Dissertations, Theses, and Capstone Projects

Date of Degree

2-2023

Document Type

Dissertation

Degree Name

Ph.D.

Program

Speech-Language-Hearing Sciences

Advisor

Valerie L. Shafer

Committee Members

Loraine K. Obler

Nancy Eng

Subject Categories

First and Second Language Acquisition | Psycholinguistics and Neurolinguistics

Keywords

agreement, sentence-processing, accent, cross-linguistic influence, bilingualism, acceptability judgments

Abstract

Speech produced with an unfamiliar accent is quite commonly encountered, especially in higher education settings, in which a large number of international students are found. As a result, it is inevitable for students and teachers alike to be exposed to different varieties of English across multiple settings. The current study examines the effects of such accents on grammatical processing in bilinguals.

English requires grammatical agreement between the subject and the verb (e.g., the 3rd person singular “s” on the verb in The dog jumps); that is, subject-verb (S-V) agreement is obligatory. S-V agreement must hold even in cases where there are intervening structures (e.g., a prepositional phrase) between the subject and verb. Native speakers of English are sensitive to S-V agreement violations while listening to speech (e.g., The dog jump, where * indicates ungrammatical).

Unlike English, Chinese has no inflectional marking, and thus, no overt S-V agreement. Due to these differences, Chinese learners of English tend to drop the plural marker -s in their production. This cross-linguistic difference in linguistic typology makes for a useful bilingual probe.

In this study, 87 bilingual speakers of Mandarin and English between the ages of 18 and 31 years participated in two acceptability judgment tasks in two experiments. In Experiment 1, participants listened to 224 sentences, consisting of unmodified subjects and intransitive verbs. In Experiment 2, they listened to 256 sentences, in which a prepositional phrase (PP) follows, and modifies the subject. Half of the sentences with intervening PPs included nouns of a different number than the subject and that could induce attraction errors (i.e., incorrect agreement marking on the verb triggered by the PP noun, for example, *The dogs with a bone barks). Each audio stimulus was delivered in one of three accent types (i.e., heavy Chinese-accented English, light Chinese-accented English, and Standard American English). After each trial, participants judged the sentences on a Likert scale (0-5) with 0 being least acceptable and 5 being most acceptable. Responses were converted into hits (H), misses (M), false alarms (FA), and correct rejections (CR), which were then used to calculate A’ for the dependent measure in the models.

The results from this study demonstrate that Chinese-English bilinguals can detect S-V violations in both basic (noun + verb) (Experiment 1) and sentences with PPs (noun + prepositional phrase + verb) (Experiment 2). Furthermore, in Experiment 2, they showed attraction errors even when the local interfering noun was inanimate and semantically implausible with the verb. While different types of accents did not significantly predict performance on the grammatical processing task in either experiment, an effect of familiarity of accent was seen when comparisons were made between the hit rates and false alarm rates of the two experiments. Furthermore, accent led to significant differences in acceptability (irrespective of experiment or grammaticality) and these differences might be driven by familiarity with the accents. Language background factors such as English use and English proficiency predicted grammatical processing. The findings also revealed that people’s self-perception of their accent in the second language was correlated with performance on grammatical processing in that language.

Download

This work is embargoed and will be available for download on Saturday, February 01, 2025

Share

COinS