Understanding Second Language Acquisition (2nd ed)

by Rod Ellis

  Older learners, however, may have an advantage where vocabulary is concerned especially if the source and target languages share a number of cognates. Older learners, with a richer L1 vocabulary, are better equipped to benefit from positive transfer.

  Language aptitude and L1 transfer

  There are good reasons to believe that learners’ aptitude for learning an L2 influences the transfer of L1 features. For example, those learners with strong phonetic coding ability and phonological memory may be better equipped to process L2 sounds in the input and so may rely less on their L1 sounds. Conversely, individuals with little aptitude for mimicry of L2 sentences may be more likely to rely on their L1 (Odlin 1989). Learners with strong language analytical abilities will be able to identify the similarities and differences between their L1 and L2 grammars which may facilitate positive transfer and prevent negative transfer. However, there have been few studies examining the role of aptitude in L1 transfer. Jarvis and Pavlenko (2008), in their overview of research investigating crosslinguistic influences, did not consider language aptitude at all.

  Two studies, however, indicate that language aptitude does mediate transfer effects. Sparks, Patton, Ganschow, and Humbach (2009), drawing on the Linguistic Coding Difference Hypothesis—which claims that both first and second language learning depend on the same basic language-learning mechanisms—provided evidence to show that there is a long-term, cross-linguistic transfer of skills from L1 to L2 and that this transfer is mediated by language aptitude. Learners with lower levels of L1 skills have lower levels of L2 aptitude and correspondingly greater difficulty in learning an L2—at least in a classroom setting. In this case, then, language aptitude is seen as related to positive transfer of L1 skills.

  Trude and Tokowicz (2011) investigated the role played by working memory—which can be viewed as a component of language aptitude (see Chapter 3)—and language transfer in native English speakers’ learning of letter-sound correspondences in L2 Portuguese. They found that those individuals with higher working memory performed more accurately than those with lower working memory and suggested that this was because they were better able to inhibit their first language. In other words, where pronunciation is concerned, higher working memory can help learning because it enables learners to avoid negative transfer.

  Language transfer as a multifactorial phenomenon

  Having examined some of the key factors that influence L1 transfer, it has become clear that transfer needs to be considered as a complex, multifactorial phenomenon. Table 6.2 summarizes the various influences we have considered. By and large, researchers have investigated the influence of each factor separately, so little is known about how the factors interact to cause or inhibit transfer.

  Factors Commentary

  Linguistic factors

  linguistic differences The key prediction of the Contrastive Analysis Hypothesis, namely that L1/TL differences cause learning difficulty, is not fully supported. In some cases, similarities cause greater difficulty. Grammatical differences cause learning difficulty especially in cases where the target feature is not salient in the L2 input.

  language distance In cases where there the L1 and TL are more proximate, both greater positive and negative transfer can occur than in cases were the languages are distant.

  markedness There is a greater likelihood that learners will transfer unmarked than marked L1 features, especially when the equivalent TL feature is marked. However, learners also sometimes transfer marked features.

  Psycholinguistic factors

  prototypicality Learners have intuitions about the transferability of L1 features based on their assessment of whether a particular L1 feature is considered ‘core’ and ‘transparent’.

  psychotypology Learners’ perceptions of language distance (rather than actual language distance) influence transfer. Intuitions about prototypicality determine whether learners are prepared to transfer; their psychotypology determines what is actually transferred.

  Contextual factors

  macro-contextual factors In general, a formal classroom setting inhibits transfer. In natural settings, negative transfer is more likely to occur when learners do not make a clear distinction between their L1 and the L2 (i.e. in unfocused contexts). But the relationship between the formality of the context and transfer is not clear cut.

  micro-contextual factors These affect the extent to which learners pay conscious attention to form. This can sometimes lead to negative transfer (when learners elect to use their L1 resources) but this can also inhibit negative transfer (when learners are focused on using the L2 accurately).

  Developmental factors Transfer works alongside universal and developmental factors.

  overgeneralization Initially, learners are likely to draw on their L1 but later they rely more on what they have learned about the L2; transfer and overgeneralization can as seen as two manifestations of the same process—prior learning facilitates new learning.

  order of acquisition The ‘natural’ order is disturbed if a grammatical morpheme that is meaning bearing in the L2 has no equivalent in the L1.

  sequence of acquisition The effects of the L1 become evident when learners have reached a stage of ‘natural’ development that allows them to access an L1 form that is similar or equivalent to the TL stage. L1 transfer can both speed up and retard the natural sequence of acquisition.

  Individual factors

  age In general, L1 transfer occurs to a greater extent in older than in younger learners. This reflects differences in the extent to which younger and older learners depend on their L1 or on L2 input as a source of data for learning.

  language aptitude Learners with higher language aptitude appear better equipped to transfer L1 skills when learning an L2. Learners with higher working memory may be better able to inhibit their L1 and thus avoid negative transfer.

