attractive or not? A few years ago, the ethologist Karl Grammer, who teaches in Vienna, was able to contribute to
clarifying these questions with an ambitious study (Gram- mer & Thornhill, 1993). He randomly selected sixteen por- traits of women from the composite computer image used by the police. (fig. 2) By superimposing these digitalized images he was able to compute an average face. (fig. 3)
In an experiment this prototype face was the one selected by most men and women as the most attractive face. It was even seen as being more beautiful than the most beautiful individual face. Since the images were superimposed, all individual features and irregularities vanished. Another factor is that the average face is more symmetric than the individual shots were. Thus the regularity and ‘averageness’ of facial features are what account for beauty. This study also revealed that it is mainly symmetry that determines wheth- er as face is seen as “erotic” or not.
Fig. 2
Fig. 3
“Individual faces” “Average face”
In order to measure facial symmetry certain measurement lines were defined linking symmetric points on both sides. For example, the outer and inner corners of the eyes, the cor- ners of the mouth, the sides of the nose at its broadest point, the two cheek bones and the width of the cheek at the height
of the mouth corners. If a face is completely symmetric, there are no differences in the symmetric axes of these six measurement lines. In this case the axis would be a straight line. It is now possible to obtain a standard for asymmetry by determining the differences between the center points of the measurement line and the axis of sym-
metry and adding them. To make the individual faces comparable the result- ing figures must ultimately refer to the width of the face.
The red line in fig. 4 connects the central points of the measurement lines and shows the deviation from the axis of symmetry. (Grammer, 1994; p. 180)
The Eyes
Fig. 4
The middle-point line
In addition to these standards relating to the face, there is literally a visual standard that has a bearing on attractiveness.
The human pupils appear as a black point in the middle of the colored iris. The openings become larger or smaller depending on the changes in the incidence of light. In glis- tening sunlight they shrink almost to the size of a pin head. Their diameter then totals about two millimetres. At dusk they expand about a fourfold of this size. The size of the pupil, however, is not just influenced by the incidence of light but also by emotional impressions. They can change their size also when the light remains the same. The change in the diameter of the pupil can be likened to a barometer of mood. If we see something that makes us happy or
frightens us, our pupils expand more than would be expect- ed given the prevailing light conditions. If we see some- thing that we don’t like or that puts a damper on us emo- tionally, the pupils contract more. Since this is also largely something that evades our control, the pupil’s size also reflects our real feelings. (Morris, 1978; p. 252)
The pupil signals are not just “emanated” unconsciously. Their reception is also an unconscious process. Even though serious research on the subject of pupil signals has only been pursued ever since the 1950s, this signal had already been conscious- ly manipulated much earlier. Centuries ago the courtesans in Italy had dripped poisonous belladonna in their eyes to widen their pupils. This way they wanted to enhance their beauty. Consequently, this plant was also referred to as Belladonna which in Italian means beautiful woman . Someone was seen as a Belladonna, beautiful woman, when she had other facial fea- tures in addition to the larger pupils: larger eyes, fuller lips, a narrower nose and pronounced cheek bones. The face on the right-hand side in fig. 5, which most people see as more attrac- tive than the left one, was modified by the computer to where it