Inside and outside the body space: a study on environmental knowledge, interoceptive and functional body representations in the prodromal stage of dementia.

There is a rich history of research on the relation between action and perception, beginning in the late nineteenth century (Woodworth, ). While much work has focused on the use of perceptual information in guiding actions (e.g., Woodworth, ), recent research has revealed a surprising effect of actions on perception: merely producing an arbitrary action (such as a key-press) in response to a stimulus can result in enhancement of features of that stimulus. For example, Buttaccio and Hahn () showed a bias in visual search toward the color of a shape that had been responded to, but not merely observed, shortly before the search. Many other studies have shown a variety of different effects of action including enhancement of individual features of the acted-on object such as color and shape (Wang et al., ), an influence on the production of eye movements as well as manual responses (Wang et al., ), and strengthening of word-picture priming (Weidler & Abrams, ). According to one explanation, the stimuli that are present when an action is made are, by virtue of the production of the action, important, and deserve the benefit of prioritized processing (Weidler & Abrams, ). For example, people often produce repeated actions toward a given object (e.g., hammering a nail, or peeling a fruit), and any prioritization of the features of the acted-on object would presumably facilitate those repeated actions. Recently it has been reported that actions may also influence the extraction of summary statistical information from a display of multiple objects – an ability known as (Knox et al., ). Ensemble processing is thought to play an important role in our ability to rapidly assess key attributes of complex scenes (Whitney & Yamanashi Leib, ). Knox et al. () had participants judge the average size of an ensemble of eight oval shapes by selecting one of two alternative ovals as a response. Before viewing the ensemble, though, participants either made an action to a small or large rectangle, or they simply viewed a small or large rectangle. The key finding was that participants were more likely to incorrectly select a lure response alternative when that lure matched the (categorical: large vs. small) size of the earlier rectangle – but such a result only occurred when participants had produced an action – not when they had merely viewed the rectangle. The interpretation was that the action had biased attention in the processing of ensemble information to favor elements that were closer in size to the acted-upon rectangle. Such an influence of action on ensemble processing is potentially important because it contributes to a growing literature on the role of attention in ensemble perception (e.g., Chen et al., ; Kimchi & Sabary, ). Although the conclusion offered by Knox et al. () is consistent with an effect of action on ensemble processing, there are at least two other ways in which action might have affected responses in their task. First, if action biases processing of shapes congruent with the size of the acted-toward rectangle, then it is possible that the action affected the processing of the shapes that were presented in the response phase of the experiment as opposed to those in the ensemble itself. The response-alternative shapes were presented only 200 ms after removal of the ensemble, so any effect of action may still have been in force at that time. Second, action in the Knox et al. task may not have influenced perception at all – it may instead have influenced processes related to memory, decision making, or response selection. For example, after making an action to a large square, participants may simply have been biased to select the larger of the two response options. That the action effect might not have influenced perception in the task seems possible because the stimuli in the ensemble task (ovals) bore very little in common with the stimuli to which the actions were directed (rectangles). In past studies, action toward a stimulus of a particular color (Weidler & Abrams, ) or shape (Wang et al., ) were shown to bias attentional selection of items with matching colors or shapes. In the present study, the shapes to which actions were made (rectangles) were different from the shapes in the ensemble and response phases (ovals) and shared only the categorical feature of large or small size – and it is not clear if categorical size supports the action effect. Importantly, if either of these possibilities holds, then the results of Knox et al. () cannot be taken to indicate an effect of action on ensemble processing.