critical review: Second Language Acquisition and Critical Period Hypothesis

Critical Review: Second Language Acquisition and Critical Period Hypothesis

2143 00027

Postgraduate of English Education Department of UMPAR

The “Critical Period Hypothesis” refers to the idea that the ability to acquire language is related to aging and there is an ideal period of time to attain a language, after which it is no longer possible. This hypothesis was originally proposed by Penfield and Roberts (1959) and followed up by Lenneberg (1967) who suggested that it could be extended to the second language acquisition. Subsequent studies all found the same result that second language learning performance correlates negatively with the age at which the learning begins prior to puberty. Contrarily, among late learners, the correlation of age with learning performance suggests that different mechanisms are affecting adults’ proficiency other than maturation (Birdsong & Molis, 2001; Hakuta et al., 2003; Oyama, 1976).  However, it is controversial that this negative correlation between age and learning performance can be attributed to a biologically based critical period for second language acquisition that prevents older learners from achieving native-like competence in phonology as well as syntax of a non-native language.

In this literature review, the existence of a critical period for second language acquisition is investigated. It is hypothesised that native-like attainment of second language acquisition is influenced by social, educational and motivational factors such as age of immigration and amount of second language education. The following three literature reviews attempt to demonstrate and support this hypothesis.

First, in a research article by Hakuta, Bialystok and Wiley (2003) the effect of age of acquisition on second language proficiency by studying a large number of immigrants with different ages of initial exposure to English was examined. Evidence for a critical period would require a clear discontinuity in learning outcome around a specific age, independent of social and educational factors which can also affect performance. They searched for an evidence for a critical period by observing immigrants second language learning performances.

In the past, Johnson and Newport (1989, 1991) have found a strong age-related decline in proficiency for languages learned prior to puberty and random variation in second language achievement among later learners, supporting the Critical Period Hypothesis. However, the data were controversial because age of exposure, duration of exposure, and social and linguistic backgrounds of the participants are possible confounding factors (Bialystok and Hakuta, 1994).

Other researchers (e.g., Epstein, Flynn & Martohardjono, 1996; Hakuta, 2001) rejected the Critical Period Hypothesis. Identification of older learners who achieve native-like competence in a second language and behavioural evidence that fails to reveal a qualitative change in learning outcomes at the end of putative period have been used to challenge the Critical Period Hypothesis.

An alternative to the critical period hypothesis is that second language learning declines with age, because of social and educational factors, as well as cognitive aging which interfere with the adult’s ability to learn a new language (Hakuta et al., 2003). It is suggested by Bialystok and Hakuta (1994) and Flege et al. (1999) that among social factors, amount of the second language education is the strongest predictor of second language acquisition. Amount of language use varies among immigrants because they have different experiences, qualitatively and quantitatively exposed to different levels of the new language and have different opportunities for formal study of the language.

Besides social factors, there are age-related changes in cognitive processing which affect the ability to learn a new language. Therefore, older learners would have more difficulties in acquiring a second language than younger ones (Salthouse, 1992).

In the study done by Hakuta et al. (2003), data from 2.016.317 native Spanish speakers and 324.444 native Chinese speakers as identified by the 1990 U.S. Census were derived. Their minimum length of residence in the U.S was 10 years. The participants self-described their English ability. They were also asked questions about their age, year of arrival in the U.S, and educational background. The goal was to model English proficiency on age of immigration, education and existence of a critical period. Age of immigration was estimated from the current age of participants and year of arrival in the U.S.  To test for evidence for a critical period, the procedures for modeling regression discontinuities suggested by Neter et al. (1996, pp. 474-478) was followed. If a discontinuity in the regression of English proficiency on age of immigration at the point marking the end of the critical period is found then a critical period would exist. Change in mean of the regression line and a change in the slope of the regression line were tested to see if discontinuities in the regression model exist (Neter et al., 1996, p. 478).

