The second-agent effect: communicative gestures increase the likelihood of perceiving a second agent

Abstract

Background: Beyond providing cues about an agent's intention, communicative actions convey information about the presence of a second agent towards whom the action is directed (second-agent information). In two psychophysical studies we investigated whether the perceptual system makes use of this information to infer the presence of a second agent when dealing with impoverished and/or noisy sensory input.

Methodology/principal findings: Participants observed point-light displays of two agents (A and B) performing separate actions. In the Communicative condition, agent B's action was performed in response to a communicative gesture by agent A. In the Individual condition, agent A's communicative action was replaced with a non-communicative action. Participants performed a simultaneous masking yes-no task, in which they were asked to detect the presence of agent B. In Experiment 1, we investigated whether criterion c was lowered in the Communicative condition compared to the Individual condition, thus reflecting a variation in perceptual expectations. In Experiment 2, we manipulated the congruence between A's communicative gesture and B's response, to ascertain whether the lowering of c in the Communicative condition reflected a truly perceptual effect. Results demonstrate that information extracted from communicative gestures influences the concurrent processing of biological motion by prompting perception of a second agent (second-agent effect).

Conclusions/significance: We propose that this finding is best explained within a Bayesian framework, which gives a powerful rationale for the pervasive role of prior expectations in visual perception.

Figure 1. Example of a communicative signal trial.

Agent A points to an object to be picked up; agent B bends down and picks it up. B was presented using limited-lifetime technique (6 signal dots) and masked with temporally scrambled noise dots. The noise level displayed is the minimum allowed in the experiment (5 noise dots). To provide a static depiction of the animated sequence, dots extracted from 3 different frames are superimposed and simultaneously represented; the silhouette depicting the human form was not visible in the stimulus display.

Figure 2. SDT parameters for the two experimental conditions.

Error bars are 95% confidence intervals.

Figure 3. SDT parameters for the three experimental conditions.

Error bars are 95% confidence intervals.