Linguistic Adaptation to Unacceptable Sentences

Abstract:

Speakers display considerable variability in language use and representations: they may have different pronunciations of the same word, different intended meanings for the same phrases, and different sets of syntactic constraints in their internalized grammars. Comprehenders adapt to such variability by constantly updating their expectations for variants, a process termed linguistic adaptation. Linguistic adaptation has been demonstrated at different levels of linguistic representation. In this dissertation, I propose an account of the widely attested yet poorly understood phenomenon ofsyntactic satiation as linguistic adaptation to unacceptable sentences. Syntactic satiation refers to the phenomenon whereby comprehenders find initially unacceptable sentence types increasingly acceptable after repeated exposure. Using island-violating sentences as a test bed (e.g., *What did John think a bottle of __ fell on the floor?), I show that satiation demonstrates speaker-specificity, a signature property of linguistic adaptation. I further address a number of issues raised by the initial findings, including pin-pointing the representational target of adaptation during satiation, and explaining the systematically varying rates of satiation across different sentence types. The findings of this dissertation are valuable in multiple ways. First, they expand our understanding of whether and how language users engage in linguistic adaptation when they face degraded inputs, a rarely studied domain in the literature on adaptation. Furthermore, by uncovering the underlying mechanism for satiation, we can relate satiation to other psycholinguistic phenomena with shared mechanisms, and evaluate the validity of past claims made on linguistic theories that drew evidence from satiation. Finally, this study constitutes a step towards a full account of what sentence acceptability judgments represent, a crucial methodological concern for linguists.