V0 and SOV/SVO word-order in intra-sentential code-switching

Abstract

The present study is an attempt to evaluate empirical adequacy of MacSwan’s (2000, 2005, 2009, 2010) proposal that feature strength of V⁰ determines SVO/SOV order of constituents in intra-sentential code-switching (CS). In order to determine the potential of this proposal in predicting switching patterns, it employs a corpus of naturally-occurring Urdu/English CS. The corpus consists of 1767 sentences in the form of 29 recorded interactions among 42 competent Urdu/English bilinguals in natural settings. The examination of the data under consideration reveals that SVO/SOV order of constituents in Urdu/English CS is not determined by feature strength of V⁰. The data provide multiple instances of complement DP placed at pre-head position, resulting in OV order, in VP in spite of having an English V⁰s as its head which requires OV order due to its weak feature. MacSwan’s claim that V⁰ and T⁰ must be supplied by a single lexicon because of restriction on word-internal switching is also falsified by the empirical evidence from the data. The data provide numerous mixed sentences in which V⁰ and T⁰ are supplied by two different languages instead of one as claimed by MacSwan; hence, they must have same feature strength. Thus, SV/VS orders also depend upon feature strength of Urdu or English V⁰ in MacSwan’ terms. However, the data examined demonstrate that SV order remains fixed no matter it is English T⁰ sharing weak feature or Urdu T⁰ sharing strong feature of its respective V⁰. Thus, MacSwan’s proposal completely fails in predicting linear order of constituents when the languages involved have conflicting grammatical requirements.