The Effects of Age on Foreign Language Learning: The BAF Project

CARMEN MUÑOZ

The idea that there is a critical age for language learning that finishes before puberty was popularised by the Canadian brain surgeons W. Penfield and L. Roberts in their book Speech and Brain Mechanisms (1959). Penfield enthusiastically defended an early start for second language learning, basing his ideas on his studies on brain damage and his experience with his own children. According to Penfield, the time to begin schooling in second languages was between the ages of 4 and 10. Theoretical support came soon from E. Lenneberg who, in his Biological Foundations of Language (1967), noted that the rapid growth of nerve connections, which ceases at puberty, coincides with the child's acquisition of language. Lenneberg supported his neurological account of the Critical Period Hypothesis (henceforth CPH) with evidence from aphasic patients, who showed a more rapid recovery if the damage had taken place before puberty, and with feral children, children who had suffered social isolation and had not learnt language before puberty. Their inability to learn language after that time, Lenneberg argued, constituted evidence that language acquisition was impossible after the critical period.

However, the extant evidence of language learning by feral children is so scarce that it cannot be used to provide strong support for either this or opposing views. The case of Genie, a girl who had lived in social isolation until the age of 13;7, proved that language acquisition was not impossible after puberty, though it seemed to be incomplete (see Curtiss, 1977). Linguistic evidence from the Genie case is still in need of clarification (Jones, 1995), and the regression she suffered after several traumatic events underlines the fact that the social and psychological circumstances of feral children cannot constitute valid evidence for a firm conclusion regarding the critical period for language acquisition. Likewise, the pathological kind of evidence provided by aphasic patients needs to be treated with caution. More relevant evidence has recently come from the field of the acquisition of sign language, which suggests that morphology and syntax may be affected by late acquisition in the case of deaf persons who are not exposed to their first language (sign language) until later childhood or adulthood (Newport, 1990).

From a theoretical point of view, the idea of a critical period sprang from an innatist conception of language, which the prevalence of Chomsky's proposals in the field of linguistics in the second half of the 20th century strongly reinforced. A biologically determined period for language acquisition fitted perfectly well in a theory that concedes a crucial role to biology in human linguistic competence. Recently, however, the field of child language acquisition seems to be drifting away from formal linguistics proposals, disappointed by their failure to explain how human children become skilled users of a natural language. Tomasello (2003) argues that one of the reasons for this failure lies in the continuity hypothesis, which attempts to explain children's language in terms of the structures and rules used to account for adult language. The best known theoretical alternatives to generative grammar at the moment, the connectionist accounts (e.g. Elman, 2001) and the construction-based (usage-based) accounts (see Tomasello, 2003), are both data based. In the latter perspective, one of the reasons underlying children's observed advantage in second language acquisition may be the fact that they are more flexible learners than adults in skilled activities.

The CPH and Second Language Acquisition

The study of second language acquisition originated from the field of first language acquisition, and has since been fed by hypotheses and theories first developed in the parent field. Among these, the hypothesis of the existence of a critical period for (first and second) language acquisition soon motivated a wealth of empirical studies in the 1970s. The work of that decade was summarised in the following three generalisations:

(1) Adults proceed through early stages of syntactic and morphological development faster than children (where time and exposure are held constant).

(2) Older children acquire faster than younger children (again, in early stages of syntactic and morphological development where time and exposure are held constant).

(3) Acquirers who begin natural exposure to second language during childhood generally achieve higher second language proficiency than those beginning as adults. (Krashen et al., 1979/1982, reprint: 161)

These generalisations led Krashen et al. (1979) to make a very important distinction between ultimate attainment and rate. Older learners have a superior learning rate, particularly in the first stages of the acquisition of morphosyntactic aspects, while younger learners are slower at first, but eventually show a higher level of ultimate attainment. The latter was held to constitute evidence for the existence of a critical period, beyond which second language acquisition cannot reach native-like levels of proficiency.

Since then, a large number of studies have compared native-likeness among younger and older starters. The most robust evidence for the existence of maturational constraints in second language acquisition seemed to be provided by the study by Johnson and Newport (1989) of Korean and Chinese learners of English using grammaticality judgement tests (Long, 1990). Many other studies have also focused on the acquisition of second language (L2) morphosyntax (e.g. Coppieters, 1987; DeKeyser, 2000; Johnson &s Newport, 1991; Patkowski, 1980; Schachter, 1996).

However, agreement is far from complete. Johnson and Newport's findings have been questioned on both methodological and empirical grounds (but see a recent confirmatory study in DeKeyser, 2000). Methodological criticisms can be found in Bialystok (1997), and Bialystok and Hakuta (1999). Replications of Johnson and Newport's (1989) seminal study have cast some doubt on the neurobiologically-based explanation of the younger learners' advantage. For example, replications have found evidence of native language effects...