On the role of motor representations and processes in the comprehension of actions and objects

Vannuscorps, Gilles [UCL]

(eng) Every day, we view multiple actions and objects that we can effortlessly recognize. We immediately “comprehend” these actions or objects in the sense that we immediately know what is the probable purpose of the observed movements or the usual function of the seen objects. The issue addressed in this thesis was whether the motor system is involved in the comprehension process of human actions and manipulable artifacts. Within the tradition of cognitive psychology, the motor system is not assumed to play any role in the comprehension of actions or objects. When an action or an object is perceived, it is given meaning when its perceptual representation gains access to stored conceptual knowledge represented in an amodal or “symbolic” format. This traditional view has been recently challenged by two main theoretical proposals that assume that the motor system plays a critical role in the comprehension of human actions and manipulable artifacts, namely, the sensory-motor theory and the simulation theory. According to the sensory-motor theory, the motor system partly stores conceptual knowledge about human actions and manipulable artifacts. This follows from the assumption that knowledge about an action or an object is stored in the sensory and motor systems that were active during past experience with that action or that object. Comprehending an action or a manipulable artifact thus consists in “re-activating”, among others, their motor features stored within the motor system. Within the motor simulation theory, it is assumed that comprehending an action performed by another individual requires that the observer internally simulates that action, that is, covertly executes the movements implicated in that action. Likewise, identifying a manipulable artifact requires that the observer covertly performs the movements that are involved in the object’s actual use. To inform this debate, I have used the empirical paradigm of the single case study in cognitive neuropsychology and examined the performance of brain-damaged patients presenting with conceptual, praxic, or motor disorders and of an individual presenting with atypical motor development due to upper limb aplasia, in various tasks involving the identification and comprehension of actions and manipulable artifacts. With the hypothesis that motor representations play a role in comprehension of actions and/or manipulable artifacts, loss or absence of motor representations should prevent or at least hamper their comprehension. The empirical evidence gathered in these studies challenges “motor theories” of comprehension by showing that manipulable artifacts and actions may be efficiently comprehended even if the observer suffer from damage in the motor system or completely lacks motor representations of these actions and manipulable artifacts. However, it also challenges the traditional view by showing that, under degraded visual conditions, comprehending what others are doing does need the observer’s own motor representations. On this basis, we put forward the middle ground hypothesis that motor representations and processes contribute to action comprehension, but only under adverse perceptual conditions.