Of all of the contact phenomena of interest to researchers and students of bilingualism, code-switching (hereafter CS) has arguably dominated the field. Broadly defined, CS is the ability on the part of bilinguals to alternate effortlessly between their two languages. This capacity is truly remarkable and invites scientific and scholarly analysis from professionals, but, at the same time, generates a great deal of pointed discussion that reflects popular misperceptions of the nature of CS in particular and bilinguals more generally. While CS is viewed as an index of bilingual proficiency among linguists, it is more commonly perceived by the general public as indicative of language degeneration. This disparity can be best understood by reference to notions of grammar. Most laypeople define grammar as a set of statements about how we should correctly use our language. Such an understanding of grammar is properly called prescriptive, because it attempts to mandate or prescribe the way language should be used. Linguists, who study language objectively, are more interested in descriptive grammars, which represent speakers’ unconscious knowledge of their languages as manifested in their actual linguistic behavior. Bilinguals in language contact situations commonly use forms that integrate their two languages to some degree, a behavior that is disparaged by language purists, who insist that each language maintain its integrity according to prescribed norms. For the linguist, on the other hand, CS provides a unique window on the structural outcomes of language contact, which can be shown to be systematic rather than aberrant. Further, the act of CS can be studied as a reflection of social constructs and of the cognitive mechanisms that control language switching. From the perspective of linguistics, then, CS is worthy of study for a variety of reasons.
The significance of this phenomenon in illuminating bilingual cognition and behavior cannot be underestimated, first and foremost because CS is exclusive to bilinguals. Nevertheless, many controversies exist in the study of CS, in large part because the phenomenon has been approached from different disciplinary perspectives, and as a consequence has evaded a uniform definition and explanation. The purpose of this chapter is to present an overview of CS from the perspective of linguistics, with a view towards defining CS, identifying who engages in CS and for what purposes, and delineating the various approaches to the study of CS. The overarching goal of the chapter is to set out why the study of CS is important, and by so doing to dispel misconceptions regarding language alternation among bilinguals.
All speakers selectively draw on the language varieties in their linguistic repertoire, as dictated by their intentions and by the needs of the speech participants and the conversational setting. Even monolinguals are capable of shifting between the linguistic registers and the dialects they command and, as such, there are parallels that can be drawn between monolingual and bilingual language use. For convenience, we can refer to such monolingual behavior as style shifting. In turn, bilinguals have available not only different registers and dialects of one language, but of two. As is true of monolingual style shifting, it is not uncommon for bilinguals to segregate their languages, speaking exclusively in one language in certain domains (e.g. at home, with friends) while shifting to another in other contexts (e.g. school, work), a bilingual behavior commonly referred to as language shifting. Given the appropriate circumstances, many bilinguals will exploit this ability and alternate between languages in an unchanged setting, often within the same utterance; this is the phenomenon understood as CS.
CS comprises a broad range of contact phenomena and is difficult to characterize definitively. First, its linguistic manifestation may extend from the insertion of single words to the alternation of languages for larger segments of discourse. Second, it is produced by bilinguals of differing degrees of proficiency who reside in various types of language contact settings, and as a consequence their CS patterns may not be uniform. Finally, it may be deployed for a number of reasons: filling linguistic gaps, expressing ethnic identity, and achieving particular discursive aims, among others. Given these factors, it is not surprising that there exists debate in the literature concerning the precise characterization of CS and how various kinds of language contact varieties are to be classified.
An incontrovertible example of CS is to be found in the English–Spanish bilingual title of Poplack’s (1980) seminal article:
Note that there are readily identifiable constituents from English and Spanish and that their combination here does not violate the grammar of either language. This type of language alternation has been termed Classic CS (Myers-Scotton 1993a) or alternational CS (Muysken 2000), but is most widely known as intra-sentential CS (Poplack 1980). This contrasts with inter-sentential CS, as in (2), where alternation occurs at clause boundaries.Sometimes I’ll start a sentence in Spanish [sic] y termino en español
“. . . and I finish in Spanish.”
Like intra-sentential CS, inter-sentential switching requires an advanced level of bilingual proficiency as it often entails the production of full clauses in each language. However, the former, but not the latter, can offer insights into the ways in which the two grammars of the bilingual interact at the sentence level.That’s too much. Sina pesa.
