David Bessis, translated by Kevin Frey — a book report

David Bessis wants to tell you a secret that he insists is not a secret at all: mathematics is not a gift, it is a private practice of building pictures in your head, and anyone willing to sit with confusion long enough can learn it. Readers have met that promise with real warmth and real irritation, often in the same review. Across 873 ratings the book averages 4.18 stars — 46% five-star, and a stubborn 20% at three or fewer. Having read it and marked it up heavily, I think both camps are right, and their disagreement is more interesting than either verdict alone.

What converts people is the book's description of mathematical thinking from the inside. One reader wrote that the first part "contained the best description of what 'doing math' feels like...I've ever read," even though the last third lost him. Another called it "the most insightful book I read on mathematical thinking and math education." That inside view is what I highlighted most. Bessis writes that "the real pleasure of mathematics is waking up one day and realizing that you can see stars in your head, which you'd never been able to do before," and that mental images, "once constructed, allow you to see it immediately and without any effort. But it takes a lot of time and effort to construct them." He quotes William Thurston's line that "one person's clear mental image is another person's intimidation," which reframes every moment I have ever felt stupid reading a proof. The proof was never the thing. The picture behind it was, and nobody had shown me theirs.

The second thing readers love is the demolition of genius. "Anyone can do math," one five-star reviewer wrote. "Just like anyone can ride a bike." Another praised the book for "dismantling genius myths" in favor of patience and imagination. Bessis builds this case through analogy rather than argument.

The Fosbury flop is unlearnable from written instructions — "to truly learn a movement, you have to understand it beyond words. You have to feel it within your own body." Even Einstein gets recruited: "I have no special talent. I am only passionately curious." The line I keep returning to describes what reclaiming childhood learning would require — "to become once again capable of devoting ten or twenty hours to something that may or may not be impossible, without being distracted by the feeling of your own uselessness."

The intellectual core is Bessis's addition to Daniel Kahneman. Kahneman gives us System 1 (fast intuition, reliably biased) and System 2 (slow reasoning, exhausting and error-prone), and recommends memorizing the biases so System 2 can override System 1; Bessis calls this a counsel of despair. He proposes System 3: "an assortment of introspection and meditation techniques aimed at establishing a dialogue between intuition and rationality," used "each time you try to recall your dreams, to put words to the fleeting impression that left a strange taste in your mouth." The goal is not to defeat intuition but to retrain it. "In the end, it's almost always my intuition that wins. When I force it to listen to what logic is saying, it takes that into account and adjusts its position. Logic is something inert, like a pebble. My intuition is organic, it is living and growing." Grounding all this is his flat insistence that intuition "isn't a magical elixir, or your lucky star, or the hand of God on your shoulder" but "the entanglement of synaptic connections between your neurons."

Here is where the critics land, and they land hard. The sharpest two-star review says that despite the promising opening, "there are absolutely ZERO practical ideas on learning or teaching math in this book." A three-star reader agrees it "falls short after the introduction, turning into a collection of anecdotal stories...with little meaningful content." I think this is unfair on the letter and fair on the spirit. There are exercises: lying in bed reconstructing every room he has slept in, "the size and orientation of the bed, the location of the walls and ceiling"; a viewpoint-switching drill using a fixed reference point across the room; a closing five-step protocol beginning "you can reprogram your intuition" and "don't expect it all to come at once." But he refuses to systematize any of it, and tells you why in advance: "Recipes never capture all the steps you need to take," and "each time someone talks to you about 'tricks,' they're telling you to stop thinking at precisely the moment when it starts to get interesting." His instruction is deliberately unfollowable — "rather than learning my pictures by heart, train yourself to construct pictures that work for you." That is intellectually honest and also very convenient for an author who has promised a method.

The other complaints I share. Several four-star readers note heavy redundancy, a patronizing register, and "some self-aggrandizing moments"; the book circles its thesis many times. And there is a tension Bessis never resolves: after insisting for hundreds of pages that talent is a myth, he writes that "geometry has been one of my strong points since childhood," and describes a six-week "state of hyperlucidity" in which he proved a conjecture. One reviewer put the objection precisely — the book corrects a crude formalist view of mathematics but leaves open how much of Bessis's own capacity was trained and how much was simply there.

What has stayed with me is the least mathematical part. The real barrier, Bessis argues, is social: "if you acknowledge how lost you are, you'll look like a fool. The social norm is to let it go." Jean-Pierre Serre's lesson was to say plainly when he did not understand, which "requires a great command of your body and emotions, because we have the instinct to hide our ignorance." Paired with Grothendieck's warning that "the person who fears being wrong is powerless to discover anything new," it is the most usable idea in the book, and it applies far outside mathematics.

Four stars. It is an original account of what understanding feels like, wrapped in a self-help promise it cannot quite keep — and the failure is honest, because the thing it teaches may not be teachable in prose.