Vol. 3. No. 2 R-8 March 1998

Return to Table of Contents Return to Main Page

In Tutorials in Bilingualism, de Groot and Kroll give three reasons for the expanding interest in the cognitive study of bilinguals. First, bilingualism is increasingly the norm and hence any genuine account of human cognition needs to include a detailed understanding of language use not only by pure monolinguals but also by balanced and unbalanced bilinguals. The study of bilingualism likewise allows researchers to investigate the relation between language and thought and whether the thought processes of bilinguals respond to the language they are using. Finally, an understanding of the processes of bilingualism can help improve accessibility for those whose culture does not historically support the use of more than one language but who seek the benefits of broadening their scope beyond their own country and culture. The scope of Tutorials in Bilingualism is thus broad and provides the reader with insights into a rich range of research traditions. To help clarify this complex field, contributions are divided among three sections: second language acquisition (SLA), comprehension and production in two languages, and cognitive aspects of bilingualism.

Part 1 is devoted to second language acquisition and begins with a review by Harley and Wang of work on the critical period hypothesis, acknowledging not only the complexity of acquisition research but also the diverse positions taken by those working in the field. Harley and Wang also tease apart the differences between first language (L1) and second language (L2) acquisition and raise the interesting question of whether early onset L2 acquisition (in which the L1 is replaced by the L2) might not better be seen as a continuation of L1 acquisition.

In chapter 2 Ellis and Laporte tackle the formal instruction/natural exposure debate. Their review considers the role in SLA of negative evidence, implicit and explicit learning, formal instruction, focusing learners' attention on grammar, and output practice. For each issue, the authors briefly review the theories underpinning the debate and then summarise key field and laboratory research findings. This selectivity makes for readability in a complex area in which the research findings are often contradictory. The authors conclude that there are roles for the provision of negative evidence and recasts, for explicit instruction (particularly when it involves grammatical consciousness raising), and for output practice. As the research has looked at each factor in isolation, however, they [-1-] wisely warn that such "findings do not imply exclusivity of cause in real-world SLA" (p. 78).

In the third chapter, Segalowitz seeks an answer to the question of why individuals differ so much in second language attainment. He reviews three perspectives and then proposes his own framework, which assumes that three basic premises underpin these differences: (a) learners may vary with regard to the flexibility and fluency required to deal with an open skill such as language, with its evolving linguistic and social environment; (b) L2 performance involves sensitivity to environmental affordances, that is, the ability to adjust to the changing linguistic and nonlinguistic context, to perceive its functional relationship to the speaker, and to construct the communication in accordance with this evolving, dynamic context; (c) internal sources of interference, such as transfer-appropriate learning and demands on capacity and information cross-talk, affect performance attainment. The value of Segalowitz's framework is that it puts the learner "in a complex, dynamic, communicative environment" (p. 107). Thus, while taking a psycholinguistic perspective, learning is seen as a complex reflection of the whole learning situation, an interaction between individuals and the context in which they find themselves.

In the final chapter in part 1, MacWhinney breaks with mainstream SLA theory and suggests that both first and second language acquisition can be explained in terms of constructive data-driven processes that rely on universals of cognitive structure rather than Universal Grammar. This discussion of the Competition Model does not seem particularly well integrated with the preceding three chapters, and closer discussion of the model's relationship with other SLA perspectives would have been useful. MacWhinney himself recommends taking his case for an empiricist position with a healthy grain of salt, while suggesting that viewing competing positions in their most undiluted form may help navigation between alternatives. This may be true, but it may also encourage the turf wars which have bedeviled SLA research and sometimes led to a narrowing of the research focus and an ignoring of insights from related disciplines. Indeed, one of the real strengths of this collection is precisely its attempt to draw together different areas of research.

Part 2 looks at representation, comprehension and production in two languages. In chapter 5, Marilyn Chapnik Smith reviews the literature on bilingual access to lexical information. She assumes a model of word recognition with three distinct levels of representation: orthographic, lexical, and semantic. The research presented--mostly laboratory studies involving the presentation of unrelated lists of words--suggests that bilinguals have two lexicons, one for each language, and that both are activated unless the context limits the search to only one. In the next chapter, Kroll and de Groot assume a similar hierarchical model and explore how words and concepts are represented in bilingual memory. A [-2-] frustration with the fascinating research reviewed is the deliberate use of out-of-context tasks, which seem to ignore the key role context may play both in establishing and retrieving meaning. The authors are aware of this but argue that conceptualising the mapping of word forms to meaning in two languages in such tasks may provide a framework for understanding performance in contexts more representative of second language use. It is also possible, however, that these tasks involve different processes that are rarely invoked in actual communicative interaction.

