TY  - JOUR
ID  - matusz2016context
T1  - The context-contingent nature of cross-modal activations of the visual cortex
A1  - Matusz, Paweł J
A1  - Retsa, Chrysa
A1  - Murray, Micah M
JA  - Neuroimage
Y1  - 2016
VL  - 125
SP  - 996
EP  - 1004
PB  - Elsevier
UR  - https://www.sciencedirect.com/science/article/pii/S1053811915010381
M2  - doi: j.neuroimage.2015.11.016
KW  - attention
KW  - Context
KW  - control
KW  - Crossmodal
KW  - multisensory
KW  - VisualERP
N2  - Real-world environments are nearly always multisensory in nature. Processing in such situations confers perceptual advantages, but its automaticity remains poorly understood. Automaticity has been invoked to explain the activation of visual cortices by laterally-presented sounds. This has been observed even when the sounds were task-irrelevant and spatially uninformative about subsequent targets. An auditory-evoked contralateral occipital positivity (ACOP) at ~ 250 ms post-sound onset has been postulated as the event-related potential (ERP) correlate of this cross-modal effect. However, the spatial dimension of the stimuli was nevertheless relevant in virtually all prior studies where the ACOP was observed. By manipulating the implicit predictability of the location of lateralised sounds in a passive auditory paradigm, we tested the automaticity of cross-modal activations of visual cortices. 128-channel ERP data from healthy participants were analysed within an electrical neuroimaging framework. The timing, topography, and localisation resembled previous characterisations of the ACOP. However, the cross-modal activations of visual cortices by sounds were critically dependent on whether the sound location was (un)predictable. Our results are the first direct evidence that this particular cross-modal process is not (fully) automatic; instead, it is context-contingent. More generally, the present findings provide novel insights into the importance of context-related factors in controlling information processing across the senses, and call for a revision of current models of automaticity in cognitive sciences.
ER  -