by Mark Bowen
The PIs scheduled a collaboration-wide analysis meeting in Zeuthen in March 2000, and Francis asked the two mainstays of the Madison analysis, his graduate student Ty DeYoung and his post-doc Gary Hill, to visit Zeuthen for three weeks in advance of the meeting in order to try to bring Christian around. His opinion was the crucial one, since he was the leader of the Zeuthen group, and Ralf, the second-in-command, was probably unreachable anyway. As Ty writes, “Ralf’s insistence that we couldn’t publish anything until [everything about AMANDA] was completely understood was way over the top, and probably would have killed any hope of doing IceCube.” The one thing that stands out in Ty’s mind from that visit is that either owing to his concerns about keeping the analysis blind or because he viewed it as a gimmick, Christian had never observed a neutrino event with the AMANDA event viewer before. This graphic means of watching an event unfold in time was the most intuitive way to see what was going on in the instrument. Ty and Gary sat him down and showed him a few events that had passed both their analysis and Zeuthen’s, as well as a few obvious fakes that the two analyses had rejected. Christian still refused to give in—on the surface at least—but Ty believes their session together had something to do with his eventual conversion.
Two months later, at a closed meeting of the PIs during a collaboration meeting in Brussels, Buford Price engineered a small break in the logjam by suggesting they present their dual (or dueling) analyses in the form of a brief letter to Nature. Zeuthen still had some discrepancies in their analysis—the Monte Carlo simulations weren’t agreeing closely enough with the real data quite yet—which may explain why Christian agreed only reluctantly. So Madison produced a first draft, presenting their analysis alone, and sent it to Zeuthen as a kind of ultimatum. (The backbone of this draft would have been written by Ty, since his Ph.D. thesis would become the primary documentation for the Madison analysis.) Some time later, an extended draft came back from Zeuthen with their analysis added, the Germans demonstrating both their pettiness and their continued competitiveness by designating the two analyses as “A” and “B” and assigning “A” to their own, even though Madison’s had been way out in front all along.
George Smoot was in the thick of this infighting, too, of course, although at some disadvantage. He knew better than most that fame and glory would come from actually making a discovery and that you can only do that if you can analyze data. His group was very close to the data stream, of course, since they were employing the supercomputers at Lawrence Berkeley Laboratory to filter the raw data coming directly from Pole. He’d recruited John Jacobsen to his group, perhaps the most accomplished code-writer in the collaboration, and he had also tried unsuccessfully to recruit Christopher Wiebusch, the driving force behind Zeuthen’s analysis. But George hadn’t managed to put much of an analysis team together yet, so he resorted instead to the next best thing: vicious mudslinging at both Zeuthen and Madison. Actually, he was badmouthing virtually every aspect of AMANDA and IceCube to anyone who would listen, including outsiders, the potentates at NSF, and myself.
He had employed such tactics quite successfully once before, on the Cosmic Background Explorer, or COBE, experiment, back in the early nineties. The person who actually made the COBE discovery was physicist Ned Wright from UCLA, who had invented a computer algorithm, much like AMANDA’s event viewer, for analyzing the data. When Wright presented the first glimmering of the discovery at a meeting of COBE’s head scientists, George, whose group was having difficulty interpreting the data, became incensed and began impugning Wright’s work. George shut off access to the portion of the COBE data that streamed through his hands and then cut off all communication with the rest of the collaboration for two months while he traveled to Antarctica to work on a different project. By the time he returned, Wright and some others had improved their methods and were now sure that his initial hunch was valid, and the collaboration had prepared three papers to announce the results. George’s group had fallen further behind in their own analysis while he’d been away, of course. He declared angrily again that none of the papers were ready for publication and then worked to delay them. When his colleagues decided to go ahead and announce the results anyway, in three talks at a meeting of the American Physical Society in Washington, he broke the rules of the collaboration and published a press release about the results a few days before the talks were given, thereby attracting worldwide attention to himself and Lawrence Berkeley Laboratory. No one else on the project was named in the press release. Even NASA, which had funded and managed the project and whose scientists were involved at the highest levels, received only passing mention. At the same time, George’s group had not contributed in any way to the analysis. For a while there, he was perhaps the most well-known scientist in the world, rivaling Stephen Hawking, and he soon signed a book contract about COBE for a rumored $2 million.
* * *
There was still a reasonable chance that IceCube would go to LBL, although Madison seemed to hold the advantage. Among the NSF officers, only Gene Loh, who was something of an outsider at the foundation, wanted it to go to the lab. (This explains why Smoot had sent his rogue white paper to Loh before the May 1999 IceCube meeting in Arlington.)
The foundation put together an external committee to review the project, chaired by Donald Hartill, an experimental high-energy physicist from Cornell, and scheduled what Bob Paulos calls a “soup to nuts management and technology review” for the end of June. It was held in Madison, which was still being viewed as the lead institution. Most of the AMANDA/IceCube PIs were there, and several high administrators from LBL demonstrated their interest by showing up as well.
It was clear that NSF was new to this game, as they didn’t even send Bob an agenda until three days beforehand. But it was also clear that the members of the Hartill committee knew their stuff. And they weren’t afraid to level criticism.
