Neural mechanisms for response selection: representation specific or modality independent?

Abstract

Is response selection a unitary central cognitive mechanism? Some information processing theories assume it is (Pashler, 1994; Welford, 1959); whereas others suggest it may not be (Meyer & Kieras, 1997). This question was addressed by two studies using functional magnetic resonance imaging (fMRI) published in this issue of the Journal of Cognitive Neuroscience . Schumacher, Elston, and D’ Esposito (2003) manipulated responseselection difficulty in two visual‐ manual tasks using spatial and nonspatial stimuli. They found that largely distinct brain regions mediated response selection for the two tasks; with right dorsal prefrontal, and bilateral premotor and superior parietal brain regions subserving spatial response selection and left dorsal prefrontal, ventral parietal, and temporal brain regions subserving nonspatial response selection. These results led these authors to conclude that the neural mechanisms for response selection are specific to the type of information processed. Jiang and Kanwisher (2003a) reached a different conclusion. They manipulated response- selection difficulty in visual‐ manual, auditory‐ manual, and visual‐ vocal tasks. Like Schumacher et al., Jiang and Kanwisher observed fronto-parietal activations for spatial response selection. Contrary to Schumacher et al., however, Jiang and Kanwisher found activation in these same brain regions for nonspatial response selection. These results led Jiang and Kanwisher to conclude that response selection is a modality-independent unitary process. That is, the same neural mechanisms mediate response selection across a variety of tasks and modalities. Is there a way to reconcile the results of these two studies, which address the same question with similar designs, yet reach opposite conclusions?