Abstract
We regularly interact with objects in our environment in a way that requires some understanding of how objects naturally move. Little is known, however, about how we extrapolate the movements of visually perceived objects. Some theories propose that movements are experienced via mental simulation, allowing an object's trajectory to be internally generated even before it begins to actually move. We examined this possibility by asking human participants to make judgments about certain visual scenes. Specifically, participants (n = 11) were shown a display with a suspended ball located near the top and various randomly arranged "planks" spanning the middle. At the bottom were two possible "catchers" that the ball could fall into. Participants were asked to determine into which of the two catchers the ball would land, if it were to be dropped. This required subjects to assess which planks the ball would hit on its way down, and how the ball would bounce upon hitting a plank. Participants indicated their choices by pressing one of two buttons. Subjects' eye movements were tracked (Eyelink 1000) as they performed this task. We found that reaction times were correlated with the number of planks hit, as well as the total distance travelled by the ball. This finding supports the idea that subjects employed mental simulation, as these two measures would affect simulation complexity and length respectively. We also found that subjects' eye movements prior to their response mapped out a path that overlapped significantly with a computer-generated true trajectory for the ball on a given trial. Lastly, we could predict participants' responses to these displays using the eye movement information from the static "pre-movement" period. Together, these findings provide evidence for mental simulation of natural motion, and pave the way for future investigations of this phenomenon at a neural level in animals.
Meeting abstract presented at VSS 2018