The Uses of Forgetting

Most of us treat forgetting as a kind of failure. We scold ourselves for losing a name, a password or the point of a story halfway through telling it, and we tend to assume that a perfect memory would be an unmixed blessing. The science of memory, however, paints a more interesting picture. Forgetting, it turns out, is not simply what happens when memory breaks down; in many cases, it is part of how memory is supposed to work.

One of the first systematic studies of forgetting was carried out in the 1880s by a German psychologist named Hermann Ebbinghaus, who chose an unusual research subject: himself. To prevent meaning and prior knowledge from interfering with his results, he memorized long lists of nonsense syllables, short strings of letters that can be pronounced but mean nothing at all. He then tested himself after intervals of different lengths, measuring how much effort it took to relearn each list. The pattern he described has come to be known as the forgetting curve. Without review, memory faded most steeply in the hours after learning, and then the rate of forgetting gradually slowed, so that what was left declined far more gently.

Ebbinghaus also noticed something that a large body of research has since confirmed. When reviews are spread out over several days rather than crammed into a single session, material tends to stay with us for much longer. This phenomenon, known as the spacing effect, helps to explain why the familiar habit of staying up all night before an exam is such a poor investment. Cramming can produce a convincing sense of mastery the next morning, but much of what was squeezed in tends to fade soon afterward. Spaced review feels harder, partly because we have begun to forget between sessions, and many researchers believe that the effort of recalling half-forgotten material is part of what makes it stick.

Forgetting is not the only way in which memory departs from the faithful record we imagine it to be. We tend to think of remembering as replaying a recording, but memories are reconstructed each time we recall them, pieced together from fragments and filled in with whatever seems most likely. Usually this works well enough, but it leaves memory open to influence. In a series of well-known experiments, the American psychologist Elizabeth Loftus and her colleagues showed people films of traffic accidents and then asked them questions about what they had seen. Simply asking how fast the cars were going when they "smashed" into each other, rather than when they "hit" each other, led people to estimate higher speeds, and those who heard the more violent verb were later more likely to recall broken glass that had never appeared in the film. Findings like these have influenced the way many courts treat eyewitness testimony.

None of this means that memory is hopelessly unreliable. Rather, it suggests that the brain is less concerned with storing every detail than with extracting what is useful. If you remembered every dog you had ever met in perfect detail, it might actually be harder to recognize an unfamiliar animal as a dog, since no new dog would exactly match any of the old ones. Letting the particulars fade allows the general pattern to stand out. In a similar way, forgetting where you parked yesterday makes it easier to remember where you parked today. Seen from this angle, forgetting is less a flaw than a filter, clearing away what we no longer need so that what matters can come into focus.

For anyone learning a language, these findings carry a practical lesson. The words most worth keeping are usually the ones you actually meet in a book, an article or a conversation, rather than those copied from a long list. Saving such words as you read, together with the sentence in which you found them, gives each one a context to cling to. Reviewing them at growing intervals, perhaps after a day, then a few days, then a week or two, works with the forgetting curve rather than against it, since each review ideally comes just as the word is beginning to slip away. Some words will still be forgotten, and that is fine; the ones that keep reappearing in your reading will gradually earn their place.