Particle physics is badly named: Harry Cliff on fields, the Higgs, and a collider with no boss

When Lex Fridman sat down with Harry Cliff in the Royal Institution lecture theatre in London, he started with what he called one of the coolest things human beings have ever built. Cliff, a particle physicist at the University of Cambridge who works on the LHCb experiment at CERN, described the Large Hadron Collider as a 27-kilometre ring buried a hundred metres under the countryside just outside Geneva. Its job is to understand the basic building blocks of the universe. The useful picture, he said, is a gigantic microscope — and the analogy is fairly precise, except that this microscope looks at the structure of the vacuum.

Particle physics, Cliff told him, is badly named. Particles are not the fundamental ingredients of the universe at all. What physicists actually believe the world is made of are invisible, fluid-like objects called quantum fields, present everywhere in space, like the magnetic field around a magnet. It was a pointed place to say so. Michael Faraday invented the idea of the field in this building, mapping the invisible loops around a magnet with iron filings on paper. In the modern picture, a particle is a little vibration, a ripple, in one of those otherwise invisible fields.

Fridman admitted he found particles easier to visualise than fields. Cliff’s reply was that the school drawing of an atom — a nucleus of little spheres, electrons like flies in orbit — is completely misleading. An electron is not a planet around a nucleus; it is a spread-out, wobbly, wave-like thing, and that has been known since early-twentieth-century quantum mechanics. Physicists keep the word “particle” because in many experiments the things behave like marbles or bullets, leaving tracks through a detector. Study them carefully and they are ethereal disturbances in underlying fields. “Our bodies are basically made up of little knots of energy in these invisible objects all around us.”

Every atom, Cliff said, is made of electrons plus two kinds of quark — up and down — in the nucleus. Each is a ripple in its own field: there is an electron field filling space, and every electron in a body is a ripple moving in it. The same goes for the up-quark and down-quark fields. The Standard Model catalogues a known set of these fields. The most recent one discovered was the Higgs field, and the way it was found was to make a little ripple in it.

That is what the LHC did. It fired two protons into each other hard enough to disturb the Higgs field, and that disturbance showed up as the Higgs boson. The particle everyone talked about around 2012 is proof, Cliff argued, but the particle itself is not the prize. The interesting object is the field, because the Higgs field is why electrons and quarks have mass. The boson is “just our way of checking it’s there.” Getting that ripple required a huge amount of energy, which is why the machine is huge. Size, here, is not a vanity project.

There are two kinds of accelerator you can build, Cliff said: a circle and a long straight line. The circular machine’s advantage is that you can kick the particles every lap. A stretch of the LHC only about 30 metres long holds metal boxes with oscillating two-million-volt electric fields, timed so a proton is attracted as it approaches and then kicked out the other side as the field flips. Send it around again. He is not an engineer — “people much, much better at this stuff than I” built the machine — but the beams are divided into bunches he likened to swarms of bees, on the order of two thousand of them around the ring.

Stand at one point and count, and you get 40 million bunches passing every second. The accelerating voltage oscillates at hundreds of megahertz, faster than the 40-megahertz bunch rate, so the timing has to land with the right sign as each swarm arrives. Cliff called it difficult engineering that is not, in principle, a serious conceptual challenge — then added that engineers were probably screaming at him. The real reason the ring is so large is speed. Protons in the LHC travel at 99.9999991 percent of the speed of light. Bend a car that fast and you need a lot of tyre friction; bend a proton and you need a magnet.

The limiting factor is how powerful a magnet you can make. When the LHC was designed it used the strongest magnets that could conceivably be built at the time. Make the machine smaller and the bend gets tighter, so the magnets must be stronger. Make the tunnel bigger and you can live with weaker magnets. It is a toss-up between magnet strength and how large a tunnel you can afford. In principle the ledger is that simple. In the tunnel, he conceded, it is not.

Fridman asked for the catalogue of particles and the history that leads to the LHC’s hopes. Cliff offered “all of twentieth-century physics in like five minutes.” At the end of the nineteenth century, matter was atoms: indestructible spheres, one kind per chemical element. In 1897, experiments at Cambridge’s Cavendish Laboratory, where Cliff is still based, showed smaller negatively charged pieces inside the atom, later called electrons. A few years later Rutherford showed a tiny positively charged nugget at the centre. By about 1910 the schoolbook atom — nucleus, electrons around it — was in place.

