On a fall morning in 2009, a team of three young physicists huddled around a computer screen in a small office overlooking Broadway in New York. They were dressed for success—even the graduate student’s shirt had buttons—and a bottle of champagne was at the ready. With a click of the mouse, they hoped to unmask a fundamental particle that had eluded physicists for decades: the Higgs boson.
Of course, these men weren't the only physicists in pursuit of the Higgs boson. In Geneva, a team of hundreds of physicists with an $8 billion machine called the Large Hadron Collider also was in the hunt. But shortly after starting for the first time, the LHC had malfunctioned and went offline for repairs, opening a window three guys at NYU hoped to take advantage of.
The key to their strategy was a particle collider that had been dismantled in 2001 to make room for the more powerful LHC. For $10,000 in computer time, they would attempt to show that the Large Electron-Positron collider had been making dozens of Higgs bosons without anybody noticing.
“Two possible worlds stood before us then,” said physicist Kyle Cranmer, the leader of the NYU group. “In one, we discover the Higgs and a physics fairy tale comes true. Maybe the three of us share a Nobel prize. In the other, the Higgs is still hiding, and instead of beating the LHC, we have to go back to working on the LHC.”
Cranmer had spent years working on both colliders, beginning as a graduate student at the Large Electron-Positron collider. He had been part of a 100-person statistical team that combed through terabytes of LEP data for evidence of new particles. "Everyone thought we had been very thorough," he said. "But our worldview was colored by the ideas that were popular at the time.” A few years later, he realized the old data might look very different through the lens of a new theory.
So, like detectives poring through evidence in a cold case, the researchers aimed to prove that the Higgs, and some supersymmetric partners in crime, had been at the scene in disguise.
The Higgs boson is now viewed as an essential component of the Standard Model of physics, a theory that describes all known particles and their interactions. But back in the 1960s, before the Standard Model had coalesced, the Higgs was part of a theoretical fix for a radioactive problem.
Here's the predicament they faced. Sometimes an atom of one element will suddenly transform into an atom of a different element in a process called radioactive decay. For example, an atom of carbon can decay into an atom of nitrogen by emitting two light subatomic particles. (The carbon dating of fossils is a clever use of this ubiquitous process.) Physicists trying to describe the decay using equations ran into trouble—the math predicted that a sufficiently hot atom would decay infinitely quickly, which isn't physically possible.
To fix this, they introduced a theoretical intermediate step into the decay process, involving a never-before-seen particle that blinks into existence for just a trillionth of a trillionth of a second. As if that weren't far-fetched enough, in order for the math to work, the particle—called the W boson—would need to weigh 10 times as much as the carbon atom that kicked off the process.
To explain the bizarrely large mass of the W boson, three teams of physicists independently came up with the same idea: a new physical field. Just as your legs feel sluggish and heavy when you wade through deep water, the W boson seems heavy because it travels through what became known as the Higgs field (named after physicist Peter Higgs, who was a member of one of the three teams). The waves kicked up by the motion of this field, by way of a principle known as wave-particle duality, become particles called Higgs bosons.
Their solution boiled down to this: Radioactive decay requires a heavy W boson, and a heavy W boson requires the Higgs field, and disturbances in the Higgs field produce Higgs bosons. “Explaining” radioactive decay in terms of one undetected field and two undiscovered particles may seem ridiculous. But physicists are conspiracy theorists with a very good track record.
How do you find out if a theoretical particle is real? By the time Cranmer came of age, there was an established procedure. To produce evidence of new particles, you smash old ones together really, really hard. This works because E = mc2 means energy can be exchanged for matter; in other words, energy is the fungible currency of the subatomic world. Concentrate enough energy in one place and even the most exotic, heavy particles can be made to appear. But, they explode almost immediately. The only way to figure out they were there is to catch and analyze the detritus.
Modern particle accelerators like the LEP and LHC are like high-tech surveillance states. Thousands of electronic sensors, photoreceptors, and gas chambers monitor the collision site. Particle physics has become a forensic science.
It’s also a messy science. “Figuring out what happened in a collider is like trying to figure out what your dog ate at the park yesterday,” said Jesse Thaler, the MIT physicist who first told me of Cranmer's quest. “You can find out, but you have to sort through a lot of shit to do it.”
The situation may be even worse than that. To reason backward from the particles that live long enough to detect to the short-lived undetected ones, requires detailed knowledge of each intermediate decay—almost like an exact description of all the chemical reactions in the dog’s gut. Complicating matters further, small changes in the theory you’re working with can affect the whole chain of reasoning, causing big changes in what you conclude really happened.
While the LEP was running, the Standard Model was the theory used to interpret its data. A panoply of particles were made, from the beauty quark to the W boson, but Cranmer and others had found no sign of a Higgs. They started to get worried: If the Higgs wasn't real, how much of the rest of the Standard Model was also a convenient fiction?
