Bilingualism in Education

Aspects of theory, research and practice

Jim Cummins, Merrill Swain

• Publisher: Routledge
• Year: 2014
• ISBN:
  9781317869177
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Linguistics
• Index: Languages & Linguistics

This is a remarkably interesting and useful book...it makes a significant contribution to our knowledge and understanding of both bilingualism and education.'
Journal of Education Policy

Part One
The educational development of bilingual children

Section 1
Metalinguistic and cognitive development

This section is concerned with the cognitive and metalinguistic development of bilingual children. Both in the research and public -media literature, controversy about the relationship between bilingualism and cognitive functioning exists. Some have argued that bilingualism will necessarily have a negative effect on cognitive development because having, for example, two labels for each concept will be confusing and result in retarded conceptual development. Others have argued that this very same phenomenon will enhance cognitive growth: having two labels will force an early separation of word from its referent. This early separation of linguistic form from meaning, it is argued, will lead to a more analytic orientation to language and to the substance it conveys, thus enriching conceptual development. From this viewpoint, one would predict, for example, that a bilingual individual would be better able to detect linguistic ambiguities than would a unilingual individual, or be better able to restructure conceptual schemata.

In summarizing the research literature addressing this question, McLaughlin (1984) states that: ‘It seems clear that the child who has mastered two languages has a linguistic advantage over the monolingual child. Bilingual children become aware that there are two ways of saying the same thing.’ (p. 214). However, he goes on to raise the issue of whether this sensitivity to the lexical and formal aspects of language generalizes to cognitive functioning.

As Chapter 1 indicates, there is evidence to support both the position that bilingualism is negatively associated with metalinguistic and cognitive development and the position that it is positively associated with them. Chapter 2 provides a concrete example of a study which examines the relationship between bilingualism and metalinguistic development. The study is included in order that readers can see first-hand the nature of the studies which underlie the summary statements found in Chapter 1. This study, like many others discussed in Chapter 1, has its methodological weaknesses.

Indeed, one likely reason for the contradictory findings - of discovering both positive and negative associations between bilingualism and metalinguistic and cognitive development - lies in the methodology of the research. For example, the dependent variables which have been used in the studies, all of which purport to measure an aspect of metalinguistic or cognitive functioning, are themselves quite diverse. Measures have ranged from detection of structural ambiguities in isolated sentences to performance on Piagetian tasks and standardized tests of intelligence. It is not at all clear what the variety of dependent measures which have been used have in common. An understanding of task commonalities awaits theoretical advances (see, for example, Bialystok and Ryan (in press)).

Similarly, the independent measure - bilingualism - has been variously measured. Measurement of bilingualism has ranged from an examination of the surnames of the students tested to a comparison of performance on language proficiency tasks in both languages. Recent studies have tended to select as participants in the experimental (bilingual) group ‘balanced bilinguals’, that is, those bilingual individuals who perform at similar levels of proficiency on language tasks in both their languages. This also, however, has problems associated with it. It may be, as Macnamara (1970) and MacNab (1979) have argued, that balanced bilinguals are a unique class of bilinguals, who, because they are cognitively more able in the first place are more likely to become highly proficient bilinguals. In other words, the question of the direction of the relationship between bilingualism and cognitive and metalinguistic development is left begging. Is it that bilingualism affects cognitive development or conversely, that cognitive development affects bilingualism?

Research can play a significant role in addressing this issue. To address the question directly, it will need to have several characteristics. It will need to be longitudinal in design. In Chapter 1 there is only one study with this characteristic (Barik and Swain 1976a). In this study, it was found that aspects of the cognitive development of primary level immersion students with a relatively high level of second language proficiency increased significantly more over a period of several years than they did for students with a lower level of second language proficiency. This finding has subsequently been replicated by Harley and Lapkin (1984) in a study spanning six years of bilingual and cognitive growth.

Hakuta and Diaz (1984) carried out a study using a similar research design and longitudinally collected data. Using a non-verbal measure of cognitive development and Spanish and English vocabulary measures to determine degree of bilingualism, their results indicated that degree of bilingualism and cognitive ability were related at two points in time. Furthermore, when two alternative models testing for the direction of causality between bilingualism and cognitive ability were examined using the longitudinal data, ‘the model claiming degree of bilingualism to be the causal link was more consistent with the obtained data than the model claiming cognitive ability to be the causal variable5 (p. 340).

