by John Gribbin
The importance of all this confusion is that this really is the way things were in the middle to late 1960s – confused. Most people regarded the quark model as a wild idea; even Gell-Mann seemed to be at best half-hearted about it, and the one person who had vigorously promoted it found his career prospects severely damaged as a result. Gell-Mann continued to develop the idea (with fewer reservations), but as experiments at high-energy accelerators never did find any evidence of free particles with fractional charge, many physicists found it hard to believe in the reality of quarks.
Gell-Mann, by now, was nearing the end of his career as a great original thinker. He had been born in 1929, did his best work between about 1954 and 1964 (between the ages of 25 and 35), was appointed R. A. Millikan Professor of Theoretical Physics at Caltech in 1967, and received the Nobel Prize in 1969, settling down as a wise older member of the science community but making only relatively minor contributions to fundamental physics after he entered his forties. This is very much the pattern associated with an ordinary genius, and it might have been natural to expect the next leap forward to come from a member of the younger generation, like Zweig. In fact, it was made by a man who was eleven years older than Gell-Mann, and who was just entering his fifties.
It is a sign of how little faith the physics community had in the quark model that in 1969 the citation for Gell-Mann’s Nobel Prize rather pointedly avoided any reference to the idea, mentioning instead his earlier work on the classification of elementary particles and their interactions – in other words, the eightfold way and the theory of the weak interaction.7
Within a year of his encounter with James Watson in Chicago, where he relearned the lesson that the way to make progress was to disregard what others were doing and start from first principles, Feynman was getting to grips with the theory of what happens in collisions between hadrons – for example, when a proton collides, at very high speed (that is, high energy) with another proton (or an antiproton). This was in 1968, the year Feynman turned 50, and the year in which Michelle joined the family. He developed a model – a way of describing what went on in such collisions – by regarding each hadron as a cloud of point-like particles. He was deliberately agnostic about the nature of those internal constituents – they might be quarks, or they might not. As ever, Feynman was solving the general problem, for any number of particles with whatever individual properties they might happen to have, not looking at any particular special case; even at this late stage of his career, the work had all the hallmarks of a classic piece of Feynman research, right down to the mathematical toolkit he used to tackle the problem.
This wasn’t out of sheer stubbornness. Feynman’s insistence on always trying to solve problems in the most general form, with as few preconceptions as possible, was part of his philosophy, a way of ensuring that you, the researcher, stayed honest in the game of developing theoretical models to explain, or (better) predict what was happening in experiments. Giving the Caltech commencement address in 1974, smack in the middle of his last great burst of creativity as a physicist, he would tell his audience of aspiring scientists about the importance of absolute integrity in science, that ‘the first principle is that you must not fool yourself – and you are the easiest person to fool. So you have to be very careful about that. After you’ve not fooled yourself, it’s easy not to fool other scientists. You just have to be honest in a conventional way after that.’8 This is why he made no presumptions about the nature of the internal constituents of the hadrons, and dubbed those constituents ‘partons’, a rather ugly word indicating that they were parts of a hadron, but one which carried no load of expectation or preconception about the nature (or even the number) of the particles.
You might think of a swarm of such partons inside a hadron as like a swarm of bees, moving around in a roughly spherical volume of space. But when the hadron is moving at close to the speed of light, as Feynman realized, strange relativistic effects come into play. The sphere is squashed in the direction of its flight, as seen by somebody at rest in the laboratory, to become a highly flattened pancake. For example, a sphere travelling at 0.999957 times the speed of light (this has been achieved in this kind of experiment) shrinks to times its rest thickness along the line of sight, but stays the same diameter at right angles to the line of flight, becoming a pancake 108 times wider than it is thick. When two such pancakes smash into one another broadside on, according to the parton model most of the partons inside would pass right past one another and off into the sunset. But, just occasionally, two of the partons themselves would collide, slowing dramatically and releasing energy in the form of a flood of ‘new’ particles. This was the basis of Feynman’s model, in which the probability of a collision between two hadrons can be considered as the sum of the probabilities of collisions between individual partons – a mathematical formalism echoing the sum over histories idea.
Feynman worked all this out in the first half of 1968, and developed the insight into a mathematical model, containing the basis of lots of predictions that could be compared with experiment. For, of course, if it disagrees with experiment it is wrong. Just at that time, a new particle accelerator had been built at Stanford University, in northern California. It was called the Stanford Linear Accelerator Center, or SLAC, and used a straight, 2-mile-long tube to fire a beam of electrons at a target, where they collided with stationary protons, producing debris in the form of particles streaming out from the point of collision. By monitoring the showers of particles produced in this way, the researchers hoped to find out what protons were like inside. Such an experiment was a few steps short of colliding protons, but since electrons can be regarded as point-like particles, the hope was that by scattering the electrons off protons the experiment would reveal any structure inside the protons, in the same way that scattering particles at much lower energies off atoms had revealed, decades before, the existence of the nucleus inside the atom.
