Transfer and the relationship between the languages of multi-competence

Vivian Cook

Manuscript version of paper in R.A. Alonso (ed.) (2016), Crosslinguistic Influence in Second Language Acquisition, Bristol: Multilingual Matters, 24-37

This paper treats transfer as one of the central relationships between languages in the second language (L2) user’s mind, rather than as social interaction or language contact. It draws on developments in the multi-competence perspective since its first appearance in Cook (1991), in particular on a paper entitled ‘Is transfer the right word?’, given at the IPRA conference in Budapest (Cook, 2000), which concluded ‘Transfer is not enough as a term because it restricts L2 users to the position of cumulative monolinguals rather than seeing the richness of the L2 mind.’ The present paper elaborates and develops this theme, particularly in connection with the current interest in how L2 users think differently from monolinguals – bilingual cognition.

From transfer and interference to cross-linguistic influence

What distinguishes second language acquisition (SLA) research from other language disciplines is that it is concerned with the acquisition and use of language by people who already know one language. This may seem so obvious it doesn’t need stating: it is a truism to say that second language research is about second languages.

Yet this axiom is effectively ignored every time we treat second language acquisition without reference to the first language (L1) already present in the person’s mind. If SLA research pays no heed to the first language, deprived of its one unique element, L2 acquisition becomes a shadow projected from L1 acquisition. Ignoring the first language effectively dismisses it as irrelevant to SLA research, equivalent to studying how people with two legs manage to hop on one.

Some SLA research indeed treats the first language as an integral part of the picture. For example the generative SLA researchers of the 1990s debated the relative importance of the first language grammar and Universal Grammar for the second language (L2) grammar through Full Access (Epstein et al, 1996), Full Transfer/Full Access (Schwartz & Sprouse, 1996) or Failed Functional Features (Hawkins & Chen, 1997) hypotheses, sometimes called the war of the hypotheses.

Nevertheless the vast bulk of SLA research has undoubtedly minimised the role of the first language in second language acquisition, particularly in the 1970s when an influential faction saw transfer merely as a communication strategy, for example Krashen (1982). One reason might be the belief that all L2 users are coordinate bilinguals, keeping the languages in separate compartments of the mind (Weinreich, 1953). The coordinate assumption has indeed been a commandment in language teaching since the 1880s, enforced by many educational systems, with the classroom effectively an L2 ghetto from which the L1 is excluded. The claim of coordination needs to be the outcome of research, not a presupposition: it now seems highly unlikely in view of the massive evidence for links between the languages in the mind, described in say Jarvis and Pavlenko (2009). A second reason might be the view that L2 users are all the same: L2 learning is L2 learning regardless of the pairs of languages involved, a universal process, so the L1 can be disregarded. Again this needs to be tested empirically rather than presupposed; recent thinking suggests that, at best, universality may apply to certain highly abstract areas of language such as structure dependency (Hauser et al, 2002).

In the 1950s the relationship between the two languages was conceptualised as interference created by contact between two languages in one person (Weinreich, 1953): the forms of the first language influence the new forms being acquired in the second. Weinreich (1953, 8) for instance talks of the Russian Subject Object Verb order yielding I him see in Russian learners of English. Transfer was a general term for this relationship: elements from the first language were transferred to the second, invisibly when they were the same as those in the second language (positive transfer), visibly when they were different (negative transfer). Much SLA research came down to investigating how the L2 was affected by the L1, leading to generations of theses on different pairs of languages as L1 and L2, originally concerned with syntax and phonology but now extending to areas such as spelling and gestures: Japanese learners for example transfer their Consonant Vowel syllable structure to English, generating the epenthetic vowels of adavantage and difficulity.

As the term transfer had so many associations with the scorned behaviorist theories of language acquisition, Kellerman and Sharwood-Smith put forward the more neutral alternative cross-linguistic influence (CLI) as ‘the interplay between earlier and later acquired languages’, which has indeed become the most established term (Kellerman & Sharwood-Smith, 1986).

