Why do we forget so quickly what we have spent hours learning? This question has probably accompanied the history of education since its earliest days. Whether it concerns a foreign language, a university course, a professional certification or simply a personal passion, everyone has experienced this paradox: the greater the initial effort seems, the more unjust the forgetting that follows appears.
Yet for more than a century, cognitive science has provided a remarkably consistent answer to this phenomenon. We do not forget because we learn poorly. We forget because our memory functions precisely this way. And it is this reality, long considered a weakness, that ultimately gave rise to one of the most effective learning methods ever studied: spaced repetition.
Today, this approach is used in fields as varied as language learning, medical studies, competitive exam preparation and continuing professional development. It also constitutes the fundamental principle underlying modern memorisation applications like Orendy, which place spaced repetition at the heart of their design.
Hermann Ebbinghaus and the Discovery of the Forgetting Curve
The history of spaced repetition begins long before computers, smartphones and even modern psychology as we understand it today.
In 1885, the German psychologist Hermann Ebbinghaus published Über das Gedächtnis (Memory: A Contribution to Experimental Psychology), a work considered one of the first scientific studies devoted to the workings of human memory. His approach was as rigorous as it was unusual: lacking volunteers, he decided to use his own memory as the subject of his experiments.
For several years, Ebbinghaus memorised thousands of meaningless syllables in order to observe precisely the rate at which they disappeared from his memory. His results revealed a phenomenon that has since become famous: the forgetting curve.
The conclusion was unambiguous. A large proportion of newly learned information is forgotten within hours or days of its acquisition. But Ebbinghaus also observed something essential: when information is reactivated before it is completely forgotten, it subsequently resists the erosion of time far longer.
This discovery, formulated nearly one hundred and fifty years ago, remains to this day the foundation of all modern spaced repetition methods.
Why Our Intuition Leads Us to Learn Poorly
Perhaps the most surprising aspect is that spaced repetition runs directly counter to our intuition.
When an exam approaches, our natural reflex is generally to read and reread our notes, sometimes for several hours at a stretch. This strategy produces an immediate sense of mastery: the information seems familiar, accessible, almost obvious.
But this impression is often misleading.
From the 1970s onwards, the work of researchers such as Robert and Elizabeth Bjork showed that the most durable learning is often that which requires an active effort of memory retrieval. In other words, having to actively recall information contributes more to learning it than simply rereading it.
This phenomenon, sometimes called the testing effect or retrieval practice, has been confirmed by numerous studies, notably those of psychologists Henry Roediger and Jeffrey Karpicke. In simple terms, trying to remember is already a way of learning.
This is precisely the principle that spaced repetition systems exploit: they do not seek to make us repeat more, but to make us review at the most effective moment.
How Spaced Repetition Works in Practice
The general principle is remarkably simple.
When information is new, it must be reviewed quickly: sometimes the following day, then a few days later. As this information becomes anchored in memory, the intervals between reviews gradually lengthen: one week, two weeks, a month, then several months.
Take the example of learning a Japanese vocabulary word. A first review might occur after twenty-four hours. If recall is successful, the next can take place a few days later, then several weeks after that. After a few successful reviews, that same word might sometimes be retained for months before requiring reactivation.
This progression is not arbitrary. It rests on the idea that each retrieval effort progressively strengthens the memory trace. Human memory functions less like a hard drive than like a forest path: the more regularly it is used, the easier it becomes to find again.
From Cardboard Flashcards to Modern Algorithms
Long before mobile applications appeared, certain educators had already grasped the value of organising reviews over time.
In the 1970s, German journalist Sebastian Leitner popularised a system using boxes containing cards. Correctly memorised cards progressively moved into compartments consulted less frequently, while errors returned to the most frequently visited compartments.
Computing then made it possible to automate this principle. The first specialised software appeared in the 1980s, notably with the work of Polish researcher Piotr Woźniak on the SuperMemo algorithms. Later, applications like Anki would contribute greatly to bringing spaced repetition to the general public.
But while the scientific principles have remained relatively stable for several decades, their implementation continues to evolve.
Spaced Repetition: a Science, but also a Human Experience
One of the historical limitations of spaced repetition systems stems from a relatively simple paradox: it is not enough for a method to be scientifically optimal for it to be genuinely used over the long term.
Learning takes place within daily life, with its professional obligations, fatigue, periods of high motivation and others where energy is in shorter supply. A method that is perfect on paper can therefore fail if it does not account for this very human reality.
This is precisely the direction in which certain contemporary applications like Orendy are evolving. Without calling into question the fundamental principles established by cognitive research, they seek to adapt spaced repetition to the real constraints of users.
Thus, rather than treating learning as a succession of absolute successes or failures, modern systems tend to favour gradual progressions, incremental consolidation and mechanisms designed to avoid both intensive cramming and cognitive overload. They also make it possible to distinguish phases of free exploration from genuine consolidation sessions, or to take into account periods when the user wishes to temporarily slow their learning pace.
This evolution may seem secondary. It is probably essential. For the best memorisation method is not necessarily the one that optimises a few percentage points in a laboratory, but the one a user will continue to practise several years later.
One of the Few Learning Methods Validated by Over a Century of Research
In a field where miracle methods and promises of accelerated learning appear with almost seasonal regularity, spaced repetition stands as an exception.
Its fundamental principles have withstood more than one hundred and thirty years of research, experimentation and independent validation. Few pedagogical techniques can claim such scientific continuity.
Spaced repetition promises neither photographic memory nor effortless learning. It offers something more modest, but perhaps more precious: working with the natural functioning of our memory rather than against it.
And there is perhaps a certain intellectual elegance in recognising that, faced with the complexity of the human brain, one of the most effective strategies ever discovered ultimately rests on a rather simple idea: to retain information for a long time, one must accept forgetting it a little.
Sources
- Hermann Ebbinghaus, Über das Gedächtnis (Memory: A Contribution to Experimental Psychology), 1885.
- Robert A. Bjork & Elizabeth L. Bjork, Making Things Hard on Yourself, But in a Good Way, 2011.
- Henry L. Roediger III & Jeffrey D. Karpicke, Test-Enhanced Learning: Taking Memory Tests Improves Long-Term Retention, Psychological Science, 2006.
- Jeffrey D. Karpicke & Janell R. Blunt, Retrieval Practice Produces More Learning than Elaborative Studying with Concept Mapping, Science, 2011.
- Daniel T. Willingham, Why Don’t Students Like School?, 2009.
- Piotr Woźniak, Optimization of Learning, SuperMemo World.