A critical period to learn a second language?

Critical PeriodResearch findings on the critical period may influence educational policy and practice. Over recent years, many Catalan primary schools have lowered the starting age of foreign language learning and start teaching English in early childhood education (1). However, making pedagogical decisions solely on the basis of research on the critical period is problematic. In this article, we take a closer look at the issue.

The hypothesis that there is a critical period in life to learn a language was first proposed by the Canadian neurologists Penfield and Roberts in 1959. The hypothesis was then popularized by the linguist and neurologist Eric H. Lenneberg in his 1967 book “Biological Foundations of Language.” Lenneberg speculated that the critical period hypothesis not only affected first language acquisition, but could also be extended to second language acquisition. According to the critical period hypothesis, language can be acquired only within a critical period, extending from early infancy until puberty. The hypothesis is untestable for first language acquisition because no child can be intentionally deprived of language for experimental purposes. However, documented cases of feral children, victims of severe abuse, neglect and social isolation, such as Victor or Genie, deaf children of hearing parents, and children recovering from aphasia and language disorders lead us to believe that acquisition of a language is guaranteed up to the age of six, but then steadily compromised until puberty. After puberty, acquisition is rare. Some aspects of language will be learned but full mastery will not be achieved. The reason behind the critical period is thought to be of a biological (or maturational) nature and related to neurophysiological changes in the brain that allow, for example, the creation of more complex neural networks early in life (Long, 2007).

The existence of a critical period is commonly accepted for first language acquisition but it remains controversial and the subject of debate in the case of second language acquisition. The good news is that, unlike in the case of first language acquisition, the hypothesis is testable for second language acquisition. A great deal of empirical evidence has been gathered showing a link between the age of an individual’s first exposure to a second language and his or her ultimate attainment (or long-term achievement) in that language. The following are some of the conclusions of research:

  • The terms “sensitive periods” and “windows of opportunity” are more accurate labels than “critical period”: there is no sharp and sudden or abrupt decline in language development across the board but different windows of opportunity for different language aspects and domains (phonology, lexis, morphosyntax).
  • There are consecutive sensitive periods for each of the language domains, in the following order:
    1. phonology (between age 0 and age 6)
    2. lexis and collocation (between age 0 and age 10)
    3. morphosyntax (between age 0 and age 12).
  • In a study by Granena and Long (2013) with Chinese learners of Spanish in Madrid (Spain), the oldest starting ages (i.e. ages of arrival in Spain) for participants that reached nativelike Spanish pronunciation, vocabulary, and grammar were 5, 9, and 12, respectively.
  • Before age 10, there are few individual differences in ability to learn a second language. But among adults, individual variations in success are large and likely to be affected by cognitive abilities such as language aptitude.
  • A 2018 study co-authored by Steven Pinker, a professor of psychology at Harvard University, with data on 669,498 people learning a second language around the world confirmed a critical period for grammar: learners could reach nativelike scores on a quiz if they had started learning the language before the age of 17 or 18.

Be that as it may, good language acquisition may be possible after puberty and talented language learners who pass as native speakers may be very successful in certain language aspects. Similarly, someone who starts young enough may not reach nativelike levels. Neither of these cases refutes the critical period hypothesis. The single piece of evidence that could refute the hypothesis would be a learner who starts learning the second language as an adult and scores like a native speaker in a variety of tests and formats, in various language domains (phonology, lexis and grammar), both receptively and productively. This learner is still to be found (and tested).

In conclusion, it seems that a young starting age is necessary, but not sufficient, for long-term success in a foreign language context. In order to take advantage of children’s capacity to learn a foreign language, you need appropriate teaching methods and sufficient exposure to the language. In addition, there are other concerns, such as teacher training and command of the foreign language (2), the quality of foreign language materials, continuity when transitioning from primary to secondary school, etc. The potential advantages of an early start, especially in pronunciation, listening and speaking, are easily lost if inadequate resources are made available overall. The real question to ask, of course, is whether we need near native-like speakers of foreign languages. If policy makers consider high-level proficiency a goal, then they should approach foreign language instruction more holistically and allocate public funds accordingly.

  1. See, for example, the Escola Montserrat, Sant Just Desvern.
  2. For example, 60% of Australian primary school teachers surveyed by Nicholas et al. (1993) reported that they were unable to conduct a full class in the foreign language.

References to the papers cited are available on request.

Gisela Granena
Latest posts by Gisela Granena (see all)

About Gisela Granena

I am a lecturer at the Centre for Modern Languages at the UOC. I hold a PhD in Second Language Acquisition from the University of Maryland (USA), an MA in Intercultural Communication from the same university, and a Postgraduate degree in teaching English as a foreign language from the University of Barcelona. My areas of expertise include language acquisition, cognitive psychology and measurement/statistics. I have published research on individual cognitive differences in both instructed and naturalistic SLA contexts; aptitude-treatment interactions; task-based language teaching (TBLT); measures of implicit and explicit language knowledge, and the effects of early and late bilingualism on long-term L2 achievement.
This entry was posted in Language learning, Profession. Bookmark the permalink.