200 million years to convey what words cannot
Everyone knows instinctively, that human facial expressions convey something about feelings. If, for instance, a friend comes up to you with a smile, you are instantly prepared to hear some pleasant news but if he is frowning, you brace yourself to hear some unhappy comment. Most expressions, however, are fleeting. Each one typically lasts between a fraction of a second and several seconds, and is then replaced by another or by a neutral face (itself an expression). If there are words spoken, it is the words one usually remembers, not the expression. How significant then are facial expressions? Are they the equivalent of Hallmark greeting cards: pleasant to see but hard to remember a minute later? In this article, I will argue that they are much more interesting and important than that.
In thinking about facial expressions, it helps to know something about their scientific history. They were first brought to general attention in 1872 with the publication of a book. The author was Charles Darwin and the book was titled The Expression of Emotions in Man and the Animals. It was a big book, one in which Darwin had clearly invested time, thought and energy. Darwin was already a famous man, and this book was not intended to make him more famous. What was the motivation behind it?
Thirteen years previously, he had shot to fame with the publication of his book on evolution, whose long title is typically abbreviated to The Origin of Species (or even The Origin). Evolution, of course, is a major topic, relevant to everything in biology, and Darwin’s book made a big splash. Why then was he, more than a decade later, spending time on a seemingly much smaller topic, namely facial expressions in humans and (non-human) animals? In fact, it was part of a much larger program that his evolution book had made necessary. In The Origin, he had included only as much evidence for evolution and his proposed mechanism (natural selection) to make the whole argument plausible but it was only a fraction of what he knew and believed relevant. At over 500 pages, The Origin is not a small book but it had so much less content than Darwin felt it should contain that he called it, without irony or false modesty, “an abstract”.
Since the first germ of his ideas about evolution came to him in 1837, twenty years earlier, he had been planning a massive multi-volume book on evolution, and he had been accumulating evidence and constructing his conceptual framework, to be buttressed by a mountain of evidence. In June, 1858, however, he had been jolted by receiving a letter and an accompanying article from a fellow naturalist, Alfred Wallace. Wallace had come up with the same basic idea to explain evolution and wanted Darwin’s opinion on it. He had no idea that Darwin was pursuing the same ambitious goal, to give an evolutionary explanation of living things. Darwin, fearing that he had been “scooped” asked two powerful colleagues what could be done. They set up an event at which there would be a joint reading of Wallace’s paper and one that Darwin would write. That took place in August at the Linnaean Society in London.
Darwin, however, knew that that paper would not be sufficient to get serious recognition for his idea: he needed to present a much fuller account. Yet, to get it out fast, it would have to be a smaller book than the planned multi-tome opus. Under that pressure, he wrote his “abstract”, The Origin, in a year. That indeed gave him priority in the eyes of his colleagues and, incidentally, fame with the general reading public. Yet it was only a fraction of what he wanted to present, to persuade his colleagues that his idea was right. Instead of the multi-volume opus, he began a series of books, each to deal with a key issue neglected in the The Origin, which would be more effective.1
Two of the big gaps in The Origin were A) the mechanism of heredity, and B) the evolution of human beings. For the first, he had said almost nothing because he had only begun to work out his ideas about it, while on human evolution, there was precisely one sentence, which simply promised the reader that it would be returned to later. The later exposition of heredity, appearing more than a decade later, became his two-volume opus, Variation in Domesticated Plants and Animals (1872) and his treatment of human evolution, appearing near the same time (1871), The Descent of Man (its abbreviated title).
It was in this program of fleshing out his evolutionary ideas that The Expression of the Emotions in Man and the Animals came into being. Unlike the books on heredity and human evolution, which were crucial for Darwin’s theory of evolution, this work on facial expressions was a reply to and refutation of a major claim by a noted biologist in a book first published long before Darwin’s Origin of Species. The author was a noted British physiologist, Sir Charles Bell (1774-1842) and his book was about facial expressions.
Its title was Essays on the Anatomy of Expression in Painting. The first edition appeared in 1806 but Darwin apparently read the 1844 edition, which came out after Bell’s death. As the title suggests, Bell was focussing on how expressions of the face, as depicted in paintings, were made. He drew attention to the special set of muscles in the face that attach to the skin and the skull and which make possible the movements of the forehead, eyes, nose and mouth. These were the so called “mimetic muscles”. Bell’s contention was that these muscles were unique to human beings and that they had been the special gift of God to humanity. Bell’s beliefs, of course, reflected his time, well before there was a theory of evolution.
Yet the idea that the properties of living things were a product of divine creation was the antithesis of Darwin’s view. Indeed the purpose of The Origin had been to refute it. Furthermore, Darwin had spent enough time at the London Zoo, observing the animals, to know that other primates, in particular orang-utans, were capable of diverse facial expressions. Even cats and dogs, though having a smaller range of such, had facial expressions. Finally, there had been dissections of the faces of various dead mammals and these had shown directly that they too had mimetic muscles. Bell’s thesis that they were unique to humans was clearly false. For Darwin, therefore, the focus of the subject became the evolutionary “why”: why did these muscles in the face exist? His answer was that they allowed the animals to communicate their feelings to others, especially members of their own species. This is what Darwin’s The Expression of the Emotions set out to argue and explore. (More than 170 years after its publication, by the way, it is still an enjoyable and informative book.)
