Why is it so difficult for some children to learn second-languages in school and why do some children, who seem to have learned the language, nonetheless do poorly in their academic work? These questions have long puzzled educators, but they have become more salient in many parts of the world as millions of children enter school each year with limited knowledge of the language of instruction. In Europe and the United States alone it is estimated that ten million children come from families where the language of the home is different from the language of the country in which they live. In many parts of Asia and Africa it is assumed that an educated person will know one or more languages beyond the language of the home. Yet learning a second language in school is a slow and tedious task for most children – some of whom never succeed in becoming bilingual (Eckman 21). One reason why children find learning second languages in school difficult may be that they are taught in the wrong way. This is a perennial argument made by educational innovators who periodically announce some new technique that will revolutionize language teaching in school. Indeed, the teaching of second languages is a long and fascinating tale. A brief review of various developments will be presented in this paper. Another reason why children fail to learn second languages well in school relates to understanding of what it is that they have to learn. There may be more to learning a second language in a school setting (especially when that language is the language of instruction) than simply learning how to speak the language well. In addition to the question of what must be learned (the product of learning), there is the question of how learning takes place (the process of learning). The paper will deal with the cognitive and linguistic processes that the child uses in learning a second language.
The demands of the classroom
What is it that children from minority language backgrounds learning a second language in school need to learn? What language skills are needed for school? What is it that the school demands of the child? How can one characterize the language of the classroom (whether it be a bilingual or a monolingual classroom), and how does this language differ from the language the child has learned to use in natural communication settings?
When the six-year-old child comes to school, he or she has mastered the task of learning to communicate in the first language. The child has learned to use language to express needs and feelings within the social context of the everyday life of the home. The classroom presents a new complexity:
In some sense, all formal education is bilingual, since the forms and ways of expression of written language never reflect the spoken language exactly. Words, ways of speaking and forms of discourse are used in the school setting which are not used in ordinary conversation and in other nonschool settings. The first aim of formal education since its beginnings in the third millennium B.C. has always been to teach the pupils a written form of language (Ferguson, Houghton, & Wells, 1977, p. 159).
Just what are the ways of speaking and forms of discourse used in the school? There have been a number of discussions of this issue and different authors have somewhat different terminology to distinguish the language of the home and the language of school.
What of children who have to learn ‘the language of school’ in a second language? When the language of instruction is different from the language the child speaks, the child has a dual task. It is not simply a question of learning the formal, academic language of school; the child has to do this in a second language. Some children seem to have no trouble with this dual task; others find it very difficult.
One particularly influential hypothesis as to why children succeed or fail in a school where instruction is in a second language is the ‘linguistic interdependence hypothesis,’ proposed by Jim Cummins (1980). According to this hypothesis, the level of competence a child attains in a second language learned in a school context is a function of certain competencies attained in the child’s first language. In particular, Cummins suggested that the use of certain functions of language and the development of vocabulary and concepts in the first language are important determinants of success in a school situation where instruction is in a second language (Eckman 78-80).
In elaborating on this position, Cummins (1980) suggested that there are three general aspects of a child’s knowledge of language that are closely related and that constitute the basic skills that children need to realize positive benefits from a bilingual school experience. The first is what Becker (1977) has called vocabulary-concept knowledge – specifically, the child’s understanding of the concepts or meanings embodied in words. Obviously, if the child does not have any understanding (or a very limited understanding) of the concepts represented in the words on a printed page, reading comprehension is impaired.
A second basic skill involves metalinguistic insights, especially two specific insights: (a) the realization that print is meaningful, and (b) the realization that written language is different from spoken language. The first insight is necessary for the child to be motivated to read; the second helps the child give structure and predictability to written language: Unless the child realizes that written language is different from spoken language, predictions about the meaning of text are likely to be inaccurate.
The third prerequisite is the ability to decontextualize language. That is, the child must be able to take language out of its immediate context. This capacity relates to a considerable extent to experiences the child has had before coming to school. Children who have had the experience of being read to are aware that written language is different from spoken language.
The processes of second language acquisition
What are the mental representations that underpin second language acquisition (SLA)? What is the nature of the mapping processes involved in learning them? To what extent are these representations learned unconsciously? And to what extent are explicit learning or explicit instruction necessary in order to attain native-like competence, fluency, and idiomaticity? SLA in children is the learning of constructions that relate form and meaning. Some constructions and interpretations are much more frequent than others. Fluent speakers of a language implicitly know this and their processing systems are tuned to expect them accordingly. Every element of surface language form is multiply ambiguous in its interpretation, just as every meaning can be expressed in a variety of ways. Fluent language learners are tuned to these mapping strengths: They know implicitly the most likely interpretation of a linguistic cue as well as the relative likelihoods of the range of alternatives and how these change in differing contexts. Their language processing is sensitive to input frequency at all levels of grain: phonology and phonotactics, reading, spelling, lexis, morphosyntax, formulaic language, language comprehension, grammaticality, sentence production, and syntax. Thus, SLA must involve acquisition of the strengths of these associations.
