Searching for actions

Abstract

Suppose that you are searching for a teapot in your kitchen, which is crowded with objects. Current theories of attention assume that it is the perceptual properties of the sought-for object (such as its shape, colour or texture), which are stored in long-term memory and temporarily retrieved in working memory, that guide your selection. However, in a recent report, Humphreys and Riddoch 1xDetection by action: neuropsychological evidence for action-defined templates in search. Humphreys, G.W and Riddoch, M.J. Nat. Neurosci. 2001; 4: 84–88Crossref | PubMed | Scopus (85)See all References1 suggest that the potential action an object affords might also provide an effective memory template for attentional selection. They used a visual search task to compare perceptual versus action cues (‘affordances’) in a patient, ‘MP’, with hemispatial neglect following damage to the right hemisphere. Results were clear-cut. Across the different experiments, MP was significantly faster and more accurate in detecting contralesional targets defined by the action they afforded (e.g. ‘an object you would use to pour from’) than detecting targets cued by colour (‘a brown item’) or name (‘a teapot’). This held regardless of whether the action cue was given through a verbal description or a gesture made by the experimenter. The effect disappeared, however, if words instead of objects were used as targets (words do not afford a limb action), or when objects were oriented so as to make them less suited for action (handles faced away from the patient).Interestingly though, two other patients with neglect showed the opposite pattern of results, being better at detecting targets defined by their names. This suggests that action-based and perceptual-based selections may depend on functionally separated neural mechanisms. Facilitation of visual processing by action cues could operate through top-down priming signals from a network of areas in the dorsal stream, in particular the ventral premotor cortex and inferior parietal cortex. Neurons in these areas encode the relationship between visual objects and the appropriate movements required to grasp them, rather than the perceptual properties of objects per se.This study confirms the importance of action in visual processing and the highly integrated nature of visual and motor representations. As such, it fits nicely with Duncan's integrated-competition hypothesis 2xCompetitive brain activity in visual attention. Duncan, J et al. Curr. Opin. Neurobiol. 1997; 7: 255–261Crossref | PubMed | Scopus (339)See all References2, which suggests that many different biases, both visual and non-visual, influence hemispatial neglect, and with Rizzolatti et al.'s premotor theory of attention 3xReorienting attention across the horizontal and vertical meridians: evidence in favor of a premotor theory of attention. Rizzolatti, G et al. Neuropsychologia. 1987; 25: 31–40Crossref | PubMed | Scopus (989)See all References3, which maintains that preparation to act on an object enhances perceptual processing of that object. Experiments like these bring the understanding of the relationship between perception and action a step closer.