Abstract
Vision scientists have long considered vision, and visual attention specifically, to function as a sort of survival alarm system: always ready to detect novelty, salience, or changes in the environment to alert us of and orient us towards those changes. In spite of the adaptive face validity of that argument, we argue that this assumption does not hold up to close scrutiny. Here, we asked participants to do a 20-step arithmetic task. Participants initially saw a three digit number at fixation, followed every 3 seconds by an operation (displayed for 500ms), to update their current total. The central area was completely empty for 83% of the total trial duration. Critically, large photographs were presented at one of four possible peripheral locations during the trial. Each picture remained on screen for 3 seconds. A new image would onset approximately 1.8 seconds after the latest operand had disappeared from the screen, which (we confirmed) was sufficient time to complete most operations. The appearance/disappearance of peripheral images were the only visual events occurring at these times (no visual information was present at the center). Results: (1) Participants eyes stayed at the center ~60% of the total trial duration. (2) Participants only spent ~10% of trial time looking at peripheral images. (3) In spite of being substantial visual transients, participants never directed their eyes towards new onsets. If their eyes happened to be near the location of a new picture when it appeared, the eyes quickly moved away from the picture. When we are oriented towards the world, vision may play the role of alarm system; yet, in the many instances when we are inwardly oriented (focused on our thoughts), the eyes do what’s right: they work to actively minimize possible distraction from the world, by actively staying away from sources of distraction.
Meeting abstract presented at VSS 2012