Abstract
Gaze is one of the most important cues for human communication and social interaction. In particular, gaze contact is the most primary form of social contact and it is thought to capture attention. A very early-differentiated brain response to direct versus averted gaze has been hypothesized. Here, we used high-density electroencephalography to test this hypothesis. Topographical analysis allowed us to uncover a very early topographic modulation (40–80 ms) of event-related responses to faces with direct as compared to averted gaze. This modulation was obtained only in the condition where intact broadband faces–as opposed to high-pass or low-pas filtered faces–were presented. Source estimation indicated that this early modulation involved the posterior parietal region, encompassing the left precuneus and inferior parietal lobule. This supports the idea that it reflected an early orienting response to direct versus averted gaze. Accordingly, in a follow-up behavioural experiment, we found faster response times to the direct gaze than to the averted gaze broadband faces. In addition, classical evoked potential analysis showed that the N170 peak amplitude was larger for averted gaze than for direct gaze. Taken together, these results suggest that direct gaze may be detected at a very early processing stage, involving a parallel route to the ventral occipito-temporal route of face perceptual analysis.
Highlights
It is not surprising that direct gaze captures one’s attention
We investigated the neural responses to faces with direct and averted gaze using high-density EEG and topographical analysis of event-related responses in order to examine if very early differentiated neural responses to direct versus averted gaze could be observed
While classical evoked potential analysis revealed an effect of gaze direction only in the form of enhanced N170 for averted relative to direct gaze, the topographical analysis revealed differentiated ERP topographies to direct and averted gaze as early as between 41 and 80 ms
Summary
Sixteen students of the University of Geneva without any neurological or psychiatric indications participated in this experiment for course credit (8 male/8 female; mean age ± SD = 21.6 ± 3.3). All participants had normal vision and were naïve as to the purpose of the experiment. All procedures were approved by the ethics committee of the University of Geneva (commission d’ethique de la Facultede Psychologie et des Sciences de l’Education, Universitede Genève) and written consent was obtained before the experiment started. One participant (1 male) was rejected because the signal was too noisy. There were six different avatar face identities (3 male), each with direct and (right + left) averted gaze. We used a wavelet decomposition to obtain the high (> 32 cycles/image) and low (< 6 cycles/image) spatial frequency versions of the faces [47]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
