Insight Editions · No. 20
Elastic
The Power of Flexible Thinking in a Time of Change
Leonard Mlodinow
Leonard Mlodinow explains how unconscious, associative thinking supports creativity, adaptability, and insight when people face rapid change and problems that resist analytical reasoning.
A study guide to Elastic 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.
Introduction: The Demands of Change
On 6 July 2016 Niantic, a 40-person startup founded by ex-Google "Geo" staff, launched Pokémon Go, an augmented-reality game that puts virtual creatures on the real streets seen through a phone camera. Within two days it was on more than 10% of US Android phones; within two weeks it had 30 million users. iPhone owners soon spent more daily time on it than on Facebook, Snapchat, Instagram or Twitter. The phrase "Pokémon Go" briefly drew more Google searches than the word "porn." It earned about $1.6 million a day from US Apple users alone, added roughly $7.5 billion to Nintendo's market value within days (and doubled Nintendo's stock within a month), and passed 600 million downloads in six months. Facebook, by contrast, took three years (2004 to 2007) to reach 30 million users; World of Warcraft took six years to hit 12 million subscribers.
The ascent was not luck. Niantic piggybacked on phone GPS and cameras, used cloud scaling, and leaned on app-store economics (free download, paid upgrades) that did not even exist when World of Warcraft launched. A year later about 65 million people still played each month, and revenues had reached about $1.2 billion. Before the game, conventional wisdom said gamers "want to sit and play." Niantic ignored that assumption. The flip side is how fast rigid firms sink: BlackBerry, Blockbuster, Borders, Dell, Eastman Kodak, Encyclopaedia Britannica, Sun Microsystems, Sears, Yahoo. In 1958 the average S&P 500 company lived about 61 years; by the book's era that figure was about 20.
We all face similar demands. We take in about 100,000 words of information a day from various media, roughly a 300-page book, compared with about 28,000 a few decades ago. Trip planners use an average of 26 websites, wrestling with prices that change by hour of travel and by hour of search. Mlodinow's then 15-year-old daughter Olivia, while her parents were abroad, gave the house sitter a night off and texted to ask if she could invite "a few" friends; thanks to Instagram the "few" became 363. Digital availability blurs work and home; cybercrime and identity theft add new chores; free time must be defended. The book's subject is what is less often discussed: how we must think to thrive amid this pace of change.
Elastic thinking is the capacity to let go of comfortable ideas, tolerate ambiguity and contradiction, rise above conventional mindsets, reframe questions, and generate and combine new ideas. Unlike analytical thinking, which runs top-down under the brain's executive control, elastic thinking arises from bottom-up processes, largely unconscious and nonlinear, in which many threads run in parallel. No computer, and few animals, excel at it; it is built into the human brain.
Three ways to process information
The roundworm C. elegans, when its microbes run out, crawls into a slug's gut to be excreted somewhere new: a fixed program. A brooding goose that sees an egg outside her nest rolls it back with her bill; if the egg is removed mid-routine, she keeps nudging the imaginary egg, "like a mime." That is scripted processing: fast and reliable until circumstances change. Mlodinow catches himself grabbing almonds from the snack cabinet without deciding to. A hunting lioness, by contrast, must adapt to terrain, wind and prey moment by moment.
Analytical thought is what schools teach and IQ tests and entrance exams measure. Elastic thought cannot be traced from A to B to C. It evolved because, physically, humans are weak primates (a bonobo can jump twice as high) and survived by ingenuity. In the past 12,000 years of settled life, we have used it to fill the world with inventions; robins' nests, Mlodinow notes, have not gained bathrooms. Today's pace of change has moved elastic thinking from the background to the foreground.
The talents elastic thinking gathers form a diverse bouquet: letting go of comfortable ideas; tolerating ambiguity and contradiction; rising above conventional mindsets; reframing questions; abandoning ingrained assumptions; relying on imagination as much as logic; generating and integrating many ideas; experimenting and tolerating failure. Psychologists and neuroscientists treat these as aspects of one coherent cognitive style. Elastic thinking solves novel problems and overcomes the neural and psychological barriers that keep people stuck in the existing order.
Bottom-up versus top-down. A brain can calculate like a computer, from the top down, with executive structures dictating the approach. Because of its architecture it can also calculate from the bottom up: individual neurons fire in complex fashion without an executive director, with valuable input from emotional centres. That processing is nonlinear and can produce ideas that would never arise in A-to-B-to-C analysis. No computer and few animals excel at it; it is built into the human brain. Understanding the mechanisms lets us harness them at work and at home.
Rising above the nematode. Every animal has a toolbox. The roundworm C. elegans has 302 neurons and about 5,000 chemical synapses. When microbes run out it crawls into a slug's gut to be excreted somewhere new: an evolutionary hardwired response, not a desperate invention. Scripted behaviour (innate or habit) is nature's shortcut for mating, nesting and hunting; it fails under novelty. Humans also follow scripts: grabbing almonds without deciding, an automatic "no" to a child's request to stay home, the familiar driving route. For situations scripts cannot cover, evolution added analytical thought and elastic thought, most developed in mammals, especially primates, especially humans.
Why elasticity evolved. Physically, humans are weak primates: a bonobo can jump twice as high; a chimpanzee has, pound for pound, twice the arm strength; a gorilla sits on a sharp boulder while we need cushions and glasses. What saved our ancestors was elastic thinking, social cooperation and innovation. In the past 12,000 years of settled life we turned that talent toward improving everyday existence. Robins' nests have not gained bathrooms; squirrels do not store acorns in safes. We swim among products of the elastic mind: erasers, paper clips, shoes, toothbrushes, Razors, Segways, hoverboards. Each began as a never-before-imagined solution. Today's pace has moved elastic thinking from background to foreground for everyone, not only inventors and artists.
Connectome and the plan. In 2016 the NIH Human Connectome Project, using high-resolution imaging, showed far more substructures than previously believed: the dorsolateral prefrontal cortex alone resolved into about a dozen distinct elements; the project identified 97 new brain regions. The lessons have been compared to discovering that atoms contain protons, neutrons and electrons. Part I: adapting to change. Part II: how humans and animals process information. Part III: where new ideas come from. Part IV: what blocks elastic thinking and how to free it. Along the way: personality traits such as neophilia and schizotypy, and abilities such as pattern recognition, idea generation, divergent thinking, fluency, imagination and integrative thinking, plus the effects of mood, alcohol, cannabis and sleep.
S-A-E. Three modes only: Scripted (goose, habits), Analytical (school, IQ, A→B→C), Elastic (bottom-up, parallel, reframe). Pokémon Go and the falling S&P 500 lifespan are the demand; S-A-E is the toolkit.
Invert the introduction: suppose faster change mainly rewards better analysis and more data, not elasticity. Then Niantic's win is ops excellence, the nematode/goose contrast is cute biology, and the Connectome project's 97 regions do not change what schools should teach. Keep that challenge alive through Part II.
Today, label three decisions as S, A or E. For the most scripted one, force a 60-second reframe: what would change if the trigger were different? Write the reframe in one sentence. For the most analytical one, name one assumption you treated as fact. For the most elastic one, note what surprise or anomaly started the new path. Keep the three labels on a sticky note through the day.
The introduction’s pressure list is familiar: careers rewrite, information floods, crises arrive without a playbook. Mlodinow’s contribution is to name the cognitive skill that matches the pressure. Elastic thinking recombines concepts, tolerates ambiguity and drops a failing frame. It sits beside scripted habit and analytical optimisation, not instead of them. A world that only trains analysis produces experts who polish the wrong model. A world that only praises novelty produces noise. The book argues for deliberate practice of the middle skill: notice the broken script, generate remote associates, then verify with analysis.
Three processing modes recur through later chapters. Scripts handle etiquette and known paths. Analysis handles well-posed search. Elastic handles novelty that breaks the pose. Pokemon Go’s adoption curve is only the opening emblem; the deeper claim is that social readiness for novelty flipped faster than institutions adapted. Personal R&D (small experiments on your own habits and tools) becomes a survival skill, not a hobby.
| Mode | Question it answers | Blind spot |
|---|---|---|
| Scripted | “What do people usually do?” | anomaly |
| Analytical | “What is optimal inside this frame?” | wrong frame |
| Elastic | “What if the frame is wrong?” | unfiltered noise |
Generative tools raise average output quality in many tasks and can shrink idea diversity. Treat them as analytical accelerators and sometimes as association pumps, not as a replacement for unassisted recombination practice. Keep some sessions model-free so transfer does not rot.
1. The Joy of Change
The peril and the promise
In the 1962 Twilight Zone episode "To Serve Man," nine-foot Kanamits arrive bearing a book titled To Serve Man, end hunger and invite volunteers to their paradise planet. A cryptographer finally translates the book: it is a cookbook. "No such thing as a free lunch, unless you're the lunch." Novelty brings peril and promise: an animal exploring new ground may find food or become food. In a stable world, exploration is optional; in a changing one, those who have gathered information about new sites and escape routes survive.
Our environment now changes at an unprecedented pace:
- Science grows exponentially: global scientific output roughly doubles every nine years. In 2017 alone there were more than three million new papers. Between 2004 and 2014 publishers created more than 5,000 new journals just to hold the overflow. Whole industries now rest on more expertise than any person can master (Firmin Z. Sillo even wrote a 190-page book on "fuzzy logic optimization of injection molding of liquid silicone rubber").
- Websites doubled every two to three years; gay-rights acceptance swept through developed societies far faster than civil rights had, again fuelled by the young.
- Biologists see exploration vary by species: dogs explore because they descend from wolves bold enough to approach human camps; birds at the forest edge explore more than birds in stable interiors.
Bell introduced push-button dialling in 1963. It was clearly better, yet the dial phone lingered for decades. Apple's iPhone (2007) wiped out keypad and stylus phones within a few years, taking BlackBerry and Nokia with them. The difference is not the technology but us: society has shifted from resisting novelty to demanding it, and it now rewards those who adapt quickly.
The myth of change aversion
Business literature says "employees tend to resist change." But what managers call restructuring or turnaround, employees experience as layoffs or more work for the same pay. Avoiding change that carries real losses is rational. Offered a change with no downside, such as a better office with the same duties, most people welcome it. Absent negative consequences, humans are neophilic.
Attitudes depend on culture too. A few centuries ago most lives meant repetitive tasks, long solitude and little stimulation; people were suspicious of novelty and comfortable with conditions we would find extremely tedious: sixty-hour weeks chipping rock, hand-axing a fifty-foot maple, weeks in a cramped stagecoach from New York to Ohio. The word "boring" appeared in English only with the industrial revolution in the late eighteenth century, when stimulation became plentiful enough for its absence to be noticed. Electricity, radio, television, movies and new transport in the twentieth century then exposed people to other ways of living. The internet era has accelerated the appetite further.
Compared with other species, humans love novelty. "We [humans] jump borders. We push into new territory even when we have resources where we are. Other animals don't do this," says Svante Pääbo, director of the Max Planck Institute for Evolutionary Anthropology. Early Homo sapiens were not like that: two hundred thousand years ago the attitude was closer to "sit on a log and avoid unchecked areas" than to Star Trek's "boldly go." The catastrophe around 135,000 years ago and the later out-of-Africa spread changed the balance.
Our exploratory drive
Two hundred thousand years ago our ancestors stayed close to their origins in Africa. Then a catastrophe, probably climatic, around 135,000 years ago, may have reduced the human population to as few as 600 breeding individuals. Many scientists think this acted as a genetic filter favouring the bold; soon afterwards fossils show humans in Israel and China.
Dopamine carries signals in the brain's reward system. The DRD4 gene, which builds a dopamine receptor, comes in variants (2R, 3R, ... 7R). The 7R variant has been linked to stronger novelty-seeking. A 1999 study and a landmark 2011 paper (Matthews and Butler) found 7R is more common in populations that migrated farther from Africa, even after controlling for neutral genetic drift. Mlodinow cautions that many genes, and environment, shape any personality trait, and that epigenetics means experience changes how genes act.
Personal R&D and the neophilia scale
Timothy Treadwell, a Hollywood darling backed by Leonardo DiCaprio, Pierce Brosnan and Patagonia, lived among Alaskan grizzlies for 13 summers. Before Alaska, in Long Beach, he nearly died on a heroin-and-cocaine speedball and once jumped from a third-floor balcony on LSD, landing in mud. In Katmai he traded drug quests for bear country, marvelling that thousand-pound bears "can run thirty-five miles per hour," "jump eleven feet in the air" and "kill you with one swat." He believed singing and saying he loved them was the secret: "Animals rule, Timothy conquered." In 2003, not long after that boast, a bear killed and ate him and his girlfriend. He was an extreme sensation-seeker. Populations may benefit from such pioneers even when individuals die young: the group gains new resources and knowledge.
Most of us explore in milder ways, a personal form of R&D: socialising with strangers (new relationships), taking a night class (new hobby), interviewing while employed (career move), starting a business, trying Match.com. How much we invest depends on satisfaction with our current situation, our life stage and our innate neophilia. In science, art and business the same drive, applied to intellectual terrain, shapes civilised life as exploration once shaped survival in the wild.
Rate each 1 (strongly disagree) to 5 (strongly agree), then total:
- I would like to explore strange places.
- I would like to take off on a trip with no preplanned routes or timetables.
- I get restless when I spend too much time at home.
- I prefer friends who are excitingly unpredictable.
- I would like to try bungee jumping.
- I would like to do frightening things.
- I like wild parties.
- I love to have new and exciting experiences, even if they are illegal.
Average total: 24. About two-thirds of people score between 19 and 29. Scores fall with age: young adults (18 to 26) average several points above the overall mean, partly because the risk-checking prefrontal cortex matures only in the mid-twenties. Neophilia measures comfort with novelty; cognitive style (how you draw conclusions and solve problems) determines how you approach it.
Neophilia measures comfort with novelty; your cognitive style (how you draw conclusions and solve problems) determines how you approach it, which the following chapters explore.
Change as the default
Mlodinow opens on accelerating change: technologies, careers and social scripts rewrite faster than a single skill set can track. Elastic thinking is the skill of recombining concepts when the script fails; analytical thinking optimises inside a fixed frame. The book’s bet: evolution already equipped us with both, but culture over-rewards the analytical and under-trains the elastic.
Meta-analyses of generative AI and creativity report average assist effects around d≈0.27 on output quality, with diversity often falling (d≈−0.86) when models homogenise ideas. CHI 2025 work warns that LLM help can weaken unassisted transfer. Elastic practice still means generating and filtering without a model in the loop sometimes, not only prompting better.
| Pressure | Elastic demand |
|---|---|
| Career half-life shrinks | reframe identity and skills |
| Info flood | filter without freezing |
| Novel crises | recombine distant ideas |
Neophilia under pressure
Mlodinow’s opening claim is not that change is new. Humans always faced novelty. The claim is that the rate and coupling of change now outrun the career scripts, school tracks and corporate ladders that once absorbed it. Elastic thinking is the cognitive skill of treating a broken script as data, not as failure. Analytical thinking remains essential: it finishes proofs, ships code and keeps planes in the air. The book’s diagnosis is imbalance. Culture praises analysis in hiring and grading, then wonders why organisations freeze when the frame itself moves.
Pokemon Go (6 July 2016) is the book’s emblem of adoption speed: hundreds of millions of downloads in months, versus Facebook’s years to tens of millions of users. The point is not the game. The point is that social readiness for novelty flipped within a generation. Dial phones lingered for decades after push-button was clearly better; the iPhone erased keypad rivals in a few years. Neophilia (love of the new) became a measurable trait. The book’s eight-item scale averages about 24; roughly two-thirds score 19 to 29. High scorers chase novelty; low scorers conserve. Organisations need both, just as Kirton’s adaptors and innovators need each other.
Score yourself on eight novelty items (1-5): new foods, new routes, new tools, new people, new hobbies, new ideas at work, discomfort with routine, boredom with the familiar. Total 8-40. Note where you sit versus the book’s mean of 24. Then pick one domain where you are low and force one small novelty this week.
A company runs “innovation weeks” while promotion still rewards only predictable delivery. Elastic talk without elastic incentives is theatre. The book’s later chapters on frozen doctrine and idea filters name the same trap.
Intro close. Elastic thinking is the name for a skill evolution already gave us and schooling often under-trains. The rest of the book supplies mechanisms (reward, hierarchy, default network, insight labs), failure modes (freeze, filters, dogma) and levers (idle, mindset, emotion, careful chemistry, small will). Read it as a training manual, not as a cheer for chaos. Analysis still ships the plane; elastic notices when you are optimising the wrong airport.
- DRD4 and migration. The DRD4-novelty-seeking link was a classic candidate-gene finding. Large meta-analyses (for example Munafò and colleagues, 2008) found little or no association, and genome-wide studies show personality traits depend on thousands of variants with tiny effects. The migration correlation is suggestive at best.
- The 135,000-year bottleneck. The "600 people" figure comes from Curtis Marean's hypothesis about a cold glacial period (roughly 195,000 to 125,000 years ago). Genetic evidence for such a severe bottleneck is disputed, and fossils now place Homo sapiens outside Africa earlier than the book implies (Misliya, Israel, about 180,000 years ago; Apidima, Greece, possibly about 210,000).
- "100,000 words a day" comes from a 2009 UC San Diego estimate of all words heard and read, including TV; it is not 100,000 words of new information.
- The neophilia scale is essentially the Brief Sensation Seeking Scale (Hoyle and colleagues, 2002), which measures thrill-seeking more than intellectual openness.
- "Boring" appears in English in the eighteenth century, but the experience of boredom (Latin taedium, French ennui) is much older.
- Pokémon Go still had tens of millions of monthly players in 2025, when Niantic sold its games business to Scopely for about $3.5 billion.
- The fastest product adoption on record is now ChatGPT, which reached an estimated 100 million users within about two months of its November 2022 launch; generative AI is the defining change of the mid-2020s.
- GenAI creativity meta-analyses (through 2025 to 2026) find a modest assist benefit (around d ≈ 0.27) but a large drop in idea diversity when many people use the same models (d ≈ −0.86). CHI 2025 work also warns that LLM help can weaken unassisted transfer later.
- Personality genetics: twin studies put the heritability of novelty-seeking and openness at roughly 40 to 50%, spread across thousands of genes; no single "explorer gene" exists. Candidate-gene DRD4 findings remain unreplicated at genome-wide scale.
- Research on exploration versus exploitation now models curiosity explicitly; dopamine signals novelty as well as reward. Koutstaal (2025) frames flexibility and persistence as a trade-off, not a single dial.
- Grahek and colleagues on aging show control transitions: older adults often shift from flexible exploration toward more stable, experience-guided control, which matches the book's age drop in neophilia scores.
S-A-E / N-E-O. Intro: three modes Scripted, Analytical, Elastic. Chapter 1: Neophilia (not loss-aversion), Explorer gene story (DRD4, treat with caution), Out-of-Africa filter. If you can recite S-A-E and N-E-O, you have the opening architecture.
Invert Chapter 1: suppose humans really are change-averse, and neophilia is a rare mutant trait. Then Pokémon Go, iPhone adoption and out-of-Africa migration become exceptions, not the species rule, and managers who fight "resistance" are facing nature rather than poorly designed losses. The book's entire liberation programme (Parts III to IV) then applies only to a minority. Watch whether later chapters assume neophilia is normal or exceptional.
Run a personal R&D audit. List what you tried for the first time in the last three months: a skill, a place, a person, a tool. Score yourself on the eight-item neophilia scale above. If the list is short and the score is low, schedule small, low-cost experiments (a new route, an unfamiliar podcast, a conversation outside your field). When your organisation announces a change, write two columns: novelty (often welcome) and losses (rationally resisted). Address the losses first; that is usually the key to adoption. Optional: name one script you ran on autopilot today and one moment you switched to analysis or elasticity.
The world changes faster than ever, rewarding elastic thinking: bottom-up, flexible, idea-generating thought, distinct from scripts and analysis. People are not change-averse by nature; they resist losses. Neophilia is part of our evolutionary heritage, varies between people and declines with age.
2. What Is Thought?
Peeking inside the skull
In 1650 Anne Greene, a servant accused of killing her newborn (doctors thought the baby had been born too small to live), was hanged in Oxford for half an hour and given to Thomas Willis and William Petty for dissection. As they opened the coffin she rattled in her throat. They rubbed her limbs, applied turpentine, bled her and warmed her; she recovered. The authorities wanted to hang her again, but the doctors argued that her survival showed divine providence and her innocence, and she was pardoned. She earned money letting people view her in her coffin, later married and had children. Willis went on to study the brains of patients he had treated in life, linking brain damage to abnormal behaviour and founding neurology.
Today cognitive neuroscience studies living brains. One tenet: the form of thought is independent of its content; the process behind a new business, a new shampoo or a new recipe is fundamentally the same.
What qualifies as thought
Philosopher Karl Popper: "All life is problem solving." A rock makes no effort to change its fate; plants do little; animals must act. Evolution "does not create a Maserati where a motor scooter would do," so animals have three tiers: scripted, analytical and elastic. A slime mould finds the shortest path through a maze to food, and between two food sources builds efficient networks, yet it is executing chemistry. Dictionary and textbook definitions of thinking stress evaluating circumstances and generating ideas, many novel. By that convention scripts are not thinking, though the line is arbitrary.
