I’ve always thought that newborns are kind of ugly, so it didn’t surprise me that my first child looked lobster-red and angry after he was born, and not particularly adorable.

Twenty-four hours later, I had a long look at my baby and realized, to my astonishment, that he was exceptionally beautiful. I don’t believe that my baby’s appearance changed that much in one day, but my perception did. I used to think that my perceptual shift was the result of being drunk on oxytocin—a hormone involved in maternal bonding—but a new paper by Viola Störmer and George Alvarez from Harvard University suggests that my conversion may have resulted, at least partly, from my paying attention.   

Previous research had shown that attention can make an object look brighter, or more saturated in color, than it actually is. One good example is the following illusion, discovered by the Dartmouth College neuroscientist Peter Tse:

Look around the image and notice that the three semi-transparent circles are all equal in brightness. Now fix your gaze on the central dot at the intersection of the three circles. While keeping your gaze on the center, direct your attention to only one of the circles. The circle that you selectively focus on will look brighter, and also appear to be positioned in front of the other two circles. Now repeat the process with a different circle. Whichever circle you choose to attend to will appear more prominent than the others.

Störmer and Alvarez wondered whether the enhancement effects of attention might apply not only to low-level visual properties such as brightness and color, but in addition affect so called high-order properties, such as facial attractiveness. To find out, they showed pairs of faces on the left and right halves of a computer screen. The left and right faces were not aligned vertically: one was shifted upward, and the other downward, along the screen’s vertical axis. Without moving their eyes from the center of the screen, subjects had to report the vertical shift (upward versus downward) of the face they found most attractive. To bias the subjects’ attention towards one face or another, the experimenters presented a black dot (called the cue) on just one side of the screen (left or right), selected at random.

Störmer and Alvarez reasoned that, if attention had an effect on the attractiveness judgements, then participants would pick the cued location more (or less) often than the un-cued location.

The data showed that attention indeed exerts a consistent and positive influence on perceived attractiveness: subjects reported the faces at the cued (i.e. attended) location as more attractive than the faces at the un-cued (i.e. unattended) location.

Störmer and Alvarez propose that these findings are meaningful for the perception of real-world faces, and I can’t disagree: I know from personal experience that perception of attractiveness doesn’t end with the first impression. Especially when you’ve been paying loads of attention.