Insight Editions · No. 23
Feynman's Rainbow
A Search for Beauty in Physics and in Life
Leonard Mlodinow
Leonard Mlodinow recounts his conversations with Richard Feynman at Caltech, reflecting on scientific creativity, the rivalry among physicists, and the search for personal standards.
A study guide to Feynman's Rainbow by Leonard Mlodinow. Shared free for personal reading; not for sale and not endorsed by the authors or publishers. Please read and buy the original.
Preface: why physics, why Feynman
Mlodinow opens with scale. Fewer than 800 physics PhDs are granted in the United States each year, yet physics gave the world X-rays, lasers, radio, the transistor, atomic energy and atomic weapons. Feynman stands at the top of that small trade: the man whose 1985 memoir Surely You're Joking, Mr. Feynman! became a best seller, who in 1986 dipped a shuttle O-ring in ice water on television, and who was called at his death in 1988 the "Einstein of modern times."
Physics, in the preface's telling, is a small guild with outsized consequences. The same trade that produces fewer than 800 American doctorates a year also produced the devices and weapons that remade the twentieth century. Feynman is presented as the public face of that guild: the bestselling memoirist of 1985, the man who on television in 1986 preferred ice water and common sense to computer models, the figure eulogised in 1988 as an Einstein for modern times.
Mlodinow's personal stake is narrower and sharper. A Berkeley doctorate on quantum theory in infinite dimensions had bought him a Caltech fellowship and an office near two Nobel laureates. It had not bought him a next idea. Caltech folklore about the suicide rate among high achievers sat in the background. The knock on Feynman's door, shortly after the second cancer operation, was not careerism alone; it was a bid to learn method while there was still time to watch it practised.
Mlodinow arrived at Caltech after his 1981 Berkeley PhD on a research fellowship. His thesis had been on quantum theory in infinite dimensions. Now he had no ideas, a reputation to protect, and Caltech's grim folklore about the suicide rate among its high achievers. He knocked on Feynman's door shortly after Feynman's second cancer operation, and they talked, on and off, over two years.
The preface also frames the genre. This is not a technical exposition of QED or string theory. It is the story of a first year at Caltech, of Feynman's last years, of the rivalry with Gell-Mann, and of the lonely beginnings of string theory, told by someone who was there with a tape recorder and a blank research agenda. The questions he brought to the door were not about Feynman diagrams. They were about what it takes to be a scientist, how scientists think, and where creativity comes from.
BLANK-KNOCK-TAPE. Three beats for the preface: a blank mind with one good thesis behind it, a knock on a dying master's door, and years of tape that become this book. Recite blank-knock-tape and you have why the memoir exists.
Write three sentences naming your current blank: the credential you already have, the door you have not knocked on, and the question you would ask if you did. Do not solve the blank. Name it. That is the preface's move.
The preface invites you to treat proximity to great people as a cure for having no ideas. Invert it: sitting down the hall from Nobels can freeze a young researcher, raise the suicide folklore Caltech already carried, and turn "knock on the door" into a dependency. The book's later chapters answer that shadow: the cure is an internal standard, not a borrowed rainbow.
Fewer than 800 US physics PhDs a year; physics that remade technology and war; a fellow with no ideas who taped Feynman for two years. Treat the dialogue as faithful in spirit. The real names are few; most characters are composites.
Chapter I: two lives at a turning point
Cancer found in summer 1978; five-year survival under 10 percent; he is in his fourth year. About forty years earlier he had sent the Physical Review the diagrams that by late 1981 were in every particle physicist's notebook. He had one Nobel Prize and, Mlodinow argues, arguably deserved two or three. His plan for whatever time he had left: play with quantum chromodynamics (QCD), the theory of the strong force.
The physics backdrop is a split. Quantum electrodynamics (QED), Feynman's own theory of electrons and light, explains atomic physics with fantastic precision. QCD, the theory of quarks and the strong force that holds nuclei together, is believed to be correct, but nobody can calculate with it. No one could compute even the mass of the proton from it. That is the problem Feynman wanted.
QED's precision is the chapter's first contrast. Atomic physics could already be calculated to extraordinary accuracy with Feynman's theory of electrons and light. Nuclear physics, rebuilt around quarks and the strong force, had a believed-correct theory (QCD) that refused to yield even the mass of the proton on paper. That gap is why a dying man still had a research programme: not a new force, but a way to compute with the one already written down.
Mlodinow's fear is sketched in two registers. The schoolboy image (dreading the junior-high showers) names exposure. The professional image names emptiness: a physicist without ideas as one of the living dead. The Caltech appointment of 1980, won on infinite-dimensional methods applied to atomic physics, had become a stage without a play.
Mlodinow, meanwhile, had won his Caltech post in 1980 on the strength of papers applying an infinite-dimensional method to atomic physics. He describes his fear in a schoolboy image (dreading the junior-high showers) and a professional one: a physicist without ideas is one of the living dead.
The chapter's quiet cruelty is that both men are stuck, for opposite reasons. Feynman has the problem of a lifetime (QCD that nobody can compute) and a body that may not last long enough. Mlodinow has the appointment of a lifetime and nothing to work on. The "living dead" are not failures; they are people who still go to the office after the ideas stop.
QED-works / QCD-won't / me-blank. Chapter I in three tags: QED already fantastically precise, QCD believed but uncomputable (even the proton mass), and a young fellow afraid he is finished. Hold the three tags and the turning point is clear.
Name one domain where your field has a QED (a theory that already works) and a QCD (a theory everyone believes but cannot calculate with). Write one sentence on which you would spend a scarce year, and why. That is Feynman's choice in 1981.
"Play with the hardest unsolved problem" is Feynman's privilege after a Nobel and tenure. For a postdoc with one paper, the same move can become years of silence and a CV that looks empty. The chapter's energy is real; the risk of imitating a late-career master too early is the shadow.
Cancer under 10 percent five-year survival; Feynman chooses QCD; Mlodinow fears becoming one of the living dead. The book opens with two stuck lives sharing a hallway.
Chapter II: the kibbutz and the quantum
The love affair began in winter 1973 on a kibbutz near Jerusalem, where Mlodinow, aged 20, was volunteering in the labor shortage after the Yom Kippur War. He was a chemistry and mathematics double major who had found high-school physics dull. In the kibbutz library he found Feynman's The Character of Physical Law, lectures from the 1960s, and then the three volumes of The Feynman Lectures on Physics, with the photo of Feynman at his bongos.
The chapter uses that discovery to teach the quantum principle: energy and other quantities come in discrete steps, so the smooth world of everyday life is an illusion of scale. Mlodinow walks through the uncertainty principle, wave-particle duality, the double-slit experiment and Einstein's explanation of the photoelectric effect (the work cited for his 1921 Nobel). He adds a vivid modern example: neutrons bouncing on a mirror in Earth's gravity occupy discrete heights, separated by about 5/10,000 of an inch.
The neutron experiment Mlodinow cites was published by Valery Nesvizhevsky and colleagues at the Institut Laue-Langevin in Nature in January 2002: ultracold neutrons falling toward a horizontal mirror "jump from one height to another, as predicted by quantum theory." The setup has since become a precision tool (the qBOUNCE line of experiments) for testing gravity at micrometre scales. One gap he noted remains as of October 2026: there is still no tested quantum theory of gravity.
He traded his jeans for a copy of the Lectures, switched to physics, and underlined a passage in which Feynman said we live in an age of discovering the fundamental laws, a moment "like the discovery of America" that will never come again.
The chapter is also a conversion narrative. High-school physics had been dull; the kibbutz library made it personal. The underlined line about living in an age "like the discovery of America" is the book's first rainbow: someone else's sense of historical luck that Mlodinow will later have to stop chasing.
| Idea | What the chapter stresses |
|---|---|
| Quantum principle | Energy and other quantities come in discrete steps; everyday smoothness is an illusion of scale |
| Uncertainty | You cannot pin down conjugate pairs (position and momentum) without a trade-off |
| Wave-particle duality | Light and matter show both faces depending on the experiment |
| Double-slit | Interference even when "particles" pass one at a time |
| Photoelectric effect | Einstein 1921 Nobel; light comes in packets |
| Neutron gravity states | Discrete heights about 5/10,000 of an inch (Nesvizhevsky 2002) |
KIBBUTZ-BOOK-SWITCH. Winter 1973, a book (Character of Physical Law), a switch from chemistry-math to physics. The jeans he traded for the Lectures are the receipt.
Pick one popular science book that once changed your direction. Write the one sentence you would underline now, and one sentence on whether that sentence is still your rainbow or someone else's.
Conversion stories flatten cost. Switching fields after a war-volunteer winter sounds romantic; it also abandons a chemistry-math path that might have fit better. The shadow of "the book that changed my life" is sunk-cost conversion: you keep proving the switch was right long after the work has gone cold.
A kibbutz library, Feynman's lectures, and a decision to become a physicist. The quantum world is discrete; gravity still has no tested quantum theory as of October 2026.
Chapter III: Caltech, and the burden of a free hand
Mlodinow came to Pasadena in fall 1981, with Reagan-era cuts biting at Berkeley; his parents' trip to his graduation was the last family event of its kind. His strategy was simple: lie low.
The family note matters. Reagan-era cuts were biting at Berkeley; his parents' trip to graduation was the last event of its kind. Against that scarcity, Caltech offered time to do what you like. Mlodinow's strategy (lie low) is the rational response of someone who knows that freedom will be scored. The chairman's German accent and cigarette pace the tour: Athenaeum winters with Einstein, then the catalogue of prizes and inventions that make the Institute feel like a museum you might stain.
When Sperry's 1981 Nobel arrived, Caltech tied MIT at about 20 prizes despite being roughly one-fifth the size. The statistic is meant to inspire. In the naked-elevator dream it becomes a verdict delivered without clothes.
The division chairman, German accent and cigarette, gave him the tour of glory. Einstein stayed at the Athenaeum for two winters. Caltech's list included Linus Pauling's chemical bond, the Richter scale (Richter and Gutenberg), Gordon Moore's PhD, the discovery of antimatter, pioneering aviation, the age of the Earth, Roger Sperry's split-brain work and Max Delbrück's molecular biology (1969 Nobel).
| Institution | Nobel Prizes (chairman's count) | Relative size |
|---|---|---|
| Caltech | 19, then 20 with Sperry's 1981 prize | 1 |
| MIT | 20 | about 5 |
The chairman's message was freedom: fellows had no duties and could work on anything, even "design sailboats." Mlodinow heard it as pressure. That night he dreamed of being naked in an elevator.