  Table 6.2 Summary of the various factors influencing L1 transfer

  Conceptual transfer

  So far, we have examined transfer in terms of the structural influences of the L1 on the L2. However, there is growing interest in how the concepts associated with one language affect the linguistic choices made in another language. This interest can be traced back to the Sapir-Whorf Hypothesis which, in its strong version, claims that the way we think is determined by how our language is structured and, in its weaker version, that linguistic differences in languages reflect differences in how we think.

  Conceptual transfer is clearly two way: that is, the conceptual framework of the first language impacts on the use and learning of a second language, and vice versa—i.e. there can be transfer from the L2 to the L1. However, in accordance with the main goal of this chapter—to examine the role of the L1 in L2 learning—I will focus only on L1-to-L2 conceptual transfer.

  Following Jarvis and Pavlenko (2008), I will distinguish semantic transfer and conceptual transfer. Semantic transfer involves the mapping of a concept shared by two languages onto a translation equivalent. For example, a Finnish learner of L2 English needs to map the concept TONGUE onto the English word ‘tongue’: when a Finnish learner says ‘He bit himself in the language’, the error arises simply because the Finnish word ‘language’ can refer to the concepts of both TONGUE and LANGUAGE. Thus, overcoming semantic transfer only involves learning the correct target language form; there is no need to modify the underlying concept.

  Conceptual transfer occurs when the concept itself differs in the two languages. Jarvis and Pavlenko illustrated this with reference to how English and Russian conceptualize PAPER CUP. In English, this belongs to the general category of CUP as it has the same function as other types of cups. In Russian, however, it belongs to the category of GLASS as it has a similar shape to a glass. Thus, when an English learner of Russian refers to a paper cup as ‘chaska’ (= cup) rather than ‘stakanchik’ (= little glass), conceptual transfer has occurred. Conceptual transfer, then, arises as a result of ‘similarities and differences in conceptual categories corresponding to lex
ical and grammatical categories of the source and recipient languages’ (Jarvis and Pavlenko 2008: 112). Overcoming linguistic errors that arise as a result of conceptual transfer involves not just learning the correct L2 linguistic form, but also developing new concepts or modifying existing concepts. It is for this reason that Jarvis and Pavlenko claimed that conceptual transfer is more persistent than semantic transfer. However, they also acknowledged that it can be overcome as we all have the capacity for conceptual development.

  Jarvis and Pavlenko (2008) provided many examples of conceptual transfer. They emphasized that a full understanding of transfer involves not only a consideration of the cross-linguistic effects that arise because of structural differences between the source and target languages, but also from differences in ‘language-mediated concepts and language specific patterns of framing and conceptualization’ (p. 148). They argued that the way concepts are realized linguistically in one language sensitizes speakers to think in particular ways and that learning a new language involves learning to conceptualize the world in different ways.

  In some cases, conceptual modification through L2 learning may not be desirable. It should not be assumed, for example, that L2 learners need to abandon their concept of what it means to be polite when they learn a new language. Japanese learners of English, for example, express politeness by apologizing to their hosts after an enjoyable meal, rather than by thanking them. In so doing, they are drawing on their L1-mediated notion of what constitutes politeness in this situation by acknowledging the inconvenience they have caused their hosts. Simply thanking them would not be sufficiently politeNOTE 4. The failure to overcome conceptual transfer in such cases may not be because learners are unable to but because—in some fundamental way—they do not wish to. Overcoming structural transfer is non threatening, but overcoming conceptual transfer can challenge the learner’s sense of personal or cultural identity. However, this is perhaps less likely with child than adult learners—yet another reason why starting young is more likely to lead to native-like competence in an L2.

  Transfer in communication and learning

  It remains to address the important issue of whether transfer is just a communication phenomenon, or whether it is also a feature of the learning process itself. Do learners just make use of their L1 as a communication strategy to help them overcome temporary performance problems or—as Selinker (1972) originally proposed—is it one of the central processes involved in interlanguage development?

  Jarvis and Pavlenko (2008) argued that transfer leads to both performance-related and learning-related effects; that these two types of transfer are distinct; and that, therefore, it is necessary to distinguish between the two ‘in order to avoid erroneous interpretations concerning the mechanisms through which transfer operates and about its ultimate impact on our knowledge representations’ (p. 210). For example, they suggested that the setting influences performance rather than learning-related transfer, while the effects of developmental and universal processes on transfer are ‘fundamentally learning-related’ (p. 193). The roles of explicit knowledge and conscious processes—such as monitoring in transfer—are seen as both performance- and learning-related.

  The distinction between these two types of transfer is problematic, however. First, evidence for transfer requires examining some kind of performance by the learner—by eliciting learners’ perceptions of transferability—as in Kellerman (1979)—by examining learners’ ability to process input containing specific target language features—as in Jiang et al. (2011), or—most commonly—by analysing learners’ L2 production. Thus, in a sense, all transfer effects are performance related, so distinguishing the two types of transfer in the way Jarvis and Pavlenko propose is methodologically problematic.

