Crosslinguistic Influence in Second Language Acquisition

Rosa Alonso Alonso, Rosa Alonso Alonso

• Publisher: Multilingual Matters
• Year: 2016
• ISBN:
  9781783094844
• Topic: Filología
• Subtopic: Psicolingüística

This volume provides an unprecedented insight into current approaches to crosslinguistic influence (CLI). The collection investigates a range of themes including linguistic relativity, the possible contributions of neurolinguistics, the problem of cognitive development and the role of the frequency of structures in acquisition from distinct, overlapping and complementary perspectives. Chapters focusing on vocabulary, morphosyntactic categories, semantic structures, and phonetic and phonological structures feature in the volume, as do over 20 languages, in order to offer new insights into both theoretical and empirical issues in CLI, including the consequences of great or little similarity in structures between languages. The relevance of CLI research for teaching is discussed in a number of chapters, as is the phenomenon of multilingualism. The collection will appeal to researchers, graduate and postgraduate students, teachers and professionals interested in the field of CLI in SLA.

1Was There Really Ever a Contrastive Analysis Hypothesis?

Terence Odlin

Introduction¹

The terms Contrastive Analysis (CA) and language transfer abound in discussions of crosslinguistic influence, but even a cursory reading of these discussions shows a wide range of definitions, characterisations and historical claims about the origins of these notions and terms. Discussions of the so-called Contrastive Analysis Hypothesis (CAH) likewise show considerable variation, as in the examples below.

The CA hypothesis held that where structures in the L1 differed from those in the L2, errors that reflected the structure of the L1 would be produced. (Dulay et al., 1982: 97)

With the eclipse of descriptive linguistics and its concept of language learning, Robert Lado is now best remembered as the author of Linguistics across Cultures. In particular, he is represented as the architect of the Contrastive Analysis Hypothesis – essentially, the claim that whenever a learner’s native and target languages differ, the learner will face difficulty and delay in acquiring the target language. (Thomas, 2006: 302)

Lado (1957) developed the Contrastive Analysis (CA) approach to L2 acquisition. Under the Contrastive Analysis Hypothesis, learning a new language involves identifying and learning differences between the L1 and the L2. Similarities between the L1 and L2 are predicted to facilitate acquisition. L2s with more differences are predicted to take longer to learn. (Foley & Flynn, 2013: 98)

Still another characterisation of the CAH equated it with the following conditions:

Where two languages were similar, positive transfer would occur; where they were different, negative transfer, or interference, would result. (Larsen-Freeman & Long, 1991: 53)

Along with Robert Lado (1915–1995), some other names are also invoked in discussions of the history of transfer research, such as Uriel Weinreich (1926–1967), whose 1953 book Languages in Contact is referred to in the following passage:

Adopting a general view of transfer as the use of knowledge or skills from one context in a different linguistic context, Weinreich (1953) introduced the concept of transfer in L2 acquisition: use of the L1 that leads to ‘correct’ usage in the L2. Interference, in contrast, involves use of the L1 that leads to ‘incorrect’ language use. (Foley & Flynn, 2013: 98)

Although Foley and Flynn do not give any source as an example of the ‘general view’ that they ascribe to mid-20th-century notions of psychology and language, Nick Ellis (2008) does offer a very specific source: the behaviourist construct of proactive inhibition (PI):

Much of this [behaviorist] work was succinctly summarized in Osgood’s ‘transfer surface’ that draws together the effects of time of learning, similarity of material, and retention interval on negative (and positive) transfer (Osgood, 1949)…. PI underpins a variety of fundamental phenomena of language learning and language transfer, as Robert Lado proposed in his CAH… The CAH held that one could predict learner difficulty by reference to utterance-by-utterance comparison of a learner’s L1 and L2. (2008: 384)

The six quotations given above vary considerably in what they address, including the issues of errors, of prediction, of positive and negative transfer, and of proactive inhibition. Even so, it would be a mistake to conclude that the characterisations of transfer and CA are completely different. There is arguably a gestalt interpretation of the history of transfer research that many readers might construct from these or other sources. For those completely new to second language acquisition (SLA), the gestalt history might consist in part of the following statements:

•It was Lado who formulated the Contrastive Analysis Hypothesis.

•The notion of language transfer has its origins in behaviourist psychology.

•Weinreich introduced the term transfer into L2 research.

Moreover, it seems likely that not only neophytes but indeed a wide segment of the SLA community, professors and students alike, would agree with at least one of the three statements. However, none of the three is true.

Intellectual interest in crosslinguistic influence goes back many decades before Lado, and probing that earlier history is necessary in order to understand transfer research in the mid-20th century and to scrutinise the historiography of transfer in research and in textbooks on SLA. As in other domains of intellectual history, complications arise in trying to untangle the strands of the story. This chapter will untangle some of the strands involving the so-called Contrastive Analysis Hypothesis. Doing so will require not only much attention to transfer research long before Lado and Weinreich but also much attention to those two figures. The results of the investigation suggest that the phrase Contrastive Analysis Hypothesis is nebulous and contributes little to understanding the intellectual history of transfer. Beyond the critique of a questionable term, the chapter can provide, albeit in a very limited space, encouragement to members of the SLA community who care to reflect on when SLA research actually began and also to reflect on where the field might be going.