Since Chapter 1 was written, only a handful of studies employing a longitudinal design have appeared in the literature. One such study is that carried out by Bain and Yu (1980) who examined the cognitive consequences of raising children unilingually and bilingually. Their measures of cognitive development followed from the theoretical/empirical research paradigm of Vygotsky (1962) and Luria (1961) who suggest that the development of children’s ability to control their own cognitive processes is contingent upon their mastery of language. In the Bain and Yu study, bilingual and unilingual children living in Alsace, in Alberta or in Hong Kong were tested at ages 22 to 24 months, and again at ages 46 to 48 months. No differences were initially found between the bilingual and unilingual children. However, at the time of the second testing some two years later, significant differences were found in favour of the bilingual children.

From a methodological viewpoint, the Bain and Yu study still suffers from one problem: the children were not randomly assigned to experimental and control groups. Essentially they ended up in the particular group they did because their parents had ‘volunteered’ them: the parents had answered an advertisement asking for their participation in a study of speech acquisition. In order to participate in the study of bilingual speech acquisition, one but preferably both of the parents had to be bilingual. As MacNab (1979) points out, there may be differences between the families of these children that will be more powerful in explaining differences than is bilingualism itself: ‘To learn a second language is a commitment to a second culture, and people who learn to speak two languages are, therefore, very likely to be quite different from those who stay unilingual.’ (p. 243).

Achieving random assignment of children to bilingual and unilingual groups, however, is a difficult, if not impossible, methodological demand given the reality of bilingual development. One solution to this problem is to look at effects within a bilingual sample: ‘If degree of bilingualism can be reliably measured within a sample of children becoming bilingual and if this measure of degree can be shown to be related to cognitive flexibility, then one would have come one step closer to finding a pure relationship between bilingualism and cognitive flexibility.’ (Hakuta and Diaz 1984, p. 330). This is essentially the research design used by Barik and Swain (1976a), Harley and Lapkin (1984) and Hakuta and Diaz (1984) to reach their conclusions noted above.

We have gone to some pains to indicate some of the methodological problems associated with research concerned with the association between bilingualism and cognitive growth in order to suggest that some of the contradictory findings can be accounted for by them. However, having said this, we suggest that finding both negative and positive cognitive consequences associated with bilingualism is not necessarily contradictory. As will be seen again in Section 2, an underlying theoretical principle can link together the disparate findings. In this section, the key theoretical hypothesis to be advanced is the ‘threshold hypothesis’.

The threshold hypothesis proposes that there may be threshold levels of linguistic competence which bilingual children must attain in their first and second languages both in order to avoid cognitive disadvantages and to allow the potentially beneficial aspects of becoming bilingual to influence cognitive functioning. In other words, two thresholds are hypothesized: one below which cognitive growth would suffer without further linguistic development: and one above which cognitive growth would be enhanced. As we point out in Chapter 1, one implication of the threshold hypothesis is that educational programmes should aim to foster high levels of proficiency in both languages. In societal situations where there is likely to be serious erosion of the first language, it is particularly crucial that school programmes aim toward its development and maintenance. As we will argue in Section 2, such an educational programme will support second language development as well.

The threshold hypothesis has proved a useful heuristic device for making sense of seemingly contradictory evidence. It has provided a useful framework for interpreting the results. Obviously, however, what actually constitutes threshold levels needs to be further specified. Diaz (in press) notes that when first language proficiency is high, enhanced cognitive effects are associated with relatively low levels of second language proficiency. Diaz suggests that the positive effects of bilingualism may be related ‘to the initial efforts required to understand and produce a second language rather than to increasingly higher levels of bilingual proficiency’. This presents an intriguing new perspective on the notion of a threshold level.

1
Bilingualism, cognitive functioning and education^*

The purposes of this chapter are to review recent studies which have investigated the relationships between bilingualism and cognitive functioning, and to outline the implications of these research findings for educational settings.

