CONSTANTS ACROSS CULTURES IN THE FACE AND EMOTION ‘
PAUL EKMAN2
University of California, San Francisco
WALLACE V. FRIESEN
Langley Porter Ne/iropsycliialric Institute
This study addresses the question of whether any facial expressions of emotion are universal. Recent studies showing that members of literate cultures associated the same emotion concepts with the same facial behaviors could not demonstrate that at least some facial expressions of emotion are universal; the cultures compared had all been exposed to some of the same mass media presentations of facial expres- sion, and these may have taught the people in each culture to recognize the unique facial expressions of other cultures. To show that members of a preliterate culture who had minimal exposure to literate cultures would associate the same emotion concepts with the same facial behaviors as do members of Western and Eastern literate cultures, data were gathered in New Guinea by telling subjects a story, showing them a set of three faces, and asking them to select the face which showed the emotion appropriate to the story. The results provide evidence in support of the hypothesis that the association between particular facial muscular patterns and discrete emotions is universal.
Prolonged and at times heated controversy has failed to demonstrate whether facial be- haviors associated with emotion are universal for man or specific to each culture. Darwin (1872) postulated universals in facial behavior on the basis of his evolutionary theory. Allport (1924), Asch (1952), and Tomkins (1962, 1963) have also postulated universals in emotional facial behavior, although each writer offered a different theoretical basis for his expectation. The culture-specific view, that facial behaviors become associated with emotion through cul- turally variable learning, received support from Klineberg’s (1938) descriptions of how the facial behaviors described in Chinese literature differed from the facial behaviors associated
1 The authors are indebted to E. Richard Sorenson, National Institute of Neurological Diseases and Blind- ness, for collaboration in planning the early stages of this research, and providing background material rele- vant to understanding of the Fore people. The authors are also grateful to Neville Hoffman, University of Western Australia, for his extraordinary ability to enlist cooperation from subjects and his many useful suggestions about the procedure. The research was supported by Grant AFOSR 1229-67 from the Advanced Research Projects Agency and Career Scientist De- velopment Award 5-K02-MH060 92 from the National Institute of Mental Health. Patricia Garlan and Robert Kleck provided editorial help on the preparation of this report.
2 Requests for reprints should be sent to Paul Ekman, Studies in Nonverbal Behavior, University of Cali- formia, 140S Fourth Avenue, San Francisco, California 94122.
with emotions in the Western world. More recently, Birdwhistell (1963) and LaBarre (1947) have argued against the possibility of any universals in emotional facial behavior, supplying numerous anecdotal examples of variations between cultures.
Ekman (1968) and Ekman and Friesen (1969) considered these contradictory view- points within a framework which distinguished between those elements of facial behavior that are universal and those that are culture specific. They hypothesized that the universals are to be found in the relationship between distinctive patterns of the facial muscles and particular emotions (happiness, sadness, anger, fear, surprise, disgust, interest). They suggested that cultural differences would be seen in some of the stimuli, which through learning become established as elicitors of particular emotions, in the rules for controlling facial behavior in particular social settings, and in man)’ of the consequences of emotional arousal.
To demonstrate the hypothesized universal element, Ekman and Friesen (1969) conducted experiments in which they showed still photo- graphs of faces to people from different cul- tures in order to determine whether the same facial behavior would be judged as the same emotion, regardless of the observers’ culture. The faces were selected on the basis of their conformity to Ekman, Friesen, and Tomkins’s (in press) a priori descriptions of facial muscles
124
CONSTANTS ACROSS CULTURES IN THE FACE AND EMOTION 125
involved in each emotion. College-educated subjects in Brazil, the United States, Argentina, Chile, and Japan were found to identify the same faces with the same emotion words, as were members of two preliterate cultures who had extensive contact with Western cultures (the Sadong of Borneo and the Fore of New Guinea), although the latter results were not as strong (Ekman, Sorenson, & Friesen, 1969). Izard (1968,1969), working independently with his own set of faces, obtained comparable re- sults across seven other culture-language groups.
While these investigators interpreted their results as evidence of universals in facial be- havior, their interpretation was open to argument; because all the cultures they com- pared had exposure to some of the same mass media portrayals of facial behavior, members of these cultures might have learned to recognize the same set of conventions, or become familiar with each other’s different facial behavior.
