CPH - language acquisition

Second Language Acquisition and the Critical Period Hypothesis

When learning a second language a native-like level is hardly ever reached. Various research has been conducted into this phenomenon which has resulted in a number of hypotheses concerning the possible cause thereof. One of these is the Critical Period Hypothesis which states that there is an optimum period for learning a second language that ends after puberty. Can the Critical Period Hypothesis adequately explain why people acquiring a second language after the age of puberty will never accomplish native-like competence? In addressing this question this paper will delineate the theoretical background of, and some of the research evidence for, the Critical Period Hypothesis for Second Language Acquisition and summarize the criticism. Finally some other hypotheses pertaining to the cause of this phenomenon and their compatibility with the Critical Period Hypothesis will be discussed.

The Critical Period Hypothesis
The concept (Singleton 2005:270) of a Critical Period finds its origin in the biological sciences where it is used to designate a well-defined period of limited duration during which specific conduct or competences are to be achieved, or during which the specialization of cells takes place. If and when these events do not take place during this period these potentials are lost forever. This concept has been applied to linguistics. The learning of a – first – language should occur in early childhood, because that is the critical period for language learning. If, for some reason, a child does not learn a language during this period, the ability to learn a language is lost. In 1959 Penfield and Roberts (Singleton 2005:270) stated that after the age of nine the brain becomes less fit to learn languages. Eight years later Lenneberg (Singleton 2005:271) saw a connection with the lateralization process which takes place in the brain with the specialization of each hemisphere. He concludes that the Critical Period for learning a language starts at age two and ends at puberty. They were followed by a host of other researchers, each giving his own timetable of the critical period, sometimes dividing up this period into sub-periods for phonetics/phonology, syntax and semantics (Singleton 2005: 271-273).
The discovery of a then thirteen year old girl in 1970, called Genie in the literature1, seemed to corroborate the existence of a Critical Period for language learning. Because her parents had kept her locked in her room and had hardly ever spoken to her, she had a negligible command of language. Attempts to teach her language were not very successful. The Genie-case is usually mentioned as support of the Critical Period Hypothesis. Opponents however, state that in case of so-called feral children the cause of the language-inability might well be attributed to a general retarded development. However, the results of research2 which has been conducted to late learning of American Sign Language by deaf children confirmed the Critical Period Hypothesis: the older deaf children performed less well on the tests. Thus it follows that in general there is agreement to the existence of a Critical Period for learning a first language. No agreement however, exists as to the exact timetable of this period, though generally it is believed that it ends somewhere around puberty.

The Critical Period Hypothesis and Second Language Acquisition
When learning a second language it is implied that there is already some fluency in a first language. By being born into a specific culture with a specific language, the first language is learnt involuntarily; spontaneously. This contrasts with the learning of a second language: whenever one starts to learn a second language the first language will always be there as a cultural and linguistical point of reference.
The Critical Period Hypothesis states that there is an optimum period for language learning– which ends at puberty at the latest. When applied to Second Language Acquisition it follows that achieving a proper command of the second language will become more difficult when learning starts at an advanced age. The Critical Period Hypothesis applied to Second Language Acquisition claims, simplified, that adult learners of a second language will never achieve the high competence in the second language that children, learning the second language, can accomplish. Note: it states that children can accomplish native-like proficiency, not that they will.
The research conducted to test the hypothesis often tests second language users on several language tasks. The participants are grouped into categories (Long 2005:290-291) such as the age at which exposure to the second language began – called age of onset; or the age upon their arrival in the second language-country – age on arrival. The duration of residence in the second language community is also taken into account, as is the age at the time of testing. The tests vary in what and how second language skills are being tested (Long 2005:288): pronunciation, syntax, oral telling and grammar, to name only a few. Within the scope of this paper it is not possible to give a full review of the many research-data available in the literature. DeKeyser (2000: 501-506) gives a short review, from which the following examples are taken. In 1980 Patkowski tested the syntactic and grammatical proficiency of a group of non-native speakers of English. Persons who arrived to the United States before the age of fifteen had better scores on both tests. In 1989 Johnson and Newport conducted a test with native Chinese and Korean speakers who had migrated to the United States at various ages. They were asked to perform grammaticality judgement tests on sentences which were uttered orally. The younger the age of arrival was, the better their performance on the test: the results showed a gradual decline in proficiency from age six to age seventeen. There was however, no further decline for persons older then seventeen at age of arrival. Comparable results were obtained in the 1992 test with the same group, this time using written sentences. The final example given here from DeKeyser is a study by Sorace. In 1993 he tested native French and native English speakers on their proficiency in their second language: Italian. All of them learned Italian well after age eighteen and were by now near-native speakers. They were asked to perform grammatical judgement-tests. Their grammatical judgement differed considerably from that of native Italians, indicating that it is fairly impossible to gain a proper grammatical knowledge of a second language when learning this language after puberty.
But it are not just the linguistic scientists who are concerned with this hypothesis: there are also contributions from researchers in the discipline of neurobiology. Results from a research conducted by Stowe and Sabourin (2005) lead to the conclusion that the same areas in the brain are used for both the first and the second language. The conclusions from this research do not corroborate the version of the Critical Period Hypothesis which claims that the language-area which is used for learning the first language is no longer available for the second language. However, the authors themselves state that further research is necessary. Singleton would no doubt agree: in his article he gives several examples of neurobiological research and he quotes Hyltenstam and Abrahamsson who stated that ‘there is certainly no consensus with regard to neurobiological interpretation of age effects’ (Singleton 2005:276).

