1. Introduction

Since Gumperz (1982:59) defined conversational code-switching “as the juxtaposition within the same speech exchange of passages of speech belonging to two different grammatical systems or subsystems”, there have been attempts at interpreting code-switching as a meaning-making act. By now, there is consensus among theorists that code-switching is not arbitrary. However, theorists vary as to the significance they contribute to the subjective, objective, and social factors as the most salient in interpreting the meaning-making functions of code-switching.

In the quest for providing a unifying, comprehensive, and universal framework of interpreting code-switching as a meaning-making act, a new perspective has been proposed by Bhatt and Bolonyai (2011: 2), focusing on the interpretation of the meaning-making functions of code-switching from a sociocognitive perspective. Bhatt and Bolonyai’s Optimality Theory for bilingual grammar is a sociocognitive, normative community framework interpreting the meaning-making function of code-switching in consideration of the cognitive, pragmatic, and social factors interplaying in the mechanism of code-switching. Bhatt and Bolonyai’s model provides a unified theoretical framework of how the sociopragmatically meaningful instances of code-switching can be assumed to index certain social constructs and to (re)negotiate the (con)textual framework within an ethnographically specific bilingual immigrant community’s linguistic repertoire. Bhatt and Bolonyai (2011) claim that the cognitive mechanism governing the sociopragmatic meaning-making functions of code-switching is universal, but its actual instantiation in a community’s linguistic repertoire is community-specific.

The aim of the present study is to demonstrate the applicability of the Optimality Theory for bilingual grammar proposed by Bhatt and Bolonyai (2011) on the Hungarian-American immigrant speech community in North Carolina, USA.

2. Optimality Theory for bilingual grammar

Optimality Theory (OT) (Prince and Smolensky 2004) is a generative grammar-based formal framework attempting to apply generative grammatical rules in order to describe how natural languages work. Its main premise is that the significant regularities of natural languages can be understood by analyzing the output structure, the surface realizations of utterances. OT claims that relying on an algorithm-based representation of empirically observed output representations, the actual rules governing linguistic mechanisms can be understood.

Bhatt and Bolonyai (2011) adopt Optimality Theory for bilingual language use and turn it into a comprehensive model describing the sociopragmatic meaning-making mechanism of code-switching. They claim that code-switching is a sociocognitive mechanism fulfilling an array of sociopragmatically interpretable functions. In the same vein as OT, they claim that by representing in algorithmic tableaux the empirically observed sociopragmatic functions that the act of code-switching fulfills, the sociopragmatic meaning-making mechanism of code-switching can be explored.

Claimed to be universally applicable in any bi- or multilingual speech community, Optimality Theory for Bilingual Use is a framework which aims to demonstrate how socio-cognitive constraints, in interaction with each other, filter the linguistic inputs to finally select the output indexing or constructing the optimal socio-pragmatic meaning and/or fulfilling the optimal socio-pragmatic function in a given utterance. In this framework five constraints are ranked hierarchically and are universal, but their ranking is community-specific. The inputs that have to undergo this hierarchically ranked set of constraints are monolingual and code-switched candidates.

Relying on the thorough and comprehensive research of previous literature on code-switching, pragmatics, and conversation analysis, Bhatt and Bolonyai’s five constraints, create, index and decode interaction, the optimal sociopragmatic meaning of code-switching in a given linguistic utterance. These universal optimality filtering constraints are as follows: the Principle of Interpretive Faithfulness (FAITH); the Principle of Symbolic Domination (POWER); the Principle of Social Concurrence (SOLIDARITY); the Principle of Face Management (FACE); and the Principle of Perspective Taking (PERSPECTIVE). ¹

Of all sociopragmatic-related functions of code-switching in the relevant literature, various functions have been subsumed under each constraint. The main socio-pragmatic function of code-switches classified under the Principle of Faithful Interpretiveness is to express the most economically and faithfully the intended meaning of the speaker when the semantic-conceptual attribute of the monolingual candidate does not allow its most optimal meaning-making formation. All Faith-related instances enable the speaker to index or construct the most optimal interpretive conceptual, ideological, socio-cultural meaning of an utterance in a community-specific, culturally bound context. Numerous functions of code-switching listed by other researchers can be classified under the principle of Faith: to name but a few examples, Faith-related instances of code-switching might express “highly specific” cultural connotations (Backus 2001), “stylistic embroidery” (Valdes-Fallis 1976; Callahan 2004), religious invocations (Callahan 2004), or linguistic routines or clichés (Montes-Alcala 2001).