Freedom without a problem is a burden. The chairman's "design sailboats" line is meant as generosity; Mlodinow hears it as a test he can fail in private. The naked-elevator dream is the book's first body metaphor for exposure without cover.
Einstein at the Athenaeum for two winters; Linus Pauling and the chemical bond; Richter and Gutenberg and the earthquake scale; Gordon Moore's PhD; antimatter; pioneering aviation; the age of the Earth; Roger Sperry's split-brain work (Nobel 1981, tying Caltech with MIT at about 20 prizes); Max Delbrück and molecular biology (Nobel 1969). Relative size: MIT about five times Caltech, with a similar Nobel count.
19-to-20, sailboats, naked lift. Caltech's Nobel race with MIT, the free-hand offer, and the elevator dream. Three images, one chapter.
If your boss said you had no duties for a year, list three projects you would start and three fears that would stop you. Circle the fear that matches "lie low." That is Chapter III's diagnosis.
Institutional glory tours ("Einstein slept here") can become a prestige trap: you measure yourself against wall plaques instead of against a problem. The free hand then becomes free fall. Sailboat freedom without a destination is how fellows disappear into polite silence.
Fall 1981, Reagan-era cuts behind him, nineteen (then twenty) Nobels ahead. Freedom felt like pressure. He planned to lie low.
Chapter IV: next door to Murray
His office was next to Murray Gell-Mann's. Murray wore a bolo tie, pronounced foreign words, his visitor's name included, like a native, and, Mlodinow learned, the hyphen in his surname was his father's invention. His achievement was immense: in the 1960s he brought order to the dozens of subnuclear particles then pouring out of accelerators, an achievement Mlodinow compares to Mendeleev's periodic table, and he proposed quarks, which won him the Nobel. He also liked to show it, whether the topic was particle accelerators or septic tanks. When a student tested his claim to know Mayan with a sentence learned on holiday, Murray reproached him: the sentence was in Lower Mayan, and the language Murray knew was Upper Mayan.
Murray had chosen Caltech partly to be near Feynman. In the late 1960s Feynman provided key theoretical evidence that particles exist inside the proton, but because they might not be Murray's quarks he called them partons. Murray was annoyed, and not only by the name (a mixture of Latin and Greek roots). Behind it lay a deeper question Feynman would not touch: are quarks real, or just bookkeeping? Feynman, Mlodinow says, claimed to be under "doctor's orders" not to discuss metaphysics.
| Babylonian (Feynman) | Greek (Gell-Mann) | |
|---|---|---|
| Starting point | Phenomena; what works | Axioms; logical structure |
| Method | Intuition, pictures, rules of thumb | Proof, order, classification |
| Ancestry | Aristotle's empiricism | Plato's ideal forms |
| Brain analogy (Sperry) | Right hemisphere | Left hemisphere |
| Attitude to rules and customs | Disdain | Respect |
| In the department | Avoided leadership | Sought it |
The Babylonian label explains Feynman's greatest work. In 1949, aged about 30, he published papers in the Physical Review that replaced solving differential equations with a sum over histories (paths): a particle goes from A to B by every possible route, each route contributing, and the result is their sum. In his picture an electron can zigzag backward in time; "the temporal order of events is irrelevant." At a 1948 conference Bohr, Teller and Dirac attacked it. With his diagrams Feynman could do in half an hour calculations that took others months, but only when Freeman Dyson showed his method equivalent to the conventional ones did it win acceptance. Murray, the Greek, later wondered whether path integrals were the true foundation of quantum theory.
Mlodinow's own credential was modest by comparison but real. With Nikos (Nick) Papanicolaou he had shown that quantum systems look classical in many dimensions: solve the problem in infinitely many dimensions, then add corrections to get back to three. Edward Witten cited the work in Physics Today, and it drew more than 100 citations. His fear was that it was a fluke.
Home life supplied the counterweight: neighbours Edward (a weapons-research graduate student) and Ray (a garbage collector), joints by the pool, Columbo and The Rockford Files. What finally moved him was chutzpah: he knocked on Feynman's door.
The Mayan anecdote is comic and sharp: when a student tested Murray with a holiday phrase, Murray corrected him to Lower versus Upper Mayan. Showmanship and real range sat in the same person. The parton quarrel is deeper than nomenclature. Feynman would not grant Murray's quarks the status of real constituents; he would only grant evidence for something inside the proton. Metaphysics, he joked, was against doctor's orders.
In 1949, around age 30, Feynman replaced differential-equation machinery with a sum over histories: every path from A to B contributes; the classical path is where neighbouring contributions reinforce. An electron may zigzag backward in time; temporal order is not sacred. Bohr, Teller and Dirac attacked the picture in 1948; Freeman Dyson later showed equivalence to the established formalisms. Murray later wondered whether path integrals were the true foundation of quantum theory. Acceptance followed the Greek proof of a Babylonian invention.
With Nick Papanicolaou he showed that quantum systems look classical in many dimensions: solve at infinity, then correct back toward three. Edward Witten cited the work in Physics Today; it drew more than 100 citations. His fear: a fluke. Home counterweight: neighbours Edward (weapons-research graduate student) and Ray (garbage collector), joints by the pool, Columbo and The Rockford Files. What moved him was chutzpah.
B-vs-G + PATHS + PARTONS. Babylonian versus Greek; sum over paths; partons that would not be called quarks. Chapter IV's triangle.
Label your last hard decision Babylonian (what worked in the data) or Greek (what followed from principles). Then force the other pass: if you were Babylonian, write the axiom you skipped; if Greek, name the ugly fact you ignored.
Babylonian style without Dyson's later proof becomes permanent handwaving. Greek style without phenomena becomes taxonomy of nothing. The rivalry is productive only when each side can fail the other's test. Copying Feynman's disdain for metaphysics as a licence to skip definitions is the shadow.
Next door to Gell-Mann; partons versus quarks; Babylonians and Greeks; path integrals that looked crazy until Dyson made them respectable. Chutzpah opens the door.
Chapter V: lunch at the Greasy
Feynman, 63, was recovering from a 14-hour operation, his second for cancer, and looked like "a retired Brooklyn cab driver." He ate not at the Athenaeum, the faculty club, but at the student cafeteria nicknamed the Greasy, over half-cooked burgers.
Age and body open the scene. Feynman at 63 is about ten years older than Murray, recovering from a 14-hour second cancer operation, looking to Mlodinow like a retired Brooklyn cab driver. He chooses the student cafeteria over the Athenaeum without apology. Appetite is mentioned only to be refused: he used to overeat; he does not anymore.
Koestler's The Act of Creation supplies the lunch topic. Köhler's chimpanzee Nueva (Berlin, 1914) plays with a stick, then, fruit out of reach, suddenly rakes the fruit in. Galileo's Dutch toy becoming a telescope is the human twin. Discovery, on this telling, is not lightning from a private god; it is combination of pieces already on the table. Feynman's first reaction dismisses psychology; his second grants the permission the chapter exists to deliver.
Mlodinow was then reading Arthur Koestler's The Act of Creation, and told Feynman about Wolfgang Köhler's chimpanzee Nueva (1914): given a stick, she played with it idly, then, with fruit out of reach, suddenly used it to rake the fruit in. The discovery, Koestler argued, was the combination of two familiar things, as Galileo turned a Dutch toy into a telescope. Feynman's response was a classic double move: "You're wasting your time... psychology is bullshit," and then, a moment later, the line Mlodinow kept: if an ape can make a discovery, so can you.
Feynman chose the Greasy over the Athenaeum on purpose. Status was optional; half-cooked burgers were not a metaphor he needed. The Koestler story (Köhler's Nueva, 1914, stick plus out-of-reach fruit) is the chapter's engine: discovery as combination, the same move Galileo made when a Dutch toy became a telescope.
First: "You're wasting your time... psychology is bullshit." Then, a beat later: if an ape can make a discovery, so can you. Dismissal and permission in one lunch.
The lunch ends without a research programme and with something more useful: permission. Mlodinow leaves the Greasy knowing that discovery need not wait for a private myth of genius. The next chapter will demystify the rest: scientists as detectives and bodybuilders, not oracles.
APE-COMBO. Nueva's stick plus fruit; Galileo's toy plus sky. Creativity as combination of two familiar things. If an ape can, so can you.
Take two tools you already use at work and force a Nueva combo: write one sentence on a problem neither tool was built for, and how using both might rake the fruit in. Fifteen minutes. No new purchases.
"If an ape can" can become toxic egalitarianism: it denies selection, training and luck. Feynman had unusual gifts; the line is motivation, not a hiring policy. The shadow is telling every stuck junior that effort alone equals Feynman's half-hour calculations.
A 63-year-old after a 14-hour operation, burgers at the Greasy, and a chimp story that ends in permission. Combination is the creative move.
Chapter VI: the detective and the bodybuilder
Mlodinow began taping. Feynman's first message demystified the scientist. Lying takes imagination; everyone infers ("Mary at the Loaf and Ladle"); scientists differ mainly in thinking constructively instead of worrying. Science is detective work, clues from experiments assembled into a story. (Murray, naturally, was the Sherlock type: "Elementary.") On Friday nights the neighbours projected The Hound of the Baskervilles by the pool.
Feynman: creativity "is done by everybody"; scientists simply do more of it, and "only idiots like me" think about the same problem every day. He compared it to bodybuilding: finding how far human potential can go in one direction.
Exceptional output is ordinary capacity trained relentlessly in one direction. It removes the mystique (Einstein's IQ, Bohr's "direct line to God") and replaces it with a question you can act on: which direction, and for how long? Its flip side, which Mlodinow notes, is that many people train hard in the wrong direction.
Feynman's modesty is half true. He did have unusual gifts (the half-hour calculations, the intuition that Dyson needed months to formalise). "Anyone can do it" is good motivation and poor selection advice; the book never tests it, and Mlodinow's own path suggests that direction matters as much as effort.
- The problem is not intelligence but direction: a gifted young scientist with nothing to work on.
- Physics in 1981 had one theory that worked (QED) and one that could not be computed (QCD).
- Feynman's style was Babylonian: what works, by pictures and intuition; Gell-Mann's was Greek: structure and proof.
- Feynman's first lessons: the ape can do it; think constructively; train in one direction.