One of their frank judgments was that the University of Wisconsin did not know what it was getting into. Paulos remembers that he and Francis were told “the lights were going to shine very brightly.” “Every time you make a move, and every time you make a mistake, it’s just gonna be an order of magnitude brighter than what you are used to. So just be prepared for that.” Terry Millar recalls Beverly Hartline, a hardnosed project director–type from Los Alamos National Laboratory, telling them “UW simply is not in the game.” She compared them to someone who wanted to run a marathon (Terry suspected she’d run a few), “training for a few weeks at two-three miles a day, and then deciding they were ready for prime time. They might last until mile five or six, but they will not be there at mile fifteen, to say nothing of miles twenty or twenty-six.” And Francis loves to tell the story of how Hartline told him and Paulos that they’d be gone in a year or two, replaced by more experienced managers. Several years later, when she turned up at another “Hartill review,” he cheerily pointed out that he and Bob were still there. (These reviews continued for years and were very helpful.)
After holding their feet to the fire for several days, the Hartill committee gave them a surprisingly hearty endorsement, and the NSF officials made it clear that they planned to approve IceCube. Of the many reviews that AMANDA and IceCube have undergone over the decades—certainly more than a hundred—this is the most memorable for Francis. His nametag from the first Hartill review still sits on his desk in his home office.
Per Olof used to love telling the story of how they all went out to dinner at a restaurant on State Street afterward and how Francis, who is known for his epicurean tastes, chose the wine for a toast. He was so excited (not only at the positive review, he claims, but also at the news that he’d soon be fired) that he didn’t notice that the wine had gone bad. None of the others had the heart to rain on his parade by pointing this out to him, so he went ahead and ordered another round.
The tide began shifting toward Madison.
* * *
NSF had put the contract for the Polar Ice Coring Office, PICO, out for bid again, and
Bob Paulos had written up what he thought was a longshot proposal to bring it to the Space Science and Engineering Center. “If we had the drilling contract, then the hot water drill would be a big component of that, that would help IceCube, and, oh by the way, the other stuff might be interesting, too,” he explains (that “other stuff” being tens of millions of dollars’ worth of ice core drilling, mostly in the polar regions). At the end of June he found out that he’d won the lottery.
“This guy calls me out of the clear blue.… He introduces himself and then he says, ‘Well, you probably know why I’m calling.’ I go, ‘I have no idea why you’re calling.’ He goes, ‘I just wanted to let you know that you guys have won the ice coring and drilling contract.’ I’m like, ‘Really!’ And I remember my first thought was ‘Oh shit!’ because I had so much going on that I couldn’t imagine now having to deal with this.” (He already had about one and a half jobs between IceCube and his regular SSEC responsibilities.)
“I said, ‘Okay, so how’s this gonna go?’ And he says, ‘I’m here to tell you, as of today you’re authorized right now to spend a million dollars.’ And I go, ‘Well that’s really great. Are we done for now?’ And he goes, ‘Yup. ’ And I go, ‘Okay, well thanks a lot. I guess we’ll be in touch.’ And so I hang up the phone, and I go, ‘This is really kind of interesting.… We won the ice coring and drilling contract; at that time that included the development of the hot water drill for IceCube—assuming IceCube got funded; but IceCube wasn’t funded yet. But yet, we were tasked to start developing the drill.’ If that makes any sense.”
This brought things to a head.
In July, Lawrence Berkeley Laboratory made a formal request to become the lead institution for IceCube. On the first of August, Willi Chinowsky followed up with a curious and condescending e-mail to Francis.
“I make you, and IceCube, an offer you cannot refuse,” he began. “Accepting some personal sacrifice, for the good of IceCube, and to lessen your burden, I offer to take charge of … IceCube organizational matters as interim Project Director.” There were sops for Madison: Albrecht Karle would be deputy director and Bob Paulos project manager. But IceCube would be led from the lab, Smoot would be PI, and Francis was nowhere to be found. Willi also made the bizarre request that Francis be the one to share the new arrangement with the collaboration as a whole.
“Do you think that I didn’t sleep at night when I received this e-mail?” Francis asks. “No. Did it upset me? Yes.… [But] you want the science to happen; that dominates everything, right? And so then you ask, does it have a better chance of succeeding if LBL does it?”
He wasn’t sure he knew the answer to that question.
In any event, LBL’s request forced everyone’s hands. John Wiley and Terry Millar began lobbying energetically for the project, the NSF folks began backing Wisconsin, too, and in response, it seems that something in George Smoot snapped. On the twenty-seventh of August, he sent an e-mail to the entire collaboration, students included, under the title “Why I asked my name to be taken from the Nature letter draft.” “There are two reasons,” he began:
First, I have been nominated for a Nobel Prize, other awards, and a promotion and as a result all my postings and publications will be reviewed carefully by the appropriate committees. It is clear that the draft AMANDA Nature paper containing essentially no science but only assertions based on theoretical prejudice and wishful thinking rather than firm documented experimental results is more likely to be an embarrassment than an asset.…
He pointed out that this was an easy decision, since he was involved in another project that was producing good science. “Second,” he continued,
I am appalled that graduate students and post docs are being shown that it is an acceptable standard to publish work that is sloppy and lacking in scientific integrity.
