Why is the critical period hypothesis so heavily disputed, yet widely accepted; what are its major strengths and weaknesses; what other explanations exist for the perceived "critical period", if it does not exist?
Let us start with a simple, relatively informal statement: “in most cases, those who start learning a language as children become ultimately become more proficient in a language than those who start learning it later”. This is uncontroversial, and something I think the vast majority of second-language acquisition researchers would agree on. However, this is not how the Critical Period Hypothesis (CPH) is understood within the field of Second Language Acquisition (SLA). CPH is a large subject, and your question is hard to answer in a few paragraphs. Therefore, I am reusing large fragments of an assignment I wrote on this very topic for an SLA course a few years ago. Let me know if something is unclear or the style is too terse at some points.
There is no universally accepted definition of a critical period within linguistics and some of the controversies are caused by the fact that different researchers use different definitions.
There are a few key findings that are not controversial:
- early learners perform consistently well in all aspects of language use,
- as we move the starting age, they perform statistically worse and worse until puberty,
- however, the decrease in performance is not uniform.
An explanation, provided by Bialystok (1997) as an alternative to CPH, is the different learning style of children, compared to late learners.
Paradoxically, the differences (or lack thereof) between those who learn a foreign language as adults is the key factor in deciding whether CPH is true or not, and a controversial one:
- Some studies found correlations between the age adult learners started learning a language and their ultimate attainment. In other words, these studies suggest that if we compare people who have been learning a language for a very long time, the ultimate attainment of those who started at the age of 20 is statistically higher than the ultimate attainment of those who started at the age of 40. These studies argue that there is no CPH in the childhood, but rather that our abilities in learning a new language consistently decrease throughout our whole lives.
- Other studies found no clear correlations between the starting age and the ultimate attainment among adult language learners. They point out that the correlation between the starting age and ultimate attainment is clear for those who started before puberty. Based on that, they argue that there is something qualitatively different about starting to learn in an early age, and therefore conclude that it is an argument for CPH.
Definitions of the critical period used by those who argue against CPH
Controversies with the Critical Period Hypothesis (CPH) are related to the issue of ultimate attainment of early and late language learners, that is, the highest language proficiency level they can attain. The patterns in ultimate attainment may be explained by CPH, but they may also have different explanations. Some researchers support some form of the Critical Period Hypothesis (Johnson and Newport 1989, DeKeyser and Larson-Hall 2005, Patkowsky 1994, Scovel 1988), while others argue against it (Bialystok 1997).
A major problem with the Critical Period Hypothesis is that there appears to be no universally accepted definition of a critical period within linguistics. Bialystok (1997) bases her discussion of the critical/sensitive period (which she takes to be synonymous1) on a specific technical definition used in ethology, which includes 14 essential structural characteristics that describe such a period (Bornstein 1989). She argues that one of these characteristics is especially problematic – the system: “structure or function altered in the sensitive period” (Bornstein 1989:184). In other words, she argues that there is no period where a structure in the brain is modified in a way that makes subsequent language learning harder or impossible. Bialystok does, however, agree that there is an optimal period for language learning – something that can be characterised by the statement “On average, children are more successful than adults when faced with the task of learning a second language” (Bialystok 1997:117). Despite the controversy around other issues, this fact is uncontested and has been verified by numerous studies.
Bialystok (1997) rejects the existence of a critical period, because of lack of postulated structure that is modified when the period is over. She postulates that an important factor that causes differences in ultimate attainment between early and late starters is learning style: children prefer accommodation (creating new concepts) over assimilation (extending existing concepts). The question remains: why do they prefer accommodation? She suggests that “[t]his may be because children are in the process of creating new categories all the time as they are learning new information” (Bialystok 1997:132).
Definitions of the critical period used by supporters of CPH
The researchers who support some form of the Critical Period Hypothesis (Johnson and Newport 1989, DeKeyser and Larson-Hall 2005), formulate it in a form that is much weaker than Bialystok's (1997) formulation. What they postulate often resembles what Bialystok calls the optimal age.
Johnson and Newport (1989) reformulated CPH into two alternative hypotheses, in order to fit second language acquisition into the picture:
The exercise hypothesis: “Early in life, humans have a superior capacity for acquiring languages. If the capacity is not exercised during this time, it will disappear or decline with maturation. If the capacity is exercised, however, further language learning abilities will remain intact throughout life.” (Johnson and Newport 1989:64)
The maturational state hypothesis: “Early in life, humans have a superior capacity for acquiring languages. This capacity disappears or declines with maturation.” (Johnson and Newport 1989:64)
We can see that if a critical period was found for second language acquisition, we could be almost sure that it exists for L1 acquisition as well (the maturational state hypothesis). However, we cannot deduce in this way in case of the exercise hypothesis – non-existence of a critical period for L2 acquisition does not exclude in any way a possibility of such period for the first language (Bialystok 1997).
DeKeyser and Larson-Hall (2005) formulate the hypothesis as: “language acquisition from mere exposure (i.e. implicit learning) […] is severely limited in older adolescents and adults”. Their formulation is quite vague, as is the constatation that there is a “qualitative change in language learning capacities somewhere between 4 and 18 years”.
There are also definitions that restrict the Critical Period Hypothesis to specific subareas of the language faculty. The most commonly mentioned area is phonology, see e.g. Patkowsky (1994, cited in Bialystok 1997), Scovel (1988, cited in Bialystok 1997).
Age effects before and after puberty
The current consensus is that early learners perform consistently well in all aspects of language use. As we move the starting age, they perform statistically worse and worse until puberty. The decrease in performance is not uniform, and in some areas (such as phonology) it is particularly visible. Performance on the same level as early bilinguals is possible, but rare.
Probably the most controversial aspect is the performance of adult learners. After puberty there is much bigger variance in the performance, so data are more prone to different interpretations. The results obtained by Derwing and Munro (2013) suggest that comprehensibility and good accent are negatively correlated with the age of arrival, that is, the age when English language immersion started. Johnson and Newport (1989) found no correlation of starting age after puberty with ultimate language proficiency, while Bialystok (1997) re-analysed these data and found some negative correlation. A meta-analysis by DeKeyser and Larson-Hall (2005) downplays the role of post-adolescent correlations. As we can see, the jury is still out on this debate.
1 In neuroscience critical period and sensitive period are two separate concepts, see Knudsen (2004).
- Bialystok, E. 1997. The structure of age: in search of barriers to second language acquisition. Second Language Research 13(2): 116-137.
- Bornstein, M.H. 1989. Sensitive periods in development: structural characteristics and causal interpretations. Psychological Bulletin 105,179–97.
- DeKeyser, R. and J. Larson-Hall. 2005. What does the critical period really mean? In J. F. Kroll and A. M. B. de Groot. 2005. Handbook of bilingualism: Psycholinguistic approaches. Cary, NC: Oxford University Press. Pp. 109–27.
- Derwing, T. M., & Munro, M. J. 2013. The development of L2 oral language skills in two L1 groups: A 7-year study. Language Learning 63, 163-185.
- 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, 60-99.
- Knudsen, E. I. 2004. Sensitive periods in the development of the brain and behavior. Journal of Cognitive Neuroscience, 16, 1412-25
- Newport, E. L., & Supalla, T. 1987. A critical period effect in the acquisition of a primary language.
- Patkowsky, M. 1980. The sensitive period for the acquisition of syntax in a second language. Language Learning 30, 449–72
- Scovel, T. 1988. A time to speak: a psycholinguistic inquiry into the critical period for human speech. New York: Newbury House