Radioactivity then supplied bullets. Alpha particles spat from radioactive elements were fired at other atoms and knocked bits out of the nucleus: first protons, about two thousand times heavier than the electron, then, roughly a decade later, the neutron. By about 1932 the picture was tidy. Protons and neutrons stuck in the nucleus by the strong force, electrons held by electromagnetism, three building blocks for every atom. Then cloud chambers, the first devices that could image subatomic tracks, were used on cosmic rays, and a zoo appeared: antimatter, and a load of other particles nobody had asked for.

The positron, the electron’s positively charged counterpart, was a rare case of theory leading. Paul Dirac combined quantum mechanics and relativity and predicted the anti-electron; it was found about a year later. Cliff called Dirac probably the second-brightest physicist of the century after Einstein, and nowhere near as well known. When an electron meets a positron they annihilate, mass turning into energy, usually photons. By around 1960, accelerators hundreds of metres across — enormous compared with university-lab kit — had produced more than a hundred strongly interacting particles that were not ingredients of atoms. Physicists, Cliff said, were tearing their hair out. Physics is supposed to simplify, and the simple thing had become a mess.

Fridman wanted to know how you even aim a photon or an electron. Early accelerators, Cliff said, were mostly fixed-target machines: accelerate around a ring, fire out the side into something. Ernest Lawrence’s first cyclotron in 1932 was tiny. Head-on colliders, two beams meeting, came much later and are a harder problem. At the LHC a bunch is not a tight ball. It is about 30 centimetres long and thinner than a human hair, with a cross-sectional diameter around a dozen microns, and it holds some 100 billion protons. Two such swarms fly through each other in the middle of a detector. Maybe ten protons actually hit.

You cannot line up particle against particle and wait. You rely on probability, and you squash the beams as narrow as you can, like two flocks of birds: far apart, they miss; packed densely, they collide. The whole accelerator is computer-controlled from a CERN room Cliff compared to NASA mission control, because nobody can be in the tunnel while it runs. Software for steering the machine, in his experience, is kept relatively unfancy so it does not fall over. The cleverness is saved for analysing data.

He offered an anecdote from 2008, when the LHC was first switched on. There was a launch, a press conference, a party. About ten days later, during tests, a huge explosion in the tunnel badly damaged about half a kilometre of the machine. Engineers in the control room watched screens go red. Their first assumption, one of them later told Cliff, was that the software was wrong. Nothing that catastrophic was supposed to be possible.

Back at 1960, with a hundred particles on the table, Murray Gell-Mann noticed symmetries. Arrange the particles in certain ways and they related to one another; he used those patterns to predict particles that accelerators then found. A little later he and George Zweig proposed that the patterns existed because the particles were made of smaller things: quarks. For a long time many people treated quarks as a theoretical convenience that happened to fit the data. Experiments tried to knock a quark out of a proton. It never happened. It still has not.

What convinced people was a three-kilometre electron gun at Stanford. Electrons near the speed of light were fired at protons, and at high energy they bounced off small, hard objects inside — an X-ray of the proton, Cliff said, pinging off three ball-bearing-like things. Those were the quarks Gell-Mann and Zweig had predicted. They remain stuck inside other particles because the strong force is too strong to liberate them. Picture two quarks held by a powerful spring. You pull, energy stores in the bond, the spring snaps, and the energy becomes two new quarks on the broken ends. You started with two. You end with four.

So how is a quark “real”? From scattering angles, from a model that predicts new particles, and from the fact that the theory of the strong force matches what the LHC sees when quarks try to escape and spawn jets of bound particles. An isolated quark is not in the data because it cannot be. Around the same decades, Richard Feynman, Julian Schwinger, and Sin-Itiro Tomonaga built the first quantum field theory of electromagnetism, in which the photon is a ripple in the electromagnetic field. A similar theory of the strong force predicted gluons. Gravity sits outside the Standard Model; a quantum theory of it, Cliff noted, is still a Nobel Prize waiting to happen.