The transfer relationship between the languages in the mind can, however, go in both directions. The Voice Onset Time (VOT) for the plosive consonant /t/ is longer for French speakers who know English than for French speakers who don’t – in their first language (Flege, 2002); the meaning of the Japanese word bosu is different for Japanese people who speak English than those who don’t – in Japanese (Tokumaru, 2002); English speaking children learning heritage Cherokee over-regularise English past tense forms such as taked more than their monolingual contemporaries (Hirata-Edds, 2011). In other words the L2 has discernible effects on the L1.

This ‘reverse’ transfer was indeed inherent in Weinreich’s first formulation of interference – ‘those instances of deviation from the norms of either language which occur in the speech of bilinguals as a result of their familiarity with more than one language’ (Weinreich, 1953: 1). Yet until recently few people have taken this point on board (Cook, 2003); the research into L2 on L1 effects amounts to a tiny fraction of that into L1 on L2 effects. (One might speculate an unconscious motivation was that this could undermine the status of bilinguals as proper native speakers of their L1).

Now much research has gone beyond people who know second languages to the other languages that multilinguals may speak. A third language (L3) may be influenced by the L2 or by the L1; Kulundary and Gabriele (2012) found most influence on the acquisition of L3 English relative clauses by speakers of L1 Tuvan came from L2 Russian. Similarly an L3 may influence an L2 or an L1: Wrembel (2011) has for instance shown that L3 VOTs are affected by both L1 and L2. And so on, up to the indefinite number of languages (Ln) a particular multilingual may know. The complexity of relationships dealt with in SLA research is ever increasing, leading to research questions about which languages have most influence on others, such as the Cumulative Enhancement Model which sees all previous language learning as contributing to L3/Ln learning (Flynn et al, 2004).

A further crucial distinction made by Weinreich (1953) was between cross-linguistic influence during actual speech and as part of language knowledge, parole versus langue: ‘In speech interference is like sand carried by a stream; in language it is the sedimented sand deposited on the bottom of a lake’ (Weinreich, 1953: 11). This was restated in De Groot’s definition as ‘The influence of the non-selected language on the selected language in language use by bilinguals (and multilinguals) or the influence of an earlier acquired language (e.g. the L1) on the acquisition of a new language (e.g. L2)’ (De Groot, 2010: 449), the first clause summing up use, the second acquisition. This distinction has not, however, been made in most research on cross-linguistic influence; it is absent from Odlin’s definition ‘Transfer is the influence resulting from similarities and differences between the target language and any other language that has been previously (and perhaps imperfectly) acquired.’ (Odlin, 1989: 27) or indeed from Hu’s ‘Transfer refers to how previous learning influences current and future learning’ (Hu, 2013: 732).

SLA research thus usually refers to transfer as an influence on acquisition, not on the speech process, even if most research data are directly products of the speech or writing process, only indirectly of learning. Perhaps the researchers no longer accept a distinction between speech and language knowledge or between langue and parole, seeing it as a version of Chomsky’s competence and performance (Chomsky, 1965). Nevertheless the logical default is surely that the relationships between languages that shape their systems in the mind are different from the relationships manifested in speech production: arguments need to be provided for the abolition of this learning/use distinction.

The key element in SLA research is the relationships between the languages in the mind. Transfer is one type of relationship, conceived in one particular way. The rest of this chapter will explore some issues with whether transfer is the central relationship between the languages in the L2 user’s mind and with how other relationships can be taken into account.

 Multi-competence and transfer

The concept of multi-competence initially complemented the independent grammars assumption that learners create languages of their own, crystallised in the term ‘interlanguage’. Rather than inefficiently imitating the target language, L2 learners create their own language out of the resources they have available to them, such as their learning strategies, the language provided by their teacher and indeed their first language (Selinker, 1972): they do not have a defective copy of the L2 in their minds so much as an interlanguage of their own making. An L2 learner thus possesses an L1 and an L2 interlanguage. But there was no overall term for the totality of [L1+interlanguage] in their minds, only for its component elements, the L1 and the interlanguage. Multi-competence was proposed as a way of referring to the sum of these elements, initially defined as ‘the compound state of a mind with two grammars’ (Cook, 1991).