To overcome this difficulty in the interpreta- tion of previous results, it is necessary to demonstrate that cultures which have had minimal visual contact with literate cultures show similarity to these cultures in their interpretation of facial behavior. The purpose of this paper was to test the hypothesis that members of a preliterate culture who had been selected to insure maximum visual isolation from literate cultures will identify the same emotion concepts with the same faces as do members of literate Western and Eastern cultures.
METHOD Subjects
Members of the Fore linguistic-cultural group of the South East Highlands of New Guinea were studied. Until 12 years ago, this was an isolated, Neolithic, material culture (Gajdusek, 1963; Sorenson & Gaj- dusek, 1966). While many of these people now have had extensive contact with missionaries, government workers, traders, and United States scientists, some have had little such contact. Only subjects who met criteria established to screen out all but those who had minimal opportunity to learn to imitate or recognize uniquely Western facial behaviors were recruited for this experi- ment. These criteria made it quite unlikely that sub- jects could have so completely learned some foreign set of facial expressions of emotion that their judgments would be no different from those of members of literate cultures. Those selected had seen no movies, neither
spoke nor understood English or Pidgin, had not lived in any of the Western settlement or government towns, and had never worked for a Caucasian (according to their own report). One-hundred and eighty-nine adults and 130 children, male and female, met these criteria. This sample comprises about 3% of the members of this culture.
In addition to data gathered from these more visually isolated members of the South Fore, data were also collected on members of this culture who had had the most contact with Westerners. These subjects all spoke English, had seen movies, lived in a Western settlement or government town, and had attended a missionary or government school for more than 1 year. Twenty-three male adults, but no females, met these criteria.
Judgment Task In a pilot study conducted 1 year earlier with mem-
bers of this same culture, a number of different judg- ment tasks were tried. The least Westernized subjects could not be asked to select from a printed list of emotion terms the one that was appropriate for a photograph, since they could not read. When the list was repeated to them with each photograph, they seemed to have difficulty remembering the list. Further, doubts remained about whether the meaning of a particular emotion concept was adequately conveyed by translating a single English word into a single South Fore word. Asking the subject to make up his own story about the emotions shown in a picture was not much more successful, although the problems were different. Subjects regarded this as a very difficult task, repeated probes were necessary, and as the procedure became lengthy, subjects became reluctant.
To solve these problems, it was decided to employ a task similar to that developed by Dashiell (1927) for use with young children.3 Dashiell showed the child a group of three pictures simultaneously, read a story, and told the child to point to the picture in which the person’s face showed the emotion described in the story. The advantages of this judgment task in a pre- literate culture are that (a) the translator recounts well-rehearsed stories which can be recorded and checked for accurate translation; (i) the task involves no read- ing; (c) the subject does not have to remember a list of emotion terms; (d) the subject need not speak, but can point to give his answer; and (e) perfect translation of emotion words is not required since the story can help provide connotations.
Emotion Stones With the exception of the stories for fear and surprise,
those used in the present study were selected from those which had been most freqxiently given in the pilot study. Considerable care was taken to insure that each story selected was relevant to only one emotion within the Fore culture, and that members of the culture were agreed on what that emotion was. Since the stories told
3 Carrol E. Izard brought Dashiell’s procedure to our attention. This method has also been used in recent studies of referential communications (e.g., Rosenberg & Gordon, 1968).
126 PAUL EICMAN AND WALLACE V. FEIESEN
by the pilot subjects for fear and surprise did not meet these criteria, the authors composed stories for these emotions based on their experience within the culture. The stories used are given below:
Happiness: His (her) friends have come, and lie (she) is happy.
Sadness: His (her) child (mother) has died, and he (she) feels very sad.
Anger: He (she) is angr)-; or he (she) is angry, about to fight.
Surprise: He (she) is just now looking at something new and unexpected.
Disgust: He (she) is looking at something he (she) dislikes; or He (she) is looking at something which smells bad.