These experiments were being carried out jointly by a team of researchers from MIT and SLAC, led by Jerome Friedman, Henry Kendall and Richard Taylor (similar investigations were carried out at about the same time by researchers at the Deutsches Elektronen-Synchrotron, or DESY, in Germany). The early results from these experiments were being interpreted by a Stanford theorist, James Bjorken, who has described Feynman’s arrival on the scene, and his influence on the development of particle physics at the end of the 1960s, in his contribution to Most of the Good Stuff.
Bjorken obtained his PhD from Stanford in 1959, and recalls how, like many other physicists, as a graduate student in the late 1950s he learned quantum electrodynamics the old-fashioned way, ploughing through what was essentially a 1930s style course with ‘a seemingly endless, gloomy, turgid mass of field-quantization formalism’. But then came a revelation – ‘when Feynman diagrams arrived, it was the sun breaking through the clouds, complete with rainbow and pot of gold. Brilliant! Physical and profound! It was instant conversion to discipleship.’
Something very similar happened with partons. Bjorken joined Stanford University just after completing his PhD, and soon became a tenured member of the faculty there. By 1967, he was a full professor at SLAC. At the time Feynman arrived on this particular scene, Bjorken had been developing a theoretical description of what was going on in the electron–proton collisions at SLAC using a highly mathematical formalism known as current algebra, largely developed, as it happens, by Gell-Mann. He had plotted out graphs of what went on at different energies during the collisions, but had no simple physical picture of what was going on. In the summer of 1968, Feynman happened to be visiting his sister, Joan, who was living near SLAC at the time, and in August he went over to SLAC to see what the experimenters were up to. Bjorken was away, but the experimenters and the other theorists showed Feynman the results Bjorken was coming up with, as well as the raw data. The key feature of this work was that the data looked the same – the graphs had the same shape – whatever the energy of the interactions. This is known as scale inv
ariance. Although Bjorken’s colleagues at SLAC could not explain to Feynman where Bjorken had come up with this prediction, which matched the experimental results, Feynman realized that it echoed his own work on partons, using the relativistic pancake description of particle interactions.
‘It took Feynman only an evening of calculation with his partons to interpret what was going on’, says Bjorken. He returned to SLAC just before Feynman was due to leave, and:
found much excitement within – and beyond – the theory group there. Feynman sought me out and bombarded me with queries. ‘Of course you must know this … Of course you must know that …,’ he kept saying. I knew about some of the things Feynman mentioned; others I didn’t know. And there were things I knew at the time but he did not. What I vividly remember was the language he used: it was not unfamiliar, but it was distinctly different. It was an easy, seductive language that everyone could understand. It took no time at all for the parton model bandwagon to get rolling.9
Everyone could understand. Just as Schwinger had sneered at the way Feynman’s version of QED brought computation to the masses, so Gell-Mann sneered at what he called ‘Feynman’s put-ons’, which made particle physics theory accessible even to people who could not handle the complexities of current algebra.
Feynman went back to SLAC in October 1968 to give a talk about his ideas, and the parton model swept through the team there like wildfire. Over the next few years, experiment and theory developed hand in hand, and it gradually became clear that a version of parton theory in which the partons were identified with quarks could best explain the experimental results. But the power of the parton model was that it also allowed for the possibility of other entities, besides quarks, residing inside protons and neutrons. Feynman was convinced from the outset that quarks – if they did exist – could not be isolated particles any more than an electron could be an isolated particle. Electrons, remember, are surrounded by clouds of virtual photons, the carriers of the electromagnetic force; the current picture of the situation inside a proton or a neutron is that the quarks are associated with clouds of ‘gluons’, the carriers of the strong force, which holds them together. Parton theory automatically took account of this kind of possibility.
The early version of the theory was largely developed by Bjorken and his colleague Emmanuel Paschos at SLAC; the verification of the reality of quarks was acknowledged in 1990 by the award of the Nobel Prize to Friedman, Kendall and Taylor for the experimental side of the work. Feynman would have approved of this recognition that in physics experiment is king. In 1988 he said, ‘I am now a confirmed quarkanian!’10 As Bjorken has put it, referring to Feynman’s eventual espousal of quarks, ‘it was data that forced the commitment (for both of us)’.11
It was well into the 1970s, though, before the blending of quarks and partons was complete, and Feynman himself still had a significant contribution to make to the development of the quark model. As unconcerned as ever about the inglorious rush for priority, he did not hasten to publish these ideas (although he did give several talks on parton theory at scientific gatherings), and his first paper on the subject, written with two of his students, only appeared in the Physical Review in 1971, with the cautious comment, ‘a quark picture may ultimately pervade the entire field of hadron physics’.12
But there was still a major problem with quark theory. If particles existed with a charge one-third or two-thirds of the size of the charge on an electron, why had nobody seen them? Of all the properties they could have, the fractional charge was a distinctive feature that could be observed in very simple experiments. If quarks were real, the only reason that fractional charges were never seen in nature must be because somehow they were kept locked up, or confined, inside hadrons, and could not roam freely about in the world. In that case, you could always ensure that the combined charge on a meson, made of a pair of quarks, added up to round numbers such as zero (+⅓ together with –⅓) or 1 (+⅔ together with +⅓), and similarly for baryons suitable triplets of quarks would give, for example, (+⅔ together with –⅓ and –⅓) or (+⅔ together with +⅔ and –⅓).
