Broken affordances, broken objects: A TMS study

Abstract

It is well known that specific components of motor programs are automatically activated when they are afforded by object related pragmatic features. Among these features the handle appears to be particularly salient for interacting with an object. The aim of this study was to test the modulation of the motor system when object features, particularly relevant for the action, like the object's handle, are violated. In order to address this issue a TMS paradigm was used in which familiar objects with a whole or broken handle, positioned to the right or to the left, were centrally presented. A control condition was also included in which a symbol (‘#’ character) was shown in the right or in the left visual field. Participants had to watch stimuli carefully. The left hemisphere hand motor area was magnetically stimulated 200 ms after stimulus presentation. Results showed that MEP areas were larger when the handle was located to the right side consistent with the visuomotor role of this feature, but only when the handle was complete. The present data (1) suggest a more active role of the dorsal stream in building up object knowledge and (2) allow one to rule out the role of any asymmetrical aspect of an object in motor coding.

Introduction

When interacting with the environment, appropriate behaviour requires a functional link between perceptual and motor systems. The concept of affordance, first introduced by Gibson, 1977, Gibson, 1979/1986, refers to the properties of a surface or an object in the environment, that potentiate within a perceiver, specific actions upon it. Consistent with this view, recent behavioural and neurophysiological studies demonstrate that the mere observation of an object involves accessing motor programs for interaction with the object, even in the absence of explicit intentions to act. More specifically, it has been shown that pragmatic features of an object automatically trigger components of specific actions, such as reaching or grasping (Craighero et al., 1999, Ellis and Tucker, 2000, Phillips and Ward, 2002, Tucker and Ellis, 1998, Tucker and Ellis, 2001, Tucker and Ellis, 2004). Among these object features, the handle appears to be particularly salient for interaction. In fact, even if a cup can be grasped from the top or the body, there is no doubt that the handle is the component that most commonly triggers the coherent motor program to reach the cup, grasp it and eventually bring it to the mouth. In keeping with this, Tucker and Ellis (1998) showed that the orientation of the handle (right or left), although irrelevant for the task, lead to faster responses when the hand was ipsilateral to the position of the handle. Similarly, Phillips and Ward (2002) showed that in a paradigm in which the target was superimposed on an object, the handle of the object being oriented to the right, left or in a neutral position, participants were faster when the handle orientation was coherent with the response position.

These behavioural results have their counterpart in neurophysiological studies showing that, both in humans and in monkeys, neural circuits involving sectors of the parietal and premotor cortex are devoted to coding the pragmatic features of objects. In the monkey, within an area inside the intraparietal sulcus (AIP), neurons are endowed with visual and motor properties like the shape, size and orientation of a specific observed object (Murata et al., 1997, Sakata et al., 1995). In a sector of the premotor cortex (area F5) neurons which discharge during the execution of specific goal directed actions like grasping, holding, manipulating specific objects, have been found (Rizzolatti et al., 1988). A set of F5 neurons also have visual features and selectively respond to the observation of objects with pragmatic features coherent with the action motorically coded by the neuron (canonical neurons). It has been hypothesised that AIP and F5 interact strictly in affording the most suitable motor program for acting upon an object, when selected (Jeannerod, 1995).

As for humans, in an early PET study Grafton, Fadiga, Arbib, and Rizzolatti (1997) showed that the observation of common objects leads to the activation of the left premotor cortex thus implying an automatic recruitment of the motor system during object observation, even in the absence of any motor output. In a further fMRI study, Binkofski et al. (1999) showed that the manipulation of an object recruits brain regions in humans including the ventral premotor cortex, the inferior frontal gyrus, and an area inside the intraparietal sulcus, the latter two areas possibly corresponding to the human homologues of F5 and AIP, respectively. In keeping with these results, Chao and Martin (2000) also identified a fronto-parietal circuit during the observation of pictures showing tools. Grèzes, Tucker, Armony, Ellis, and Passingham (2003) showed that the degree of activation within the fronto-parietal circuit during the execution of power or precision grip responses covaried with the action afforded by the object.

This experimental evidence fits the well-known theoretical model introduced by Milner and Goodale (1995) which claimed that visual input is segregated into two main pathways: a ventral stream and a dorsal stream, each responsible for a different analysis of visual information. According to this model, the ventral stream (from occipital visual areas to temporal lobe) is mainly involved in processing semantic knowledge of visually presented objects, while the dorsal stream (from the occipital visual areas to premotor cortex through the parietal lobe) is responsible for visuomotor transformations necessary to act (for a review, see Milner & Goodale, 2008). The dorsal stream appears to have a specific role in on-line control during the course of an action. Fischer and Dahl (2007) showed observers an animation of a rotating cup and found spontaneous lateralized motor preparation more pronounced for the right than the left-hand, thus showing that vision for action is rapidly updated. This raises the issue of how the motor system is modulated when the most important features of an object relevant for action are violated. In other words what happens in our motor system when we look at a cup with a broken handle? Given the privileged role of some object components in affecting the motor response and the role of these in visuomotor transformations, we could expect that the motor programs triggered by the handle are violated as well. Since the excitability of primary motor cortex reflects the activity of the premotor–parietal circuits known to be involved in sensorimotor transformations (Rizzolatti & Luppino, 2001), we employed single pulse TMS technique to address this issue.