This definition has developed over the years into the current ‘the overall system of a mind or a community that uses more than one language’ (Cook, 2015), extending its scope to the community as well as the individual, and to any other languages that are known beyond the first, and talking more generally about the language system rather than the grammar. Multi-competence is not so much a theory or a model as a perspective, viewing second language acquisition from the L2 user’s standpoint rather than from the native speaker’s.

In a sense the crucial point is what is left out of the definition of multi-competence – the native speaker. An L2 user has an independent language system of their own, not a combination of two monolingual states. In Grosjean’s terms (Grosjean, 2008), it is a bilingual ‘wholistic’ interpretation of bilingualism as opposed to a monolingual ‘fractional’ interpretation. The proficiency, or lack of it, of the L2 user is defined in their own terms, not in terms of how successfully they mimic monolingual native speakers. Interesting comparisons can be made between L2 users and monolingual native speakers – both are after all examples of human language acquisition – just as it is interesting to compare apples and pears, but this is incidental rather than the core issue. What cannot be claimed is that L2 users are deficient compared to native speakers, since this assumes the native speaker to be the norm rather than the multi-competent speaker, any more than one can say that an apple is a rather inferior pear. Measuring the L2 user against the native speaker at best reveals the similarities and differences but not the unique qualities of the L2 user, any more than describing a pear in terms of an apple captures the unique essence of pears.

An important element in the multi-competence perspective described in Cook (2015) is thus the premise that ‘multi-competence concerns the total system for all languages (L1, L2, Ln) in a single mind or community and their inter-relationship’. This asserts that all the languages in the individual mind or community form a whole system at some level. Since bilinguals do not have two heads, this is trivially true; the question is at what level, if any, does the whole system divide into two or more languages?

Multi-competence therefore emphasises the dynamic inter-relationship between the languages in the mind. To take some examples: if L2 users are shown pictures of objects named in one language, their eyes are attracted by objects that have similar names in the other (Spivey & Marian, 1999; 2003); producing cognates activates both phonological systems (Hermans et al, 2011; Friesen & Jared, 2011). L2 users never switch off either language entirely; the word coin has different meanings in English and French but, whichever language you are speaking, you cannot turn off the meaning in the other (Beavillain & Grainger, 1987). The question of which language to use in speech is not a question of either/or but of more/less, whether conceived positively as activation of one language or negatively as deactivation of one language (Green, 1998).

Seeing the relationships between the languages in the mind as evolving in both the short and long terms brings multi-competence within the same compass as dynamic systems theory (DST). Here the main characteristic of the system is that it is always in a state of flux, changing from minute to minute and day to day: any description is a single frame taken from a continuous movie, as described by De Bot (2015). This does not mean that it is impossible to describe, simply that its workings are always fluid and changeable; a snapshot is an arbitrary moment in time stolen from a continuous process.

Aronin & Singleton (2012: 59) have put forward the idea of Dominant Language Constellation (DLC): ‘A complex of languages shared on a day-to-day basis by an entire community …’. The relationships between the languages of multi-competence form an inner constellation of languages in active use, surrounded by a repertoire of other languages the person knows but does not currently use, surrounded in turn by languages the person is merely aware of to some degree. DLC presents a complex set of relationships between three or so main languages, another four or five minor languages and an indefinite number of peripheral languages. DLC is one useful way of looking at multilingualism from a multi-competence perspective.

 Other relationships in multi-competence

The overall question which naturally arises for multi-competence is ‘How do the relationships between the languages of multi-competence affect them?’ Under the rubric of transfer/cross-linguistic influence, these relationships were defined in terms of direction – L1 affects L3, L2 affects L1 and L3, and so on – and in terms of positive and negative effects.

 The metaphor involved in the term transfer creates a particular problem (Cook, 2000). The central dictionary meaning of transfer is 'To convey or take from one place, person, etc. to another' (OED, 2009). In other words transfer means something moving from Point A to Point B, as reflected in the titles of articles such as ‘Transfer to somewhere’ (Andersen, 1983) and ‘Transfer to nowhere’ (Kellerman, 1995). The English verb transfer typically has four arguments, namely an Actor, an Object, a source location and a destination; John transferred his account from the NatWest to Barclays. The verb reifies three distinct objects – the source language, the destination language and the linguistic entity that is transferred – and one process – the act of transferring. (The analysis of the dictionary meanings of transfer is developed further in Dechert (2006)).

