How We Perceive Falling and Rising Objects

Our motion perception is remarkably well tuned to detect small changes in speed and direction. For example, soccer goalkeepers need to precisely judge the speed, direction, and curvature of an incoming free kick in order to block it successfully. Baseball players need to make a snap judgment to determine if a pitch will dip, curve to the right, or stay flat. These judgments happen quickly and, for the most part, unconsciously.

However, our experience with moving objects is primarily based in a gravity-constrained real-world context. Objects do not spontaneously accelerate, decelerate, or curve midway through their path. Rather, objects systematically follow the laws of gravity: Falling objects always accelerate downward at 9.81 meters per second squared, while rising objects always decelerate at the same rate. Does our experience with these regularities in the motion of real-world objects transfer to digitally rendered objects?

A study of acceleration perception

A new study by Mai Huong Phan and colleagues published this month in Perception set out to address this question: Is the ability to judge the acceleration of an object influenced by the direction of its movement? More specifically, do people judge the acceleration of falling objects the same way as rising objects?

In the study, participants observed a basketball moving in one of the four cardinal directions (up, down, left, or right) on a computer screen. The basketball was embedded in a naturalistic scene with walls, doors, and people, to provide a sense of scale. Across different trials, the experimenters manipulated the initial speed of the ball and its acceleration (from negative values corresponding to deceleration, to positive values indicating acceleration). Participants were asked to judge whether the ball was accelerating or decelerating.

A bias for falling objects

To compare behavior across the different cardinal directions, the researchers calculated participants’ point of subjective constant velocity (or PSCV), as the acceleration rate that was perceived as being constant motion; in other words, the acceleration rate that led to equal proportions of “acceleration” and “deceleration” judgments across trials. Although the researchers found no difference between the PSCV for rightward and leftward moving balls, they did find a significant difference when comparing the PSCV for upward versus downward moving balls. Specifically, falling balls needed to be accelerating more than rising balls to be judged as moving at a constant speed. This bias is consistent with the idea that we are used to seeing balls accelerate downward, but not accelerate upward. If a ball is moving downward at a constant speed, it may actually appear to be decelerating (because it is not accelerating as we expect).

The study revealed two additional findings: First, the authors found that when the initial speed of the ball was faster, observers were more likely to judge it as accelerating, even if it was moving at a constant speed. That is, observers showed a bias that incorrectly associated speed with acceleration. Second, the authors found that although there were systematic differences in bias between up and down motion, there were no differences in sensitivity. That is, people’s ability to detect a change in acceleration was comparable across upward and downward motion directions. That suggests that our experience with falling objects leads to perceptual biases, but not changes in perceptual precision.

Exploring the role of gravity

Being the first study to reveal this bias for falling objects, many questions remain to explore in future work. For example, previous research has shown that people integrate visual and vestibular cues when making perceiving-oriented objects, words, and faces (Davidenko & Ambard, 2018; Davidenko, Cheong, Waterman, Smith, Anderson, & Harmon, 2018; Davidenko & Flusberg, 2012). In those studies, observers show systematically different behavior when they are lying down sideways versus sitting upright while completing a perceptual task. A future study could examine whether the bias for falling objects depends on the observer’s orientation relative to gravity. If an observer is lying face-up, looking at a moving ball projected onto the ceiling, will the same bias for falling objects still emerge, even when the falling ball is not going in the direction of gravity?