Age and the Acquisition of English as a Foreign Language

María del Pilar García Mayo, María Luisa García Lecumberri

• Publisher: Multilingual Matters
• Year: 2003
• ISBN:
  9781788920384
• Topic: Languages & Linguistics
• Subtopic: Literacy
• Index: Languages & Linguistics

This book provides an overview of current research on the age factor in foreign language learning, addressing issues, which are critical for language planning. It presents new research on foreign language learning within bilingual communities in formal instruction settings focussing on syntax, phonology, writing, oral skills and learning strategies.

Part 1

Theoretical Issues

Chapter 1

Critical Period or General Age Factor(s)?

DAVID SINGLETON

Introduction

The question of whether there is an age factor in language development is perennially a topic which attracts wide interest and generates fierce debate. The reasons why it continues to be so energetically discussed are both theoretical and practical in nature. On the theoretical front there is an interaction between the notion of maturational constraints on language acquisition and the idea that language development is underpinned by special bioprogramming; and on the practical level the claim that younger L2 beginners have an advantage over older beginners is constantly invoked and disputed when decisions are being taken about the optimal starting point for L2 instruction in schools.

In fact, few L2 researchers now question the proposition that those learners whose exposure to the L2 begins early in life (and whose exposure to the language is substantial) for the most part eventually attain higher levels of proficiency than those whose exposure begins in adolescence or adulthood. The question that continues to divide the field, however, is whether age effects in L2 acquisition constitute a manifestation of a preprogrammed critical period specifically related to language or whether they reflect other, more general, factors which may militate against the learning of new skills and which happen to be concomitants of increasing age. The present chapter addresses this question first by looking at some early work on the age factor, second by looking at the notion of the critical period and some relevant evidence in respect of L1 acquisition, third by examining the L2 evidence for three different interpretations of the Critical Period Hypothesis bearing on L2 acquisition, and finally by exploring some explanations of age effects which do not rely on the idea of a critical period for language.

Anecdote and Assumptions in the Genesis of the Critical Period Hypothesis

Whereas in recent times the issue of maturation and language acquisition has been addressed with a high degree of empirical rigour, in the past discussion of this matter (especially in an L2 context) was largely based on anecdote and assumption. For example, Tomb(1925) referred simply to the 'common experience' of hearing English children in Bengal (in the days of the British Raj) fluently conversing in English, Bengali, Santali and Hindustani with various members of the household, while their elders had scarcely enough Hindustani to give instructions to the servants. Stengel (1939), for his part, proposed a highly sophisticated Freudian analysis of the age factor in L2 learning but, again, his ideas about children's language learning were based simply on impressionistic observation.

Science appeared to enter the picture in the 1950s, when the neurologist Penfield took an interest in the discussion. Penfield cited evidence (Penfield & Roberts, 1959: 240) suggesting that children are normally able to re-learn language when injury or disease damages speech areas in the dominant language hemisphere(usually the left), whereas speech recovery in adults is much more problematic, and that whereas in young children the speech mechanism is often successfully transferred from the injured dominant hemisphere to the healthy minor hemisphere, such transfers do not seem to occur in the case of adults. He used such evidence as a basis for asserting that 'for the purposes of learning languages, the human brain becomes progressively stiff and rigid after the age of nine' (Penfield & Roberts, 1959: 236). He went on to advocate that children should be introduced to L2s early in life, asserting that 'when languages are taken up for the first time in the second decade of life, it is difficult . . . to achieve a good result' (p. 255). However, despite the fact that Penfield made much in his writings of the 'unphysiological' nature of language learning beyond the childhood years, thus implying that his advocacy of early L2 instruction was firmly rooted in his neurological expertise, in fact, as Dechert (1995) demonstrates, his views in relation to L2 learning owed more to his personal experience (successful, he claimed) of immersing his own children in FLs at an early age than to his work as a scientist.

