Most language apps let you cram 50 words in a session and forget 45 by Thursday. The science of memory says there is a better way, and it starts with forgetting on purpose.
Most language apps let you cram 50 words in a session and forget 45 of them by Thursday. You feel productive in the moment. The progress bar fills up. You earn a badge. Then a week later you cannot remember the Spanish word for "kitchen" and you start to wonder if something is wrong with your brain.
Nothing is wrong with your brain. This is exactly how memory works. The problem is the app, not you. There is a better approach, and it starts with forgetting on purpose.
Ebbinghaus and the forgetting curve
In 1885, a German psychologist named Hermann Ebbinghaus decided to turn himself into a lab rat. He memorised lists of nonsense syllables (things like "DAX", "BUP", "ZOL") and then measured how quickly he forgot them. What he found was remarkably consistent. Within 20 minutes, he had lost roughly 40 per cent of what he had just learned. Within a day, more than 60 per cent was gone. Within a month, the number settled around 80 per cent loss.
That pattern became known as the forgetting curve, and it has been replicated hundreds of times in the 141 years since. The shape barely changes across ages, subjects, or types of material. Your brain discards most new information within hours unless something intervenes.
The speed of forgetting is remarkably consistent: roughly 60 per cent of new material is lost within the first 24 hours without intervention.
Ebbinghaus, H. (1885). Memory: A Contribution to Experimental Psychology.
This is not a flaw. It is a filter. Your brain sees thousands of new inputs every day. Holding onto all of them would be catastrophic. So it keeps what seems important and lets the rest go. The question for language learners is: how do you convince your brain that "cocina" is worth keeping?
Why cramming fails: the spacing effect
The instinct when learning vocabulary is to repeat it many times in one sitting. See the word, say the word, write the word, repeat. This feels effective because you can recall the word five minutes later and you think you have learned it. But you have not. You have activated a short-term memory trace that will fade on the same curve Ebbinghaus documented.
The alternative is called the spacing effect. Instead of repeating a word 10 times today, you review it once today, once tomorrow, once in four days, once in two weeks. The total number of exposures is lower. The long-term retention is dramatically higher.
In 2006, Cepeda, Pashler, Vul, Wixted, and Rohrer published a meta-analysis of 184 separate studies on distributed practice. Across every domain they examined, spacing produced better retention than massing. The effect was not marginal. In many studies, spaced learners retained twice as much material over longer periods compared to learners who studied the same amount in concentrated blocks.
Cepeda et al. (2006) — Meta-analysis of 184 studies in Review of General Psychology, 10(4), 354-380. Found that distributed practice produced superior retention across all conditions tested. Optimal gap between sessions increased as the target retention interval lengthened.
The reason spacing works is counterintuitive. Each time you encounter a word just as you are starting to forget it, the act of retrieving it strengthens the memory trace more than seeing it when it is still fresh. You need to struggle slightly. That struggle is the signal your brain uses to decide this information matters.
Each time you retrieve a word just as you are starting to forget it, the memory trace grows stronger.
From Habla Journal
How spaced repetition algorithms work
Paul Pimsleur was one of the first to formalise this into a system for language learning. In 1967, he proposed what he called "graduated interval recall": review a new item after 5 seconds, then 25 seconds, then 2 minutes, then 10 minutes, then 1 hour, then 5 hours, then 1 day, then 5 days, and so on. Each successful recall pushes the next review further into the future.
Modern spaced repetition systems automate this. They track every item you are learning, record how easily you recalled it, and schedule the next review at the moment the research predicts you are about to forget. Get a word right easily? It comes back in a week. Struggle with it? It comes back tomorrow. Get it wrong entirely? It comes back in the same session.
Absolute spacing enhances learning regardless of the number of intervening items between repetitions.
Karpicke, J.D. & Bauernschmidt, A. (2011). Journal of Experimental Psychology: Learning, Memory, and Cognition, 37(5), 1250-1257.
The algorithm is doing something you could not do manually: tracking hundreds or thousands of vocabulary items, each on its own schedule, and presenting the right word at the right time. You do not choose what to study. The system does. Your job is to show up and answer.
What this looks like in Habla's Review module
Habla's Review module is a spaced repetition flashcard system. Every word and phrase you encounter in other modules (Listen, Talk, Watch) gets fed into your review deck. The algorithm tracks your performance on every item and schedules reviews accordingly.
Words you get wrong come back sooner. Words you recall easily get pushed further out. Over time, the intervals grow. A word you consistently recall correctly might not appear again for three weeks, then six weeks, then three months. That is not neglect. That is the algorithm trusting your long-term memory, which is exactly what the Cepeda meta-analysis identified as optimal.
The practical result is that most of your review time is spent on words you are genuinely at risk of forgetting, not words you already know well. This is far more efficient than flipping through the same stack of flashcards in the same order every evening.
The retention numbers
Researchers typically measure retention at specific intervals: 1 day, 1 week, 1 month, 6 months. The numbers vary by study, but the pattern is consistent. Cramming produces high 1-day retention (often 70 to 80 per cent) that collapses rapidly. Spaced repetition produces slightly lower 1-day retention (60 to 70 per cent) but sustains 80 per cent or higher retention at the 30-day mark and beyond.
Bahrick and colleagues conducted a 56-year longitudinal study on foreign language vocabulary retention (published in Psychological Science, 1993). They found that items reviewed with longer spacing intervals were retained for decades. Not weeks. Decades. The memory traces formed through spaced retrieval were functionally permanent.
This is what we mean by "a memory trick that lasts a lifetime." It is not a metaphor. The research literally shows retention across a human lifespan when spacing is applied correctly.
The forgetting curve is real. It is also solvable. Every modern language learner has access to algorithms that work directly against it, and the evidence for those algorithms is not ambiguous or contested. Spaced repetition is one of the most replicated findings in cognitive psychology.
The only question is whether the tools you are using actually implement it. Most language apps do not. Habla's Review module does, because the research left very little room for argument.