Most people meet a new word, look it up, feel like they’ve learned it, and never see it again. Two weeks later it’s gone. That’s not a discipline problem. It’s the documented shape of human memory, and once you know the shape you can work with it instead of against it.
Forgetting is a curve, not a cliff
In 1885, Hermann Ebbinghaus spent months drilling himself on nonsense syllables and timing how long they took to relearn. Memory for new material, he found, decays steeply in the first hours and then flattens: most of the forgetting of a freshly learned list happened within the first day, the bulk of it in the first hour. A 2015 replication with modern controls traced nearly the same curve. The decay isn’t a defect; it’s your brain concluding, reasonably, that something met once isn’t worth keeping. You change its mind by meeting the word again, and when matters enormously.
The useful review is the difficult one
Review five minutes after learning and recall is effortless, and teaches almost nothing. Wait three months and you’re relearning from scratch. The productive moment sits between: recall that’s hard but still possible. Researchers call this retrieval practice, and its edge over re-reading is one of the most replicated results in memory science, and it is not subtle. In a Science study on Swahili–English vocabulary pairs, students who kept testing themselves recalled about 80% a week later; those who kept re-studying the same pairs managed about 35%. A meta-analysis covering more than 14,000 learners puts spaced study at 47% recall against 37% for massed. A major review of ten study techniques rated exactly two of them high-utility, practice testing and distributed practice, with re-reading and highlighting at the bottom.
Spaced repetition is those two winners combined and automated: test each word at the moment it’s about to slip.
Why the intervals grow
Each successful hard recall makes the memory more durable, buying a longer gap: a day, three days, a week, a month. A word retrieved after a two-month silence is genuinely yours, and that, not “I’ve seen it 40 times,” is what knowing a word means. Mature words leave the queue, so daily reviews stay roughly flat while your vocabulary compounds: the practice gets cheaper per word over time.
The entire practice, in three habits
Learn words you actually met, because context is retrieval scaffolding. Review a little most days; the schedule is per-word and doesn’t care about your calendar. And let the algorithm decide what comes back: left to ourselves we re-review what we already know, because easy recall feels like progress. In the studies above, the re-readers were the most confident about how much they’d remember. They remembered the least.
References
- Murre, J. & Dros, J. (2015). Replication and Analysis of Ebbinghaus’ Forgetting Curve. PLOS ONE.
- Ebbinghaus, H. (1885/1913). Memory: A Contribution to Experimental Psychology.
- Karpicke, J. & Roediger, H. (2008). The Critical Importance of Retrieval for Learning. Science.
- Roediger, H. & Karpicke, J. (2006). Test-Enhanced Learning. Psychological Science.
- Cepeda, N. et al. (2006). Distributed practice in verbal recall tasks. Psychological Bulletin.
- Dunlosky, J. et al. (2013). Improving Students’ Learning With Effective Learning Techniques. PSPI.