Fear: He (she) is sitting in his (her) house all alone, and there is no one else in the village. There is no knife, axe, or bow and arrow in the house. A wild pig is standing in the door of the house, and the man (woman) is looking at the pig and is very afraid of it. The pig has been standing in the doorway for a few minutes, and the person is looking at it very afraid, and the pig won’t move away from the door, and he (she) is afraid the pig will bite him (her).4
Pictures and Emotions The six emotions studied were those which had been
found by more than one investigator to be discriminable within any one literate culture (cf. Ekman, Friesen, & Ellsworth, in press, for a review of findings). The photo- graphs used to show the facial behavior for each of the six emotions had been judged by more than 70% of the observers in studies of more than one literate culture as showing that emotion. The sample included pictures of both posed and spontaneous behavior used by Ekman and Friesen (1968), Frijda (1968), Frois-Wittmann (1930), Izard (1968), Engen, Levy, and Schlosberg (1957), and Tomkins and McCarter (1964). A total of 40 pictures were used of 24 different stimulus persons, male and female, adult and child. The photographs were prepared as 3 X 5 inch prints, cropped to show only the face and neck.
Story-Photographs Trial A single item consisted of an emotion story, a correct
photograph, in which the facial behavior shown in the photograph was the same as that described in the story, and either one or two incorrect photograph (s). Adult subjects were given two incorrect pictures with each correct picture; children were given only one be- cause of a shortage of copies of the stimuli.
Because of a limitation on the number of available photographs, and upon the subjects’ time, not all of the possible pairings of correct and incorrect photo- graphs were tested. Instead, the subjects were presented with some of the presumably more difficult discrimi- nations among emotions. The emotion shown in at least one of the incorrect photographs was an emotion which past studies in literate cultures had found to be most often mistaken for the correct emotion. For example, when anger was the emotion described in the
* The fear story had to be long in order to eliminate possibilities for anger or surprise being associated with the story.
story, the incorrect choices included disgust, fear, or sadness, emotions which have been found to be often mistaken for anger. The age and sex of the stimulus persons shown in the correct and incorrect photographs were held constant within any trial.
No one subject was given all the emotion discrimina- tions, because again the stimuli would have been too few and the task too long. Instead, subjects from differ- ent villages were required to make some of the same and some different discriminations. Subjects were shown from 6 to 12 sets of photographs, but no picture ap- peared in more than 1 of the sets shown to any one particular subject.5 A subject’s task included making at least three different emotion discriminations; the same story was told more than once, with differing correct and incorrect photographs, and often requiring discrimination among differing sets of emotions. For example, the anger story might have been read once with Anger Picture A, Sadness Picture B, and Fear Picture C; the same anger story might have been read again to the same subject, but now with Anger Picture D, Disgust Picture E, and Surprise Picture F.
Procedure
Two-person teams conducted the experiment. A member of the South Fore tribe recruited subjects, explained the task, and read the translated stories; a Caucasian recorded the subjects’ responses. Three such teams operated at once within a village; one team with a male Caucasian worked with male adult subjects; the two others with female Caucasians worked with the female adult subjects and the children. In most in- stances, almost all members of a village participated in the experiment within less than 3 hours.
Considerable practice and explanation was given to the translators. They were told that there was no correct response and were discouraged from prompting. Re- peated practice was given to insure that the translators always repeated the stories in the same way and resisted the temptation to embellish. Spot checks with tape re- cordings and back translations verified that this was successful. The Caucasians, who did know the correct responses, averted their faces from the view of the subject, looking down at their recording booklet, to reduce the probability of an unwitting experimenter bias effect. Data analysis did not reveal any systematic differences in the responses obtained with different translators.
RESULTS
No differences between male and female subjects were expected, and no such differences had been found in the literate culture data. In this New Guinea group, however, the women
6 The number of sets of photographs shown varied among villages, because a limited number of photo- graphs were available in this field setting; the need to assure that the three pictures in any one set were com- parable (in terms of the configuration of the mouth, the tilt of the head, and the age of the stimulus persons) restricted the number of sets which could be composed for some of the combinations.
CONSTANTS ACROSS CULTURES IN THE FACE AND EMOTION 127
were more reluctant to participate In the ex- periment, and were considered by most out- siders to have had less contact with Caucasians than the men. The number of correct responses for each subject was calculated separately for males and females and for adults and children. The t tests were not significant; the trend was in the direction of better performance by women and girls. The data revealed no syste- matic differences between male and female subjects in the discrimination of particular emotions, or in relation to the sex of the stimulus person shown on the photographs. In the subsequent analyses, data from males and females were combined.