Instances of code-switching that enable the speaker to express or negotiate socio-cognitive structures or relational frames according to the perceived or desired social status of participants in interrelation to one another have been subsumed under the constraint of Power. As a higher social status (dominance, power) cannot exist without presupposing a lower one (concurrence, solidarity), the constraint of Power is in a complementary position in relation with the constraint of Solidarity responsible for social concurrence.

A code-switch complying with the constraint of Power is a linguistic resource drawn upon to index, in unequal social relations, a higher, dominant, or superior position among the participants of a linguistic situation. To mention some of the socio-pragmatic functions of code-switches classified under the constraint of Power: Power includes functions expressing “they-code” (Gumperz 1982), “authority” (Lin 1990; Canagarajah 1995), “topping” an argument (Gal 1979), and “power-wielding” (Auer 1998; Jorgensen 1998; Wei 1998; Esdahl 2003).

Under the constraint of Solidarity, linguistic resources, specifically, instances of code-switching, serve as means of expressing either a perceived lower position in an unequal situation or as means of expressing solidarity with or a sense of belonging to a group. In the Optimality Theoretical framework, numerous instances of code-switches cited by other theorists have been subsumed under the constraint of Solidarity. These instances express the disposition of the participants, acting as social actors in an interaction where the roles are hierarchical, based on affiliation, equality or solidarity rather than on domination, power, or authority. Instances expressing “we-code” (Gumperz 1982), “intimacy” (Bhatt and Bolonyai 2011), “inclusion” (Canagarajah 1995), and the “default language” (Meeuwis and Bloomaert 1998) have been subsumed under this constraint.

Face is the constructed and approved public façade of a person that determines their social status and their interpersonal relationships. Face-work is the embracing term for all social and interactional practice that an individual gets engaged in to achieve or orient themselves to a desired social status. Face-work is a bidirectional activity: it involves certain social practices that challenge the self’s face schemas by others (face-threatening acts) as well as the practices deployed by the self in order to minimize or avoid face threat (mitigating, minimizing, avoiding face-threatening acts). Code-switched instances which pose a potential threat to the speaker’s face needs as well as all those mitigating these threats are listed under the constraint of Face management. For example, instances of code-switching avoiding “risking loss of face” (Gumperz 1982), “mitigating or defusing face threats” (Heller 1988; Myers-Scotton and Bolonyai 2001), “dampening directness” (Gardner-Chloros and Finnis 2003), and “mitigating request” (Zentella 1997).

The main function of code-switching complying with the constraint of Perspective is to accentuate some aspect of bi- or multilingual reality against some other aspect either by contrasting them, by placing them into simultaneous vision, or by bringing them into a common focus. Hence, the constraint of Perspective enables the speaker to construct and focus on some aspect of reality from the speaker’s prominent point of view. The act of code-switching under the principle of Perspective fulfills its main discourse-related function, that is, constructing and focusing on the desired aspect of discursive reality (the time, the place of the setting, the voice of participants) relying on its conversational resources, such as “quotations” (Gal 1979; McClure and McClure 1988; Auer 1995), emphasis (Gumperz 1982; Callahan 2004), and “contextualization cues” (Gumperz 1982; Wei 1994; Auer 1995).

Not only does the principle of Perspective enable the speaker to construct and put into focus one aspect of reality, it also enables them to position themselves, to take a stance, in the discursive reality. Therefore, such discourse-related functions as “irony”, “sarcasm”, “parody” (Woolard 1988; Pandey 1995), “role-shift” (Auer 1995; Zentella 1997) which position the speaker in a distance from the constructed reality, are also subsumed under the constraint of Perspective.