Yet, for fully a century after it appeared, that is largely where the subject stood, with little subsequent work by others. Yet, it was ripe for further exploration. The implicit prediction of Darwin’s idea was that particular facial expressions involved the contraction of particular mimetic muscles. There should be specific muscles connected to specific expressions. And this is what several scientists, especially the psychologist Paul Ekman, set out to investigate.
There are 21 distinct kinds of mimetic muscle in the human face, the great majority in pairs to match the bilateral symmetry of the face, and Ekman and his colleagues, were able to show that particular expressions were correlated with the contraction of specific facial muscles. Some expressions involve only a single mimetic muscle while many are associated with the contraction of several of these muscles. Altogether the set of detectable expressions correlated with 33 mimetic muscle contractions; each mimetic muscle-expression association is termed an “action unit” or AU. There are also a further set of head or eye movements that can contribute to the expressions, which are powered by a deeper set of muscles, termed “action descriptors” or AD. The entire system has been named the “Facial Action Coding System” or FACS and it has become the standard means of measuring and systematizing facial expressions in humans. With such a system, one can take the topic of facial expressions out of the realm of qualitative impressions, story-telling, anecdotes and inference, into a more quantitative and precise description.2
One major long-standing question to be addressed was: are human expressions universal, one shared by all people, regardless of culture, ethnicity, sex, and age? Darwin had in fact gone into this and his provisional answer was “yes” but it was clear, even then, that the matter was complicated. Take the matter of age: it is clear that babies produce far fewer expressions than adults and probably respond to fewer of the more subtle expressions. Here, we should note that human facial expressions can be divided into “major” expressions, corresponding to the major emotions, and so called “microexpressions”, namely the more subtle and, often, more fleeting ones. Infants can display the major expressions but the microexpressions come along later.
The major expressions correspond to: anger, fear, disgust, sadness, happiness and surprise. Darwin investigated the question of their universality by writing to his various correspondents around the world about how these feelings are expressed in their native lands. Of course this was an imprecise form of research and quite fragmentary but the information that he received indicated that the forms of expression of these big emotions were basically the same across cultures and ethnicities, from bushmen in Africa to London or New York sophisticates. Furthermore, individuals blind from birth also make these facial expressions in accordance with their emotions, as judged from things they are saying or the context. Hence, the major expressions corresponding to these six major feelings seem to be shared by all people.
BUT, as modern research has shown, different cultures can modify the form of these expressions. This is not surprising. The expressions are under the neurological control of both the basal nuclei in the brain, which control spontaneous reactions, and the cortex, which supplies a degree of voluntary control over what the face shows. With some degree of cortical and voluntary control, there is bound to be some influence of culture on the mode of expression. For instance, several investigators who have looked into this report that, amongst native Japanese, muscles around the mouth are more important in modulating expressions while in Europeans, the eyes are more important. This is almost certainly a cultural effect rather than a genetic one. Hence the degree of universality, and the role of culture, has been the basis of some debate in this field.
Microexpressions are underlain by more automatic neural circuitry and probably show less influence from culture but they are harder to study. They typically last a half second or less and are involuntary and often betray a feeling that the person may want to hide. As far as I know, there has not been a quantitative study of the form of microexpressions, using FACS, across cultures but perhaps this is underway now. Because of the transience of these expressions, high speed photographs of people experiencing some form of cognitive dissonance are necessary but that is perfectly feasible. A trickier question might be the ethics of such experiments, which would involve inducing some degree of discomfort in the subjects taking part in this research.
Let us return to the evolutionary questions. By Darwin’s time, it was clear that the mimetic muscles were not the exclusive property of human beings. Indeed, modern research shows that the mimetic muscles are found in all mammals and therefore the potential for facial expressiveness is universal. Even platypuses, with their bills and rigid mouth, have mimetic muscles. Nevertheless, the set of these muscles is not identical amongst mammals but many mammals “below” the great apes, contain fewer than humans. Indeed, there is a rough correlation between numbers of the mimetic muscles and expressiveness in various mammals. In that connection, it is worth noting that chimpanzees, our closest animal cousins, have the same set of mimetic muscles as humans and comparable major expressions. (We can judge those expressions as corresponding to comparable feelings in humans from the context of the associated behaviours.)
What human faces have, however, is bare skin, unlike most mammals. This makes it much easier to detect microexpressions in our fellow humans than it would be if our faces were covered in fur. The great apes, namely chimpanzees, bonobos, gorillas and orang-utans, also have more bare skin on their faces and therefore the potential to convey more information in their expressions to their conspecifics than if they did not.
FACS is now being applied to other mammals. Unsurprisingly, the findings show a great range of facial expressiveness. The implication is important: many mammals almost certainly pay more attention to the faces of their conspecifics than had been believed, especially for clues as to how they are feeling about something. Of course, we have always known that other animals pay attention to each other but it is now a good bet that, to some degree, they are attuned to some extent to each other’s feelings through their facial expressions, something that had not been appreciated.