Poulisse in chapter 7 reviews a number of models that look at language production in bilinguals and try to account for the entire speech production process, from the conceptualisation of messages to their articulation. Bilingual language production differs from L1 production in three key ways. First, bilingual speakers are usually working from an incomplete knowledge base. Second, they lack the L1 speaker's automaticity and so production tends to be more hesitant and contain more slips of the tongue. The final difference is that the speech of L2 speakers may be mixed, either intentionally or unintentionally. It is this possibility for mixing and the ability to separate that is problematic for existing monolingual models of language production. Poulisse reviews five current models and the questions to which they still give rise.

Grosjean begins his discussion of processing mixed languages by pointing out how much more complex studies of bilingualism are than studies of monolingualism. Factors which may influence outcomes--and explain conflicting or inconclusive results--include the subjects themselves and the status of their bilingualism, the language modes in which they are functioning, the stimuli and tasks proposed, and the models underpinning the research. Production studies show that the topic and person addressed are key factors in determining where speakers find themselves on the language mode continuum and hence how much they codeswitch or borrow. Grosjean then explores the highly complex factors involved in the recognition of guest words, suggesting these include monolingual factors (context and word properties) and bilingual factors (listener factors such as fluency, language mode, and so on). In the final section of the chapter Grosjean presents his own interactive activation model, in which he assumes that bilinguals have two independent and interconnected language networks. The elements of each language form separate networks of connections, a subsystem within a larger system. In the monolingual language mode one language network is strongly activated, the other weakly. While not proving that there is no language switch or monitor mechanism, this model does show how the processing system could do without one.

In the final chapter in part 2, Durgunoglu explores the cognitive components of bilingual literacy. Literacy, whether in L1 or L2, requires the activation and orchestration of linguistic knowledge, literacy knowledge, and background knowledge. In bilinguals, [-3-] however, the number of knowledge sources is doubled, as all three are available in both languages. A key issue, therefore, becomes what transfers from L1 to L2 reading. There is some evidence that metalinguistic and metacognitive awareness may transfer more easily because they are relatively language-independent. Durgunoglu argues also that bilingual literacy cannot be studied purely from a cognitive perspective, because literacy acquisition and development are strongly affected by sociocultural contexts and political concerns. For this reason, also, it is important that research across different bilingual readers be integrated.

The third section of Tutorials in Bilingualism focuses on the consequences of bilingualism for thought and for special forms of language processing. In chapter 10, Cook distinguishes between the monolingualist view that one language is the norm for human beings and the multilingualist view according to which the monolingual is deficient, deprived of the natural ability to speak more than one language. Cook then examines the evidence available to support the monolingualists' subtractive view and the multilingualists' additive view of bilingualism, and concludes there are indeed both costs and gains. He then questions whether such comparisons are appropriate. If multilingualism were the norm, research would focus on why monolinguals do not attain the normal potential of a human being rather than looking at the deficiencies of bilinguals. Because the choice of norms is rooted in preconceptions about the nature of human beings and of human society, this then becomes an ideological issue.

In chapter 11, Dufour reviews studies which have looked at sign language and bilingualism and concludes that the differences between sign and spoken languages are such that current theories of bilingual language representation may need to be expanded to enhance the similarities and differences in the mental architecture between spoken forms and bilingualism that involves sign languages.

In the final chapter, Paradis looks at the cognitive neuropsychology of bilingualism, and argues that factors such as degree of proficiency, manner of acquisition, degree of affective involvement, context of use, and structural differences between the languages spoken may all influence the representation and/or processing of a bilingual's languages. Paradis concludes that the two languages in the brains of bilinguals are neurofunctionally independent, although their representation is not anatomically distinct and does not show evidence of differential lateralisation. In non-brain-damaged speakers, each language is connected to, and interacts with, the conceptual system, with one language assumed to be inhibited when the other is in use. Both are more or less easily activated according to frequency and recency of activation.

As this brief summary suggests, Tutorials in Bilingualism covers a broad range of issues not usually brought together. The reader is [-4-] thus provided with a clear introduction to the complexity and controversy surrounding research in this area. While the emphasis is very much on the psycholinguistic perspectives of the subtitle, the role of context emerges in several chapters and in the underlying richness of the somewhat different European and American perspectives presented. Readers new to the field might wish for tighter cross-referencing among the contributions, but are still provided with a very useful understanding of how research in this area is developing and being integrated.

Jane Crawford
Queensland University of Technology
<j.crawford@qut.edu.au>