Spaced Repetition for Language Learning

Spaced repetition is a simple idea that quietly does most of the heavy lifting in modern vocabulary learning: review each word just before you are about to forget it, and stretch the gap a little longer every time you get it right. Instead of drilling a list ten times in one sitting and losing most of it by next week, you see each item again at the exact moment it starts to fade, which resets the memory and buys a longer break before the next review. Do this consistently and you can keep hundreds of words alive for a few minutes of daily review. This article explains the science, the algorithm, and how to use it without burning out.

What is the forgetting curve?

The whole method rests on one observation, first measured by the German psychologist Hermann Ebbinghaus in 1885. Learn something new and your ability to recall it decays over time along a predictable shape called the forgetting curve. The drop is steepest in the first hours and days: without any review, you might lose more than half of what you learned within a day or two, and most of the rest over the following week.

The forgetting curve is not a bug to be fought with brute repetition. It is a schedule to be exploited. Ebbinghaus, and every researcher since, found that each time you successfully recall something, the curve flattens. The memory decays more slowly after the second exposure than the first, more slowly still after the third, and so on. So the winning strategy is not to review constantly, which wastes effort on words you already know, but to review at the precise point where recall is still just possible but about to fail. That effortful, successful recall is what strengthens the memory most.

Why is spaced repetition better than cramming?

Cram a word list the night before and you will remember it in the morning and forget most of it by the weekend. This is massed practice, and it produces a strong but brittle memory. Spaced practice takes the same number of repetitions and distributes them across expanding intervals: day 1, day 3, day 8, day 20, and so on. Each review is slightly harder because more time has passed, and that extra effort is exactly what makes the resulting memory more durable.

This is the spacing effect, one of the most reliably replicated findings in all of cognitive psychology. For the same total study time, spaced review produces far more long-term retention than massed review. The counterintuitive part is that spaced practice feels harder and less productive in the moment, because you struggle to recall, whereas cramming feels smooth and successful. That feeling is exactly backwards. The struggle is the mechanism. Easy review builds weak memory; effortful, well-timed recall builds strong memory. Learners who chase the comfortable feeling of fluent cramming are optimising for the sensation of learning rather than the result.

How does the SM-2 algorithm schedule reviews?

Doing this by hand is tedious, which is why algorithms exist. The dominant one is SM-2, designed by Piotr Wozniak in 1987 for the SuperMemo software and now the foundation of Anki and most flashcard apps, including the one on this site.

The mechanism is straightforward. Every card carries an ease factor and a current interval. When a card comes due, you review it and rate how well you recalled it, typically on a scale from "forgot completely" to "trivially easy". SM-2 uses that rating to update the card:

Recalled easily: the interval is multiplied outward (times the ease factor), so the next review might be weeks or months away, and the ease factor nudges up.
Recalled with effort: the interval still grows, but more modestly.
Failed: the card resets to a short interval (you will see it again within a day) and the ease factor drops, so it recurs more often until it stabilises.

Over time, cards you know well drift out to intervals of months, while stubborn cards keep resurfacing until they stick. The algorithm ensures you spend your review minutes almost entirely on the words that are actually at risk of being lost, and barely at all on the ones already secure. That efficiency is the entire point.

The spaced-repetition flashcard tool on this site runs an SM-2 schedule and saves your progress by word, so you can build a review deck from the Core 1,000 lists and let the algorithm decide what you see each day. It is the retention layer that sits underneath the lessons and stories, catching the words that would otherwise quietly slip off the forgetting curve.

How do I use spaced repetition without burning out?

The failure mode of spaced repetition is not the algorithm, which is nearly perfect. It is human discipline, and it fails in two predictable ways.

The first is adding too many new cards. Enthusiasm on day one says "I will learn 100 words a day". But every new card creates future reviews, and a week later those reviews all come due at once. The backlog balloons, review sessions swell to an hour, the whole thing starts to feel like a punishment, and the learner quits. The fix is boring and reliable: cap new cards at 10-20 a day. That builds to a steady mature review load of 15-30 minutes daily, which is sustainable for years.

The second failure is skipping days. Because reviews are scheduled to catch words at the point of forgetting, a missed day means those words decay past the ideal review point, and a missed week means a wall of overdue cards. Short daily contact is the design assumption. If you must miss time, it is better to reduce your new-card intake to zero for a while and just clear the reviews than to stop entirely.

Three habits keep it healthy. Do reviews daily, even a short session. Meet words in context first, in a lesson, a sentence, or a story, before they become flashcards, so the card is jogging a real memory rather than teaching a bare translation. And keep the deck honest: delete cards you no longer need and do not hoard thousands of words you never actually encounter. Spaced repetition is a maintenance tool for vocabulary you have already met, not a substitute for meeting it.

Frequently asked questions

The detailed answers to the most common questions about spaced repetition are collected in the FAQ block above. The essentials: it schedules each review just before you would forget, the forgetting curve is the reason it works, SM-2 is the algorithm that automates the timing, 10-20 new cards a day is a sustainable pace, and it suits vocabulary and discrete facts better than open-ended grammar.

Cross-references

The spaced-repetition flashcard tool is the SM-2 review layer for your vocabulary.
The word drill covers active recall through four exercise types, the complement to flat flashcards.
Why the first 1,000 words matter explains which words are worth putting into your deck first.
Active recall vocabulary drills covers why testing yourself beats re-reading.

Frequently asked

What is spaced repetition and how does it work?

Spaced repetition is a review schedule that shows you each item just before you are likely to forget it, then increases the interval every time you recall it correctly. It works with the forgetting curve rather than against it: memory decays predictably over time, but each successful recall resets and flattens that decay, so the next review can be pushed further out. Over weeks the intervals stretch from days to months, meaning you keep hundreds of words alive with only a few minutes of review a day.

Why is spaced repetition better than cramming?

Cramming (massed review) packs many repetitions into one session, which produces a strong but short-lived memory that decays fast. Spaced repetition distributes the same repetitions over expanding intervals, which produces a slightly harder but far more durable memory. This is the spacing effect, one of the most replicated findings in cognitive psychology (Ebbinghaus first documented it in 1885). For the same total study time, spacing yields substantially more long-term retention than cramming, which is exactly what a language learner needs.

What is the SM-2 algorithm?

SM-2 is the scheduling algorithm Piotr Wozniak designed for SuperMemo in 1987, and it is the basis of Anki and most modern flashcard apps. After each review you rate how easily you recalled the card; SM-2 uses that rating to adjust an ease factor and calculate the next interval. Easy cards have their gaps multiplied outward quickly; hard cards are shown again sooner. It is a simple set of rules but it captures the core insight: successful recall earns a longer break, failure resets the clock.

How many flashcards should I review a day?

For vocabulary, adding 10-20 new cards a day is a sustainable pace that most learners can maintain long term. That builds to a review load of maybe 100-200 cards a day once the deck matures, taking 15-30 minutes. The temptation to add 50 or 100 new cards a day is a trap: the review backlog compounds a week later and becomes unmanageable, which is the single most common reason people abandon spaced repetition. Slow and daily beats fast and abandoned.

Does spaced repetition work for grammar as well as vocabulary?

It works best for discrete facts with a clear prompt and answer, which vocabulary and set phrases are. Grammar as a system of rules is learned better through exposure and practice than through flashcards, but specific grammar points can be drilled with cards: a card that prompts a verb form, a gender, or a preposition works well. The rule of thumb is that if you can phrase it as a clean question with one right answer, spaced repetition can hold it for you. Rules and intuition come from reading and speaking.