Terminology

The term ‘bilingualism’ has not been used in a consistent way among researchers and theoreticians. Definitions vary considerably. Macnamara (1967), for example, defines bilinguals as those who possess at least one of the language skills (listening, speaking, reading and writing) even to a minimal degree in their second language. At the other end of the scale, bilinguals have been defined as those who demonstrate complete mastery of two different languages without interference between the two linguistic processes (Oestreicher 1974) or who have native-like control of two or more languages (Bloomfield 1933). The tendency has been to focus on speaking and listening skills (e.g. Haugen 1953; Pohl 1965; Weinreich 1953).

Other definitions of bilingualism have considered the age at which the second language is learned (simultaneous versus sequential; early versus late)\ the contexts in which the two languages have been learned (compound versus coordinate (Osgood and Sebeok 1965), artificial versus natural (Stem 1973)); or the domains in which each language is used (e.g. Fishman 1968; Oksaar 1971).

It is clear, then, that there is little consensus as to the exact meaning of the term bilingualism, and that it has been used to refer to a wide variety of phenomena. Research associated with bilingualism reflects this semantic confusion. It is essential, therefore, in reconciling contradictory results associated with bilingualism, to be aware of the levels of bilingualism attained by the experimental students, and the social and psychological factors which lie behind the particular ‘bilingualism’ attained.

This review is specifically concerned with the association between bilingualism and cognitive functioning. Cognitive functioning is used in this chapter to refer to measures involving general intellectual and linguistic skills such as verbal and non-verbal IQ, divergent thinking, academic performance and metalinguistic awareness.

For the purposes of this review it is also important to introduce two terms associated with educational programmes. Both terms -‘immersion’ and ‘submersion’ - relate to situations where the child is required to use in school a language that is different from that used in the home. Immersion refers to a situation in which children from the same linguistic and cultural background who have had no prior contact with the school language are put together in a classroom setting in which the second language is used as the medium of instruction. Submersion, on the other hand, refers to the situation encountered by some children wherein they must make a home-school language switch, while others can already function in the school language. Within the same classroom, then, one might find children who have no knowledge of the school language, varying degrees of facility in the school language through contact with the wider community, and native speakers of the school language (Swain 1978c).

In the next two sections, studies which report a negative association between bilingualism and cognitive functioning and studies which report a positive association between bilingualism and cognitive functioning will be reviewed. Following that, several of the factors which appear to differentiate the positive and negative studies will be reviewed, and the resultant implications for educational programmes will be summarized.

Studies reporting negative associations

Several comprehensive reviews exist of studies conducted prior to 1960 (see Darcy 1953; Macnamara 1966; Peal and Lambert 1962) and consequently these studies will not be considered here. Although the majority of these early studies had serious methodological defects, taken together they seemed to indicate that bilinguals suffered from a language handicap when measured by verbal tests of intelligence or academic achievement.

The general findings of earlier studies are supported by those of Skutnabb-Kangas and Toukomaa (1976) who reported that children of Finnish migrant workers in Sweden tended to be characterized by ‘semilingualism’, i.e. their skills in both Finnish and Swedish (as measured by standardized tests) were considerably below Finnish and Swedish norms. The extent to which the mother tongue had been developed prior to contact with Swedish was strongly related to how well Swedish was learned. Children who migrated at the age of ten maintained a level of Finnish close to Finnish students in Finland and achieved Swedish language skills comparable to those of Swedes. However, children who were seven to eight years of age when they moved to Sweden or who moved before starting school were most likely to achieve low levels of literacy skills in both languages. Skutnabb-Kangas and Toukomaa argue on the basis of these results that the minority child’s first language (LI) has functional significance in the developmental process and should be reinforced by the school.

Vernacular education is also supported by Macnamara’s (1966) study of bilingualism in Irish primary education. Macnamara reported that Irish primary school children, whose home language was English but who were instructed through the medium of Irish, were eleven months behind in problem arithmetic relative to other Irish children taught through the medium of English. No differences were found between the Irish immersion g...

Table of contents

• Cover Page
• Half Title page
• Series Page
• Title Page
• Copyright Page
• Acknowledgements
• Contents
• Preface
• Part One The educational development of bilingual children
• Part Two Bilingual proficiency