“. . . I don’t have [much] money.”
(Myers-Scotton 1993a:41)
Muysken (2000) advances a typology of CS patterns, suggesting that bilinguals employ three distinct strategies: alternation, where the two languages remain relatively separated in an A–B configuration, as exemplified in (1) and (2) above; congruent lexicalization, in which the two languages share a common grammatical structure that can be filled with lexical elements from either language, as in (3); and insertion, which involves the embedding of a constituent – usually a word or a phrase – in a nested A–B–A structure, as in (4).
wan heri gedeelte de ondro beheer fu gewapende machten
one wholepart cop under control of armed force
“One whole part is under control of the armed forces.”
(Bolle 1994:75, cited in Muysken 2000:139)
Congruent lexicalization is most prevalent between languages that are closely related typologically (Sranan in (3) is a Dutch-based creole). Alternations such as in (3) have been analyzed as constituting a composite matrix language (Myers-Scotton 2003), which arises “when speakers produce structures for which the source of structure is split between two or more varieties (2003:99).” Myers-Scotton further maintains that composite structures arise in contexts of language shift. For this reason alone, congruent lexicalization differs from Classic CS, i.e. intra-sentential CS, for which it is assumed bilinguals fully maintain both language systems. Insertion, as in (4), can also arguably be viewed as distinct from intra-sentential CS, as it has much in common with lexical borrowing, which does not necessitate bilingual proficiency. Similarly, tag-switching may also occur among bilinguals with limited abilities in one language, as it is defined by the insertion of a formulaic expression from language B (e.g. so, well, d’accord?) into an utterance in language A, primarily for pragmatic effect, as in (5).xob pas falsk-an pesa-â
well then false-cop3pl boy-pl
“Well then boys are false.”
(Naseh Lotfabbadi 2002:101)
Les autres pourraient [sic] parler français comme lui, ya know
“The others could speak French like him, . . .”
(Bullock fieldnotes)
In brief, although all of the above forms can be classified as CS, it is Classic or intra-sentential CS that may reveal the most about language structure. Consider, again, the example in (1). Because Spanish and English have similar surface structures for this expression, the lexemes can be aligned more or less in a one-to-one fashion. Thus, numerous other CS patterns should be possible. However, consider the hypothetical examples in (6):
In (6a), the alternation between English and Spanish occurs at every other word. In (6b, 6c), the switching is less frequent, allowing for longer stretches of English and Spanish. Significantly, though, none of these sentences would be attested nor accepted among Spanish–English bilinguals because each is in violation of core principles of CS.(6)
a. *Sometimes yo will empezar a oración in inglés and termino in español.
b. *Sometimes I’ll empezar una oración en inglés y I finish in Spanish.
c. *A veces yo will start a sentence in English and I termino en español.
Clearly, CS is not the random mixing of two languages, as is popularly assumed. Nevertheless, this misperception endures, as evidenced by the various metaphors and terms ascribed to bilingual speech varieties. For instance, metaphors whereby contact varieties are likened to a mix of grains are common (e.g. trasjanka for mixed Russian–Byelorussian speech, literally “hay and straw,” and surzhyk for mixed Ukranian–Russian, literally “wheat and rye”). Also common are portmanteau creations such as finnglish, ingleñol, franglais, portinglês, and so on. While these terms are playful they often carry pejorative connotations that the speech varieties they reference are nothing but a linguistic hodgepodge and that the speakers who use them are uneducated and incapable of expressing themselves in one or the other language. However, a significant body of research has amply demonstrated that CS does not represent a breakdown in communication, but reflects the skillful manipulation of two language systems for various communicative functions. This is articulated through a different – and to our mind, more apt – metaphor offered by Valdés: “[I]t is helpful to imagine that when bilinguals code-switch, they are in fact using a twelve-string guitar, rather than limiting themselves to two six-string instruments (1988:126).”