Only two Standard Model particles are massless, Cliff said: the photon and the gluon. All twelve matter particles — six quarks and six leptons, including the electron, its heavier cousins the muon and tau, and three neutrinos — have mass. So do the weak force’s carriers, the W plus, W minus, and Z, at 80 to 90 times the proton’s mass. The heaviest of all is the top quark, about 175 proton masses, for reasons nobody knows. The weak force is short-range because those heavy fields die off fast; electromagnetism, with a massless photon, follows an inverse-square law you can feel across a room.

In the 1960s, putting massive force fields into quantum field theory produced infinities. Even an electron with mass was a problem. The way out, associated most famously with Peter Higgs though about six people hit on it at around the same time, was to say that the particles do not have mass intrinsically. They acquire it. A new field bunches around an electron; the energy stored in that crowding is what we measure as mass. Turn the Higgs field off, he said, and all the particles in nature would become massless and fly at the speed of light.

Once the Higgs field is in the picture, electromagnetism and the weak force become two aspects of one interaction: electroweak theory, written in the mid-1970s, predicted four new particles — W plus, W minus, Z, and the Higgs. The W and Z were found in 1983–84 at CERN’s seven-kilometre Super Proton Synchrotron. From the 1980s on, people were confident the Higgs had to exist, because the structure does not work without it. The 27-kilometre tunnel was not even dug for the LHC. It was built for the Large Electron–Positron Collider, which ran from the late 1980s until 1999, studied the W and Z with high precision, and never had quite enough energy for the Higgs. By shutdown, the jigsaw had one gap. Finding the Higgs was, in one mood, “a confirmation of the obvious.” What makes it interesting, Cliff said, is that the Higgs is unlike every other Standard Model particle — theoretically troublesome, and a reason to expect company.

Measure the electromagnetic field in a shielded box in empty space and it goes to zero, aside from quantum fluctuations. The Higgs field does not. It has a value everywhere, like a background temperature of space, and that stored energy is what gives electrons and quarks mass. The precise value has to be carefully tuned if interesting things are to happen. Naive calculation, Cliff said, offers two likely fates: the field collapses to zero, and you get a universe of massless particles that never form atoms; or it explodes to the Planck scale of quantum gravity, where even an electron would collapse into a black hole. Landing in between takes fine-tuning of the other fields. People, he said, find this deeply fishy.

Fridman offered a designer, or a multiverse in which we happen to live in the habitable universe. Cliff did not rule a creator out, but called it scientifically untestable. He mentioned Lee Smolin’s idea that universes are born inside black holes, with a Darwinian bias toward laws that make more black holes, and then set it aside. As an experimentalist he finds such pictures interesting and not really testable. There was a solution that did not need a fiddler or a multiverse: supersymmetry, a still-unseen symmetry between force particles (bosons) and matter particles (fermions). Every Standard Model particle would have a superpartner. Extra fields would cancel the Standard Model’s pull on the Higgs and park it at a sensible value without tinkering.

Supersymmetry also predicted a dark-matter particle and hinted that the strong and electroweak forces might unify at high energy. Before the LHC turned on, expectations were high. If superpartners stabilise the Higgs, their masses should sit near the energies the machine probes. Discover the Higgs, and you should probably discover superpartners — ripples in superfields that, in Cliff’s opening language, already fill the vacuum. They have not arrived. After a decade of LHC data there are no superpartners, and no new particles beyond the Standard Model at all. Older “technicolor” theories that made the Higgs a bound state of new strongly interacting particles were largely ruled out by the Higgs discovery. Partial compositeness, which some of Cliff’s Cambridge colleagues work on, still tries to explain both the Higgs’s awkward value and why matter particles come in three repeating generations — a periodic table whose deeper pattern nobody has found.

The dream of one simple object underneath, Cliff agreed, is still alive for many people. He pointed to Steven Weinberg’s *Dreams of a Final Theory*, written when string theory’s “superstring revolution” made a vibrating loop that was also a quantum theory of gravity feel close. Three decades on, he said, string theorists would probably admit that nobody really knows what string theory is yet. Worse, most of its predictions live at the Planck scale, around 10^18 GeV — some fourteen or fifteen orders of magnitude above the LHC’s 14 TeV collisions. With today’s technology, reaching that energy would mean an accelerator that circled the Milky Way. Baring an almost unimaginable breakthrough, a final unified theory’s most distinctive claims will not be tested in a laboratory.

