Professor Masami K. Yamaguchi's Laboratory (Department of Psychology, School of Humanities, Chuo University), to which I belong, conducts research into perception, such as visual and olfactory (smell) perception, in infants aged from two to eight months and looks at how their ability to see and process faces develops. Infants living in the vicinity of the University come with a member of their families to participate in a 30-minute psychological experiment. The infants who come for these “exams” are known as “Baby Researchers” and we have more than 150 little Researchers visit us annually, on a once or twice a month basis. Thus, we have between two and four infants and their families, and on a busy day as many as 8 little Researchers, visiting the University almost every day to cooperate with our research.

My work in all this is to research the face recognition abilities of infants. Faces are used in communication between humans to identify who the other person is, and in emotion understanding to make a conjecture about the other person’s emotional state, and are said to communicate a greater amount of information than even voices or language. However, the ability to see and process faces is not something that we are born with. It is an ability that is considered to develop gradually after birth, by gaining experience from looking at various faces.

The ability of infants to see and process faces

The ability of infants to see and process faces develops at a remarkable rate in their first year. The ability that they have at birth is limited to searching for a face-like pattern. To us adults, if three dark spots are placed in a white background, it looks like a face (Figure 1, left-hand picture). We know that newborns also are attracted by and look for longer at this type of face-like pattern of black spots. However, newborns are not looking at it because it is a face. Newborns will look longer at any pattern that is top-heavy, that is, any pattern with elements clustered into the top part of the outline, even if it is not a face but some other geometric figure. Thus, newborns seem to prefer and stare for longer at patterns which happen to share some characteristics with faces, but it seems unlikely that they are actually seeing these top-heavy patterns as faces. Until they are about three months old, infants stare at patterns that look like faces, without actually knowing that they are faces, and in this way gain experience of looking at faces.

Once they are three to five months old, infants stop staring as much at unmoving faces and stare for longer at moving faces. The faces that infants see are mostly those of their family members or other adults. When face to face with an infant, people’s faces constantly move: talking to the baby, smiling or laughing. From gaining experience in looking at moving faces, infants develop so that they can better see and process real faces. Further, from five-months old onward, they start to be able to see and process faces in more detail, in terms of the spatial location information of eyes, nose and mouth, such as how the eyes are separated from the nose, and use this as clues in distinguishing between faces. Further, from seven-months old, they become able to recognize facial expression, such as smiling faces and angry faces.

Measuring infant brain activity

The discussion in this article so far has been on experiments that investigate the behavior of infants when looking at and processing faces. However, it is not possible to fully determine from such behavioral testing whether or not an infant is looking at a certain face for longer than other faces because s/he knows it is a face. For this purpose, we use near-infrared spectroscopy (NIRS) to estimate whether an infant is looking at a face because s/he knows it is a face by measuring the infant’s brain activity when s/he is looking at the face. From previous research, we know that when adults look at faces, there tends to be greater activity on the right-hemisphere of their brain than the left-hemisphere of their brain. If brain activity increases with right-hemisphere dominance in infants too, we can conclude that they know they are looking at a face.

We discovered in 2013, from conducting brain activity measurements, that when infants of five to six months old look at faces, they use the distinctive features of the eyes to judge that they are looking at a human face. In fact, the contrast of a dark iris in a white sclera (the dark part of the eye on the white part of the eye) is a characteristic unique to humans, not found among any other primates. If this contrast is reversed, the “Blair illusion” (Figure 2) occurs, whereby even a famous person becomes instantly unrecognizable. In addition, the illusion gives such a bizarre impression that the face no longer even appears to be human (Anstis (2005), who discovered this illusion, went so far as to describe it as “vampirish enough to scare little Tory children.”) We showed 13 five to six-month old infants images of faces with normal black on white eye contrast and faces with contrast-reversed eyes created using image-processing software, and measured the brain activity of the infants in response to each. The results were that their brain activity increased when they were looking at faces with normal eye contrast, but did not increase when they were looking at the faces with contrast-reversed eyes. The right occipitotemporal area of the brain showed a particularly high level of brain activity when looking at normal eyes. These infants were able to recognize that only faces with eyes of normal contrast were human faces. The ability of infants to recognize faces develops remarkably in the first eight months of life.

The ability to look at faces and be looked at

In the above we looked at the ability of infants to see and process faces, but I think it can be also be said that infants have the ability to have their own faces “looked at” too. The communication abilities of infants are limited. In particular, since they do not understand language, they cannot express precisely what they want or do not like. However, we do see them suddenly break into a smile at the appearance of a face before them. The British psychoanalyst, John Bowlby, who is famous for originating Attachment Theory, argues that the smiles of infants enhance the motivation of the mother and other persons rearing the child in child-raising. But this is something that anyone who comes into contact with infants understands naturally, and scarcely needs Bowlby to tell them. The smiles of infants encourage and support the family members they live with, as well as the members of our team as we continue to conduct our research into infants.