So, when someone transfers language, they move some of the properties of one language to another. While the term cross-linguistic influence may substitute influence for transfer , it still accepts the same pair of L1 and L2 objects. But the transferred object does not actually move and leave a gap where it came from. The metaphor does not work literally – when you transfer your funds from one bank to another, you don’t leave funds behind. Rather some kind of copy is made and stored in the other language. In Odlin (1989) this is indeed used to distinguish the meaning of transfer in behaviourist psychology, where the original becomes extinct, from that in SLA research where the first language does not vanish – when it does, it is called attrition, not transfer, as we see below.

A further logical difficulty occurs when what is transferred is the lack of something in one language that exists in another. Chinese does not have articles and inflections; the effects of this on the learning of English by Chinese speakers is readily evident in sentences such as I used to be cook in peking hotel and computer take the new life into our world. To Weinreich, what is transferred is rarely just an item, a discrete object: ‘a language is a system of oppositions’ (Weinreich, 1953, 8): it is a structure or system. A lack of articles for instance can be transferred as part of a grammatical system, not as an individual object.

Let us now turn to other relationships between the languages of multi-competence than transfer, such as codeswitching and attrition. In codeswitching the L2 user employs two languages in the same discourse, as in the advertising slogan for Stella Artois, Into a chalice not a glass. C’est cidre not cider. Codeswitching has been studied from many angles: what they have in common is the attempt to relate the two languages in the user’s mind, whether through vocabulary chalice/glass, syntax C’est cidre not cider, or many other areas of language.

Codeswitching is important because it is unlike anything that a monolingual can do. Two languages combine together to produce one sentence; the relationship between the languages extends to every aspect. It is an extreme case of Grosjean’s bilingual mode (Grosjean, 2008) in which the person has access to both languages simultaneously rather than the monolingual mode in which they have access to only one language, whether the first or the second. Codeswitching in itself shows the necessity for studying both languages in the L2 user’s mind rather than the second language in isolation.

Attrition is defined as ‘the non-pathological decrease in a language that had previously been acquired by an individual’ (Kopke & Schmid, 2004: 5): parts of a language are ‘lost’ to the user whether L1 or L2. L2 attrition is perhaps all too familiar to school language learners: school Latin becomes a dim memory apart from a few tags from Horace. Attrition has become a major field of research of its own, concentrating more on describing the complexity of factors that affect it rather than the over-riding importance of any single common factor (Seton & Schmid, 2015), the underlying cause being ‘a lack of inhibitory control’ between the two languages, i.e. a malfunction in the relationship between the languages.

The idea of transfer does not then include decline; taking something from A and putting it in B is different from something dying out. Multi-competence includes this possibility by insisting on a complex network of relationships between languages, among which transfer is only one. The attrition relationship of multi-competence allows language growth and decline in the network. Transfer, on the other hand, denotes a relationship between existing elements, not bringing a language into being or into non-existence.

So what is it that changes in first language attrition? We have seen that the transfer relationship goes in both directions and involves systems rather than items. Under pressure from the VOT in one language, the VOT in the other language changes; ‘any of the languages in a multilingual system may change’ (Opitz, 2013: 753). This is a change in the system, not a loss. But how far can such change go before it constitutes loss, before the person can no longer be said to command their first language or other attriting language?