Becoming mindful
Ellen Langer and colleagues asked, "How much behaviour can go on without full awareness?" A researcher approached people about to use a photocopier:
| Request | Agreed |
|---|---|
| "Excuse me, I have five pages. May I use the machine?" | 60% |
| "... May I use the machine, because I'm in a rush?" | 94% |
| "... May I use the machine, because I have to make copies?" | 93% |
The empty reason worked almost as well as the real one. People followed a script: "if a reason is given, comply." The paper was titled "The Mindlessness of Ostensibly Thoughtful Action."
Clinical psychologists see scripts constantly: couples fall into cycles where one partner presses the other's "raw spot," triggering a predictable reaction, which in turn triggers the first. Therapists teach couples to spot and interrupt the script together. The first step to better thinking, analytical or elastic, is to notice when you are on autopilot. William James: "Compared with what we ought to be, we are only half awake."
- Body scan (10 to 20 minutes): attend to each part of the body in turn, noticing sensations without judging them.
- Mindfulness of thoughts (up to 20 minutes): after quieting the mind with the breath, let thoughts drift in and out, observing them like clouds without following them.
- Mindful eating (5 minutes): examine a raisin or, Mlodinow's preference, a piece of chocolate; look, smell, hold it on the tongue, then chew slowly.
Research suggests doing an exercise three to six times a week produces measurable change within about a month.
The laws of thought
Analytical thought seems objective and emotion-free, which makes it simpler to study. In 1854 George Boole, dean at Queen's College Cork, published The Laws of Thought, reducing logic to algebra with true/false values. A century later his algebra became the logic of digital computers, once called "thinking machines." Boole also campaigned for limits on working hours; he died at 49 after walking in the rain to a lecture.
Charles Babbage designed the Analytical Engine, with punched-card input, a "store" (memory), a "mill" (processor) and output: nearly every principle of the modern computer. Ada Lovelace wrote that it "weaves algebraic patterns just as the Jacquard loom weaves flowers and leaves," and imagined it composing music. But playing chess by rules and composing an original symphony differ enormously.
The non-algorithmic elastic brain
In the 1950s, AI pioneers thought a meeting of experts could produce human-level intelligence. The General Problem Solver (Newell, Shaw and Simon, 1957) manipulated symbols and rules but could not cope with the open-ended real world. Even in 2018, Google veteran Andrew Moore, dean of computer science at Carnegie Mellon, called computers "the equivalent of really smart calculators." Computer "Mozart" and "Charlie Parker" pieces recombine human-compiled signatures rather than creating new styles.
"Aoccdrnig to a rseheearcr at Cmabrigde Uinerevtisy, it deosn't mttaer in waht oredr the ltteers in a wrod are..." Humans read this easily; text-reading software of the time choked. Your brain notices something is wrong, works out the rule, and reconstructs words from context, without being told. Similarly, MIT economist David Autor notes that any schoolchild recognises a chair, but defining "chair" by rules for a computer is nearly impossible because the category includes endless novel designs.
Google built a neural network that learned, without supervision, to recognise cats in YouTube videos (2012). Brains and computers differ in architecture: computers are switches understood through circuit logic; brains combine top-down and bottom-up processing and can produce useful non sequiturs.
Defence lawyer Sanford Perliss, in a murder trial with no body, announced that the supposed victim had been found and would walk into the courtroom. Every juror turned to the door. No one came. "You all looked," he said; that showed reasonable doubt. It was a leap no step-by-step reasoning would have produced. (In some retellings, the jury convicted anyway, noting that the defendant was the only one who did not look.)
Elastic, bottom-up thinking is how riddles are solved, how J. K. Rowling imagined Hogwarts and how Chester Carlson conceived the photocopier.
Peeking inside the skull (expanded)
In 1650 Anne Greene, a "fat, fleshy woman," was hanged in Oxford for a crime doctors doubted she had committed. Cut down after half an hour and taken for dissection, she was found still alive. Physicians revived her; the event made a sensation and fed early curiosity about what goes on inside the head. Opening the skull remained the only window on the living brain for centuries. The chapter's point is not gore: thought became a scientific object only when tools (from Berger's EEG to fMRI) made the inside measurable without a saw.
What qualifies as thought
Karl Popper wrote that "all life is problem solving." Brains evolved as better problem-solving machines. Fixed scripts handle routine triggers; analytical thought chains rules; elastic thought recombines when the world will not fit a script or a proof. Ellen Langer's late-1970s paper "The Mindlessness of Ostensibly Thoughtful Action" showed that complex social routines often run on autopilot, not full awareness. Rising above scripts is the first step toward either analysis or elasticity.
The laws of thought and the non-algorithmic brain
George Boole's The Laws of Thought (1854) cast reasoning as algebra; Lady Lovelace warned that machines follow what they are told. At the 1956 Dartmouth summer project, pioneers assumed artificial intelligence could rival human thought if experts met and coded hard enough. They did not separate analytical from elastic processing. Decades later, games such as chess and Go fell to machines that search and learn patterns, yet reframing a poorly posed problem, inventing a new category, or noticing that the rules themselves should change remain human strengths. AlphaGo's move 37 looked creative to observers; it was still optimisation inside a fixed game. Elastic thinking often means changing the game.
Anne Greene, Langer, Boole, Dartmouth (expanded)
Anne Greene climbed the Oxford gallows in 1650 still proclaiming innocence; after half an hour she was cut down for dissection and found alive. Revival made a public sensation and fed curiosity about the living brain. For centuries the only window remained a saw. The chapter's arc runs from that crude curiosity to the claim that thought has kinds: scripted, analytical, elastic.
Ellen Langer's late-1970s "mindlessness" work showed complex social interactions often run as scripts with little awareness. Rising above scripts is necessary but not sufficient: the next rung is analysis, praised for detachment from emotion and criticised for the same reason. Boole algebraised thought; Lovelace warned machines do only what they are told. The 1956 Dartmouth summer project gathered pioneers who expected artificial intelligence to rival human thought if the right experts coded hard enough. They did not distinguish elastic recombination from analytical search. Chess and Go later fell to machines that optimise inside fixed rules. Reframing the rules, inventing a category, or noticing that the problem is mis-posed remains the scarce human move. AlphaGo's famous unusual moves still lived inside the game of Go.
Scripts in relationships and the Xerox demand (pass 3)
Demand-withdraw cycles in couples are scripted: one partner requests talk; the other withdraws; the request amplifies; anger hits a raw spot; both escalate. Therapists coach recognition and joint interruption of the script, the same move available at the Xerox machine if someone notices the automatic "May I use the copier?" compliance. Driving offers the everyday version: autopilot until a siren forces disengagement.
Rising above scripts is necessary but not enough. Analysis is the next rung, praised for emotional detachment and criticised for the same reason. Elastic recombination sits above both when the rule set itself must change. Dartmouth 1956 optimists coded for analytical search; AlphaGo's unusual moves still lived inside Go. Reframing the game remains the scarce human move.
Boole, Lovelace and the missing elastic rung (pass 4)
Boole algebraised thought into symbols that machines could later grind. Lovelace warned that machines do only what they are told. Dartmouth 1956 gathered pioneers who expected coded effort to yield mind-like intelligence; they under-specified recombination outside fixed rules. Chess and Go fell to search and learning inside rule sets. The scarce human move remains inventing or noticing that the rule set is wrong. That is the elastic rung between script and analysis.
Scripts, analysis, elastic
Anne Greene (1650): hanged at Oxford, cut down for dissection, found alive; the sensation fed curiosity about the living brain when the only window was a saw. Ellen Langer’s late-1970s mindlessness work: complex social life often runs as scripts. Rising above scripts is necessary but not enough; the next rung is analysis, praised for detachment and criticised for the same reason.
Boole algebraised thought; Lovelace warned machines do only what they are told. 1956 Dartmouth summer: pioneers expected coded AI to rival human thought if experts worked hard enough. They did not separate elastic recombination from analytical search. Chess and Go later fell to machines inside fixed rules. Reframing the rules remains the scarce human move.
| Mode | Job | Failure mode |
|---|---|---|
| Scripted | habit, etiquette | miss the anomaly |
| Analytical | optimise in-frame | freeze the frame |
| Elastic | recombine / reframe | noise without filter |
From scripts to machines that think
Thought, for Mlodinow, is not a single process. It stacks. At the base sit scripts: Langer’s mindlessness studies showed people can run social routines so automatically that an empty “because” still gets them to the front of a copier queue for a short ask. Rising above scripts into analysis is the Enlightenment ideal Boole and Lovelace already debated: can thought be algebra? Lovelace’s answer still bites. Machines do what they are told. Dartmouth 1956 expected coded AI to match human thought if experts worked hard. Chess and Go later fell inside fixed rules. The scarce human move is still reframing the rules.
Anne Greene’s near-dissection in 1650 Oxford is the chapter’s macabre reminder that brains were once only known through trauma and saws. Today we have imaging and lesion studies, yet the functional split the book needs is already visible in behaviour: scripted habit, analytical optimisation inside a frame, and elastic recombination when the frame fails.
SCRIPT → ANALYSIS → ELASTIC. Habit runs the day; analysis tunes the day; elastic remakes the day when the map is wrong.
Thought stack in practice. Catch yourself in a script today (email greeting, meeting opener, commute). Break it once on purpose and note the discomfort. That discomfort is the cost of leaving mindlessness. Then take one work problem and ask whether you are analysing inside a frame or still choosing the frame. If you cannot name the frame, you are probably stuck in analysis theatre.
- Willis and Petty did not have "permission from King Charles I": Charles I was executed in 1649. Oxford anatomists had a long-standing right to the bodies of executed criminals.
- The scrambled-text "Cambridge researcher" is an internet myth; no such study exists. Cambridge psychologist Matt Davis has shown that reading scrambled text gets much harder when words are long or the letter swaps are large.
- Langer's Xerox effect held only for small requests (five pages). For 20 pages, the placebo reason did not work, which shows people were weighing the cost after all.
- The Perliss story is a widely told legal anecdote with several versions; treat it as illustration.
- Mindfulness research shows modest average benefits; a 2014 JAMA Internal Medicine meta-analysis found moderate effects on anxiety, depression and pain, and weaker evidence elsewhere.
- Large language models write fluent text and solve many CRA-like association tasks by search over training data; that is powerful pattern completion, not proof of elastic reframing.
- GenAI creativity meta-analyses (2025 to 2026) find a modest assist gain (about d ≈ 0.27) but a large drop in collective diversity when many creators use the same models (d ≈ −0.86).
- CHI 2025 studies warn that LLM help can weaken later unassisted performance: the tool substitutes for practice rather than teaching the skill.
- Game AI (AlphaGo, chess engines) still optimises inside fixed rules; changing the rules or inventing a new game remains a human elastic move.
S-A-E again, with teeth. Chapter 2: Scripts (Langer's mindlessness), Analysis (Boole, Lovelace, Dartmouth), Elastic (non-algorithmic recombination). Machines own more of A each year; E is the scarce skill.
Invert the chapter: suppose elastic thinking is just messy analysis plus luck, and every "reframing" can be restated as a larger search problem. Then Dartmouth was only early, not wrong, and the human edge is temporary compute. Watch whether later chapters (insight, DMN) give mechanisms that cannot be flattened into search.
Pick one routine you ran on autopilot today (email triage, a commute choice, a meeting script). Rewrite it three ways: (1) keep the script, (2) analyse it with explicit criteria, (3) reframe the goal so a different action becomes sensible. Note which rewrite felt hardest. That friction is where elastic work begins.
Thinking means evaluating circumstances and generating ideas, unlike scripts. Much human behaviour is mindless (Langer's copier). Boole and Babbage mechanised analytical thought; elastic thought, bottom-up and associative, resisted mechanisation for decades, though by 2026 learned AI systems have narrowed the gap.
3. Why We Think
Desire and obsession
Pat Darcy (a pseudonym) was diagnosed with Parkinson's disease at 41 in 1994. Parkinson's kills dopamine neurons in the substantia nigra, impairing movement. Her neurologist prescribed a dopamine agonist, a drug that mimics dopamine. Her symptoms eased, and then she began to paint compulsively: tables and canvases everywhere, lilies on her Maytag washing machine, walls, even the toilet seat. She spent nights painting and neglected everything else. Because the drug could not be targeted, it boosted not only the movement circuits but the reward system, which motivates thought and creativity.
Kurt Vonnegut: we "have to constantly be jumping off cliffs and developing our wings on the way down." When Mlodinow told his son Alexei that a half hour more study a day would turn Bs into As, Alexei replied, "Why would I want to do that?" A computer starts when you switch it on; the brain's "on" switch is internal. Darcy's doctors cut her dose, her tremors worsened, and she had deep brain stimulation surgery; her compulsion eased and her art became more balanced.
When thought goes unrewarded
EVR, raised on a farm, was a comptroller at a home-building firm by 29. At 35 surgeons removed a benign orbitofrontal meningioma along with much of his ventromedial prefrontal cortex. Afterwards his IQ was about 120, and his personality and knowledge tests were normal, yet he could not decide anything: he might spend a whole day debating whether to file documents by date or by size, and hours choosing a restaurant. He lost his job, went bankrupt in a business venture, divorced and remarried unwisely. Doctors blamed "adjustment problems." The tests measured only analytical ability. What he lacked was feeling: without emotional reward to weigh options, he could reason endlessly but never choose.
Emotions, including the pleasure of reward, let us evaluate options. Faced with an unfamiliar obstacle, the reward system guides us toward elastic thinking, stimulating us to generate alternatives.
Choice overload
Even healthy people are drained by constant decisions. William James: "There is no more miserable human being than one... for whom the lighting of every cigar, the drinking of every cup, the time of rising and going to bed every day... are subjects of express volitional deliberation." The remedy, from psychologist Barry Schwartz: be a satisficer, accepting the first good-enough option, rather than a maximiser searching for the best. Maximisers analyse more and feel more regret.
How good feelings happen
At McGill in 1954, postdoc Peter Milner, working in Donald Hebb's lab, taught new arrival James Olds to implant electrodes. Olds missed his target; the rat kept returning to the corner where it was stimulated. X-rays showed the electrode in what is now identified with the nucleus accumbens region. Given a lever, rats pressed it obsessively; males ignored females in heat and mothers abandoned pups.
The reward system predicts consequences and rewards each step, so you enjoy every bite of lasagna, and every step of a promising line of thought, not just the end. As a goal is reached, satiety signals reduce the reward. Signals from the ventral tegmental area (VTA) to the accumbens regulate this; Olds's rats bypassed satiety by stimulating the accumbens directly.
Thirteen statements ("When I'm doing well at something, I love to keep at it," "I crave excitement and new sensations," "It would excite me to win a contest," "I often act on the spur of the moment") are rated 1 to 4. Average: 41; about two-thirds score 37 to 45 (maximum 52). Low scorers are less reward-driven and may focus more on the journey; very high scorers are strongly goal-driven. Persistent drive toward goals is what psychologists call grit.
The rewards of art
Rote "drill or kill" maths is unrewarding; finding your own route to a solution is not. Brains reward idea generation, pattern recognition, divergent thinking and imagination, which is why people have always made art without material reward. Dostoyevsky, writing under deadline for money, felt his art suffer. Many studies suggest that paying for creative work can undermine the intrinsic motivation behind it.
Animals display creativity in courtship: bowerbirds decorate elaborate bowers. Martie Haselton and Geoffrey Miller gave 41 women descriptions of a creative but poor man and a rich but uncreative one. Near peak fertility, women rated the creative poor man more desirable for a short-term affair; fertility made no difference to ratings of the rich man.
Attention deficit, elasticity surplus
In the early 1990s Bonnie Cramond (University of Georgia) noticed that descriptions of ADHD children resembled descriptions of creative children: restless, impulsive, daydreaming, ignoring rules. Testing ADHD children for creativity, and gifted children for ADHD traits, she found striking overlap. ADHD involves weaker signalling in the VTA-to-accumbens reward circuit: routine feels boring, attention flits, and stray thoughts intrude, sometimes producing unusual combinations.
A theory holds that ADHD traits suited nomadic hunter-gatherers. Among Kenya's Ariaal, some herders settled to farm about 40 years before the study. Men carrying the ADHD-linked DRD4-7R variant were better nourished among the nomads but worse nourished among the settled farmers, where sustained focus pays.
Everyone falls somewhere between exploring and exploiting. Psychologist Michael Kirton distinguished adaptors (focused, prudent, improving tried methods, "stuffy" to others) from innovators (undisciplined, challenging assumptions, "abrasive" to others). Organisations and couples need both.
The pleasure of finding things out
A Homo habilis rolls onto a sharp rock, cuts himself, and gets the idea of a stone blade. In the early 1990s Jerry Hirshberg, head of Nissan's California design studio, saw a bus and a truck, and the image shaped the Nissan Quest minivan. Different eras, same process: association of unrelated ideas, rewarded by the brain along the way.
A former student wrote to Richard Feynman that he was working on humble problems and felt "nameless." Feynman replied: "A problem is grand in science if it lies before us unsolved and we see some way for us to make some headway into it... No problem is too small or too trivial if we can really do something about it. You say you are a nameless man. You are not to your wife and to your child." The joy is in the process.
Desire, obsession and the switch for thought
Pat Darcy (pseudonym), diagnosed with Parkinson's in 1994 at age 41, was treated with dopamine agonists. Like some patients, she developed compulsive behaviours: shopping, gambling, hypersexuality. The same chemical that restores movement can overdrive the reward system, which does not only make us feel good; it switches thinking on. Without expected reward, effortful thought stalls.
When thought goes unrewarded: Patient EVR
Patient EVR was a high-performing comptroller who, after surgery that damaged pathways linked to emotion and reward, remained intelligent on tests yet made disastrous life decisions: bankruptcies, bad partnerships, endless indecision over trivial choices. He could list pros and cons but felt no pull toward one option. Elastic and analytical effort both need a reason to start; the reward system supplies that reason.
Choice overload and how good feelings happen
Too many options drain the same motivational fuel. Peter Milner and James Olds at McGill in 1954, aiming at sleep circuits, accidentally found rats that would press a lever to stimulate a brain site thousands of times per hour: the pleasure/reward pathway. Art, puzzles and games hijack the same system with patterned surprise. Bonnie Cramond in the early 1990s noticed that descriptions of ADHD children overlapped gifted profiles: distractibility and appetite for stimulation can look like elasticity surplus when filters are loose.
EVR, choice and satiety (expanded)
EVR recovered in three months after tumour removal; doctors expected "no major dysfunction." He spent whole days debating filing schemes (date versus length versus relevance) and hours on restaurants: seating plan, menu, atmosphere, management, even driving to see how busy each was. IQ about 120, MMPI normal, moral-judgment interview healthy, fluent on foreign affairs and finance. Physicians blamed a "compulsive personality style" and "adjustment problems" amenable to psychotherapy. Later controlled work showed he had little capacity for feelings. Among the tissue removed was most of the orbitofrontal cortex, part of the reward system. Without conscious pleasure he had no goals, so open-ended choices ("Where would you like to vacation?") paralysed him, while closed questions with an external correct answer ("Where is Paris?") did not. He was fired, went bankrupt, divorced and moved back with his parents.
The contrast is the chapter's hinge: analytical tests can look fine while elastic, preference-driven thought is gone. Emotions evolved so organisms can evaluate positive and negative implications; without them, novelty and change have no steering signal. Mlodinow floats a speculative AI moral: perhaps machines will need something like enjoyment before they face open worlds the way humans do.
Choice overload is the everyday cousin. Barry Schwartz counted, in one medium grocery, 85 cracker brands, 285 cookies, 61 suntan lotions, 150 lipsticks, 175 salad dressings and 20 types of Goldfish crackers. Herbert Simon's 1956 satisficing (satisfy + suffice) is the remedy for most choices: take the first good-enough option. Maximising fits rare high-stakes picks (doctor, life partner); for shoes, shows and appliances it breeds regret.
Olds and Milner built a lever so rats could self-stimulate. Some pressed about a hundred times a minute. Males ignored females in heat; mothers abandoned nursing pups; animals had to be unhooked to prevent starvation. Direct accumbens stimulation skipped the VTA's satiety brake: desire without fading. Darcy's agonist flood is the human rhyme; healthy brains sit between EVR's drought and that flood.
Maximizers, dopamine dials and goal fuel (pass 3)
Maximizers scan endlessly for the best show, the best jacket, the best restaurant; Schwartz finds they report more regret and less satisfaction even when their objective picks look better. Satisficers set a threshold ("a comedy under two hours that starts soon") and stop. Elastic work often needs a temporary maximizer mode for rare, high-leverage choices, then a return to satisficing so the default network and insight systems have fuel left.
Reward chemistry is not a binary on/off. Healthy VTA → nucleus accumbens signalling rises for novelty and predicted reward, then settles as satiety and habituation kick in. Direct electrical self-stimulation bypassed that settle, so rats pressed until collapse. Agonist floods (Darcy's case) mimic the bypass; EVR's lesion removes the appraisal that makes any option worth choosing. Everyday elastic practice sits in the middle: enough pleasure and curiosity to explore, enough satiety to finish and ship.