Taping begins. Feynman's demystification is thorough: lying takes imagination; everyday inference ("Mary at the Loaf and Ladle") is already model-building; scientists mainly think constructively instead of worrying. The detective metaphor (clues from experiment into a story) sits beside Murray as Sherlock saying "Elementary." Friday nights by the pool: The Hound of the Baskervilles.
DETECTIVE + BODYBUILDER. Science as casework; expertise as ordinary capacity trained in one direction. Two images, one chapter.
For one week, when you catch yourself worrying about a problem, write the next constructive question instead (what clue would change my mind?). Count how often worry returns after the constructive question is written.
The bodybuilder model ignores wrong gyms. Training relentlessly in one direction is disaster if the direction is a dead theory (Chew in Chapter XII) or a tuned lattice (Constantine in Chapter XXI). Direction selection is the missing half of the metaphor.
Creativity is common; scientists do more of it. Train in one direction; think constructively; treat science as detective work. Mystique is optional.
Act II. The Forces (Chapters VII to XI)
The second act moves from the person to the work. Mlodinow learns how physicists pick, survive and enjoy problems, through three models: the strong force as a metaphor for Caltech's social pressures, Feynman's account of his own persistence, and John Schwarz's apparently hopeless devotion to string theory.
Chapter VII: the strong force, at work and in physics
The chapter opens at lunch with Stephen Wolfram, who ate a pound of rare roast beef: Oxford-educated, a published paper at 15, a Caltech PhD at 20, later founder of a software company and author of a book on cellular automata. Caltech was full of such people, and full of pressure.
Mlodinow uses the strong force as his metaphor. Without it, every nucleus heavier than hydrogen would blow apart under the repulsion of its protons. Its strangest property is asymptotic freedom: the force between quarks grows with distance. Inside the proton, close together, quarks move almost freely; try to pull one out and the binding gets stronger. "To escape, move closer." Caltech's social forces worked the same way: the closer you were to the centre, the freer you felt, and no one was free to fail.
The property Mlodinow describes was discovered in 1973 by David Gross and Frank Wilczek and, independently, David Politzer (then a Harvard graduate student, later at Caltech). They received the 2004 Nobel Prize in Physics for it, a year after the book appeared. It remains the reason QCD is believed: at short distances it can be calculated and matches collider data closely.
The pressure had a map. At the centre of the hierarchy of respectability sat elementary particle theory; Murray called solid-state physics "squalid state physics." Mlodinow compares it to Saul Steinberg's classic New Yorker cover looking west from Manhattan: particle theory in the foreground like Manhattan's buildings, everything else shrinking into the distance. Feynman simply ignored these forces: he worked on physics in a strip club, did not drink, and paid no attention to status.
Around the metaphor sit two portraits. Stephen Wolfram at lunch ate a pound of rare roast beef: paper at 15, Caltech PhD at 20, later Mathematica (1988) and A New Kind of Science (2002). Neighbour Ray treated a woman problem as a "field of study," talked suicide and a gun, drank martinis. Murray yelled at a bank clerk. The department's social physics was not gentle.
Wolfram Research's Mathematica shipped in 1988; A New Kind of Science appeared in 2002, a year before Feynman's Rainbow. The lunch portrait is of a young prodigy who would build a private computational empire rather than a conventional academic career.
GROW-WITH-DISTANCE. Asymptotic freedom: quarks freer when close, bound harder when pulled. Same map for Caltech status. Escape by moving closer.
Sketch your organisation's respectability ladder (top: glamorous work; bottom: unglamorous dependencies). Mark where you sit. Ask whether the bottom layers are where value actually lives, as solid-state physics was for the transistor.
"To escape, move closer" can justify staying in a toxic centre because periphery feels worse. Helen Tuck's counsel (bounce back, or find something else) is the escape hatch the metaphor hides. Persistence without an exit is how one-trick ponies become gardeners in bitterness.
Asymptotic freedom as social map; particle theory at the centre; Gardening and Constantine; Helen on dry spells. Persistence correlates with graduate success; exit remains allowed.
Chapter VIII: persistence, imagination and the inside track
Feynman had contributed in many fields: low-temperature physics, optics, electrodynamics. Each came after a dry spell. Asked how he found problems, he used the centipede who, asked which leg moves first, cannot walk. Sometimes the creative act is finding the problem; sometimes the problem is obvious and the solution is hard. Einstein's general relativity is the second kind: the problem (reconcile the speed-of-light limit with gravity) was clear; the clue was that a falling person feels weightless.
QCD was the hard kind. Physicists had their "noses rubbed in it." He had tried everything for two years and was now trying to think about how the theory behaves roughly.
To keep going on a hard problem, Feynman said, you must believe you have an inside track, even if that means fooling yourself a little, like warriors who beat drums before battle. Persistence requires being convinced. "I haven't gotten anywhere... it will all be by imagination and persistence." Tenure let him spend years on the hardest problems; he took only those.
Mlodinow draws the lesson that physics is an art, like law or medicine: judging what is essential and what can be approximated. Hydrogen is the only atom solvable exactly, and even hydrogen in a magnetic field cannot be solved, which is where his infinite-dimension method came in.
"Believe you have an edge" is good advice for Feynman and dangerous for most people. The same conviction kept Geoffrey Chew working alone on a dead theory (Chapter XII) and, Feynman would say, the Montreal visitor on his twelve-year theory. The book's answer, believe it only if you can defend it as your own harshest critic, is the necessary counterweight.
The centipede joke lands: ask which leg moves first and walking stops. Finding the problem and solving the obvious hard problem are different arts. Einstein's clue for general relativity (a falling person feels weightless) is the chapter's model of a hard, clear problem.
In the mid-1950s Feynman explained liquid helium (it does not boil; it creeps up beaker walls) by waving hands, analogies, pictures and plausible guesses. Exact solutions were unavailable; judgment of essence was the craft. Hydrogen is the only atom solvable exactly; even hydrogen in a magnetic field is not, which is where Mlodinow's infinite-dimension method entered.
The book says QCD would yield to ever more powerful computers. The BMW collaboration and others computed light hadron masses from lattice QCD around 2008, including a proton mass in agreement with experiment within uncertainties. The "noses rubbed in it" problem of 1981 is no longer blank; it is a computational industry. That does not erase Feynman's point: the breakthrough was imagination plus persistence, then machines.
INSIDE-TRACK + DRUMS. Believe you have an edge; beat drums before battle; tenure buys years on the hardest problems. Pair with "be your own worst critic" from Chapter XII.
Pick one stuck project. Write (a) the story that says you have an inside track, and (b) the harshest one-paragraph critique of that story. Persist only if (a) survives (b).
"Fool yourself a little" without the critic becomes delusion with tenure. Chew's lonely S-matrix persistence (Chapter XII) and the Montreal visitor's twelve-year theory are the same fuel misaimed. Inside-track talk is necessary and dangerous.
Dry spells, the centipede, QCD's uncomputability, liquid helium by pictures. Persistence needs conviction; conviction needs a critic.
Chapter IX: four forces and one dream
Mlodinow lays out his strengths (mathematical, rebellious) and then the physics that will tempt him: unification.
| Force | Relative strength | What it does | Theory in 1981 |
|---|---|---|---|
| Strong | about 100 times electromagnetism | Holds nuclei; powers atomic energy | QCD |
| Electromagnetic | about 1,000 times the weak force | Atoms, materials, chemistry, technology | QED, part of electroweak theory |
| Weak | billions cubed times gravity | Beta decay | Electroweak theory |
| Gravity | weakest | Stars, planets, the cosmos | General relativity (not quantum) |
The stakes are concrete. Less than a third of an ounce of uranium converted to energy destroyed Hiroshima. Electromagnetism is so strong that if electrons and protons in two people a yard apart were out of balance by 1 percent, the force between them would exceed the weight of the Earth.
Unification has a history. Thales knew of magnetism and static electricity; around 1820 it was found that currents create magnetism; in 1865 James Clerk Maxwell (five foot four, Mlodinow notes) unified electricity and magnetism and found that light is an electromagnetic wave travelling at light speed, which led to Einstein and to Hertz's radio. Feynman called Maxwell's laws the most significant event of the nineteenth century. By 1981 electromagnetism and the weak force had been unified as the electroweak theory; QCD was separate; gravity resisted everything. A unified field theory, Einstein's dream, would show all forces flowing from a single principle, a Greek aspiration.
String theory, the laughing stock
String theory, proposed in 1970 by Yoichiro Nambu and Leonard Susskind, replaces point particles with tiny vibrating strings; each mode of vibration, like a note on a violin string, is a different particle. Its great promise is that it includes gravity. Its problems were that it could not be tested directly, that its best hope was "postdiction" (deriving known numbers such as the electron's charge or the quark masses), that it was riddled with inconsistencies in the 1970s, and that it required six extra, tiny dimensions. Almost everyone dropped it except John Schwarz and a tiny group; colleagues treated it like Area 51.
A Caltech research fellow, not a professor, after nine years without tenure ("He's a mule," Constantine said). Asked what he worked on, Schwarz said "string theory," and handed Mlodinow preprints instead of explaining: "I don't really have the time... just read these." Within a few years, Mlodinow notes, his work would be revered as a breakthrough.
Mlodinow's self-portrait (mathematical, rebellious) explains why Schwarz's outsider math appealed. Unification is the Greek dream: one principle under all forces. The history lesson runs from Thales to Maxwell (1865, five foot four) to electroweak theory. Feynman called Maxwell's laws the most significant event of the nineteenth century.
Nambu and Susskind, 1970: particles as vibrating strings; modes like violin notes. Promise: includes gravity. Costs: untestable at ordinary energies; hope of "postdiction" (electron charge, quark masses); 1970s inconsistencies; six extra tiny dimensions. Almost everyone left; Schwarz stayed nine years without tenure. Constantine: "He's a mule." Colleagues treated it like Area 51. Within a few years the joke would invert.
Michael Green and John Schwarz shared the 2014 Breakthrough Prize for anomaly cancellation and the first superstring revolution. As of 2026 the theory still has no confirmed experimental test; characteristic effects sit at unreachable scales, and critics still call it "not even wrong." David Gross received a 2026 Special Breakthrough Prize citing lifetime work including heterotic strings; that is honour for theory, not evidence from nature. Quanta's March 2026 survey records the same stalemate Mlodinow already smelled in 1981: beauty without a checked prediction.
