You CAN Learn a New Language After Age 10! Response to Steven Pinker’s Fascinating “Critical Period” Research (and Bogus News Reports)
Up until what age can you become very fluent in a language? It’s a good question, and it’s one I’m going to answer, looking at some new research from linguist and psychologist Steven Pinker.
First, I’d like to bust a few myths about the “critical window” for language learning that have popped up in the news following the publication of the research.
The study itself is really interesting (though I don’t buy into all of it). What I find frustrating is the way it’s been presented in the media. Attention-grabbing headlines aren’t always the most truthful.
I’ve done a ton of research myself about how people learn. I also have plenty of my own experience learning languages as an adult, and I’ve seen the results of thousands of learners from the Fi3M community, so I can give a pretty informed perspective about what adult learners are capable of.
But I’m getting ahead of myself. Let’s start at the beginning.
What’s all this “Critical Period” Fuss About?
I was inspired to write this article after seeing this news report making the rounds on the internet: “Becoming fluent in another language as an adult might be impossible.” The first sentence reads: “If you haven't started learning a new language by the age of 10, you have no chance of achieving fluency.”
Wow. That must be one interesting study that has conclusively proven that it's impossible to learn a language to fluency after the age of 10!
Obviously, some of these results, and the way they’ve been reported, are misleading. That's what I want to address here. The research paper that all of the fuss has been about has to be purchased if you want to read the entire thing. I recommend you take a look at it if you’re interested in understanding the results for yourself.
I do want to emphasise that there were several interesting scientific results that have come out of this study, and there are many parts that I agree with. But there are also a couple of conclusions that I wouldn't be so confident about stating with certainty.
The main problem, however, is how the media are interpreting these results. I’ve got plenty of evidence, which I’ll lay out below, that shows that this interpretation is utter nonsense.
First of all, even if everything stated in the study is true, it's impossible to draw the conclusions that you'll see in a lot of the news reports. I'm going to try to explain how this happened, while giving you some encouragement about your language learning goals. Because despite what the media says, it’s not hopeless. Not by a long shot!
How I (Unwittingly) Contributed to the Study
This study is titled “A critical period for second language acquisition: Evidence from two-thirds of a million English speakers”. Believe it or not, I unknowingly contributed to this study back in 2014, by encouraging those on my email list to try out this language game, which became the basis for the study. My email list was around 100,000 strong at the time, so I probably sent quite a few readers over to participate!
This language game had you answer some English questions, and then it tried to guess which dialect of English you speak, if you were a native speaker. If you weren’t a native speaker, it would try to guess your native language. It was presented as a game, and shared a lot online, and eventually over 600,000 people participated.
“Start Before Age 10 to Reach Fluency” — It’s Not the Truth, it’s Click-Bait
I’ve read the research study myself, and there are several results that I actually find quite encouraging. More on that in a minute. Before we get into the results themselves, let’s see how journalists interpreted them — and by “interpreted” I mean the angle they used to create headlines. Remember, headlines are written to grab attention, to press your emotional buttons, and to be “click-bait” so you’ll read the article.
In the Guardian article I mentioned earlier, the writer begins by saying that even though fluency is impossible after the age of 10, he's going try anyway. But then the strategy he opts for is 40 minutes per day learning French and Japanese on Duolingo (20 minutes per language). That's it. That's the only thing he did for eight months. Not surprisingly, he is not very impressed with his progress.
I’m sorry, but you simply can’t reach any level of fluency by restricting yourself to a single resource that teaches you a language in one, quite narrow, way. You need a lot more varied experience in the language.
As for the statement that “if you haven’t started a new language by the age of 10, you have no chance of achieving fluency,” well, let’s go back to the actual research study.
The TRUTH: Here’s What the Results Actually Say
First of all, the words “fluency” and “fluent” never even appear in the original study. Not only that, but the researchers actually don't make any claim whatsoever about a “biological ceiling”. So the phrasing in these news articles, saying it's impossible to get fluent in a language you start learning after the age of 10…that’s just poor journalism, in my opinion. They’re implying that there is something that biologically prevents us from being able to learn a language after age 10, when the research says nothing of the sort.
Here’s what the research actually says about the term “critical period” (bold emphasis is mine): “We use the term critical period as a theory-neutral descriptor of diminished achievement by adult learners, whatever its cause.”
In other words, the researchers used the term “critical period” in their paper to simply give a name to that age window beyond which people don’t seem to learn languages as well – but they make no claims that adults are poorer learners because they’re adults, OR that adults are incapable of becoming better learners.
This is important. The research study doesn’t state why people might change their language learning abilities as they age. So there’s still no proof that there’s a biological or other insurmountable reason for why people tend to get worse at learning languages as they age.
