Cue-shifting between acoustic cues: Evidence for directional asymmetry

Meng Yang,Megha Sundara

doi:10.1016/j.wocn.2019.04.002

Abstract

Previous research shows that experience with co-varying cues is neither sufficient nor necessary for listeners to integrate them perceptually. Auditory Enhancement theorists explain this by positing that listeners integrate two cues more readily if the cues enhance each other’s percept. To isolate the role of enhancement from that of experience, we forced English adult listeners to shift attention between two enhancing cues that they do not use phonemically, pitch and breathiness, by reversing the informativeness of the two cues in a cue weighting experiment. Listeners were able to shift attention from pitch to breathiness and vice versa if the two cues were in an enhancing relation. When this relationship was reversed, listeners could shift attention from pitch to breathiness but not in the opposite direction. Clearly, both the change in informativeness and the enhancing properties of the cues influenced the listeners’ re-weighting of these cues. However, the directional asymmetry was not predicted. Moreover, the same asymmetry was observed in two new groups of listeners who have native language experience with either pitch or breathiness. We discuss the consequences of such asymmetric enhancement effects, rising from either processing limitations or articulatory contingencies, for language change.

Highlights

Speech sound categories are multidimensional and often signaled by two or more acoustic cues
It has been demonstrated that listeners attend to cues other than voice onset time (VOT) when making a voicing distinction
Kingston and colleagues have argued that voicing co-varies with low f0 because the two cues jointly contribute to the percept of low frequency energy (Diehl & Molis, 1995; Diehl et al, 1995; Kingston & Diehl, 1994; Kingston, 2011)

Summary

Introduction

Speech sound categories are multidimensional and often signaled by two or more acoustic cues. Kingston and colleagues have argued that voicing co-varies with low f0 because the two cues jointly contribute to the percept of low frequency energy (Diehl & Molis, 1995; Diehl et al, 1995; Kingston & Diehl, 1994; Kingston, 2011) Cue pairs such as these are enhancing, a specialized term Kingston and colleagues define to refer to cues that reinforce a single auditory effect. They argue that enhancing cues are represented by a higher-level auditory unit, an integrated perceptual property (IPP), mediating between individual acoustic correlates and the features they distinguish In the rest of the paper, we will use the term enhancing in the Kingstonian sense – for cue pairs that converge on a single IPP – and we will use the term integral to describe the fact that these cues cannot be perceived independently

Methods

Findings

Discussion

Conclusion