Abstract
How does vision interface with the rest of cognition? We showed observers an animation in which the main event was somewhat ambiguous (a red car moves behind a blue car, which then moves and stops next to a pedestrian, who then falls), and asked them to judge who was to blame for the outcome. Below this key event, we showed an irrelevant contextual event (a circle moves and stops next to a second circle, which then moves at the same velocity as the first). People tend to perceive the first circle as causing the second circle to move, and such contextual events have been shown to increase perceived causality in an ambiguous event that occurs simultaneously. Thus, we hypothesized that the contextual event would make participants more likely to see the red car as causing the blue car to move, and therefore blame the driver of the red car for the incident. Indeed, observers were more likely to blame the driver of the red car when there was a causal-contextual event than when there was not, suggesting that causal phenomenology steered their moral judgments (E1). The severity of blame judgments scaled with the contextual event's duration (E2) and temporal asynchrony relative to the car event (E3), and the effects also generalized to a potential boundary condition — a mini car bumps into a huge trash truck — where explicit knowledge might have prevented the effect (observers could reason that a mini is unlikely to cause a huge truck to accelerate to the same speed from rest; E4). Further, only 30% of observers predicted that such contextual events would change their blame judgments about the drivers (E5), indicating that these effects are counterintuitive. Collectively, the findings suggest that causal phenomenology elicited by irrelevant contextual events leaves systematic fingerprints in moral judgment.
Meeting abstract presented at VSS 2017