Meanings of language and transfer

De Saussure once remarked, 'D’autres sciences opèrent sur des objets donnés d’avance et qu’on peut considérer ensuite à différents points de vue; dans notre domaine, rien de semblable’ (De Saussure, 1915/1976: 23) (trans 1959: 8: ‘Other sciences work with objects that are given in advance and that can then be considered from different viewpoints; but not linguistics.'). Everything in SLA research therefore depends upon what is meant by language, as argued in Cook (2010): language is not a primitive term given in advance. The language used in studying bilingual networks (Li Wei, 1994) is hardly the same as that in psycholinguistics research like Sebastián-Gallés & Bosch (2005), post‑modern studies of discourse such as Block (2007), Vygotskyan studies such as Lantolf (2000), or procedural/declarative models like Ullmann (2001). The differences between generative-based multilingualism research (Cabrelli Amaro et al, 2012), and usage-based theorists (Ellis, 2002) are irreconcilable because of their different understanding of language itself.

According to Chomsky, language is a derived notion – an epiphenomenon: 'The grammar in a person's mind/brain is real ... The language (whatever that may be) is not' (Chomsky, 1982: 5). To refer to something in the mind as language is a convenient shorthand label for whatever this mental representation consists of – rules, items, patterns, or whatever. In this sense, language is a countable noun, so the mind might contain many languages. The word language is needed to be able to talk about first language, second language and so on. This does not establish these ‘languages’ as discrete components in the whole language system in the mind, where they may in fact be totally intertwined; it is a convenience in order to be able to distinguish subsystems within the whole. Claiming that existing ‘language’ A in the L2 user’s mind has a cross-linguistic influence on emerging ‘language’ B or vice versa does not commit one to a belief that these language systems are countable objects in the mind.

In another sense the word language labels an abstract system independent of the mind of the speaker, reflected in published grammars and dictionaries of English, not the individual mind. In this sense a language belongs to Popper’s third world of abstract ideas (Popper, 1972) or the linguistic realism of Katz and Postal (1991); it is language in an ideal form, which does not even require a community of speakers, as in ‘dead’ languages like Latin. This is the meaning of language that became associated in eighteenth century Europe with ideas of national identity (Anderson, 1983); French is a proud possession of the French, German of the Germans. The notions of two languages in the mind and two languages in the world of ideas are radically different, yet both are covered by language. A person who speaks English and French does not possess the essentially unknowable abstractions of the English and French languages. Transferring something from English to French is not a matter of relating an item recorded on p.295 of someone’s grammar of English to one on p.342 of someone’s grammar of French but of seeing how a node in one subsystem in the mind affects a node in another subsystem: the two meanings of language and the two processes are not commensurate.

Neither of these meanings, however, brings in the meaning of language as ‘an instrument used by the members of the community to communicate with each other’ (Labov, 1972: 277). Language is for interacting with other people, at the same time communicating information and building and maintaining social relationships with them, ‘making sense of our experience, and acting out social relationships’, the ideational and interpersonal metafunctions (Halliday & Mattheissen, 2013: 30). The relationship between two languages in this sense is then that the L2 user can carry out these functions with more than one community of speakers. Some of this communicative ability may be transferrable from one language to another, some may not; in Halliday’s terms, unlike the native child, the L2 learner does not have to learn how to mean (Halliday, 1973).

It is clear that any theory of transfer has to specify just what aspect of ‘language’ is involved. Traditionally transfer has been seen as the carrying over of mental elements of the language from one language to another. But this is not carrying over elements from one abstract language to another: it is fatal to assume that the English grammatical structures in the mind are the same as those in the grammar of English in the grammar book.

The relationships between mental languages in the mind of the L2 user are then difficult to pin down. Transfer is not taking an item from Language A and putting it in Language B, moving it from one pigeonhole in the mind to another. Attrition is not an item vanishing from Language A; codeswitching is not taking an item from Language A and putting it in Language B. Rather they are all aspects of an overall system in the mind.

 Creating new aspects of cognition

So far we have seen the relationships between the languages in the mind as links between some kinds of mental entities, which increase or decrease in strength in diverse ways. This means that the entire relationship can be stated in terms of things that already exist; if we can fully describe what someone knows in the first language and in their other languages, we can see everything as interactions between them. Nothing new needs to be added as it is all describable somewhere within the existing overall system. In a sense transfer is not about the acquisition of new knowledge or behaviour, i.e. learning, but about the rejigging of existing knowledge or behaviour into new configurations. It is a partial explanation of second language acquisition and use because it cannot account for the acquisition of new information or for the loss of old information, or as we see below, for the unique state of L2 users unrelated to L1 or L2.