Analyses of data showed a strong main effect of education was found. A moderate effect for the age of immigration and education was found (Chinese: marginal R²=.0932; Spanish: marginal R²=.0632) which suggests that the relation between age of immigration and English proficiency change with different levels of education. The results indicated that length of residence (10 years or more) does not affect English proficiency. For both Spanish and Chinese speakers, no significant change in the mean or in slope of the regression line was found. Interactions between both the mean and slope change variables and the education variables were tested, as well as the size effects which would be evidence for regression discontinuities. There was little evidence for these discontinuities (No change in R² more than .0018). Moreover, to better understand the data, Hakuta et al. (2003) tested the data with a local regression model. The local regression plot also found no evidence for discontinuity.

All of these findings combined confirm the hypothesis that the negative correlation of age with learning performance is due to the social and educational factors such as age of immigration and amount of second language education, rather than a biologically based critical period (Hakuta et al., 2003).

Next, in a research article by Birdsong and Molis (2001), the idea that second language acquisition is constrained by maturational factors that include a critical period for native-like attainment is investigated. Johnson and Newport (1989) proposed a maturational model of second language attainment based on behavioural evidence. In their study, 46 Chinese and Korean immigrants living in the U.S., with Age of Arrival (AoA) ranging from 3 to 39 years, were administered a grammaticality judgements of 276 randomly ordered English sentences, half being grammatically correct. The sentences were in twelve types of structures, representing surface features of English morphology and syntax. Participants also asked about their language background, attitudes about the U.S., and motivation to learn English. Among early arrivals (AoA ≤ 15), a strong negative correlation was found between the second language performance and AoA. However, among late arrivals (AoA ≥ 17), a weak negative correlation was found between performance and AoA, indicating that near-zero incidence of native-like attainment was observed. The results were consistent with a maturational account and indicated that a critical period exists for second language acquisition. 

Similarly, Shim (1994) investigated the limits of second language attainment from a maturational view by testing English acquisition of Korean subjects who varied in their age of second language exposition. Shim concluded that exposure to the second language before age 5 assures native-like performance.

On the contrary, post maturational age effects were found by Bialystok and Hakuta (1994, 1999), Birdsong (1992) and Flege (1999) who did not support a maturational view. Bongaerts (1999), Cranshaw (1997), and van Wuijtswinkel (1994) have testified significant numbers of late learners who perform like natives on various linguistic tasks. Bongaerts (1999), Cranshaw (1997) and Flege (1999) suggest that results are not generalizable, as the incidence of native-like attainment appears to depend on particular pairings of first-second languages. Moreover, Bialystok and Miller (1999) studied the acquisition of second language English morphosyntax by native speakers of Spanish and Chinese. Biologically based second language acquisition constraints were not predicted. Age and performance of second language were negatively correlated for a range of AoAs, however an evidence for a critical period was not found.

Flege, Yeni-Komshian, and Liu (1999) investigated end-state proficiency in English pronunciation by Korean natives of varying AoAs in the U.S. With increasing AoA, there was a decline in pronunciation accuracy and performance, but not in morphosyntax. The present research by Birdsong and Molis (2003) examines the predictions of a maturational account of second language by replicating the Johnson and Newport (1989) study with a sample of native Spanish speakers whose second language is English.

The participants were 61 native speakers of Spanish, of 29 were Early Arrivals (AoA <= 16) and 32 were Late Arrivals (AoA >= 17). The mean length of residence in U.S. was 10 years. All the participants had obtained a bachelor's degree, and were students, faculty members or employees at major U.S universities. All materials and procedures were identical to J&N89 study except that 274 sentences were asked. Participants were given three items, one to self-asses their fluency in English, the other two to estimate current use of Spanish and English.

