Semantic modulation of time-to-collision judgments

Observers are able to make generally accurate judgments of the time-to-collision (TTC) of approaching stimuli. Traditional theories have emphasized the role of optical cues about the expansion of the retinal image in this ability. Recent work, however, has further emphasized the role of semantic information about the object. Here we investigate the role of semantic information in TTC judgments by presenting a range of real-world objects, which varied widely in size, weight, and hardness. Our results show that the physical characteristics of looming stimuli predict observers' TTC estimations. Bigger, heavier, and harder objects were underestimated more, relative to smaller, lighter, and softer objects. As expected, actual TTC and stimulus size were also significant predictors of TTC judgments. In estimating the arrival time of looming stimuli, observers automatically take into account several characteristics of the stimuli, even though these characteristics are completely task irrelevant. This suggests that semantic properties of seen objects and the consequences of their impact on the observer's body are processed automatically.