Abstract
Previous studies have demonstrated that humans have an attentional priority for detecting animals compared to inanimate objects. While these findings are consistent with the claim regarding the evolutionary significance of animals in human history, it is unclear whether the attentional advantage for animals depends on visual or semantic differences between the categories. Indeed, animals and inanimate objects vary greatly in shapes, and the visual and semantic features are often confounded. To distinguish visual and semantic influences on the attentional priority for animals, we conducted two visual search experiments (both N=28). Participants searched for the presence of either an animal or a man-made object on a 6-item display, with a non-target category item (e.g., man-made object when searching for an animal) appearing as a distractor in half of the trials. The rest of the distractors were fruits/vegetables. Experiment 1 used images that differ in visual characteristics across the three categories, and revealed faster and better search for animals, compared to man-made objects, replicating previous findings. Also, search performance for either category was comparably reduced with the presence, compared to the absence, of a non-target category item. Experiment 2 used instead images that were equated in image statistics and shapes (either round or elongated) across the three categories. Two aspects of the results are critical: 1) the overall advantage for animal search, compared to object search, was minimized, but 2) an interaction in response times revealed that the presence of an animal as a non-target category distractor slowed down the search performance for object search, whereas the presence of an object distractor did not slow down animal search. These results suggest that while differences in visual features between animate and inanimate categories indeed contribute greatly to the attentional advantage for animals, part of the advantage may also be driven by semantic influences.
Meeting abstract presented at VSS 2017