According to the study, “The critical period cannot be attributed to neuronal death or syntactic pruning in the first years of life, nor to hormonal changes surrounding adrenarche [pre-puberty] or puberty.”
To sum this up: there’s no basis whatsoever for the big scary headlines saying that no one beyond the age of 10 can ever hope to learn a language to fluency.
Add to that the fact that the research study identifies 17-18 as the age period beyond which language learning abilities start to decline – not 10 – and you start to really appreciate just how misleading all these news headlines have been.
So where did the number 10 come from? Basically, what the researchers observed is that around age 17-18, people’s language abilities would start to decline. And since reaching native-like fluency can take quite a few years, they conclude that you need to start by age 10 in order to have enough time to reach fluency before that decline.
From the study: “Both traditional, ultimate attainment analyses, and permutation analyses indicated that learners must start by 10 to 12 years of age to reach native level of proficiency. Those who begin later literally run out of time before the sharp drop in learning rate around 17 to 18 years of age.”
But a drop, even a “sharp drop”, in learning rate, is not an instant drop down to zero! So even if you start learning after age 10, all that means is you have a lower chance of reaching a native skill level. Not zero chance.
And most language learners aren’t aiming for native-like skills, anyway. So whatever the conclusions say about reaching a native level, they say nothing about reaching a very high skill level, even mastery, in the language, well beyond that “decline” that occurs at age 17-18.
Adult Learners ARE Different — But Not for the Reasons You Might Expect
Why would there be a decline at all, though? Well, the study didn’t try to find any conclusive reasons for that. They did speculate a little, and I also have my own opinions formed from years of observing adult learners. So let’s look at the possibilities.
According to the study, the critical age identified – 17-18 years old – “coincides with a number of social changes, any one of which could diminish one's ability, opportunity or willingness to learn a language. In many cultures, this age marks the transition to the workforce, or to professional education which may diminish opportunities to learn.”
They only put forth social reasons as a possible reason for declining language learning ability, not biological ones.
I definitely agree with their suggestion that adults simply have fewer opportunities to learn. Language learning just isn’t mandatory beyond a certain age. The research study focused exclusively on English learning, and in many countries, English is a mandatory subject in school. Kids are forced to study it, often up until the age of about 17-18 (do those numbers sound familiar?). After that, they're not forced to learn it, and they have other priorities in life. So fewer people will try to learn English beyond that age. It’s just simple math. When more people are forced to study a language, more people will progress in the language. That makes sense to me.
Consider this: If it was mandatory for 25 to 35-year-olds to study English, then 25 to 35-year-olds would be the ones who learn the most English. But going that extra step to conclude that they’re naturally the best learners would be absurd. And yet, that’s what the newspapers have done with their conclusions about children being naturally the best learners.
The Quiz was Seriously Flawed
When I was researching this article, I looked at the questions that they asked participants in the study. There were some pretty weird ones! Some of them would be confusing even to native English speakers.
The level of grammar they test in some of these questions doesn’t really say anything about your fluency in the language. As I said earlier, fluency isn’t even mentioned in the study. The study really seems to measure grammar ability for a few types of grammatical construct.
In fact, some of the questions were phrased strangely enough that not even native English speakers got them right. Older native speakers tended to do better than younger ones. Generally, native-speaker participants aged 30 and up did the best. So, if someone can live exclusively in English from birth to nearly age 30, and still not get all the answers right, then the language game isn’t really a great way to test if someone has a native level in English.
Don’t get me wrong. This was still a fascinating study, and it produced a lot of results that I found very interesting – from a grammatical point of view. But a test of someone’s true level in a language needs to extend beyond a narrow set of grammar questions and focus instead on what you can do in the language.
What I Consider to be Fluency in a Language
If someone says they’re fluent in a language, I would expect them to have reached level B2 in the Common European Framework of Reference (CEFR) (click the image to enlarge).
A level of B2 is all that you need to use a language fluently. If you can interact with people the same way that a native speaker could in most everyday situations, then you can be considered fluent.
Now suppose you go beyond B2, past C1, and all the way to C2, the highest level you can reach in the CEFR. If you reach this level, then you’ve mastered the language. Would that make you indistinguishable from a native speaker?
No, it wouldn’t. There is nothing in the CEFR that says you need to sound exactly like a native speaker once you reach the most advanced level. You just need to be able to do all the things that native speakers can do in all situations (not just everyday interactions, as in B2 level).