3. Data and the method of analysis

The data presented in this paper come from sociolinguistic interviews conducted by the author and Ágnes Bolonyai in the Hungarian-American community of North Carolina in 2007 and 2008. ² In the course of `semi-structured` interviews, which were informal dinner conversations at Bolonyai’s home, there were narrative elicitation types of questions enquiring about the subjects’ experience of being a Hungarian-American. Altogether, 25 sociolinguistic interviews were conducted with 39 Hungarian-Americans.

The minimum time length of the interviews was 45 minutes, while the longest interview lasted 4 hours. Prior to the interviews, all subjects were informed that the interviews would be recorded, and they all gave their consent to this. Altogether, the whole sample consists of 54 hours of recorded sociolinguistic interviews. The conversations were transcribed to provide a text of 2,174 pages (in 12-point Times New Roman, double-spaced). A similar portion (5 pages, written in 12-point, Times New Roman, double-spaced, approximately 7,200 characters with spaces or 1,100 words) of the transcript was taken from each participant’s interview. When selecting the text to be analyzed, the middle part of the interview as well as the one toward its end was preferred. By the middle of the interview, subjects became comfortable with the interview situation and were open to speak about their immigrant experience. Toward the end of the interview, they became even more relaxed and some spontaneous conversations about varied topics occurred. The more relaxed atmosphere of the conversations prompted more spontaneous speech patterns and code-switched instances, which were of particular interest to us.

All instances of code-switching from the similar portion of text were counted and then analyzed qualitatively. The aim of the qualitative analysis is to demonstrate that the community specific ranking of the sociopragmatic constraints of the optimality bilingual grammar proposed by Bhatt and Bolonyai (2011) can be applied for the Hungarian-American bilingual community in North Carolina. For the discussion of the community specific ranking of Optimality Theory, only the socio-pragmatically meaningful instances of code-switching – which can be interpreted as serving a particular sociopragmatic function in light of the given context – were examined. Code-switched instances prompted by a lack of appropriate Hungarian competence as well as sociopragmatically not meaningful instances – borrowings, proper nouns – were excluded from the scope of my examination.

In the examples, plain type is used to indicate Hungarian or American English; bold italicized type is used to highlight code-switched instances. The original quotes are indicated with quotation marks, and the English translations are either inserted in brackets – if only some words need to be translated – or below the original quotes. The translations provided are my translations and I left the non-standard grammatical forms unchanged. With regard to transcription inventions, they were conducted with great detail since they are used for the qualitative analysis. However, as the transcriptions are semi-structured, informal dinner conversations, words were described as uttered by the subjects, e.g. as don’t, can’t, etc.

4. The interaction of constraints

According to the framework of the Optimality Theoretical analysis of bilingual grammar, the code-switching mechanisms of all bi- and multilingual speech communities can be described as the result of a conflict between a monolingual and a code-switched candidate that has to go through a hierarchical ranking of five universal socio-cognitive constraints. While the five constraints are supposed to be the same in any bi- and multilingual communities, the hierarchy according to which these constraints are ranked depends on an array of socio-cultural norms, the historical and structural context of the multilingual community in the macro social setting as well as on the communities’ collective speech practices (Bhatt and Bolonyai 2011: 5).

As the ranking of constraints varies in different bilingual settings, there have been attempts at setting up community-specific rankings. Bhatt and Bolonyai (2011) set up a proposed ranking applicable in the Hungarian-American bilingual immigrant community in North Carolina, which is as follows: {FAITH, PERSPECTIVE} >> SOLIDARITY >> {FACE, POWER}.

The aim of this study is to show how Bhatt and Bolonyai’s proposed ranking on the Hungarian-American bilingual immigrant community in North Carolina can be applied to describe the interaction of constraints governing the meaning-making mechanism of code-switching.

The interaction between the five sociopragmatic constraints is illustrated in tableaux. In these tableaux, the constraints that are violated by the competing – code-switched or monolingual – candidates are indicated with asterisks. The constraints are arranged in the order following the hierarchy proposed by Bolonyai with the highest ranked constraint placed in the left side of the tableaux and the lowest at the extreme right. The candidates undergo the array of the hierarchically arranged constraints, and if they violate one particular constraint, it is marked with an asterisk. Violating the highest ranked constraint is lethal, marked with exclamation points, which means that the surface realization of the violating candidate is disqualified. The actual output representation, the successful candidate, is indicated by a horizontal arrow.

