Am I too old to learn a new language?
Dobar dan! Halo!Welcome back to Adventures in Language! Did you know that if you start learning a language as a kid, you’ll end up a lot more successful than if you start learning as an adult?
Well, yeah. You probably did know that! But while the relationship between age and language learning may seem obvious, it’s not always so clear. In this article, we’re exploring what we know about how age affects language outcomes, and why it’s complicated. We’ll even talk about one of the most hotly contested topics in Second Language Acquisition: The Critical Period Hypothesis.
You’ve probably figured out by now that people who learn second languages (L2s) as children usually end up sounding like native speakers, but people who learn L2s as adults usually, well… don’t. At the very least, most adult learners find that foreign accents are pretty hard to shake, and some grammar errors can be really persistent.
The effects of age start to emerge pretty early in the language learning process, even for first language (L1) learners. Babies are born able to recognize and distinguish speech sounds in any language, but by about 10 months old, they can only distinguish sounds that exist in languages that they’ve been exposed to (Werker & Hensch, 2015).
But what would happen if they weren’t exposed to any languages? Is there a point at which it’s just too late to learn a language? Enter the Critical Period Hypothesis.
What is the Critical Period Hypothesis?
A critical period is a period in development during which a skill or ability is most easily acquired, given adequate exposure. Once the critical period is over, it’s impossible to develop that ability either at all, or at least to its full potential. There are known critical periods for things like vision and hearing. Fortunately, it’s very rare for children to go without early exposure to language, so this isn’t the easiest thing to study.
But there have sadly been a handful of cases of isolated children, and many deaf children of hearing adults actually don’t get exposure to sign language in their early years. So the experiences of these children have led researchers to propose that there is a critical period for L1 learning — children with no exposure to language before around puberty may not ever be able to fully acquire their L1, and even those exposed as young as 2 to 4 years old may not achieve native-like proficiency.
Are L2s subject to this same critical period?
It could be that simply having experience learning an L1 tunes your language learning “muscles,” so to speak, so that they are still usable for language learning later in life. The Critical Period Hypothesis says that age is strictly tied to ultimate attainment, which is the end point of learning — it’s essentially the highest proficiency that an individual can reach in a language. And if you don’t start learning a language by a certain age, your ultimate attainment will never match that of a native speaker.
The proposed cutoff for a second language critical period is usually around puberty, but some estimates are as low as age 7, or even younger, especially for pronunciation. If you’re listening to this episode and missed the boat on learning a second language in childhood, that might sound pretty bleak!
But opponents of the Critical Period Hypothesis point to examples of individuals who started learning their second languages later in life — as adults — and are indistinguishable from native speakers. While these learners are no doubt exceptions, their mere existence would disprove the Critical Period Hypothesis.
In return, Critical Period supporters claim that more sensitive language tests can reveal that such speakers aren’t native like after all. But there’s a fundamental problem with this line of evidence — it’s common to see monolingual native speakers used as a benchmark for evaluating gifted late-L2 learners in critical period studies. But gifted L2 learners aren’t monolingual. Bilingual native speakers perform differently from monolinguals on many tests of language, so it seems misguided to expect late bilinguals to perform like monolinguals (Ortega, 2014).
Open Questions about Age and Second Language Learning
Two things that Critical Period supporters and opponents agree on are 1) that ultimate attainment for early learners is likely to be higher than for late learners, and 2) that native-like or not, some late learners do achieve very high proficiency. But there are so many open questions.
1. How exactly do language learning outcomes relate to age?
This is intricately tied to the Critical Period Hypothesis. One option is there is no Critical Period, but learning outcomes decline steadily and continuously as age of acquisition (AOA) increases. We don’t know how steep this decline is, but basically the younger you are, the higher your ultimate attainment. Maybe after a certain age, ultimate attainment is low, on average, but it no longer declines with age. A 16 year old, 30 year old, and 70 year old are all on equal footing.
If a critical period does exist, it most likely looks something like this.
Ultimate attainment remains uniformly high for AoAs during childhood and then abruptly starts decreasing after a certain age, with age effects perhaps leveling off. But when is this cutoff? Puberty? Early adulthood? Early childhood? Are there different cutoffs for different aspects of language? Grammar tends to be harder to master than vocabulary, and phonology (that part of language responsible for your accent) seems to be especially difficult to learn, even for the youngest learners.
For more on the nuances of the different possible relationships between age and ultimate attainment, as well as how this is different from the relationship between age and learning ability, see Birdsong (2018) and Hartshorne (2022).
2. Why do young learners tend to have better language outcomes?
One very plausible explanation for age effects is biology. Hormones that affect learning and memory change around puberty, which is a commonly cited critical period cutoff. Children’s brains also work differently from adults’. They have more neural plasticity, which is a fancy way of saying that it’s easier for their brain cells to make new connections. Some evidence suggests that these brain-based differences might account for a critical period for learning grammar but not vocabulary…though even adult learners have been shown to process grammar similarly to native speakers!