The researchers found weak negative correlation between early AoAs and judgement accuracy (r=-.23). Conversely, strong negative correlation was found between late AoAs and judgement accuracy (r=-.69). This result contrasts with J&N89 findings where a weak correlation was found between the AoA with language performance among late arrivals. J&N89 study and the replication of J&N89 study indicated significantly different results in terms of onset of age effects, the correlation of scores with early and late AoA and the regression discontinuities for early and late AoAs.  Increasing variability in performance was found with later AoA in J&N89 study. This greater variability after maturation is consistent with the Critical Period Hypothesis (Johnson & Newport, 1989, pp. 96-97). However, the replication study did not reveal variations in their performance among late arrivals. In addition, 13 out of 32 late AoAs (AoA ≥17) showed native-like performance in English, refuting the Critical Period Hypothesis which supports the criterion that among late learners near-zero incidence of native-like attainment can be observed. With native speakers of Spanish, Birdsong and Molis (2001) found that among Spanish speakers the onset of age effects differs from what J&N89 found for Chinese and Korean speakers. In their study, Birdsong and Molis (2001) were particularly concerned with the end-state. Sorace (1993) argues that end-state second language grammars differ qualitatively as a function of the first language. As a result, the hypothesis that, if a maturational based critical period determines the level of attainment, then the findings should not be unique to the first-second language pairings is not supported by the replication study. Among Late AoAs and over all subjects, amount of current English use is found to be a strong predictor of performance. Both J&N89 and the replication study found that AoA is the strongest predictor of performance.

Consequently, these results confirm the hypothesis that the negative correlation of age with learning performance is due to the social and educational factors such as age of immigration and amount of second language education, rather than a biologically based critical period. These criteria can be generalized to any first-second language pairings (Birdsong & Molis, 2001)

Lastly, the third literature accounts adult’s poor performance in second language learning, as opposed to children, to motivational factors such as desire to identify with native speakers at the same time supporting a sensitive period for second language acquision. A research paper by Oyama (1976) investigated the existence of a sensitive period for acquisition of the phonological system of a non-native language. Past research has found that only early learners of a second language are able to achieve native-like performance in pronunciation (Krashen, personal communication; Asher & Garcia, 1969). Oyama (1976) tested the hypothesis that there is a developmental period beginning approximately from 18 months to the end of maturation, during which it is possible to fully acquire the phonology of at least one non-native language, and after which complete acquisition is extremely unlikely.

Participants were 60 Italian-born male immigrants, from New York metropolitan area, who learned English as a second language upon their arrival in the U.S. To minimize the social and cultural factors and dialectal variations, the participants were taken from the upper educational groups to ensure that participants had ample exposure to Standard English. Participants were distributed along two independent variables (IVs), age at arrival in the U.S (6-20 years) and length of stay in the U.S (5-18 years).

Participants were asked to read a sample paragraph in English to test their pronunciation. Then, they were asked to recount a frightening episode in their life to observe their stylistic variation in causal speech. It is observed by Labov (1966c) that when people engage in casual speech tends to be informal. Both the paragraph readings and the stories were recorded in tapes with control samples mixed at irregular intervals. Finally each subject filled a questionnaire about their method of learning English, relative amount of first and second language use and certain attitudes about the second language. The tapes were judged for degree of accent by two native graduate linguistic students and defined as non-native English speakers.

The analyses of the data showed that age at arrival was a strong predictor of accent while length of stay had very little effect. Unexpectedly, casual speech showed fewer accents than the paragraph readings and age at arrival had a stronger effect for the paragraphs than for the casual speech. Paragraph samples yielded results that stylistic variation for early age at arrivals seems to resemble that of native speakers, more than later ones. These results indicate that there is a strong relationship between age and language which provides a theory for the postulation of a sensitive period for second language acquisition.

This paper has reviewed the empirical literature available as to the existence of a Critical Period for second language acquisition and the reasons that constraint adults from achieving a native-like performance. All of the literature studied supports the hypothesis that the native-like attainment of second language acquisition is influenced by social and educational factors and points out the importance of early age of arrival and large amount of formal education of the second language to come to a native-like level in non-native languages (Birdsong & Molis, 2001; Hakuta et al., 2003; Oyama, 1976).

References

Asher, J. J., and Garcia R. (1969). The optimal age to learn a foreign language. Mod. Lang. J. 53:

334-342.

Bialystok, E., & Hakuta, K. (1994). In other words: The science and psychology of second-

language acquisition. New York: Basic Books.

