Organizer and Chair: Miwa Nishimura, The American University
Discussant: S. N. Sridhar, State University of New York
In many bilingual communities, alternate use of two languages often arises in a single situation, code switching. The alternation is often rapid and smooth, taking place even within a sentence. Socio oriented linguists have considered the functional aspects of code switching in a given community. Theoretical linguists have attempted to formulate the 'universal' or 'near universal' constraints on intrasentential code switching using various frameworks. This panel examines code switching between English and four Asian languages-Hindi, Korean, Japanese, and Marathi-based on data collected in four communities in North America. Code switching between English and these Asian languages is of theoretical interest, because English and these languages are typologically different. For sociolinguists, it is interesting to find out how these Asians living in North America, whether they are American born or recent arrivals, use their 'old' language and English in their family and community life.
Yoon studies Korean/English code switching that occurs in intergenerational communication between the Korean born first generation and the American born second generation. Yoon first identifies the three patterns of Korean/English code switching using the concepts developed in speech accommodation theory (convergence, divergence, and compliance). Yoon then demonstrates how politeness, which is important in Korean language use, shapes these patterns of code switching observed in the intergenerational communication.
Bhatt takes a formal approach to intrasentential code switching within the framework of the government binding theory. Challenging the constraint posited by Belazi et al. (1994), which disallows switching between the functional head (Complementizer, Inflection, Modal, & Determiner) and its complement (sentence, subject NP, VP, & NP), Bhatt argues that Case theory accounts for Marathi/English and Hindi/Marathi code switching.
Nishimura, based on Japanese/English data collected among the niseis (second generation) in Toronto and San Francisco, proves that a common word order is not a requirement for code switching to take place, as some researchers claim; borrowing (of nouns), therefore, is not an alternative to code switching in language contact involving two typologically different languages. Nishimura, analyzing the data within the framework of 'who speaks what language(s) to whom,' demonstrates that the niseis do intrasentential code switching when they use both Japanese and English talking to a kika Nisei (Japan educated, Japanese dominant nisei) or to a mixed group of native Japanese and niseis in their in group interactions.
Keumsil Kim Yoon, William Paterson College
This paper presents an analysis of important factors that influence codeswitching strategies for intergenerational bilingual communication. It represents an attempt to integrate two areas of research, namely, speech accommodation theories and linguistic politeness theories
The paper explores codeswitching patterns of the first generation (Gl) Korean-English bilinguals and the second generation (G2) Korean English bilinguals. In conversations between Gl and G2 bilinguals, each of whom is a second language speaker of the other's first/predominant language, three different local patterns of codeswitching emerge: convergence (each speaker attempts to speak the interlocutor's native language), divergence (speakers maintain their own language or diverge linguistically from their interlocutors), and compliance (each speaker agrees with and complies with the interlocutor's code choice). These patterns interact with one another in complex ways to influence codeswitching strategies.
To explain the identified patterns, the paper investigates the area of linguistic politeness and examines linguistic characteristics of the language pair in contact; Korean is a honorific language in which the choice of formal linguistic forms is obligatory in the light of social convention whereas English is a non honorific language in which the choice is neutral, which implies egalitarian verbal interaction. The paper raises a question as to whether the pattern of convergence can be interpreted as a factor related to speakers' assessments of relative proficiency of the interlocutor's second language or a factor related to 'discernment' (almost automatic observation of socially agreed upon rules) and 'volition' (considerations of cost and benefit as well as face that the speaker displays) aspects of linguistic politeness.
Finally, the paper discusses codeswitching within the notion of 'politic verbal behavior' (socio culturally determined behavior directed towards the goal of establishing and/or maintaining in a state of equilibrium), and proposes a model for the study of codeswitching strategies that would account for the complexity of communication in Korean English bilingual settings.