4-FORCES, 6-DIMENSIONS, 1-MULE. Four forces; six curled dimensions; one research fellow who would not leave. Chapter IX's census.
List a "string theory" in your field: an elegant framework that includes everything and predicts nothing you can check this year. Write one postdiction it already fits and one near-term test that would change your mind.
Unification lust can starve near-term work that actually ships. (enterprise ontologies, AGI roadmaps) that absorb talent while operational messes sit unfixed. The shadow of Einstein's dream is neglect of the transistor layer.
Four forces, Maxwell's merger, electroweak success, gravity the holdout. String theory in 1981 was a laughing stock with one mule in the stable.
<
div class="fig af">
Chapter X: nonsense, and when to work on it
Schwarz had four or five graduate students, most of the department's string theorists; 99.9 percent of physicists, Mlodinow estimates, thought them crackpots. Feynman's view at a Saturday talk was precise: work on a theory everyone thinks is nonsense "only if you don't think it is nonsense." He added that too much education can cause trouble, since great discoveries are often made young (Newton's calculus, Einstein's relativity, his own diagrams, all at roughly the same age), and that he always derived results himself, keeping a beginner's outlook.
The numbers are meant to sting. Schwarz's group is most of the department's string effort and still a rounding error against the field: Mlodinow's 99.9 percent crackpot estimate. The Saturday talk answers the moral question underneath the joke. Unpopular work is allowed only when you do not think it is nonsense. Education, past a point, can erase the beginner's licence that let Newton, Einstein and Feynman hit their signature results young. Deriving everything yourself is how he kept that licence.
The Athenaeum wedding buffet is not a side gag. It is the same rule in social form: ask whether the posted boundary is real, then act if it is not. "We represent the physics department" is both joke and method.
The chapter ends in comedy that is also a lesson in ignoring rules. Walking past a wedding buffet in the Athenaeum courtyard, Feynman asks whether not being invited really means you aren't welcome, and joins the line. Asked "Bride's side or groom's?" he answers: "We represent the physics department."
99.9 percent crackpot odds, four or five graduate students, and a Saturday rule that became the book's ethic: work on nonsense only if you don't think it is nonsense. Too much education can blunt the beginner's eye; Newton, Einstein and Feynman hit their signature results young. He kept deriving results himself.
Walking past an Athenaeum wedding buffet, Feynman asks whether not being invited really means you are unwelcome, joins the line, and when asked "Bride's side or groom's?" answers: "We represent the physics department." Rule-breaking as method, not as lifestyle brand.
ONLY-IF-YOU-DON'T. The whole chapter collapses to one clause. Pair it with "be your own worst critic" (XII).
Name one idea your peers call nonsense. Write two columns: reasons they are right; reasons you still do not think it is nonsense. If column two is empty, drop it. If column two is only status anxiety, drop it.
"Only if you don't think it is nonsense" can become a filter bubble: you recruit students who already agree, and the 99.9 percent outside never get a vote. Schwarz needed Murray's protection and, later, a real calculation (anomaly cancellation). Belief is necessary; contact with hostile seminars is the disinfectant.
Work on unpopular ideas only when you yourself do not think they are nonsense. Keep a beginner's derivation habit. Sometimes join the buffet.
Chapter XI: play, and changing the laws
This is the book's chapter on play. In high school Feynman invented the half-derivative, defined only by its properties; in university he checked it; at Los Alamos he used it to solve an equation numerically. Mlodinow gathers other voices on the creative "attic": Tchaikovsky, Mary Shelley (invention comes "out of chaos"), the poet Stephen Spender.
Feynman's favourite game was changing the laws. He kept a notebook on physics in two space dimensions, where atomic spectral lines would differ. He played with physics with two times with his son on the beach. Murray's quarks came from the same move: "What if everything were made of three particles?" The K-meson would not fit unless the particles had fractional electric charges, which seemed absurd: since George Stoney coined "electron" in 1891, every observed charge had been a whole multiple of it.
So unsure was Murray that he avoided the prestigious Physical Review fearing its editors and referees, and published in a journal of lesser prestige (Physics Letters). Feynman was sceptical at first and admired the idea later. The lesson for Mlodinow: even giants are unsure at the moment of creation, and the idea that later wins a Nobel can look like a joke when it is new. He began to wonder: why not more dimensions?
A practical creativity drill from Feynman's notebooks: take a system you know, change one fundamental rule (two dimensions instead of three, fractional charges, two time axes) and work out the consequences carefully. Most variants teach you why the real rule holds; occasionally one turns out to be the real world.
- Strong forces bind hardest at a distance: in physics (asymptotic freedom) and in institutions.
- Persistence needs conviction, and conviction needs honest self-criticism to stay sane.
- Unification is a Greek dream; in 1981 string theory was its least respected candidate.
- Play (changing the laws, inventing a half-derivative) is how Feynman and Gell-Mann both found real things.
Play is not break time; it is how Feynman trained. The half-derivative he invented in high school, checked in university, and used at Los Alamos for a numerical solve is the emblem. Creative "attic" voices (Tchaikovsky, Mary Shelley "out of chaos," Spender) sit beside the physics notebook with two space dimensions and the beach game with two times with his son.
Murray asked: what if everything were made of three particles? The K-meson would not fit without fractional charges. Since Stoney coined "electron" in 1891, every observed charge had been an integer multiple. Murray hedged: he avoided Physical Review and published in Physics Letters. Feynman was sceptical, then admiring. Tentative publication of a crazy-looking idea later won a Nobel.
CHANGE-ONE-LAW. Two dimensions; two times; three quarks. Vary one rule; work it out to the end.
Take a system you know (a payment flow, a model pipeline, a team ritual). Change one fundamental rule and write consequences for a page. Most variants explain why the real rule holds. Circle any variant that deserves a real experiment.
Endless law-changing without finishing the calculation is aesthetic doodling. Feynman worked variants carefully to the end. Play that never closes becomes the attic Mary Shelley warned against: chaos without a creature.
Half-derivatives, two-dimensional notebooks, quarks from "what if three?" Play is serious when you finish the sum.
Act III. Rainbows (Chapters XII to XVI)
The third act is about choosing. Mlodinow watches three ways of living a scientific life up close: Schwarz's stubborn devotion, Murray's Greek grandeur and private grief, and Feynman's insistence that the point is beauty and fun.
Chapter XII: the detective picks a case
Spring came. Ray met a woman in Bellevue, Washington, who worked for "a small software company called Microsoft." Explaining physics to Ray, Mlodinow reaches again for the detective metaphor, now with a twist: a detective must first choose the case. You can chase the Ten Most Wanted, as everyone else does, or find a problem only you realise is important, and then chase the leads with mathematics. He quotes Einstein, by way of Zen and the Art of Motorcycle Maintenance, on the scientist's urge to form "a simplified and intelligible picture of the world."
In a CERN cafeteria, Schwarz and Michael Green realised that string theory is a theory of gravity. It made no front pages because nobody believed it. Schwarz gave more seminars than anyone; about ten people would come. His Berkeley adviser, Geoffrey Chew, had championed S-matrix theory, once the hottest idea in particle physics; when the field abandoned it, Chew kept going alone and, in Mlodinow's words, ended his once brilliant career in oblivion. Schwarz had seen what persistence can cost.
Caltech seminars were brutal. Murray might challenge a speaker on the tiniest point or, if bored, pull out a newspaper and read it conspicuously; Feynman was brash and unwilling to accept sloppy thinking. At least one future Nobel laureate hesitated to lecture there. Before one seminar Feynman needled Murray over his insistence on pronouncing Montreal "Mon-ray-al" ("Where's that?"), Murray noted that there were many well-known cities Feynman did not recognise; Feynman replied that, logically, either he was an ignoramus or Murray said them in a funny way. "Logically speaking, it could also be both," said Murray. Then Feynman turned on the speaker of the day: "Hey Schwarz, how many dimensions are you in today?" Afterwards he called it "the only nonsense theory we've got going in this department." Ray asked why, then, he had come. "I came for the cookies."
A visitor from Montreal had spent twelve years on one theory. Feynman's verdict on such efforts: be your own worst critic; do not keep at it unless you really believe it. "That's exactly what I mean about wasting your time." Privately he had "recognised him as a pompous ass."
The chapter also shows Feynman unguarded: he had tried marijuana and LSD; he found the Nobel Prize unfair and a distraction, and told the reporter who woke him with the news to call back later. And he had talked with Ray, telling him what physicists do and that he must "find your own answer."
Spring. Ray met a woman in Bellevue, Washington, working for "a small software company called Microsoft." The detective metaphor sharpens: you must choose the case. Better to find a problem only you realise is important than to join the Ten Most Wanted everyone already chases. Leads are chased with mathematics.
Schwarz and Michael Green realised at a CERN cafeteria that string theory was a theory of gravity; it was not front-page news because nobody believed. Schwarz gave more seminars than anyone; about ten people came. His Berkeley advisor Geoffrey Chew had ridden S-matrix theory when it was hot, then persisted alone into what the book calls oblivion. Caltech seminars were brutal: Murray read the newspaper; Feynman marched out; a future Nobelist hesitated to lecture there.
Feynman: "Hey Schwarz, how many dimensions are you in today?" Also: "It's the only nonsense theory we've got," and "I came for the cookies." A Montreal visitor with a twelve-year theory got the critic's full force: do not waste your time unless you really believe; Feynman "recognized him as a pompous ass." Feynman had tried marijuana and LSD; he found the Nobel a distraction and told a reporter to call back. Ray's summary: Feynman told him what physicists do; "find your own answer."
PICK-THE-CASE. Detectives who only work the Ten Most Wanted never find the case only they see. Choose, then criticise yourself.
List the Ten Most Wanted problems in your field. Cross out any you chose because others chose them. Circle one problem only you seem to care about. Write the harshest paragraph against working on it. Decide.
"Find a problem only you see" can become contrarianism for brand. Chew's lonely persistence is the cautionary twin of Schwarz's. The difference, the book implies, is whether hostile contact and self-criticism still bite.
Choose the case; Schwarz and Green see gravity in strings; Chew's warning; be your own worst critic; find your own answer.