Unfortunately, what many news reports have done is take the reasonable suggestion that sounding exactly like a native speaker 100% of the time is difficult (if not impossible) for adult learners, and changed the message to convince readers that it’s impossible to reach any useful level of fluency beyond the cut-off age. Now, speaking as someone who has reached C2 level in Spanish, which I only began studying at age 21, I’m proof that the idea of a 10-year-old “cut-off age” as defined by these headlines is downright false.
You can learn a language to a level better than you'll ever need it, no matter how old you are when you start learning. As for reaching the point where you’ll be mistaken for a native speaker, I’d have to say it’s unlikely, but again, not impossible. There may be a “cut-off” point in your life right now where you can’t realistically invest the time you need. But there are adult learners who do succeed in that. So it clearly can’t be impossible.
Is There a Cut-off Age for Certain Parts of the Language (Accent, Pronunciation, Grammar or Hearing New Sounds)?
This is an interesting question. In this study, they’re using the concept of critical age in a very narrow context: grammatical accuracy. The conclusions have nothing to do with how many things you can do in the language, or your accent, or the mistakes you make, or whether you'd get a job if you were being interviewed in another language. It’s all about grammatical accuracy – and grammatical accuracy to quite a ridiculous degree, too, as I showed earlier.
But to answer the question, personally I don't think there is a cut-off age. This whole concept of a cut-off age only has weight in a few very specific situations.
As I mentioned earlier, it’s common sense that children tend to learn languages more often than adults. Adults just don't invest the amount of time that children invest in language learning, because it's not mandatory for adults. That's a very big, crucial difference when you're comparing how children and adults learn languages. What matters isn't whether they're children or adults, it's what they're actually doing with their time.
For instance, a child in an immersive environment may play in the language, have friends in the language and go to school in the language. Adults don’t generally have the chance to live in such a total immersion environment. A child not in an immersion environment, simply studying the language at school, will still study it for several hours a week, whereas an adult learner may just use a language app for 10 minutes a day. In these cases, age isn’t a factor; one of them just has a better method than the other.
If you gave a child a smartphone and said, “Use this app for 10 minutes a day,” then 10 years later, they're not going to be fluent in the language, are they? But the actual things children tend to do in the language make their strategies better than adults' strategies. And this doesn't mean adults can't use those strategies too. As adults, we can take inspiration from how children learn languages.
Then there’s the question of confidence. I think children tend not to be afraid of making mistakes. A child is more likely to have a perfect accent, because they're OK with getting corrected, and they're OK with saying things wrong many, many times and having native speakers push them in the right direction. Adults tend to be more fearful than children are of trying to use a language.
Adults can take inspiration from that. An adult who has a very, very good strategy can potentially improve their accent, pronunciation and other parts of the language to a native degree.
You Can Pass as a Native Speaker — Even as an Adult
While it’s demonstrably possible to master a language to native-like ability even if you started learning it as an adult, I think it’s impossible to convince native speakers 100% of the time that you’re a native speaker.
Some of the reasons for this include the sheer difficulty of investing the necessary amount of time when you’re an adult (which I mentioned earlier), as well as all the cultural aspects outside of the language that you would have to master (which I’ll touch on a bit later).
However! It’s definitely possible to pass yourself off as a native speaker some of the time.
I did this myself once. I challenged myself in Brazil to learn to speak like a Carioca, a resident of Rio de Janeiro. I never reached the point where I could always convince everyone I was a native speaker. But for short periods, I was able to convince multiple Brazilian people that I was Brazilian. I always slipped up eventually, either with a cultural mistake or a language mistake, but until that point, I was successfully passing myself off as a native speaker.
I’ve also met polyglots around the world who have convinced me, for a time, that they were native speakers. The accent is there, even a lot of the cultural references are there, but then there’s a slip-up that tips me off that they’re not a native speaker. If you weren’t born into that culture, it’s going to happen eventually.
But you know something? This isn’t that big a deal. And you shouldn’t lose any sleep over not being a perfect replica of a native speaker. Because…
Do You Really Need to Pass Yourself Off as a Native Speaker? (You’re Not James Bond!)
This is a question overlooked by a lot of language learners. If you can reach B2 fluency in a language, or even C1-C2 mastery, why would you need to go that extra mile to try to pass yourself off as a native speaker?
Passing yourself off as a native speaker goes beyond a perfect accent and grammar. You also need to have all the cultural knowledge of that country, too. All the subtle ways people interact with each other, all the little catch phrases and cultural references that were popular there years ago, etc.
For example, I grew up watching Teenage Mutant Hero Turtles in Ireland, whereas Americans my age grew up with the Teenage Mutant Ninja Turtles. You could master an Irish accent and convince me beyond any doubt that you’ve never left Ireland…until you accidentally mention watching the Ninja Turtles as a child, which would instantly tip me off that you aren’t from Ireland.
