What question did this study set out to answer?

This research examines how listeners adapt their attention to different acoustic cues while learning an unfamiliar accent.

May 14, 2026

The role of attention in perceptual learning through acoustic dimension reweighting

Key Points

This research examines how listeners adapt their attention to different acoustic cues while learning an unfamiliar accent.
Participants complete a same/different judgment task based on pitch or duration.
Groups are trained to reweight the importance of F0 and VOT cues during the task.
Data collection is ongoing, with a planned sample size of N=40 per group.
Listeners who attended to pitch showed a higher reliance on F0 than those who focused on duration.
The attention-based training helped participants to adjust the significance of VOT and F0 based on their initial focus.
Data analysis is still in progress, with comparisons expected to reveal significant cue reweighting.

Abstract

This study investigates how listeners adapt to unfamiliar accents by using an artificial English accent in which F0 (pitch) following a stop is the primary cue to the b/p contrast (as in Korean-accented English) while VOT (the duration of the aperiodic noise after stop release), which is the primary cue to the contrast in native English, does not matter. Specifically, lexical feedback indicates that VOT is uninformative and the listener has to learn to rely exclusively on F0 (as in Harmon et al., 2019). This experiment tests the central prediction of learned selective attention theory (Kruschke, 1992), that it is easiest to learn about cues one is already attending to. Native English listeners complete a same/different judgment task in which some judge non-linguistic stimuli as same if they match in pitch, while others judge them same if they match in duration. Both groups are then trained to reweight VOT and F0. We expect both groups to reweight VOT and F0 when their informativities are reversed. However, participants that pay attention to pitch should upweight F0 more than participants who pay attention to duration, who in turn should downweight VOT more. Data collection is ongoing, with a planned N=40 per group.

Ask AI

Helpful

Bookmark