Rhesus monkeys use both eye and head gaze to reallocate covert spatial attention facilitating visual perception

Nonverbal cues, particularly eye-gaze, significantly shape human social interactions. Although nonhuman primates reliably follow head gaze, their capacity to use eye-gaze alone for inferring the other's focus of attention remains debated. We investigated this question using a realistic rhesus monkey head avatar that directed its gaze toward one of two LEDs (left or right), employing either eye movements alone or combined eye and head movements. After a randomly chosen interval (range: 50-400 ms) from gaze presentation, one LED transiently increased its luminance to near-threshold levels. Rhesus monkeys were trained to detect and report this luminance change via a saccade to the corresponding LED, independent of the avatar's gaze direction, to receive rewards. Our results showed that head-gaze cues robustly directed covert attention toward gaze-congruent targets with short delays, indicative of reflex-like, stimulus-driven orienting. In contrast, eye-gaze alone, at comparable amplitudes, did not affect attentional shifts. However, increasing the avatar's size and eye-gaze amplitude, simulating a close-range interaction, made eye-gaze cues effective in guiding attention. These findings demonstrate that rhesus monkeys possess the capacity to use eye-gaze cues to determine conspecifics' attentional targets, and validate and underscore the utility of 3D animal models as powerful tools for generating realistic yet precisely controlled stimuli. Our study supports the idea that eye-gaze following is not uniquely human but is an evolutionarily ancient ability, likely shared across Old World monkeys and apes that diverged more than 30 million years ago.