You've probably heard some version of the 10,000-hour rule: put in that many hours at anything, and mastery follows. It's a clean number and an easy to follow story, but hours only tell part of it. The psychologist whose research it's built on has spent years publicly correcting it. What his actual research found is less catchy and considerably more useful, because it doesn't say how many hours you need. It says what has to be true about the hours you spend, and most self-directed language practice doesn't deliver

The study behind the myth, and what it actually found

The 10,000-hour rule traces back to a 1993 study by psychologist Anders Ericsson and colleagues, which compared violin students at a Berlin music academy across three skill tiers: an elite "best" group on track to become soloists, a "good" group headed for orchestral careers, and a group headed toward music education. The researchers asked each student to estimate the solo practice hours they'd accumulated since starting the instrument.

What the violin study actually measured: By age 18, the best group had accumulated roughly 7,410 hours of solo practice on average, the good group about 5,301 hours, and the music-education group about 3,420 hours. By age 20, the best group's average reached close to 10,000, which is where the popular version of the number comes from. Half of that top group had not yet reached 10,000 hours at age 20. It was never a threshold every top performer had crossed. It was an average, with real variation between high performers.

Some key nuance got lost in translation as the 10,000 hour rule got pop science visibility. The researchers behind the study have said plainly, in their own words, that theirs became "a popularized but simplistic view," suggesting that anyone who logs enough hours in any domain automatically becomes an expert. That was never the finding. Among students already admitted to a competitive conservatory, the ones who improved fastest had spent their hours differently, not just more of them.

The real variable: what kind of practice, not how much

The finding that actually explains the gap between the three violin groups has four parts. Practice reliably produces improvement when it has a well-defined goal, when the practitioner is motivated to improve rather than just to perform, when feedback arrives promptly enough to act on, and when there's room for repetition and gradual refinement. A fifth condition, elaborated in later work, matters just as much: this kind of practice typically needs an expert who designs the training and delivers the feedback, since it's hard to accurately diagnose your own weak points. Solo, goal-directed practice without that outside expert, think targeted vocabulary drilling or focused pronunciation work you design for yourself, is a real and valuable category that gets close.

Why hours alone stop working: the automaticity trap

Ericsson's research draws a sharp line between mere experience, doing a task repeatedly until it becomes automatic, and deliberate practice, which actively works against that automaticity. Once a skill reaches the point of comfortable, effortless execution, additional repetition of it stops producing improvement. The performance has plateaued; the experience keeps accumulating anyway. Automaticity is a very important skill in language learning, but it can also hold us back from pushing past simple patterns we default to as beginners and intermediate learners.

The tennis analogy from that research is helpful: a player will not improve their backhand by playing more matches, because a match rewards consistent execution, not targeted correction of a specific flaw. A coach who feeds increasingly difficult, unpredictable shots at that one weakness, with feedback after each attempt, is the thing that actually changes the backhand. In recreational skills generally, people tend to hit a comfortable, "good enough" plateau after well under a hundred hours, at which point they stop being able to consciously adjust the specific things they're doing.

Sound familiar? Years of conversations in your target language, real ones, with real people, and the same handful of errors keep showing up. That's not a contradiction. Having a conversation is closer to a tennis match than a coaching session: it rewards keeping the exchange moving, not stopping to isolate and fix the exact thing that's wrong. The practice and the performance can both be valuable. They are not the same activity.

What deliberate practice looks like specifically in language learning

A 2025 framework paper by researchers Wedlock and Binnie applies this distinction directly to second-language learning, and draws the same line Ericsson draws elsewhere: deliberate practice is individualized training, designed by an expert, targeting a specific weakness, with immediate feedback attached. Purposeful practice is the solo version, structured and goal-directed but without that expert feedback loop. Both are more effective than unstructured exposure. Only the first counts as deliberate practice by the strict definition, and the framework identifies specific activities, isolated pronunciation drilling, targeted grammar production, retelling, reading aloud, shadowing, and dictation, as the kinds of practice that meet the bar when feedback is attached to them.

