Testing mental computations of center of mass using real-world stability judgments

In many of our daily activities, we arrange objects to create stable configurations. For instance, when placing glasses on the tray, we tend to avoid clustering all glasses at one edge or grouping heavy and light glasses on opposite sides. These decisions are intuitive and effortless, although they might not be flawless. What are the mental computations underlying our everyday stability judgments? We hypothesized that people rely on a Newtonian mental model, computing the center of mass -CoM- to judge the stability of complex systems, rather than relying on heuristics guided by mainly perceptual cues (object size, position and shape). To test this idea, we designed a paradigm that allows us to assess when people employ CoM versus other heuristics. Our real-life experiment consisted of a tilted platform supporting 6 blocks -3 heavy and 3 light-, all of them at different positions, resembling a tray full of glasses. Participants were tasked with making the platform as flat as possible by removing only one block. We manipulated the variables “weight” and “position”, in such a way that the best block to remove varied in position and weight trial by trial. With this design, no single visual heuristic would yield successful performance, allowing us to test whether participants nonetheless employed heuristic-based strategies. Our results show that participants predominantly employed a strategy utilizing CoM computations, consistently choosing blocks that moved the CoM of the platform to a balanced state while exhibiting a slight bias toward choosing heavier blocks. People may not execute a flawless computation of the center of mass, but their approach is remarkably close. Our findings shed light on the cognitive processes involved in stability judgments in everyday tasks, emphasizing the nuanced interplay between physics-based computations and perceptual heuristics.