New Perspectives on Transfer in Second Language Learning

Liming Yu, Terence Odlin

• Publisher: Multilingual Matters
• Year: 2015
• Print ISBN: 9781783094325
• eBook ISBN: 9781783094356
• Topic: Languages & Linguistics
• Subtopic: Psycolinguistics
• Index: Languages & Linguistics

When people attempt to learn a new language, the language(s) they already know can help but also hinder their understanding or production of new forms. This phenomenon, known as language transfer, is the focus of this book. The collection offers new theoretical perspectives, some in the empirical studies and some in other chapters, and consists of four sections considering lexical, syntactic, phonological and cognitive perspectives. The volume provides a wealth of studies on the influence of Chinese on the acquisition of English but also includes studies involving Finnish, French, Hindi, Korean, Persian, Spanish, Swedish and Tamil. It will be of great interest to researchers and students working in the areas of crosslinguistic influence in second language acquisition, language pedagogy and psycholinguistics.

1 Introduction

Terence Odlin and Liming Yu

What is Language Transfer?

The word transfer has had many specialized uses and not just in linguistics: for instance, transfer and transference have long appeared in psychology, with different movements (e.g. psychoanalysis and behaviorism) using the words as terms with quite different meanings. Likewise, in linguistics the technical meanings of transfer are far from uniform. Some of the varying uses will be mentioned a little further on, but for now, the working definition that will inform this chapter is as follows: ‘Transfer is the influence resulting from the similarities and differences between the target language and any other language that has been previously (and perhaps imperfectly) acquired’ (Odlin, 1989: 27).

The definition deliberately includes ‘any other language’ because there are many cases of people learning not only a second language (L2) but also a third (L3). For example, in China many native speakers of Uighur (a Turkic language) have Mandarin Chinese (a Sino-Tibetan language) as their L2 when they begin to study English as their L3, and so similarities and differences between Chinese and English as well as similarities and differences between Uighur and English might affect such learners’ acquisition of the L3. Although relatively little study has yet been done on this particular trilingual situation, the steadily growing research field dealing with multilingual settings has documented many cases of both first language (L1) and L2 influence on an L3 as well as the influence of an L3 on an L4, etc. (De Angelis, 2007; De Angelis & Dewaele, 2011; Gabryś-Barker, 2012; Hammarberg, 2009). In our volume, Chapters 3, 11 and 13 consider trilingual or multilingual cases, but the other chapters focus on L1 → L2 transfer.

Even when only two languages are involved, cross-linguistic influence (which is a synonym for transfer) can be manifested in different ways, as will be seen. Moreover, while an L1 can influence an L2, the reverse is also common (thus constituting L2 → L1 transfer). For instance, Porte (2003: 112) investigated the English of several native speakers of English who were teaching English as a foreign language (EFL) in Spain and found many examples of L2 → L1 influence: e.g. I was really shocked when I first saw how molested some teachers got at my criticising the system, where molested has the less pejorative meaning of Spanish molestar (‘annoy’) as opposed to the English molest, which is often used to denote criminality (e.g. child molester). The teachers in Porte’s study seem to have been influenced both by some direct knowledge of Spanish and by their relatively long residence in Spain, especially since the molested example could just as easily come from a native speaker of Spanish using English (Nash, 1979). It is also clear that L2 → L1 influences in grammar are rather common. For example, Pavlenko (2003) found that Russian speakers residing in the USA sometimes used the perfective/imperfective system of verbs in Russian in ways quite different from monolingual speakers in Russia and in ways quite like those found in the L2 Russian of L1 English speakers. Sometimes, the difference between the L2 → L1 influence and what is called code-switching is subtle or even non-existent, and there likewise exists a large body of research on code-switching (e.g. Isurin et al., 2009), though none of the chapters in our volume focuses on that phenomenon.

In many immigrant situations, both L2 → L1 and L1 → L2 transfer are likely in the same community, but an opportunity to study a different kind of bidirectional transfer comes from international schools, where, for example, Italian children who might or might not become permanent residents of England study English, and where English-speaking children study Italian in international schools in Italy. Rocca (2007) investigated just this kind of parallel transfer (L1 Italian influence on L2 English, L1 English influence on L2 Italian) with regard to tense and aspect structures in the target language.

The most typical cases of transfer – and usually the ones that preoccupy language teachers – involve divergences between the source language (whether the L1 or perhaps the L2 in cases of L3 acquisition) and the target language (i.e. the language that learners are seeking to acquire). Such divergences can result in negative transfer, which is often evident in vocabulary problems as when a native speaker of Spanish uses molest in English as a synonym for annoy. Along with such vocabulary problems, negative transfer is often evident in syntactic structures, as in the following sentence from a native speaker of Vietnamese: She has managed to rise the kite fly over the tallest building (=She has managed to fly a kite over the tallest building), where the use of rise…fly indicates the influence of Vietnamese grammar (Helms-Park, 2003). The pronunciation and spelling patterns of L2 learners likewise show many instances of negative transfer related to pronunciation problems, as seen in a spelling error of a Finnish student who writes crass instead of grass. The Finn’s misspelling of grass with either the letter <c> or the letter <k> reflects a phonological fact about the native language: Finnish does not have a phonemic contrast between /k/ and /g/, and learners of English in Finland thus have to learn a new consonant contrast.¹

