The fastest route to a low level is not a shorter version of the route to a high level.

Almost every argument about how to learn a language is somebody describing the method that happened to work for them. They are usually telling the truth — and that is the problem. Given enough hours, every method works, so every method has people who swear by it.

What the methods actually differ in is how fast you see progress, how high you can eventually get, what they cost you, and how likely you are to quit. Three of those four are about you, not about the language.

Find out which one is yours Read the three methods

Why the advice contradicts itself

You cannot fail at learning a language. That is the source of the confusion, not the cure for it.

Speak from day one, says one camp, and here are the people who did it and now speak Spanish. Never speak until you understand, says another, and here are the people who did that and now speak Japanese. Learn ten words a day. Throw away your flashcards. Study grammar. Grammar is a waste of your life.

Every one of them has students who succeeded, and none of them are lying. The hours produced the result. The method was the thing they happened to be holding while the hours went by.

So “which method works” is a dead question. The live one is narrower and more useful: how good do you want to get, and by when?

The three ways

Pick the one that matches your target, not the one that sounds most impressive.

1

Expression‑first

Target: A1, fast

Speak from the first day. Vocabulary and grammar up front, a tutor early. What almost every school, app and course sells.

What it costs
The most effort per hour. Conversation is tiring in a way watching and reading is not.
Where it stops
Basic level quickly, then a plateau that is expensive to leave.
Most people 2

The combination

Target: B1

Start with listening and reading. Add grammar and writing. Add speaking last, rising in intensity towards your deadline.

What it costs
A few months of low visible progress — short enough that motivation survives it.
Where it stops
Comfortably conversational, on a foundation you can build on later.
3

Input‑first

Target: B2 and above

Months of listening and reading before you produce anything. Output comes late, and then comes quickly.

What it costs
Long stretches where you cannot demonstrate anything, to anyone, including yourself.
Where it stops
Highest ceiling of the three. Wrong choice if you need visible progress to keep going.
Each one in detail

The number everyone quotes

600 hours to learn French — and what that figure actually measures.

The Foreign Service Institute trains US diplomats and publishes how long its students need to reach professional working proficiency. It is the best dataset anyone has, and it is quoted almost everywhere without the conditions attached to it.

CategoryClass hoursLanguages
I600–750French, Spanish, Italian, Portuguese, Norwegian, Swedish, Dutch, Romanian
II~900German, Indonesian, Swahili, Malay
III~1,100Russian, Hindi, Greek, Hebrew, Polish, Turkish, Thai, Finnish, Hungarian, Vietnamese
IV~2,200Arabic, Mandarin, Cantonese, Japanese, Korean

What the number does not say. Those are classroom hours. FSI students also do around 17 hours a week of directed self‑study, which puts the real total for a Category I language closer to 1,000–1,200 hours.

They are also selected, motivated adults studying full‑time, in small groups, with professional native‑speaker instructors, and no job. If you are doing an hour a day around work, the honest expectation is years. That is normal, not failure.

The book

3 Ways to Learn a Language is being written now.

Three methods, what each one costs, where each one stops, and how to work out which is yours. Written for adults teaching themselves, not for classrooms.

Leave your email and I will send it when it is finished.

Your email is stored only for this waiting list. Nothing else, a few emails a year. No tracking on this site — no cookies and no analytics.