Adults learn languages faster than children — for the first year

"Children are sponges. Adults are too old." It is the most common reason people give for not starting, and on the timescale that anyone actually plans around, it is backwards.

The Dutch study

Snow and Hoefnagel-Höhle's 1978 study is the one everything else is measured against. They followed English speakers of several ages who had moved to the Netherlands and were learning Dutch in the same environment.

Over the first months, the adolescents and adults progressed fastest. The youngest children were slowest. By the end of the first year the 8–10 and 12–15 groups had the best control of the language.

That result has held up across the literature since: given comparable exposure, older learners acquire grammar and vocabulary faster in the early stages. The usual explanation is not mysterious — adults deploy conscious strategy, can be told a rule and apply it, and transfer what they already know about how a language works.

Where the folk belief comes from

It is not invented. Younger starters do better in the long run, and there is a real effect concentrated at the top end — accent especially, and ultimate attainment generally. The critical-period question is still being studied at scale; Hartshorne, Tenenbaum and Pinker's 2018 analysis of two-thirds of a million English speakers is the largest attempt at it.

But notice which claim survives and which does not:

• "Children reach a higher ceiling, especially in accent" — supported.
• "Children learn faster" — not supported for the early stages, which is the part any adult learner is deciding about.

The confusion comes from watching a child in an immersive environment for three years and an adult in a classroom for three months, and attributing the difference to the age rather than to the three years.

Why this matters for planning

The FSI figures that everything on this site rests on are measured on adults — screened, motivated, full-time adult students. Six hundred hours for Spanish is an adult figure. There is no separate, worse adult column, because there was never an adult penalty in the data to begin with.

If you are 40 and wondering whether the numbers apply to you, they were derived from people like you.

What is actually different about being an adult

Not the machinery. The circumstances, and they cut both ways.

Against you: less time, competing obligations, and an intolerance for feeling incompetent in public that children mostly lack. The last one is real and underrated — the willingness to say things badly is the single biggest predictor of who gets past B1, and it is a disposition rather than a capacity.

For you: you can be told a rule instead of inferring it over years. You know what a case is, what an aspect is, what a subjunctive is for. You can choose your method deliberately, and method choice dominates almost everything else — the gap between tutored and app-only study is a factor of three and a ceiling, which is far larger than any age effect anyone has reported.

And a ceiling you may not care about. Most of the well-attested age advantage sits at native-like attainment and accent. If your target is B2 — work, study, argue a point — you are not operating near the region where the effect is strongest.

The honest summary

Start now, because the alternative is starting later, and the research says nothing that makes later better.

What it does say is that you will move fastest at the beginning, which is worth knowing for the opposite reason: the early speed is not a preview. Each CEFR level costs about as much as everything before it, and the point where you stop feeling fast is the point where most people quit — usually while blaming their age for something that is happening to everybody.