A plateau is a method that has outlived its conditions. Every way of studying catches one class of failure and lets the rest through, and while the thing it catches is what is blocking you, progress is fast and obvious. The moment something it does not catch becomes the binding constraint, the same effort produces nothing, and from the inside that feels exactly like losing the ability to learn.
So the diagnosis is a single question and it is not about you. What can my current method not catch? The answer is usually visible in ten seconds and it names the next method, which is why the correct response to a stall is almost never to work harder at the thing that stopped working.
The six that explain and break it
- the plateau is structural, not a failure of motivation. It happens to nearly everybody at the same point.
- each level costs roughly twice the last, so identical effort produces visibly less progress as you climb.
- comprehension outruns production, and more input widens that gap rather than closing it.
- the fix is harder input and forced output, together. Either alone is what most people try and it is why most people stay.
- get corrected. Without correction errors stabilise, and stabilised errors are what a plateau is made of.
- set a task, not a level. Give a talk, write a report, read a novel, because concrete tasks produce progress that level targets do not.
What each way of studying lets through
An app stops you from never starting and hands you the first thousand words, and everything longer than one of its exercises walks straight past it. A deck stops you forgetting what you have met, and cannot help with anything you never met or with getting a word out in half a second. Volume of material fixes a narrow vocabulary and slow reading, and produces nothing you can say. A weekly hour with a teacher ends the silence and catches the errors nobody was correcting, and does nothing about the other six days.
None of those is a bad method and none of them is finished business. They are layers, each with a list of what it stops and a list of what it never sees, and reading the second list is how you choose what to add. A learner who has been on an app for a year and is stuck has not failed at the app. They have finished with it.
Four camps, and the exit from each
The first arrives around month two and it is the one that ends the most attempts. The free tools were built for month one, where everything is new and progress is visible every single day, and by week six the sentences you meet have outgrown the exercises. Nothing feels like movement for about six weeks. The exit is a change of material rather than a change of effort, and the specific change is to something with whole sentences in it that you can nearly follow.
The second is the wall that gets named more than any other, and its cause is arithmetic. Everything you learned up to that point was frequent, so every hour paid out immediately in things you could suddenly understand. What is left after that is individually rare: each item turns up once a month rather than once a page, so the same hour buys the same amount and almost none of it shows. Nothing has slowed down except the visibility. The exit is to change what you measure, from things learned to situations survived, and to accept a feedback delay of weeks. That is the whole subject of understood everything, can say nothing: the standard trap, and the same wall in one language is why progress stalls at B1, and what moves it.
The third is comprehension running away from production. Input is comfortable and output is not, so a reasonable person keeps choosing input, and after a year they follow films and freeze in a shop. The exit is forced output with a deadline in it, which in practice means another person, since nobody imposes a deadline on themselves at eleven at night.
The fourth is the quietest and it is caused by success. You can do everything the language is currently asked to do. Nothing in your week requires it to get better, so it does not, and the person stays there for a decade. The exit is to change what you need it for, which is a decision about your life rather than about study, and it is why the language people call fossilised is usually just finished.
The response that cannot work
Here is the trap this page exists for. Almost everybody reads a plateau as a motivation problem, because that is what it feels like, and the response is to try harder at the current method. That is the one response guaranteed to fail, since the method is not underperforming. It is doing exactly what it does, and what it does is no longer the thing in the way.
Part of the illusion is measurement. At the beginning, a week of study is visible in a week. Past the second camp, the smallest interval over which improvement is detectable is about three months, and nobody has a memory of their own ability accurate enough to compare across three months. So progress becomes invisible at precisely the point it becomes slow, and the two get confused.
Where people stop
Four camps, and what is blocking the climb at each
The gaps are not equal and the last two are the long ones. Not one of the four is a shortage of motivation.
- 01Month twoThe free tools were built for month one and the sentences have outgrown them
- 02The intermediate wallWhat is left is individually rare, so the same hour shows less
- 03Understood everything, said nothingInput is comfortable, output is not, and a reasonable person picks input
- 04Good enoughNothing in your week now requires the language to improve
The ground rises because the hours behind each visible step really do grow. What the figure cannot show is that most people who give up do it at the first camp, which is also the cheapest one to get past.
Two ways to measure something you cannot feel
Record two minutes of yourself talking about your own week, on the same day of every month, and keep the files. It replaces judging your progress by how the last session felt, which is a measurement of the session. Free, on the phone in your pocket, and the value arrives late: the recording from three months ago is the only object in this subject that can prove a plateau was not one.
The second is a fixed benchmark. Choose two minutes of difficult native audio, keep it, and come back to exactly it every three months. What you catch this time against what you caught last time is a real comparison, because the material is identical. It replaced taking another placement test, which is a slightly different test measuring a slightly different thing each time and cannot detect a small change at all. Free, and the only rule is not to study the clip.
The mistake: starting again
The stall arrives, and the response that feels best is a fresh start: a new app, a different textbook, back to the first chapter, a clean deck. It is enormously comfortable, because the first chapters are the material you already know, and for about six weeks everything is easy again and feels like progress.
Then the same wall arrives at the same place, because it was never about the resource. Restarting has a real cost that is easy to miss: it spends the months in which the actual blockage would have had to be faced, and it can be done repeatedly for years. If you have started three times, the honest reading is that you are not a beginner three times over, you have hit the second camp three times. The thing that does move is the one that was uncomfortable, and it is generally the one with another person in it, which is finding someone to speak to, and what to do in the hour.
The stall nobody plans for
The plateaus above all happen while you are working. There is one more that happens when you stop, and it is not the same shape: a language that is not used fades in a particular order, with production going first and comprehension lasting years, and the amount needed to hold it is far smaller than the amount needed to build it. That is what fades first, and how little it takes to hold on, and it is worth reading before a break rather than after one. If the camp you are at is the second one, the mechanism underneath the slowdown is in why you forget, and the schedule that beats forgetting, and the material change that gets you out of the first one is filling your day with the language without moving country.