The exciting possibilities at this energy, Cliff said, are still large even if new particles are scarce. About 95 percent of the universe, from cosmology, is dark matter and dark energy — names for things we do not know. The LHC might still produce a dark-matter particle, or supersymmetry might still be hiding in a neglected corner of the data, or the processes might simply be rare enough to need another decade. A lot of people, he added, would now say the chance of a direct discovery in the near future is quite slim. That is why his own experiment does something else.

Four big detectors sit around the ring. ATLAS and CMS, with about three thousand people each, are the general-purpose machines that found the Higgs and hunt for supersymmetry and dark matter. Cliff’s experiment, LHCb, studies B quarks — bottom or beauty; they prefer beauty, “because it sounds sexier.” They can make hundreds of billions of them and measure their properties with high precision. Imagine hunting an elephant in a jungle, he said. You can wander and hope to meet the one elephant, or you can look at the ground for footprints. LHCb is in the footprint business: looking for the impressions that undiscovered heavy fields leave on ordinary B-quark decays, tiny deviations from Standard Model predictions. At the time of the conversation, beauty-quark decays were the only place in high-energy physics with fairly compelling hints of something beyond the Standard Model.

The collisions are the same; the detectors are not. ATLAS and CMS are barrel-shaped onions: silicon trackers near the beam pipe, calorimeters for energy, muon chambers at the edge, catching particles that fly out in every direction. ATLAS is about 25 metres high and 40 metres long. B quarks tend to fly close to the beam, so LHCb is a cone looking in one direction and ignoring the rest. B particles live about 1.5 trillionths of a second — an eternity next to a Higgs, which lives something like a trillionth of a trillionth of a second — and fly a few centimetres before decaying. The inner silicon sits about seven millimetres from a beam with as much energy as a jumbo jet taking off, enough to melt a ton of copper: a furiously powerful thing next to a delicate sensor.

Some beauty particles oscillate, flipping between matter and antimatter as they travel. If the symmetry were exact they would spend equal time in each state and decay accordingly. They do not. That gives LHCb a laboratory for the matter–antimatter imbalance. Fridman brought back the question he had parked earlier: if the symmetry were perfect, the universe should have destroyed itself. In the Standard Model, Cliff said, the Big Bang should have made equal amounts of matter and antimatter. In every collider, a particle comes with an antiparticle; they appear together and annihilate together. No process yet found makes more of one than the other.

A few seconds after the Big Bang there should have been a Great Annihilation: particles and antiparticles smacking into light, leaving a black, empty universe of photons. Instead there is stuff, apparently all matter. One idea ties the surplus to a phase transition when the Higgs field switched on in the first trillionth of a second, bubbles forming in a hot plasma of massless particles, matter and antimatter bouncing off the walls differently. Another involves exotic neutrinos that decay with a bias toward matter. LHCb and planned neutrino experiments are trying to test those pictures. At the moment, Cliff said, there is no clear evidence for which idea is right. In Fridman’s outro, Cliff’s line is sharper: you and I are leftovers, every particle in our bodies a survivor of that shootout, only about one in a billion from the beginning still here.

Fridman wanted the human lesson: brilliant people with egos, countries with budgets, one machine. Cliff called CERN possibly the best example he could think of of international collaboration that is not for an unpleasant purpose. He arrived as a student in 2008 with the LHC basically finished, and has huge admiration for the people who spent two and a half decades — first conversations in the late 1970s, ground broken in the early 2000s — campaigning and fundraising. Scientists pull in the same direction because they want the ingredients and the laws. They also compete. On LHCb you have on the order of eight hundred collaborators who are also rivals for the exciting paper. It works, he said, amazingly well.

ATLAS has about three thousand people and a nominal head called the spokesperson, elected by the collaboration, who cannot fire anyone and is not anyone’s boss. Cliff’s boss is a professor at Cambridge; the head of the experiment cannot really tell him what to do. Independent academics, their own bosses, get things done by consensus, discussion, and lots of meetings. McKinsey or a similar firm, he recalled, once went into CERN to try to understand how the organisations function. He also pushed back on genius. You have to be relatively bright, but many of the most successful experimental physicists are not the biggest brains. The important skills are working with others, exchanging ideas, and working hard — determination more than virtuoso intellect. A lot of experimental physics is engineering by another name.