Penfield's notion about the 'unphysiological' nature of later language learning was very much echoed in the work of Lenneberg, the person who is generally acknowledged as the 'father' of the Critical Period Hypothesis relative to language acquisition. Lenneberg saw the human capacity for language acquisition as constrained by a critical period beginning at age two and ending around puberty, this period coinciding with the lateralization process – the specialisation of the dominant hemisphere of the brain for language functions. He adduced a wide range of evidence pointing to changes in the brain that were occurring during this period. However, his claim that lateralisation ends at puberty has been significantly undermined by later studies which reinterpret the data in question as indicating that the process is already complete in early childhood (see, e.g., Kinsbourne & Hiscock, 1977; Krashen, 1973). Moreover, that part of Lenneberg's argument which referred to L2 learning, namely his suggestion that after puberty the learning of L2s required 'labored effort' and foreign accents could not be 'overcome easily' (Lenneberg, 1967: 176) was of dubious status in scientific terms. While his arguments in relation to the maturation of the brain development were supported with a range of neurological evidence (some of which has, as has been noted, since been reinterpreted), no evidence was offered in respect of his claims regarding post-pubertal L2 learning, which relied instead simply on an implicit appeal to popular assumptions.

The Concept of Critical Period

Before we proceed further in our discussion of the Critical Period Hypothesis (henceforth CPH), it may be worth reminding ourselves how the concept of critical period is usually understood in the biological sciences. The example that is usually cited in this connection is that provided by Lorenz (1958), who noted that new-born goslings became irreversibly attached to the first moving object they perceived after hatching. Usually, the first moving object in question is the gosling's mother. However, any other moving object will trigger the relevant reaction if it comes into the gosling's line of vision in the post-hatching period. The period during which the attachment of this kind may be effected is limited in duration, and beyond that period the gosling will no longer fix its following behaviour in the way described. Indeed, when this particular period ends, goslings will retreat from rather than follow moving objects.

Critical periods in biology can, on this basis, be characterised as follows.

(1) They relate to very specific activities or behaviours.

(2) Their duration is limited within well-defined and predictable termini.

(3) Beyond the confines of the period in question the relevant behaviour is not acquired.

If language acquisition in human beings is rigidly constrained by the limits of a critical period of this kind, the implication is that L1 development begins only at the onset of this period and that unless it gets under way during the period in question it will not happen at all. A further implication may be that even if L1 development begins within the critical period it does not continue beyond the end of that period.

The CPH in respect of L1 acquisition is not, one has to say, particularly well supported by the available evidence. With regard to the starting point of the critical period, Lenneberg (1967: 155) claims that whereas 'children deafened before completion of the second year do not have any facilitation [in relation to oral skills] in comparison with the congenitally deaf', those who lose their hearing after having been exposed–even for a short time–to the experience of [oral] language subsequent to this point 'can be trained much more easily in all the [oral] language arts' (p. 155). He interprets this as indicating that the critical period is to be seen as beginning around the age of two years. However, his own synthesis of the language acquisition timetable undermines his interpretation; thus, his summary of development between 4 and 20 months is 'from babbling to words' (p. 180). Also, research relative to the acquisition of phonology suggests that there is no sharp break in the developmental progression from prespeech to speech (see, e.g., Stark, 1986: 171), and research into conceptual and lexical development indicates that comprehension of linguistically mediated communicative functions is established early in the second half of the child's first year (see, e.g., Griffiths, 1986). Lenneberg's own evidence is somewhat vague and anecdotal in nature and bears interpretations other than the one he proposes (see, e.g., Singleton, 1989: 44).

In relation to the notion that unless L1 development begins during the critical period it will not happen at all, two kinds of evidence have been cited in this connection: evidence from'wolf-children'–children who have grown up in isolation from normal human society and who have then been rescued – and evidence from the late acquisition of sign language. Some wolf-children have been brought into contact with language only around the age of puberty, the point at which some researchers, including Lenneberg, claim the critical period for language acquisition ends. Two much discussed cases of this kind are those of Victor (see e.g., Lane, 1976), and Genie (see, e.g., Curtiss, 1977; Rymer, 1993). The problem with such instances is that the evidence is extremely difficult to interpret, especially since there is often a deficiency of information about the child – the precise amount of exposure to language he/she experienced, the extent of the trauma induced by his/her experience, etc. The typical pattern is that some post-rescue progress in language development is observed – but of a limited and abnormal kind. Some researchers see this as 'first language acquisition after the critical age of puberty' (De Villiers & De Villiers, 1978: 219), while others consider that it indicates 'specific constraints and limitations on the nature of language acquisition outside of . . . the critical maturational period' (Curtiss, 1977: 234). Interestingly, Lenneberg himself is sceptical about evidence from this kind of source, his comment being that all one can conclude from such cases is that 'life in dark closets, wolves' dens, forests or sadistic parents' backyards is not conducive to good health and normal development' (Lenneberg, 1967: 142).

