THE HIDDEN MEANING OF CODESWITCHES IN SPANISH ENGLISH CONVERSATIONS

El análisis de la conversación en dos lenguas desde los puntos de vista de Myers-Scotton, Auer y otros investigadores se usa como base para analizar la conversación en español e inglés en una comunidad hispana de Georgia, EEUU. Indicamos cuáles son las fuerzas y las debilidades en el análisis de estos investigadores para mejorar nuestro análisis, con implicaciones para futuros estudios.


INTRODUCCIÓN
Bilingual conversations can be analyzed from various perspectives. This study takes perspectives from Myers-Scotton's Markedness model (1993b) of codeswitching (CS) and Auer's (1995) conversation analysis as the bases from which to analyze Spanish and English bilingual conversation data.
The conversation excerpts in this study are mostly from conversations in a Hispanic community in northeast Georgia. The speakers in the community range from mostly monolingual Spanish to mostly monolingual English. Almost all the speakers are to some degree Spanish/English bilinguals. Gumperz' (1982;Blom and Gumperz 1972) 'interactional/interpretative' model and Hymes' (1962;1967;1972a,b) 'ethnography of speaking' led to the study of CS via 'Conversation Analysis' (Myers-Scotton 1993b: 55-59). Myers-Scotton's Markedness model is based on which language is 'marked' or 'unmarked', 'marked' denoting special meaning and 'unmarked' denoting the given or default whenever the 'marked' is not used for a given conversation. Auer's model of conversation analysis focuses on the alternation of languages in a bilingual conversation. The following study shows how all three of these models' perspectives come together to examine the same thing, the structure of a bilingual conversation. These perspectives are studies of the pragmatics of bilingual speech.

THE MARKEDNESS MODEL OF CODESWITCHING
Based on the 'Co-operative Principle,' Myers- Scotton's (1993b: 113) Markedness model of conversation analysis emanates from her 'negotiation principle': Choose the form of your conversation contribution such that it indexes the set of rights and obligations which you wish to be in force between speaker and addressee for the current exchange.
The language of "unmarked choice" is associated with the social "norms" sometimes called a "rights-and-obligations" set. Choosing the "marked" language constitutes a speaker's "negotiation" for another rights-and-obligations set different from the social norm or the status-quo. Myers-Scotton lists four types of CS under her Markedness model: (1) "CS as a sequence of unmarked choices," in which each language used is the unmarked choice due to changing social or other conversational context requirements; (2) "CS itself as the unmarked choice," which may occur, for example, between bilingual interlocutors who are peers; (3) "CS as a marked choice," in which an interlocutor switches codes in order to not comply with an expected social or contextual norm; and (4) "CS as an exploratory choice," or CS to 'explore' or 'negotiate' the unmarked choice between interlocutors when the choice of code is not clearly apparent, given the situation (Myers-Scotton 1993b: 7, 84, 113-114, 119, 131, 142-143;1993a: 478-480, 492-493 (1) "CS as a sequence of unmarked choices," in which each language used is the unmarked choice due to changing social or other conversational context requirements (1993b: 114). The first part of the conversation is exclusively in English due to the presence of the English monolingual speaker, AP. The latter half of the conversation is mostly in Spanish because Spanish is the dominant language of JM, who is more comfortable conversing in Spanish with DS because he knows that DS is fluent in Spanish. Therefore, English is unmarked for the first half of the conversation and Spanish is unmarked for the second half of the conversation.
In this conversation in example [3], in which there is no need to use English, English usage 'marks' items ('record,' and 'insurance') that are closely connected to the surrounding English-speaking culture. These items could have been said in Spanish and in this conversation 'insurance' is repeated as 'seguro.' The expressions 'You don't know?' and 'I don't know' also serve to mark their respective speakers as participants in the larger English-speaking community, because the norm would dictate that they be said in Spanish, given that the vast majority of the conversation is in Spanish. We cannot say exactly what motivated these English expressions in an otherwise overwhelmingly 'unmarked' Spanish conversation, but the choice of English over Spanish is conspicuous and can hardly be unnoticed or deemed meaningless. In other words, exactly why they Normas (ISSN: 2174-7245) | https://ojs.uv.es/index.php/normas/index Diciembre 2020 | Volumen 10 | Número 1| Pág.87 are 'marked' by English is unknown, but they are, nonetheless, 'marked' as a consequence of these speakers' contact with English in the wider Anglo culture in which they participate. The speech 'marked' by English in this conversation may 'mark' or signal to interlocutors that the speakers recognize the importance the English of the larger community around them has in their lives, as if to indicate to each other that they are 'in tune with' the world and not relegated to their own little corner.
Example [4] is representative of (4) "CS as an exploratory choice" or CS to 'explore' or 'negotiate' the unmarked choice between interlocutors when the choice of code is not clearly apparent, given the situation (Myers-Scotton 1993b: 114, 142-143). AR and DS are both proficient in both Spanish and English and can communicate with ease in either of the two languages. This conversation represents indecision between the two speakers as to which language to use as the principle means of communication, that is, the ML or 'unmarked' language. It is almost a 'duel' of sorts, because AR insists on continuing the conversation in English while DS insists on keeping his turns in Spanish before finally relinquishing to AR's use of English by the end of the conversation when all turns by both speakers are in English. English dominated in this case probably because it was easier for DS to submit to AR's English usage since the school assignment discussed was taught and written in English. The conversation began, however, as a negotiation between Spanish and English use because DS usually talked to Hispanics in the community in Spanish and it was also the expected norm for him in this conversation.
AR=37, Hispanic adult female; DS=0, Anglo adult male; both AR and DS are bilingual; DS is helping AR with a college assignment which is taught in English and for which the written materials are all in English; AR is Ecuadorian and is a friend of DS's brother and sister-in-law and an acquaintance of DS at a Hispanic church. Other CS researchers such as Auer (1998: 8-13) and Meeuwis and Blommaert (1998: 77-80) have found Myers-Scotton's Markedness model too simplistic because of its reliance on identifying languages with fixed sets of social information.