In the section below, I provide examples to demonstrate how the five constraints interact with each other in the examined set of data. To illustrate the interaction between Perspective and Solidarity, example (1) has been analyzed:

(1) The interaction of PERSPECTIVE and SOLIDARITY 

A: ”Jaj, jöttem visitbe, egy családhoz, és amíg itt voltam, kaptam egy …”
A: `Yeah, I came to visit a family, and while I was here, I got one …`
(source: the author’s own data collected in 2008-2009)

In this utterance, the speaker remembers the first time she came to the USA. When recalling the circumstances, she switches to English to say that she came to visit somebody. The switch to `visit` contextualizes the original purpose of her coming to the USA. By switching to English in the middle of a Hungarian sentence, she accentuates the fact that she originally came to the USA visiting somebody and not with the purpose of immigrating. The contextualizing function of code-switching is listed among the functions of Perspective.

As Hungarian is the unmarked language of the interview, the language shared with the interviewers, by switching to English, the speaker moves away from that language of solidarity, violating the constraint of Solidarity. Apparently, the need for the switch to English as a contextualization cue seems to be stronger than complying with the constraint of Solidarity. Therefore, it shows that Perspective is a higher ranked constraint than Solidarity. Faith, Face and Power are not relevant in this utterance. The interaction of the constraints in this utterance is illustrated in Tableau 1.

Tableau 1: The interaction of PERSPECTIVE and SOLIDARITY (PERSPECTIVE >> SOLIDARITY)

It can be seen in Tableau 1 that there is an interaction between two linguistic inputs, two candidates. While the code-switched candidate (a) fulfills the function of Perspective (acting as a contextualization cue), the monolingual Hungarian candidate (b) fulfills the function of Solidarity. As only one surface representation is possible, the more optimal candidate will be the actual linguistic output. The tableau shows that the monolingual candidate fulfills the function of Solidarity as the unmarked language of the interview is Hungarian but violates the function of Perspective as it does not fulfill the function of contextualizing the story. The code-switched candidate, however, fulfills the function of Perspective, contextualizing the purpose of the speaker’s coming to the USA, but it violates the constraint of Solidarity, as it moves away from the unmarked language of the interview. As the actual output is the code-switched one, it can be inferred that Perspective is a higher ranked constraint than Solidarity. According to the Optimality Theory for the analysis of Bilingual Grammar, the violation of a higher ranked constraint is lethal, indicated by an asterisk, so Solidarity is marked with an asterisk in the tableau.

Example 2 illustrates the interaction between Solidarity, Power and Face.

(2) The interaction of SOLIDARITY and POWER, FACE
A: ”Szeretem a történelmet. Gyűjtöm a fiamnak az információt. Ő szereti a törénelmet.” (`I like history. I gather the information for my son. He is keen on history.`)
…
I1: ”Mi az értelme? Mért fontos, hogy tudjuk a történelmünket?” (`What’s the point in it? Why is it important to know our history?`)
A: ”Minden fontos.” (`Everything is important.`)
I1: ”Miért?” (`Why?`)
A: ”Mert akkor jobb a a …” (`Because then it is better the the …`)
B: ”I just can’t believe that you said that you are gathering that for Daniel. I am so offended.”
A: ”Everybody who is interested.”
B: ”I am so offended, so offended.”
A: ”Everybody who is interested. Well, you never showed a whole lot of interest.”
(source: the author’s own data collected in 2008-2009)