It has been suggested that the clearest evidence for a critical period in respect of L1 development comes from studies of deaf subjects who are deprived of language input in their early years and who then acquire sign language as their L1 at a later stage. Long (1990: 258f.) cites a number of studies indicating that the later acquisition of sign language as L1 is characterised by deficits of various kinds. Thus Woodward (1973) found that some American Sign Language (ASL) rules were acquired more often by individuals exposed to sign before age 6, and Mayberry et al. (1983) note similar long-term advantages for individuals who began acquiring ASL in childhood over adult beginners have been reported for the processing of ASL. Likewise, the study of 'Chelsea', a deaf adult who began acquiring ASL as a first language in her early thirties reveals good lexical and semantic abilities after six years of exposure but impaired morphology and syntax (Curtiss, 1988), and a large-scale project reported in Newport (1984) and Newport & Supalla (1987) shows that late/adult learners first exposed to ASL after age 12 fell far short of native standards in their signing (cf. also Neville et al., 1997). Two observations come to mind in relation to such research. First, the studies in question do not appear to indicate that language completely fails to develop after a given maturational point, which is what one might expect in the case of a critical period for language. Second, with regard to the abnormalities and deficits recorded in late beginners, it is now clear (see, e.g., Peterson&Siegal, 1995) that deprivation of language input during the phase in a child's life when cognitive development is at its most intense has quite general psychological/cognitive effects, and it may be these general effects that are reflected in later language development (see, e.g., Lundy, 1999) rather than effects relating specifically to a critical period for language.

Finally on the subject of an upper age limit to L1 development, a strong version of the CPH would predict that even if such development starts within the critical period, the process does not continue beyond the end of this period. In a three-year observational study of 54 Down's syndrome subjects Lenneberg et al.(1964)were able to record progress in language development only in children younger than 14. This is taken by Lenneberg (1967: 155) to indicate that 'progress in language learning comes to a standstill after maturity'. Alternative interpretations of these data (see Singleton, 1989: 58f.) are (1) that what Lenneberg and his colleagues were observing was a general developmental cut-off point (widely reported in the literature on mental retardation); (2) that what was involved was not a complete arrestation but a temporary plateau (which is again referred to in the literature on mental retardation); and/or (3) that the halt in progress was due to the absence of the right kind of stimulation. In any case, with regard to normal L1 development, there are ample indications that this continues well past puberty. For example, in a study of L1 morphology in Dutch between ages 7 and 17, Smedts(1988) found that his 7-year-old subjects displayed a mastery of, on average, only 14% of a range of Dutch morphology, that his 13-year-olds knew just 51% of the rules tested, and that even his 17-year-olds demonstrated a command of no more than 66% of these rules. In fact, at least some aspects of L1 development extend well into adulthood. Thus, for instance, Carroll (1971: 124) concludes from his review of a number of lexical studies that L1 vocabulary tends to increase significantly up to at least the age of 40 or 50, while Diller (1971: 29) reports research which suggests that there is no point before death at which L1 vocabulary acquisition can be predicted to cease (cf. also, e.g., Singleton, 1989: 54–8).

With regard to L2 acquisition, the way in which the CPH is interpreted varies according to researchers' theoretical predispositions. Three commonly advanced views – which are not incompatible but which are advocated to different extents by different schools of thought – are the following:

(1) after a certain maturational point the L2 learner is no longer capable of attaining native-like levels of proficiency;

(2) after a certain maturational point successful L2 learning requires markedly more effort than before this point; and

(3) after a certain maturational point L2 learning is no longer subserved by the same mechanisms that subserve child language acquisition.

Since, however, the notion of a critical period inherently carries with it a claim regarding a marked qualitative change in learning capacity at a particular stage of maturation, all interpretations of the CPH predict that at the maturational stage in question a sharp decline in L2 learning potential will be observable (which is of its nature different from the more gradual age-related declines in the organism's general learning capacity). The sections that follow will address these issues with reference to rel...

Table of contents

• Cover
• Title page
• Copyright
• Contents
• Introduction
• Part 1: Theoretical Issues
• Part 2: Fieldwork in Bilingual Communities