Abstract
People naturally move both their head and eyes to attend to information. Yet, studies of attentional orienting normally immobilize the head in a chin rest in order to focus on eye movements, thus questioning their ecological validity. Here, using a virtual reality headset with built-in eye tracking, participants were asked to view indoor and outdoor, fully immersive, 360-degree scenes while their head and eyes were tracked. In order to investigate eye-head coordination, participants viewed the scenes through a small moving window that was yoked either to their head or eye movements. We found that perturbations induced by the head- or gaze-contingent windows affected head and eye movements differentially, in line with their distinct roles in head-eye coordination. Compared with windowless viewing, gaze-contingent viewing was more disruptive than head-contingent viewing, indicating a functional separation between head and eye. Indeed, gaze-contingency actually decreased the coupling between head and gaze movements in 360-degree scene exploration, while head-contingent viewing looked more like windowless viewing. These data dovetail with the nested effectors hypothesis, which proposes that the head prefers exploration into non-visible space while the eyes prefer to exploit visible areas delivered by the head. It also suggests that real-world orienting may be much more head-based than previously thought. We discuss our findings in relation to the cognitive repercussions of eye vs. head movements, as well as highlighting the utility and ecological validity of unconstrained eye and head tracking in virtual reality.