Chapter XIII: a theory found by accident
A Berkeley classmate's string theory textbook, projected in 1988 for June 1989 "plus or minus one month," came out in 1998. That joke frames Mlodinow's point: nobody fully understands string theory. It was not derived from a principle, as relativity was from the constancy of the speed of light; it was found by accident, like penicillin, and physicists since have been like paleontologists digging out a skeleton one bone at a time.
| Year | Event |
|---|---|
| Summer 1967 | Gell-Mann lectures at Erice (Ettore Majorana centre) on regularities in S-matrix data |
| About a year later | Gabriele Veneziano, an Italian graduate student working in Israel, finds a formula that fits, with no physical derivation |
| 1970 | Nambu and Susskind show the formula arises if particles are vibrating strings |
| 1970s | Inconsistencies; nearly everyone leaves |
| 1974 | Schwarz and Joel Scherk propose strings as a theory of gravity (Breakthrough Prize citation) |
| 1980 to 1984 | Schwarz and Green develop superstring theory |
Mlodinow pressed Feynman. What does string theory promise? The mass of the proton? "No." Its only prediction was ten dimensions, and those were explained away by rolling six of them up. Feynman's answer is the book's sharpest statement of the Babylonian creed: "I don't want anything. Nature has nothing to do with what I want!... I don't tell nature what to do." He suspected Mlodinow of something more personal: "You hit a dry spell... scrambling." Choosing a research problem, he said, "isn't like climbing a mountain"; if Mlodinow really believed in strings, he would come in excited, not asking permission. Helen added that string theory was "just as bad" as philosophy, and that Schwarz survived because Murray protected him.
Feynman's objection (no principle, no prediction) was strong in 1982 and remains the central criticism. But his test, only principle or observation leads to the right theory, would also have rejected quarks in their first years, when they had neither direct observation nor a deriving principle. Mlodinow lets both men be partly right.
A Berkeley classmate's string textbook, projected for June 1989 "plus or minus one month," appeared in 1998 (eleven years late). Nobody fully understands string theory; like penicillin, it was found by accident. Paleontologists digging one skeleton.
| When | Who / what |
|---|---|
| Summer 1967 | Murray lectures at Erice (Ettore Majorana centre) on S-matrix regularities |
| About 1968 | Gabriele Veneziano finds a formula that fits (math that works, no derivation) |
| 1970 | Nambu and Susskind: vibrating strings underlie the formula |
| 1970s | Exodus; nearly everyone leaves |
| 1974 | Scherk and Schwarz: strings as a theory of gravity |
| 1980 to 1984 | Schwarz and Green on superstrings |
| 1984 | Anomaly cancellation; Witten converts |
Feynman to Mlodinow: what does it promise? Mass of the proton? No. Only prediction: ten dimensions, explained away by rolling up. "I don't want anything. Nature has nothing to do with what I want! ... I don't tell nature what to do." On dry spells and scrambling: selecting a research problem is not like climbing a mountain. If you really believed, "you'd come telling me." Helen Tuck: string theory "just as bad" as philosophy; Schwarz was protected by Murray.
ACCIDENT-NOT-WANT. Found like penicillin; do not tell nature what to want. Two checks before you join a beautiful theory.
For a framework you admire, write what it predicts that could fail this year. If the list is empty, label it postdiction-only and decide whether you are doing science or aesthetics.
"Found by accident" can excuse never stating a principle. Veneziano's formula worked; that is not yet a theory of nature. Accident is a beginning; Feynman's demand for principle or observation is the gate.
Veneziano to strings by accident; Feynman refuses to want what nature has not shown; Helen ranks strings with philosophy; Murray's protection keeps Schwarz alive.
Chapter XIV: Murray, the Greek in mourning
Murray's world was not as orderly as his physics. His daughter Lisa had joined a Marxist-Leninist party and admired Albania; Murray derided Reagan as "Ray-gun," but Lisa crossed a line by singing "Down with Ronald Reagan, Chieftain of Capitalist Reaction!" They had been estranged for years. Cranks haunted top departments (one at Berkeley with a knife; at Columbia a gunman killed a secretary). Mlodinow saw Murray broken: his wife Margaret had died of colon cancer that spread to her liver. Murray had fought it with experimental treatments, as Feynman, by contrast, accepted the data on his own disease.
They bonded over birds: Mlodinow's brother Steve had written a book on Chicago-area birds, and Murray was a devoted birder. Then Murray gave the Greek case for strings.
| Gell-Mann | Feynman | |
|---|---|---|
| Why care | It could unify all forces, Einstein's dream | Only a principle or an observation leads to the right theory |
| Status of the idea | "Too beautiful not to be" true; a discovery, not an invention | A theory with no predictions |
| What he did | Brought Schwarz to Caltech about ten years earlier; kept Caltech as a "nature reserve for endangered theories" | Mocked it in seminars; respected Schwarz |
| Style | Greek: beauty and structure | Babylonian: what nature shows |
Mlodinow wants a peacemaker, someone like Dyson who reconciled Feynman's methods with the standard ones, and finds the metaphor in quantum physics itself: particle and wave are both true. At 52, after Margaret's death, Murray had mellowed. Schwarz was promoted to senior research fellow on a three-year contract: progress, of a sort.
Murray broken: wife Margaret dead of colon cancer that spread to the liver; he chased experimental treatments while Feynman accepted data. Daughter Lisa in a Marxist-Leninist party, Albania fan; "Ray-gun" for Reagan. Cranks at top schools (Berkeley knife; Columbia gun that killed a secretary). Murray the birder recognised Mlodinow's brother Steve's Chicago bird book. On strings: unify all forces, Einstein's dream; he brought John about ten years earlier; too beautiful not to be; discovery not invention; Caltech as a "nature reserve for endangered theories."
| Feynman | Murray (here) | |
|---|---|---|
| What leads to the right theory | Principle or observation | Beauty that feels like discovery |
| Role toward Schwarz | Mockery and cookies | Protection and a nature reserve |
| Grief style | Accept the data | Fight with experimental treatments |
TOO-BEAUTIFUL / NATURE-RESERVE. Murray's two phrases for why strings lived at Caltech through the lean years.
Name one idea you protect because it is beautiful. Write the observation that would make you withdraw protection. If you cannot name one, you are running a nature reserve without an exit criterion.
"Too beautiful not to be" is how entire fields freeze around an aesthetic. Nature reserves for theories can become tax shelters for unfalsifiable work. Murray's protection of Schwarz was historically productive (1984); the same instinct without a later miracle is Chew.
Grief, beauty, and a nature reserve. Schwarz promoted to senior research fellow on a three-year contract. Murray at 52, mellowed and mourning.
Chapter XV: computers, spiders and a rainbow
Constantine had made his name doing lattice QCD: replacing continuous space with a grid of points so a computer can grind out what the equations cannot be solved for by hand. He had computed the mass of the proton this way. "Computers are the future." Ray agreed: computers would augment humans, like HAL in 2001. Mlodinow distrusted black-box numerical answers, which you cannot follow step by step, but weighed the career advantages of being "in."
Then the title scene. Mlodinow found Feynman gazing at a rainbow. Murray, he knew, could recite rainbow myths: Navajo good fortune, a bridge between living and dead, the gold at its end that only a naked man can reach. Mlodinow, the student, explained Descartes's geometry: each colour reaches the eye along a cone at a fixed angle, and the whole bow can be worked out from a single drop.
Feynman cared less for the geometry than for the motive. Descartes's inspiration, he said, "was that he thought rainbows were beautiful." And: "Remember, it's supposed to be fun."
Constantine bets on lattice QCD: computers are the future; he has computed a proton mass. Ray imagines computers that will "augment" humans like HAL. Mlodinow distrusts black-box numbers but weighs career ("in"). Spider Guy keeps hundreds of spiders in plastic cups on card tables and sleeps under them. A 7.2 quake near Eureka shakes the week.
Murray knows myths: Navajo good fortune; a bridge between living and dead; gold at the end only a nude man reaches. Descartes explained the rainbow; Mlodinow recalls the cone and single-drop geometry. Feynman's line lands as the book's title ethic: his inspiration was that he thought rainbows were beautiful. Then the rule under everything: Remember, it is supposed to be fun.
Lattice groups (notably BMW, 2008) obtained light hadron masses consistent with experiment. Separately, Feynman's 1981 Endicott/"Simulating Physics with Computers" keynote argued that classical computers are the wrong machines for quantum systems, a seed of quantum computing. The 2025 Nobel Prize in Physics went to John Clarke, Michel Devoret and John Martinis for superconducting circuits that made quantum bits controllable, a line of work that cites that seed. Computers did become the future; so did a new kind of computer Feynman sketched.
LATTICE / SPIDERS / RAINBOW. Computation, eccentricity, and beauty that is supposed to be fun.
Before your next analysis that "agrees with the known answer," write the tuning knobs you touched. If you cannot list them, you are at risk of Constantine's move (Chapter XXI). Then look at one beautiful thing in the problem and ask Feynman's question: is it still fun?
"Computers are the future" without understanding is how black boxes ship. "It is supposed to be fun" without standards is how hobbies eat careers. Hold both lines at once.
Lattice QCD, Spider Guy, a quake, Descartes's rainbow, and Feynman's order to remember fun. Beauty starts the work; experiment finishes it.
Chapter XVI: still a child
Feynman avoided administration; junior physicists who had worked with him received a form letter two years after leaving Caltech saying he would no longer write them recommendations, since he had not followed their research. Even dying he remained a child, the same boy who had yelled fake curses in made-up Italian at Brooklyn drivers fifty years earlier. As a boy he had a home lab: a photocell that rang a bell, radios repaired for neighbours during the Depression. It was all playing.
Avoiding administration included a form letter that refused recommendations two years after someone had left. The child Feynman is not childish about physics; he is allergic to the adult theatre around it. Fake Italian curses at Brooklyn drivers sit beside the Depression kid who repaired radios and built a photocell bell because the work was playing.
The MIT exit from pure mathematics is a values scene. Higher math that exists only to teach more higher math failed his test. He wanted facts that had been proved, and the pleasure of finding them, more than the machinery of proof. Physics won because it was more fun than anything else, a criterion Chapter XXI will turn into a self-graded exam.
At MIT as a freshman he asked a mathematics professor, "What is the use of higher mathematics if not to teach more higher mathematics?" The reply: "If you have to ask, don't go on in mathematics." He did not. What interested him were the facts that were proved, not the proofs; physics was more fun than anything else.
A Babylonian rule for any technical field: learn what is true and how to use it before learning how to prove it. Proof is for when the stakes need certainty; intuition about verified facts is what makes you fast. The Greek counter-rule, which Murray embodies, is that structure reveals what intuition misses. A good team needs both.