So if you want to truly pass yourself off as a native speaker, not only do you need the perfect accent, grammar, slang and everything, but you would also have to spend months watching all the films, TV shows and even memorable commercials that were popular in that country when you were younger. That’s taking fluency to absurd levels, if you ask me. I think it’s very fair if you’re not willing to go to those lengths.
Unless you’re a secret agent like Jason Bourne or James Bond, where you need to blend into the country, infiltrate the bad guys, save the world and get the girl, then do you really need to take fluency to such lengths?
I doubt it. If you can get to a level of fluency where you can achieve anything that a native speaker can, why would you need more?
Unicorns! The Mythical “Language Gene”
Let’s go back to the research study and its findings that people older than a certain age become more limited in what they achieve in language learning. The study only reported on what the participants did achieve, not what they’re capable of achieving. Considering that many adult learners in the study did achieve native-like proficiency, you should never take anyone – like the news articles – saying “Here's what your limitations are,” and believe it, unless it's an actual, measurable biological limitation.
Biologically, I have a limitation that I can't fly, because I don't have wings built into my back. That's just the way it is. But you can’t find a hard-and-fast biological rule for, say, being incapable of learning chess because you’re “too old”. What is that based on? You could try to invent some possible reasons, but it’s just hearsay. There hasn't been any good, solid reason given for why people can't learn languages (or chess!). And to be honest, there hasn't been enough research into successful language learners to see what makes them different. I think that would be a lot more interesting, because there are lots of adult language learners who fail, but until that mythical “language gene” is discovered, you can’t use their experience to deduce that these are strict limitations dictated by your biology.
The “Average Adult” in the Study Doesn’t Represent You!
Regardless of what can or can’t be proven when it comes to our limitations in learning languages as we get older, just remember that the results are always averaged out over all participants.
Two-thirds of a million people from all walks of life participated in this study. Since the average adult doesn’t spend any time in language learning, it’s safe to say that the average adult in this study also doesn’t spend any time in language learning. (Remember that the study was open to people who had studied English at any point in their life, even just for a year or two as a child in school.)
Most adults don’t have any motivation to continue with language studies after they stop being mandatory in school. So they give it up, and their skills decline, and this is reflected in their results on this language quiz.
But then there’s you.
If you’re reading this, you’re probably very interested in learning other languages. And if you’re also an adult, that makes you different to most other adults. You’re part of a very distinct portion of the population that are passionate language learners, and passionate language learners tend to do things differently.
If this study had taken passionate language learners and compared them to child learners, then things would start to get a lot more interesting. The gap in ability between adults and children would narrow quite a bit.
This is already supported by other studies, such as this one, which says (bold emphasis is mine): “Studies that compare children and adults exposed to comparable material in the lab or during the initial months of an immersion program show that adults perform better, not worse, than children (Huang, 2015; Krashen, Long, & Scarcella, 1979; Snow & Hoefnagel-Höhle, 1978), perhaps because they deploy conscious strategies and transfer what they know about their first language.”
The Importance of Body Language, Personality and Other Social Cues in Passing Yourself Off as a Local (My Experiences in Spain and Egypt)
I’m going to switch gears here a little. I mentioned earlier that a perfect accent isn’t the only thing you need if you want to pass yourself off as a native speaker. You need to understand all the pop culture references as well, which is nearly impossible if you didn’t grow up in that country.
But there are other things you can do fairly easily, which won’t make you pass as a native speaker all the time, but which will definitely make you fit in better with locals so they feel more at ease around you.
The main things are body language, personality and dress. These have nothing to do with language, but they shouldn’t be overlooked when you’re travelling abroad.
When I was in Spain, I met a man from Sweden. When he spoke, I swear he sounded like a Spanish radio announcer. His Spanish accent was absolutely perfect. If I closed my eyes, I could be convinced I was talking to a native Spaniard.
I was still studying for my Spanish C2 at the time, and this guy’s level practically put mine to shame. And yet, many Spanish people actually told me that I seemed more Spanish than he did. It's in large part because of my body language. This guy kept more distance between himself and others, the way he made eye contact was a bit unusual, and he just generally put people off. No matter how good his Spanish was, his behaviour made locals treat him differently.
So if you try to talk with native speakers in your target language, and they always reply in English, don’t automatically assume it’s your skill level in the language. Take a look at your personality and body language, and ask yourself if these are playing a role.
Your clothing and outward appearance also play a role. When I first went to Egypt, I wasn't planning to make anyone believe I was an Egyptian, but I did want to practise my Arabic. Those first few days, whenever I would try to speak to someone in Arabic, even if I didn’t stumble or make mistakes, they would cut me off and say, “It's OK. We can speak in English.”