What does this mean for the increasingly important area of bilingual cognition – how people think who know more than one language? A testcase can be how people see colours, a topic pioneered in SLA research by Athanasopoulos (2009). In Bassetti and Cook (2011) four possible scenarios for L2 users are spelled out, illustrated in Figure 1. Suppose that we are dealing with monolingual native speakers of a language that recognises two distinct colours, for example Greek ble, in English a light Cambridge blue, and ghalazio, in English a dark Oxford blue (Athanasopoulos, 2009). What happens when they acquire a language with a single colour spanning both their colours, like English blue? 

 

 i) The one-concept scenario. In this, people do not think differently when they learn another language; the same concept is used across languages regardless of which they are speaking: the original, L1-related concept is used in both languages. So a Greek person who has learnt English will still think in terms of two colours equivalent to ble and ghalazio even when using the English word blue.

ii) The double-concepts scenario. In this, L2 users switch concept according to the language they are speaking, using L1-related concepts when speaking the first language, L2-related concepts when using the second. Their minds hold two sets of concepts, which come into play when required. So the Greek/English speaker will think in terms of two blues when using Greek but one blue when speaking English.

iii) The one-integrated-concept scenario. Here the speakers adopt a single concept that integrates the L1-related concept and the L2-related concepts. Their thinking differs from monolingual native speakers of both languages. So the Greek/English speaker will have neither the two Greek blues nor the single English blue but a colour that is a compromise between the two.

iv) The original-concept scenario. The final logical possibility is that L2 users devise a new concept that is not so much intermediate between the L1-related and L2 concepts as something different. The Greek/English speaker will not think of blue in terms of the dark/light difference, say, but might have a blue that is characteristic of neither first language. Indeed the Greek-English L2 users have a concept of ‘ghalazio’ (‘light blue’) that is lighter than Greek monolinguals (Athanasopoulos, 2009).

These four scenarios portray four different outcomes to the relationship between L1 and L2 concepts. In (i) the user is unaffected by the L2 concepts, in (ii) the user switches between the L1 and the L2 concepts, and in (iii) the L1 and the L2 concepts are merged so that the L2 user does not think like an L1 user whichever language is used. All of these are indeed statable in terms of relationship between the existing concepts in the mind so we can talk of transferring a concept, of losing a concept and switching between concepts.

Scenario (iv) however is different: something is created that is not predictable from the relationships between L1 and L2: something new has come into being. The interesting problem for SLA research is describing things that cannot be anticipated from the given, the cases when 2+2=5. We have seen that measuring the L2 by the L1 of monolinguals gives an unfair result; we now see that accounting for the joined L1 and L2 of the L2 user in terms of the combined L1 and L2 is also inadequate as the unique constructs of the L2 user elude an analysis built only on the L1 and L2; comparison is not enough.

 Conclusions

This paper has tried to explore different issues in incorporating transfer within the multi-competence perspective. It has succeeded mostly in raising difficulties with the typical use of transfer as a relationship in which something in one language affects something in another. Transfer needs to be seen as more than a one-way relationship between the first and second language but as consisting of multiple directional relationships between multiple languages. It also means seeing transfer as one among several relationships between the languages of multi-competence, such as attrition and codeswitching. Additionally it concerns the very nature of language as a reality in the L2 user’s mind. Crucially the study of the relationships between the languages of multi-competence involves seeing how the state of multi-competence is not the same as two monolinguals combined but something of its own, not statable solely in terms of the thinking or language of monolinguals.