Goals are not luxuries layered on top of cold calculation. Without felt valence, open questions ("Where should we vacation?") have no ranking function; closed questions with external answer keys still work. That is why EVR could discuss foreign policy fluently and still freeze over lunch. Teams that starve emotional appraisal (endless KPI grids without felt purpose) recreate a mild EVR pattern: lots of analysis, little commitment.
Try mapping one stalled decision this week: is the stall maximizer regret, missing valence, or a genuinely under-specified problem? Those three causes need different cures.
Olds, Milner and Hebb's wet-behind-the-ears postdoc (pass 4)
Peter Milner at McGill was studying sleep regulation when Donald Hebb asked him to show the ropes to a new postdoc, James Olds. The sleep question and the reward discovery were unrelated; serendipity put an electrode in the nucleus accumbens (literally "nucleus adjacent to the septum"). The lever that delivered stimulation turned desire into compulsion: press rates near a hundred a minute, ignored mates and pups, risk of starvation if not unhooked. That is the laboratory cartoon of a reward system without a brake.
Pat Darcy's agonist flood painted lilies on a Maytag and then turned "uncontrollable creativity" destructive. Kurt Vonnegut's line that we are what we pretend to be is the soft cousin: pretend long enough at a role and the reward system learns it. EVR could answer "Where is Paris?" and freeze on "Where would you like to vacation?" because the second question demands self-made criteria fueled by felt preference. Elastic thinking needs that fuel; analytical tests alone will not detect its absence.
From undercooked beaver to Queso Fiesta (pass 5)
Schwartz's aisle counts (crackers, cookies, lotions, lipsticks, dressings, twenty Goldfish types) dramatize how far choice has moved from "happy eating an undercooked beaver" to obsessing over cheddar versus queso. The evolutionary gift of preference becomes a modern tax when every SKU asks for maximising. Use satisficing as the default operating system; reserve maximising for the few choices where regret would be lasting. If a decision has consumed more than twice the time of similar past choices with no new information, you are probably maximising by habit, not by stakes.
Darcy, EVR and the steering signal
Pat Darcy (forty-one in 1994): arm pain, tremor, Parkinson’s. Dopamine neurons die in the substantia nigra (“black substance,” named 1791). Medications and agonists reshape reward; some patients develop impulse-control excesses (gambling, shopping, hypersexuality) when desire floods without satiety brakes. Darcy’s story sits beside EVR: orbitofrontal damage left IQ ~120 and fluent analysis but no feelings to steer open-ended choice. He debated filing schemes for days, went bankrupt, divorced, moved home.
Barry Schwartz’s grocery census (85 crackers, 285 cookies, 61 lotions, 150 lipsticks, 175 dressings, 20 Goldfish types) makes everyday choice overload. Herbert Simon’s 1956 satisficing is the remedy for most picks. Olds and Milner: rats self-stimulating ~100/minute, ignoring mates and pups until unhooked. Healthy brains sit between EVR’s drought and that flood. Emotions are not noise; they are the evaluation channel novelty needs.
EVR: no pleasure signal → paralysis on open choices. Darcy (agonist flood): too much desire signal → compulsion. Elastic thought needs a tuned reward dial, not a smashed one.
Desire, dopamine and open choices
Why we think is not only “to be right.” Desire and evaluation steer which thoughts get airtime. Pat Darcy’s Parkinson’s story and the agonist-driven impulse-control cases show what happens when the desire dial sticks high: gambling, shopping, hypersexuality without the usual satiety brake. EVR shows the opposite drought: high IQ, fluent debate, no feelings to end an open choice. He could analyse filing schemes for days and still ruin a life. Between flood and drought sits healthy motivation.
Olds and Milner’s self-stimulating rats (~100 presses/minute) ignored mates and pups. Barry Schwartz’s grocery aisle counts make the everyday version of EVR’s problem: too many options without a stopping rule. Herbert Simon’s satisficing (1956) is the practical fix for most picks. Explore-exploit trade-offs (Kirton’s adaptors versus innovators; Ariaal herders versus farmers and the contested DRD4-7R story) show the same dial in culture and ecology. Candidate-gene claims are fragile; the behavioural contrast between restless attention and sustained focus is not.
| Brain state | Signal | Choice pattern |
|---|---|---|
| EVR-like | little pleasure/pain guidance | endless analysis, little commitment |
| Agonist flood | excess desire | compulsion, poor brakes |
| Tuned | enough evaluation | explore when needed, stop when good enough |
For your next non-critical choice (lunch, playlist, weekend plan), set a two-minute timer, pick the first option that is “good enough,” and move on. Notice the urge to keep searching. That urge is the explore dial; practice turning it down on low-stakes picks so you have fuel for high-stakes reframes.
Kirton’s adaptors and innovators, Ariaal ecology contrasts and dopamine stories converge on one design problem: how much of today is for exploit (harvest known strategies) versus explore (pay for information). EVR cannot stop exploring options; agonist floods cannot stop exploiting a compulsion. Healthy days schedule both. Careers that only exploit decay when the niche moves; careers that only explore never compound.
Adjust the mix after two weeks based on whether you feel frozen or scattered.
- EVR was described by Eslinger and Damasio (1985) and inspired the somatic marker hypothesis. Later critics (for example Dunn and colleagues, 2006) questioned parts of the evidence for that hypothesis, but the decision-making deficit is well established.
- Choice overload. Meta-analyses (Scheibehenne and colleagues, 2010) found the average effect near zero; overload occurs only under specific conditions (complex choices, unclear preferences, time pressure; Chernev and colleagues, 2015).
- Ovulation and mate preference. Large preregistered studies (Jones and colleagues, 2018; Stern and colleagues, 2021) found little evidence that women's preferences shift across the cycle. The Haselton and Miller result, with 41 women, should be treated as unconfirmed.
- ADHD and creativity. Meta-analyses find a weak and inconsistent link: ADHD traits relate modestly to divergent thinking but not to real creative achievement in most samples.
- The Ariaal study (Eisenberg and colleagues, 2008) was small; its DRD4 finding inherits the candidate-gene caveats noted in Chapter 1.
- The scale shown is drawn from Carver and White's BAS (behavioural activation) scales.
- Computational neuroscience now treats curiosity as a formal drive alongside reward; dopamine tracks novelty and prediction error, not only pleasure.
- Impulse-control disorders on dopamine agonists remain a recognised clinical risk in Parkinson's care guidelines.
- Choice-overload findings are mixed at scale; effects depend on expertise and how options are structured, not only on raw count.
- ADHD and creativity: meta-analyses show a small average link with divergent thinking, larger in adults than children, and confounded by comorbidity and measurement.
R-E-W-A-R-D. Reward switches thought on; EVR shows thought without feeling fails; Wanting ≠ liking; Agonists can overdrive (Darcy); Rats self-stimulate (Olds/Milner 1954); Distractibility can be elasticity surplus (Cramond).
Invert: suppose reward is noise and cold analysis is enough. Then EVR should have thrived, Darcy's shopping is irrelevant to science, and schools that drain curiosity do no cognitive harm. The chapter's claim is that without felt value, the engine stalls. Test that against your last abandoned hard problem.
For one week, before each deep-work block, write the felt reward in one line (curiosity, status, relief, play). If you cannot name one, add a small immediate reward (timer race, shared progress note) or cut the task. Separately, audit one domain of choice overload (apps, groceries, tabs) and cut options by half for three days; note whether thinking feels freer.
We think because the reward system makes thinking feel good. Too little reward (EVR) paralyses choice; too much (Pat Darcy) breeds compulsion. Reward flows along the path to a goal, so the journey matters. Explorers and exploiters, innovators and adaptors, ADHD and focus, all reflect different reward settings.
4. The World Inside Your Brain
How brains represent the world
Anything you think about must be represented in a neural network, a point Aristotle anticipated. Try reciting a familiar phone number backwards: it is hard, because the brain does not store it as a written string but as a learned sequence. How information is represented determines what you can do with it.
| IBM Deep Blue | Garry Kasparov | |
|---|---|---|
| Representation | A tree of possible moves | Meaningful patterns of pieces |
| Method | Top-down brute-force search (about 200 million positions a second) | Bottom-up recognition, a few positions deeply |
| Strength | Speed of hardware | Understanding and intuition |
Deep Blue won the 1997 rematch. It won by speed, not by thinking like a grandmaster.
In 2011 IBM's Watson, with four terabytes of storage and 200 million pages of content, beat the champions of Jeopardy!. Computer scientists then began imitating the brain's bottom-up neural networks, producing a program that beat the world's best Go player and a much improved Google Translate. But each needed its own machine; a single human brain does both while balancing on one foot.
How brains create meaning
A doorbell's physical reality is air vibrations. A microphone represents it as a varying current, a radio as electromagnetic waves, your brain as activity in the auditory cortex, which you experience as ringing. People with synaesthesia may also see a colour. Mammal brains add layers of meaning: a doorbell may mean a visitor, a delivery, or the dog going wild.
The trick is to group elements into compound units, and those into higher-level units: concepts. Spot your grandmother unexpectedly and your brain does not process pixels and search a database; it recognises clusters of features, as Kasparov recognised clusters of pieces, and activates your "grandmother" concept.
Recordings from electrodes implanted in epilepsy patients found single neurons that fire for one person or place (famously, Jennifer Aniston or Halle Berry, whether seen in a photograph, a drawing or written as a name). These concept cells are building blocks of thought. Because neurons are shared among networks, activating one concept primes related ones: the root of association.
The bottom-up brilliance of ants
Social insects (ants, bees, wasps, termites) are just 2% of insect species but most of insect biomass. No ant is in charge, yet a colony adjusts its foraging to a shrinking or expanding territory, army ants run raids of up to 200,000 workers, and weaver ants form living chains to pull leaves together. Stanford's Deborah Gordon studies harvester ants: each year, on the same day, every colony releases winged males and virgin queens to a mating ground. Intelligence emerges from simple local rules: bottom-up processing, the opposite of a corporation's chain of command or a programmed computer.
- Top-down: the brain's executive (the prefrontal cortex) directs thought toward goals, suppressing irrelevant ideas.
- Bottom-up: many simple units interact without a supervisor; ideas emerge from associations.
Human thought is the interplay of both: organised and focused, yet capable of surprise.
Your brain's hierarchy
We have about 86 billion neurons, roughly 200,000 times the ants in a typical colony, each connected to thousands of others. The cortex has two hemispheres, each with four lobes. The prefrontal cortex, found only in mammals (birds have an analogous structure), lets us rise above scripted responses: it supervises thinking, sets goals and inhibits impulses, as when it stops you from telling off a boss who is unfairly berating you. It contains finer structures such as the primate-only lateral prefrontal cortex, and connects with subcortical structures such as the VTA and substantia nigra.
An intellectual adventure
Academic science works bottom-up: young researchers choose problems and ideas compete. Can a company work that way?
Nathan Myhrvold, a physics PhD who worked with Stephen Hawking, took a summer job at Microsoft, founded Microsoft Research and left in 1999 rich (he showed Mlodinow an expensive fossil collection). His Intellectual Ventures was structured like a brain, with interconnected inventors encouraged to pursue strange ideas, then patent and license them:
- Turning discarded coffee-cherry hulls into a gluten-free flour (growers earn about 10 cents profit per pound of coffee selling for $15).
- The Photonic Fence, a laser that detects and shoots down mosquitoes, to fight malaria.
IV makes money from patents rather than manufacturing, which critics say stifles innovation.
How brains represent and create meaning
Concepts are not library cards; they are networks of neurons that can share units. "Dog" and "wolf" overlap; "dog" and "justice" barely do. Shared neurons make association cheap and metaphor possible. Meaning is built as association cortices bind features into usable wholes: edges become letters, letters become a warning on a NO TRESPASSING sign.
Bottom-up brilliance and hierarchy
Ant colonies solve complex logistics with local rules and no CEO: a living demo of bottom-up intelligence. Brains stack a hierarchy: primary sensory areas detect features; association areas combine them; executive regions bias which coalitions win. Elastic thinking thrives when lower and middle layers explore combinations while the top layer loosens its veto. Deep Blue beat Kasparov in 1997 with brute search plus evaluation; human grandmasters still rely on pattern webs that look more like concept networks than tree search.
An intellectual adventure
Mlodinow frames everyday perception as an adventure in construction: the brain does not receive a finished world; it builds one from spikes and priors. When two people share a label ("risk," "quality," Elastic teams make those networks visible with examples, not slogans. That is why concept overlap is not only neuroscience; it is a collaboration tool.
Hierarchy, ants and Deep Blue (expanded)
Ant colonies allocate foragers and nest work through local pheromone rules with no project manager: a textbook bottom-up system. Brains stack levels: primary sensory maps → association cortices → executive bias. When two colleagues share a word ("aligned," "risk") but not a network, they need examples, not more slogans. Making overlap visible is practical elastic hygiene.
Grandma, pixels and shared concept webs (pass 3)
Recognizing Grandma is not a pixel-by-pixel database scan across infinite lighting and hat conditions. The brain works with feature clusters and concept webs: eyes, gait, voice, the fruit hat as optional detail. Deep Blue could brute-force chess positions; Kasparov chunked patterns. Elastic thinking lives in the middle layers where features recombine into new concepts, while top-down bias temporarily softens.
When two colleagues share a label ("aligned," "risk") but not a network, meetings stall. Make the web visible: draw three nodes each person attaches to the word, then look for overlap and gaps. That is applied concept-network hygiene, not mysticism.
Eighty-six billion ants and the doorbell's many meanings (pass 4)
Roughly 86 billion neurons: about two hundred thousand times a large ant colony. Bottom-up processing among them produces unorchestrated elastic thought; hierarchy adds executive bias on top. A doorbell is not only a ringing sound; it carries associations of interruption, social connection or gratification depending on which concept web is active. Aristotle noted that thought needs an internal representation; modern neuroscience fills that claim with networks rather than a private video screen. If there is no screen that plays raw optical data, then every percept is already a translation. Elastic work is the art of re-translating: same doorbell, new meaning; same chess chunks, new plan; same product category, new question.
Deep Blue's guts versus Kasparov's chunks (pass 5)
IBM's first design task for Deep Blue was internal representation: how the machine would hold the game. Brute position trees worked at silicon speed. Kasparov held clusters of pieces as meaningful units. Human elastic thought is closer to Kasparov's chunks than to Deep Blue's exhaustive leaves: new ideas rearrange chunk boundaries. When a team is stuck, ask whether you are searching deeper inside the wrong representation. Sometimes the win is a new encoding, not a longer search.
Hierarchy, ants and meaning
Brains do not store a photographic world; they build hierarchical representations. Bottom-up brilliance (edge → object → scene) meets top-down expectation. Ant colonies show local rules, global intelligence without a CEO ant. Your cortex stacks similar tricks: association areas turn edges into meaning. An “intellectual adventure” in the chapter is recognising that perception is already inference.
PERCEPTION=MODEL. What you “see” is a controlled hallucination constrained by sense data.
Shofty et al., Brain 2024: stimulating default-mode network nodes can reduce originality, evidence for causal DMN roles in idea generation. Luchini/Beaty 2025 (Cerebral Cortex): default-executive coupling neurofeedback raised originality the next day. Elastic training is becoming a measurable circuit story, not only a metaphor.
Ants, hierarchies and controlled hallucination
About 86 billion neurons build a world that is not a camera dump. Bottom-up feature detection meets top-down expectation. Ant colonies show local rules yielding colony-level intelligence without a CEO. Your cortex stacks similar tricks: association areas turn edges into objects and objects into meaning. Perception is already inference, a controlled hallucination constrained by sense data.
Machine history in the chapter tracks the same move: Deep Blue (1997) brute-forced chess trees; Watson (2011) searched huge text; Go and AlphaZero leaned on brain-like learning. Representation changed what search could do. Elastic thought often begins by changing the representation, not by searching harder in the old one.
Shofty et al., Brain 2024 linked stimulation of default-mode nodes to lower originality. Luchini/Beaty 2025 (Cerebral Cortex) reported that default-executive coupling neurofeedback raised originality the next day. The book’s idle-mind chapters now sit next to causal and training evidence, not only correlational PET stories.
A team can have brilliant individuals and still freeze if every promotion rewards only in-frame optimisation. Hierarchy of cortex is not hierarchy of org chart. Elastic organisations need dissent channels and safe anomaly reporting, not only smarter hires.
Changing representation is the chapter’s quiet thesis. Ants and neurons both show intelligence as emergent from local rules. Hierarchical cortex turns spikes into edges into objects into scenes; top-down prediction fills gaps and also creates illusions. When Deep Blue searched trees and AlphaZero learned policies, the leap was not only compute. It was a new way to encode the game. Elastic thinkers do the same on messy life problems: invent a diagram, a metaphor, a stakeholder map, then search inside it.
| Level | Example content |
|---|---|
| Features | edges, tones, phonemes |
| Objects | faces, words, tools |
| Scenes / scripts | kitchen, meeting, commute |
| Models | “this market,” “this marriage,” “this proof” |
| Elastic work often rewrites the model row while analysis polishes the object row. |
World-building caveat. Hierarchical brains predict so well that they can ignore disconfirming pixels. Optical illusions are cute; organisational illusions are expensive. Build a habit of asking “what would I see if my model were wrong?” once per project review. That single question is elastic hygiene for predictive minds.
- Concept cells (Quiroga and colleagues, 2005) are sparse, not unique: each concept is represented by a small population of neurons, and each neuron may respond to several related concepts. The "grandmother cell" is a convenient simplification.
- Intellectual Ventures became one of the world's largest patent holders and was widely criticised as a "patent troll" for licensing and litigation; its coffee-flour product never became a mass success.
- Deep Blue's 1997 victory involved controversy: Kasparov suspected human intervention (never substantiated), and IBM retired the machine without a rematch.
- The prefrontal cortex is not uniquely mammalian in function; corvids and parrots solve complex problems using the nidopallium caudolaterale, which the book's footnote acknowledges.
- Modern neural nets learn distributed representations that echo the book's concept-network picture; interpretability work still struggles to map those dimensions to human concepts.
- Multimodal models bind vision and language, a rough analogue of association cortices, yet still fail at many reframing tasks that children solve.
- Kasparov's loss to Deep Blue (1997) looks quaint beside engine strength today; the elastic lesson stands: search wins fixed games; representation change wins open worlds.
N-A-H. Networks of overlapping neurons encode concepts; Ants show bottom-up smarts; Hierarchy lets executive bias which coalitions win. Elasticity = explore overlaps while the top layer eases its grip.
Invert: suppose concepts are crisp symbols, not overlapping nets. Then metaphor is a bug, insight is noise, and symbolic AI should have won. The chapter's bet is distributed representation. Later DMN and CRA findings either support that bet or leave it metaphorical.
Pick two concepts from your work that "should not" connect. List five shared features or metaphors (forced is fine). Circle the least ridiculous link and spend fifteen minutes building a project, joke or diagram on it. You are rehearsing overlap-driven association.
Brains represent the world as hierarchies of concepts built bottom-up from features. Shared neurons create associations. Like an ant colony, much of our intelligence emerges without a supervisor, while the prefrontal executive directs it top-down. Organisations can be designed to work the same way.
5. The Power of Your Point of View
A paradigm shift in popcorn
In the 1960s David Wallerstein, an executive at the Balaban & Katz cinema chain in Chicago, could not get customers to buy more popcorn. His insight: people wanted more but did not want to be seen buying two bags, which looked gluttonous. So he offered a bigger bag, the jumbo. Popcorn sales soared, and so did sales of Coca-Cola. Wallerstein had found a basic law of the food industry: people will eat enormous portions if "enormous" is one of the sizes on offer.
Yet the idea was heresy. When Wallerstein moved to McDonald's in the mid-1970s, Ray Kroc refused to introduce larger fries: "If people want more fries, they can buy two bags." McDonald's adopted the strategy only in 1990, by which time "supersizing" was conventional wisdom. It took the food industry longer to accept the law of gluttony than it took physicists to accept relativity.
Thomas Kuhn's The Structure of Scientific Revolutions (1962) described paradigm shifts: not incremental advances but changes in the framework of thought itself. They leave behind successful people whose rigidity makes them cling to the old framework despite the evidence. Paradigms govern businesses and personal lives as much as science.
The structure of personal revolutions
Mlodinow writes strongly held beliefs on slips of paper, picks one at random and imagines being told it is wrong. Why does he hold it? Who disagrees, and why might respectable people do so?
The son of Polish-Jewish Holocaust survivors who emigrated after the war, he had always favoured generous immigration. He drew the slip "The supporters of building a wall with Mexico are evil," rolled his eyes, then watched Fox News to find the other side's logic: we have immigration laws, and if we dislike them we should change them, not ignore them. He did not change his position, but he saw that opponents were not evil, and their argument had substance.
The more we argue, the further apart we can drift; quick initial judgements rest on our paradigm. In business the same rigidity is costly. Apple is classified as a manufacturer yet makes virtually nothing itself, avoiding factories and staying nimble. Nike, partnering with HP from 2016, used 3D printing for prototypes and envisioned in-store, custom-made shoes.
