized the result of a discussion of the situation with colleagues who like him resented the way the interaction with the media had been handled. He explained that the spokespersons had the right to speak on behalf of the experiment, but spokespersons were not supposed to confuse their own opinions with those they were meant to represent.

    Avi's summary is believed to have had one notable effect when in 1997 the first term of Bellettini expired: he declared his availability to serve for another term, but he lost the election, which in CDF worked with the democratic rule "one head, one vote". The media-wary soul of the collaboration did not forgive him for the way newspapers and outreach magazines had covered and over-hyped the jet ET excess. In retrospect, I believe Giorgio was guilty: he was guilty of believing that creating attention on particle physics and explaining to laymen the excitement of a possible new important CDF discovery was in the interest of the experiment. There lay the mistake: it was in the interest of the progress of science, but against the interest of some of his colleagues.


    Bravo Tommaso for sharing this important piece of CDF history so nicely ! From the vantage point of the LHC, where the majority of published papers are on searches for kinds of new physics branded "Exotic", this episode at CDF now seems very quaint and conservative. Now many LHC publications have titles that include words like, "Quantum Black Holes", "Extra Dimensions", "Quark Compositeness", which a CDF scientist at the time would have been embarrassed to put in the title of his publication. At the time at CDF searches for new physics were done as almost an afterthought, and the main emphasis was on solid measurements of the standard model. The average CDF collaborator took it as given that a prudent scientist simply measures nature, and assumes the standard model is the explanation for natures behavior until a significant deviation is found and then proven to be new physics, and that is when the new physics should be discussed, even as a possibility, and not before. We would not talk about the excitement of searching for new physics: it was just too speculative, and not how a real scientist spent their time. I remember one of my references at the time, a renowned and senior Fermilab scientist, very kindly and frankly telling me, with a bit of amazement in his voice, that I "made a reputation searching for strange particles using jets". He thought of the new models of physics as "strange particles", and I think he was a bit amazed that I had made any kind of reputation at all searching for them. He respected the jet measurements for certain, but it is unclear what he really thought about using these measurements to constrain the models of new physics that were then considered by many experimentalists as the playground of theorists. Although experimentalists continue to rightly insist on a very high standard of proof of new physics, times have changed a lot since then in our willingness to early on entertain the possibility of alternatives to the standard model. Cheers, Robert Harris.