Present Tense Auxiliary and Copula Choice in Banarsi Bhojpuri Susan B. Das University of Pennsylvania This paper studies the changing relationship in Banarsi Bhojpuri between two sets "be"-verbs. Banarsi Bhojpuri, spoken in and around the Indian city of Varanasi, possesses two complete sets of present indicative "be"-verbs. These sets are based on different roots: baa- and ha- (e.g., 1st person "baai" and "hai", 2nd fem. pl. "baau" and "hau", 3rd masc. pl. "baayan" and "hauwan"). In this paper, I examine evidence that the forms in baa- are declining in use, and I consider how the semantic distinctiveness of baa- versus ha- might have affected and been affected by this decline. The data for this study is drawn from sociolinguistic interviews with sixty-four members of the Hindi/Bhojpuri bilingual community of Varanasi, a city in North India about midway between Delhi and Calcutta. The interviews were conducted and transcribed with the help of Bhojpuri-speaking research assistants, and these transcriptions have been transferred to computer files. The main aim of this paper is to describe how the social variable of age correlates with variation in the verb root. However, there are at least three independent linguistic variables that may mask the effect of this social variable. Much of my paper will be devoted to an in depth examination of these linguistic variables, specifically, the auxiliary/copula's (1) semantic function, (2) person agreement (first, second, or third), and (3) honorificity (normal, honorific). Previous descriptions of various Bhojpuri dialects (Tiwari 1960, Simon 1986) note that the baa- and ha- rooted forms serve somewhat unique semantic functions. For example, the forms in baa- are favored in existential/locative constructions and as auxiliaries for the Present Continuous tense/aspect. The forms in ha- are preferred in copular constructions with predicate nouns. Such distinctions are not uncommon within the Indic language family (Masica 1991:337). By running computer searches of the transcriptions, I have located each present indicative "be"-verb in my data sample. The tokens are already coded for speaker and root used. I am presently coding them for their semantic function and agreement characteristics. Once the coding is completed, I will tabulate the number of tokens baa- and ha- for each of the different conditioning factors under study and examine the statistical significance of each. Preliminary results indicate that the semantic differentiation between baa- and ha- does indeed to hold true for some speakers in my sample. One elderly Brahman female used baa- rooted forms 46% of the time, when all semantic categories were taken together. As expected, the percentage was higher than this average for Present Progressive tense (73%) and Existential constructions (62%), and lower for Copular constructions with Predicate Nouns (11%). -------------------------------------------------- References: Masica, C. 1991. The Indo-Aryan Languages. New York: Cambridge University Press. Simon, B. 1986. Bilingualism and language maintenance in Benaras. U of Wisconsin-Madison PhD thesis. Tiwari, U. 1960. The origin and development of Bhojpuri. Calcutta: The Asiatic Society.