Reimagining our frames of thought
A rabbi told his noisy congregation that socialising while praying was disrespectful, but praying while socialising was always welcome. Framing changes everything.
Fewer than half of subjects solved each one:
- A man keeps reading after the lights go off in a pitch-dark room (not an e-book).
- A magician throws a ping-pong ball so that it stops and reverses, without bouncing it off anything or tying anything to it.
- Two mothers and two daughters caught three fish, yet each caught her own.
- Marsha and Marjorie have the same birthday and parents but are not twins.
Answers: he was blind, reading braille; the ball was thrown upward; the women were three generations; they were triplets. Each riddle plants a default picture (a man staring at a page, a horizontal throw, four women, a pair) that our associative brain accepts automatically.
The dog-and-bone problem
Psychologists call the reframing of a problem restructuring. Douglas Hofstadter illustrates it: a bone lands in the neighbour's yard behind a fence, with an open gate some distance away. A dog using a geographic map runs straight at the bone and stops at the fence, inches away but stuck. A smarter dog switches to distance in problem space: the path to the bone, through the gate.
How mathematicians think
Remove two diagonally opposite corners of an 8×8 board, leaving 62 squares. Can 31 dominoes, each covering two adjacent squares, cover it? Trial and error never proves impossibility. Reframe: search the space of laws, not placements. Each domino covers one black and one white square. The removed corners are the same colour, leaving 32 of one colour and 30 of the other. Covering is impossible.
In the sixteenth century Rafael Bombelli asked: suppose a number whose square is -1 exists; does it lead to a contradiction? It did not, and he worked out its rules. Imaginary numbers are now essential to describing waves, electromagnetism and quantum mechanics, and taught to high school students, though medieval scholars could not accept them.
The influence of culture
Our professional, social and cultural milieu shapes our frameworks, even our perception. Shinobu Kitayama's framed-line test showed participants a square with a vertical line and asked them to redraw the line in a different-sized square, either at the same absolute length or the same proportion. Americans did better at the absolute task; Japanese participants, from a culture that stresses context, did better at the relative task.
Studies ranking the US and 13 comparably wealthy European countries by patents per capita (1971 to 1980, and 1995 to 2005) found ranks were stable over time, suggesting cultural factors in inventiveness. A strong cultural identity can lock in an approach; exposure to other cultures loosens thinking.
A paradigm shift in popcorn
Movie theatres once sold popcorn in bags behind the counter. A reframing (popcorn as a high-margin product to push, with better smell, bigger containers, visible popping) remade the economics of the whole outing. The point is Kuhnian in miniature: the data (kernels, customers) stayed; the frame changed, and with it the profitable questions.
Personal revolutions, dog-and-bone, mathematicians, culture
Thomas Kuhn's Structure of Scientific Revolutions (1962) named paradigm shift for science; Mlodinow applies the same pattern to personal and business frames. The dog-and-bone puzzle (how does the dog get the bone across the fence?) rewards seeing a new path, not more force on the old one. Mathematicians often progress by restating a problem until it becomes a different, solvable one. Culture supplies default frames: what counts as polite, risky, or even visible. Elastic work often begins by naming the frame you did not know you were using.
Popcorn, Kuhn, dog-and-bone, culture (expanded)
Theatre popcorn shifted from a sideline bag to a high-margin, smell-forward, size-tiered product engineered into the outing. Kernels and customers were not new; the frame was. Kuhn (1962) named paradigm shifts for science; Mlodinow applies the pattern to personal and commercial revolutions: anomalies pile up inside an old question set until a new question set makes progress cheap again.
The dog-and-bone style puzzle rewards seeing a path around the fence, not pulling harder on the chain. Mathematicians often win by restating until the problem becomes one they already know how to solve. Culture loads default frames: what counts as ambitious, polite, masculine, risky or even visible. Elastic work starts by naming the frame you did not know you were using, then trying on another long enough to act inside it.
Nike, synagogues and restating the problem (pass 3)
Nike's move toward on-the-spot 3-D printing and knitting for custom feet is not a gadget story; it is a frame story. Competitors who treat past success methods as sacred face an existential challenge when the question set changes from "Which mass size sells?" to "Can we fit this foot today?" Apple's category inventions rhyme: invent the question, then the product.
The rabbi's joke about praying while socializing versus socializing while praying is a frame flip in miniature: same behaviors, different primary purpose, different judgment. Elastic work often starts there. Mathematicians restate until the problem becomes one they already know; product teams restate until the customer job becomes visible. Culture loads defaults (what counts as ambitious, polite, masculine, risky). Name the default frame, try another for a week, and watch which actions become cheap.
Marsha, Marjorie and the quiet church (pass 5)
The Marsha/Marjorie riddle (same day, same parents, not twins) forces a frame break: they are triplets. Analytic age problems ("Bob's father is three times…") stay inside systematic steps. The synagogue social-noise sermon and the quiet church contrast show culture loading what counts as respectful. Elastic practice: collect three "not twins" style assumptions in your domain this week and write the triplet-style alternative for each.
Apple, assumptions and existential challenge (pass 6)
Competitors who treat past-success methods as sacred meet category inventors as existential threats. The cure is not cynicism about what worked; it is scheduled assumption audits: list three methods that "always work," invent the customer question that would make each obsolete, and run a cheap probe. Popcorn, Nike custom feet and the rabbi's purpose-flip are the same move at different scales.
Assumption audit worksheet (pass 7)
Write three methods that "always work" in your domain. For each, invent the customer or nature question that would make it obsolete. Design a one-day probe that would falsify the method cheaply. Popcorn smell, Nike custom feet and synagogue purpose-flips are templates: change the primary question, and the old success metrics stop being the scoreboard.
Popcorn, paradigms and dog-and-bone
A paradigm shift in popcorn (how we classify and market a snack) stands for Kuhn-like personal revolutions: the data did not change, the frame did. Structure of personal revolutions: anomaly → crisis → new frame → textbook rewrite of memory. Dog-and-bone geometry problems reward changing the diagram. Mathematicians “think” by inventing representations; culture loads which frames feel natural.
When stuck, change the representation before you push harder inside the old one: redraw the axes, rename the variables, swap the stakeholder.
Frames, popcorn and checkerboards
A paradigm shift in popcorn (jumbo bags, Ray Kroc’s refusal, eventual supersizing) is Kuhn for consumers: the data about hunger did not change; the acceptable frame did. Personal revolutions follow a similar arc: anomaly, crisis, new frame, rewritten memory of what always seemed obvious. The mutilated checkerboard teaches the same lesson in logic. Remove two same-colour corners; thirty-one dominoes cannot cover the board. Search placements forever or reframe to colour parity in one step.
Dog-and-bone geometry problems reward changing the diagram. Mathematicians invent representations; culture loads which frames feel natural. Point of view is not decoration. It is the search space.
| When | Move |
|---|---|
| 1960s | jumbo popcorn experiments |
| 1962 | Kuhn names paradigm shifts |
| mid-1970s | Kroc refuses larger fries (“buy two”) |
| 1990 | McDonald’s adopts supersizing norms |
When stuck for twenty minutes, force a representation change before more effort: redraw axes, rename stakeholders, invert the goal (“how would I cause this problem?”), or solve a toy version with ugly numbers.
You do not need a scientific revolution to feel a paradigm shift. Career identity, parenting scripts and scientific pet theories all accumulate anomalies until a crisis forces a rewrite. Popcorn supersizing and the mutilated checkerboard are comic and crisp versions of the same move: the facts were available; the acceptable question was not. Keep a private anomaly log. When three anomalies rhyme, schedule a reframe session before the crisis does it for you.
ANOMALY LOG → REFRAME. Three rhyming surprises beat one dramatic failure as a teacher, if you write them down.
- Wallerstein's jumbo popcorn story is widely repeated (notably by Michael Pollan) and plausible, but rests largely on anecdote.
- Kuhn's paradigm concept has been criticised as vague (one critic counted 21 senses of the word), and many scientific changes are incremental rather than revolutionary.
- Cultural differences in perception are real but smaller and less consistent in replications than early studies suggested; Western, educated samples still dominate (the "WEIRD" problem).
- Patent counts are a crude proxy for creativity; they reflect legal systems, industrial structure and incentives as much as culture.
- Nike's in-store 3D-printed shoes did not materialise at scale; 3D printing remains mainly for prototyping and limited editions.
- "Paradigm shift" remains overused in business writing; Kuhn's stricter sense (incommensurable scientific worlds) is rarer than product rebrands.
- Cross-cultural cognition work continues to show framing differences in attention and categorisation; effect sizes vary and are easy to overclaim.
F-R-A-M-E. Find the unspoken question; Restate it; Anomalies signal a bad frame; Mathematicians rename problems; Economics of popcorn (and Kuhn) show frames move markets and minds.
Invert: suppose frames are just words and the "real problem" is frame-invariant. Then popcorn tricks are marketing, Kuhn is sociology, and more data always helps. Try solving your current stuck project without renaming it. If you stay stuck, the chapter's claim gains force.
Write your stuck problem in one sentence. Then rewrite it five times, each time forbidding the main noun of the previous sentence. Pick the rewrite that makes a new first action obvious. Do that action for twenty-five minutes before returning to the original wording.
The framework through which we view a problem determines which solutions we can see. Restructuring, from jumbo popcorn to imaginary numbers, is often more powerful than better analysis. Culture shapes our frames; questioning our beliefs and exposure to other cultures loosen them.
6. Thinking When You're Not Thinking
Nature's plan B
In June 1816, during a cold, rainy summer near Lake Geneva, eighteen-year-old Mary Godwin (later Shelley) stayed with Percy Shelley, Lord Byron and others. Byron proposed that each write a ghost story. Everyone produced one except Mary, who for days had to give "a mortifying negative." One night the group discussed the "nature and principle of life," including experiments attributed to Erasmus Darwin and galvanism. Lying in bed after 2 a.m., not trying, "my imagination, unbidden, possessed and guided me." She saw a student kneeling beside the thing he had assembled. That vision became Frankenstein.
Elastic thinking is not a linear chain of steps. Ideas seem to appear from nowhere, but they come from associative processes that run when the conscious mind is quiet. Neuroscience now shows that quiet periods are a "plan B" for problem solving.
The dark energy of the brain
Marcus Raichle calls it the brain's "dark energy": background activity that consumes a large share of the brain's energy yet was long ignored. His 2001 papers named the default mode; his first article has been cited more than 7,000 times. But the story starts earlier.
A shy, self-critical psychiatrist at Jena, Berger secretly pursued an audacious goal: link the brain's physical energy to mental experience, inspired partly by a belief in telepathy. After a near-fatal fall from a horse, he received a worried telegram from his sister's urging, sent to his family at that moment, and wanted to explain it.
- He measured brain activity in patients with holes in their skulls from tumours or injuries, starting with a 23-year-old with an 8 cm gap from bullet surgery.
- By the mid-1920s he could record electrical activity from the scalp: the electroencephalogram (EEG), published in 1929 when he was 56.
- He found the brain is busy even when resting and that its electrical rhythm changes with mental effort. Few listened.
- Forced out by the Nazis in 1938 and depressed, he took his own life on 30 May 1941.
The symphonies in idle minds
Nancy Andreasen, an English PhD (on John Donne) who in the late 1960s turned to medicine despite discrimination against women, became a pioneer of PET imaging. Imaging experiments subtract a control condition, often "let your mind go empty," from a task condition. In 1995 Andreasen noticed that at rest "there wasn't just a little bit of activity, there was a great deal" in specific regions, and "it wasn't just a din, it was a symphony": distant areas firing in synchrony. She had found a coalition of structures, later called the default network.
Smarts by association
The default network lies in association cortices. The primary visual cortex detects edges and light; association areas give meaning, so that the lines of a NO TRESPASSING sign become a warning. Animals such as rhesus monkeys seem to make few abstract associations. "When your mind is at rest," Andreasen says, "what it is really doing is bouncing thoughts back and forth."
After a stroke in her frontal lobe damaged her default network, Patient J lay alert, answered questions and followed instructions, but never initiated conversation. Asked how she liked the hospital food, she would say "It's not very good" and nothing more. She later said that at the time her mind had simply been empty: no inner dialogue.
The importance of being aimless
Mlodinow co-wrote books with Stephen Hawking, who, paralysed by ALS, composed speech at about six words a minute by twitching a cheek muscle. At first Mlodinow waited impatiently; then he began to use the minutes to reflect, and found the conversations deeper. Our usual instant replies come from a superficial place.
- Participants asked not to text for two days felt "annoyed," "anxious" and "agitated."
- iPhone users kept from answering a ringing phone showed rising anxiety, heart rate and blood pressure.
- 73% of smartphone users said they panic when they misplace their phone.
- David Greenfield: "The Internet is the world's largest slot machine, and the smartphone is the world's smallest." Unpredictable rewards drive compulsive checking.
Constant activity leaves little default-mode time, which feeds our sense of self, connects divergent information and lets us step back. A relaxed mind explores novel ideas; an occupied mind settles for the most familiar. Ecopsychologists recommend unplugging: walks without phones, showers, jogging. Schedule mindless tasks when stuck, and clear your mental to-do list. Procrastination correlates with creativity. Leonardo sometimes spent half a day lost in thought while painting The Last Supper; he told an impatient prior that "the greatest geniuses sometimes accomplish more when they work less."
Berger's path in fuller detail
Hans Berger (1873 to 1941) at Jena was shy, routine-bound and secretly audacious. A near-fatal fall from a horse, followed by a telegram prompted by his sister's sudden fear, convinced him thought-energy might travel; telepathy is false, but the obsession drove rigorous measurement. Inspired by Angelo Mosso, he used blood flow as a proxy for energy (the principle behind later fMRI), fitting a water-filled rubber bladder into an 8 cm skull gap of a factory worker. He also fired a gun behind subjects to measure "involuntary attention." By the mid-1920s he reversed the apparatus: read the brain's own electricity from the scalp. EEG papers began in 1929 ("On the Electroencephalogram of Man," fourteen numbered sequels). He showed the resting brain is electrically active and that the signature shrinks when voluntary concentration begins. Colleagues treated the signal as noise. Fired by Nazi authorities in 1938, laboratory dismantled, he took his own life on 30 May 1941. A grandfather's poem hung on his study wall: each person faces an image of what they are meant to become.
Andreasen, Patient J, aimlessness
Nancy Andreasen, English PhD on John Donne turned physician despite discrimination, saw in 1995 PET controls a "symphony" of synchronised idle regions: the future default network. Patient J, after frontal stroke damaging that network, answered but never initiated; her mind felt empty. Mlodinow's waits with Stephen Hawking (about six words a minute) taught him that instant replies are shallow. Constant phone checking (unpredictable rewards as "the world's largest slot machine") crowds out default-mode time. Leonardo told an impatient prior that geniuses sometimes accomplish more when they work less.
From blood bladder to EEG to default mode (expanded)
Berger's first subject was a 23-year-old factory worker with an 8 cm skull gap after two surgeries to remove a lodged bullet. Cognitively intact despite seizures, he allowed a water-filled rubber bladder in the gap, linked to a volume recorder. Tasks included mental arithmetic and counting wall spots; for "involuntary attention" Berger stepped behind him and fired a gun. Early blood-flow work was weak; a 1910 book on brain temperature drew self-doubt. By 1920 he stimulated cortex with electrodes; in June 1924, working on a 17-year-old student, he reversed the setup to read brain electricity. Scalp EEG followed; thousands of readings included his own son. Fourteen papers after 1929 shared one title, numbered. He showed rest is electrically active and that the signature shrinks when voluntary concentration begins. Colleagues called it noise needed for heartbeat and breathing. On 30 September 1938 Nazi-approved authorities fired him; the lab was dismantled. His last published plea asked colleagues to notice the amplitude drop when "active conscious activity" begins. Few did.
Nancy Andreasen's path was unusual: Renaissance literature professor, then medicine after a hard pregnancy left her bedridden and thinking about mind and brain. PET subtraction used "empty mind" as control; she saw not a din but a symphony of distant areas in synchrony. The coalition later named the default network sits in association cortices that turn edges into meaning. Rhesus monkeys seem to make few abstract associations of the human sort. "When your mind is at rest," she says, "what it is really doing is bouncing thoughts back and forth."
Patient J answered and obeyed but never started talk; hospital food was "not very good," then silence. She later said her mind had been empty: no inner dialogue. That is what constant connectivity risks in milder form. Participants barred from texting for two days felt annoyed and anxious; iPhone users blocked from a ringing phone showed rising anxiety, heart rate and blood pressure. David Greenfield: the internet is the world's largest slot machine, the smartphone the world's smallest. Unpredictable rewards drive checking. Schedule mindless chores when stuck; clear the mental to-do list so the default network has room. Moderate procrastination can help idea quality; chronic delay does not.
Daydreaming, Mary Shelley and the empty mind (pass 3)
Berger's EEG showed that the "idle" signature shrinks the moment voluntary concentration starts. The scientific culture of his day treated residual brain activity as housekeeping for heart and breath; the patterned rest signal looked like noise. That paradigm blindness delayed recognition of the default network by decades. Andreasen's PET work later made the coalition visible: association cortices trading meaning when the task network stands down.
Mary Shelley's Frankenstein origin story (the waking dream after days of talk about galvanism and life) is the literary emblem of productive mind-wandering. The point is not laziness; it is that recombination needs off-task time. Patient J shows the opposite: obedient, fluent when prompted, empty when left alone. Constant connectivity risks a milder empty-mind: the phone supplies external prompts so the inner dialogue never starts.
Practical schedule: protect one daily block with no inbox and no "quick check." Put mindless chores (dishes, walking a known route) after hard analytical blocks so the default network can remix. Moderate, intentional delay on a hard creative problem often raises idea quality; chronic avoidance does not. Greenfield's slot-machine metaphor is useful only if it changes behavior: turn off badges, batch messages, and treat unpredictable pings as the addictive design they are.
Berger the brooder and Andreasen the double doctor (pass 4)
Colleagues called Berger shy, reticent, inhibited, brooding, detail-oriented and highly self-critical; fond of instruments more than people. That temperament built the EEG and also left him isolated when paradigms shifted past him. Fired 30 September 1938, lab dismantled, he still asked readers to notice the amplitude drop when active conscious work begins. Few did.
Andreasen near eighty: M.D. plus Ph.D. in English, close-cropped brunette, unusual even inside neuroscience. Literature first, then medicine after a hard pregnancy left her bedridden and thinking about mind. PET "empty mind" controls revealed synchrony across association cortices, not noise. Patient J's stroke emptied the default mode: answers when prompted, no self-started talk, later report of an empty mind. Recovery made the deficit visible in hindsight. Phone addiction studies (two-day text bans; blocked ringing iPhones) show anxiety and autonomic rise when the external prompt stream stops. Greenfield's largest/smallest slot-machine line is memorable because variable rewards are the mechanism, not because phones are literally casinos.
Schedule for the idle symphony (pass 5)
Treat mind-wander as a calendar item, not a leftover. After a hard analytical block, take a walk without podcasts, or do dishes without a screen. Keep a capture pad for associations that surface. If Patient J's empty mind is the pathology of no default dialogue, the modern risk is outsourced dialogue to the feed. Recover the inner turn on purpose.
Mary Shelley, Berger, Andreasen
June 1816, Cologny near Geneva: eighteen-year-old Mary Godwin cannot invent a ghost story while Byron’s circle can. After talk of Erasmus Darwin’s moving vermicelli and the principle of life, she lies awake past 2 a.m.; the image of a pale student kneeling beside a creature assembled from parts arrives as insight during restless idle. Frankenstein is Plan B of the default mind.
Hans Berger: factory worker with an 8 cm skull gap; water bladder; gunshot for involuntary attention; June 1924 EEG on a 17-year-old; fourteen papers after 1929 with one numbered title. Fired 30 Sep 1938; lab dismantled. Rest is electrically active; concentration shrinks the signature.
Nancy Andreasen: literature professor → medicine; PET “empty mind” control reveals a symphony, later the default network. Patient J answered but never started talk; mind felt empty. Smartphones as pocket slot machines (Greenfield). Moderate procrastination can raise idea quality; chronic delay does not.
Shelley’s creature, Berger’s rest rhythm and Andreasen’s default symphony agree: elastic work often happens when you stop forcing the analytical gear.
Default minds: Shelley, Berger, Andreasen
Mary Godwin at Cologny (June 1816) could not invent a ghost story on command. After talk of Darwin’s vermicelli and the principle of life, past 2 a.m., the pale student and assembled creature arrived as insight in restless idle. Frankenstein is Plan B of the default mind.
Hans Berger turned a factory worker’s skull gap into a research window, then recorded the first human EEG (June 1924). Rest is electrically active; concentration changes the signature. Fired 30 Sep 1938, lab dismantled, he still left the rhythm that later mapped onto default-network talk.
Nancy Andreasen moved from literature to medicine. PET “empty mind” controls showed a symphony, later named the default network. Patient J answered but never started talk; the mind felt empty. Smartphones act as pocket slot machines (Greenfield). Moderate procrastination can raise idea quality; chronic delay does not. The exploratory drive chapter also floats a bottleneck hypothesis (few breeding humans, selection for boldness) that the A closer look should keep disputed.
