CANCELLED // Physics Colloquium : "Social motion - The informational cues and inference processes underlying human cooperation and competition scenarios"

Date: 
Mon, 23/03/2020 - 12:00 to 13:30

Location: 
Levin building, Lecture Hall No. 8,

Lecturer: Dr. Yuval Hart, from the Psychology Department at HUJI.
Abstract:
At the basis of our social interactions lies a cohort of subtle movements that signal invitations, affects, and intentions. While social interactions are highly complex, in both cooperative and competitive scenarios people recognize and act upon motoric cues in surprising accuracy and speed. In this talk, I will present two studies that aim to reveal the informational cues and inference mechanisms that propel our decision-making process in social contexts. In the first study, we asked how people bridge between their individual differences to create joint, synchronized motion? We showed that while each participant has a unique motion signature, when participants synchronize together they use universal, optimal motion which consists of smooth and predictable movements (minimal jerk motion). In the second study, we asked what are the informational cues that underlie human quick and accurate inference of the other's motion onset in an experiment where Blockers need to react quickly to Attackers' motion onset. We used concepts from criticality theory to show that people's motoric decision making is akin to a critical transition and thus is accompanied by early warning signals. The early warning signals - a sharp rise in motion's autocorrelation at lag -1 and a sharp rise in the autocorrelation decay time - occur ~130 ms before motion onset and strongly correlate with Attackers' motion onset and with Blockers' response times. The divergence of motion's variance near the transition point indicates that the transition is of a fold-bifurcation type. Lastly, we simulated the motion of the fold-bifurcation events to show that people react faster to motion driven by fold-bifurcation dynamics than to its uncorrelated counterpart. Together, our findings suggest people can recognize motoric decision making by inferring its early warning signals, allowing for a fast and accurate response.

image can not be displayed