CS is to be distinguished from other types of contact phenomena, although it is not always the case that clear-cut distinctions can be drawn. For instance, as noted above, insertional CS can be equated with borrowing. However, the term borrowing has been used to describe many different forms, from the transfer of structural features (e.g. phonemes, suffixes) to that of whole clauses. Lexical borrowing normally involves the morphological and phonological integration of a single lexeme, as in the Japanese word basubaru, from English “baseball,” which is fully established in the monolingual Japanese lexicon. But unassimilated loan words, also called nonce borrowings (Poplack et al. 1988), can occur spontaneously in the speech of bilinguals, blurring any boundary that can be drawn between these contact forms on structural criteria alone. It is evident that nonce borrowing is akin to CS because both are attested in the speech of bilinguals and unlikely to be found in that of monolinguals; hence some researchers (e.g. Treffers-Daller 1991; Myers-Scotton 1993a) view borrowing and CS as falling along a continuum (see Bullock, Treffers-Daller, this volume).
Other contact forms are more easily differentiated from CS, although they too implicate the transfer of material from one language into the other (see Treffers-Daller, this volume, on the relation between transfer and CS). Loan translations or calques, as in (7), involve the importation of foreign patterns or meanings with the retention of native-language morphemes (see Backus and Dorleijn, this volume). Also attested in bilingual speech are cross-linguistic semantic extensions, where a word from language A takes on additional meanings that are modeled by language B, as in (8).
(7)
a. US Chicano Spanish
escuela alta “high school”
literally “school high”
(cf. Spanish secundaria)
b. US French
étudiant gradué “graduate student”
literally “student graduated”
(cf. French étudiant de troisième cycle)
(8)
a. US Spanish
mayor [madƷor] “mayor”
literally “older”
(cf. Spanish alcalde)
b. US French
enregistrer “register (for a course)”
literally “check a bag”
(cf. French s’inscrire)
CS is also distinct from mixed languages, which are contact varieties that derive components of their grammatical systems from diverse genetic sources. For example, Media Lengua, spoken as a native language in the highlands of Central Ecuador, has been described as a prototypical mixed language (Muysken 1988, 1996). The general properties of Media Lengua include Quechua morpho-syntax combined with Spanish lexical stems, as shown in (9).
Media Lengua is structurally distinct from Quechua and Spanish and is not intelligible to monolingual speakers of those languages. Unlike mixed languages, CS does not constitute a composite or hybrid system. However, it is conceivable that mixed languages may have arisen within communities where bilingual CS was prevalent; indeed, this is explicitly argued to be the case for at least one mixed language, Gurindji Kriol, spoken in the Northern Territory of Australia (McConvell and Meakins 2005). However, the origins of most mixed languages are not well understood and whether CS lies at their source remains an issue of debate within contact linguistics.Unu fabur-ta pidi-nga-bu bini-xu-ni
one favor-acc ask-nom-ben come-prog-1sg
“I come to ask a favor”
cf. Quechua: Shuk fabur-ta maña-nga-bu shamu-xu-ni
one favor-acc ask-nom-ben come-prog-1sg
cf. Spanish: Vengo para pedir un favor
I-come for ask-inf a favor
(Muysken 1981:68–69)
Finally, CS should not be confused with diglossia. Diglossia describes a community where languages or language varieties are functionally compartmentalized. Within such a situation, each language form is associated with a particular social function. A well-cited example is the functional distribution of languages in Paraguay, where Spanish is used in official and institutional contexts, and Guaraní is relegated to informal domains. In diglossic settings, the selection of which language to use is not free, but determined by community norms; that is, diglossia is socially imposed. In contrast, CS is understood as an individual phenomenon wherein a speaker chooses when, why, and how to alternate between languages.
Any healthy individual who speaks more than one language has the capacity to select the appropriate language in a given situation. Only in certain instances of brain damage is language selection impaired with pathological switching as a result (see Kutas et al., this volume). CS, in the normal case, is under the conscious control of the speaker and, significantly, not all bilinguals are observed to engage in CS. Thus, a relevant domain of inquiry is to examine the individual, discursive, and social conditions under which a bilingual deploys both languages simultaneously. Since CS is manifested only in the speech of the bilingual it is also necessary first to ask, who is a bilingual?
