Skip to main content

The Forgetting Curve

How We Discovered the Math of Memory​

Article 1 · Spaced Repetition Series

Imagine spending an entire evening memorizing fifty foreign words, only to discover the next morning that you can barely recall a dozen. By the end of the week, perhaps three or four remain. This is not a failure of intelligence or willpower. It is how human memory works by default, and understanding the mathematics behind this decay turns out to be the single most powerful insight for anyone who wants to learn a foreign language efficiently.

For more than a century, cognitive scientists have studied why we forget and, more importantly, how we can strategically interrupt that forgetting. The technique born from this research, known as Spaced Repetition, has become one of the most rigorously validated methods in all of educational psychology. This article traces its origins from a solitary researcher in nineteenth-century Berlin to the practical systems millions of language learners use today.

Hermann Ebbinghaus and the Birth of Memory Science​

In 1879, a young German psychologist named Hermann Ebbinghaus set out to do something no one had attempted before: to measure memory with the precision of a physicist measuring gravity. At the time, memory was considered a philosophical subject, far too slippery and subjective to be captured in numbers. Ebbinghaus disagreed.

His method was as ingenious as it was grueling. He invented roughly 2,300 nonsense syllables — meaningless combinations like "DAX," "BUP," and "ZOL" — specifically to strip away any prior knowledge or emotional associations. He then became his own sole experimental subject, memorizing lists of these syllables, testing himself at precise intervals, and meticulously recording every result. This self-experimentation lasted more than two years.

Why nonsense syllables? Ebbinghaus realized that if he used real words, his existing knowledge would contaminate the results. A word like "house" already carries meaning, imagery, and emotional resonance. A syllable like "WID" carries nothing. It was, in a sense, the purest raw material for studying how the brain encodes and retrieves new information — precisely the kind of material a language learner faces when confronting a truly foreign word for the first time.

"Psychology has a long past, yet its real history is short." — Hermann Ebbinghaus, Memory (1885)

The results of his painstaking work were published in 1885 in a slim monograph titled Über das Gedächtnis (On Memory), and they contained a finding so robust that it has been replicated hundreds of times over the following 140 years.

The Forgetting Curve: Visualizing Memory Decay​

Ebbinghaus discovered that newly learned information follows an exponential decay function. In mathematical terms, the probability of recalling an item can be approximated by the formula

R=e−t/SR = e^{-t/S}

where R is the retention rate, t is the time elapsed since learning, and S is the "stability" of the memory — a measure of how deeply it has been encoded.

In practical terms, this means that forgetting is not linear. You do not lose ten percent of your memories each day in a steady drip. Instead, the steepest drop happens in the first hours. Ebbinghaus found that within the first twenty minutes after learning, retention could fall to around 58 percent. After one hour, it was roughly 44 percent. After a single day, only about 34 percent remained. The curve then flattens, with the surviving memories becoming increasingly resistant to further decay.

The Ebbinghaus Forgetting Curve Figure 1. The Ebbinghaus Forgetting Curve. Without active review, newly learned material fades rapidly, with the steepest loss occurring in the first 24 hours.

For vocabulary learners, this has a direct and sobering implication: the traditional study session, in which a student sits down the night before an exam and crams a list of words, is working against the fundamental architecture of the brain. Most of those words will be gone within days.

The Spacing Effect: A Counter-Intuitive Discovery​

But Ebbinghaus also discovered something far more hopeful. When he relearned the same lists after intervals of time, the forgetting curve flattened. Each successive review made the memory more durable, as if the act of retrieving a fading memory strengthened the neural pathway. Moreover, he noticed that the timing of review mattered enormously.

This phenomenon — that learning is greater when practice sessions are spaced apart rather than massed together — came to be known as the Spacing Effect. It is one of the most robust findings in the entire history of experimental psychology. A comprehensive meta-analysis by Cepeda and colleagues in 2006, reviewing 254 studies involving more than 14,000 participants, confirmed that distributed practice consistently outperformed massed practice for long-term retention across a wide variety of materials and age groups.

The Spacing Effect in Action Figure 2. The Spacing Effect in action. Each review (colored dots) resets retention to near 100% while progressively flattening the forgetting curve, meaning longer intervals between reviews become possible over time.

The counter-intuitive insight is this: reviewing material just before you are about to forget it is far more efficient than reviewing it while you still remember it well. The slight effort of retrieval during that moment of near-forgetting appears to be the signal that tells the brain to consolidate the memory more deeply. Cognitive psychologists call this phenomenon "desirable difficulty." The struggle is not a sign of failure — it is the mechanism of learning itself.

"Without a struggle, there is no strength. The effort of recall is what builds long-term memory." — Robert A. Bjork, UCLA

From Theory to Practice: Pimsleur and Leitner​

Paul Pimsleur and Graduated-Interval Recall​

While Ebbinghaus had identified the underlying principle, it took nearly a century before anyone systematically applied it to language learning. In the 1960s, the American linguist Paul Pimsleur developed what he called "Graduated-Interval Recall," an audio-based language instruction method that introduced new vocabulary and then prompted learners to recall it at precisely timed, expanding intervals: five seconds, twenty-five seconds, two minutes, ten minutes, one hour, five hours, one day, five days, twenty-five days, and so on.

Pimsleur understood intuitively what the research confirmed: each successful recall at the edge of forgetting consolidated the word more firmly, allowing the next interval to be longer. His audio courses, still in print today, remain one of the most effective commercially available tools for building a basic spoken vocabulary in a new language.

Sebastian Leitner and the Flashcard Box​

Around the same period, the German science journalist Sebastian Leitner proposed an elegant physical system for applying spaced repetition to flashcard study. Published in his 1972 book So lernt man lernen (Learning to Learn), the Leitner system uses a series of boxes to organize flashcards by difficulty.

The Leitner Box System Figure 3. The Leitner Box System. Cards advance to the next box when recalled correctly and return to Box 1 when recalled incorrectly, creating a natural spacing algorithm.

The mechanism is simple. All new cards start in Box 1, which is reviewed daily. When a card is answered correctly, it advances to Box 2, which is reviewed every two days. A correct answer in Box 2 moves it to Box 3 (reviewed every four days), and so on. However, any card answered incorrectly at any stage returns immediately to Box 1. This creates a beautifully self-correcting system: difficult words receive the most practice, while well-known words are reviewed less frequently but never entirely abandoned.

For vocabulary learners, the Leitner Box was revolutionary because it solved a practical problem that Ebbinghaus's research alone could not: how to manage hundreds or thousands of words, each at a different stage of consolidation, without relying on a computer. The physical box provided its own algorithm.

Why This Matters for Language Learners​

The story of the forgetting curve is not merely a historical curiosity. It is the scientific foundation upon which every modern spaced repetition application — from Anki to SuperMemo to the algorithm powering Duolingo's word review — is built. Understanding why these tools work, not just how to use them, empowers learners to make better decisions about their study habits.

The core lessons from more than a century of research can be distilled into three principles for effective vocabulary acquisition. First, new words will be forgotten rapidly unless actively reviewed; the forgetting curve is not optional. Second, the timing of review matters more than the volume of review; a single well-timed repetition can be worth more than five poorly timed ones. Third, the difficulty you feel when trying to recall a word on the verge of being forgotten is not a problem to avoid but rather the very engine that drives long-term retention.

In the next article in this series, we will examine how these principles were translated into digital algorithms — and how modern spaced repetition software calculates the optimal moment to show you each flashcard.

References & Further Reading​

Scientific Literature​

  1. Ebbinghaus, H. (1885). Über das Gedächtnis: Untersuchungen zur experimentellen Psychologie. [English translation: Memory: A Contribution to Experimental Psychology]. Duncker & Humblot.

  2. Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., & Rohrer, D. (2006). Distributed practice in verbal recall tasks: A review and quantitative synthesis. Psychological Bulletin, 132(3), 354–380.

  3. Pimsleur, P. (1967). A memory schedule. The Modern Language Journal, 51(2), 73–75.

  4. Bjork, R. A. (1994). Memory and metamemory considerations in the training of human beings. In J. Metcalfe & A. Shimamura (Eds.), Metacognition: Knowing about knowing (pp. 185–205). MIT Press.

  5. Leitner, S. (1972). So lernt man lernen: Der Weg zum Erfolg [Learning to Learn]. Herder.

Online Resources​