The picture that emerged was one in which the force that binds quarks together must get stronger when the quarks are farther apart. This is both strange and quite natural. In physics, we are used to dealing with forces between two objects that, like gravity or magnetism, are stronger for objects that are closer together. On the other hand, in the everyday world we have a simple example of a force that gets stronger as distance increases. Try stretching an elastic band, and you will literally get a feel for the force between quarks.
Imagine a collision between two quarks which are components of relativistic pancakes travelling in opposite directions. Consider just one of the quarks, happily sitting in a triplet, that receives energy from a head-on collision with a quark in the other pancake and recoils, moving away from its partners. At first, it moves away freely. But the further the quark is going to move, the more energy will be required to drag it apart from its companions. If there is not enough energy available, the quark will snap back into place, like a stretched elastic band snapping back when it is released. But if there is enough energy in the collision, the quark will break the bonds that bind it to the other quarks, breaking free, like an overstretched elastic band snapping in half. But does this mean that we now have a free quark? No! For by ‘enough energy’ we mean that there is so much energy in the collision that it can create a pair (at least) of new quarks, one on each side of the ‘break’ in the ‘elastic band’ (really, the strong force) trying to hold the recoiling quark in place. Instead of a single quark escaping, you have a pair of quarks (forming a meson), or even a new triplet; instead of two quarks from the original triplet being left behind, a new companion appears on the other side of the ‘join’ and remains alongside them.
This is a slightly oversimplified picture. At very high energies, instead of a simple break with one new quark appearing on each side of the join, the process of breaking the grip of the strong force would produce a shower of new particles, manufactured out of pure energy, forming a jet moving in the direction of the escaping quark. But what matters is that nowhere in that jet of particles emerging from the site of the original collision is there an isolated quark; the particles in the jet are formed out of a train of quark pairs and triplets created by repeatedly breaking the bonds between other quarks.
From 1972 onwards, experimenters at CERN were able to observe such jets in collisions between beams of particles travelling in opposite directions; this is exactly the ‘colliding pancake’ situation that Feynman had described theoretically a few years earlier. Throughout the 1970s, researchers at CERN and elsewhere found more and more examples of this kind of behaviour, as they probed to higher and higher energies. The important point is that the jets can come out of the collisions almost at right angles to the line of flight of the colliding pancakes, and this is only possible because at the moment of collision the quarks hardly feel the strong force restraining them at all. When they are close together, they do not notice that they are confined (a property known as asymptotic freedom); it is only when they try to escape that they feel their restraint.
Richard Field, a postdoctoral researcher at Caltech, became interested in the properties of these quark jets, and persuaded Feynman to join him in investigating the jet properties theoretically. Using the language of what is now known as quantum chromodynamics (QCD), involving quarks exchanging gluons in an analogous way to electrons exchanging photons in QED (see Figure 15), and with asymptotic freedom included as part of the package, Feynman and Field were able to make predictions about the kind of jets that should be observed. According to Field,13 Feynman kept them honest by insisting that they only calculate the behaviour of jets in experiments that had not yet been carried out, so that the experiments would provide a genuine test of the theory; as the higher-energy experiments were carried out, they produced jets of exactly the kind that the two Caltech theorists had predicted.
> Figure 15. Also using QED as its template, QCD describes interactions between quarks. Here, two quarks on diverging paths exchange a gluon and are pulled back towards one another.
This work, some of which also involved another theorist, Geoffrey Fox, was being carried out in the second half of the 1970s. As Bjorken puts it, ‘as the evidence for QCD grew, Feynman (with Richard Field) worked out the modifications to the “naive” parton model phenomenology implied by QCD, and grappled with the fundamental properties of QCD that might explain confinement’. Some of the work with Field was published in 1977, and some in 1978 – the year of Feynman’s sixtieth birthday. Physicists simply don’t make major contributions to their field in their late fifties, yet here was Feynman, still (or again) at the heart of new developments in particle physics. It wasn’t just that his own theory of QED, developed more than 30 years before, provided the archetype on which QCD was based, but that Feynman himself was actively involved in establishing QCD as the best theory we have of the strong interaction.