Codeswitching and the Egyptian Arabic construct state

Abstract

In this paper, I assume a grammatical approach to codeswitching (MacSwan, 2012), which predicts a ban on codeswitching below the head level. Previous literature has analyzed this ban largely at the word-level, terming it a ban on word-internal codeswitching. In this paper I argue that the said ban can also be extended from the lexical domain to certain syntactic domains that act as one word. I test MacSwan’s theory in the context of codeswitching within construct state nominals, a genitive construction prevalent in Semitic languages. The construct state is particularly relevant for discussions on the syntax-phonology interface within codeswitching because it is a complex and productive syntactic structure that is mapped onto one word. The results of an acceptability judgment experiment show lowered acceptability for sentences where a codeswitch occurred within a construct state nominal vs. sentences where the codeswitch occurred outside of a construct state nominal. The lowered acceptability for such codeswitched judgments suggests that the ban on word-internal codeswitching is not in fact limited to words but can be extended to complex syntactic units that prosodically function as one word.