Where a person in a conversation looks: at the partner's face for
seeded glances, away between them, more while listening than while
speaking.
Acarturk et al. (2021, J Eye Mov Res 14(1):1, mobile eye tracking of mock
job interviews) measured both parties: the listening interviewer looked
at the face for 3.89 s on average and away for 0.29 s, the speaking
interviewee at the face for 1.82 s and away for 0.83 s (Table 3), so a
listener spends about 93 % of the time on the partner's face and a
speaker about 69 %, more while listening than while speaking as Kendon
(1967, Acta Psychol 26:22) and Argyle and Cook (1976) observed. Each
participant's gaze alternates glances at the face and looks away, both
durations drawn from exponential distributions of those means, the study's
standard deviations being about their means. The listener's aversions go
diagonally, down and to a seeded side; the speaker's follow the study's
mix after a look at the face: right 39 %, left 19 %, down 20 %, diagonal
19 % and up 3 % (Table 5). Each aversion turns the gaze by the eyes' own
share of a gaze shift (HumanGaze.EYES, under which the head stays put,
Freedman 2008), split evenly between aside and down on a diagonal, to a
point at the partner's distance, the personal conversation gap (Hall
1966), so the eyes keep their vergence. A turn boundary starts a new
glance at the partner, as speakers look up at the end of their turn
(Kendon 1967).
The result is a gaze layer's requests, so the conversation's gaze never
cuts the bodies' actions.
Where a person in a conversation looks: at the partner's face for seeded glances, away between them, more while listening than while speaking.
Acarturk et al. (2021, J Eye Mov Res 14(1):1, mobile eye tracking of mock job interviews) measured both parties: the listening interviewer looked at the face for 3.89 s on average and away for 0.29 s, the speaking interviewee at the face for 1.82 s and away for 0.83 s (Table 3), so a listener spends about 93 % of the time on the partner's face and a speaker about 69 %, more while listening than while speaking as Kendon (1967, Acta Psychol 26:22) and Argyle and Cook (1976) observed. Each participant's gaze alternates glances at the face and looks away, both durations drawn from exponential distributions of those means, the study's standard deviations being about their means. The listener's aversions go diagonally, down and to a seeded side; the speaker's follow the study's mix after a look at the face: right 39 %, left 19 %, down 20 %, diagonal 19 % and up 3 % (Table 5). Each aversion turns the gaze by the eyes' own share of a gaze shift (
HumanGaze.EYES, under which the head stays put, Freedman 2008), split evenly between aside and down on a diagonal, to a point at the partner's distance, the personal conversation gap (Hall 1966), so the eyes keep their vergence. A turn boundary starts a new glance at the partner, as speakers look up at the end of their turn (Kendon 1967).The result is a gaze layer's requests, so the conversation's gaze never cuts the bodies' actions.