In the near term the LHC is getting a luminosity upgrade: not more energy, more collisions per second, an order-of-magnitude jump in data rate against diminishing returns. LHCb’s detector was being rebuilt to swallow that rate, partly to test intriguing anomalies in B-quark decays that all seem to point the same way, still below a discovery threshold. If those footprints are real, the next step is a bigger machine to make the particle of the new field. Europe’s options included stronger magnets in the existing tunnel (cheaper, widely disliked as a dead end), CLIC, a compact linear electron–positron collider, and the Future Circular Collider: a 100-kilometre tunnel under the Geneva region, under Lake Geneva and around the Alps, as large as the local geology allows. First an electron–positron Higgs factory, later a much bigger proton machine to hunt dark matter or recreate the early-universe phase transition. Optimistic timelines put data after 2040. Cliff put the LHC at around 10 to 12 billion euros all-in, the FCC at about 30 billion over half a century — in principle fittable in CERN’s existing budget if the lab did little else.

CERN itself, the European Organization for Nuclear Research, was founded in the 1950s as a kind of scientific Marshall Plan: bring European science back together for peaceful, curiosity-driven work after the wartime brain drain to the United States. Countries join as members and pay a small fraction of GDP; the UK contribution, Cliff thought, was on the order of 100 or 200 million pounds a year. “It’s a beautiful idea,” he said — countries collaborating to do good science when there is no war on the line.

Most LHC data is thrown away. Record it all, Cliff said, and you would fill every computer in the world in a matter of days; they keep on the order of one collision in ten thousand. A trigger on the detector, 40 million times a second, has to decide whether a collision might contain something interesting — a Higgs, a dark-matter particle — and used to do it with blunt cuts on momentum and energy. The push now is to put machine learning earlier, even on raw hits rather than reconstructed tracks, treating the detector as a strange three-dimensional image. Triggers are inefficient because they have no time for full reconstruction. Smarter algorithms could keep more useful events and reject background without new colliding hardware, though they would still need GPU farms. Cliff uses off-the-shelf packages in his own analysis; he is a physicist who had to learn the tools, not a machine-learning expert.

The conversation was recorded at the Royal Institution, about two centuries old, founded to communicate science to the public. Humphry Davy demonstrated nitrous oxide and electrolysis there; Faraday lectured and experimented in the basement. Cliff has given a couple of lectures in that theatre, his favourite a talk called “Beyond the Higgs” on the Institution’s YouTube channel. He claimed no great theory of communication: he is interested and excited, he has told the stories many times, and the explanations that work have survived contact with school talks and even a bit of standup comedy that “went reasonably well, even if it was terrifying.” He keeps a story in his head, some images, and gets up.

A Feynman line he likes is “What I cannot create, I do not understand.” He does not think of himself as particularly super-bright, so he has to break ideas into simple elements, and telling them to other people is how he finds out he does not understand them. Writing a book, he had just spent two days in textbooks after realising he did not really grasp a fundamental theoretical point in his own subject. The more you understand, the simpler the thing you finally say — which is why Feynman’s lectures feel accessible. Some ideas, including the Higgs mechanism, he still cannot properly explain, because they live in the mathematics; a graduate degree is the honest path, and a public talk can only offer a flavour that is intelligible without being misleading.

Asked for the most beautiful idea in physics, Cliff gave a specific and a general answer. The specific one, from undergraduate days: the forces of nature arise in the theories as consequences of symmetries. Assume the universe respects a particular symmetry, not far in spirit from the rotations of a cube, and you are forced to introduce a force with the properties of electromagnetism; another symmetry gives the strong force, another the weak. The general wonder is simpler still. Down close, particles bounce off each other through force fields and move in simple ways. From those few ingredients come stars that cook matter in their gravitational energy, planets, and people, in a story you can trace to a trillionth of a second after the Big Bang. Going from those laws even to a molecule is already hard; molecule to human being is not a problem physics can currently tackle. That, he said, is why there is chemistry and biology. Complexity from simple rules is beautiful, mysterious, and not a calculation anyone knows how to finish.