Babies as young as three months old will follow the eyes of an adult to look at the same thing the adult is looking at. This behavior makes sense from an evolutionary perspective: if a predator or other danger looms, we can learn from the actions of others (though it's unclear exactly what a three-month old would do to escape a ravenous bear).
But if the gaze-following behavior is really a survival adaptation, wouldn't we be more likely to follow someone's gaze if they also had a fearful facial expression? After all, if someone's glancing to the side with a cheerful smile, we don't expect they're looking at a jaguar ready to pounce. There's some evidence to support this notion. Several researchers have found that in photos of crowds, people are quicker to spot angry faces than happy faces.
A team led by Andrew Mathews showed volunteers photos of faces looking either to the left of the right, and with fearful or happy expressions. Then a letter appeared to one side of the face. While participants were faster at identifying the letter when it was on the side the face was looking at, they weren't any faster when the face had a fearful expression. So does this mean people never pay attention to the facial expression when following someone's gaze?
Mathews's team suspected that while not everyone is affected by the expression of the face, perhaps people who had higher than normal levels of anxiety would react differently. They tested a number of college students using the State-Trait Anxiety Inventory, which rates anxiety on a scale of 20 to 80, with average levels of anxiety at a score of about 40. Forty-five people were selected from this group: those with scores above 45, and below 35.
They then tested these participants using the same protocol they had tried before: viewers focused on a cross at the center of the screen. Then a photo of either a happy, neutral, or fearful face appeared. Next, the eyes of the face moved to look either to the left or right, and finally, a target letter appeared, half the time in the same direction as the photo's gaze, and half the time in the opposite direction. Viewers were told their task was simply to identify the letter that appeared as quickly as possible, and that the direction of the face's gaze should be ignored, since it did not predict where the letter appeared. Here are the results:
All participants were quicker to recognize the letter when it appeared in the direction of the gaze. However, as before, those with low anxiety did not recognize the letter any faster with the fearful face. But high-anxiety participants had a significantly different pattern. When the face looked toward the letter, these viewers identified it significantly faster when the face was fearful, compared to a neutral face.
One thing Mathews's team didn't do is manipulate the orientation of the head: gaze direction was achieved solely through eye movement. In one of the earliest articles we wrote for Cognitive Daily, we discussed a study showing that head orientation is more important in determining gaze direction, so it would be interesting to see this study repeated with different head orientations. Perhaps in that case, we would all react faster to fearful faces.
Mathews, A., Fox, E., Yiend, J., & Calder, A. (2003). The face of fear: Effects of eye gaze and emotion on visual attention. Visual Cognition 10(7), 823-835.