Table 1 shows the results for the least Westernized adults for each emotion dis- crimination. Within each row, the percentage of subjects who gave the correct response for a particular discrimination between three emo- tions was calculated across all subjects shown that particular discrimination, regardless of whether the photographs used to represent the three emotions differed for individual subjects. Within each row, each subject contributed only one response, and thus the sum of responses was derived from independent subjects. However, the rows are not independent of each other. Data from a given subject appear in different rows, depending upon the particular dis-
TABLE 1 ADULT RESULTS
TABLE 2 RESULTS POR
Emotion described in the story
Happiness
Anger
Sadness
Disgust (smell story) Disgust (dislike story) Surprise
Fear
Emotions shown in the two incorrect
photographs
Surprise, disgust Surprise, sadness Fear, anger Disgust, anger Sadness, surprise Disgust, surprise Fear, sadness Anger, fear Anger, surprise Anger, happiness Anger, disgust Disgust, surprise Sadness, surprise Sadness, surprise Fear, disgust Happiness, anger Anger, disgust Sadness, disgust Anger, happiness Disgust, happiness Surprise, happiness Surprise, disgust Surprise, sadness
No.
62 57 65 36 66 31 31 64 26 31 35 35 65 36 31 31 92 31 35 26 65 31 57
choosing correct
face
90** 93** 86**
100** 82** 87** 87** 81** 81** 87** 69* 77** 77** 89** 71* 65* 64** 87** 86** 85** 48 52 28»
Emotion described in the story
Happiness
Anger Sadness
Disgust (smell story) Disgust (dislike story) Surprise
Fear
Emotion shown in the one
incorrect photograph
Surprise Sadness Anger Disgust Sadness Anger Surprise Disgust Fear Sadness Sadness Happiness Disgust Fear Sadness Anger Disgust
No. 5s
l ie 25 25 25 69 60 33 27 25 19 27 14 14 19 25 25 H
% choosing
the correct
face
87* 96*
100* 88* 90* 85* 76* 89* 76* 95* 78*
100* 100* 95* 92* 88*
100*
* p < .05. ** p < .01. "Subjects selected the surprise face (67%) at a significant level (p < .01, two-tailed test). criminations he was asked to make. If a group of subjects was requested to discriminate the same emotion from the same two other emo- tions more than once, only one randomly chosen response was included in the table. A binomial test of significance assuming chance performance to be one in three showed that the correct face was chosen at a significant level for all of the discriminations (rows) except that of fear from surprise. Twice, fear was not discriminated from surprise, and once surprise was chosen more often than fear, even though the story had been intended to describe fear. A binomial test assuming chance to be one in two (a more conservative test, justified if it was thought that within a set of three pictures, there may have been one which was obviously wrong) still yielded significant correct choices for all but the fear-from-surprise discriminations. The results for the most Westernized male adults were almost exactly the same as those reported in Table 1 for the least Westernized male and female adults. The number of correct responses for each subject was calculated; the t test showed no significant difference between the most and least Westernized subjects. Again, the only failure to select the correct picture occurred when fear was to be dis- tinguished from surprise. Table 2 shows the results for the children, tabulated and tested in similar fashion. The children selected the correct face for all of their discriminations. Through an oversight, the one discrimination which the adults could not 128 PAUL EKMAN AND WALLACE V, FRIESEN make, fear from surprise, was not tried with the children. The percentages reported in Table 2 are generally higher than those in Table 1, but this is probably due to the fact that the children were given two photographs rather than three, and chance performance would be 50% rather than about 33%. Six- and 7-year-old children were compared with 14- and 15-year-olds, by the same procedures as described for comparing males and females. No significant differences or trends were noted. DISCUSSION The results for both adults and children clearly support our hypothesis that particular facial behaviors are universally associated with particular emotions. With but one exception, the faces judged in literate cultures as showing particular emotions were comparably judged by people from a preliterate culture who had minimal opportunity to have learned to recognize uniquely Western facial expressions. Further evidence