When do faces capture attention? Evidence from eye movements

A recent visual search study showed that the presence of an upright distractor face slows the search for a butterfly target while a butterfly distractor does not affect the search for a target face, suggesting that faces capture attention automatically (Langton et al, 2008 Cognition 107 330-342). To further test this hypothesis, we recorded eye movements during a cued target search task. When the search target alternated between a face and a butterfly (Experiment 1), faces were found faster and with less saccades than butterflies. The presence of the opposite distractor (eg a face during a search for a butterfly) slowed down the search but to a greater extent when the distractor was a face. Moreover, faces captured the eyes more frequently than butterflies. Inverted face targets were also found more efficiently than inverted butterfly targets and captured the eyes more than butterflies when presented as distractors (Experiment 2). However, when upright or inverted faces consistently appeared as irrelevant distractors during a butterfly or a flower target search (Experiment 3), they did not cause significant disruption. These results challenge the view that faces capture attention automatically and suggest that faces only attract attention when their processing is relevant during a search task.