Shelley’s creature, Berger’s rest rhythm and Andreasen’s default symphony agree: elastic work often arrives when you stop forcing the analytical gear. Schedule idle on purpose; do not only steal it from exhaustion.
IDLE ≠ OFF. Default-mode activity is work. Protect some of it from pings.
Idle design. Default-mode recombination is not the same as fatigue scrolling. Design idle: walk without podcasts, shower without planning the next slide, stare at a window after a failed proof. Berger’s rest rhythm and Andreasen’s symphony are reminders that the offline brain is working. Steal idle from the calendar, not only from midnight.
- The default mode network does more than daydreaming: it is involved in memory, self-reflection, social reasoning and some focused tasks. "Rest = DMN, task = executive" is a simplification.
- Berger's telegram anecdote supports telepathy no more than coincidence does; it motivated him but was not evidence.
- Smartphone "addiction" is not a formal diagnosis in DSM-5-TR. The famous "73% panic" figure comes from a commercial survey. Large studies (Orben and Przybylski) find small average links between screen time and well-being, though heavy use and notifications do disrupt attention.
- Procrastination and creativity: evidence is mixed; Adam Grant and Jihae Shin's work suggests moderate procrastination helps idea quality, while chronic procrastination harms performance.
- Patient J illustrates akinetic mutism from frontal (medial) damage; attributing it specifically to the default network is an interpretation.
- Intracranial work: stimulating the DMN in epilepsy patients reduced number and originality of alternate-uses ideas (Shofty and colleagues; Brain 2024), strong causal evidence.
- 2025 Cerebral Cortex (Luchini, Beaty and colleagues): covert neurofeedback strengthening default-executive coupling raised originality a day later.
- Roger Beaty's program shows creative people flexibly couple default, executive and salience networks; resting connectivity predicts subset of creative achievement.
- Incubation meta-analyses (Sio and Ormerod 2009, later updates) find modest benefits, larger for divergent tasks and undemanding breaks.
- School phone bans (2023 to 2025) aim to restore attention; early achievement effects are mixed.
B-A-R-D. Berger saw rest electricity; Andreasen heard the symphony; Raichle named default mode; Downtime (and Hawking waits) is the practical lever. Idle is plan B.
Invert: suppose the default network is only mind-wandering waste, and every breakthrough comes from longer deliberate practice. Then Mary Shelley's bed, Leonardo's pauses and phone-free walks are romantic myths. Causal DMN stimulation results (2024) make that inversion harder to hold.
Protect one idle block daily: a 20-minute walk with no phone and no podcast. Before it, review the stuck problem for two minutes, then forbid yourself from solving it. Keep a pocket notebook for what surfaces in the hour after. At work, batch notifications into two-hour windows. Once a week, schedule a "Leonardo half-day": half the usual output target, double the aimless incubation.
Ideas often arrive when we are not trying, because the default network links remote concepts in idle moments. Berger glimpsed it, Andreasen saw its symphony, Raichle named it. Constant connectivity crowds it out. Downtime is not wasted time.
7. The Origin of Insight
When the unimaginable becomes the self-evident
On 21 December 1941, two weeks after Pearl Harbor, President Roosevelt told his Joint Chiefs that Japan must be bombed soon, to lift American morale and shake Japanese confidence. No bomber had the range. Weeks later submarine captain Francis Low, watching bombers practise on a carrier-shaped outline painted on a Norfolk, Virginia airfield, had an idea: launch army bombers from an aircraft carrier.
Every expert knew why it could not work: carriers were for light fighters, bombers were too heavy for a short deck, and they could not land back on one. Low nervously took the idea to Admiral Ernest King. In desperate times it was adopted. Stripped-down B-25s with extra fuel were flown by 80 volunteers in 16 planes, led by Jimmy Doolittle, on 18 April 1942. They bombed Japan and flew on to China.
Sometimes the most powerful realisation is that circumstances have changed and the old rules no longer apply. Carriers and bombers did not fit together under normal rules; after Pearl Harbor, they did. Low called his idea "fortuitous." Neuroscience now shows that such sudden ideas are the end of a long hidden process.
Splitting the brain
In the late 1950s Roger Sperry studied cats whose corpus callosum, the bridge between the hemispheres, had been cut. Conventional wisdom held that the left hemisphere did the real work and the right was nearly useless. With one eye covered, cats learned to tell a triangle from a square; tested through the other eye, they had to learn it afresh. Each hemisphere seemed to be a separate mind.
Neurosurgeon Joseph Bogen had argued for splitting the human brain to treat severe epilepsy. In 1962 he operated on the first of 16 patients Sperry tested. One woman, asked how many seizures she had had, raised two fingers with her right hand while her left hand pulled it down and showed one; the hands fought "like angry children." She called her left hand the maverick, speaking for the left hemisphere, which controls speech and the right hand. The right hemisphere understands words but cannot speak. Sperry shared the 1981 Nobel Prize.
The trial of the hemispheres
John Kounios, studying language with EEG-based ERPs (event-related potentials), saw an analogy: understanding a sentence is like solving a puzzle. Words have many meanings, and the brain holds them in working memory until context picks one. "The cooking teacher said the young children made bad snacks" versus "The cannibal said the young children made bad snacks": one word changes the meaning of "made." Our unconscious notes the remote meanings but does not usually show them to us.
At the NIH, Mark Beeman studied patients with right-hemisphere damage, who struggled with jokes, metaphors and inferences. His theory: the left hemisphere focuses on the dominant, literal meaning, while the right keeps remote associations alive. Jokes depend on the remote meaning, as in Conan O'Brien's line about a rapper switching to a "more traditional" insult.
The lessons of CRAs
In late 2000 Kounios (EEG, good timing) and Beeman (fMRI, good location) teamed up. They used compound remote associates (CRAs): find one word that forms a compound with each of three. Try pine, crab, sauce. Two words pull you toward strong but wrong associates (pine tree, crab cake). Subjects had 30 seconds and solved 59%. They reported whether the answer came by insight or by step-by-step analysis. (Answer: apple.)
Key findings:
- Insight solutions were preceded by a burst of gamma waves in the right anterior superior temporal gyrus, about a third of a second before subjects knew the answer, the moment a remote association broke through.
- Just before that, alpha waves over the right visual cortex signalled the brain shutting out visual input, like closing your eyes to think.
- The anterior cingulate cortex (ACC) monitors competing candidate answers and, when the obvious ones fail, broadens attention.
- Even before a puzzle appeared, some subjects' brains showed an insight-ready mindset, and a positive mood made insight more likely.
As a young physicist, impatient to finish a tedious but sure approach, Mlodinow kept grinding. Only when he relaxed on a date did an elegant solution pop into his head, and he had to ask a captivating woman to wait five minutes while he checked it. Lesson: impatience locks you into a suboptimal approach.
Zen and the art of ideas
A Zen meditator at Kounios's lab failed CRA after CRA because his focus was too narrow; then, apparently broadening his attention at will, he solved nearly all the rest. Kounios had never seen anyone improve within a session. A 2012 study found that open-monitoring meditation enhances the ability to broaden focus.
Rate from 1 (almost always) to 6 (almost never): breaking things through inattention; rushing without noticing the way; missing physical tension; forgetting names instantly; listening with one ear. Scores range 5 to 30; average about 15; two-thirds score 12 to 18. Higher means more mindful. (The items come from the Mindful Attention Awareness Scale, MAAS.)
Environment matters too: darkened rooms, closed eyes, high ceilings and expansive spaces widen thinking; low ceilings and windowless offices narrow it. Remove time pressure and above all interruptions: even expecting a message pulls the mind outward.
Language, hemispheres and CRAs in fuller form
John Kounios saw sentence comprehension as puzzle-solving: words hold many meanings until context picks. "The cooking teacher said the young children made bad snacks" versus "The cannibal said..." flips "made." Mark Beeman found right-hemisphere damage hurts jokes, metaphors and inferences. Theory: left hemisphere favours dominant literal meaning; right keeps remote associations alive.
They met in late 2000. Tool: compound remote associates (CRAs). Example: pine, crab, sauce → apple (pineapple, crab apple, apple sauce). Subjects had 30 seconds, solved about 59%, and reported insight versus analysis. Insight trials showed a gamma burst in right anterior superior temporal gyrus about 0.3 s before awareness, after an alpha burst over right visual cortex (brain "closing its eyes"). The ACC monitors conflict and broadens attention when obvious answers fail. Positive mood and a pre-trial insight-ready state raised odds.
Zen, environment, Mlodinow's date
A Zen meditator first failed CRAs through over-focus, then broadened attention and solved nearly all remaining: Kounios had never seen within-session recovery like it. 2012 work linked open-monitoring meditation to broader focus. Darkened rooms, closed eyes, high ceilings and low interruption widen thought; time pressure and expected messages pull attention outward. Young physicist Mlodinow, grinding a tedious sure method, found the elegant solution only when relaxed on a date and had to ask her to wait five minutes while he checked it.
CRAs, ACC and the insight mindset (expanded)
Kounios brought EEG timing; Beeman brought fMRI location. Parallel labs, same puzzles. They avoided the acronym "CRAP" for compound remote associates. Design: three words, one solution that compounds with each. Strong lure associations (pine→tree/cone, crab→cake/meat) must be dropped so weaker links can win. Analytical attack generates associates one by one; insight relaxes until apple appears. About 40% more puzzles were solved by insight than by analysis in their reports. Beeman's subjects lay in a scanner; Kounios's sat in a hot lab with broken air-conditioning, sweaty electrode caps interfering with readings. The sweat was worth it: the study became a classic.
Deconstruction: left hemisphere raises obvious answers; right (especially right anterior superior temporal gyrus, a fold above the right ear) raises oddball ones. The ACC, above the corpus callosum, monitors and, when obvious paths fail, broadens attention. It arranges for the right visual cortex to quiet, like closing your eyes, so right-aSTG ideas can strengthen and burst into awareness. That is why grit at impasse matters: frustration is the moment the judge may loosen the leash. Separately, some brains show an insight mindset seconds before the puzzle appears: a relaxed, positive state that primes the right hemisphere to be heard. Impatience locks the suboptimal path; a date-night linguine order can unlock the elegant one.
The Zen meditator failed until he deliberately broadened, then nearly cleaned the set. Across hundreds of subjects Kounios had never seen within-session practice improve CRA success; only this meditator's control of mindset did. 2012 open-monitoring meditation work supports trainable broadening. Environment: dark, quiet, high ceilings, few interruptions. Expected messages pull the mind outward even before they arrive.
ERP timing, remote associates and the Aha sequence (pass 3)
Kounios came from ERP work on language comprehension: event-related potentials timed how the brain builds meaning from words. He noticed an analogy between sentence understanding and insight: both need remote associations to win over local ones. Beeman's fMRI lab and Kounios's EEG lab then attacked the same CRA puzzles from two angles. The design forces conflict: strong associates of each cue word are wrong; the weak triple-link (pine / crab / sauce → apple) is right.
Sequence under insight (simplified): left-hemisphere obvious paths fail → ACC detects conflict and broadens → right visual cortex quiets (inward turn) → right aSTG strengthens oddball candidates → sudden awareness with the Aha feeling. Analytical solvers stay in left-dominant serial search longer. About 40% more puzzles fell to insight than analysis in their reports, but the deeper claim is mechanism, not a scoreboard.
Mindset precedes the item: seconds before a puzzle appears, a relaxed positive state predicts insight success. Impatience and threat lock the suboptimal path. The Zen meditator's within-session jump, which Kounios had not seen from practice alone across hundreds of subjects, shows that control of breadth can be trained. Environment still matters: dark, quiet, high ceilings, few expected interruptions. An expected ping pulls attention outward even before it arrives.
Sperry's two minds and the CRA naming dodge (pass 4)
Roger Sperry's split-brain work, half a century earlier, framed the hemispheres as semi-independent. Left-side processing favors strong, local associates; right-side (especially right aSTG) favors oddball remote ones. Insight needs both: the left to propose, the right to surprise, the ACC to referee when conflict rises. Before imaging and ERP timing, insight research stalled on self-report; the Aha feeling was vivid and scientifically thin.
The naming dodge matters culturally: psychologists already had the remote associates test (RAT); Kounios and Beeman's compound remote associate (CRA) variant avoided the acronym that would have spelled a word nobody wants in a paper title. Design seriousness underneath the joke: three words, one compound solution, strong lure associates that must be dropped. Subjects report method after each solve; insight led to about 40% more solutions than analysis in their reports. Beeman's scanner subjects and Kounios's sweaty EEG subjects jointly mapped timing and place. The Zen meditator's session remains the chapter's training parable: focused mind failed; deliberately broadened mind nearly cleaned the set.
Before the puzzle: the insight mindset (pass 5)
Seconds before a CRA appears, EEG can already show a relaxed, positive state that predicts insight success. That finding reframes "talent" as partly preparable state. Threat, hurry and expected interruptions pull the opposite state: locked local search. Practical prep is boring and effective: silence badges, sit where ceilings feel high, accept a minute of mind-wander before the hard item, and treat the first failed obvious answer as a signal to broaden rather than as personal failure. The ACC's conflict signal is information, not shame.
Doolittle, CRAs and the insight mindset
21 Dec 1941: Roosevelt demands Japan be bombed; range seems impossible. Captain Francis Low sees bombers practicing on a painted carrier outline and ignores “expert” impossibility: launch B-25s from a carrier. Doolittle Raid follows. Ignorance (or pretending what you know is false) unlocked the frame.
Kounios (EEG timing) and Beeman (fMRI location) use compound remote associates; about 40% more puzzles solved by insight than analysis in their reports. Left hemisphere raises obvious associates; right anterior superior temporal gyrus raises oddballs; ACC broadens when paths fail. Insight mindset seconds before the puzzle appears predicts success. Zen open-monitoring (2012 work) trains broadening. Environment: dark, quiet, high ceilings, few pings.
| Piece | Role |
|---|---|
| CRA lure words | strong wrong associates |
| Right aSTG | oddball remote links |
| ACC | loosen failed path |
| Pre-trial mood | insight mindset |
Insight labs and the Doolittle reframe
21 Dec 1941: Roosevelt wants Japan bombed soon; range seems impossible under old carrier doctrine. Captain Francis Low sees bombers practicing on a painted carrier outline and ignores expert impossibility. Launch B-25s from a carrier. The Doolittle Raid (18 Apr 1942, sixteen aircraft, eighty volunteers) follows; Midway later compounds Japan’s carrier losses. Ignorance of the “impossible” unlocked the frame.
Kounios (EEG timing) and Beeman (fMRI location) use compound remote associates. Roughly 40% more puzzles fall to insight than analysis in their reports. Left hemisphere raises obvious associates; right anterior superior temporal gyrus raises oddballs; ACC broadens when paths fail. Insight mindset seconds before the puzzle appears predicts success. Zen open-monitoring trains broadening. Environment: dark, quiet, high ceilings, few pings. The unimaginable becomes self-evident only after someone drops the expert rule that forbade the move.
| Path | Signature (lab sketch) | Best when |
|---|---|---|
| Analytical | stepwise, verbalisable | well-defined search |
| Insight | sudden, oddball associates, prior broad mindset | remote links needed |
Default-executive neurofeedback and DMN stimulation studies (see Ch 4 update) make “insight mindset” less mystical: state before the trial matters, and some states are trainable.
Insight conditions checklist. Before a remote-associate style task: reduce pings, dim lights if you can, sit with a slightly open question for two minutes, then start. After the flash, switch to analytical verification. The lab story is not that insight is magic. It is that state before the trial and hemisphere-scale association patterns change odds.
- Doolittle Raid facts: Japanese leaders were shocked, and the raid did push them to attack Midway, where in June 1942 they lost four fleet carriers to the US Navy, which had broken their codes. The book's "sank all but one of their carriers" overstates; Japan still had other carriers. The raid did little physical damage; its effect was on morale and strategy.
- Left brain/right brain: popular claims of "left-brained" and "right-brained" people are myths (Nielsen and colleagues 2013 found no such individual differences). Kounios and Beeman's specific right-temporal finding is more modest and has had mixed replication.
- The "high ceilings" effect (Meyers-Levy and Zhu 2007) comes from small priming studies of a kind that often fails to replicate.
- Mindfulness and creativity: meta-analyses find small positive effects, mostly for open-monitoring meditation and divergent thinking.
- The ACC as judge is a theory, as Mlodinow acknowledges.
- Aha moments boost memory (Becker and colleagues 2023 to 2025): stronger hippocampal and dopaminergic signatures, linking Chapters 3 and 7.
- Insight answers are not always correct but beat analytical guesses under time pressure more often than chance (Salvi, Beeman and colleagues).
- Brief drowsy N1 sleep tripled discovery of a hidden rule (Lacaux and colleagues 2021), echoing Edison-style naps.
- LLMs solve many CRA items by association search; the phenomenology of sudden insight is not required for the behavioural score.
- Split-brain: Pinto and colleagues (2017) argue for more unified awareness than the classic "two minds" textbook story.
P-F-I-B-A. Insight sequence: Prepared mind → Focused search → Impasse → Broadening (alpha) → Aha (gamma, ~0.3 s). ACC is the judge that loosens the leash.
Invert: suppose "insight" is just fast analysis we fail to introspect. Then gamma bursts are epiphenomena, CRAs are vocabulary tests, and Doolittle is ordinary ops research. The chapter needs the hidden sequence to be causally special, not only subjectively sudden.
Work a stuck problem for 25 focused minutes, write the failed approaches, then change state: eyes closed in a quiet room, or a walk in a high-ceiling space, phone silenced. After the break, note whether anything "popped" versus whether you resumed grinding. Practise ten online CRA items and label each solution insight or analysis.
Insight feels instant but follows a hidden sequence: focused search, impasse, broadened attention, and a burst of right-hemisphere activity as a remote association breaks through. A relaxed, positive, uninterrupted mind is primed for it.
8. How Thought Freezes Over
Building lives and candleholders
Jonathan Franzen planned his first life early: marry his Swarthmore sweetheart and write big, socially important novels that would change the culture. His first novel, The Twenty-Seventh City (1988), drew good reviews but modest sales; the second, Strong Motion (1992), did worse. His marriage collapsed. To reinvent himself he had to overcome functional fixedness: the difficulty of seeing a tool used in a new way.
Given a candle, a box of tacks and matches, attach the candle to the wall so it burns properly. People try to tack the candle or glue it with wax. The answer: empty the box, tack it to the wall, and stand the candle in it. Under a deadline of several minutes, about three-quarters of adults fail. Young children, with fewer fixed ideas about what boxes are for, often do better. When the tacks are presented beside the box rather than in it, the solution becomes far easier.
Once on a path, we tend to stay on it, not because we fear change but because we are used to thinking a certain way.
The momentum of thought
Functional fixedness is one case of a broader momentum of thought. A change of circumstances can supply the force to redirect it: a nudge for some, a brick wall for others. Franzen hit the wall. "I realized I was trying to be a kind of writer that I wasn't best at being." He stopped writing for the masses and wrote for people who love books, about the problems of a set of characters. The Corrections (2001) won the National Book Award, and he landed on the cover of Time as "Great American Novelist."
"The difficulty lies, not in the new ideas, but in escaping from the old ones, which ramify... into every corner of our minds."
When thought freezes over
Around 1900, Sir James Jeans helped derive a theory of blackbody radiation from Newton's laws and electromagnetism; it predicted infinite energy at high frequencies. Max Planck introduced the quantum principle, which fitted the data. Jeans resisted for years. Hannah Arendt called such unquestioned principles "frozen thoughts," akin to thoughtlessness.
Frozen thinking is a particular danger for experts. Deep knowledge is valuable for usual problems, but it can blind you to novel ones.
- A JAMA study of ten years of data found that 30-day mortality among high-risk cardiac patients was about one-third lower when top cardiologists were away at national conferences. The authors linked many errors to forming opinions quickly from prior experience.
- An Israeli study of 119 geriatric nursing-home patients taking an average of seven drugs a day carefully discontinued about half the medications. No one suffered serious harm, most reported better health, and mortality was far below a control group.
- A paediatrician about to refer a six-year-old for ADHD testing learned the boy was using his asthma inhaler more. He recognised that the inhaler can cause hyperactivity, adjusted the asthma treatment, and the problem disappeared.
Destructive doctrine
The military is prone to frozen thinking because authority and doctrine are prized. General Stanley McChrystal, who commanded US and international forces in Afghanistan and Joint Special Operations Command, transformed operations by flattening decision-making and attacking networks. His successor David Petraeus told Mlodinow: "it is often the side that adapts the fastest that prevails."
From August 1973, Israeli intelligence saw Syria moving Soviet anti-aircraft missiles to the border; Egypt moved reserves and convoys to the Suez Canal. By Yom Kippur, Syria had an eight-to-one advantage in tanks on the Golan. Intelligence chiefs Eli Zeira and Yona Bandman accepted Egypt's "training exercise" story. On 6 October Israel was surprised; by 8 October Moshe Dayan warned Golda Meir the state was in danger.