was obtained in another experiment, in which the facial behavior of these New Guineans was accurately recognized by members of a literate culture. In that study, visually isolated members of the South Fore posed emotions, and college students in the United States accurately judged the emotion intended from their videotaped facial behavior. The evidence from both studies contradicts the view that all facial behavior associated with emotion is culture specific, and that posed facial behavior is a unique set of culture-bound conventions not understandable to members of another culture.6 The only way to dismiss the evidence from both the judgment and posing studies would be to claim that even these New Guineans who had not seen movies, who did not speak or understand English or Pidgin, who had never worked for a Caucasian, still had some contact with Westerners, sufficient contact for them to 6 If posed behavior were simply a set of arbitrary conventions, it would be unlikely that the same con- ventions would be utilized in the cultures discussed here. That does not, however, imply that posed facial behavior is identical with spontaneous bshavior. Ekman, Friesen, and Ellsworth (in press) have suggested that most posed behavior is similar in appearance to that spontaneous facial behavior which is of extreme intensity and unmodulated, although it may still differ in onset, duration, and decay time. learn to recognize and simulate culture-specific, uniquely Western facial behaviors associated with each emotion. While these subjects had some contact with Westerners, this argument seems implausible for three reasons. First, the criteria for selecting these subjects makes it highly improbable that they had learned a "foreign" set of facial behaviors to such a degree that they could not only recognize them, but also display them as well as those to whom the behaviors were native. Second, contact with Caucasians did not seem to have much influence on the judgment of emotion, since the most Westernized subjects did no better than the least Westernized and, like the latter, failed to distinguish fear from surprise. Third, the women, who commonly have even less contact with Westerners than the men, did as well in recognizing emotions. The hypothesis that there are constants across cultures in emotional facial behavior is further supported by Eibl-Eibesfeldt's (1970) films of facial behavior occurring within its natural context in a number of preliterate cultures. Evidence of constants in facial be- havior and emotion across cultures is also consistent with early studies which showed man)' similarities between the facial behavior of blind and sighted children (Fulcher, 1942; Goodenough, 1932; Thompson, 1941). Uni- versals in facial behavior associated with emotion can be explained from a number of nonexclusive viewpoints as being due to evolu- tion, innate neural programs, or learning experiences common to human development regardless of culture (e.g., those of Allport, 1924; Asch, 1952; Darwin, 1872; Huber, 1931; Izard, 1969; Peiper, 1963; Tomkins, 1962, 1963). To evaluate the different viewpoints will require further research, particularly on early development. The failure of the New Guinean adults to discriminate fear from surprise, while succeed- ing in discriminating surprise from fear, and fear from other emotions, suggests that cultures may not make all of the same distinctions among emotions, but does not detract from the main rinding that most of the distinctions were made across cultures. Experience within a culture, the kinds of events which typically elicit particular emotions, may act to influence the ability to discriminate particular pairs of CONSTANTS ACROSS CULTURES IN THE FACE AND EMOTION 129 emotions. Fear faces may not have been distinguished from surprise faces, because in this culture fearful events are almost always also surprising; that is, the sudden appearance of a hostile member of another village, the unexpected meeting of a ghost or sorcerer, etc. The growing body of evidence of a pan- cultural element in emotional facial behavior does not imply the absence of cultural differ- ences in the face and emotion. Ekman (1968) and Ekman and Friesen (1969) have suggested that cultural differences will be manifest in the circumstances which elicit an emotion, in the action consequences of an emotion, and in the display rules which govern the management of facial behavior in particular social settings, Izard (1969) agrees with the view that