“Bilingual” is a cover term that encompasses speakers who fall along a “bilingual range,” a continuum of linguistic abilities and communicative strategies (Valdés 2001). As a consequence, there may be a relationship between a speaker’s place in the bilingual continuum and the quality and quantity of CS attested. Therefore, a careful consideration of how a bilingual is defined is in order. The layperson’s definition holds that a bilingual is an individual who has native-like control of two (or more) languages (a definition also offered by the linguist Bloomfield in 1933). Specialists, too, have employed terms such as balanced bilingual, true bilingual, and symmetrical bilingual to describe such a person. But consider what this would involve: no accent, no non-target word selection, and the ability to converse on any subject with any interlocutor at any time in either language. Such a bilingual would be like the putative “two monolinguals in one,” a metaphor made current by Grosjean (1998). However, monolingual-like control of two languages over all aspects of linguistic knowledge and use within all domains is rare, if possible at all. Most bilinguals show disparate abilities in their component languages, for a myriad of reasons, including age of second language acquisition, the quality of linguistic input received, the language most used, and the status of the language in the community.
Speakers who have been exposed to two languages from birth or early childhood – simultaneous or early bilinguals – and who have maintained the use of their languages throughout their lifespan most closely approximate what is meant by true bilingual. These speakers possess advanced linguistic and communicative abilities in both languages, and are able to deploy each as required. The examples in 10, taken from Köppe and Meisel (1995:285), show the language alternations of Ivar (age 2;05), as he interacts with a French-speaking interviewer (F) and a German-speaking interlocutor (G).
As demonstrated in (10), Ivar is clearly able to separate his languages yet, at the same time, he is also reported to code-switch. Veh (1990) and Meisel (1994) report a high rate of language mixing for Ivar until around the age of 2;05, mostly between deictic elements and nouns (11a) and between verbs and nouns (11b). In (11c), Ivar uses translation equivalents, a common strategy for emphasis among bilinguals.(10)
Iv (to G): oh der kann nich fahr(en) der auto
“Oh this one can’t move the car.”
F: qu’est-ce qu’elle a fait l’auto la voiture?
“What has it done the auto the car?”
Iv: peut peut pas rouler
“can can not move”
(11)
a.
das bateau (2;00,02)
“this boat”
(Köppe and Meisel 1995:291)b.
sent füβe (2;04,09)
“smell feet”
(Köppe and Meisel 1995:291)c.
j’ai trouvé – i gefunden diese! (2;08,15)
“I have found – I (have) found these.”
(Schlyter 1990:114)
CS, then, is not indicative either of the bilingual’s inability to separate his languages or of a lack of proficiency. Rather it is an additional communication resource available to bilinguals. CS also speaks to a bilingual’s competence in each of the two languages. Various researchers have demonstrated that the ability to switch at the intra-sentential level correlates with increased mastery of linguistic structures. In particular, Genesee and his colleagues in Canada and Meisel and his colleagues in Germany have charted the language patterns of bilingual children and demonstrated that development in each language proceeds independently and that as the acquisition of the syntax of their component languages progresses, their language mixing patterns become more adult-like (see Müller and Cantone, Miccio et al., this volume).
Simultaneous bilingualism is frequently encountered in immigrant and guest-worker communities. Such communities also give rise to second generation, or heritage, bilinguals who, unlike their parents, may be dominant in the majority language. As their contact with the majority language increases, their use of and exposure to the home language may become more restricted. Thus, in addition to CS, their productions may also demonstrate features that are typical of language attrition including loan translations or calquing, semantic extensions and convergence (see Sebba, Bolonyai, this volume). It is usually the case that by the third generation the descendants of immigrants will have shifted to the dominant language, retaining only residual, formulaic traces of the heritage language in their speech. Oftentimes, this may be reflected in their CS patterns, which have been reduced to lexical insertion and/or tag-switching. In this respect, heritage speakers’ linguistic forms come to resemble those attested among second language learners.
