The man who memorised nonsense to measure forgetting

In the 1880s Hermann Ebbinghaus learned lists of meaningless syllables and timed how fast they slipped away. Along the way he found a cheaper way to keep them.

Much of what you learn today will be hard to recall in a week. That sounds discouraging, but the first person to measure it carefully also found a way to push back, and his finding still holds.

Hermann Ebbinghaus, a German psychologist, wanted to show that memory could be studied by experiment. Real words carry meanings and associations that make some easier to learn than others, so he built a set of about 2,300 syllables, each a consonant, a vowel and a consonant, chosen to mean nothing. He drew lists of them at random and used one test subject: himself. He read each list to the beat of a metronome until he could recite it without a mistake, and he published the results in 1885 in a book called Über das Gedächtnis, On Memory.

To measure forgetting, he used a clever trick. After a delay, he learned the same list again and noted how much less effort it took the second time. He called the difference savings. Even when he couldn't recall a single syllable, relearning was quicker, so something had stayed.

The savings dropped fast and then levelled off. After 20 minutes he saved about 58 percent of the original effort, after an hour about 44 percent, after a day about 34 percent, and after 31 days still about 21 percent. Most of the loss came early. Drawn as a graph, these numbers make the forgetting curve.

Spreading beats cramming

In the same book he tested what happens when practice is spread out. Sixty-eight repetitions of a 12-syllable list, made one after another on a single day, let him recite it again the next day after 7 more readings. Thirty-eight repetitions, spread over the three days before, did just as well. He judged it probable that spreading repetitions out over time is decidedly better than massing them together.

This is now called the spacing effect, and it is one of the sturdiest findings about learning. An analysis of earlier studies published in 2006 by Cepeda and colleagues found that spaced practice beat massed practice in 259 of 271 comparisons.

Ebbinghaus's results came from one person, but they have held up. In 2015 Jaap Murre and Joeri Dros of the University of Amsterdam repeated his forgetting experiment with a single subject, who spent about 70 hours learning and relearning lists, and found a curve similar to the original.

The practical lesson is simple. A few short reviews with gaps between them keep something longer than the same time spent all at once.