there are cultural differences in the antecedent and consequent events, and has also found evidence suggesting differences in attitudes about particular emotions. REFERENCES ALLPORT, F. H. Social psychology. Boston: Houghton Mifuin, 1924. ASCII, S. E. Social psychology. Englewood Cliffs, N. J.: Prentice-Hall, 19S2, BIRDWHISTELL, R. L, The kinesic level in the investi- gation of the emotions. In P. H. Knapp (Ed.), Expression of tlie emotions in man. New York: In- ternational Universities Press, 1963, DARWIN, C, The expression of the emotions in man and animals. London: Murray, 1872. DASHIELL, J. F. A new method of measuring reactions to facial expression of emotion. Psychological Bulletin, 1927, 24, 174-175. EiBL-EiBESFELDi, I. Ethology, tile biology of behavior. New York: Holt, Rinehart & Winston, 1970. EKMAN, P. Research findings on recognition and display of facial behavior in literate and nonliterate cultures. Proceedings of the 76th Annual Convention of the American Psychological Association, 1968, 3, 727. (Summary) EKUAN, P., & FKIESEN, W. V. Nonverbal behavior in psychotherapy research. In J. Slilien (Ed.), Research in psychotherapy. Vol. 3. Washington, D. C.: Ameri- can Psychological Association, 1968. EKMAN, P., & FKIESEN, W. V. The repertoire of non- verbal behaviorCategories, origins, usage and coding. Scmiotica, 1969, 1, 49-98. EKMAN, P., FRIESEN, W. V., & ELLSWORTH, P. Emotion in the human face: Guidelines for research and integra- tion of findings. New York: Pergamon Press, in press. EKMAN, P., FKIESEN, W. V., & TOMKINS, S. S. Facial affect scoring technique: A first validity study. Sewiotica, in press, EKMAN, P., SORENSON, E. R., & FRIESEN, W. V. Pan-cultural elements in facial displays of emotions. Science, 1969, 164, 86-88. ENGEN, T., LEVY, N., & SCIILOSBERG, H. A new series of facial expressions. American Psychologist, 1957, 12, 264-266. FRIJDA, N. H, Recognition of emotion. In L. Berko- witz (Ed.), Advances in experimental social psychology. New York: Academic Press, 1968. FROIS-WITTJIANN, J. The judgment of facial expression. Journal of Experimental Psychology, 1930, 13, 113-151. FULCHER, J. S. "Voluntary" facial expression in blind and seeing children. Archives of Psychology, 1942, 38, 272. GAJDUSEK, D. C. Kuril. Transactions of the Royal Society of Tropical Medicine and Hygiene, 1963, 57, 151-169. GOODENOUGII, F. L. Expression of the emotions in a blind-deaf child. Journal of Abnormal and Social Psychology, 1932, 27, 328-333. HUBEB, E. Evolution of facial musculature and facial expression. Baltimore: Johns Hopkins Press, 1931. IZARD, C. E. Cross-cultural research findings on de- velopment in recognition of facial behavior. Pro- ceedings of the 76tli Annual Convention of the A merican Psychological Association, 1968, 3, 727. (Summary) IZARD, C. E. The emotions and emotion constructs in personality and culture research. In R. B. Cattell (Ed.), Handbook of modern personality theory. Chicago: Aldine Press, 1969. KLINEBERG, 0. Emotional expression in Chinese literature. Journal of Abnormal and Social Psy- chology, 1938, 33, 517-520. LABARRE, W. The cultural basis of emotions and gestures. Journal of Personality, 1947, 16, 49-68. PEIPER, A. Cerebral function in infancy and childhood. New York: Consultants Bureau, 1963. ROSENBERG, S., & GORDON, A. Identification of facial expressions from affective descriptions: A proba- bilistic choice analysis of referential ambiguity. Journal of Personality and Social Psychology, 1968, 10, 157-166. SORENSON, E. R., & GAJDUSEK, D. C. The study of child behavior and development in primitive cultures. A research archive for ethnopediatric film investi- gations of styles in the patterning of the nervous system, Pediatrics, 1966, 37(1, Ft. 2). THOMPSON, J. Development of facial expression of emotion in blind and seeing children. Archives of Psychology, 1941, 37, 264. TOMKINS, S. S. A/ect, imagery, consciousness. Vol. 1. The positive affects. New York: Springer, 1962, TOMKINS, S. S. Affect, imagery, consciousness. Vol. 2. The negative afects. New York: Read more Applied Sciences Architecture and Design Biology Business & Finance Chemistry Computer Science Geography Geology Education Engineering English Environmental science Spanish Government History Human Resource Management Information Systems Law Literature Mathematics Nursing Physics Political Science Psychology Reading Science Social Science Home Homework Answers Blog Archive Tags Reviews Contact twitterfacebook Copyright © 2022 SweetStudy.com