- Choose the case before chasing the leads; the best problems are the ones only you see.
- String theory was found, not derived: beautiful to the Greeks, empty to the Babylonians.
- Persistence can end in a breakthrough (Schwarz) or in oblivion (Chew); conviction must be honest.
- The rainbow lesson: beauty is the motive, and fun is the fuel.
Feynman avoided administration; he used a form letter refusing recommendations two years after someone left. He remained, in Mlodinow's eyes, a child: fake Italian curses at Brooklyn drivers; a kid who built a photocell bell and repaired radios in the Depression because it was "playing." At MIT, a freshman math course pushed him out of pure mathematics: "What is the use of higher mathematics if not to teach more higher mathematics?" The answer he got: "If you have to ask, don't go on in mathematics." He wanted facts proved, not proofs for their own sake; physics was more fun than anything else.
PLAY-NOT-ADMIN. Radios in the Depression; form letters for recommendations; physics over pure proof. Stay a child at the work, not at the responsibilities you owe people.
Block two hours this week for play on a real problem (no slides, no status update). Separately, answer one recommendation or review you owe. Do not let play become an excuse to stiff people.
"Still a child" romanticises refusal of stewardship. Organisations need Gell-Manns who protect endangered work and answer mail. A department of pure Feynmans is a myth and a management failure.
Play built the physicist; administration never did. Interest in facts over formalisms kicked him from math to physics.
Act IV. What Matters (Chapters XVII to XXIV)
The last act compresses a lifetime of questions into a few weeks. A sneering lunch, a cancer diagnosis that turns out to be wrong, a friend caught fudging, and a final exam from Feynman force Mlodinow to decide what he actually values. The epilogue follows Feynman to his death and string theory to its triumph.
Chapter XVII: the God of Experiment
What Mlodinow calls the worst day of his life began late, in a tradition that goes back to Descartes, who never rose before noon. Caltech offered distractions grad school never had: learning about Mexican professional soccer from the friendly night janitor, and a secret project, a parody screenplay of The Hound of the Baskervilles in the style of Airplane! (which he had seen five times). He had been writing since he was nine; this effort, he admits, ranked "below zero" on the lowbrow scale.
At lunch his former undergraduate mentor, "Professor Breadcrumb" (for the crumb that settled in his beard), lectured him on managing his career and keeping his research "coherent," asked what else he would possibly do, and called the screenplay "worthless Hollywood crap," an insult to the gifts Mlodinow owed it to himself to use.
Feynman, asked about fiction, was more interesting. An English professor once told him, "All you have to do is write." He tried a Grimm-style fairy tale and found he could only recombine what he already knew.
In science, Feynman argued, you must imagine what is actually there, and your imagination is checked. The God of Experiment says "incorrect," or at best "so far so good"; an idea can be lovely and still fail: "That's pretty, my friend, but it's not real." Artists have no such judge; imagine, he said, a great God of Aesthetics who pronounced every painting "good" or "bad" no matter how it satisfied you, and the artist's task would become developing a sense that fits the god, which is the scientist's situation. And science accumulates, "a big monstrous being built by people together," while each novelist starts over. He loved Madame Bovary's craftsmanship and could not stand modern novels.
Feynman's contrast flatters physics. Literature has its own checks (readers, time, the canon) and its own accumulation (forms, techniques, genres). The useful core is narrower: ideas that face an external test improve faster.
"Worst day of my life" opens the chapter. Late hours are tradition (Descartes never rose before noon); Caltech distraction includes a night janitor and Mexican soccer. Mlodinow is writing a parody screenplay of The Hound of the Baskervilles in the Airplane! style (he had seen that film five times); he has written since age nine; the screenplay sits "below zero on the lowbrow scale."
Lunch with a former undergrad mentor: manage your career, keep coherence; "What else would you do?" On the screenplay: you owe it to yourself not to waste talent on "worthless Hollywood crap." The insult names the book's tension: physics prestige versus the writing life Mlodinow already loves.
Scientific imagination is harder than artistic imagination because it is checked. You imagine what is there; the God of Experiment says "incorrect" or, at best, "so far so good." There is also a God of Aesthetics in thought experiments, but science is cumulative ("a big monstrous being built by people together") in a way a single novel is not. Feynman liked Madame Bovary and could not stand modern novels. His one try at fiction (recombining Grimm) failed.
CHECKED. Art can be only beautiful; science must survive "incorrect." So far so good is the highest grade.
Take one idea you have not tested. Write the cheapest experiment that could return "incorrect." Run it or schedule it. If you cannot design a failing test, you are doing aesthetics.
Worship of the God of Experiment can become philistinism: Breadcrumb's insult as policy, killing every side project that is not a Physical Review letter. Mlodinow's later writing career is the rebuttal. The god judges scientific claims, not whole lives.
Breadcrumb's insult; fiction versus checked imagination; "so far so good." Science is collective and falsifiable; that is why it is hard.
Chapter XVIII: the lumps
That afternoon a doctor Mlodinow calls "Dr. Smiley-Face" found lumps on both testicles, "almost certainly cancerous," rare enough to be "publishable." Mlodinow's maternal grandmother had died of cancer at 40 in 1930s Poland; the rest of the family had been killed by Hitler. A blood test came back negative, which, the doctor said, "doesn't mean anything." The plan was to remove both testicles without a biopsy.
He reacted like a physicist and a patient at once. He walked in a daze, cried, found the olive trees on campus suddenly precious. He went to the USC medical library to read up, consoled by a study claiming a physics PhD is on average about 25 percent smarter than his doctor; the books backed the doctor. He thought about the two ways cancer had been met at Caltech that year: Feynman's calm acceptance of his data, Murray's fight for Margaret.
Ray revealed why he was good at mathematics and hated it: his father, an engineer from Spanish Harlem, hit him when he got arithmetic wrong. "That's why I hate it and that's why I'm good at it." Mlodinow saw himself in both of them, forcing himself and beating himself up for not producing.
Dr. Smiley-Face, past president of a professional society, finds lumps on both testicles: "almost certainly cancerous," rare enough to be "publishable." Family history: mother's mother died of cancer at 40 in 1930s Poland; screams; Hitler killed family. That year at Caltech, two cancer styles: Feynman's calm acceptance versus Murray fighting. Blood test negative "doesn't mean anything"; the plan was to remove both testicles without biopsy. Mlodinow reads USC medical books; a study claimed PhDs are about 25 percent smarter. He feels like a cat's mouse. Ray's father (engineer from Spanish Harlem) hit him for wrong arithmetic: "That's why I hate it and that's why I'm good at it." Mlodinow sees himself as both Ray and the father: forcing himself, beating himself up.
The olive trees outside, suddenly precious, and Feynman staring at a rainbow earlier in the year, become the chapter's visual anchors. Blood work that "doesn't mean anything" and a plan for irreversible surgery without biopsy push him into the USC medical stacks. The cat-and-mouse feeling is not melodrama; it is what "almost certainly" does to a mind trained to trust experts.
ALMOST-CERTAINLY. The phrase that empties the week. Pair with Chapter XX's "saved by symmetry."
When you next hear "almost certainly" from an expert, ask for the base rate, the alternative explanations, and whether a less irreversible test exists. Write the three answers before consenting to the irreversible step.
Medical and scientific authority share a failure mode: rare cases that flatter the physician ("publishable") can rush irreversible action. The shadow of expertise is narrative capture by an interesting diagnosis.
Bilateral lumps, family cancer history, USC library, and a mind that punishes itself. The scare begins.
Chapter XIX: the support group of two
He went to Feynman, who said "Go away," and then got in with a ploy: a quantum optics question Feynman surely "hadn't thought about in years." With Mark Hillery, then in New Mexico, Mlodinow had been modelling laser light in crystals. The accepted theory treated the crystal as a continuum with bulk properties like density and refractive index. Mlodinow's infinite-dimension method modelled the individual atoms, and the results conflicted with a 15-year-old paper.
Cancer worry makes the visit. "Go away" is the first line; the quantum-optics ploy is the key. Mark Hillery in New Mexico studies laser light in crystals. Mlodinow's infinite-dimension method treats atoms individually where the accepted continuum theory uses bulk density and refractive index. Their results disagree with a paper about fifteen years old. Feynman's half hour re-derives the living core and discards the rest as old bullshit. The support group does not discuss tumours. It discusses whether a calculation can be owned.
In thirty minutes Feynman gave him more insight than two months of work. Asked whether he had followed the old paper, he said that following was not the point: "When you can derive it yourself, then you understand it." Of the established approach: "Not only is it bullshit, it is old bullshit."
The talk worked better than any therapy: for an hour Mlodinow forgot he might be dying.
He visits Feynman as a "support group of two." "Go away" is met with a ploy: a quantum optics question Feynman "hasn't thought about in years." Mark Hillery (New Mexico) works laser light in crystals; Mlodinow's infinite-dimension method models individual atoms against an old continuum theory (bulk density, refractive index); results conflict with a fifteen-year-old paper. Feynman gives thirty minutes that beat two months: "When you can derive it yourself, then you understand it"; the old paper is "not only bullshit, it is old bullshit." Work displaces terror.
DERIVE-IT-YOURSELF. Thirty minutes beat two months. Understanding is reproduction, not assent.
Take one result you cite weekly. Close the paper and derive it on a blank page. Where you stall is where you do not understand. That stall is the syllabus.
Using work to escape fear can become denial with a notebook. The chapter's gift is real; it is not a substitute for medical second opinions (Chapter XX). Derive the optics; still see the doctors.
A support group of two; quantum optics as lifeline; derive it yourself; old bullshit named. Terror pauses.
Chapter XX: saved by symmetry
Three doctors later, the specialists' verdict: the lumps were mirror images of each other, congenital, not tumours. "Saved by symmetry." The fear did not leave entirely for years.
The medical plot resolves in one short chapter because the evidence is geometric. Three doctors, then specialists: the lumps are mirror images of each other, congenital, not tumours. What looked like a rare bilateral cancer was a developmental symmetry the first physician had narrated as "publishable." Mlodinow was saved by symmetry, and he says the fear still returned for years. The chapter's length is part of its point: when the right observation arrives, the story collapses to a sentence, and the nervous system takes much longer to catch up.