Criticism of the Critical Period Hypothesis for Second Language Acquisition
If there is a Critical Period for the learning of a Second Language, one should expect a sharp decline in second language proficiency when this language is learned after puberty. A research conducted by Hakuta, Bialystok and Wiley (2003) led them to conclude that there is no evidence for the existence of a critical period for learning a second language. Using the responses of Chinese or Spanish-background immigrants given in the 1990 United States Census it appeared that the results showed no such sharp decline. There were however, strong relations between level of education and age of immigration on the one hand and the test-results on the other.
Long (2005) lists the evidence against the Critical Period Hypothesis in order to refute these claims. He starts by pointing out that authors do not agree on the exact nature of the Critical Period: they do not agree on the critical age period (both onset age and offset age). There is also no agreement on the various language faculties such as phonology, syntax and lexical definitions which should be tested. Many authors prefer combinations of the two categories. For example: Ruben sets the onset for phonetics/phonology at the sixth month of foetal life, and the offset on age one (Singleton 2005:272-273). Long concludes that the Critical Period Hypothesis for Second Language Acquisition does not exist.
Not only is there no agreement as to the exact nature of the Critical Period Hypothesis; several of the studies claiming to have refuted the Critical Period Hypothesis are flawed. Long discusses a number of these studies. Some studies indicate that adults are able to learn a second language faster than children. However, to conclude that adults are thus better performers is a misjudgement: the Critical Period Hypothesis is concerned with the results of language proficiency in the long term. Other studies have not selected their subjects well, or they place their subjects in wrong categories by confusing age of onset, age on arrival, age of testing and length of residence. Long lists more examples of ill-conducted research or too hastily drawn conclusions making interesting remarks here which every researcher should read. According to Long, none of the studies he describes can refute the Critical Period Hypothesis because, as he himself emphasized: there is no agreement to the exact content of the Critical Period Hypothesis. Singleton concludes that it cannot ‘be regarded as a scientific hypothesis either in the strict Popperian sense of something which can be falsified or indeed in the rather looser logical positivist sense of something that can be clearly confirmed or supported. As it stands it is like the mythical hydra, whose multiplicity of heads and capacity to produce new heads rendered it impossible to deal with.’(Singleton 2005:280).

