DS0804 -

Perception changes with the evolution of action representations – PERCERA

How do you perceive this object? It depends on what you can do with it

The project raises the general issue of the interrelations between perception, action, and cognition, which are viewed as a source of change. We hypothesize that the way people perceive and represent their environment depends on their own action repertoire and action abilities. Consequently, changes in the action repertoire and action abilities will entail modifications in the perception and representation of the environment.

Consequences of the involvement of multiple action representations on perception

Theories of embodied cognition posit the existence of intimate relations between perception, action and object representations. Perception of manipulable objects or perception of object-directed actions would activate the same action representations as those involved in actual action execution. Moreover, a single object may be associated with different gestures to achieve various goals. Multiple action representations may therefore be simultaneously activated. When people act, they have to select an integrate the relevant action representations and these processes affect action performance. If action representations are also recruited in the absence of real action, then perceptual processing should also be influenced by the range and monitoring of the different action representations activated. Thus, the aim is to characterize the mechanisms underlying competition and integration between action representations during object and action perception. Since action representations and competition monitoring abilities are dynamic, we further aim at highlighting perceptual modifications that may be caused by developmental, interindividual and contextual changes in action processing.

One protocol assesses the perception of objects in virtual environment. Participants perform different types of perceptual judgements on objects presented at different distances. Half of the objects are conflictual, namely associated with distinct grasp and use gestures. When conflictual objects are within reach, the distinct gesture representations activated compete with one another, which slows down perceptual processing. We evaluated how the cost of the competition between gesture representations changed between 8 and adulthood. We also studied the neural correlates of this processing cost using electroencephalography and how it may be modulated by the presentation of action verbs.
A second protocol measures the identification of actions containing gesture and/or goal violations. It evaluates activation and competition between the gesture and goal representations of the same action. Participants judge whether action photographs are correct or not. Action recognition was assessed at the behavioral and brain levels using electroencephalography and transcranial brain stimulation. In one study, questionnaires evaluating individual social functioning were added.

Object perceptual processing is slowed by the competition between gesture representations. This cost does not follow a linear development, it is visible in young children and adults but disappears during adolescence. At the brain level, the cost is reflected by a suppression of motor resonance during object perception, more or less importantly depending on the verbal context. Recognition of object-directed actions follows a predictive model, even when contextual cues are minimal. A representation of the action goal is first activated before being integrated with gesture information. Goal priority is modulated by the social functioning of the observer. Results demonstrate the dynamic involvement of action representations of gesture and goal in perception.

The present work highlights three importance sources of changes: development, individual personality characteristics, and current situational context. These changes may affect both the specification and selection of action representations, two complementary mechanisms initially described in the action planning literature. Therefore, we propose to enlarge the challenges and issues of models of affordance selection to all situations of hypothetical interaction with the environment. Important challenges for future research will be to better understand the role higher-level goal representations (expectations about the actor’s intention, task demands, etc.) in the specification and selection of action representations in perceptual situations and to investigate interindividual variability in a more systematic manner. The results of the project also stress the relevance to consider action-perception relations in a developmental perspective for both theory and practice. Transition periods such as adolescence or retirement should receive particular attention. Moreover, the phenomenon of competition between action representations has implications in the domains of robotics, ergonomics, and brain computer interfaces. Finally, the role of verbal context on action-perception relations is important to consider for images created for education, marketing or culture.

Findings of the present project have been published in 6 articles in international peer-reviewed Journals; additional publications are anticipated. They have been presented and discussed in many national and international conferences. The present work is at the origin of novel national and international collaborations in the domains of high-quality image processing and perception-action relations. Finally, the project is the foundation of two PhD theses in cognitive psychology.

The project raises the general issue of the interrelations between perception, action, and cognition, which are viewed as a source of change. If the way people perceive and represent their environment depends on their own action repertoire, then changes in the action repertoire will entail modifications in the perception and representation of the environment. Research on action planning indicates that multiple action representations guide the execution of goal-directed actions and that competition between action representations affects action performance. In addition, similar processes underlie the actual execution of goal-directed actions on the one hand, and the perception of actions and the perception of manipulable objects on the other hand. Thus, we hypothesize that competition between action representations also affects object and action perception. Moreover, the recruitment and monitoring of action representations may evolve during lifespan development and change at a given age under the pressure of contextual and social factors. Consequently, the present project aims at identifying the mechanisms underlying competition between action representations during the perception of objects and object-directed actions. It further seeks to highlight the changes in perceptual processing that can be caused by developmental and contextual modifications in action processing.
The project considers a multi-level model of action representations and focus on three specific aims. Each aim combines the questions and methods of cognitive psychology, developmental psychology, social psychology, and neurosciences through the complementary expertise of four young researchers situated in Northern France. The first aim is to investigate the developmental, contextual, and social changes in the evocation and monitoring of gesture-level representations during object perception. Preliminary data indicate that perception of objects associated with conflicting structural and functional gestures entails a processing cost. In three behavioral studies, we will assess how this cost evolves during life span (childhood, adulthood, elderly), and how it can be moderated by object visual presentation and social variables (e.g. power). Object perception will be examined using a recently developed paradigm in virtual reality. The second aim is to test the hierarchical organization of action representations and specify how and when gesture-level and intention-level action representations participate in action understanding. In four studies, we will identify the temporal dynamics of action decoding at the behavioral and neural levels and further assess how the visual and social contexts modify those temporal dynamics. Action pictures containing gesture or intention violations have been designed to examine this issue. The third aim is to identify the neural markers of conflict between competing action representations in perception using electroencephalography. One study will focus on the conflict between competing gestures during object perception (cf. aim 1) and a second study will focus on the conflict between incongruent gesture and intention during action perception (cf. aim 2).
At the scientific level, findings will fuel theoretical models on perception, action semantics, and embodied cognition. They will further help determining specific periods of the life span during which perception is most affected by competing action representations and identifying factors that can reduce the cost of competition between action representations. At the societal level, findings will have an important impact in the domains of education, neurological rehabilitation, and technological innovations
related to image creation and diffusion.

Project coordinator

Madame Solene Kalenine (Université de Lille - Sciences Humaines et Sociales)

The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.

Partner

UDL - SHS Université de Lille - Sciences Humaines et Sociales

Help of the ANR 203,130 euros
Beginning and duration of the scientific project: - 36 Months

Useful links

Sign up for the latest news:
Subscribe to our newsletter