The Agranat Commission blamed faith in "Ha'Conceptzia" (the Concept): Egypt would not attack without air superiority, and Syria would not attack without Egypt. The Arabs had redefined air superiority as surface-to-air missile cover. Bandman had even suppressed a subordinate's dissenting report. McChrystal: "The essence of military deception is to figure out what the enemy is preset to believe, and to play to that."
McChrystal saw the same in Iraq: expecting terrorists to have certain limits, the US was outpaced until he dismantled and rebuilt his organisations. He was fired in 2010 after aides' remarks to Rolling Stone.
Handicapping the expert brain
Psychologists showed chess experts board positions. On a one-solution board, all found a clever five-move checkmate. On a two-solution board, which also allowed a familiar but longer mate (the "smothered mate"), experts locked onto the familiar plan and failed to find the better one, even while believing they were searching. This Einstellung effect reduced performance by the equivalent of about 600 rating points, roughly three standard deviations (like 45 IQ points).
The benefits of discord
Psychologists call frozen thinking dogmatic cognition: processing information so as to confirm prior opinions. The ideal expert keeps a beginner's mind. One effective thaw is dissent.
Serge Moscovici, a Holocaust survivor, showed groups blue slides. In one group, planted confederates called them "green." No one was persuaded; everyone answered "blue." But in a later "unrelated" test sorting blue-green colour chips, those exposed to the dissenters labelled more chips "green." A minority view no one accepted had shifted their boundaries unconsciously. Other studies show dissent improves thinking even on unrelated problems.
Dogmatic leaders often punish dissent, as Bandman did. Diversity in universities and companies helps not only through new ideas but because the presence of different views makes everyone think more flexibly.
Franzen, candles, Jeans, doctrine, experts, discord
Jonathan Franzen describes a first life-script (literary marriage, masterworks by thirty) that froze until he rebuilt another. Duncker's candle problem (1945) shows functional fixedness: the matchbox is seen as a container, not a shelf. Sir James Jeans's beautiful blackbody theory failed until quantum ideas unfroze the frame. Military doctrine, says General Stanley McChrystal, institutionalises authoritative thought; the Yom Kippur War (1973) surprise is a classic of frozen expectations. Expertise helps until the distribution shifts: some medical studies suggest less-experienced doctors can outperform on unusual cases. Zen beginner's mind is the antidote to dogmatic cognition. Discord and dissent keep frames from icing over.
Frozen lives, candles, physics, war, experts (expanded)
Jonathan Franzen's first life-script (Swarthmore sweetheart, literary masterworks by thirty, teaching jobs) froze when the marriage and plan cracked; the second life required dropping the old blueprint. Duncker's candle: fix a candle to a wall so wax does not drip on the table, using a candle, matches and a box of tacks. People who see the box only as a container miss using it as a shelf. Functional fixedness is one face of momentum of thought: minds in motion stay in motion until a force (dissent, anomaly, beginner's mind) acts.
Sir James Jeans built a beautiful blackbody theory from Newton and electromagnetism; data refused it. The maths was fine; the frame was not meant for that regime. Quantum ideas thawed the freeze. In war, doctrine hardens faster: General Stanley McChrystal notes the military institutionalises authoritative thought. The 1973 Yom Kippur surprise punished Israeli expectations that had calcified after prior victories. Expertise helps on the training distribution and hurts when the distribution shifts; some medical studies find less-seasoned clinicians relatively stronger on rare presentations. Dogmatic cognition processes information to reinforce priors. Zen beginner's mind is the named opposite: meet even routine situations as if new. Discord is not rudeness; it is a thaw protocol.
Experts, referees, Lange and hospital novices (pass 3)
Frozen thought is not only a personal vice; it is a professional hazard. Paper referees who already "know" what the authors mean misread hurried. Golfers cannot easily alter a stroke wired into motor cortex; professional thinkers cannot easily shed prefrontal conventions. Dorothea Lange: knowing ahead what you look for means photographing your own preconceptions.
Medicine sharpens the cost. A 2014 JAMA Internal Medicine analysis found that, for certain hospital outcomes, patients of relative novices sometimes fared better than those of the most experienced physicians, a result that jars intuition and invites careful reading rather than slogan. The mechanism Mlodinow stresses is not that experience is worthless; it is that experience encodes scripts that fail when the case leaves the training distribution. Public-health researchers are still unpacking when that tradeoff bites.
War and doctrine accelerate freeze: after victories, expectations harden; 1973 Yom Kippur punished Israeli assumptions. Businesses that refuse to question past success meet Nike- and Apple-style reframes as existential threats. The thaw protocol is institutional as well as personal: invite dissent, rotate reviewers, run red-team sessions, and practice beginner's mind on routine cases so rare ones do not arrive as total surprises.
Franzen's second life and Jeans's beautiful wrong theory (pass 4)
Franzen's first script (literary masterworks by thirty, teaching posts, the Swarthmore plan) shattered with the marriage; the second life required dropping the blueprint, not polishing it. Sir James Jeans built an elegant blackbody account from Newton and classical electromagnetism; data refused the elegance. The maths was fine; the frame was wrong for that regime. Quantum ideas thawed the freeze. Personal freeze and scientific freeze share a structure: momentum of thought plus identity invested in the old path. Discord, anomaly and beginner's mind are the external forces that can act.
Red teams and beginner's mind as thaw protocol (pass 5)
Institutional freeze needs institutional thaw: rotate referees, run red-team sessions, and practice beginner's mind on routine cases so rare ones do not arrive as total surprises. Personal freeze needs a smaller version: one trusted dissenter, one anomaly log, one weekly assumption you deliberately reverse for an hour. Momentum of thought is physical metaphor and practical diagnosis: without an external force, the path continues.
Franzen, candles and frozen doctrine
Jonathan Franzen’s first life-script (Swarthmore sweetheart, masterworks by thirty, Victorian house, political novels) froze when marriage and plan cracked; second life required dropping the blueprint. Duncker’s candle: box of tacks must become a shelf; functional fixedness is momentum of thought. Sir James Jeans’s beautiful classical blackbody theory failed until quantum frames thawed it. McChrystal: military institutionalises authoritative thought; 1973 Yom Kippur punished Israeli expectation after prior victories. Expertise helps on the training distribution and hurts when it shifts. Zen beginner’s mind is the named opposite of dogmatic cognition.
“Just be creative” without dissent channels fails in hierarchies. Freeze is often incentive design, not only personality.
Frozen scripts: Franzen, candles, doctrine
Jonathan Franzen’s first life-script (Swarthmore sweetheart, masterworks by thirty, Victorian house, political novels) cracked with marriage and plan. Second life required dropping the blueprint. Duncker’s candle: the box of tacks must become a shelf; functional fixedness is momentum of thought. Sir James Jeans’s beautiful classical blackbody theory failed until quantum frames thawed it. McChrystal notes that military culture institutionalises authoritative thought; 1973 Yom Kippur punished Israeli expectation after prior victories. Expertise helps on the training distribution and hurts when it shifts. Zen beginner’s mind names the opposite of dogmatic cognition.
Freeze is often incentive design, not only personality. Hierarchies that punish dissent get silent anomalies until crisis.
“Just be creative” without dissent channels fails in hierarchies. If promotion kills messengers, elastic talk is decoration.
Past success freezes doctrine. After a win streak, treat “what would make this strategy fail next time?” as a required exercise, not pessimism.
Unfreezing drills. Write the doctrine you are most proud of at work. List two conditions under which it would fail. Assign someone to argue those conditions in the next planning meeting. If no one is allowed to argue them, you do not have a team; you have a chorus. Yom Kippur and Franzen both punish unbroken scripts.
- The JAMA cardiology study (Jena and colleagues, 2015) is observational; a leading explanation is fewer elective, high-risk procedures during conferences, not only frozen thinking. It does not show that top doctors are worse.
- The Israeli deprescribing study (Garfinkel and Mangin) was small and not randomised; later randomised deprescribing trials show it is generally safe but mortality benefits are uncertain.
- Inhaled asthma drugs (albuterol) can cause jitteriness; hyperactivity is a recognised but uncommon side effect.
- Chess Einstellung (Bilalić, McLeod and Gobet 2008) used small expert samples; "600 points" applied to the specific task, not general skill.
- The minority-influence effects of Moscovici are real but modest and context-dependent in replications.
- Franzen's reinvention also involved luck and Oprah's 2001 book club controversy, which boosted sales.
- "Beginner's mind" is widely co-opted in business training; evidence for short workshops transferring to real diagnostic skill remains thin.
- Expert underperformance on unusual cases is real in some medical and forecasting literatures, but average expert advantage on typical cases remains large.
- Structured dissent (red teams, premortems) outperforms unstructured "be creative" exhortations in organisational studies.
C-D-E-B. Candle (functional fixedness), Doctrine (military freeze), Expertise handicap on novelty, Beginner's mind + discord as thaw.
Invert: suppose expertise almost never freezes thought, and dissent mostly adds noise. Then candle problems are toys, Yom Kippur is unique intel failure, and organisations should maximise credentialed consensus. Watch whether your own last miss was ignorance or a frozen frame.
Name one tool or role you always use the same way (the matchbox-as-box habit). List three forbidden uses. Run a premortem on your current project: assume it failed in six months because the team was expertly blind; write the blind spot. Invite one person who disagrees with your frame to a fifteen-minute dissent slot; capture their frame without rebutting.
Expertise and habit create momentum: functional fixedness, frozen thoughts, doctrine and the Einstellung effect. Experts are especially vulnerable in novel situations. Dissent, newcomers and deliberate challenges to your "concept" help thaw frozen thinking.
9. Mental Blocks and Idea Filters
When believing means not seeing
Mlodinow's father, who fought in the Jewish underground during the Second World War, taught him a Yiddish saying: "When a worm sits in horseradish, it thinks there's nothing sweeter." His father added: "And if the worm sits there long enough, everything looks like horseradish."
At a late-night Monopoly game at an English conference, Mlodinow, as banker, openly put in $20 or $50 bills and took out $100s. No one noticed. After he won and confessed, some colleagues refused to believe him. Their eyes had registered it, but their brains filtered it out: no one expected a physicist to cheat.
The conscious mind processes roughly 40 to 60 bits per second; the visual system alone handles about 10 million. Between them stand cognitive filters that guess what matters, using expectations rooted in beliefs and experience. Ideas are filtered the same way. Tiling a bathroom, your unconscious may associate marble, ceramic, and also Peppermint Pattie wrappers. Filters suppress the silly ones, but sometimes they censor good ones too, such as bamboo or rubber.
Thinking outside the box
In 1916 groceries still worked as in the nineteenth century: customers asked clerks behind counters to fetch goods. Grocer Clarence Saunders, returning by train from a failed mission to help a department store, saw a sow with six piglets feeding themselves at a pig farm. Why not let customers serve themselves? He invented the pieces: shopping baskets, price labels, aisles of shelves, checkout counters. He opened the first Piggly Wiggly in 1916 and patented the design in 1917; within six years there were 1,200 stores in 29 states. The chain still exists, mostly in the South.
The nine-dot problem: connect nine dots in a 3×3 grid with four straight lines without lifting your pencil. Often zero, and almost always fewer than one in ten, people solve it. The solution requires lines to go outside the imaginary square formed by the dots, which nobody told you was there.
Success rises when two extra dots are added outside, or a larger box is drawn around the grid: the brain drops its imagined boundary. Our filters are shaped daily as some neural responses strengthen and others weaken, leaving a brain adapted to its environment but wired to see through the lens of past experience.
Our idea-filtering system
In 2012 two Australian scientists gave the nine-dot problem to 30 people; only one solved it, a man with a severe childhood head injury. Could the injury have weakened his filter? They then used transcranial direct current stimulation (tDCS), sponge electrodes on the scalp, to temporarily dampen part of the filtering system. Subjects had all failed earlier. Controls, given sham stimulation, still failed; 40% of stimulated subjects solved it.
A key filter is the lateral prefrontal cortex, found only in primates. It suppresses ideas that break rules. After a stroke, Mlodinow's father, walking past a diner's table, grabbed a handful of fries: his brain no longer censored the idea. We use all our brain (the 10% claim is a myth), but we can benefit by adjusting our filters.
Long live the sophomoric
Mlodinow's daughter Olivia, eleven, told her ninety-year-old grandmother her "face looks like a raisin." Some associations are best kept quiet, but too much "think before you talk" stops ideas arising at all. To have original thoughts, let ideas flow first and judge them later. Value is hard to judge: cold fusion sounded nutty and was; metamaterials (materials with properties not found in nature, even near-invisibility) sounded nutty and were not. Bob Kearns invented a self-dispensing hair-tonic comb (nutty) and the intermittent windscreen wiper (worth tens of millions). Linus Pauling: "The way to get good ideas is to get lots of ideas and throw out the bad ones."
The story conference for Raiders of the Lost Ark (Lucas, Spielberg, Kasdan) shows unfiltered brainstorming at its rawest: alongside the ideas that made the film, the transcript records some that were deeply inappropriate and rightly discarded. Seth MacFarlane, creator of Family Guy, told Mlodinow that fear of social-media mobs makes it hard to stay creatively free; critics called his mind "singularly sophomoric." Children are elastic thinkers partly because they have absorbed less culture and partly because their prefrontal cortex matures last, into the mid-twenties. The beginner's mind is the strength of youth; expertise is the strength of age.
Horseradish, Piggly Wiggly, nine dots, sophomores
Mlodinow's father valued a Yiddish saying: when a worm sits in horseradish it thinks nothing is sweeter; sit long enough and everything looks like horseradish. Clarence Saunders in 1916 studied an innovative grocery and then opened Piggly Wiggly, the self-service model that reframed shopping. The nine-dot problem (connect three rows of three dots with four lines without lifting the pen) punishes staying "inside the box." In 2012 Australian researchers found only one of thirty subjects solved it; that subject had childhood traumatic brain injury, a clue that weakened filtering might help. tDCS studies claimed temporary filter loosening improved nine-dot solving; treat with caution. Children and sophomores sometimes blurt useful remote associations adults filter out; "think before you talk" can be over-learned into silence.
Horseradish, Saunders, nine dots, filters (expanded)
The Yiddish horseradish saying is about habitat of mind: stay long enough in one pungent world and every other world smells like home. Clarence Saunders, first a holiday shop-boy at nine, later a wholesale seller, was sent in late summer 1916 to spy on an innovative grocery in Terre Haute. He then built Piggly Wiggly: customers walked the aisles and selected goods themselves instead of handing a clerk a list. Self-service reframed labour, layout, pricing and trust.
The nine-dot problem punishes the assumption that lines must stay inside the square. Australian researchers in 2012 slipped it at the end of an unrelated study: one of thirty solved it; that person had childhood traumatic brain injury, a clue that weakened filtering might help (not a prescription for injury). tDCS claims of temporary filter loosening should be read as provisional. Children blurt raisin-face metaphors adults filter; over-learned "think before you talk" can silence useful remote associations. The skill is a dial: low for generation, high for shipping.
Nine dots, Monopoly cheats and the filter dial (pass 3)
Most people cannot connect nine dots with four continuous lines without lifting the pencil; many experiments find success rates under one in ten, sometimes zero. Even after seeing the solution, more than a third cannot reproduce it a week later. The difficulty is the assumed boundary: lines must stay inside the square. Habit and expectation decide what is even visible. Accustomed to honest Monopoly players, you miss the cheat; accustomed to clerks fetching goods, you miss self-service as a design.
Children blurt raisin-face metaphors; adults filter them. Over-learned "think before you talk" can silence remote associations that CRA-style insight needs. Treat filtering as a dial: low during generation, high during critique and shipping. Provisional tDCS claims of temporary loosening remain provisional; sleep, walk and off-peak timing are safer first tools.
Piggly Wiggly, horseradish habitat and Monopoly vision (pass 4)
Clarence Saunders spied an innovative Terre Haute grocery in late summer 1916, then built Piggly Wiggly: customers walked aisles and chose goods instead of handing a clerk a list. Self-service reframed labour, layout, pricing and trust. The Yiddish horseradish saying is about habitat of mind: stay long enough in one pungent world and every other world smells like home. Accustomed to clerks fetching, you miss self-service; accustomed to honest Monopoly players, you miss the cheat. Nine-dot failure is the same mechanism on paper: the assumed square boundary hides the four-line path. Filters are not villains; over-learned filters that never dial down are.
tDCS, children and the shipping gate (pass 5)
Provisional brain-stimulation claims of temporary filter loosening should stay provisional. Safer first tools: sleep, walk, off-peak timing, and an explicit generation phase where "think before you talk" is paused. Children blurt raisin-face metaphors adults suppress; keep a private generation notebook so odd associations survive until the shipping gate, where filters go high again.
Nine dots, horseradish and filters
Nine-dot problem: connect without lifting the pen; solution exits the implied square. Horseradish and supermarket tales show category blindness. Idea filters protect against noise and also block sophomoric seeds that become breakthroughs. Piggly Wiggly-style retail innovations look obvious after the filter moves. Sophomores (literal and figurative) sometimes see past expert filters.
FILTER≠TRUTH. A tight filter raises precision and kills recall of distant associates; dial it to the task.
Nine dots, horseradish and idea filters
The nine-dot problem: connect without lifting the pen; the solution exits the implied square. Horseradish and supermarket tales show category blindness: you fail to see a use because the label hid it. Idea filters protect against noise and also block sophomoric seeds that become breakthroughs. Piggly Wiggly-style retail moves look obvious after the filter moves. Sophomores (literal and figurative) sometimes see past expert filters precisely because they have not yet installed them.
Filters are not villains. Without them, brainstorms are sludge. The skill is dialling: tight for execution, looser for exploration. Teams that never loosen never invent; teams that never tighten never ship.
| Mode | Filter | Risk |
|---|---|---|
| Explore | loose | noise, false starts |
| Exploit | tight | missed reframes |
| Handoff | explicit switch | staying in the wrong mode |
Run a ten-minute “bad ideas first” round before critique. Write three ideas you would normally filter out. Only then score them. Notice which filter rules fired earliest.
Idea filters are precision-recall in cognition. Tight filters raise precision (fewer stupid ideas) and kill recall of distant associates. Loose filters raise recall and drown you. Nine-dot solvers who stay inside the implied square have a filter that feels like geometry but is only habit. Horseradish and supermarket mis-categorisations show labels hiding uses. Retail self-service looked absurd until Saunders’s filter moved; then it looked inevitable.
Sophomores help when experts overfit. Experts help when sophomores cannot finish. Teams need an explicit handoff between explore and exploit modes instead of a permanent average that does neither well.
Sticky notes without a later tighten step are filter-free theatre. Elastic practice includes the ugly edit that turns remote associates into a testable plan.
Filter journal. For one week, capture three ideas you almost did not write down. Tag why the filter fired (status, feasibility, aesthetics, fear). At week’s end, revive one. Nine-dot and Piggly Wiggly rhyme here: the square and the counter were imaginary walls.
- The tDCS nine-dot study (Chi and Snyder 2012) targeted the anterior temporal lobes (inhibiting the left, exciting the right), not the lateral prefrontal cortex. It had 33 subjects and has not been robustly replicated; meta-analyses find tDCS effects on cognition small and inconsistent.
- The "40 to 60 bits per second" figure is a rough estimate; a 2024 Neuron paper by Zheng and Meister estimated conscious behaviour runs at about 10 bits per second.
- Piggly Wiggly: the self-service idea had precursors, and the pig-farm epiphany is Saunders's own story, possibly embellished.
- Kearns earned his fortune largely through decades of patent lawsuits against Ford and Chrysler (about $30 million), dramatised in the film Flash of Genius.
- The nine-dot "outside the box" framing is too simple: studies (Weisberg and Alba 1981) found that telling people to go outside the square helped only modestly, so other constraints also block solvers.
- Nine-dot and tDCS creativity claims often fail replication; treat single-lab neuromodulation effects as provisional.
- Latent inhibition (weak filtering) still correlates modestly with creative achievement in some samples, especially when paired with high intelligence.
- Self-service retail history is solid; Saunders's Piggly Wiggly patents and business drama remain a clean reframing case study.
H-P-N-S. Horseradish (everything looks like home), Piggly Wiggly (reframe the store), Nine dots (leave the box), Sophomoric blurts (filters can be too tight).
Invert: suppose filters should always stay high, and "thinking outside the box" is mostly error. Then Saunders is a fluke, children's metaphors are noise, and brain-injury clues are dangerous to generalise. The chapter needs a context-sensitive dial, not a cult of looseness.
Attempt the nine-dot problem cold, then with explicit permission to draw outside the square. Separately, run a two-pass idea meeting: pass one forbids criticism (filters low); pass two ranks and kills ideas (filters high). Never mix the passes. Note one "horseradish" assumption in your field (something everyone treats as obviously sweet).
Cognitive filters decide what reaches awareness, guided by expectation. They protect us from nonsense but also block good ideas, as in the nine-dot problem. Loosening filters (the beginner's mind, quantity before judgement, changing the frame) helps generate originality; judgement can follow.
10. The Good, the Mad, and the Odd
It's a mad, mad world
In 1951 entomologist Jay Traver, an expert on mayflies at the University of Massachusetts, published a detailed paper describing house dust mites infesting her own scalp, resisting DDT and every treatment. Dust mites do not colonise humans. Scientists later concluded she suffered from delusional parasitosis; one called her an "authentic mad scientist."