 References

Andersen, R. (1983) Transfer to somewhere. In Gass, S. and Selinker, L. (eds.) Language Transfer in Language Learning (177-201). Rowley, Mass.: Newbury House

Anderson, B. (1983) Imagined Communities. New York: Verso

Aronin, L. and Singleton, D. (2012) Multilingualism. Amsterdam: John Benjamins

Beauvillain, C. and Grainger, J. (1987) Accessing interlexical homographs: some limitations of a language-selective access. Journal of Memory and Language, 26, 658-672

Block, D. (2007) Second Language Identities. London: Continuum

Cabrelli Amaro, J., Flynn, S. and Rothman, J. (eds.) (2012) Third Language Acquisition in Adulthood. Amsterdam: John Benjamins

Chomsky, N. (1965) Aspects of the Theory of Syntax. Cambridge, Mass.: MIT Press

Chomsky, N. (1982) Some Concepts and Consequences of the Theory of Government and Binding. Cambridge, Mass.: MIT Press

Cook, V.J. (1991) The poverty-of-the-stimulus argument and multi-competence. Second Language Research, 7, 2, 103–17

Cook, V.J. (2000) Is transfer the right word? Unpublished paper given at IPRA 2000 in Budapest

Cook, V.J. (ed.) (2003) Effects of the L2 on the L1. Clevedon: Multilingual Matters

Cook, V.J. (2010) Prolegomena to second language learning. In P. Seedhouse, S. Walsh and C. Jenks (eds.) Conceptualising Language Learning (6-22). Basingstoke: Palgrave MacMillan

Cook, V.J. (2015) Premises of multi-competence. In V.J. Cook & Li Wei (eds.) The Cambridge Handbook of Linguistic Multi-competence. Cambridge University Press

De Bot, K. (2015) Multi-competence and dynamic/complex systems. In V.J. Cook & Li Wei (eds.) The Cambridge Handbook of Linguistic Multi-competence. Cambridge University Press

De Groot, A. (2010) Language and Cognition in Bilinguals and Multilinguals. Hove, UK: Psychology Press

De Saussure, F. (1916/1976) Cours de Linguistique Générale, edited by Bally, C., Sechehaye, A. and Reidlinger, A. Reprinted (1976) Paris: Payot. Translation by W. Baskin (1959) Course in General Linguistics. London: Peter Owen

Dechert, H.W. (2006) On the ambiguity of the notion ‘transfer’. In Arabski, J. (ed) Cross-linguistic Influences in the Second Language Lexicon (3-11). Clevedon: Multilingual Matters

Ellis, N.C. (2002) Frequency effects in language processing – a review with implications for theories of implicit and explicit language acquisition. Studies in Second Language Acquisition, 24, 143-188

Epstein, S.D., Flynn, S. and Martahardjono, G. (1996) Second language acquisition: theoretical and experimental issues in contemporary research. Brain and Behavioral Sciences, 19, 4, 746-752

Flege, J.E. (2002) Interactions between the native and second language phonetic systems. In Burmeister, P., Piske, T. and Rohde, A. (eds.) An Integrated View of Language Development (217-244). Trier: Wissenschaftliche Verlag

Flynn, S., Foley, C. and Vinnitskaya, I. (2004) Cumulative-Enhancement Model for language acquisition; comparing adults’ and children’s patterns of development in first, second and third language acquisition of relative clauses. International Journal of Multilingualism, 1, 1, 3-16

Friesen, D.C. and Jared, D. (2011) Cross-language phonological activation of meaning: evidence from category verification. Bilingualism: Language and Cognition, 15, 1, 145-156

Green, D.W. (1998) Mental control of the bilingual lexico-semantic system. Bilingualism: Language and Cognition, 1, 77–82

Grosjean, F. (2008) Studying Bilinguals. Oxford: Oxford University Press

Halliday, M.A.K. and Mattheissen, C. (2013) An Introduction to Functional Grammar. 4th edition. London: Hodder Educational

Hauser, M.D., Chomsky, N. and Fitch, T.M. (2002) The faculty of language: what is it, who has it and how did it evolve. Science, 298, 1569-1579

Hawkins, R. and Chen, C. (1997) The partial availability of Universal Grammar in second language acquisition: the ‘failed functional features’ hypothesis. Second Language Research, 13, 3, 187-226

Hermans, D., Ormel, E., van Besselaar, R. and van Hell, J. (2011) Lexical activation in bilinguals’ speech production is dynamic: how language ambiguous words can affect cross-language activation. Language and Cognitive Processes, DOI:10.1080/01690965.2010.530411

