A candidate in a job interview looked at the ceiling for four seconds before answering a question about her biggest professional failure. Two of the three people on the panel wrote something down. Afterwards, one of them described her as evasive.
She was doing something else entirely. She was hunting for the right words, and the three faces watching her were making that hunt harder.
Gaze aversion has a public relations problem. In most offices it reads as shifty, bored or badly raised. The research points somewhere duller and more useful: looking away is often what people do to think.
A human face is a lot of information
Faces are the most information-dense object in most rooms. Eyes, micro-expressions, the question of whether a look was a frown or a squint. Processing all of it takes capacity, and that capacity has to come from somewhere.
Arthur Glenberg and colleagues ran five experiments on this in 1998 and reported three findings. People avert their gaze more as questions get harder. Embarrassment alone does not explain the behaviour. And it works: looking away improved performance. Their proposed mechanism was that averting the gaze disengages a person from their surroundings, which leaves internally directed thinking to run more efficiently.
Later work at Kyoto University narrowed things down. Shogo Kajimura and Michio Nomura asked participants to generate verbs for nouns while watching videos of faces that either looked at them or looked away. Direct gaze slowed word-finding down, though only when both retrieval and selection demands were high. That was a single experiment, and it used video rather than a live conversation, a narrower test of the idea.
The harder the question, the more the eyes wander
Children are not born doing this. Gaze aversion barely features at age five, then climbs steeply through the first years of primary school and settles into adult patterns by around eight.
Gwyneth Doherty-Sneddon and Fiona Phelps questioned 36 eight-year-olds, some face to face and some over a live video link, using questions that varied in type and difficulty. Looking away increased in the face-to-face condition, and difficulty had a very strong effect in both. High levels of aversion signalled a child who found the material hard and was still working on it.
The same group later found the pattern held in autistic children and children with Williams syndrome, where looking away was again associated with thinking. Worth remembering before too much gets read into a child avoiding an adult’s gaze.
The rules depend on where someone learned them
Norms vary, and they vary more than most business travellers appreciate. A team at the University of Tokyo, working with a colleague in Finland, measured heart rate and impressions while Japanese and Finnish participants viewed neutral faces with direct or averted gaze.
The physiological response was similar across both groups. The judgements did not. Japanese participants rated the direct-gaze faces as angrier and less approachable than Finnish participants did. The authors put the gap down to cultural display rules rather than anything hard-wired. Two countries are not the whole world, but the finding still undercuts the idea of one correct amount of eye contact.
The shifty-eyes theory of lying keeps failing
Ask a police officer how to spot a liar and gaze aversion tends to come up first. In a survey of 99 British officers, 73 per cent said that liars look away, well ahead of any other cue.
Samantha Mann’s team tested the claim on 338 passengers at an international airport, some telling the truth about their trip and some lying about it. Gaze aversion did not differ between the two groups. What did differ was deliberate eye contact, which the liars used more, presumably because they wanted to look convincing. The researchers cautioned against treating a single instance of deliberate eye contact as proof of deception.
What to do with any of this
Let the pause happen. When someone looks at the wall after a hard question, they are usually mid-thought, and filling the silence just interrupts it.
Move sideways when the topic is heavy. Walking, driving, cooking and washing up all take the face out of the equation, which is part of why difficult conversations so often happen in cars.
Look away deliberately when the thinking is yours. Saying so helps. A quick “give me a second” buys the time without the room reading it as retreat.
Retire eye contact as an influence tactic. Frances Chen and colleagues found that more eye contact predicted less attitude change in listeners hearing arguments they already disagreed with. Participants instructed to watch a speaker’s mouth shifted their views more than those instructed to hold the eyes. Both studies ran in a lab, and a sales pitch is a noisier setting, though the direction of the effect should trouble anyone trained to lock eyes while making a case.
Eye contact still does real work in a conversation. It also costs something to maintain, and the colleague staring at the ceiling has often decided to spend that budget on the answer.