The chapter proceeds from definitions of two key terms, language transfer and contrastive analysis, to looking at the origins of the term transfer independent of any work in psychology. It then considers the term contrastive analysis and proceeds from there to a critique of the so-called Contrastive Analysis Hypothesis. The word habit is also relevant to the intellectual history of transfer, and there will be considerable attention to uses of the term in linguistics and psychology and to recent notions such as automaticity that involve similar concerns. The concluding section of the chapter summarises the main points and discusses implications of the findings.

Preliminary Definitions

Throughout the chapter, the following definitions will serve as a reference point for the diverse ideas considered:

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)

Contrastive analysis: Systematic comparison of two or more languages. (Odlin, 1989: 165)

These definitions seem accurate enough, but they certainly do not tell all that may interest SLA specialists. The definition of CA does depart from uses of the term as a historical designator and certainly from uses of the phrase Contrastive Analysis Hypothesis. Moreover, the definition of transfer characterises it as influence (and it is often called crosslinguistic influence), but what influence might actually mean remains a major research question. Even so, the notion of influence has a long history. In 1884, the German linguist Hugo Schuchardt (1842–1927) used Einfluss (influence) in his classic study of German and Italian in contact regions inhabited by speakers of various Slavic languages (especially Slovenian, Czech and Polish). Schuchardt had already considered language contact in pidgin and creole settings, and many linguists today remember him as a founder of the field of creolistics. In his work on contact in Central Europe, Schuchardt focused on what he considered to be examples of L1 influence, but he also occasionally commented on the psychology of SLA. Work on language contact in the decades after Schuchardt also reported crosslinguistic influence, but as the discussion in Weinreich’s Languages in Contact will show, the psychology of bilingualism remained an under-explored area. As interest in transfer grew among contemporaries of Weinreich, such as Lado, so did the controversies over the general psychology of language and the specific psychology of L2 acquisition. Before the discussion of controversies, however, the origins of the term transfer require attention.

Transfer and its Independent Origins in Linguistics

In English, the use of transfer to denote crosslinguistic influence goes back at least to the 19th century, well before the behaviourist research cited by Ellis in the quotation in the introduction. The term appears in the writing of two American linguists, William Dwight Whitney (1827–1894) and Aaron Marshall Elliott (1844–1910). Both had studied in German universities, the former in Berlin and Tübingen and the latter in Munich, and both were probably influenced by German terms to be considered in the next paragraph. In a discussion of language contact in historical linguistics, Whitney declared, ‘By universal consent, what is most easily transferred from one language to another is a noun’ ([1881] 1971: 184). This assertion might suggest that Whitney considered crosslinguistic influence to be widespread, but in fact he shared the scepticism of many linguists of his time about the extent of such influence. Elliott used a variant of the term (transference) in an 1885 review of Schuchardt’s book in the American Journal of Philology (and the review is reprinted in the 1971 edition of Schuchardt’s work). Elliott employed influence as well as transference in the review, and he also used transfer in an 1886 article on language contact in Canada.

Since Whitney and Elliott had studied linguistics (aka philology) in Germany, they probably adopted transfer and transference as translations for two words used by German linguists: hinübertragen and übertragen, both of which can also translate as ‘carry over’. Hinübertragen goes back at least to 1836, appearing in a famous study of mind and language by Wilhelm von Humboldt (1767–1835), and the metaphoric notion of carrying over is the same as in the Latin word transferre (Odlin, 2008; Odlin & Yu, 2016). With regard to übertragen, it goes back at least to 1875, appearing in an analysis of loanwords by a Finnish linguist, August Ahlqvist (1826–1889); although some uses of übertragen in his book involve simple cases of semantic extensions not related to language contact, other uses clearly refer to crosslinguistic influence (e.g. Ahlqvist, 1875: 51) in the language contact setting of Finland.² Schuchardt used übertragen at least a year before his book on language contact in Central Europe, as the term denotes crosslinguistic influence in one of his creolist studies (Schuchardt, 1883), and it also appears a number of times in his 1884 book as well as in an 1885 review of the book, the historical linguist Hermann Paul being the reviewer (also reprinted in the 1971 edition of the Schuchardt work). That the use of transfer comes from German linguistics rather than from any work in psychology seems all the more likely because neither Whitney nor Elliott discusses any psychological research, behaviourist or otherwise. Indeed, behaviourism was in its infancy in the late 19th century (Graham, 2010; Levelt, 2013; Singley & Anderson, 1989). Weinreich, it should be added, cites both the Whitney article and the Schuchardt book, and he calls the latter ‘unexcelled’ (1953a: 111).

In the several decades between Whitney and Weinreich, the term transfer did not disappear. Both Edward Sapir (1884–1939) and Otto Jespersen (1860–1943) used the term in their introductions to linguistics:

We may suppose that individual variations arising at linguistic borderlands – whether by unconscious suggestive influence of foreign speech habits or by the transfer of foreign sounds in the speech of bilingual individuals – have gradually been incorporated into the phonetic drift of a language. (Sapir, 1921: 200)

…it is, of course, a natural supposition that the aboriginal inhabitants of Europe and Asia were just as liable to transfer their speech habits to new languages as their descendants are nowadays....