3
AUER'S ANALYSIS OF CONVERSATIONAL CODESWITCHING Romaine (1995) and Auer (1988;) use Goffman's (19741981) concept of 'footing' in a Conversation Analysis of CS. Romaine (1995: 166, 172) describes "footing" as a socio-pragmatic "alignment" between interlocutors, for example to convey the relationships of "identity," "power," and "transaction" (Scotton and Ury 1977;McConvell 1988). Romaine explains that a realignment between interlocutors or a "change in footing," recasts a speech event in a different socio-pragmatic "frame." This change can be accomplished, for example, by monolingual style changes or bilingual CS. An example is the use of Spanish between Hispanics to indicate shared identity and a switch to English (where English is the language of power and prestige) for business transactions. Auer (1998: 17) indicates that a code-switch "introduces a new footing" as at least one of its functions. Along this same line of thinking, Koike (1987: 153) notes in bilingual Chicano narratives that code switches pattern "along ... idea units or breaks in the discourse." Thakerar, Giles, and Chesire (1982); Bourhis, Giles, Leyens, and Tajfel (1979); and Genesee andBourhis (1982, 1988) have treated CS to varying extents within 'speech accommodation theory.' They use the terms 'accommodation' or 'convergence' 6 to describe changes or switches in a speaker's language to accommodate to the interlocutor's language and 'divergence' to change or switch away from the other's language in order to distance oneself from the interlocutor. Auer (1995) focuses on the sequential arrangement of alternating turns in conversation and the necessity of considering the meanings assigned to CS according to conversational context (p. 132). In his "theory of code-alternation," he notes patterns of code convergence 7 and divergence associated with changes, for example, in conversation topic or interlocutors. His patterns also include those that serve to negotiate language choice between interlocutors and include insertions of one language into the other within turns. He notes that a 'base' or 'unmarked' language may be determined in a given conversation but that interlocutors sometimes keep open the choice of which language to use. In such cases, determination of which language is the 'base language' is not possible (pp. 124-126). This problem is also present in Myers-Scotton's Markedness models.
We must also take into consideration the problems such discourse analysts as Auer (1995Auer ( , 1998 and Li Wei (1998) find in Myers-Scotton's Markedness model in that she attaches socio-pragmatic information to the marked/unmarked character of each language. For example, in detailed conversation analysis, Auer (1995: 119-120) sees that a switch itself, regardless of to or from which language, is significant apart from the https://ojs.uv.es/index.php/normas/index Diciembre 2020 | Volumen 10 | Número 1| Pág.89 socio-pragmatic information that may be attached to a particular language. Example [5] is an instance from our data of switching from Spanish to English and from English to Spanish.
It is difficult to assign significance to the language used in line 4, in which the turn begins in English, then switches to Spanish, then back to English. 'He's gone now' serves to reiterate 'Ya se fue,' and the reiteration is emphasized or marked because it is in the language other than that of the original statement 'Ya se fue.' It is also difficult to surmise the function of the first switch in the same turn, from 'He's doin' that thing' to 'Ya se fue.' The fact that a switch has occurred may indicate, as Auer (1995: 119-120) claims, more than to or from which language the switch was made. In our data, however, Spanish is the unmarked language for the majority of the conversations. English, therefore, is inserted more as a marked language than is Spanish. In example [5] English is used to mark an exclamation in line 1 ('here he comes!') and to mark a reiteration in line 4. Thus, English in these instances is used to mark or emphasize. Therefore, along with Myers-Scotton (1993b), we claim that not only the switch itself but also the language of the inserted or switched elements is of significance.
Analysis of CS within the context of conversations is a necessary part of a complete understanding of CS where it occursin sentences, conversations, and the community as a whole. In a given bilingual community, one language may be perceived as being the "base language" (Auer 1995: 124-126) in one conversation, and in another conversation in the same community the other language appears to be base language.
In some conversations, neither language may be perceived to be the base language due to approximately equal quantities of utterances in each language. Therefore, our concept of what is the base language and what is the social meaning behind use of the other language should be context sensitive. In other words, a sentence may have a base of one language in a conversation consisting mostly of the other language. In turn, this conversational base language may not be the main language of the community as a whole. From this perspective we may seek the goals of conversation analysts at the level of the conversation within frameworks such as speech accommodation theory or Auer's sequence/alternation approach, the results of which may show more general patterns when several conversations from one community are studied together.