The first verdict skipped the bilateral-mirror pattern that congenital features often show. Irreversible surgery (remove both testicles, no biopsy) was on the table before the geometric alternative was heard. Second opinions were not politeness; they were the God of Experiment applied to medicine.
Three doctors. Specialists: the lumps are mirror images, not tumours, congenital. He was saved by symmetry. Fear lingered for years.
What the chapter teaches beyond medicine is epistemic: a rare bilateral pattern flattered a publishable story and hid a common developmental one. Symmetry was not mysticism; it was the observation the first narrative had no room for. Years of lingering fear after a clean verdict are part of the data too.
MIRROR. Congenital bilateral symmetry, not cancer. The universe's joke after "almost certainly."
In your next design review, ask whether a scary anomaly is a symmetric feature of the system (expected in pairs, inherited from setup) before you treat it as a one-sided failure. Document the check.
"Saved by symmetry" can become magical thinking: you expect aesthetics to cancel every threat. Sometimes the mirror is cancer on both sides. Symmetry was evidence here, not a charm.
Mirror-image congenital lumps. Saved by symmetry. Relief that does not erase vigilance.
Chapter XXI: followers, and a final exam
Elated, he told Ray, and they partied for two days on pizza, burgers, joints and beer. Then Ray dropped his own news: he was moving to Bellevue to be with the Microsoft woman and learn computer programming, finally using his mathematical talent.
At Caltech, a scandal. Constantine, about to leave for a job in Athens, was rumoured to have computed the proton mass by working backward: knowing the answer, he had tuned how he set up the lattice until it came out right. Lattice theory involves choices (there is no unique way to turn continuous equations into a grid), so the difference is subtle, "but it is important to disclose." Constantine waved it off ("Everybody does that"), puffing short, joyless puffs.
Feynman would not psychoanalyse Constantine. What should bother you as much as whether your friend fudged his work, he said, "is that a lot of people read it and couldn't tell the difference... too many followers, too few leaders."
Tuned or not, lattice QCD has since become a test of honest computation at the frontier. In June 2025 Fermilab's Muon g-2 experiment released its final value of the muon's anomalous magnetic moment, precise to 127 parts per billion. The 2025 Theory Initiative prediction, which for the first time used lattice QCD for the hardest hadronic piece, agreed with it, erasing a much-publicised discrepancy; the theory uncertainty is still about four times the experiment's, and tensions remain among data-driven methods. Independent lattice groups are the institutional answer to Constantine's temptation.
Mlodinow told Feynman about his scare. Feynman told him about his own surgeon, the one doctor in the country brave enough to operate: "Of course, chances are he may have missed some." He looked old and drained. "You're thinking I don't look so good... Don't bullshit me... You don't look so good either."
They talked about Ray's father. Feynman brightened: he and his son Carl loved talking mathematics. Mlodinow said he liked thinking hard, being creative, solving problems. "Well then, that's the answer you've been searching for." But it isn't specific to physics. "So?"
Feynman's final exam had one question: look closely at an electron-microscope photograph of an atom and ask, "Does it make your heart flutter?" Yes or no, no equations allowed. "I don't need to know. You need to know. This exam is self-graded. And it's not the answer that counts, it's what you do with the information." It is the cleanest test in the book for whether a field is yours.
Mlodinow asked whether he had regrets. His eyes filled: "I regret that I might not live to see my daughter, Michelle, grow up."
Elation; near accidents; Ray gets a hug that smells of garbage; a two-day party (pizza, burgers, joints). Ray leaves for Bellevue to learn programming (the Microsoft woman). Constantine returns to Athens under a rumor: he worked backward from the known proton mass and tuned lattice parameters; "Everybody does that"; joyless puffs. Feynman will not psychoanalyze; what should bother you is that many read the work and could not tell: too many followers, too few leaders.
Feynman's surgeon was "brave enough"; "Don't bullshit me... you don't look so good either." Son Carl loves math. Mlodinow realises he likes thinking hard, creating, solving: "that's the answer"; not specific to physics: "So?" Helen and a mail ruse. The exam: look at an electron-microscope photo of an atom. "Does it make your heart flutter?" Self-graded. What matters is what you do with the information. Regret on the table: he might not see daughter Michelle grow up.
Fermilab's final Muon g-2 result (June 2025) reached 127 ppb precision (world average about 124 ppb) and later shared the 2026 Breakthrough Prize in Fundamental Physics. The point for this chapter is procedural: blind analysis, disclosed corrections, independent cross-checks. That is what "could tell" looks like when followers are not enough.
FOLLOWERS / FLUTTER. Too many followers; does the atom make your heart flutter? Two tests, one chapter.
Grade yourself: open a picture or demo from your field that once thrilled you. Note the flutter (or its absence) without performing enthusiasm. Then audit one analysis you shipped: could a careful stranger tell if you tuned to a known answer?
"Heart flutter" as sole career guide ignores mortgages and teams. "Too many followers" can become contempt for necessary replication work. Leaders without followers leave no cumulative science.
Tuned lattices, too few leaders, a self-graded exam about flutter. Joy returns with a standard attached.
Chapter XXII: who are you?
The book's one long monologue is Feynman's answer to the question he disliked most: who are you as a person, and how has science shaped your character? He did not know what "find out who I am" meant. He knew he was a scientist, that the excitement is finding something beautiful that has always been there, and that he could not tell whether his scepticism made him a scientist or science made him sceptical. "I want to know what is true."
Identity talk bored him. He did not know what "find out who I am" meant as a project. What excited him was finding something beautiful that had always been there. Skepticism was not a pose layered on science; it was the same appetite: "I want to know what is true."
Arlene's timeline is the emotional spine: met at 13, engaged at 19, married at 26, died of tuberculosis after he married her knowing the prognosis. Friends warned. He reports no anger (no God to prosecute; bacteria are not moral agents) and no remorse. The household rule she left him, think when unhappy and do not when happy, is as close as the book comes to a personal algorithm. "I had already had it all" closes the marriage the way the rainbow closes the physics: the good was already present.
Feynman met Arlene at 13; they were engaged at 19 and married at 26, though she had tuberculosis and friends told him he no longer had to marry her. He married her because he loved her. She taught him that one has to be irrational sometimes, which does not mean stupid. His rule for happiness: when unhappy, think about it; when happy, don't. "After Arlene, the rest of my life didn't have to be so good... because I had already had it all."
Who are you; how did science shape character? Feynman does not know what "find out who I am" means. Excitement is finding something beautiful that was always there. Skepticism and science intertwine: "I want to know what is true." Arlene: met at 13, engaged at 19, married at 26; she taught him one must be irrational sometimes; women and the emotional side. Rule: when unhappy, think about it; when happy, don't. Arlene died of tuberculosis; he married knowing; friends warned; no anger (no God to blame; cannot be mad at bacteria), no remorse. "I had already had it all."
UNHAPPY-THINK / HAPPY-DON'T. Arlene's household law. Pair with wanting what is true.
For 48 hours, when unhappy, schedule fifteen minutes of honest thought; when happy, refuse to poke the mood apart. Journal whether the rule improved either state.
"When happy, don't think" can become avoidance of needed course corrections. "No remorse" can harden into refusal to learn from harm. Feynman's grief ethics were earned in a marriage under a death sentence; they are not a universal HR policy.
Truth over identity talk; Arlene; think when unhappy. He had already had it all.
Chapter XXIII: chasing other people's rainbows
Mlodinow's conclusion. He had been living with an external focus, Murray's way: accomplish, impress, lead. It was "like chasing a rainbow. Even worse, it was like chasing other people's rainbows." Feynman's focus was internal: satisfaction in discovery even if others already knew it, even in re-deriving someone else's result, even in playing with a child. That internal focus gave him freedom. Our culture, by Feynman's terms, is Greek; Feynman was a Babylonian, in physics (sum over paths, diagrams) and in life (the Greasy, the strip club, research for love).
The title arrives as a verdict on external focus. Murray's style, in Mlodinow's reading, tracks other people's standards of beauty and importance. Feynman's internal focus allows satisfaction from re-deriving a known result or from playing with a child, because the standard is owned. Culture runs Greek; Feynman remains Babylonian. The practical choices follow: drop string theory; keep quantum optics with Mark after Feynman showed the accepted approach was flawed; admit the writing life Breadcrumb had insulted. Even the office move ("Murray wants the office") is a small surrender of corridor prestige. The hope that remains is to write something he himself would admire.
| External focus (Murray's way) | Internal focus (Feynman's way) | |
|---|---|---|
| Measure of success | Recognition, leadership, coherence | Satisfaction, beauty, understanding |
| Choice of problem | Where the field's centre is | Where your curiosity is |
| Response to criticism | Defensive, competitive | "What do you care what other people think?" |
| Risk | Chasing | Being seen as eccentric, scattered |
| Control of happiness | In others' hands | In your own |
So he dropped string theory, went back to the quantum optics with Mark (Feynman had been right: their theory held up and the accepted approach was flawed), and came out about his writing. Colleagues rolled their eyes; soon Helen told him to move offices, because "Murray wants the office next to him for one of his own people." He hoped one day to write something Feynman would admire, and then, better, something he himself would admire.
What matters: know yourself, be honest, accept. External focus (Murray's way) is chasing . Internal focus (Feynman's way) is freedom: satisfaction even re-deriving known results or playing with a child. Culture is Greek; Feynman is Babylonian. Mlodinow chooses: drops string theory; continues quantum optics with Mark (Feynman was right; the accepted approach was flawed); comes out about writing. Moves office ("Murray wants the office"). Hope: write something he himself admires.
OWN-RAINBOW. Internal standard over borrowed beauty. Drop the chase; keep the fun.
List three goals you can name only because a peer, league table or famous person has them. Rewrite each as an internal test ("I would still want this if nobody knew"). Drop any that fail the rewrite.
Internal focus can become self-sealing arrogance: you ignore every external signal as "someone else's rainbow." Markets, reviewers and patients are not Murray. The book's point is ownership of standards, not deafness.
Stop chasing . Drop strings; own the writing; judge by your own admiration.
Chapter XXIV: epilogue
Feynman, weak from cancer, was the single scientist on the presidential commission investigating the Challenger explosion. Impatient with the bureaucracy, he investigated on his own, then dipped an O-ring in ice water at a public meeting on February 11, 1986, and showed it lost its resilience. The launch-morning temperature had been 29°F; the lowest at any previous launch was 53°F. Managers had ignored engineers' warnings. The commission tried to suppress his report as embarrassing to NASA; he fought, and it appeared as an appendix.