In this utterance speaker A is speaking about the importance of gathering all the historical records of his descendents for his son. He makes this statement in Hungarian. However, speaker B, his daughter, who is a second-generation Hungarian-American, makes an English comment on this. She, as his daughter, feels offended by her father’s remark that he collects all the family records for his son without mentioning his daughter. The daughter feels that she is excluded from this and gives voice to her disappointment in English. For her, English – although she understands and speaks some Hungarian – is the default language of communication. When her father reacts to her remark, he switches from Hungarian to English. He feels that his face as a good father is threatened by his daughter’s remark, so he tries to come up with an explanation defending his case by saying that the family records are for everybody interested. His daughter is not satisfied with this explanation and repeats how offended she is. The father wants to end this embarrassing argument going on in front of the two interviewers and reproaches his daughter for not showing too much of an interest in the family’s history. It is interesting that he makes the final statement in English, which is the language of solidarity with his daughter, and not in Hungarian, which is his stronger language, the father’s mother tongue. The switch to Hungarian would promptly enable the father to gain back his role of an authoritative father, topping the argument, as well as mitigating the threat against his face as a competent father. Therefore, the switch to Hungarian would optimally fulfill the function of Power, topping the argument, and that of Face, mitigating the face threat, which are the sub-functions of Power and Face respectively. However, he tops the argument in English, which fulfills the function of Solidarity as English is the `we-code` with his daughter. Therefore, expressing solidarity with his daughter in English is a stronger socio-pragmatic need for the father than expressing his authority in Hungarian.

Tableau 2: The interaction of SOLIDARITY and POWER, FACE (SOLIDARITY >> FACE, POWER)

It can be seen in Tableau 2 that there are two candidates competing for surface realization: the monolingual English candidate (a) and the switch to Hungarian (b). The monolingual candidate optimally fulfills the function of Solidarity, while the switch to Hungarian fulfills that of Power and Face. According to the Optimality Theory for Bilingual Grammar, the actual surface representation is the most successful candidate, the one that the most optimally fulfills the sociopragmatic function instantiated by the situation. Relying on this logical premise, the monolingual English candidate is the actual surface representation, so the function that it actually fulfills is a higher ranked constraint than the one that its competing Hungarian candidate fulfills (Face, Power). As the monolingual candidate fulfills the function of Solidarity, while the switch to Hungarian fulfills the function of Power, in this particular interaction, Solidarity outranks Power and Face. Solidarity is a higher ranked constraint, while Face and Power are not in conflict, so they are equally ranked. Perspective and Faith are not activated in this situation.
The example above shows the interaction between Faith and Solidarity.

(3) The interaction of FAITH and SOLIDARITY
A: ”… Van egy, ööö, Theonak van most egy új munkatársa, aki tíz évvel ezelőtt feltalált egy kis ketyerét, így fogom nevezni, mert nem tudom igazából, page keeper, ő page keepernek nevezi”
A: `… Now, Theo has a new colleague, who ten years ago invented a gadget, I will call it like this because I do not really know its name, page keeper, he calls it a page keeper.`
(source: the author’s own data collected in 2008-2009)

In (3) the speaker is talking about a gadget that an acquaintance of hers invented. She is not familiar with the Hungarian equivalent of this term, so first she refers to this thing as “some gadget” in Hungarian. She also explains that the reason why she calls it ketyere “this gadget” is because she does not know its name. However, as she feels that the Hungarian term ketyere is hardly specific, she switches to English to specify this invention.Giving the English name of this gadget expresses this specificity with the greatest economy. The switch to English, therefore, serves the function of filling in this particular semantic gap, and as such, it fulfills one sub-function of Faith. However, the switch to English violates the constraint of Solidarity. As the default language of the interview is Hungarian, and the shared mother tongue of the four participants is also Hungarian, the switch to English is a move away from the `we-code` of this situation. Nevertheless, as the greatest semantic specificity is ensured by the switch to English, and the main socio-pragmatic aim of the speaker is to inform the other participants of the situation about this gadget, she switches to English. First, she hesitates, she tries to give the Hungarian equivalent of this thing, in compliance with Solidarity, that is why she says “I am going to call it this gadget”, but then she resorts to the English switch as an option which expresses the thing with the greatest economy.

Tableau 3: The interaction of FAITH and SOLIDARITY (FAITH >> SOLIDARITY)

It can be seen in Tableau 3 that in this situation two candidates compete for surface realization: the switch to English (a) and the monolingual Hungarian one (b). The monolingual Hungarian candidate fulfills the constraint of Solidarity, as it is the default language of the interview, the `we-code` of the participants. The switch to English, however, fulfills the constraint of Faith, as it expresses meaning with the greatest specificity. As the actual surface representation is the switch to English, it is the more optimal choice for surface representation. Relying on the OT logical premise, no successful candidate, the surface representation, can violate a higher ranked constraint, so Faith must outrank Solidarity.