Another possibility is that children are just better learners! But actually, when directly compared, adults and older children learn languages faster than young children. Children do eventually edge out adults, but only after up to 5 years of learning in immersion settings. In formal classroom settings, the adult advantage actually persists for much longer!
It could be the circumstances in which children learn languages. Children usually have an urgent need to learn languages, especially if they’re living in a country where that language is spoken. Fitting in on the playground and understanding what’s going on at school are pretty big motivators! They also tend to get more rich, varied, and ample exposure than adults do. Some studies have found that language learning abilities hold pretty steady even into the late teens, suggesting that societal factors related to school and not having the pressures of, well, “adulting” might contribute to age effects.
It might simply be that kids, whose first languages are less developed, experience less interference when learning second languages. Or maybe there’s some other explanation!
3. Many of the key variables involved in age and language learning —like AoA, years of exposure, and age at testing — are highly interrelated.
People who start learning as kids have more time to master the language. We can try to control for years of exposure — say we only look at people who have been learning for 30 years, which should be sufficient for reaching ultimate attainment. Well, the ones who started learning in early childhood will be in their thirties. But after 30 years, late learners will be older adults. So now we’re testing people at different ages, and their performance on language tests may be influenced by things other than language ability, like general cognitive abilities that change with age, or familiarity with the testing environment.
In order to disentangle variables like AoA, years of exposure, and age at testing — not to mention other factors that affect language learning and are also related to age, like motivation, learning context, learning strategies, aptitude…we’d need to look at tons of learners. Recent estimates based on statistical models suggest that tens of thousands of learners would be needed to understand these effects…which isn’t exactly your typical linguistics research sample!
As it turns out, this seemingly simple relationship between age and language learning is, in fact, not so simple!
To Recap:
If there’s one thing everyone can agree on, it’s that children have an advantage over adults when it comes to language learning success.
This advantage doesn’t necessarily translate into “the younger the better.” It’s possible that there is a continuous relationship between age and ultimate attainment, which would mean that starting at age 3 is better than 5 which is better than 7 and so on. But if there’s a critical period, anyone who starts learning before the critical period ends can achieve native-like proficiency.
Where do age effects in second language learning come from? Biological changes in hormones and neural plasticity, superior learning ability, higher motivation, richer learning environments, and less interference from the first language may all contribute.
It is extremely difficult to make conclusions about the relationship between age and language learning because relevant variables like age of acquisition, years of exposure, and age at testing are so highly interrelated. Also, it’s time to acknowledge that monolingual native speakers might not be the most accurate benchmark for evaluating critical period effects. Newer methods using big data and statistical modeling might be the key to disentangling these issues.
And if you are a late second language learner, don’t lose heart! It might require a bit more effort, but you can still make great strides in your language learning journey!
Thanks for reading!
If you liked this article, let us know! Want more engaging language content like this? Subscribe to our YouTube channel and podcast feed, and follow us on Instagram @MangoLanguages! Or visit us at Mangolanguages.com!
Start LearningFill out the form below for some free materials on age effects and the critical period, including an interactive quiz to test your knowledge! Doviđenja! Hati-hati!We look forward to seeing you back here for our next article. To embark on your next language adventure, visit us at mangolanguages.com!
Wondering what languages were used in today’s article?
Dobar dan (DOHbahr dahn) and Doviđenja (DOHveeDJEHnyah) mean “Hello” and “Goodbye” in Croatian, a Slavic language spoken in Croatia and surrounding countries.
Halo[hah-lo]and Hati-hati[ha-tee ha-tee]mean “Hello!” and “See you later!” in Indonesian, the standardized official language of Indonesia.
Want to know more about the scientific research underlying this episode? Here’s some of the research we consulted and/or mentioned in this episode.
For recent reviews of the vast age effects literature that dive into the nuances of this research, check out:
For a seminal paper on age effects of language learning, see: Johnson, J. S. & Newport E. L. (1989). Critical period effects in second language learning: the influence of maturational state on the acquisition of English as a second language. Cognitive Psychology, 21(1), 60–99.
And for a replication and exploration of how aptitude may overshadow age in older learners, see: DeKeyser, R. M. (2000). The robustness of critical period effects in second language acquisition. Studies in Second Language Acquisition, 22(4), 499–533.
For a recent big data approach to age effects, see: Hartshorne, J. K., Tenenbaum, J. B., & Pinker, S. (2018). A critical period for second language acquisition: Evidence from 2/3 million English speakers. Cognition, 177, 263–277.
For a neurological perspective on age effects, see: Ullman, M. T. (2020). The Declarative/Procedural Model: A Neurobiologically-Motivated Theory of First and Second Language. In B. VanPatten, G. D. Keating, & S. Wulff (Eds.), Theories in Second Language Acquisition (3rd ed., pp. 128-161). Routledge.