Fields that favour elastic thinking seem to contain more than their share of eccentrics:
- William Blake believed spirits dictated his works.
- Howard Hughes sat naked for hours in a "germ-free" hotel room.
- Buckminster Fuller logged his life every fifteen minutes from 1920 to 1983.
- David Bowie lived on milk and peppers in his most productive 1970s years.
- Nikola Tesla credited alternating current to a vision and, late in life, declared his love for a white pigeon: "I loved her as a man loves a woman, and she loved me."
In the 1960s behavioural geneticist Leonard Heston studied children of schizophrenic mothers adopted at birth. Half of the healthy ones were unusually artistic and eccentric, with "imaginative adaptations to life which were uncommon in the control group." Perhaps a small inherited "dose" of schizophrenia-like traits helps creativity.
Measuring doses of madness
Yes/no items include "People sometimes find me aloof and distant," "Have you ever noticed a common event or object that seemed to be a special sign for you?", "I sometimes use words in unusual ways," "I am an odd, unusual person." In about 1,700 people the average was about 6 yes answers. Two or fewer puts you in the bottom quarter; 13 or more in roughly the top 10%. High scorers tended to be both eccentric and strong at divergent thinking. (This is the SPQ-B of Raine and Benishay, 1995.)
Schizotypy describes a spectrum of schizophrenia-like traits in healthy people. The proposed mechanism is the cognitive filter of Chapter 9: a lax filter brings original thinking and nonconformity; a stringent one, cognitive inhibition, brings conventional thought.
John Nash, whose game theory won a Nobel Prize, had low enough filters to generate remarkable ideas, and then schizophrenia. After recovering, a friend asked how he could have believed aliens had recruited him to save the world. "Because the ideas I had about supernatural beings came to me the same way that my mathematical ideas did. So I took them seriously." Imaging studies link belief in telepathy and lucky charms with low activity in the lateral prefrontal cortex. Great original thinkers often have inhibition high enough to function but low enough to entertain ideas others reject.
Elastic personalities, from the arts to science
Creative fields need different blends. Jazz improvisers must lower inhibitions on stage; composers of symphonies must plan and edit (even Mozart reworked ideas laboriously, contrary to myth). Psychologist and former musician Geoffrey Wills studied 40 pioneers of jazz's golden era (1945 to 1960) and found them reckless far beyond other creative fields; Chet Baker was a heroin addict.
Science sits at the other end: ideas must agree with experiment. A recipe for turning mercury into gold works or does not (physicists actually did it in 1941 by neutron bombardment). Scientists are eccentric in milder ways: one ate only condiments at lunch; Einstein sniffed cigarette butts after his doctor banned his pipe; Newton searched the Bible for coded prophecy. Every field needs some ordered analysis to turn novel ideas into a useful product.
The Dr Jekyll and Mr Hyde inside
Judith Sussman, an imaginative girl of the 1940s, became a 1950s suburban wife and mother. She began writing between laundry and tuna casserole; her husband found it subversive and friends were unsupportive. She published her first book in 1969 as Judy Blume (she divorced in 1976) and became a hugely bestselling author. "Once I began to write, I suddenly was anxious to get up in the morning. Writing saved me." Her normal life is orderly; when she writes, "it is like I'm another person."
William James and Freud both argued that analytical and imaginative modes coexist in everyone. In an fMRI study, people judged logically identical syllogisms ("All A are B..." versus "All dogs are pets...") using different brain regions: abstract ones engaged parietal reasoning areas, meaningful ones engaged language and semantic areas in the left temporal lobe. Within each of us are a logician and a poet.
Schizotypy, Tesla, arts to science, Jekyll/Hyde
Schizotypy names a spectrum of unusual ideas, magical beliefs and perceptual quirks short of schizophrenia; questionnaires score the "dose." Creative groups often score higher on mild schizotypy than controls, especially in the arts; science tends to need stronger analytical brakes. Nikola Tesla is the book's emblem of brilliant oddity. Judith Sussman (later a working imaginative adult after an analytical marriage) illustrates the inner Dr Jekyll (control) and Mr Hyde (loose association) that must trade the wheel. Elastic personalities are not "mad genius" cartoons; they are tunable mixes of loose generation and tight selection.
Measuring the dose and the creative mix
Schizotypy questionnaires typically probe unusual perceptual experiences, odd beliefs, social withdrawal and nonconformity. Mild elevation often co-occurs with comfort generating remote associations; high elevation tips toward distress and disorganisation. Jazz improvisers need fast filter-lowering; proof-oriented mathematicians need strong reconvergence. The practical lesson is not to chase eccentricity but to schedule generation and selection as separate modes (the poet/logician switch of the whole book).
| Mode | Filter | Job |
|---|---|---|
| Poet (Hyde) | Low | Generate remote, odd, plentiful ideas |
| Logician (Jekyll) | High | Test, kill, shape, ship |
| Failure mode | Stuck low | Noise, offense, unfinished work |
| Failure mode | Stuck high | Competent sameness, missed Piggly Wigglys |
Nash, jazz, Mozart and the filter lifespan (pass 4)
Imaging ties high schizotypy to quiet cognitive filtering (lateral prefrontal and related circuits). Lax filters admit remote associations; stringent filters produce cognitive inhibition and conventional action. High scorers often adapt well to novelty; at the extreme, coherence fails. John Nash generated Nobel-winning game-theory ideas while filters were low enough for originality, then lost years to full schizophrenia when wild ideas drowned usable ones. Belief in telepathy, rituals and luck charms correlates with unusually quiet filtering circuitry. Supernatural belief tends to fall as children's lateral prefrontal cortex matures and to rise again as that vigor declines in old age.
Eccentric exemplars (Blake's spirits, Hughes's germ-free hotel room, Fuller's prune-Jell-O years and fifteen-minute diary from 1920 to 1983, Bowie's chili-pepper diet) are illustrations, not goals. Jazz improvisers need low inhibition in performance; symphony composers and working scientists need more executive shaping. Mozart's letters describe immersion, experiment, study and reflection, not lightning whole-works. Scientists Mlodinow has known (condiment-only cafeteria lunches, orange hair and Apple tattoos, snowflake or banjo obsessions) stay eccentric inside a moderated band. Higher IQ appears to help hold a barrage of odd thoughts without social collapse; that is one proposed difference between schizotypic productivity and schizophrenia. The practical moral is a dial: lower filters to generate, raise them to edit and ship.
Dose, not romance (pass 5)
Treat schizotypy scores as a dial reading, not a destiny. Generation needs some laxity; shipping needs inhibition. Fields differ: jazz performance tolerates more oddity than proof-writing. Organisations that fetishise eccentricity without selection waste time; organisations that crush mild oddity starve insight. Match the dose to the task phase.
Schizotypy, dose and Jekyll/Hyde
Creativity research tracks schizotypy (odd associations without full schizophrenia) more than romantic “madness.” Dose matters: too little oddness, conventional output; too much, unusable noise. Artists and scientists differ in average profiles yet share a need for controlled leakage through filters. Tesla myths and jazz improvisation illustrate productive oddness under craft constraints. Nash’s aliens rhyme with Mr. X’s marijuana associations: the filter failure that feels insightful when lucky and delusional when not.
Work on cognitive flexibility versus persistence (Koutstaal 2025) and aging control transitions (Grahek et al.) reframes “elastic personality” as a trainable trade-off, not a fixed trait label.
Oddness with craft: schizotypy and dose
Creativity research tracks schizotypy (odd associations without full schizophrenia) more than romantic madness. Dose matters: too little oddness, conventional output; too much, unusable noise. Artists and scientists differ in average profiles yet share a need for controlled leakage through filters. Tesla myths and jazz improvisation illustrate productive oddness under craft constraints. Nash’s aliens rhyme with Mr. X’s marijuana associations: filter failure that feels insightful when lucky and delusional when not.
Jekyll/Hyde is the chapter’s metaphor for the same brain under different chemical and social controls. The practical lesson is not “seek madness.” It is learn your personal dose of unusual association, then bind it with craft, sleep, critique and feedback.
Koutstaal 2025 and related aging/control work reframe elastic personality as a trainable trade-off between flexibility and persistence, not a fixed type label. Train both; schedule which one you need.
ODD + CRAFT. Leak associations; bind them with skill. Leak without craft is noise; craft without leak is cliché.
Schizotypy scales (book: ~22 yes/no items; average near 6; top tenth from about 13) quantify mild oddness without romanticising illness. Jazz improvisers need on-stage disinhibition bound by years of craft. Composers and scientists need more planning and contact with external checks (score, experiment). The “mad genius” story collapses once you separate association looseness from verification. Nash’s tragedy and Wilson’s arc warn that chemistry and illness can loosen filters past the point of use.
| Profile | Association | Verification |
|---|---|---|
| Conventional expert | tight | strong |
| Productive odd | loose enough | strong craft |
| Unusable noise | very loose | weak |
| Aim for the middle row on purpose; do not confuse the bottom row with genius. |
Oddness budget. Give yourself a weekly quota of deliberately strange analogies in a safe draft doc. Cap them. Edit ruthlessly before they meet an audience. Controlled leakage plus craft beats both bland expertise and unfiltered rant.
- Mad genius: large Swedish registry studies (Kyaga and colleagues 2011, 2013) found creative professions were over-represented among relatives of people with schizophrenia and bipolar disorder, and bipolar disorder among writers, but most creative people have no disorder. Effects are modest.
- Schizotypy and creativity: meta-analyses (Acar and Sen 2013) find a small positive link, mostly for positive schizotypy (unusual experiences); negative schizotypy (social withdrawal) is unrelated or harmful. The quiz above mixes both.
- Heston's 1966 observation was qualitative and based on a small sample.
- Tesla's pigeon and Hughes's illness (obsessive-compulsive disorder, drug use) remind us that eccentricity often means suffering, not just colour.
- Jay Traver: a 2011 analysis argued her paper should never have been published, illustrating weak peer review in the 1950s.
- The syllogism imaging study (Goel and colleagues 2000) was small and is a classic example of belief bias, consistent with dual-process theory.
- Large registry studies (especially Nordic) find some links between creative careers and psychiatric diagnoses in families, but effect sizes are small and bidirectional selection confounds remain.
- The "mad genius" stereotype harms patients and overpredicts creativity from illness; most people with schizophrenia are not unusually creative, and most creators are not ill.
- Schizotypy scales still predict divergent-thinking scores better than they predict lifetime creative achievement.
D-O-S-E. Creativity tracks a Dose of schizotypy, not full illness; Oddity helps generation; Selection (analytical Jekyll) shapes it; Each field needs a different mix (jazz ≠ proof).
Invert: suppose eccentricity and creativity are unrelated, and every apparent link is romantic selection bias. Then Tesla stories are entertainment, schizotypy questionnaires are distractions, and organisations should ignore personality oddity. Hold that view against the dose-response data the chapter cites, and against Chapter 11's filter-loosening results.
On a 1 to 5 scale, rate your usual generation looseness and selection strictness. For your next creative task, deliberately mismatch: if you over-select, impose a ten-minute wild-idea rule; if you under-select, impose a kill-criteria checklist before sharing. Note which mismatch improved the output.
Creativity and eccentricity share a root in lax cognitive filters, measured as schizotypy. Too lax brings disorder; too strict brings rigidity. Fields differ in the blend they need, from jazz to physics. Within everyone, a poet and a logician compete, and we can learn to switch between them.
11. Liberation
Let's go get stoned
An anonymous young scientist described his marijuana experiences in an essay: bizarre associations that felt reasonable, deeper appreciation of music and art, "a feeling of communion with my surroundings," and a flood of free association. He ended: "The illegality of cannabis is outrageous, an impediment to full utilization of a drug which helps produce the serenity and insight, sensitivity and fellowship so desperately needed in this increasingly mad and dangerous world." The author was Carl Sagan.
Early survey studies of users report similar effects: "The ideas that come to my mind are much more original"; "I am more willing to accept contradictions between two ideas."
- Cannabis (2012): 160 users were tested sober and after smoking. High scorers on divergent thinking when sober were unaffected; low scorers improved, rising toward the level of the creative group. THC suppresses prefrontal filtering.
- Alcohol (2012): 40 social drinkers recruited via Craigslist. Half drank vodka-cranberry to just below the legal limit. Tipsy subjects solved about 60% of remote-associates problems versus 40% for sober ones, and faster, more often by insight.
- Psychedelics: imaging found LSD and psilocybin disrupt the default network, matching reports of "ego dissolution."
- Ayahuasca: effects start 45 to 60 minutes after drinking; users report visions and rapid idea generation, and score higher on divergent thinking while their convergent thinking worsens.
The catch: loosening can derail thought. Brian Wilson of the Beach Boys credited marijuana with his move from simple rock to orchestral pop. In 1982 he was diagnosed with schizoaffective disorder, possibly triggered by heavy use. Those already high in schizotypy risk psychosis. Isabel Allende found ayahuasca terrifying, facing demons. Silicon Valley microdosing of LSD is the "performance psychedelic" counterpart to Adderall.
The silver lining of fatigue
In 2015 French researchers fatigued participants' executive brains with a tedious arrow task (focusing on a central arrow while suppressing flanking ones). Afterwards, the depleted subjects generated more and more original uses for household objects.
In 2011 at Michigan State, 223 students classified as morning or evening types solved problems at 8:30 a.m. or 4:00 p.m. At their peak time they solved more analytic problems; at their off-peak time they solved more insight riddles, thanks to "more diffuse attentional focus." Mlodinow, a night person, does science at night but writes best in his morning fog.
Don't worry, be happy
In 1930 young nuns wrote short autobiographical essays. Sixty years later, Kentucky researchers studying ageing found 180 of the writers among their subjects. Nuns whose essays expressed the most positive emotion lived about ten years longer than the least positive. The nun study helped spark positive psychology.
Negative emotions come with urges to act (flee, fight), narrowing what filters let through; sad film clips reduce novel word associations. Barbara Fredrickson's broaden-and-build theory says positive emotions widen thought and action, building resources. Volunteers who watched a funny video or ate tasty snacks did better on elastic-thinking tests, and solving problems creatively boosts mood in turn: a virtuous circle.
- Do pleasurable things: a novel, a hot bath; savour and share good news.
- Gratitude exercise: regularly write three things you are grateful for.
- Prosocial spending: money spent on others cheers us more than money spent on ourselves.
- For negative thoughts: accept rather than suppress them, then ask what advice you would give a friend having the same thought.
Happiness is not only an end in itself; it is a strategy for mental productivity.
Where there's a will
Mlodinow's mother, a Holocaust survivor in her late eighties, ordered a cheap blender online. It arrived broken, and the return policy was opaque. Her motto: "Where there's a will, there's a way." Best Buy would not exchange it without a receipt, so she bought an identical one, then returned the broken one with the new receipt, asking for a refund instead of an exchange, netting about $3.17 minus four bus fares. Neophilia, reward, the default network and her executive brain all worked together on a small everyday problem.
She is now 95, and a fog has rolled in as neural connections weaken with age. The lesson: use and nurture elastic thinking while you can.
Survival of the elastic
During the Second World War Mlodinow's father, himself a slave labourer, supervised child workers in a German munitions factory, lined up in groups of thirty each morning for the Gestapo count. One day a boy of about nine, believed killed with his parents, appeared in line, making thirty-one. "You have one extra," said the officer. Before anyone could answer, the boy stepped forward and fluently invented a story about being in the infirmary. The officer noted it and moved on. Decades later his father wept telling it: "He created a tale, like Isaac Bashevis Singer or Malamud."
Our era, like war, is one of disruption. Analytical skill matters, but elastic thinking, generating new ideas and adapting, is what change demands.
Sagan, wine, fatigue, nuns, blender, Gestapo count
Carl Sagan's anonymous cannabis essay praised original ideas, tolerance of contradiction and "fellowship." Survey users echo him; lab work (2012) found cannabis lifted divergent thinking mainly for low scorers, and alcohol just below the legal limit raised remote-associates success (~60% versus ~40%). Psychedelics disrupt the default network (ego-dissolution reports); ayahuasca can raise divergent scores while harming convergent ones. Costs are real: Brian Wilson's trajectory, psychosis risk at high schizotypy, Allende's terrifying ayahuasca night.
Safer levers: French 2015 executive fatigue raised originality on alternate uses; Michigan State 2011 (n=223) found morning/evening types solved more insight problems at off-peak hours. The nun study (essays from 1930, follow-up decades later) linked positive emotion language to ~ten more years of life and helped launch positive psychology. Fredrickson's broaden-and-build: positive mood widens thought-action repertoires. Toolkit: savour, gratitude lists, prosocial spending, accept-then-advise negative thoughts.
Mlodinow's mother ("Where there's a will, there's a way") turned a broken blender and opaque returns into a Best Buy manoeuvre that netted a few dollars after bus fares: neophilia, reward, default mode and executive control on a small stage. His father, supervising child slave labourers in a German munitions factory while in the underground, watched a boy of about nine invent a fluent infirmary story when the Gestapo count showed thirty-one instead of thirty; decades later he wept: the child "created a tale, like Isaac Bashevis Singer or Malamud." Elastic thinking is survival gear under disruption.
Wilson, vodka, ayahuasca, nuns and will (expanded)
Brian Wilson began recreational marijuana in 1964 and soon used it for creative work, crediting it with the move from simple rock to orchestral pop and "playing the studio." Voices had begun in 1963; after heavy use they worsened. He obsessed over tile counts and peas on a plate; by 1966 he would interview only in his swimming pool, sure the house was bugged. 1982 brought a schizoaffective diagnosis. The chapter treats him as caution, not as proof that cannabis causes genius or madness alone.
Alcohol testimonials run from Aristophanes (424 BCE) to modern musicians. The 2012 vodka-cranberry CRA study (n=40, Craigslist twenties) put half just below the legal limit: tipsy solvers about 60%, sober about 40%, and faster. Too loose, and ideas fall off the tracks. Psychedelic research nearly stopped after the 1971 UN psychotropic treaty; recent imaging (Imperial College London; the book sometimes says Oxford) links LSD and psilocybin to default-network disruption and ego-dissolution reports. Ayahuasca effects begin 45 to 60 minutes after the tea; fluency and divergent scores rise, convergent thinking falls, as top-down control weakens across cortical hierarchies. Isabel Allende broke writer's block and faced demons, a terrified four-year-old self, retching for days: "I don't ever want to do it again." Silicon Valley microdosing is the street twin of campus Adderall; placebo-controlled trials through 2023 undercut the creativity hype.
Simon task fatigue (2015 France): forty minutes suppressing flanking arrows exhausts prefrontal control; afterwards alternate-uses scores (number and originality) rise versus controls. Schedule imaginative work after tedious focus, not only when "fresh." Michigan State 2011: morning types and evening types, tested at 8:30 a.m. or 4:00 p.m., solved more analytic items at peak and more insight riddles off-peak. Mlodinow writes science at night and prose in morning fog for that reason.
Nun study: 22 September 1930 letter asked young nuns for ~300-word life essays; sixty years later Kentucky researchers found 180 of the writers in their ageing sample. Most-positive essayists lived about ten years longer than least-positive. Negative emotions come with action urges that narrow filters; sad clips hurt novel associations. Fredrickson's broaden-and-build: positive emotion widens thought and action, builds resources; funny videos and tasty snacks lift elastic-test scores; creative success lifts mood in turn. Toolkit: savour, share good news, gratitude lists, kindness tallies, accept-then-reframe negative thoughts as advice to a friend.
The blender story and the Gestapo-count story close the book: elastic will on a small stage and on a lethal one. Analytical skill still matters; survival under disruption needs the poet as well as the logician.
Pills, peaks, fatigue and the positive-emotion loop (pass 3)
Microdosing and campus stimulants sit on opposite sides of the same cultural wish: buy the right brain state. Focus drugs (Adderall, Vyvanse) tighten analytical grip; psychedelic microdoses were sold as creativity helpers. Placebo-controlled work through 2023 largely undercuts the microdosing creativity claim. Imperial College imaging still matters for full-dose mechanisms (default-network disruption, ego-dissolution reports), not as a home recipe. FDA rejected MDMA-assisted therapy for PTSD in 2024 after advisory concerns; the regulatory path is slower than the hype cycle. Brian Wilson died in June 2025; his story remains a caution about mixing creative myth with untreated illness.
Off-peak insight (Wieth and Zacks 2011) and French Simon-task fatigue (2015) give drug-free levers: do imaginative work when executive control is slightly loose, and schedule generation after exhausting suppression tasks. Positive emotion (nun essays; Fredrickson) widens filters; sad clips narrow them. The toolkit is ordinary: savour, share good news, gratitude lists, kindness tallies, advise yourself as you would a friend. The blender repair and the Gestapo-count survival stories end the book because elastic will is not only for artists; it is for anyone facing a script that must break.
Blender will and the count of thirty-one (pass 4)
Mlodinow's mother, late eighties, refused a Best Buy trip as "too much trouble," then ordered the cheapest blender online after her first computer lesson. The unit arrived defective; return postage looked worse than the win. She bought an identical replacement in-store (all day on arthritic legs), then used the working receipt path to walk the broken twin back for exchange: two identical blenders as a workaround. "Where there's a will, there's a way" was her lifelong line, forged in surviving a Nazi slave-labor camp and arriving at Ellis Island penniless. At ninety-five, fog thickened; neural connections that once supported neat collaboration among neophilia, reward, default network, executive focus and filters weakened. The book ends by urging use of those abilities while they last.
The father's roll call is the lethal twin of the blender story. Supervising groups of thirty children at a factory farm, he one morning faced thirty-one. Before he could invent a cover, the Gestapo counted: "You have one extra." Seconds of blank mind; an extra child, or he, or all of them, could be shot. The fugitive boy stepped forward and spun an infirmary yarn until the officer grew impatient, noted thirty-one, and left. Elastic will on a kitchen counter and under a clipboard: both need neophilia, reward drive, associative generation and enough filter to keep the plan coherent.
Carl Sagan, Mr. X and the balanced cocktail (pass 5)
The anonymous "Mr. X" cannabis essay (later linked to Carl Sagan) argued creative benefit while protecting an academic career. The chapter pairs that hope with Wilson's caution and with alcohol's inverted-U on CRA performance. A future "balanced cocktail" of focus drugs plus creativity aids would raise fairness and safety fights; Mlodinow leaves the ethics open. Until such pills are real and safe, the free stack remains: off-peak timing, post-fatigue generation, positive-emotion broadening, protected mind-wander and the will illustrated by a blender and a count of thirty-one.
Mr. X, Wilson, nuns and will
Anonymous Mr. X (Carl Sagan, later revealed) described marijuana associations that would have been censored sober: a Volkswagen in plant shadows. Analytical brain still labeled illusion; he liked the loosened filter. Brian Wilson used cannabis for orchestral pop leaps; voices and obsessions worsened; 1982 schizoaffective diagnosis. (Wilson died June 2025; the cautionary arc stands.)
2012 vodka-cranberry CRA study (n=40): tipsy ~60% vs sober ~40%. Psychedelic imaging (Imperial) links LSD/psilocybin to default-network disruption; ayahuasca raises divergent scores, lowers convergent. FDA declined MDMA therapy (Aug 2024); microdosing hype still outruns placebo-controlled trials through 2023-25.
Simon task fatigue (2015 France): forty minutes of suppression raises later alternate-uses originality. Morning/evening chronotypes: analytic peak on-peak, insight off-peak (Michigan State 2011). Nun study (1930 essays): most-positive lived ~ten years longer. Fredrickson broaden-and-build: positive emotion widens thought. Toolkit: savour, share good news, gratitude, kindness, reframe as advice to a friend. Blender story and Gestapo-count story: elastic will on small and lethal stages.
MDMA assisted-therapy setback (FDA Aug 2024) and mixed microdosing evidence keep the book’s caution intact: chemistry can loosen filters; it is not a creativity subscription.
Liberation toolkit: chemistry, will and widen
Anonymous Mr. X (Carl Sagan, later revealed) described marijuana associations that would have been censored sober: a Volkswagen in plant shadows. Analytical labeling still called illusion; he liked the loosened filter. Brian Wilson used cannabis for orchestral pop leaps; voices and obsessions worsened; 1982 schizoaffective diagnosis. (Wilson died June 2025; the cautionary arc stands.)
2012 vodka-cranberry CRA study (n=40): tipsy ~60% vs sober ~40%. Psychedelic imaging (Imperial) links LSD/psilocybin to default-network disruption; ayahuasca can raise divergent scores and lower convergent ones. FDA declined MDMA therapy (Aug 2024); microdosing hype still outruns strong placebo-controlled trials through 2023-25. Chemistry can loosen filters; it is not a creativity subscription.
Simon task fatigue (2015 France): forty minutes of suppression raises later alternate-uses originality. Morning/evening chronotypes: analytic peak on-peak, insight off-peak (Michigan State 2011). Nun study (1930 essays): most-positive lived ~ten years longer. Fredrickson broaden-and-build: positive emotion widens thought. Toolkit: savour, share good news, gratitude, kindness, reframe as advice to a friend. Blender story and Gestapo-count story: elastic will on small and lethal stages.
| Lever | Elastic effect | Caution |
|---|---|---|
| Protected idle | default-mode recombination | not endless scroll |
| Off-peak insight windows | broader associates | still verify analytically |
| Positive broaden | wider thought-action repertoire | not toxic positivity |
| Mild disinhibition | temporary filter loosen | addiction, psychosis risk |
| Will on small acts | trains reframe muscle | not a substitute for sleep |
This week: (1) one phone-free idle block of twenty minutes after a hard problem; (2) one gratitude or kindness act before a creative session; (3) one off-peak hour for remote-associate style work, then an on-peak hour to verify.
Liberation without romance. Substances that loosen filters have real risks and weak average evidence for reliable creativity gains. Prefer sleep, idle, off-peak windows, kindness and small will experiments. Use analysis to cash the cheques elastic thinking writes.
- Ego depletion: the idea that tiring the executive boosts creativity rests on a concept that failed a large multi-lab replication (Hagger and colleagues 2016) and later tests. The 2015 French study (Radel and colleagues) is suggestive, not settled.
- Time-of-day insight (Wieth and Zacks 2011) is a single study with modest effects; replications are limited.
- The alcohol study (Jarosz and colleagues 2012) used a blood alcohol level of about 0.075; benefits vanish and harms dominate at higher doses.
- The cannabis study (Schafer and colleagues 2012) did not randomise; THC strength has risen sharply since, and high-potency cannabis is linked to higher psychosis risk (Di Forti and colleagues 2019).
- The psychedelic imaging studies were led by Robin Carhart-Harris at Imperial College London, not Oxford.
- Nun study (Danner and colleagues 2001): a powerful correlation, but positive writing could reflect health, education or verbal style.
- Broaden-and-build is well supported in broad terms; Fredrickson's specific "positivity ratio" of 2.9 was retracted after mathematical criticism (2013).
- Microdosing: placebo-controlled trials (Szigeti 2021; Murphy 2023) find microdosing LSD or psilocybin ≈ placebo for mood and creativity.
- Psychedelic therapy: US FDA declined MDMA-assisted therapy in August 2024; psilocybin depression trials continue; Oregon, Colorado and Australia allow forms of supervised use.
- Cannabis: wider legalisation; Hjorthøj and colleagues (2023) estimated a substantial fraction of schizophrenia cases in young men might be preventable by avoiding cannabis use disorder.
- Brian Wilson died in June 2025 at 82.
- Generative AI as a filter-loosener: useful for wild associations, best when humans generate first and use models to extend; diversity collapse remains the main group-level risk (see Intro update).
- Gratitude interventions: meta-analyses show small, real well-being benefits.
L-O-W F-I-L-T-E-R. Liberation = Loosen filters via Off-peak / Well-being first; Fatigue can help; Intoxicants cost; Life stories (blender, Gestapo count) show stakes; Trade poet and logician; Eschew ego-depletion overclaims; Reward positive mood loops.
Invert: suppose nothing liberates elastic thinking except talent and hours, and mood, timing and substances are footnotes. Then the nun study, off-peak insight and Sagan's essay are colour. The chapter's practical claim is that filter tone is controllable. Test it with one green lever for a week before dismissing it.
Map your rhythm for a week: peak versus foggy hours. Put analysis at peak and brainstorming or reframing at off-peak. Before each creative session add one mood lever (walk, music, funny clip). Nightly: three-line gratitude note. Do not add substances as an experiment; if you already use alcohol or cannabis, note next-day originality honestly rather than romantically.
Elastic thinking can be liberated by anything that loosens prefrontal filters: substances (with real costs), fatigue and off-peak hours, and, best of all, positive mood. From a broken blender to a Gestapo count, the will to find a way, plus a flexible mind, is how humans survive change.
Synthesis
25 takeaways
- The pace of change makes elastic thinking as important as analysis.
- There are three modes of processing: scripted, analytical and elastic.
- Humans are not inherently change-averse; neophilia varies like a personality trait.
- Exploring is risky but pays off when the environment changes.
- Mindfulness counters scripted, automatic behaviour.
- Computers excel at analysis and pattern learning, but humans still lead at flexible reframing across domains.
- The reward system switches thinking on; too little paralyses (EVR), too much compels (Pat Darcy).
- Reward flows along the way to a goal, so design work with visible progress.
- Satisfice on small decisions to save energy for big ones.
- Brains store concepts in hierarchies; shared neurons create associations.
- Intelligence emerges from both top-down executive control and bottom-up association, like an ant colony.
- Framing decides which solutions you can see (jumbo popcorn, the checkerboard).
- Restructuring a problem often beats more analysis.
- Culture shapes perception and frameworks; exposure to other cultures loosens them.
- The default network recombines ideas when we are idle.
- Constant connectivity crowds out the idle time creativity needs.
- Insight follows a hidden sequence: search, impasse, broadening, breakthrough.
- A relaxed, positive, uninterrupted mind is primed for insight.
- Expertise can freeze thinking (Einstellung, doctrine, "the Concept").
- Dissent thaws frozen thinking even when no one is persuaded.
- Cognitive filters protect us from nonsense but also block good ideas.
- Generate first, judge later: quantity before quality.
- Creativity and eccentricity share lax filters; too lax brings disorder.
- Positive mood, off-peak hours and fatigue loosen filters safely; substances do so at a cost.
- Within each of us are a poet and a logician; learn to switch.
Frameworks to reuse
| Framework | Use it when | Ch |
|---|---|---|
| Scripted, analytical, elastic | Diagnosing which mode a task needs | 2 |
| Explore versus exploit | Balancing new options against proven ones | 3 |
| Satisficer versus maximiser | Facing a decision with many options | 3 |
| Top-down versus bottom-up | Designing teams and organisations | 4 |
| Problem space restructuring | Stuck on a problem despite effort | 5 |
| Default mode breaks | Scheduling incubation time | 6 |
| Insight sequence | Deciding when to stop grinding and step away | 7 |
| Red team / beginner's mind | Testing an expert consensus | 8 |
| Generate then filter | Running brainstorms | 9 |
| Poet and logician scheduling | Splitting creative and editing work | 10, 11 |
The book chapter by chapter
| Ch | One-line summary |
|---|---|
| Intro | Rapid change makes elastic thinking the key skill of our era |
| 1 | Humans love novelty more than we think; neophilia varies and can be cultivated |
| 2 | Thought ranges from scripts to analysis to elastic, associative processing |
| 3 | Reward makes us think; too little or too much distorts it |
| 4 | Concepts built bottom-up, steered top-down, are the substance of thought |
| 5 | Framing and restructuring unlock problems analysis cannot solve |
| 6 | The idle brain's default network is where associations form |
| 7 | Insight is the end of a hidden process that relaxation helps |
| 8 | Habit and expertise freeze thought; dissent and novices thaw it |
| 9 | Cognitive filters censor ideas; loosening them aids originality |
| 10 | Eccentricity and creativity share lax filters within limits |
| 11 | Mood, timing, fatigue and substances liberate thought; the will finds a way |
Questions for discussion
- Where in your work do you rely on scripts that may no longer fit the situation?
- Mlodinow argues society shifted from resisting novelty to demanding it. Is that true in your industry?
- EVR could reason but not decide. What does that say about purely logical decision-making, in people or machines?
- Which "Concept" in your organisation would be most dangerous if it were wrong?
- How could you build dissent into decisions without slowing everything down?
- Given that ego depletion and several priming effects failed to replicate, which of the book's practical tips do you still trust, and why?
- Does generative AI make human elastic thinking more or less valuable? Where does it homogenise ideas?
- How would you redesign your week to give the default network more time?
- Is the "mad genius" link helpful or harmful to talk about?
Whole-book drill (pass 2)
Walk one live problem through the stack: (1) label S/A/E, (2) name the frame (ch. 5), (3) check reward fuel and maximiser tax (ch. 3), (4) schedule a default-network block (ch. 6), (5) at impasse, broaden like the Zen CRA session (ch. 7), (6) invite one discordant voice to thaw freeze (ch. 8), (7) dial filters down for generation and up for shipping (ch. 9-10), (8) use off-peak or post-fatigue timing and a positive-emotion nudge before the hard generative hour (ch. 11). If any step is skipped, note which failure mode returned.
Closing pact
Keep the poet and the logician on speaking terms. Schedule both. Measure progress by problems reframed and shipped, not by hours of strain.
Synthesis deepen (pass 6)
Elastic thinking is not anti-analysis. It is the skill of noticing when analysis is polishing the wrong frame. Scripts run social life; analysis tunes known games; elastic remakes the game. Desire and emotion steer which thoughts survive; EVR and agonist floods show the dial’s edges. Brains build hierarchical models; idle default activity recombines them; insight labs and wartime reframes show sudden oddball associates beating stepwise search when the path is remote. Freeze comes from functional fixedness, victory disease and incentives that punish dissent. Filters save us from noise and hide breakthroughs; schizotypy and chemistry remind that dose and craft matter. Liberation is a toolkit: idle, mindset, emotion, careful chemistry, small acts of will.
GenAI assist effects with homogenisation risk; DMN causal and neurofeedback studies; MDMA therapy setback; flexibility-persistence training work. The book’s core still holds: train recombination, protect idle, dial filters, keep analysis for verification.
One-page operating loop.
- Detect script failure (anomaly, repeated surprise, gut “this frame is wrong”).
- Loosen filters (idle, off-peak, positive broaden, bad-ideas-first).
- Recombine (remote associates, new representation, outsider prompt).
- Tighten (analysis, experiment, critique, ship).
- Update incentives so dissent and anomaly reporting survive success.
Print the five-step loop. Next stuck project, mark which step you are skipping. Most stuckness is skipping loosen or skipping tighten, not lacking talent.
Carry card. Detect → Loosen → Recombine → Tighten → Update incentives. When stuck, name the skipped step. When succeeding, ask which frozen doctrine the success might be hiding. Elastic is a loop, not a mood.
- Replication. Several vivid studies are weaker than presented: ego depletion, high ceilings, ovulation shifts, choice overload, tDCS and the nine-dot study, DRD4 candidate genes. The core neuroscience (default network, reward, insight sequence) is sturdier.
- Neat anatomy. The book sometimes maps functions onto single regions (the ACC "judge," the lateral prefrontal "filter," the right hemisphere for insight). Modern work emphasises interacting networks.
- Anecdote and celebrity. Many chapters lean on memorable personal stories and famous eccentrics, which illustrate but do not prove.
- AI claims have aged. The 2018 view that computers could not reframe or cross domains understated how fast foundation models would advance, though human-style reframing is still debated.
- Self-scoring scales (neophilia, reward, mindfulness, schizotypy) are drawn from real research tools but give only a rough self-portrait.
- Strengths: clear, warm writing; a sound core message that flexible thinking can be understood and cultivated; and practical advice that mostly matches 2026 evidence.
Who's who
| Person | Role |
|---|---|
| Nancy Andreasen | PET pioneer; saw the resting brain's "symphony" |
| Mark Beeman | Right-hemisphere language and insight (fMRI) |
| Hans Berger | Inventor of the human EEG (1929) |
| Joseph Bogen | Neurosurgeon of split-brain operations |
| Judy Blume | Novelist; the poet and logician within |
| Bonnie Cramond | Overlap of ADHD and creativity |
| Antonio Damasio | Studied patient EVR |
| Jimmy Doolittle, Francis Low | Carrier-launched bombing of Japan (1942) |
| Richard Feynman | "No problem is too small" letter (1966) |
| Barbara Fredrickson | Broaden-and-build theory |
| Person | Role |
|---|---|
| Jonathan Franzen | Novelist who reinvented his career |
| Deborah Gordon | Harvester ant colonies |
| Leonard Heston | Children of schizophrenic mothers (1960s) |
| Douglas Hofstadter | The dog-and-bone problem |
| Garry Kasparov | Chess champion beaten by Deep Blue |
| John Kounios | EEG studies of insight |
| Thomas Kuhn | Paradigm shifts |
| Stanley McChrystal | Adaptive command; lessons of Yom Kippur |
| Serge Moscovici | Minority dissent and blue slides |
| Nathan Myhrvold | Intellectual Ventures |
| Person | Role |
|---|---|
| John Nash | Game theorist; schizophrenia |
| James Olds, Peter Milner | Self-stimulating rats (1954) |
| Marcus Raichle | Named the default mode (2001) |
| Carl Sagan | Anonymous essay on cannabis |
| Clarence Saunders | Invented the self-service grocery |
| Mary Shelley | Conceived Frankenstein in a relaxed state |
| Roger Sperry | Split-brain research; Nobel 1981 |
| Nikola Tesla | Inventor and eccentric |
| David Wallerstein | Introduced jumbo popcorn |
| Eli Zeira | Israeli intelligence chief, 1973 |
Glossary
| Term | Meaning |
|---|---|
| ADHD | Attention deficit hyperactivity disorder, linked to weaker reward signalling |
| Alpha waves | EEG rhythm associated with inward attention; burst before insight |
| Analytical thinking | Top-down, logical, step-by-step thought |
| Anterior cingulate cortex (ACC) | Midline region that monitors conflict and may broaden attention |
| Association cortex | Brain areas that link sensory data into meaning |
| Bottom-up processing | Intelligence emerging from many interacting units without a supervisor |
| Broaden-and-build | Fredrickson's theory that positive emotions widen thinking |
| Cognitive filter | Unconscious process deciding which ideas or perceptions reach awareness |
| Cognitive inhibition | Strong filtering, producing conventional thought |
| Compound remote associates (CRA) | Puzzles needing one word that joins three others |
| Term | Meaning |
|---|---|
| Concept cells | Neurons that fire for a specific person, place or idea |
| Corpus callosum | Bridge of fibres between the hemispheres |
| Default mode network | Brain network active during rest, mind-wandering and self-reflection |
| Divergent thinking | Generating many varied ideas from one prompt |
| Dogmatic cognition | Processing information to confirm prior beliefs |
| DRD4-7R | Dopamine receptor gene variant once linked to novelty seeking |
| EEG | Electroencephalogram, recording the brain's electrical activity |
| Einstellung effect | A familiar solution blocking a better one |
| Elastic thinking | Bottom-up, flexible thought that reframes and recombines ideas |
| Executive function | Prefrontal control of attention, goals and inhibition |
| Term | Meaning |
|---|---|
| Frozen thought | Arendt's term for unquestioned principles |
| Functional fixedness | Difficulty seeing new uses for familiar objects |
| Gamma waves | Fast EEG rhythm; burst marks the moment of insight |
| Grit | Persistent passion for long-term goals |
| Incubation | Setting a problem aside so unconscious processing continues |
| Insight | Sudden solution from an unconscious reorganisation |
| Lateral prefrontal cortex | Primate-only region that enforces rules and filters ideas |
| Maximiser | Someone who searches for the best option |
| Mindfulness | Non-judgemental awareness of the present moment |
| Neophilia | Attraction to novelty |
| Term | Meaning |
|---|---|
| Nucleus accumbens | Reward hub in the ventral striatum |
| Paradigm shift | Kuhn's term for a change in the framework of thought |
| Prefrontal cortex | Front of the brain; the executive |
| Restructuring | Reframing a problem to reveal a solution |
| Reward prediction error | Dopamine signal of better- or worse-than-expected outcomes |
| Satisficer | Someone who accepts the first good-enough option |
| Schizotypy | Spectrum of schizophrenia-like traits in healthy people |
| Scripted processing | Fixed routines triggered by cues |
| tDCS | Transcranial direct current stimulation of the scalp |
| Ventral tegmental area (VTA) | Midbrain source of dopamine for reward |
| Bottom-up processing | Neural coalitions form without executive dictation; basis of elastic thought. |
| Top-down processing | Executive control directs step-by-step analysis. |
| Functional fixedness | Seeing an object only in its usual use (Duncker's candle). |
| Latent inhibition | Tendency to ignore familiar stimuli; lower values can admit more remote ideas. |
| Beginner's mind | Zen-derived stance of dropping preconceptions. |
| CRA | Compound remote associates puzzle used in insight research. |
| Default mode network (DMN) | Interacting regions active when not focused on an external task. |
| Broaden-and-build | Fredrickson's theory that positive emotion widens thought and builds resources. |
Study sheet
Recite S-A-E, then poet/logician, before each chapter review.
| Term | One-line cue |
|---|---|
| Satisficer | First good-enough option; Simon 1956 |
| CRA | Three words, one compound link; insight vs analysis |
| Default network | Association cortices active in rest and mind-wander |
| ACC | Conflict monitor that can broaden attention |
| Right aSTG | Oddball remote associates in insight |
| Schizotypy | Dose of unusual traits; filter laxity, not full illness |
| Functional fixedness | Box-as-container blindness (Duncker candle) |
| Cognitive inhibition | Stringent filter; conventional thought |
| Neophilia | Attraction to novelty that starts elastic search |
| Broaden-and-build | Positive emotion widens thought and resources |
Drill: cover the cue column, say the term; then reverse.
About this book
Elastic: The Power of Flexible Thinking in a Time of Change by Leonard Mlodinow, 2018. 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