The model had at least one troubling feature beyond a missing Higgs: For matter to be capable of forming planets and stars, for the fundamental forces to be strong enough to hold things together but weak enough to avoid total collapse, an absurdly lucky cancellation (where two equivalent units of opposite sign combine to make zero) had to occur in some foundational formulas. This degree of what's known as "fine-tuning" has a snowball's chance in hell of happening by coincidence, according to physicist Flip Tanedo of the University of California, Irvine. It's like a snowball never melting because every molecule of scorching hot air whizzing through hell just happens to avoid it by chance.
So Cranmer was quite excited when he got wind of a new model that could explain both the fine-tuning problem and the hiding Higgs. The Nearly-Minimal Supersymmetric Standard Model has a host of new fundamental particles. The cancellation which seemed so lucky before is explained in this model by new terms corresponding to some of the new particles. Other new particles would interact with the Higgs, giving it a covert way to decay that would have gone unnoticed at the LEP.
If this new theory was correct, evidence for the Higgs boson was likely just sitting there in the old LEP data. And Cranmer had just the right tools to find it: He had experience with the old collider, and he had two ambitious apprentices. So he sent his graduate student James Beacham to retrieve the data from magnetic tapes sitting in a warehouse outside Geneva, and tasked NYU postdoctoral researcher Itay Yavin with working out the details of the new model. After laboriously deciphering dusty FORTRAN code from the original experiment and loading and cleaning information from the tapes, they brought the data back to life.
This is what the team hoped to see evidence of in the LEP data:
As light particles hurtled through the detector's many layers, detailed information on their trajectory was gathered (see sidebar). A tau particle would appear in the data as a common origin for a few of those trails. Like a firework shot into the sky, a tau particle can be identified by the brilliant arcs traced by its shrapnel. A Higgs, in turn, would appear as a constellation of light particles indicating the simultaneous explosion of four taus.
Unfortunately, there are almost guaranteed to be false positives. For example, if an electron and a positron collide glancingly, they could create a quark with some of their energy. The quark could explode into pions, mimicking the behavior of a tau that came from a Higgs.
To claim that a genuine Higgs had been made, rather than a few impostors, Beacham and Yavin needed to be extremely careful. Electronics sensitive enough to measure a single particle will often misfire, so there are countless decisions about which events to count and which to discard as noise. Confirmation bias makes it too dangerous to set those thresholds while looking at actual data from the LEP, as Beachem and Yavin would have been tempted to shade things in favor of a Higgs discovery. Instead, they decided to build two simulations of the LEP. In one, collisions took place in a universe governed by the Standard Model; in the other, the universe followed the rules of the Nearly-Minimal Supersymmetric Model. After carefully tuning their code on the simulated data, the team concluded that they had enough power to proceed: If the Higgs had been made by the LEP, they would detect significantly more four-tau events than if it had not.
The team was hopeful and nervous as the moment of truth approached. Yavin had hardly been sleeping, checking and re-checking the code. A bottle of champagne was ready. With one click, the count of four-tau events at the LEP would come onscreen. If the Standard Model was correct, there would be around six, an expected number of false positives. If the Nearly-Minimal Supersymmetric Standard Model was correct, there would be around 30, a big enough excess to conclude that there really had been a Higgs.
“I had done my job,” Cranmer said. “Now it was up to nature."
There were just two tau quartets.
“Honey, we didn’t find the Higgs,” Cranmer told his wife on the phone. Yavin collapsed in his chair. Beacham was thrilled the code had worked at all, and drank the champagne anyway.
If Cranmer's little team had found the Higgs boson before the multi-billion-dollar LHC and unseated the Standard Model, if the count had been 32 instead of 2, their story would have been front-page news. Instead, it was a typical success for the scientific method: A theory was carefully developed, rigorously tested, and found to be false.
“With one keystroke, we rendered over a hundred theory papers null and void,” Beacham said.
Three years later, a huge team of physicists at the LHC announced they had found the Higgs and that it was entirely consistent with the Standard Model. This was certainly a victory—for massive engineering projects, for international collaborations, for the theorists who dreamt up the Higgs field and boson 50 years ago. But the Standard Model probably won't stand forever. It still has problems with fine-tuning and with integrating general relativity, problems that many physicists hope some new model will resolve. The question is, which one?
“There are a lot of possibilities for how nature works,” said physicist Matt Strassler, a visiting scholar at Harvard University. “Once you go beyond the Standard Model, there are a gazillion ways to try to fix the fine-tuning problem." Each proposed model has to be tested against nature, and each test invariably requires months or years of labor to do right, even if you're cleverly reusing old data. The adrenaline builds until the moment of truth—will this be the new law of physics? But the vast number of possible models means that almost every test ends with the same answer: No. Try again.