The last example, (4), serves as an illustration of the interaction between Faith and Perspective.

(4) The interaction of FAITH and PERSPECTIVE
A: ”elkezd egy beszélgetést, nekem mindig az az érzésem, hogy hi, how are you, hi, how are you, ez olyan először olyan nagyon furcsának gondoltam, hogy mit érdeklődik ez …”
A: `and they start a conversation, I always have the feeling that this hi, how are you, hi, how are you, first I found this so strange, why they would want to enquire … `
(source: the author’s own data collected in 2008-2009)

In this utterance, the speaker highlights one significant difference between American and Hungarian conversation practices. She cites one common and set linguistic expression hi, how are you as a typical example, as according to her, it well illustrates the different cultural connotations embedded in the American English and Hungarian languages. In Hungarian, `hi, how are you` usually expresses genuine interest, to which a detailed response is acceptable. For Americans, though, it is rather a way of greeting or starting a conversation than expressing real interest in how the other person is, and no detailed responses are expected. When illustrating the striking difference between American and Hungarian speech patterns, she switches to English to quote this characteristic example. By switching to English, she fulfills the function of Faith, as the switch to English the most optimally expresses the culture specific connotation of the American term. The Hungarian equivalent could not fulfill this function for its different underlying connotation. The switch to English fulfills another function as well, that of Perspective, as it is a quotation from Americans, expressing their voice. The switch to English, hence, fulfills the function of Faith and Perspective. However, as Hungarian is the default language of the interview, the `we-code` shared by the participants of the interview, it violates the constraint of Solidarity.

Tableau 4: The interaction of FAITH and PERSPECTIVE (and SOLIDARITY) (FAITH = PERSPECTIVE >> SOLIDARITY)

It can be seen in Tableau 4 that the two competing candidates are the switch to English (a) and the monolingual Hungarian one (b). The monolingual Hungarian fulfills the constraint of Solidarity but violates that of Perspective and Faith. The switch to English fulfills the constraints of Faith and Perspective but violates Solidarity. As the actual surface representation is the switch to English, it must be the more successful candidate fulfilling a higher ranked constraint. Consequently, Faith and Perspective outrank Solidarity, while they are not in contrast with each other, so they are equally ranked. Face and Power are not activated in this situation.

5. Conclusion

In this study, I have set out to show how the sociopragmatic constraints governing the sociopragmatic meaning-making function of code-switching interact in a model proposed by Bhatt and Bolonyai (2011). I have analyzed examples gathered in the Hungarian-American bilingual community in North Carolina and represented them in tableaux. Relying on the examples provided, the algorithmic ranking of the constraints can be set up as follows (5):

(5)
PERSPECTIVE >> SOLIDARITY (Tableau 1)
SOLIDARITY >> FACE = POWER (Tableau 2)
FAITH >> SOLIDARITY (Tableau 3)
FAITH = PERSPECTIVE >> SOLIDARITY (Tableau 4)

From the algorithmic representation above, it can be concluded that FAITH and PERSPECTIVE are ranked above SOLIDARITY, while SOLIDARITY ranks above FACE and POWER. As no evidence has been found for a conflict in the interaction of FAITH, and PERSPECTIVE, or in that of FACE and POWER, they are ranked equally.

Therefore, my findings reinforce Bhatt and Bolonyai’s (2011) ranking of constrains in the Hungarian-English community grammar, which is as follows:
{FAITH, PERSPECTIVE} >> {SOLIDARITY} >> {FACE, POWER}