(The other principal investigators, by and large, held quite the opposite view. They were appalled that George would have aired such dirty laundry in front of students, since it would undermine their morale.)
He ended by asking rhetorically if AMANDA or IceCube would ever become assets rather than embarrassments.
Of the several responses to this missive, the most entertaining came from Per Olof Hulth, always a constructive voice, who also knew a thing or two about the Nobel Prize. As a member of the Royal Swedish Academy of Sciences, he voted every year on the winners.
Dear George,
I do not think you need to worry to sign any “sloppy” paper because you are nominated for the Nobel price [sic]. I do not know about the other awards which you are up for. But for the Nobel price you can for sure sign several papers with wrong results, much worse than any “sloppy” paper and still get the price for one single paper which have the real discovery! It is the discoveries or the genius ideas confirmed by experimental results which counts, not the mistakes. You can e.g. think about a very good scientist like Carlo Rubbia […]
Anyway, I disagree with your comments about the Nature letter. It has never been the idea that it should be anything else than the announcement that we think we have a working detector and that the forthcoming papers with all details will come in the near future. The version which Bob now has edited is nice and fulfills this limited goal.
Looking forward to see you in Stockholm.
Cheers,
Per Olof
Ironically, George’s capable and diplomatic LBL colleague, Bob Stokstad, had incorporated the many beneficial comments submitted by others into the most recent draft of the paper.
The letter was submitted to Nature, absent George’s authorship, about two weeks after his outburst. This was more than twenty months after Gary Hill and Phil Romenesko had experienced their moment of truth late that summer night in Madison. It was published the following March. Ten years of work by more than a hundred people were boiled down to three pithy pages, under the title “Observation of high-energy neutrinos using Čerenkov detectors embedded deep in Antarctic ice.”
* * *
Two months later, Gene Loh called Francis to tell him that IceCube had received formal approval and Madison would run it. Francis responded just as Bob Paulos had when he’d learned about the drilling contract: “Oh shit!,” a rare four-letter word coming from him.
“Halzen,” said Loh, “sometimes you get what you ask for.”
18. No New Starts
Except that it didn’t happen. That formal approval was given by the Bill Clinton administration only three months before Clinton left office. Within days of George W. Bush’s inauguration in January 2001, the new administration began disseminating a series of directives exercising an unprecedented level of control over the way scientists and scientific agencies throughout the federal government would now be doing business. One of the first of these diktats was nicknamed “no-new-starts”: the National Science Foundation would not fund any new capital projects in the coming year. IceCube lost its funding even before it began.
At a collaboration meeting at the University of Delaware in March, Francis told me, “We were in Clinton’s budget. If Al Gore had won the election, we’d be drinking champagne right now.”
* * *
When he had first discussed the idea of running IceCube out of Madison with John Wiley, the dean of the Graduate School, and Terry Millar, associate dean for physical sciences, Francis had exacted a promise that it would not be business as usual: rules would be broken if they got in the way. Not only did the two administrators live up to that promise, Wiley became considerably more capable of doing so on January first, a few weeks before Bush was inaugurated, when he became chancellor of the Madison campus. He and Millar responded to the no-new-starts diktat by loaning IceCube $4.5 million to keep it afloat while the funding quandary sorted itself out. The money came from WARF, the Wisconsin Alumni Research Fund, and Wiley backed it with a guarantee that the university would repay WARF from its general fund if the National Science Foundation didn’t come through. He later said it was the best loan he ever made.
* * *
The main task during this unstable period was to design and build a “production” drill. Bob Morse and Bruce Koci had been on the case since mid-1999, actually, and they’d been having some fun with it. Both were opera buffs, and Bruce and his wife, Ann, were dedicated Wagnerians. In the time I knew them they made separate pilgrimages to New York, Seattle, and Chicago to attend the complete Ring Cycle. (Ann points out that Bruce was the only person she ever met who could grovel in the squalor of a drilling camp for three months, come home, change into his tuxedo and rush off to the opera, and look right at home.) Thus inspired, Bruce named the drill Wotan, after the king of the gods in the Ring, and Bob actually referred to it as such in the drilling section of the first IceCube proposal, which he wrote.
Wotan would need to be an entirely different creature from the AMANDA drill. The plan for IceCube was to deploy no less than eighty strings, each a mile and a half long and comprising sixty optical modules, spread out over a square kilometer. If they were to accomplish this gargantuan task in the five years called for in the proposal they would need to deploy sixteen strings per season, almost three times the level of their best season with AMANDA. Wotan’s heating plant would be more than twice as powerful as AMANDA’s, five megawatts versus two; the hose would have a larger diameter, two and a half inches versus one and a half; and they would “remove the current necessity of changing reels during hole drilling” by winding a single hose nearly two miles long onto a reel so large that it would take two Hercs to ship it to Pole. It would be shipped in pieces and assembled once it got there. When the hose was filled with water, Wotan’s reel would weigh nearly fifty tons (see photographs 20–22).