Alternative hypotheses
Some of the main other hypotheses to the Critical Period Hypothesis will be discussed here. The first is the Universal Grammar system. First described by Chomsky in 1965 (Herschensohn 1998:611) and later termed Universal Grammar by him, this hypothesis assumes that there is an innate sense of grammar in human beings which determines both the form or the principals of grammar and the procedures or strategies to be used when acquiring language.
A strong advocate of the theory of Universal Grammar, though slightly modified, is Lydia White (Hawkins 2005). Her argument can be summarised as follows. When learning a first language Universal Grammar is available. When learning a second language the first language grammar is readily available, and will be used in dealing with the second language. This is called the Full Transfer Hypothesis. But when this grammar-set is not applicable to the second language Universal Grammar has to be fallen back onto. This Universal Grammar is still available: this is called the Full Access Hypothesis. Combined, this hypothesis is termed Full transfer/Full Access Hypothesis. According to this hypothesis Universal Grammar is available, also when learning a second language. Therefore this hypothesis diverges from the Critical Period Hypothesis in that it implies that there is no such thing as a critical period: language acquisition is always possible, there is no critical period beyond which access to Universal Grammar is no longer possible.
In 1989 Robert Bley-Vroman emphasised that there is a huge difference in the way one learns a first and a second language (DeKeyser 2000:500; Herschensohn 1998:611). According to his Fundamental Difference Hypothesis only first language learners have access to Universal Grammar. When learning a second language, as an adult, Universal Grammar is no longer available. Therefore learners have to resort to other, general, problem-solving capabilities. Because they no longer have access to Universal Grammar they use the first-language grammar as substitute. The Fundamental Difference Hypothesis thus states that both strategy and form of Universal Grammar are no longer available to adult learners of a second language. This hypothesis does not conflict with the Critical Period Hypothesis: both hypotheses assert that learning a second language at a later age will be more difficult; the first because Universal Grammar is no longer available, the second because the critical period for language learning has passed.

The existence of a Critical Period for learning a first language seems well founded, though there is no agreement concerning the exact time-table and the time-table for specific competences. When applied to Second Language Acquisition matters become more complicated. To a large extent this is due to the ill-defined contents of the hypothesis: there is no agreement to the exact definition. This makes it difficult to test this hypothesis – hence some alleged flawed research - and, by implication, impossible to confirm or falsify it. Of the alternative hypotheses the Fundamental Difference Hypothesis is compatible with the Critical Period Hypothesis; the Full Transfer/Full Access Hypothesis is not. The answer to the central issue of this paper then should be that the Critical Period Hypothesis cannot adequately explain why people acquiring a second language after the age of puberty will never accomplish native-like competence.

Works Cited

“Critical period hypothesis” Wikipedia: The Free Encyclopedia. 20 February 2008. 26 March 2008.
DeKeyser, R.M. (2000). The robustness of critical period effects in second language acquisition. Studies in Second Language Acquisition 22. 499-533.
“Genie (feral child)” Wikipedia: The Free Encyclopedia. 25 March 2008. 26 March 2008.
Hakuta K., E. Bialystok and E. Wiley (2003). Critical evidence: a test of the critical-period hypothesis for second-language acquisition. Psychological Science 14. 31-38.
Hawkins, R. (2005). Rev. of Lydia White, Second Language Acquisition and Universal Grammar, Cambridge 2003. Language: Journal of the Linguistic Society of America 81. 754-757.
Herschensohn, J. (1998). Universal grammar and the critical age. Behavioral and Brain Sciences 21. 611-612.
Long, M. (2005). Problems with supposed counter-evidence to the critical period hypothesis. International Review of Applied Linguistics in Language Teaching 43. 287-317.
Singleton, D. (2005). The Critical Period Hypothesis: a coat of many colours. International Review of Applied Linguistics on Language Teaching 43. 269-285.
Stowe L.A. and L. Sabourin (2005). Imaging the processing of a second language: effects of maturation and proficiency on the neural processes involved. International Review of Applied Linguistics in Language Teaching 43. 329-353.

Cited: DeKeyser, R.M. (2000). The robustness of critical period effects in second language acquisition Hawkins, R. (2005). Rev. of Lydia White, Second Language Acquisition and Universal Grammar, Cambridge 2003 Herschensohn, J. (1998). Universal grammar and the critical age. Behavioral and Brain Sciences 21 Long, M. (2005). Problems with supposed counter-evidence to the critical period hypothesis. International Review of Applied Linguistics in Language Teaching 43 Singleton, D. (2005). The Critical Period Hypothesis: a coat of many colours. International Review of Applied Linguistics on Language Teaching 43 Stowe L.A. and L. Sabourin (2005). Imaging the processing of a second language: effects of maturation and proficiency on the neural processes involved