Differences between Monolinguals and Bilinguals in Phonetic and Phonological Learning and the Connection with Auditory Sensory Memory

Abstract

Bilingualism has been linked with improved function regarding certain aspects of linguistic processing, e.g., novel word acquisition and learning unfamiliar sound patterns. Two non mutually-exclusive approaches might explain these results. One is related to executive function, speculating that more effective learning is achieved through actively choosing relevant information while inhibiting potentially interfering information. While still controversial, executive function enhancements attributed to bilingual experience have been reported for decades. The other approach, understudied to date, emphasizes the role of sensory mechanisms, specifically auditory sensory memory. Bilinguals outperformed monolinguals in tasks involving auditory processing and episodic memory recall, but the questions whether (1) bilinguals' auditory sensory memory skills are also enhanced, and (2) phonetic skill and auditory sensory memory are correlated, remain open, however. Our study is innovative in investigating phonetic learning skills and auditory sensory memory in the same speakers from two groups: monolinguals and early bilinguals. The participants were trained and tested on an artificial accent of English and their auditory sensory memory was assessed based on a digit span task. The results demonstrated that, compared to monolinguals, bilinguals exhibit enhanced auditory sensory memory and phonetic and phonological learning skill, and a correlation exists between them.

Keywords: auditory sensory memory; bilingualism; phonetic and phonological learning.

Figure 1

Sensorimotor systems involved in speech (adapted here from [21]).

Figure 2

Examples of the 4 features in baseline (left) and Model Speech (right). The VCV sequence for tapping was extracted from the word ‘chilly’. The tracks of formants 1–4 are obtained from the vowel in the word ‘Ben’. The spectrograms for epenthesis are obtained from the word ‘spinning’. Pitch tracks for the sequence ‘wasn’t he?’ illustrate the intonation change feature.

Figure 3

Mean of accent scores grouped by the number of novel accent features (1, 2, or 4) per sentence obtained by monolinguals and bilinguals in Baseline, Training and Testing.

Figure 4

Mean of accent scores for each novel accent feature obtained by monolinguals and bilinguals in Baseline, Training and Testing.

Figure 5

Digit span task: proportion of group who reached each list length.

Figure 6

Memory task: mean accuracy for monolinguals and bilinguals for each sequence length.

Figure 7

Digit span task: mean accuracy at each serial position for 7-digit (right) and 8-digit sequences (left). The terminal item is labeled with a 0, and each item preceding it is labeled in terms of its distance from the terminal item (e.g., −1 for the penultimate item, −2 for the antepenultimate item, etc.).

Figure 8

Regression plots for pairs of variables reflecting phonetic and phonological learning (Testing AS, Overall AS) and variables associated with auditory sensory memory (Max Digit, Digit Accuracy, and Score Algo). AS = accent score. Max Digit = maximum sequence length reached. Digit Accuracy = overall accuracy obtained in the digit span task, Score Algo = the corrected algorithm-based score obtained by taking into account permutation errors, with bigger penalties for items displaced at longer distances.