Automatic domain-general processing of sound source identity in the left posterior middle frontal gyrus

Bruno L Giordano,Cyril Pernet,Ian Charest,Guylaine Belizaire,Robert J Zatorre,Pascal Belin

doi:10.1016/j.cortex.2014.06.005

Abstract

Identifying sound sources is fundamental to developing a stable representation of the environment in the face of variable auditory information. The cortical processes underlying this ability have received little attention. In two fMRI experiments, we investigated passive adaptation to (Exp. 1) and explicit discrimination of (Exp. 2) source identities for different categories of auditory objects (voices, musical instruments, environmental sounds). All cortical effects of source identity were independent of high-level category information, and were accounted for by sound-to-sound differences in low-level structure (e.g., loudness). A conjunction analysis revealed that the left posterior middle frontal gyrus (pMFG) adapted to identity repetitions during both passive listening and active discrimination tasks. These results indicate that the comparison of sound source identities in a stream of auditory stimulation recruits the pMFG in a domain-general way, i.e., independent of the sound category, based on information contained in the low-level acoustical structure. pMFG recruitment during both passive listening and explicit identity comparison tasks also suggests its automatic engagement in sound source identity processing.

Highlights

Natural objects in the environment produce variable sounds: a speaker utters different phonemes; a guitar plays different tones; a drinking glass produces different impact sounds depending on how it is struck
A conjunction analysis revealed that the left posterior middle frontal gyrus adapted to identity repetitions during both passive listening and active discrimination tasks. These results indicate that the comparison of sound source identities in a stream of auditory stimulation recruits the pMFG in a domain-general way, i.e., independent of the sound category, based on information contained in the low-level acoustical structure. pMFG recruitment during both passive listening and explicit identity comparison tasks suggests its automatic engagement in sound source identity processing
Significant effects of the sequence type/N repetitions factor emerged in three left temporal clusters, in the left Rolandic operculum (RolOp), in a comparatively larger right cluster in the middle temporal plane, extending to the STG, and in the posterior aspects of the left middle frontal gyrus

Summary

Introduction

Natural objects in the environment produce variable sounds: a speaker utters different phonemes; a guitar plays different tones; a drinking glass produces different impact sounds depending on how it is struck. Recognizing the identity of a sound source in the face of this acoustical variability is fundamental to developing a stable and meaningful representation of the auditory environment. With the dual-stream model of auditory processing (Rauschecker & Scott, 2009; Romanski et al, 1999), these studies reveal that the cortical processing of source identity relies on regions part of the ventral “what” stream such as the anterior temporal sulcus (aSTS, Andics et al, 2010; Belin & Zatorre, 2003; Zatorre et al, 2004), and the posterior aspects of the inferior prefrontal cortex/premotor cortex (Latinus et al, 2011; Von Kriegstein & Giraud, 2006; Zatorre et al, 2004). The close proximity of the above-mentioned posterior inferior prefrontal regions to regions that implement contentindependent control and working-memory processes in the premotor/middle prefrontal cortex (Power & Petersen, 2013; Rottschy et al, 2012) may suggest an involvement in domain-general identity processing, i.e., an involvement in identity processes not modulated by the particular category of the sound stimulus

Methods

Results

Discussion

Conclusion