My initial instinct was to think, “OK. Maybe my accent is so bad that those first few words are giving me away,” but actually it's because I just completely and utterly did not look Egyptian.
To solve this problem, I actually sat down in a café with a pen and paper, and I noted everything I could see that was different about Egyptian men around my age, and I noticed I was doing everything wrong. The way I was walking, my facial hair, my shorts and t-shirt (instead of heavier clothes which are the norm), my cap, even my shoes! Everything about my screamed tourist!
So I made a few changes. And right away, people became a lot more inclined to speak Arabic with me, even though I had a strong accent and made quite a few mistakes as I spoke. How I presented myself made me seem more Egyptian, and so speaking with me in Arabic just felt more natural for locals.
In some of my language missions, I would watch TV not to hear what they're saying, but to see their facial expressions, hand gestures, clothing, personal grooming, and other behaviour like how much distance they keep between one another. And I would try to mimic that. And that made all the difference.
”Years of Experience” is a Poorly-Defined Measure of Length of Time Studying a Language
Scott Chacon published a very informative article on Medium along the same lines of what I'm trying to do here, to show you that it’s not hopeless like the headlines are saying. He thoroughly analysed the study’s data (which were released in full by the researchers) and noticed some very interesting findings.
What he found was that there were thousands of participants in the study who outperformed native speakers on the language quiz, even though they began learning after the age of 20! That’s the opposite of the conclusions reported by the press.
Keep in mind that the study’s definition of “native level” is incomplete since it’s only based on those very specific, pedantic grammar questions. But as outlandish as many of the questions were, there were still plenty of adult learners who answered them just fine.
When you take this language quiz, there are some survey questions afterwards about your native language and language-learning history. But only one question is related to how long you’ve been studying English. It asks “When did you start studying English?” and you can answer “from birth” or input the age you were. You also input your current age, and it uses those two numbers to calculate “how long” you’ve been learning the language.
This is not a precise measure at all of how long a person has been studying another language. Think of all the scenarios it fails to consider! For example, suppose you’re 30, and you studied English from age 10 to 12. The survey will wrongly think you’ve been learning for 20 years (since you started at age 10 and are now 30). It doesn’t ask if you ever stopped studying. So it can’t get a true measure of the number of years you’ve been studying.
But even if it did ask you if you ever quit studying, the calculation of your experience would still be inaccurate. “Years of experience” in a language is never a good indication of ability. In my example above, even if the survey knew that you quit studying at age 12, it would consider that to be “two years” of experience. But it would fail to look at the quality of those years.
What were you actually doing to learn the language in those years? Were you sitting in language class for two hours a week, feeling bored and doing the minimum work required? Or were you a passionate learner, spending hours of your free time every week trying to learn as much as you could?
I took German for five years in school, with very little to show for it at the end. I could barely order food or buy a train ticket in German. If I was a pessimist, I would say, “See? I’m crap at languages! I studied German for five whole years and didn’t learn anything!”
But this would be completely false. I didn’t “study” German for five years. I attended classes for two hours a week, nine months a year for five years. I could accumulate the same number of study hours in a few months if I studied for 2-3 hours per day. And the quality of those hours matters too. I spent most of my German classes with my eyes glazed over, thinking about the Teenage Mutant Hero Turtles. You simply can’t equate an hour of intensive, passionate language study with an hour of half-arsed, mandatory study.
When people come to me and ask, “How many years does it take to learn [insert language here],” I tell them they’re asking the wrong question. Hours – and the quality of those hours – are what matter most in your language learning.
Now, how do you find the hours you need for language learning? In my most intensive language missions, I used a few approaches:
- I would work longer hours at my job as a translator to save up some money so I could then devote almost every hour of the day to my language for a while.
- I would give up TV shows and any other dead time during the day that kept me from studying.
- I would even sometimes give up socialising with friends in languages that I didn’t need to practise.
Add all these hours up over a 3-month mission, and I would make far more progress than I did in those “five years” of German.
Conclusion: You Can Become Fluent at Any Age!
I hope I’ve managed to convince you that the big scary headlines saying “if you haven’t started studying a language by age 10, you’ll never ever be fluent” simply don’t stand up to critical analysis, common sense, or just plain old observation of successful learners in the real world.
Always be sceptical of any news stories that state hard-and-fast rules about people’s language-learning ability. There are exceptions to every rule, so don’t let anyone else define your limitations. Prove them wrong!
Finally, I recorded a one-hour live stream about this news story and the study behind it. You can watch the video here: