The Ebbinghaus curve

The experiment that produced a law

Hermann Ebbinghaus worked largely alone, at the University of Göttingen and later Berlin, during the 1880s. He wanted to bring memory inside the scope of experimental science — something most of his contemporaries thought impossible for so private and variable a mental event. His method was to construct lists of meaningless consonant-vowel-consonant syllables (DAX, BUP, LOC) specifically because they carried no prior associations that would contaminate the measurement. He memorised these lists himself, tested himself, and tracked how much he had forgotten after intervals ranging from twenty minutes to a month.

The result he reported in Über das Gedächtnis (1885) — translated into English as Memory in 1913 — was the forgetting curve: a steep initial drop in retention followed by a progressively shallower decline. The mathematics Ebbinghaus fitted to his data described retention as falling as a function of the logarithm of time since learning. Roughly, he found that after about an hour he had forgotten more than half of what he had learned; after a day, the curve was already flattening; after a month, only a residue remained unless he had revisited the material.

What the curve actually says — and what it doesn’t

The shape is robust. The specific equation he proposed has been refined — researchers including John Anderson have fitted power functions that outperform the logarithmic version — but the qualitative picture survives.

What the curve is often taken to say, however, goes beyond what it shows. Ebbinghaus used nonsense syllables on a single subject: himself. That subject was rigorous, but he was not representative. The self-experiment that produced the data is both psychology’s founding quantitative achievement and a standing methodological caution.

Meaningful material — prose, faces, skilled procedures — is forgotten on a different timescale, often far slower. Frederic Bartlett’s later work at the University of Cambridge made this explicit: memory for meaningful content is reconstructive, shaped by prior knowledge, and far stickier than a list of syllables would suggest.

The curve also describes forgetting under a specific condition: a single learning episode, with no subsequent practice. It is not a verdict on what memory can do, but a baseline for what it does when left alone. That distinction matters enormously, because the same Ebbinghaus who charted the decay also documented the savings method — measuring not whether a person could recall a list, but how much less effort it took to relearn one. Even when explicit recall had collapsed entirely, relearning was substantially faster. Some trace persisted that the forgetting curve’s surface reading would miss.

Decay over time is one account; interference from subsequent learning is another, and has strong experimental support.

What the curve gave subsequent research

The forgetting curve set the template for every quantitative memory study that followed. Its most direct descendant is distributed practice: Ebbinghaus also showed that the same number of repetitions produced far better retention when spread across days than when massed in a single session — an early demonstration of the spacing effect that has since replicated across an unusually wide range of materials and ages.

The curve also underpins the logic of spaced-repetition software, which attempts to schedule review just before forgetting erodes it, catching memory before it falls below the retrieval threshold. Whether the exact timing algorithms used by such systems matter as much as their marketing implies is a separate and contested question, but the underlying premise — that forgetting is predictable enough to be gamed — comes directly from Ebbinghaus’s 1880s notebooks.

What the curve cannot tell us is why forgetting happens. Endel Tulving’s work on encoding specificity added a third angle: retrieval depends on matching the conditions of encoding, so “forgotten” material may be inaccessible rather than gone. The forgetting curve describes the phenomenon with admirable precision. The explanation remains genuinely open.