The animals for which FACS has been applied so far are four primate species and three domesticated animals. The primates are chimpanzees, orangutans, gibbons, and rhesus macaques, while the non-primates are dogs, cats, and horses. For all of these species, the anatomy of their mimetic muscles is known, and it is possible to correlate the muscles involved with the expressions, as required in FACS. (The context then allows one to infer the corresponding emotions.) The data are now voluminous but suffice it to say that the findings on these animals are fully concordant with what has been learned via FACS about human expression-making. One suspects that Darwin would have been tremendously pleased to have this confirmation of his thinking.
Two last points I would make here, the first about evolution, the second about significance. The evolutionary question is: was there a selection pressure to promote the mimetic muscles and if so, how did it work? One can answer this only partially but the answer suggests something interesting.
In general, if a physical feature persists over hundreds or thousands of generations or more, it is doing something useful. The reason is that it takes energy to make biological structures and if a particular structure exists without doing something for the organism, it becomes just a cost without a benefit, and it tends to be lost. Hence, the mimetic muscles and their underlying neural circuitry, which go back to the evolutionary origin of the mammals, 200-150 million years ago, must be doing something of value. And that something must be facial expression. In contrast to properties that have a direct value for helping the animal to survive, such as obtaining a certain kind of food or running faster or sensing danger more efficiently or warding off infection, facial expressions only promote social interactions and those must be of indirect benefit to survival. In this case, it would be in building social bonds.
Is there any evidence for selection of a mimetic muscle for its expressive capacity? Not much but some. Dogs, who are the domesticated version of wolves, have a special muscle near the eyes that makes eyes look rounder and more puppy-like. We know that such juvenile states have appeal for us humans. Our domestic dogs have a kind of baby-like appeal for us and their round eyes are part of that. In wolves, this muscle is greatly reduced, hardly there really, and cannot give this effect. If this thinking is right, we are talking about an indirect and social benefit from a physical change in a mimetic muscle.3
Such indirect benefits are, by their nature, harder to quantify and the population genetics that would be needed for them hardly exist. If the reasoning about the value of facial expressions is valid, however, the population genetics for such indirect benefits to be selected needs to be developed.
Finally, let us return to the question of significance of facial expressions. Is their significance as fleeting as the expressions themselves are? I would say definitely “no”. They are an important aspect of animal communication and probably always have been. Individually, each expression is transient and of short-lived significance but collectively, they help create and strengthen social bonds within an animal group, probably far more than has been suspected.
In humans, they are particularly important in face-to-face communication in speaking. We tend to believe that the words we use in communicating with someone convey the important meaning. That, however, is only part of what is going on. Speaking directly to someone, face-to-face, the facial expressions convey the emotional sub-text of what we are saying. Those feelings are just as, or even more important, than the meaning of the words. Why, for instance, is email so less effective than directly speaking to one another? It is because that emotional subtext cannot be conveyed by the words alone. That is why we put in emojis, a poor substitute for actual expressions, but better than nothing.
Hence, there is a practical message that comes out of this subject: if you have something important to say to someone, do not send it by email! Say it directly to them so that they can see your face, with its accompanying facial expressions. Think of such direct, face-to-face speech as drawing on the benefits of 200 million years of mammalian evolution.
Supplementary reading:
Ekman, P., Friesen, W.V., Hager, J.C. (2002). Facial Action Coding System. Research Nexus, Salt Lake City.
A book-length, hence full, account of FACS, though of course with references to the earlier work.
A.S. Wilkins (2017). Making Faces: the evolutionary origins of the human face. Belknap Press of Harvard University Press, Cambridge.
This book provides an account of evolutionary history of the human face, in all its aspects. Chapter 7 provides a detailed discussion of human facial expressions.
Waller, B.M., Julle-Daniere, E., Micheletta, J. (2020). Measuring the evolution of facial expression using multi-species FACS. Neurosci. Behav. Rev. 113: 1-11.
A very good overview of how FACS has been applied to various mammalian species and what one can, and cannot, learn from such measurements. The authors take the interesting point of view that one should not speak of facial “expressions”, which implies knowledge of the underlying feelings, but only of facial “behaviours”.
Burrows, A. M. (2008). The facial expression musculature in primates and its evolutionary significance. BioEssays 30: 212-225.
A good early review of the nature and significance of facial expressions in primates.
Darwin did not give up writing about evolution generally, however. He wrote successive editions of The Origin of Species, for a total of six. Each one was larger, having more material, and probably less effective than the preceding edition. It is, I suspect, the first edition, the shortest, which is still the one most often read and cited.
The successive development of Ekman’s methodologies, including FACS, was not without its controversies, as one might expect. The Wikipedia article on facial expressions gives a good overview of some of these disputes.
For this study, see Kaminski, J. et al. (2019). Evolution of facial anatomy in dogs. Proc. Natl Acad. Sci. doi.org/10.1073/pnas.1820653116.