The published hour figures are the numerator and they are not where people go wrong. The divisor is: somebody takes a number, divides it by the hours they intend to study, and arrives at a date built on a quantity that has never been measured.
Six hundred hours at five hours a week is a little over two years. At ten it is fourteen months. At twenty it is seven months. Those three answers describe the same language and the same learner, and the only variable is the number nobody checks.
Eight numbers, and what they assume
- 600 to 750 hours for a close language to professional working proficiency, for an English speaker learning Spanish, French or Dutch.
- 2200 hours for a distant one, for Arabic, Chinese, Japanese or Korean, which is roughly a factor of three or four.
- those are classroom hours plus study, from intensive full time programmes, and they are not hours of casual app use.
- the estimates assume good teaching and motivated adult learners studying full time, which is not most people's situation.
- your starting language changes everything. The same figures for a Persian speaker learning Arabic, or a Korean speaker learning Japanese, are far lower.
- each level costs roughly twice the last, so the hours are not spread evenly across A1 to C1.
- 20 minutes a day is 120 hours a year, which reaches a usable A2 in a close language and is a realistic long term pace.
- an hour a day is 365 hours a year, which puts B2 in a close language within about two years.
Three adjustments, in this order
First, find your real weekly figure by measuring rather than estimating, over a fortnight. Everybody over-reports, in the same direction, by a lot.
Second, discount for density. A published hour is usually a classroom hour with a teacher and other students, which is a denser hour than one spent alone with an app on a train. That does not make solo hours worthless; it means the exchange rate is not one to one and pretending it is produces a date that arrives and passes.
Third, adjust for what you already speak. A language closely related to one you have takes a substantial slice off, and a language unrelated to anything you know does not, and this adjustment is larger than any study technique you will ever adopt.
Where the published numbers come from
The Foreign Service Institute in the United States has been teaching languages to adults full time for decades and publishes the hours each of its language groups typically takes to reach a working professional level. Those figures are free to read and they are the source behind almost every number quoted second-hand on the subject.
They replaced a figure from a forum with no source attached, which is what most people are working from and which is usually somebody's own experience generalised.
The limitation is in who was measured, and it is severe. These are full-time adult learners, selected, taught in small groups by professionals, with a job depending on the outcome, and they are native English speakers. If you are none of those things, the figure is a floor rather than an estimate, and the gap between a floor and your actual number is the three adjustments above.
Where it stops: dividing by the intended number
The stall is a plan built on a divisor that was chosen rather than observed. Somebody intends ten hours a week, divides by ten, and gets a date. The real figure turns out to be four, the date passes with a third of the hours done, and the conclusion is that the published number was wrong or that they are slow.
Neither is true and the arithmetic was fine. What was invented was the denominator, and the failure surfaces a year later when it is too late to be useful.
Measuring for a fortnight fixes it permanently and almost nobody does it, because the result is unflattering and because intending ten hours feels like nearly doing them.
What it looks like when it goes wrong
The recognisable failure is an exam booked from a calculation. The date was set by dividing a total by an imagined weekly figure, the preparation ran at the real one, and the fee is spent finding that out.
The second is comparing yourself to somebody whose divisor was completely different. A person who reached a level in eight months was doing twenty hours a week, and hearing the eight months without the twenty is how a reasonable schedule starts to look like failure.
How to find the hours in the first place, and why a plan is a subtraction from a week that is already full, is turning a vague intention into hours on a calendar, and where the biggest single adjustment to the numerator comes from is difficulty, usefulness and the honest reasons people pick.
The open question is what the hours should contain, since the same number spent entirely on input and entirely on retrieval produce very different people at the end of it: seven days built around the time you actually have.