Works cited

• Auer, Peter. 1995. The pragmatics of code-switching: A sequential approach. In: Milroy, Lesley and Pieter Muysken, eds. One speaker, two languages: Cross-disciplinary perspectives on code-switching. Cambridge: Cambridge University Press, 115-135.   
• Auer, Peter. 1998. Introduction: Bilingual conversation revisited. In: Peter Auer, ed. Code-switching in conversation. London and New York: Routledge,  1-24.
• Backus, Ad. 2001. The role of semantic specificity in insertional code-switching: Evidence from Dutch-Turkish. In: Rodolfo Jacobson, ed. Code-switching worldwide. Berlin: Mouton, 125-154.
• Bhatt, Rakesh M. and Ágnes Bolonyai. 2011. Code-switching and the optimal grammar of bilingual language use. Bilingualism: Language and Cognition, 14 (3): 1–25.
• Callahan, Laura. 2004. Spanish/English code-switching in a written corpus. Amsterdam: John Benjamins Publishing Company.   
• Canagarajah, A. Suresh. 1995. Functions of code-switching in ESL classrooms; socializing bilingualism in Jaffna. Journal of Multilingual and Multicultural Development, 16 (3): 173-195. 
• Esdahl, Trine. 2003. Language choice as a power resource in bilingual adolescents` conversations in the Danish Folkeskole. Journal of Multilingual and Multicultural Development, 24 (1-2): 76-89.
• Gal, Susan. 1979. Language shift: Social determinants of linguistic change in bilingual Austria. New York: Academic Press.
• Gardner-Chloros, Penelope and Katerina Finnis. 2003. Code-switching, gender and politeness: A study in the London Greek-Cypriot community, Estudios de Sociolinguistica, 4 (2): 505-532.
• Gumperz, John J. 1982. Discourse strategies. Cambridge: Cambridge University Press.
• Heller, Monica. 1988. Strategic ambiguity: code-switching in the management of conflict. In: Monica Heller, ed. Code-switching: Anthropological and sociolinguistic perspectives,. Berlin: Mouton de Gruyter, 77-96.
• Jorgensen, J. Normann. 1998. Children`s acquisition of code-switching for power yielding. In: Peter Auer, ed. Code-switching in conversation: Language interaction and identity. London, New York: Routledge, 237-258.
• Lin, Angel. 1990. Teaching in two tongues: Language alternation in foreign language lassrooms. Research Report, 3. City Polytechnic of Hong Kong.
• McClure, Erica and Malcolm McClure. 1988. Macro- and micro-sociolinguistic dimensions of code-switching. In: Monica Heller, ed. Code-switching: Anthropological and sociolinguistic perspectives. Berlin: Mouton de Gruyter, 25-51.
• Meeuwis, Michael and Jan Blommaert. 1998. A monolectal view of code-switching: Layered code-switching among Zairians in Belgium. In: Peter Auer, ed. Code-switching in conversation: Language interaction and identity. London, New York: Routledge, 76-98.
• Montes-Alcala, Cecilia. 2001. Written code-switching: Powerful bilingual images. In: Rodolfo Jacobson, ed. Code-switching worldwide II. Trends in Linguistics. Studies and Monographs 126. Berlin, New York: Mouton de Gruyter, 193-219.
• Myers-Scotton, Carol and Ágnes Bolonyai. 2001. Calculating speakers: Code-switching in a rational choice model, Language in Society, 30 (1): 1-28.
• Pandey, Anita. 1995. The pragmatics of code alternation in Nigerian English, Studies in the Linguistic Sciences. 25 (1): 75-117.    
• Prince, Alan and Paul Smolensky. 2004. Optimality Theory: Constraint interaction in Generative Grammar. Massachusetts, USA: Blackwell. 
• Valdes-Fallis, Guadalupe. 1976. Code-switching in bilingual Chicano Poetry. Hispania, 59:877-886.
• Wei, Li. 1994. Three generations, two languages, one family: Language choice and language shift in a Chinese community in Britain. Clevedon, England: Multilingual Matters.
• Wei, Li. 1998. The `why` and `how` questions in the analysis of conversational code-switching. In: Peter Auer, ed. Code-switching in conversation. London and New York: Routledge, 156-176
• Woolard, Katherine A. 1988. Code-switching and comedy in Catalonia. In: Monica Heller, ed. Code-switching: Anthropological and cociolinguistic perspectives. Berlin: Mouton de Gruyter, 53-76.
• Zentella,Ana C. 1997. Growing up bilingual: Puerto Rican children in New York. Malden, MA: Blackwell.

Notes

¹ For the limited scope of this paper, the list of these functions is by no means comprehensive. ↩

² I had the opportunity to conduct these interviews as a visiting Fulbright researcher. ↩

Copyright (c) 2011 Tímea Kovács

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.