SUMMARY AND CONCLUSION
Bilingual conversations bring the possibility of conveying underlying pragmatic messages not conveyed in just the semantics. The switching from one language to another presents challenges to the analyst as to what these messages actually are. We have illustrated with Spanish English bilingual conversation data from northeast Georgia, U.S.A. the marked/unmarked dichotomy and how it is presented in Myers-Scotton's Markedness Model. These include (1) "CS as a sequence of unmarked choices," in which each language used is the unmarked choice due to changing social or other conversational context requirements; (2) "CS itself as the unmarked choice," which may occur, for example, between bilingual interlocutors who are peers; (3) "CS as a marked choice," in which an interlocutor switches codes in order to not comply with an expected social or contextual norm; and (4) "CS as an exploratory choice," or CS to 'explore' or 'negotiate' the unmarked choice between interlocutors when the choice of code is not clearly apparent, given the situation. This analysis depends on attaching social or contextual information to the choices of each language used. The language used is unmarked if it is the norm given the situation or unmarked if it goes against the norm. "CS as an exploratory choice" is when the norm is being negotiated by the speakers; the conversation involves an attempt to establish a norm. We have also illustrated from the same data set Auer's emphasis on the switch itself between languages being sometimes more important than assigning any particular pragmatic significance to the languages themselves. Studies of the pragmatics of bilingual conversations are best analyzed with sensitivity to at least the approaches of both Myers-Scotton and Auer.
Further study of speaker intuitions of what they think the messages are behind their CS along with attention to current and future socio-political associations made to bilingual speech will serve to further refine understanding of the pragmatics of CS and specifically that of Spanish and English in the U.S.