The influence of the first language
The evidence from research on second language learning in children is that connection between languages is not as inevitable or ubiquitous as was once supposed. Contrastive analysis, in its traditional form, was not able to account for the vast majority of errors that second-language learners made; in fact, learners from quite different language backgrounds appeared to make the same types of mistakes in the target language. Research in the early 1970s (especially Dulay ; Burt, 1973) suggested that regardless of their first language, children learning English as a second language made similar kinds of mistakes. If, as contrastive analysis supposed, first-language structures were the major source of a second-language learner’s errors, one would expect that children from such structurally dissimilar first languages as Chinese and Norwegian would make dissimilar mistakes in English. However, the research seemed to indicate that they did not, but instead made the same kinds of errors – errors that were similar to those made developmentally by children acquiring English as a first language.
Subsequent studies (e.g., Wode, 1978) revealed that transfer from the first language does occur in the speech of children from certain first-language backgrounds and at certain times in the learning process. It is an exaggeration to say that transfer from the first language is minimal and unimportant. The acquisition of phonological, syntactic, and morphological structures in a second language involves an interplay of both developmental and transfer factors. Transfer errors do occur and are extremely interesting for the researcher because of what they reveal about the learner’s strategies.
Nonetheless, the influence of the learner’s first language is more indirect and restricted than was once supposed. The evidence suggests that preschool children approach the task of second-language learning in much the same way they approached the task of learning their first language. Some authors speak of the reactivation of children’s facility for language acquisition (Corder 76-90) or of a creative construction process (Dulay ; Burt 89). The idea behind these notions is that children seem to be guided in second-language learning, as in first-language learning, by strategies that cause them to formulate certain types of hypotheses about the language system being learned. They reconstruct the rules for the speech they hear on the basis of these hypotheses, until the mismatch between the target language they are exposed to and their own speech productions is resolved.
SLA is the learning of constructions relating form and meaning
The task of the language learner is to make sense of language. Understanding is built, or falls, depending on the adequacy of the learner’s construction set for meanings. Language construction sets are as infinitely combinatorial and creative as are Lego and Meccano, and as limiting also. Without the right piece, the support buckles and the structure crashes. Without preparatory organization and practice, activity focuses on searching for the right block rather than the process of building itself. Less tangible than plastic or metal, the language learner’s kit consists of constructions that map forms and meanings—the recurrent patterns of linguistic elements that serve some well-defined linguistic function. They may be complex structures, like Lego arches, trucks, or houses. Some frequent, smaller structures, like generic Lego arches, walls, and wheeled axles, are abstract patterns—the noun phrase, the prepositional phrase, and so forth. Others come preformed, like Lego windows, doors, and beams (where kit frequency inversely relates to beam size)—formulas like “how are you?”, “I think I’ll…”, “a great deal of…”, and “survival of the fittest.”
A construction is part of the linguistic system, accepted as a convention in the speech community, and entrenched as grammatical knowledge in the learner’s mind. Constructions may be complex or simple. Hence, “morphology, ” “syntax, ” and “lexicon” are uniformly represented in construction grammar, unlike both traditional grammar and generative grammar. Constructions are symbolic: In addition to specifying the defining properties of morphological, syntactic, and lexical form, a construction also specifies the semantic, pragmatic, and discourse functions associated with it. Constructions form a structured inventory of a speaker’s knowledge of language, in which schematic constructions can be abstracted over the less schematic ones that are inferred inductively by the learner in acquisition. A construction may provide a partial specification of the structure of an utterance.
If language is represented as a community of constructions, induced from exemplars and evidencing classic prototype effects, then the understanding of language acquisition can be informed by classic psychological research on category formation, schema learning, and classification. Construction-based theories of child language acquisition (Tomasello 2000) emphasize the piecemeal learning of concrete exemplars and widespread lexical-specificity in LI grammar development. A high proportion of children’s early multiword speech is produced from a developing collection of slot-and-frame patterns based around chunks of one or two words or phrases (e.g., I can’ + verb; where’s + noun + gone?). Children are very productive with these patterns and both the number of patterns and their structure develop over time. They are, however, lexically specific: A child who consistently uses two patterns, / can’ + X and / don 7 + X, will typically show little or no overlap in the verbs used in the X slots of these two constructions (Tomasello 2000).
