Abstract ID: 917
Part of General Paper Session (Other abstracts in this session)
Authors: Chatterjee, Tridha
Submitted by: Chatterjee, Tridha (University of Michigan, United States of America)
New codeswitching (CS) data from Bengali-English bilinguals show bilingual compound verb (BCV) constructions that are somewhat different from the structure of BCVs that have been reported so far. BCVs occur widely in CS between various language pairs. These constructions are characterized by two verbs occurring contiguously, the first verb being an embedded language lexical verb and the second verb being a matrix language helping verb such as ‘do’ or ‘make’ (Muysken 2000, Backus 1996, Edwards and Gardner-Chloros 2007, Romaine 1986). Until now this structure in Bengali‐English CS has not been documented.
This paper presents Bengali-English CS data (collected during my fieldwork in India in 2010) and explains how the BCVs that occur in this data are different from the BCVs that have been reported so far. For example, educate kora ‘to do educate’ is a regular BCV in Bengali-English where educate is in English, and kora, a helping verb is in Bengali. But, some BCVs in Bengali-English bilingual speech have three verbs occurring adjacent to each other, where each verb performs a specific function. An example of a three-verb construction is given below.
1) Ora park-ta renovate kore dieche
3pl park-DEF renovate do-PRF P give.PRF.3P
They have renovated the park (for someone).
Here, the English verb renovate, gives meaning, kore expresses the action of doing renovating, and the third verb dieche, provides aspectual information of perfectivity and semantic information of doing the action for someone. This construction, however, still expresses one single action. Also through syntactic tests on the three-verb compounds, I show that the three verbs together form one single syntactic unit.
This paper specifically focuses on the third verb in these constructions and examines the function that these verbs have in these constructions. The verbs that occur in the third position (also called vector verbs) of the bilingual compound, come from a list of 12 to 16 verbs, and have different degrees of grammaticalization and semantic bleaching (Basu 2010, Paul 2003) and the degree to which the third verb contributes meaning to the compound depends on how grammaticalized the verb is.
I also explain how the three-verb constructions draw from two monolingual constructions, the Bengali compound verbs (V1 + V2) and Bengali conjunct verbs (noun + helping verb + V2). Bengali compound verbs express one action with the help of two verbs, occurring contiguously. The first verb gives the semantic content while the second verb often loses its own meaning and gives aspectual information to the compound (Thompson 2010). The example below illustrates this construction.
2) ami or naamta bhule gechi
1sg 3sg-GEN name-DEF forget-PRF P go.PRF.1P
I have forgotten his/her name.
Overall this paper shows that the third verb is an important component of the compound, whenever it occurs in a compound, and that it performs a range of functions from giving aspectual information to semantic information to being highly grammaticalized.