Second language acquirers or late bilinguals are those who have a linguistic system fully in place when their exposure to the second begins. Clearly, under this definition, we find a vast range of patterns of acquisition and outcomes. Naturalistic or folk bilinguals who learn a second language without formal instruction (e.g. immigrants and guest-workers) will differ greatly from so-called elite bilinguals whose language learning is primarily classroom based. Not only do these two types of bilinguals differ according to the context of second language learning, but they may also differ in terms of motivation. For many naturalistic bilinguals, second language learning is a necessity, as they cannot function easily in the dominant society without such knowledge. Elite bilinguals, on the other hand, often choose to learn a second language for personal or professional gain. Among speakers of both groups, particularly in the early stages of acquisition, CS results from an inability to produce a target form. Due to temporary or permanent lapses in knowledge, learners may switch to the native language, a process referred to as crutching. But as their proficiency develops, CS among second language learners and folk bilinguals, if attested, will resemble that of more fluent bilinguals. Thus, even among incipient bilinguals, CS patterns may be used as a measure of bilingual ability, rather than deficit. In fact, the degree of language proficiency that a speaker possesses in two languages has been shown to correlate with the type of CS engaged in. Poplack (1980) observes that adult bilinguals who reported to be dominant in one language tended to switch by means of tag-like phrases; in contrast, those who reported and demonstrated the greatest degree of bilingual ability favored intra-sentential switches. Similar patterns were attested among the school-age children studied by McClure (1981), who concludes that
. . . just as the monolingual improves his control over his verbal resources with age, so too does the bilingual. Further, just as there is a developmental pattern in the monolingual’s syntactic control of his language, so too may such a pattern be found in the bilingual’s control of the syntax of code-switching, which begins with the mixing of single items from one code into discourse in the other and culminates in the code changing of even more complex constituents (1981:92).
Despite the fact that CS has been shown to index bilingual linguistic and communicative skills rather than shortcomings, CS remains largely stigmatized. Nevertheless bilinguals do choose to code-switch, a decision that is influenced by a number of social and discursive factors (see Gafaranga, Gardner-Chloros, Khattab, this volume). At the community level, the persistence of CS may reflect the covert prestige ascribed to this linguistic behavior. In particular, CS may serve as a marker of group membership and solidarity. Importantly, bilinguals only code-switch with other bilinguals with whom they share a dual language identity. For many, CS is a speech form that allows for the expression of their membership in two cultures: the dominant and the minority. Within some strata of bilingual communities, CS carries overt prestige. For example, Sankoff (1980) reports that in some areas of lowlands New Guinea, villagers are trilingual in Buang, Tok Pisin, and Yabem, and switching among them is the most prestigious form of public-speaking and is expected of persons in possession of power. In other instances, switching into a particular language may confer status on a speaker. For example, in Bulgaria, trilingual Muslim Roms who speak Romani, Bulgarian, and Turkish will code-switch into Turkish, as it has higher prestige than the other languages they command (Kyuchukov 2006).
There are also discursive functions that motivate the presence of CS in bilingual conversation. These pertain to the speaker’s communicative intentions. Gumperz, in his seminal work on bilingual discursive strategies (1976, 1982a), describes many important functions served by CS. The premise underlying his and subsequent studies is that CS is a conscious choice on the part of the speaker, used to mark quotations, emphasis, realignment of speech roles, reiteration, and elaboration, among others. In (13), from Romaine (1995:162), a girl from Papua New Guinea inserts an English quotation from a cartoon into a Tok Pisin utterance. In (14), from Frenchville, PA (USA), the speaker uses English for translation/repair of an ill-formed French sentence that he is quoting. In (15), from Zentella (1997:94), the speaker switches from Spanish to English to mark a role shift. Finally, in (16), a Japanese–English bilingual uses Japanese to introduce the discourse topic (Nishimura 1985a, cited in Romaine 1995:163).
Lapun man ia kam na tok, “oh you poor pusiket,” na em go insait.
“The old man came and said, . . ., and then he went inside.”
Elle m’a dit, “il pleuve [sic] maintenant.” It’s raining now. That’s not good French, is it?
“She said to me ‘it’s raining now [cf. il pleut]’ . . .”
Mi nombre es Lourdes. Now we turn to my sister.
“My name is Lourdes. . .”
A great deal of research has been devoted to analyzing bilingual speech and ascribing particular functions to specific language alternations (see Gafaranga, this volume). However, it merits pointing out that not all language alternations in bilingual speech do signal a particular communicative intent or purpose; for many bilinguals, CS merely represents another way of speaking; that is, some bilinguals code-switch simply because they can and oftentimes may not be aware that they have done so.Yano-san-wa he was speaking all in English.
“As for Mr. Yano, he was speaking all in English.”