Cues to intention: The role of movement information

Abstract

Body movement provides a rich source of cues about other people’s goals and intentions. In the present research, we investigate how well people can distinguish between different social intentions on the basis of movement information. Participants observed a model reaching toward and grasping a wooden block with the intent to cooperate with a partner, compete against an opponent, or perform an individual action. In Experiment 1, a temporal occlusion procedure was used as to determine whether advance information gained during the viewing of the initial phase of an action allowed the observers to discriminate across movements performed with different intentions. In Experiment 2, we examined what kind of cues observers relied upon for the discrimination of intentions by masking selected spatial areas of the model (i.e., the arm or the face) maintaining the same temporal occlusion as for Experiment 1. Results revealed that observers could readily judge whether the object was grasped with the intent to cooperate, compete, or perform an individual action. Seeing the arm was better than seeing the face for discriminating individual movements performed at different speeds (natural-speed vs. fast-speed individual movements). By contrast, seeing the face was better than seeing the arm for discriminating social from individual movements performed at a comparable speed (cooperative vs. natural-speed individual movements, competitive vs. fast-speed individual movements). These results demonstrate that observers are attuned to advance movement information from different cues and that they can use such kind of information to anticipate the future course of an action.

Introduction

Perception of others’ action is not simply a post-hoc reconstruction of the visual input, but an intrinsically predictive activity. When we observe the movements of other people, we cannot help but anticipate their future course. At the most basic level, from seeing the start of a movement, we can predict how it will end (Frith & Frith, 2006). For instance, by observing a person throwing a dart on a target board, we can predict the landing position of the dart on the board (Knoblich & Flach, 2001). Similarly, we can anticipate the depth of a volleyball or a tennis serve (Abernethy and Zawi, 2007, Abernethy et al., 2008), predict the success of free shots at a basket (Aglioti, Cesari, Romani, & Urgesi, 2008), or determine whether a player is about to throw the ball or mimic a throw (Sebanz & Shiffrar, 2009). In more complex situations, predictive coding allows us to understand others’ intentions and to predict what they will do next (Frith & Frith, 2006). For example, from seeing someone approaching her hand at a cup of coffee, we can anticipate not only the closing of her fingers on the handle, but the intention to drink.

A current controversy concerns the possibility to understand the intention of others by simply observing their movements (Jacob, 2008, Jacob and Jeannerod, 2005, Kilner et al., 2007). It has been proposed that inferences regarding the intentions associated with a movement are made at the start of the movement and tested by predicting how the movement will continue (Wolpert, Doya, & Kawato, 2003). But is information from movement sufficient to make an inference regarding the intention associated with an action? Someone grasping a cup may grasp it to drink, to hand it to another person or to examine the cup itself. Is it possible to understand her intention by simply observing the start of her movement?

In the present research, we investigate how well people can distinguish between social and non-social intentions based on movement information. Specifically, we asked whether by observing the initial phase of a two-stage action, observers would be able to understand whether the movement was associated with a cooperative, competitive, or individual intent. Both cooperation and competition necessitate the ability to anticipate the action of the interacting partner. In cooperative situations, understanding the partner’s intention might be important to optimally adapt to her action in service of a common goal (Sebanz, Bekkering, & Knoblich, 2006). In competitive situations, intention understanding might be equally important to prevent the other person’s move in service of a conflicting goal (Ruys & Aarts, 2010).

We conducted two psychophysical experiments to determine how well observers can discriminate between cooperatively, competitively, and individually motivated actions under temporal and spatial occlusion conditions (Abernethy & Zawi, 2007). In Experiment 1, we used a temporal occlusion procedure to determine whether advance information picked up during the initial phase of the movement allowed observers to infer the model’s intention. In Experiment 2, we examined what kind of cues observers relied upon for the detection of social intentions. This was done by masking selected spatial areas of the video clips representing the model’s movement.