26
Why some learners reach a very high level and others stall for years at the same point
The plateau is real and it is not talent. Comparing people who got past it with people who did not, the differences cluster.
What the ones who got through actually did
- Vast amounts of input at a level slightly above them. Not comprehensible-and-comfortable, but material where roughly one word in twenty is unknown. Comfortable input maintains; slightly hard input builds.
- Output with correction. Speaking a lot without anyone ever correcting you fossilises errors. The people who improved had someone, a tutor or a patient friend, who would interrupt.
- They noticed things deliberately. Not just reading, but stopping to ask why a native chose that word rather than the obvious one. This is the single behaviour that separated people at the same hour count.
- They moved from studying the language to using it for something else. Reading about a hobby, doing work in it, watching things they would watch anyway. That is what produces the hours.
What kept people stuck
- Staying in the material they can already handle, which feels like progress and is maintenance.
- Apps past the early stage. Excellent for the first months and a plateau machine after, because the difficulty stops rising.
- Avoiding the specific things they are bad at, which is universal and invisible.
- Treating vocabulary as the bottleneck when it is usually listening speed.
Two things worth separating
Accent and fluency are different skills with different training, and people conflate them. Someone can be near-native in comprehension and vocabulary with a strong accent, and that is not a plateau, it is a separate skill they did not train.
The honest part: the hour counts involved are large. Most plateaus are a hundreds-of-hours problem being treated as a method problem.
@input_hours · 3w ago · 4 replies
Hours explain most of it and people badly underestimate their own. Three years of study can be two hundred hours or two thousand, and those are different situations entirely.
Try counting a normal week honestly. Most people who feel stuck at three years are at a few hundred hours total, which is early rather than plateaued. That reframe alone changed how I approached it, because the problem stopped being "my method is wrong" and became "I have not done enough yet".
Reply
Report
@two_thousand_hours · 2w ago · 3 replies
I counted mine and it was the most useful thing I did in four years. Rough total across three years of "studying": about 340 hours. I had been telling people I had studied for three years, which is true and completely uninformative.
The people I know who got past the plateau were all at four figures. Not because they are different, because they did more hours. That is a much less interesting explanation than talent and it fits the evidence better.
Reply
Report
@input_hours · 2w ago
Counting also changes behaviour on its own. A week of logging honestly tends to produce more hours the following week.
Reply
Report
@stalled_at_b2 · 2w ago
This is the comment I did not want to read. Four years at the same level and I have never once counted, which probably means I already know what the number is.
Reply
Report