While divergences involving pronunciation, vocabulary and grammar naturally compel teachers’ attention, a topic just as important for anyone wishing to understand transfer is the complementary phenomenon of positive transfer, which does not involve errors. For example, it now seems clear that some errors such as omitting articles are less likely to come from speakers of some L1s in comparison with others (e.g. Luk & Shirai, 2009; Master, 1987; Oller & Redding, 1971). Such research indicates that although speakers of languages with articles may still have problems with articles in a new language, they have fewer than do speakers of languages that do not have any articles. In other words, omissions and other article problems are less characteristic of some groups, as in a sentence taken from a corpus of EFL writing² that focused on events seen in a film: Old woman say to baker: Girl take it, not man (with four articles being omitted along with the problems of punctuation and the erroneous verb forms). In principle, omissions might be found among any L1 group, but in reality they are much less common among native speakers of Swedish, a language that has articles, than among native speakers of Finnish, a language that does not. The example just given comes from a speaker of Finnish, whereas the facilitating influence (i.e. positive transfer) of L1 Swedish is reflected in many passages in the same corpus, cases that would not, as far as article use is concerned, cause any concern for language teachers: e.g. But the old woman came and she said to the man that it was the woman ho tok [took] the bread. The advantage of L1 Swedish speakers over Finnish speakers in regard to the positive transfer of articles is now amply documented (e.g. Jarvis, 2002; Odlin, 2012a; Ringbom, 1987).

A number of researchers including Corder (1983: 92) have criticized the term transfer for its metaphoric suggestion when in fact ‘nothing is being transferred from anywhere to anywhere’. The notion of movement inherent in transfer has sometimes been seen as one manifestation of a larger group of metaphors which Reddy (1979) dubbed ‘the conduit metaphor’. Corder’s criticism might be best viewed as a warning against an unreflective use of metaphors, which can indeed cause misconceptions, as discussed by various specialists (e.g. Lakoff & Johnson, 1980, 1999; Vervaeke & Kennedy, 2004). Even so, transfer has long served as a useful cover term for a variety of phenomena that require detailed non-metaphoric analyses. The conduit metaphor is also evident in two other terms: translation and (somewhat ironically) metaphor. The latter seems less closely related to transfer, but Dechert (2006) observes that the etymologies of both transfer and metaphor suggest that something is carried from one place to another, transfer coming from the Latin trans (across) and ferre (carry) and metaphor from the Greek meta (across) and phor (carry). The verb translate has essentially the same etymology as transfer, since the word latus is a participial form of the irregular Latin verb ferre. Of course, there are undeniable differences between the phenomena of transfer, translation and metaphor; nevertheless, the semantic kinship of the terms seems significant. Specialists such as Lakoff and Johnson have emphasized that metaphors permeate the everyday use of language, as people often seek to comprehend the world and their own lives, and it is thus not surprising that the image of carrying over has seemed useful to both specialists and non-specialists in discussing language, whether the topic is translation or transfer (Odlin, 2008).

The phenomena of translation and transfer do in fact show more than simply an etymological overlap. The molest example given above illustrates a pitfall for both translators and language learners, who are often warned about ‘false friends’. Sometimes, the overlap is more subtle, as in the prepositional error seen in They sit to the grass (=They sit on the grass), which was written by a native speaker of Finnish in describing a scene in a film and which reflects the semantic influence of a Finnish inflection –lle (meaning ‘to’ or ‘onto’) as in the word nurmikolle, translatable literally as ‘grass-to’ (Jarvis & Odlin, 2000). Yet, even though many cases of lexical and grammatical transfer involve some kind of translation or, as Weinreich (1953) preferred to phrase it, some ‘interlingual identification’, it would be a mistake to conclude that the two phenomena are really one and the same thing. Not all transfer behaviors involve translation. For example, hypercorrections such as a Finnish writer’s use of gomes (=comes) differ in a remarkable way from the error crass discussed earlier. As already noted, Finnish has a voiceless velar /k/ but no voiced counterpart /g/. The crass example involves a simple carrying over, so to speak, of the voiceless /k/ to an inappropriate target language context, but the use of the letter <g> in gomes represents a rejection of both the Finnish letter <k> and the Finnish phoneme /k/, even though comes or even the misspelling komes would constitute a valid interlingual identification between an English and a Finnish phoneme. Unlike translation errors such as molest and to the grass, where learners overestimate the similarity between the L1 and L2, hypercorrections arise when learners underestimate the actual similarity.

Although errors involving mistranslation or hypercorrection naturally loom large in the concerns of language teachers, some manifestations of negative transfer lie on the borders of language itself. One such area involves what is sometimes called contrastive rhetoric and sometimes contrastive pragmatics. For example, Blum-Kulka (1982) found that requests in L2 Hebrew by L1 speakers of English were often too indirect by the norms of the target language. Such research entails comparing cultural norms, which, of course, vary a great deal from one society to the next. Not surprisingly, the norms sometimes prove challenging to study, due in part to the fact that they can vary over time, as seen in a study of Chinese rhetoric through the centuries (Bloch & Chi, 1995). Along with the cross-linguistic variation in rhetoric and pragmatics (aka discourse), the complexities of cultural norms are also manifest in paralinguistic behaviors, with the variation sometimes conducive to transfer as evident in recent work on L2 gestures. A wide repertory of gestures is typically available to native speakers as when, for instance, French speakers often provide information about the path of a moving object, a pattern in contrast to that of Dutch speakers, who tend to provide information about the object as such (Gullberg, 2011). Transfer of L1 gestural patterns has been evident in a wide range of studies ...

Table of contents

• Cover-Page
• Half-Title
• Series
• Title
• Copyright
• Contents
• Contributors
• Acknowledgments
• 1 Introduction
• 2 The Scope of Transfer Research
• Part 1: Lexical Perspectives
• Part 2: Syntactic Perspectives
• Part 3: Phonological Perspectives
• Part 4: Cognitive Perspectives
• Index of Persons Cited