Human preferences for faces are determined via their association with hormones that metabolize to pheromones, which determine the social preferences of other mammals whose interest in not in facial expression but in chemical cues of nutrient-dependent reproductive fitness.
Those who think that the development of face preferences can can be meaningfully interpreted in the context of only visual cues, place themselves firmly in an ethological world that is known to ignore the importance of olfactory/pheromonal input in some species (e.g. avian species) to further the less relevant study of audiovisual input as a primary influence on behavior. New evidence from a study of the males of one avian species will no doubt be extended to humans during the next decade or so, as more is learned about olfaction in birds. I am, however, impatient to offer it as an example of what is wrong with basic research in the context of avian behavior.
This evidence suggests that social context regulates both sexual signaling (e.g., colorful ornamentation) and testosterone, which is typically associated with more colorful ornamentation in avian species and other species. When testosterone levels were manipulated in a social context (e.g., male competition), changes in colorful orientation and sexual signaling appeared to be the result of increased testosterone.
Levels of sex steroids, like testosterone, metabolize to androgenic “male” pheromones in mammals, and testosterone is associated with darker complexion (a more reddish hue) of human males compared to human females. This represents an extension of testosterone dependent sexual signaling via color that can be found — with variations — in the males and females of all mammals. All of the variations can be directly linked to sex hormones and sex differences in pheromones, which are responsible for the control of reproduction in species from microbes to man.
What this means is that even in species that do not have any colorful ornamentation, the pheromones control reproduction. What that means is the evolutionary theorists cannot, without question, make the assumption that we are primarily visual creatures. The basis for that assumption was a non-existent avian model, one that has gradually gone missing from the literature, only to leave a gaping hole in theories about the development of human behavior that fail to incorporate what’s known about the role of olfactory/pheromonal input in every other species on this planet.