Eye Contact and Presence: What the Research on Gaze Actually Shows

What gaze signals: approach versus avoidance

Gaze is a biological signal with deep evolutionary roots. In mammalian social behaviour, direct gaze is an approach signal, an indicator of interest, engagement, or dominance. Gaze aversion is an avoidance signal, indicating submission, discomfort, or disengagement. Human gaze behaviour carries both these signals simultaneously, in complex combinations that vary by context, relationship, and culture.

In social interactions, the basic function of eye contact is mutual monitoring: both parties use it to assess the other's attentional state, emotional tone, and intentions. The person who maintains steady eye contact is perceived as confident, engaged, and honest, partly because sustained eye contact requires a level of comfort that anxious individuals find difficult to sustain, and this difficulty is accurately read by observers as a signal of the underlying state.

Kleinke's 1986 review of the gaze literature found that, in Western studies, people who make appropriate eye contact are generally rated more favourably, for example as more attentive, credible and likeable, than people who avoid it. Culture matters, though. Akechi and colleagues (2013) compared Finnish and Japanese participants facing a real person: eye contact raised arousal in both groups, but Japanese participants rated a face making eye contact as angrier, less approachable and more unpleasant than Finnish participants did. "More eye contact is better" is a Western norm, not a universal rule.

The optimal range: what the duration research shows

Classic observational studies of two-person conversations, summarised by Michael Argyle and Mark Cook in Gaze and Mutual Gaze (1976), give rough averages: mutual gaze, both people looking at each other at once, fills about a third of the conversation. Listeners look at the speaker considerably more, around 70–75% of the time, than speakers look at listeners, around 40%. These are averages from Western samples and vary a lot between people and situations.

How long should a single moment of eye contact last? Binetti and colleagues (2016) showed several hundred visitors to the London Science Museum videos of an actor looking at them for different lengths of time. On average, people preferred about three seconds of direct gaze, with most preferences falling between two and five seconds. Preferences did not depend on gender or personality, but people who preferred longer gaze showed faster pupil dilation, a sign of arousal. Much longer than this, and gaze starts to feel like a stare.

This duration range, however, is a starting point rather than a target. The more experienced social reader is looking not for a specific duration but for a specific quality of gaze, and the qualitative dimension matters as much as the quantitative one.

Quality versus quantity: the soft gaze

Duration is not the whole story. The distinction below between a hard stare and a soft gaze comes from coaching practice rather than from controlled studies, but it captures something most people recognise. It is easiest to see by comparing two kinds of eye contact that last about the same time. The hard stare is characterised by muscular tension around the eyes, the brow furrowed, the eyes slightly narrowed, the jaw set, combined with high-intensity focus on a specific feature of the other person's face. It is experienced as confrontational, evaluative, or aggressive.

The soft gaze is characterised by relaxed muscles around the eyes, a slightly wider visual field (peripheral vision included alongside the focal point), and an absence of the evaluative intensity. It is experienced as warm, receptive, and genuinely attentive. The critical distinction is not duration but the internal state it expresses: the hard stare is associated with high-arousal vigilance; the soft gaze is associated with calm, open attention.

This is directly relevant to the trainability of eye contact. The person who holds sustained eye contact but does so from an anxious, high-arousal internal state will tend to produce a hard, evaluative gaze quality, which observers experience negatively despite the "correct" duration. The person who maintains less technically precise duration but does so from a genuinely calm, receptive state produces a quality of gaze that observers experience as warm and engaging. Duration is a necessary but insufficient condition.

Gaze and social dominance

Research on the "visual dominance ratio" shows an asymmetry linked to status. Most people look at their partner much more while listening than while speaking. People with more power tend to look about as much while speaking as while listening. Dovidio and Ellyson (1982) showed that observers use this pattern: people who looked more while speaking, relative to while listening, were judged as more powerful.

The practical implication is that eye contact while you speak, looking at your audience rather than down at notes, is a stronger signal of confidence and authority than eye contact while you listen.

Why eye contact training is difficult

Eye contact training is among the more counterintuitive of the charismatic behaviour skills because the difficulty does not lie in the mechanical execution, looking at someone's eyes is not physically demanding, but in the internal state required to sustain it comfortably.

Eye contact is arousing. Faces making eye contact are detected faster and processed more deeply than faces looking away, an "eye contact effect" that involves social brain regions including the superior temporal sulcus and the amygdala (Senju & Johnson, 2009), and eye contact raises self-reported arousal (Akechi et al., 2013). For someone who is already socially anxious, this can feed a loop: the eye contact feels uncomfortable, which increases arousal, which makes the eye contact harder to sustain, which increases avoidance, which increases the signal of anxiety the other person receives.

Attempting to force eye contact through willpower under these conditions typically produces the hard-stare quality described above, the duration may be technically correct, but the quality is wrong, and observers experience it as uncomfortable or aggressive rather than warm and engaging.

The effective training sequence for eye contact follows the same pattern as all other charismatic behaviour skills: establish the physiological baseline first (reducing the arousal that makes gaze activating), train the specific skill with gradually increasing duration in low-pressure environments, and progressively increase the social pressure, not the reverse. The goal is not to endure uncomfortable eye contact. It is to reach a state from which eye contact is naturally comfortable, because the internal state it requires has become the default.