Rakesh M. Bhatt, University of Tennessee
It is perhaps common knowledge now that intra sentential code switching (ICS) is not random, but rather subject to regular linguistic constraints. What is not clearly understood is the nature of the properties of the linguistic constraints. What remains largely contentious however, is whether the proposed constraints are universal, or language specific (cf., Poplack 1980, Sankoff and Poplack 1981, Woolford 1983, Joshi 1985, DiSciullo, Muysken, and Singh 1986, among others). Relatively recently, Bokamba (1989) and Myers Scotton (1993) have argued that none of the proposed constraints are universal. In the latest volume of Linguistic Inquiry however, Belazi, Rubin and Toribio (1994) propose a new "Functional Head Constraint" (FHC) which disallows ICS between functional heads and their complements. They claim that any functional head must obey FHC. In this paper I will provide evidence that FHC makes wrong predictions for Hindi English and Marathi English ICS. I will further argue that ICS is constrained by the same set of linguistic principles that are provided by Universal Grammar.
The net effect of FHC, abstracting away from the formal instantiations of the constraint, is that switches such as between a complementizer (=C0) and a sentence (=IP), or between determiner (=D0) and a noun phrase (=NP) are prohibited. Such predictions are falsified in English Hindi ICS between D0 and NP (1). Further, the grammaticality contrast between (2) and (3) in Marathi English ICS between a quantifier (Q0) and NP remains unexplained in the theory of Belazi et al.
(1) Of all the places John has hidden kuch (some) books bathroom meN (in).
(2) kaahi (some) chairs.
(3) *some khurcyaa (chairs)
The data in (4) is evidence from Hindi English ICS between C0 and IP, contra FHC.
(4) I told him (that [Ram ko ghar jaanaa paRega] )
Ram-dat home go obliged
In addition to the empirical problems with Belazi et al.'s proposal, I will also point out some conceptual problems with their theory, especially their account of Case theory to explain the problematic Marathi English data (5).
(5) * some chairs-war (On some chairs)
Finally, I will propose, as an alternative, that mixing (ICS) is possible only at (Case or Theta) licensed sites in the phrase structure. In the remainder of the paper, I will discuss the whole array of the predictions of my proposal, in addition to explaining why (5) is ungrammatical.
Miwa Nishimura, The American University
In many bilingual communities, two languages are often used in a single situation, even within a sentence. There have been many studies of syntactic aspects of intrasentential code switching (ICS) (e.g., Belazi et al. 1994; di Sciull et al. 1986; Joshi 1985; Nishimura 1985; 1990; Poplack 1980; Sridhar & Sridhar 1980; Woolford 1983; Yoon 1992 among many others). There is a theory which claims that a common word order is necessary for ICS to take place as in Spanish and English, predicting that ICS does not occur in language pairs such as Japanese/English and others as presented in this panel (Poplack et al. 1989; Sankoff et al. 1985; Sankoff et al. 1990). This theory, furthermore, claims that borrowing (of nouns) is an alternative to CS in the typologically different language pairs; the data base for this claim is the Finnish speech into which English nouns are incorporated (Poplack et al. 1989).
This paper proves that ICS does take place in Japanese/English data, and that borrowing (of nouns) is not an alternative to ICS. This paper demonstrates that one can locate both ICS and borrowing in the Japanese/English data collected among the Niseis(second generation Japanese) in Toronto and San Francisco. Borrowing (of nouns) takes place when the bilingual Niseis use Japanese when talking to native Japanese. ICS, switching of various other syntactic categories such as noun phrases (NP), prepositional phrases (PP), and adverbs, etc., takes place when the same Niseis use both Japanese and English actively in discourse; that is when the Niseis talk to a group of native Japanese and fellow Niseis or when they talk to the kika Niseis (Japan educated Niseis with native Japanese proficiency). Also, ICS in the Japanese/English data takes place in sentence environment of both Japanese and English.
From our finding, we can conclude that the occurrence of ICS is not typologically conditioned, but it is conditioned by the type of bilingual speech. When bilingual speakers use both languages(without choosing one specific base language), ICS takes place.
Would you like to return to the Interarea, Library, & Teaching Table of Contents? Choose another area?