The everyday version of the naive-practice trap in language learning looks like this: a learner has conversations in their target language for years, without anyone flagging the specific errors they keep making. They're accumulating real, valuable experience. Their communicative strategies get better at working around gaps rather than closing them, and the underlying errors have every opportunity to become permanent habits. This is the same mechanism as the tennis player who's played thousands of matches and never fixed their backhand.

The supporting evidence, briefly

A few adjacent findings back this up from different angles, without needing to become the center of the argument.

d ≈ 0.64 the medium-to-large effect size a major meta-analysis found for corrective feedback on second-language accuracy, larger still in foreign-language classroom settings than in immersion

Large effects the size of gains a meta-analysis of nearly 50 studies found for explicit, focused instruction over implicit exposure

Consistently faster how deliberately learning vocabulary, rather than picking it up incidentally, compares across every study that has directly tested the two

In addition to how language learners practice, how much they practice is also very important for their progress. Research on the spacing effect finds that distributed review consistently beats cramming for long-term retention, and research on retrieval practice finds that actively producing an answer builds more durable memory than passively re-reading or re-listening to it, even when the passive version feels more productive in the moment. Psychologist Robert Bjork's concept of "desirable difficulties" names why: conditions that feel harder while you're in them, spacing sessions out, mixing topics instead of blocking them, varying who you speak to and what register you use, produce more durable learning than the easier-feeling alternative. The one caveat Bjork is careful to add: a desirable difficulty only helps if you can actually overcome it. Push a learner past what they're currently capable of, and the difficulty stops being desirable and starts being discouraging, which is exactly why this kind of practice needs to stay calibrated to the learner's real level.

What this means for how you practice

None of this argues against input, conversation, or the hours you've already put in. It argues for spending some of your practice time differently.

1. Set a specific, narrow goal. Not "get better at French," but "stop dropping the subjunctive after expressions of doubt this week."

2. Push slightly past what's comfortable. If an exercise feels easy, it's probably mere experience. Deliberate practice should feel like effort.

3. Get feedback specific enough to act on. Not just what was wrong, but the rule behind why, so the correction transfers to the next sentence.

4. Repeat on purpose, with correction attached each time, rather than moving on the moment something is technically understandable.

5. Keep your conversations and input too, and recognize them for what they are: valuable, but a different activity from the targeted work that closes a specific gap.

Dioma is built around supplying the piece most self-directed learners can't generate alone: the expert-designed targeting and immediate feedback loop that turns practice into the deliberate kind. Every exercise is built against a specific point in an expert-designed curriculum, and every correction names the rule you missed rather than just marking the sentence wrong, so the effort you're already putting in has somewhere specific to land.

The real shift

The 10,000-hour version of this story asks how much time you have to give. The actual research asks a better question: is the time you're spending structured around a specific goal, pushed to the edge of what you can currently do, and followed by feedback precise enough to fix the exact thing that's wrong. Hours you've already logged were not wasted. They built the comfort and vocabulary that make focused, corrected practice possible now. What moves you forward from here is turning some of that experience into practice that's actually deliberate.

Frequently asked questions

Is the 10,000-hour rule true for learning a language?

Not as it's usually stated. The original violin research found an average, not a universal threshold, and half of the top performers in that study hadn't reached 10,000 hours by the point the popular version claims they should have mastered their instrument. The more useful finding from the same research is about the kind of practice that produces improvement, not a specific hour count.

What's the difference between deliberate practice and just practicing a language?

Deliberate practice has a specific goal, pushes just past your current ability, and comes with feedback precise enough to correct the exact thing you're getting wrong. General practice, like having conversations or consuming content, is valuable but tends to reward keeping communication going rather than fixing a specific weakness.

Why haven't I improved even though I practice my target language constantly?

Once a skill becomes comfortable enough to execute automatically, more repetition of it stops producing improvement on its own. If your practice has mostly been comfortable rather than targeted, and hasn't included specific feedback on your actual errors, you may be accumulating experience without accumulating new skill.

Do I need a teacher for deliberate practice, or can I do it alone?

Solo, structured practice with a clear goal, sometimes called purposeful practice, is genuinely valuable and better than unstructured exposure. The research suggests the fastest improvement comes with an expert providing the targeting and feedback, since accurately diagnosing your own blind spots is difficult to do for yourself.