Hirata-Edds, T. (2011) Influence of second language Cherokee immersion on children’s development of past tense in their first language, English. Language Learning, 61, 3, 700–733

Hu, A. (2013) Transfer. In Byram, M. and Hu, A. (eds.) The Routledge Encyclopedia of Language Teaching and Learning (732-734). Abingdon: Routledge

Jarvis, S. & Pavlenko, A. (2009) Crosslinguistic Influence in Language and Cognition. Abingdon: Routledge

Katz, J.J. and Postal, P.M. (1991) Realism vs. Conceptualism in linguistics. Linguistics and Philosophy, 14, 5, 515-554

Kellerman, E. (1995) Crosslinguistic influence: Transfer to nowhere. In W. Grabe (ed.), Annual Review of Applied Linguistics, 15, 125-150

Kellerman, E. and Sharwood-Smith, M. (eds.) (1986) Crosslinguistic Influence and Second Language Acquisition. Oxford: Pergamon Press

Kopke, B. and Schmid, M. (2004) Language attrition: the next phase. In Schmid, M., Kopke, B., Keijzer, M. & Weilemar, L. (eds.) First Language Attrition (1-43). Amsterdam: John Benjamins

Krashen, S. (1982) Principles and Practice in Second Language Acquisition. Oxford: Pergamon Press

Kulundary, V. and Gabriele, A. (2012) Examining the role of L2 syntactic development in L3 acquisition: a look at relative clauses. In Cabrelli Amaro et al (eds.), 195-222

Labov, W. (1972) Sociolinguistic Patterns. University of Pennsylvania Press

Lantolf, J. (ed.) (2000) Sociocultural Theory and Second Language Learning. Oxford: Oxford University Press

Li Wei (1994) Three Generations, Two Languages, One Family: Language Choice and Lan­guage Shift in a Chinese Community in Britain. Clevedon, Avon: Multilingual Matters

Odlin, T. (1989) Language Transfer: Cross-Linguistic Influence in Language Learning. Cambridge University Press

Oxford English Dictionary (2009), Oxford: Oxford University Press, online at http://www.oed.com/

Opitz, C. (2012) A dynamic perspective on late bilinguals’ linguistic development in an L2 environment. International Journal of Bilingualism, 17, 6, 701–715

Popper, K.R. (1972) Objective Knowledge. Oxford: Clarendon Press

Schwartz, B. and Sprouse, R. (1996) L2 cognitive states and the Full Transfer/Full Access model. Second Language Research, 12, 1,40-72.

Sebastián-Gallés, N. and Bosch, L. (2005) Phonology and bilingualism. In J.F. Kroll and A.M.B. de Groot (eds.) Handbook of Bilingualism: Psycholinguistic Approaches (68-87). Oxford: Oxford University Press

Selinker, L. (1972) Interlanguage. International Review of Applied Linguistics, X, 3, 209-231.

Seton. B. and Schmid, M. ( 2015) Multi-competence and first language attrition. In V.J. Cook & Li Wei (eds.) The Cambridge Handbook of Linguistic Multi-competence. Cambridge University Press

Spivey, M.J. and Marian, V. (1999) Cross talk between native and second languages: partial activation of an irrelevant lexicon. Psychological Science, 10, 181-84

Spivey, M.J. and Marian, V. (2003) Competing activation in bilingual language processing: Within- and between-language competition. Bilingualism: Language and Cognition, 6, 2, 97–115

Tokumaru, Y. (2002) Cross-linguistic influences of L2 English on L1 Japanese in Japanese-English bilinguals. ACTAS/Proceedings II Simposio Internacional Bilinguismo, 399-406

Ullman, M. (2001) The neural basis of lexicon and grammar in first and second language: the declarative/procedural model. Bilingualism: Language and Cognition, 4, 1, 105-122

Weinreich, U. (1953) Languages in Contact. The Hague: Mouton

Wrembel, M. (2011) Crosslinguistic influences in third language acquisition of Voice Onset Time. ICPhS, XVII, 2157-2160