How We Learn
Benedict Carey
How We Learn is science journalist Benedict Carey's tour of a century of memory research, and its verdict is uncomfortable: habits we trust, such as rereading, cramming and long single-topic sessions, often work worse than methods that feel awkward. Recalling before you look, spacing your reviews, mixing topics, taking real breaks and sleeping on it all make learning stick.
Most of us study the way we were taught: highlight, reread, repeat until the page feels familiar. Carey's point is that familiarity is a trap. A text feeling fluent in front of your eyes is not the same as being able to recall it with the book closed.
The book in essence
Carey, a New York Times science reporter, builds the book around research by psychologists such as Robert and Elizabeth Bjork, whose idea of “desirable difficulties” runs through every chapter: a bit of struggle during learning, such as having to retrieve, having partly forgotten, or having to choose a method, is exactly what makes memory durable.
It is not a rigid system but a set of adjustments: close the book and recall, spread the same hours over more days, mix problem types, step away from a stuck problem on purpose, and treat sleep as part of studying rather than its competitor. The findings are not equally strong. Retrieval practice and spacing are among the best-replicated results in learning science, while claims about sleep stages and incubation deserve more caution, and Carey is mostly candid about the difference.
Who it's for
- Students facing high-stakes exams who suspect they study long rather than well
- Working adults learning a language, coding or a professional skill on the side
- Teachers, trainers and parents helping children study
Who it's not for
- — Readers who want a step-by-step study schedule or ready-made templates
- — People already following memory research at an academic level; the book stays at popular-science depth
Key lessons
01
Close the book and recall
Rereading feels productive because the material becomes familiar, but recognition is not recall. Pulling an answer out of memory before checking it strengthens that memory far more than another pass through the page. Carey describes experiments in which students who spent part of their time on small self-tests beat those who only reread, even though the rereaders felt more confident. A practice test is not just a measurement; it is the studying itself. Even getting an answer wrong first and then checking it helps.
Dana is preparing for her driving theory test. Instead of reading the road-sign booklet a third time, she covers each page, says out loud what each sign means, then lifts her hand to check. The first round is humbling: she misses a third. By the end of the week the misses are rare, and she has spent fewer hours than her friends who kept rereading.
02
Space your reviews
The same total hours, spread across days, beat one long session. When you return to material after a gap you have forgotten a little, and the effort of recalling it rebuilds the memory more strongly than reviewing it while it is still fresh. Gaps can grow: a day, a few days, a week, longer. Cramming can rescue tomorrow morning's test; spacing is what still works next month. The forgetting between sessions is not wasted time but the condition that makes each review count.
Tom has three weeks to learn basic Spanish before a trip to Madrid. Rather than one six-hour Saturday marathon, he does twenty minutes on his commute on days 1, 2, 4, 8, 15 and 21, always recalling old phrases before adding new ones. On arrival he can still order dinner and ask for directions without reaching for his phone.
03
Shuffle the deck
Practising one problem type in a block produces smooth progress and false confidence, because you always know which method comes next. Mixing types in one session feels slower and messier, yet studies of maths, motor skills and even recognising painters' styles found that mixed practice led to better scores on later tests. The reason is simple: real exams and real life do not announce the problem type. Interleaving trains the skill of telling problems apart and choosing the right tool.
A physics student usually works through a chapter of kinematics problems, then a chapter of energy problems. Before her final, she copies past questions from five chapters onto index cards and draws them at random. The first evening is frustrating. But in the exam she no longer freezes when a question gives no hint of which chapter it came from.
04
Step away, forget a little, sleep on it
Hard problems need two rhythms: concentrated work that loads the problem into your mind, then a genuine break. Carey covers incubation research suggesting that stepping away once you are stuck can help a solution surface. He also reframes forgetting: it mostly dims access to a memory rather than erasing it, and relearning something partly forgotten makes it stronger. Sleep, finally, appears to consolidate and reorganise what you studied. An all-nighter trades away the very process that would have fixed the material.
A junior developer stares at a bug for two hours. Instead of pushing on until midnight, she writes down what she has already ruled out, goes for a walk and gets to bed early. Next morning, rereading her note over coffee, she spots the wrong assumption within minutes, one her tired eyes had skipped past again and again.
Try this today
Pick something you studied this week. Close your notes, write everything you remember on a blank page for ten minutes, then check what you missed. Put the same exercise in your calendar for three days from now.
Quick check
According to the research Carey describes, which study method usually strengthens memory most?
Why does mixed (interleaved) practice tend to beat blocked practice?
Selected quotes
“A curious peculiarity of our memory is that things are impressed better by active than by passive repetition.”
“While we teach, we learn.”