The study, published in the journal Cognition, found that it’s “nearly impossible” for language learners to reach native-level fluency if they start learning a second tongue after age 10 — though that doesn’t seem to be because language skills go downhill at this age. “It turns out you’re still learning fast,” says study co-author Joshua Hartshorne, an assistant professor of psychology at Boston College. “It’s just that you run out of time, because your ability to learn starts dropping at around 17 or 18 years old.” Those who start a few years after age 10 may still become quite good at a language, the paper notes, but are unlikely to reach total fluency.
Why the drop in learning ability happens at the threshold of adulthood is still unclear, Hartshorne says. Possible explanations could include changes in brain plasticity, lifestyle changes related to entering the workforce or college or an unwillingness to learn new things — potentially while looking foolish in the process — that mounts with age.
Though that may seem discouraging — age 10 is far in the past for many hopeful language learners — it was heartening for scientists to learn that the critical period for language acquisition might be considerably longer than they previously thought. Some scientists believed that the brief window closes shortly after birth, while others stretched it only to early adolescence. When compared to those estimates, 17 or 18 — when language learning ability starts to drop off — seems relatively old.
The study used a unique method to reach that new finding. To compile the large and diverse group of people required for a language acquisition study, the researchers created a user-friendly grammar quiz intended to go viral. The 10-minute quiz, called “Which English?,” hooked people by guessing their native language, dialect and home country based on their responses to English grammar questions. At the end of the quiz, people were asked about their actual native language, if and when they had learned any others and where they had lived.
The gimmick worked. The quiz was shared more than 300,000 times on Facebook, hit the front page of Reddit and reached trending status on 4chan. Almost 670,000 people took it, giving the researchers huge amounts of data from native- and non-native English speakers of all ages, some of whom spoke other languages and some of whom didn’t. Analyzing participants’ responses and mistakes allowed them to draw unusually precise conclusions about language learning. (Hartshorne is currently spearheading other online language experiments, which are open to anyone.)
In addition to insights about the critical period, Hartshorne says the quiz results clearly showed that students fared better when they learned a new language by immersion, rather than simply in the classroom. Though he acknowledges it’s easier said than done, “you’d be better off moving to a country as an adult and trying to learn a language than taking it all throughout school.”
If uprooting your life isn’t an option, Hartshorne recommends mimicking an immersive environment as much as possible — that is, finding ways to have actual conversations with native speakers, rather than trying to pick up your skills from books. If you can do so, it’s perfectly possible to become conversationally proficient, if not completely fluent, even as an adult, he says.
That should be encouraging for those well beyond their elementary school years, Hartshorne says. The adult brain seems to be better at learning than researchers previously thought—even if it’s unlikely that you’ll become fluent at a language you learn later in life.
“We’re finding that you don’t start to see lack of plasticity until late adolescence, early adulthood, mid-adulthood,” Hartshorne says. “As a scientist, it’s always fun when there’s undiscovered country, but it also reminds us to be careful about the things that we don’t know. It makes me wonder, what other things do we not know about?”