Second and foreign language acquisition is different from LI acquisition in numerous respects. First, it differs in conceptual development: In child language acquisition, knowledge of the world and knowledge of language are developing simultaneously, whereas adult SLA builds upon preexisting conceptual knowledge. Moreover, adult learners have sophisticated formal operational means of thinking and can treat language as an object of explicit learning, that is, of conscious problem solving and deduction, to a much greater degree than can children (Ellis 1994). Second, it differs in language input: The typical LI pattern of acquisition results from naturalistic exposure in situations where caregivers naturally scaffold development (Tomasello 2000), whereas classroom environments for second or foreign language teaching can distort the patterns of exposure, function, medium, and social interaction (Ellis 1994). Third, it differs in transfer from LI: Adult SLA builds on preexisting LI knowledge.
People are very good at judging whether or not nonwords are native-like and young children are sensitive to these regularities when trying to repeat nonwords. Phonotactic competence simply emerges from using language, from the primary linguistic data of the lexical patterns that a speaker knows (Bailey & Hahn, 2001). Bailey asked native speakers to judge nonword stimuli for whether they were more or less like English words. The nonwords were created with relatively high or low probability legal phonotactic patterns as determined by the logarithm of the product of probabilities of the onset and rime constituents of the nonword. The mean wordlikeness judgments for these nonword stimuli had an extremely strong relationship with expected probability (r – .87). An emergentist account of phonotactic competence is thus that any new nonword is compared to the exemplars that are in memory: The closer it matches their characteristics, the more wordlike it is judged (Eckman 87). The gathering of such relevant distributional data starts in infancy. Saffran, Aslin, and Newport (1996) demonstrated that 8 month old infants exposed for only 2 minutes to unbroken strings of spoken nonsense syllables (e.g., ‘bidakupado’) are able to detect the difference between threesyllable sequences that appeared as a unit and sequences that also appeared in their learning set but in random order.
Children’s awareness of the sounds of their language, particularly at the segmental levels of phoneme, is important in their acquisition of literacy. It is an awareness that develops gradually. De Cara and Goswami (2002) show that 4- to 7- year-old children are better able to identify the word with the odd sound (e.g., “bag, rag, jack”), rather than when the stimuli came from sparse ones (e.g., “pig, dig, lid”). The children were also better in short-term memory span tasks at remembering nonword triples from dense phonological neighborhoods like “cham, shen, deek” than triples like “deeve, chang, shem” derived from sparse ones. These phonological neighborhood density effects are driven by vocabulary age, not by chronological age. Schmidt (1990) proposed a Lexical Restructuring Hypothesis of these effects whereby, as vocabulary increases, more and more similar words are acquired.
Context and Second Language Acquisition
One incontrovertible fact that scholars in the field of second language acquisition can agree on is that language is not learned in isolation. What is not clear is what the role of context is as learners move forward to learn forms, to leam meanings, and to make the necessary connections between those forms and meanings.
Nearly 22 years ago Gass ; Madden (1985) published the first book dealing with input in second language acquisition. Given the history of the field of second language acquisition, the role of input had never received much attention until that time. In the following years, the role of input, which is of course central to any discussion of context and form-meaning connections, has endured and yet is still not entirely clear. Today’s research world is dealing with greater theoretical and methodological sophistication and a greater focus on psycholinguistic aspects of form-meaning relationships than in the 1980s.
Researchers argue that input is necessary for acquisition, defined as the development of an underlying mental representation. Certainly, input is necessary, but interaction plays an important role for acquisition because it facilitates the attentional link that is crucial to understanding how learners extract information from the environment and use it in the development of their second language grammars. This is very similar to conclusion about the roles of input and of output. Acquisition is input dependent. As stated earlier, there can be no acquisition without input. Output is similarly important, but is not a sine qua non. However, as shown in the study described here, the combination leads to greater learning than either one alone. What seems to be emerging is that there are numerous factors that guide second language acquisition. They can be investigated in isolation and their significance can be determined, but they should also be investigated as interacting and converging factors to truly see how they operate in the learning of a second language for children.
As many have said before, selective attention is a crucial mechanism in the development of second language knowledge. Importantly, it is what links the context with internal learning mechanisms. But part of the controversy as to whether it is essential for learning may be obscured by the fact that no one has looked carefully enough at the contributing factors to attention of which there are many. To just take the examples, it is necessary to understand the what of attention—that is, what parts of language can be attended to and can benefit from focal attention—and to pay attention to the when of attention. For example, is it in conjunction with large doses of input, the socalled input-flooding with regard to positive and negative evidence? Is it the same at all proficiency levels? What sort of input is necessary? Is input through interaction necessary? Is input through specific explanation necessary? But, this too, is only a drop in the bucket. In sum, this paper considers the questions of whether or not context is relevant and begins to examine the nature of the role of the environment. An initial attempt was made to understand what internal mechanisms are necessary for linkage with the linguistic context.