Bialystok, E., & Miller, B. (1999). The problem of age in second language acquisition:

Influences from language, task, and structure. Bilingualism: Language and Cognition,

2, 127–145.

Birdsong, D. (1992). Ultimate attainment in second language acquisition. Language, 68, 706–

755.

Birdsong, D. & Molis, M.  (2001).On the evidence for maturational Constraints in second-language

acquisition. Journal of Memory and Language . 44, 235-249.

for second-language acquision. American Psychological Society. 14, 31-38.

Bongaerts, T. (1999). Ultimate attainment in foreign language pronunciation: The case of very

advanced late foreign language learners. In D. Birdsong (Ed.), Second language acquisition and the Critical Period Hypothesis(pp. 133–159). Mahwah, NJ: Erlbaum.

Cranshaw, A. (1997). A study of Anglophone native and nearnative linguistic and metalinguistic

performance. Unpublished Ph.D. dissertation, Université de Montréal.

Epstein, S., Flynn, S., & Martohardjono, G. (1996). Second language acquisition: Theoretical

and experimental issues in contemporary research. Behavioral and Brain Sciences, 19, 677–758.

Flege, J. E. (1999). Age of learning and second-language speech. In D. Birdsong (Ed.), Second

language acquisition and the Critical Period Hypothesis (pp. 101–131). Mahwah, NJ: Erlbaum.

Flege, J. E., Frieda, A. M., & Nozawa, T. (1997). Amount of native-language (L1) use affects the  

pronunciation of an L2. Journal of Phonetics, 25, 169–186.

Flege, J. E., Yeni-Komshian, G. H., & Liu, S. (1999). Age constraints on second-language

acquisition. Journal of Memory and Language, 41, 78–104.

Hakuta, K., Bialystok E. & Wiley, E. (2003).Critical Evidence: A test of the critical-period

hypothesis

Jakobovits, L. A. (1970). Foreign Language Learning: A Psycho-linguistic Analysis of the

Issues, Newbury House, Rowley, Mass.

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.

Johnson, J. S., & Newport, E. L. (1991). Critical period effects on universal properties of

language: The status of subjacency in the acquisition of a second language. Cognition, 39, 215–258.

Krashen, S. (1973). Lateralization, language learning, and the critical period: Some new

evidence. Lang. Learning 23: 63-74.

Labov, W. (1966c). The Social Stratification of English in New York, Center for Applied

Linguistics, Washington, D.C.

Lambert, W. E. (1963). On second language learning and bilingualism. Mod. Lang. Z

47: 114-121.

Larew, L. (1961). The optimum age for beginning a foreign language. Mod. Lang. s

45: 203-206.

Lenneberg, E. H. (1976). Biological foundations of language. New York: Wiley.

Lenneberg, E. H. (1960). Review of Penfield and Roberts' Speech and Brain Mechanisms.

Language 36: 97-112.

Lenneberg, E. H. (1966). Speech development: Its anatomical and physiological concomitants.

In Carterette, E. C. (ed.), Brain Function, Vol. Ill." Speech, Language and Communication, University of California Press, Berkeley, pp. 37-66.

Neter, J., Kutner, M., Nachtsheim, C., & Wasserman, W. (1996). Applied linear statistical

models. Chicago: Irwin.

Oyama, S. (1976).A Sensitive period for the acquisition of a nonnative phonological system.

Journal of Psycholinguistic Research. 5, 261-283.

Penfield, W., & Roberts, L. (1959). Speech and brain mechanisms. Princeton, NJ: Princeton

University Press.

Pulvermüller, F., & Schumann, J. H. (1994). Neurobiological mechanisms of language

acquisition. Language Learning, 44, 681–734.

Salthouse, T.A. (1992). Mechanisms of age-cognition relations in adulthood. Hillsdale,

NJ: Erlbaum.

van Wuijtswinkel, K. (1994). Critical period effects on the acquisition of grammatical

competence in a second language. Unpublished thesis, Katholieke Universiteit,

Nijmegen, The Netherlands.