Linguistic accommodation and code switching during Luo/Luhyia and Luo/Gusii contact
Kebeya, Hilda Udali
MetadataShow full item record
This study set out to investigate and then compare the linguistic behaviour of bilingual speakers in intergroup contexts in two rural speech communities in Kenya. In order to achieve this, recordings were obtained from Luo/Luyia bilinguals in Kiboswa, and Luo/Gusii bilinguals in Suneka. The languages under study differ from each other in significant ways. Luo, for example, is classified as a Nilotic language (see.Stafford 1967), while Luyia and Gush are categorized as Bantu languages (Guthrie 1967). The major questions that motivated the research were: What linguistic strategies do bilingual speakers in linguistic border points use in intergroup communication? What Sociolinguistic patterns do these speakers exhibit and what are the stylistic functions of these patterns? The study found out that speakers in Kiboswa and Suneka use three linguistic strategies in intergroup interaction: convergence, divergence and code switching. These three strategies appeared to be influenced by situational factors (e.g. the household type), extra-linguistic factors (e.g. sex and age), social-psychological factors (e.g. desire for social approval and integration or asserting of one's ethnic group) and socao-cultural factors (e.g. jamwas being used in ritualistic practises). When the linguistic behaviour of the households was compared, certain patterns emerged. Generally, the households, appeared to exhibit very high scores for the own-group code and very low scores for the out-group code. Thus when intergroup communication took place in the home domain there was a general tendency for speakers to converge towards the ethnic language of the head of the household where the recording was done. This observation appears to be in line with what Giles & Robinson (1990) propose that during intergroup contexts speakers tend to converge to their listeners rather than diverge from them. However when convergence, divergence and code switching were correlated with the social variables of ethnicity, sex and age, although it appeared that all the social groups exhibited higher degrees of divergence than convergence; there was a general tendency for male groups, in the two study areas, to diverge more than female ones, and older speakers appeared to diverge more than the younger ones. No systematic patterns of Accommodation were observed in the ethnic groups. As regards Luo/Luyia and Luo/Gusii code switching all the respondents, irrespective of their social group and household type, appeared to display very low scores. This indicates that code switching is rarely used in exchanges involving ethnic and non-standardised languages. There were socio-psychological factors that motivated speakers in Kiboswa and Suneka to converge/diverge/code switch. Convergence, for example, was found to take place when speakers desired certain rewards e.g. receiving social approval and integratio+n from their listeners; making themselves socially attractive; increasing the efficacy of communication; making sales; inducing addressees to do run errands for them; minimising the impact of offensive acts; impressing their listeners and specifying the addressee to whom a message was intended. Speakers would weigh these rewards against the costs they were likely to incur in converging (namely; expended linguistic effort and a temporary loss of one's ethnicity). If rewards for converging outweighed the costs, speakers would converge but if the costs outweighed the rewards, speakers would diverge. For instance, speakers in this study diverged in order to: assert their ethnicity, subtly show disapproval of the people they were talking to, make slanderous statements about out-group members and show disapproval of the people they were talking about. These findings indicate that the divergence in Kiboswa and Suneka is as a result of social identity processes while convergence is due to similarity-attraction and social exchange processes (cf. Giles & Robinson 1990). As regards Luo/Luyia and Luo/Gusii code switching a number of factors appeared to be responsible. To determine these factors, the study categorised all the code switched material into three classes: sequential unmarked code switching, unmarked code switching and exploratory code switching (cf. Myers-Scotton 1993a). It became apparent that each type of switching was motivated by different factors. Sequential unmarked code switching, for example, was found to occur when there had been a change in one or more of the situational factors during a conversation e.g. when a new participant joined in the conversation or when the topic changed. Unmarked switching, on the other hand, was found to serve an emblematic role in the two communities studied and could occur even when there was no change in situational factors. Whenever speakers seemed to be unsure of the code to use on their respondents (e.g. when speakers were meeting their addressee's for the first time and the norms governing the interaction were not clear) they would employ exploratory switching. Thus each of these three categories of code switching was motivated by different factors. Inter-sentential and intea-sentential switches were observed in Kiboswa and Suneka. Our study found that inter-sentential switching could take place between L codes (i.e. ethnic languages) whereas intea-sentential switching could not. Thus, all instances of intea-sentential switching in our data have an H code (i.e. English and/or Swahili) serving as either the matrix or embedded language.