  Second, there are strong theoretical grounds for claiming that transfer affects learning via performance. Many years ago, Corder (1983) rejected the idea that learners transfer directly from their L1 into their interlanguages. He saw transfer occurring initially in communication and only subsequently becoming part of the learner’s L2 knowledge system:

  … persistent communicatively successful borrowing (from the L1) works backwards, as it were, and the successfully borrowed forms are eventually incorporated into the interlanguage grammar, both the correct and the incorrect. (p. 94)

  Ringbom (1992) also claimed a relationship between communication and learning transfer:

  Transfer in communication is motivated by the learner’s desire to comprehend or produce messages, but it may also have an effect on the process of hypothesis construction and testing, which many scholars see as central to interlanguage development. In other words, transfer in communication may lead to transfer in learning. (p. 106)

  Such a position is compatible with current usage-based theories of L2 acquisition (for example, Complexity Theory—see Chapter 1). These propose that any exposure to or use of a linguistic form will contribute to change in the underlying representation of this form in the mind of the learner.

  For these reasons, it may not be helpful to maintain a clear-cut distinction between performance-based and learning-based transfer effects.

  Conclusion

  This chapter has provided a partial account of the cross-linguistic effects that arise from the interface of two languages in language use and in the minds of learners. I have elected to focus on the effects that a source language (the learner’s L1) has on the use and learning of a second language—i.e. forward transfer—and have paid scant attention to the effects that the learning and use of a second language can have on the learners’ L1 (i.e. reverse transfer). This is not because the latter effects are not well documented, but because the purpose of this book is to examine the acquisition of a second language. Thus, I have tried to show what SLA researchers have discovered about the role of the first language in second language learning and have focused on identifying the factors that govern transferability: the main findings relating to this are summarized in Table 6.2.

  The effects of L1 transfer on L2 learning are extensive, varied, and persistent. They are also illuminative of the cognitive processes involved in L2 use and acquisition: no theory of L2 use or acquisition can be complete without an account of L1 transfer. However—just as there is no single, universally acknowledged theory of L2 acquisition—neither is there a general theory of L1 transfer. Such a theory would need to account for:

  the relative strength of positive and negative transfer in learners with different source and target languages

  the differential effects of transfer in different aspects of language—for example, its prominence in L2 phonology

  why learners transfer some L1 features in a particular area of language and not others

  the extent to which transfer is a conscious or subconscious process (i.e. whether it involves explicit and/or implicit knowledge of the source and target languages)

  the general relationship between L1 and L2 proficiency

  how developing L2 proficiency influences what and how much is transferred from the L1

  how natural processes of L2 acquisition interact with L1 transfer

  how macro- and micro-contextual factors affect transfer

  the extent to which transfer differs among learners of the same L1 as a result of factors such as age and language aptitude

  how—and to what extent—learning an L2 involves learning L1-mediated concepts

  how performance-based transfer contributes to learning-based transfer.

  While it is clear that transfer is influenced by a range of linguistic, psycholinguistic, psychological, and social factors, there is no explicit theory to explain the interaction of these factors.

  I conclude with a brief account of a very different perspective on the relationship between languages in the bilingual/multilingual mind—one that disputes two basic premises of much of the research discussed in this chapter—namely, the separateness of the learner’s first and second languages and the assumption that the study of transfer involve
s a comparison of the source and target languages. Cook (2000) proposed that people who know two languages are different from monolinguals, and so need to be considered in their own right. He coined the term multicompetence to refer to the knowledge of more than one language in the same mind. He emphasized that multicompetence is not restricted to people who are highly proficient in two languages (balanced bilinguals), but is applicable to all users of an L2, irrespective of their level of achievement. The study of multicompetence calls for a holistic account of what it means to know more than one language. As Cook (2003) put it, ‘since the first language and the other language or languages are in the same mind, they must form a language super-system at some level rather than completely isolated systems’. Cook (2000) also said that he preferred to avoid using the word ‘transfer’. From this perspective, then, the distinction between the L1 and L2 becomes blurred and what is required is a focus on bilingual/multilingual language use.

  Notes

  1 Transfer effects are also clearly evident in the pragmatic and discourse levels of language. For example, learners commonly draw on their L1 when performing requests or apologies.

  2 Stockwell and Bowen (1965) also identified a hierarchy of difficulty for phonological features based on whether a given sound was phonemic, allophonic, or absent.

  3 In a self-paced reading task, participants are exposed to a computerized sentence one word at a time and are in control of when to move on to the next word. The time taken to move forward to the next word is automatically recorded. Jiang et al. (2011) argued that if an error in a word was attended to, it would result in longer being spent on that word—or possibly the following words—than if it were not attended to. In effect, then, a self-paced task allows for the investigation of transfer effects on receptive, rather than productive, knowledge.