Syntactic representation and processing of transitives in bilinguals: evidence from structural priming in Mandarin-English speakers

Abstract

This thesis discusses central focuses concerning language processing. Specifically, it discovers questions about how people represent and process transitive sentences (e.g., an active sentence, such as the dog bit the man and passive sentence, such as the rat was chased by the cat), by exploring both production and comprehension in structural priming paradigms (a fact that people tend to replicate the linguistic patterns they have recently encountered and an approach to the nature of linguistic representation), and investigating through the differences between native and non-native language. This thesis seeks to unpack the layers of how transitive sentences are syntactically represented. Central to this exploration is a deep dive into the interplay between syntax and the lexicon, emphasising the relationship of syntactic information and open- and closed-class elements, and how these syntactic and semantic representations interact to arrive at an interpretation during comprehension (i.e., syntactic representation is independent or in conjunction with semantic information).

In Chapter 2, this thesis explores the shared representation of voice among Mandarin-English bilinguals. Numerous studies suggest that bilinguals process a shared syntactic representation. However, the extent to which they integrate their two languages remains a subject of investigation. To address this question, Chapter 2 presents a series of structural priming experiments to examine whether bilingual speakers have shared representation in passive voice, and the relationship between passive voice and lexical items, explicitly focusing on Mandarin-English speakers. Mandarin-English participants first listened to a sentence (the prime sentence) that could be either active or passive. Then, they described a picture depicting a transitive event. The cross-linguistic findings suggest that Mandarin-English bilinguals maintain a shared representation of passive voice independent of lexical repetition despite differences in word order. Interestingly, the varying priming effects between priming from two different types of passive voice with different function words indicate an association between function words and specific languages.

In Chapter 3, this thesis focuses on how people decide on a final interpretation of an implausible sentence. People sometimes misinterpret (nonliterally interpret) such sentences.

They might either syntactically reanalyse or semantically interpret this sentence for the final misinterpretation. For example, people may adjust the original syntactic representation, or semantically reinterpret the sentence by directly swapping the thematic roles. The evidence for how people syntactically analyse an implausible sentence is mixed. Three structural priming experiments were conducted in transitive structures to investigate the processing of a nonliteral interpretation for an implausible transitive sentence. Chapter 3 reported results that participants were more likely to nonliterally interpret a sentence after an implausible prime than a plausible prime, which suggests plausibility information affects comprehension. Importantly, the finding is different from syntactic reanalysis. Equivalent priming after implausible and plausible sentences suggests that people are more likely to maintain their initial syntactic representation rather than to change it. In conclusion, interpreting implausible sentences relies on syntactic analysis, but people may sometimes syntactically reanalyse the initial interpretation.

Furthermore, some researchers suggest that non-native speakers rely more on non-syntactic information than syntactic information. A structural priming experiment was carried out similarly to Chapter 3 to investigate the processing of an English implausible transitive structure in Mandarin-English speakers. Chapter 4 shows that participants were more likely to nonliterally interpret a sentence after an implausible prime than a plausible prime, which suggests plausibility information affects comprehension in non-native speakers. The effect of plausibility on nonliteral interpretations is stronger in passive sentences compared to active sentences, suggesting the interaction between syntactic and semantic representation. Equivalent priming between implausible and plausible sentences suggests that non-natives tend to preserve an initial syntactic representation of an implausible passive sentence rather than overwrite its syntactic representation. The comparison between non-native and native speakers shows no differences in syntactical processing.

Together, these results update a shared syntactic representation model in bilinguals and refresh an account for nonliteral interpretation processing in both native and non-native speakers.