He had further operations in October 1986 and October 1987, still taught a course in QCD, and in his last months learned string theory from Murray in a weekly private "seminar." On February 3, 1988 he entered UCLA Medical Center with a failing kidney, refused dialysis, and died on February 15, 1988, months before his seventieth birthday. Feynman saw dying as a final discovery and expected it to be "IN-ter-ES-ting." He had survived ten years, beating the odds he had looked up, and lived to see Michelle reach adulthood: his one regret was not realised.
Murray paid tribute in Physics Today with what Mlodinow calls a "mixed review": he praised Feynman's lack of pomposity and original ideas, but objected that Feynman "surrounded himself with a cloud of myth" and thought in terms of "you and me" rather than "us." Ironically, Murray was then doing landmark work built on Feynman's histories formulation. He later left for Santa Fe.
- Murray Gell-Mann died at home in Santa Fe on 24 May 2019, aged 89; he was a distinguished fellow of the Santa Fe Institute, which he helped found.
- John Schwarz became a Caltech professor at the end of 1984 (later the Harold Brown Professor of Theoretical Physics). He and Michael Green shared the 2014 Breakthrough Prize in Fundamental Physics ($3 million) for superstring theory; Schwarz's acceptance thanked Gell-Mann for supporting him "when string theory was not very popular."
- Feynman's 1981 keynote "Simulating Physics with Computers" (MIT's Endicott House, May 1981) is now credited with founding quantum computing ("Nature isn't classical, dammit"). The 2025 Nobel Prize in Physics went to John Clarke, Michel Devoret and John Martinis for macroscopic quantum tunnelling in superconducting circuits, the basis of modern superconducting qubits.
The string story ends in vindication of persistence, if not of the theory. In 1984, after five years of work, Schwarz and Green found the mathematical "miracle" that removed the last major inconsistency (anomaly cancellation). It convinced Edward Witten, and within months the laughing stock became the hottest thing in physics: hundreds of theorists and over a thousand papers in two years. An administrator joked of Schwarz's appointment that even if he had invented sliced bread, people would say he did it at Caltech, so there was no need to keep him. Schwarz won a MacArthur fellowship (1987), election to the National Academy of Sciences (1997) and the Dannie Heineman Prize (2002). He now has Feynman's old office.
As of October 2026 string theory remains the dominant framework of quantum gravity research and a rich source of mathematics, but it has no confirmed experimental prediction. The Large Hadron Collider has found the Higgs boson but no supersymmetric partners and no sign of extra dimensions at the energies probed. Feynman's 1982 complaint, a theory that predicts nothing we can check, has not been answered; Gell-Mann's bet, that something this beautiful must be true, has not been lost either.
Mlodinow ends with the Radio Shack cassettes, found decades later, and Feynman's own stated aim in the Lectures: "I wanted most to give you some appreciation of the wonderful world and the physicist's way of looking at it." Feynman's way, Mlodinow concludes, was distinctly his own: to get the most out of what the world offered and out of whatever talent he had.
- Ideas that face an external test (the God of Experiment) improve fastest.
- "Derive it yourself" is the test of understanding; following is not enough.
- Fudged results are a symptom; the disease is readers who cannot tell.
- Judge your field with a self-graded exam: does it make your heart flutter?
- Internal focus puts happiness under your own control.
| Date | Event |
|---|---|
| 11 Feb 1986 | Challenger commission; O-ring in ice water; launch at 29°F versus previous lowest 53°F; managers ignored engineers; his report suppressed, then printed as appendix |
| Oct 1986, Oct 1987 | Further operations (fourth); still teaching QCD; learning string theory from Murray in a private weekly seminar |
| 3 Feb 1988 | UCLA; failing kidney; refuses dialysis; "final discovery" |
| 15 Feb 1988 | Dies, months before 70; survived about 10 years from diagnosis; lived to see Michelle reach adulthood |
Memorial six weeks later: Murray absent (federal agents raided his home over smuggled artifacts; he forfeited, went to Peru, key to Lima). Physics Today printed a "mixed review" line about a cloud of myth. Murray later worked on decoherent histories from Feynman's path ideas, left Caltech for Santa Fe.
1984: Schwarz and Green cancel the anomaly after five years; Witten converts; hundreds jump in; more than 1,000 papers in two years; Schwarz made professor by end of 1984 ("sliced bread" quip). Later: MacArthur 1987, NAS 1997, Dannie Heineman 2002, Breakthrough Prize with Green 2014. Schwarz has Feynman's old office. Helen Tuck retired in her seventies. Gell-Mann died in 2019. Radio Shack cassette tapes surfaced about twenty years later. The Feynman Lectures epilogue still asks for appreciation of the wonderful world and the physicist's way of looking at it; the lectures are free online.
O-RING TO OFFICE. Ice water on television; death in 1988; Schwarz in Feynman's office after the 1984 miracle. The epilogue's arc.
Run an O-ring drill on a decision in your organisation: the simple, visible test anyone in the room can understand. Write who would suppress it and how you would still publish the appendix.
Heroic last acts (ice water, refusing dialysis) can become a script for recklessness. Feynman's O-ring worked because the physics was simple and public. Copying the theatre without the clarity is how people grandstand in meetings.
Challenger, last operations, death at nearly 70, string theory's sudden fashion, Schwarz in the old office. The tapes remain.
Glossary
Anomaly cancellation. The 1984 Green and Schwarz result that certain superstring theories are free of quantum inconsistencies; it started the first superstring revolution.
Asymptotic freedom. The property of the strong force that it weakens at short distances and grows at long ones, so quarks move freely inside a proton yet cannot be pulled out.
Athenaeum. Caltech's faculty club, where Einstein stayed during two winters.
Babylonian and Greek. Feynman's labels for two styles: working from phenomena and what works, versus working from axioms and proof.
Electroweak theory. The unified theory of the electromagnetic and weak forces.
Feynman diagram. A picture of a particle interaction that encodes a term in a quantum calculation.
Greasy, the. Caltech's student cafeteria, where Feynman preferred to eat.
Lattice QCD. Solving quantum chromodynamics on a computer by replacing continuous space-time with a grid of points.
Muon g-2. The measurement of the muon's anomalous magnetic moment; Fermilab's final 2025 result was precise to 127 parts per billion.
O-ring. A rubber seal in the Challenger shuttle's boosters that lost resilience in cold weather.
Partons. Feynman's name for the constituents of the proton, later identified with quarks and gluons.
Photoelectric effect. The emission of electrons by light, explained by Einstein using light quanta.
Quantum chromodynamics (QCD). The theory of quarks, gluons and the strong force.
Quantum electrodynamics (QED). The quantum theory of electrons and light, developed by Feynman and others.
Quarks. Gell-Mann's proposed constituents of protons, neutrons and other hadrons, with fractional electric charges.
S-matrix theory. A 1960s approach that described particle collisions without underlying fields; championed by Geoffrey Chew and later abandoned.
String theory. The proposal that elementary particles are vibrations of tiny strings; it includes gravity and requires extra dimensions.
Strong force. The force binding quarks into protons and neutrons and holding nuclei together.
Sum over histories (paths). Feynman's formulation of quantum mechanics, in which a particle's behaviour is the sum of contributions from every possible path.
Supersymmetry. A proposed symmetry relating bosons and fermions, used in superstring theory; no partner particles have been found.
Uncertainty principle. Heisenberg's limit on how precisely pairs of quantities such as position and momentum can be known together.
Unified field theory. A single theory from which all the fundamental forces follow.
Wave-particle duality. The quantum fact that light and matter behave as waves in some experiments and particles in others.
People and places (quick)
Helen Tuck. Secretary whose desk sat between Feynman's and Gell-Mann's offices; witness to dry spells and bounce-backs.
John Schwarz. Caltech research fellow who held string theory through the lean years; later professor, eventually in Feynman's old office.
Mark Hillery. Quantum-optics collaborator; the "old bullshit" paper episode of Chapter XIX.
Nick Papanicolaou. Coauthor on the infinite-dimension method that won Mlodinow his Caltech post.
Geoffrey Chew. Schwarz's Berkeley advisor; S-matrix persistence as cautionary twin of fruitful stubbornness.
Constantine. Composite postdoc; lattice QCD and the tuning rumor of Chapter XXI.
Ray / Edward. Composite neighbours; garbage route and weapons research; Friday-night films by the pool.
Professor Gardening / Professor Breadcrumb. Composites for early fame then drift, and for career-management insult.
Spider Guy. Eccentric biologist with cups of spiders; Caltech's permission for oddity.
Santa Fe Institute. Gell-Mann's later institutional home after Caltech.
Further reading (from the book)
Mlodinow closes with a short further-reading list rather than a full bibliography.
- Richard Feynman, The Character of Physical Law; The Feynman Lectures on Physics (free online); Surely You're Joking, Mr. Feynman! (1985).
- Richard Feynman, "Simulating Physics with Computers" (1981 keynote; seed of quantum computing arguments).
- Murray Gell-Mann on the Eightfold Way and quarks; later decoherent histories.
- Green and Schwarz on anomaly cancellation (1984) and the first superstring revolution.
- Lattice QCD hadron masses (BMW and successors) for the "computers are the future" claim.
- Fermilab Muon g-2 final result (127 ppb, June 2025) and the 2026 Breakthrough Prize to the multi-lab collaboration.
- 2025 Nobel Prize in Physics (Clarke, Devoret, Martinis) for superconducting quantum circuits.
- String theory: still without a confirmed experimental test; 2014 Breakthrough Prize to Green and Schwarz; ongoing "not even wrong" debate (Quanta, March 2026).
- Gell-Mann's death in 2019; Schwarz eventually occupying Feynman's old office, as the epilogue notes.
READ-THE-VOICES. Lectures, 1981 simulation talk, 1984 anomaly paper, 2025–26 prizes. The memoir is the door; these are the rooms.
Use the memoir to catch the method; use the primary papers and the 2025–26 results to check which rainbows held.
About this book
Feynman's Rainbow: A Search for Beauty in Physics and in Life by Leonard Mlodinow, 2003. This is a chapter-by-chapter study of the book, with summaries, maps, tables and diagrams; short quotations keep their page references. It is shared free for personal study, is not for sale, and is not endorsed by the author or publisher. Please read and buy the original.
Insight Editions · rajeshrmahapatra.com/insight-editions