Dissertations, Theses, and Capstone Projects

Date of Degree

6-2024

Document Type

Thesis

Degree Name

M.S.

Program

Cognitive Neuroscience

Advisor

Valerie Shafer

Subject Categories

Cognition and Perception | Cognitive Neuroscience | First and Second Language Acquisition

Abstract

This study aims to examine the intricate relationship between language experience and the neural processing of stop consonant speech sounds. Previous research has shown minimal differences in amplitudes and latencies of cortical Auditory Evoked Potentials (AEPs) to speech sounds across listeners from various language backgrounds. It is only towards the later latencies (around the P2 peak at 200 ms), that there have been suggestions of cross-linguistic differences. In addition, a recent study, observed an AEP difference between monolinguals and bilinguals when processing speech, specifically an "Nd effect”. The Nd effect, which is an increased negativity of the AEP is proposed to reflect active attention to the stimulus. Yet, it remains unclear whether this effect stems from bilingualism itself or is closely linked to the nonnative quality of target speech sounds, specifically to the bilingual’s second language (L2). To untangle this question, our study took a more comprehensive approach by including both monolingual and bilingual participants and speech sounds that were native and non-native to the various groups.

Three types of stop consonants were presented to participants in our more systematic investigation of cross-linguistic variations in AEPs: short lag voice-onset time (VOT) [pa], prevoiced [ba], and long-lag aspirated VOT [pha]. Bilinguals whose first language (L1) was Hindi, Spanish, or Portuguese, as well as monolingual English listeners made up our participants. All these languages have the same phonetic sound, [p] (written as “b” in English), but only Hindi and English have [pha] (written as “p” in English), while Portuguese, Spanish, and Hindi have [ba] (written as “b” in Spanish) as part of the languages’ phonologies. This selection of speech sounds gave us the opportunity to investigate potential nuances brought about by bilingualism and linguistic experience, as well as general stop consonant perception effects. We hypothesized that language group differences to native versus non-native speech sounds would be seen only for the later AEP peak, P2 (around 200 ms post stimulus onset), and not the earlier peak, N1 (around 100 ms post stimulus onset). We also hypothesized a difference in the Nd, indicating that the bilinguals covertly attend more to the speech than monolinguals).

To execute the study, EEG data were collected from 41 participants (13 Hindi, 11 English, 6 Spanish, 11 Portuguese). We recorded the electroencephalogram (EEG) from 17 scalp electrodes time-locked to the speech sounds. Participants were instructed to ignore the speech stimuli and instead watch a muted movie, with the goal of examining pre-attentive processing of the speech sounds. After data-cleaning and averaging the EEG was averaged across trials that were time-locked to five stimuli of interest. These stimuli had different VOT values (voicing lead [ba] of -100 ms and -36, ms; short-lag [pa] of 0 ms VOT, and long-lag aspirated [pha] or +36 and +96 ms VOT).

Our investigation revealed intriguing findings. For the “easy” stimuli (-100 ms, +96 ms VOT), a linear mixed effects model unveiled no significant language group differences in AEP amplitude. However, a noteworthy effect of stimulus emerged, with a larger P2 amplitude for the aspirated [pʰa] compared to the prevoiced [ba], likely because the acoustic information of aspiration was more energetic than that of prevoicing. Counter to our hypothesis, the interaction between language group and stimulus did not show significance. Similarly, in the difficult condition, no significant effects of language group or stimulus on P2 amplitude were observed. In both conditions, there were no discernible effects of the stimulus, language group, or their interaction on the N1 response. These findings suggest that linguistic experience does not have a discernible influence on the brain's processing of stop consonant speech sounds for AEPs. However, more investigation is necessary to examine whether additional factors, such as amount of experience, and/or age of acquisition of English for the bilingual participants influence the results.

Share

COinS