SOCIAL DETERMINANTS OF FACIAL DISPLAYS Nicole Chovil
ABSTRACT: A primary function of facial displays is to communicate messages to others. Bavelas and Chovil (I 990) proposed an Integrated Message Model of language in which nonverbal acts such as facial displays and gestures that occur in communicative (particularly face-to-face) interactions are viewed as symbolic messages that are used to convey meaning to others. One proposition of this model is that these nonverbal messages will be shaped by the social components of the situation. The present study attempted to delineate more precisely the components of sociality that explicitly affect the use of facial displays in social situations. Frequency of motor mimicry displays in response to hearing about a close-ca!l experience was examined in four communicative situations. In one condition, participants listened to a tape-recording of an individual telling about a close-call event. In two interactive but nonvisual conditions, participants listened to another person over the telephone or in the same room but separated by a partition. In the fourth condition, participants listened to another person in a face-to-face interaction. The frequency of listeners' motor mimicry displays was found to vary monotonically with the sociality of the four conditions. Actual presence and visual availability of the story-teller potentiated listener displays. The results support the proposition that facial displays are mediated by the extent to which individuals can fully interact in communicative situations.
Social Determinants of Facial Displays It has long been rec0gnized that facial displays' are an important means of communication. Their use by both infants and adults to infer information This study was conducted at the University of Victoria, Victoria, B.C., Canada, as part of the author's doctoral dissertation and was partly supported by a Social Sciences and Humanities Research Council of Canada doctoral fellowship. I thank Janet Bavelas for her assistance in the development of the study and her valuable comments on earlier drafts. Alan Fridlund helped in the analysis of the data and provided comments on earlier drafts. Linda Coates, Jennifer Halliday, Laura Lane, and Doug Laurie assisted in the collection of data. Laura Lane also helped in the scoring of facial displays. Correspondence and requests for reprints should be addressed to Nicole Chovil, Department of Psychology, University of British Columbia, 2136 West Mall, Vancouver, B.C., Canada, V6T IY7. i For the purposes of this study, the term "facial display" is used rather than "facial expression" because the latter term has offen been equated with expressionsof emotion. We Joumal of Nonverbal Behavior 15(3), Fall 1991 © 1991 Human 5ciences Press. Inc.
| 41
142 IOURNAL OF NONVERBALBEHAVlOR about various emotion states has been widely documented (for a review of this literature, see Ekman, 1982; Fridlund, Ekman, & Oster, 1987). The significance of facial expressions in both caregiver-infant and adult communications has been also amply noted (e.g., Darwin, 187211965; Emde, 1984; Emde, Gaensbauer, & Harmon, 1976; Izard, 1971; Malatesta, 1985; Sorce & Emde, 1981; Tomkins, 1982; Trevarthen, 1984). The present study investigated how the use of facial displays in communicative situations is shaped by the social aspects of the situation. Specifically, the goal was to delineate more precisely the components that make up sociality of communicative situations and to examine their effect on facial displays.
An Integrated Approach to Verbal and Nonverbal Communication Although nonverbal acts such as facial displays are offen acknowledged to have communicative value, the information and process by which we use nonverbal acts (including facial displays) to convey information has been considered distinct from that undertying spoken language. Spoken messages are assumed to be deliberate, intentional communication, whereas nonverbal messages are offen seen as spontaneous readouts of underlying states that can have a secondary communicative role. Dissatisfaction with using separate models of verbal and nonverbal communication has led a number of researchers to advocate a more integrated approach to the study of communication behaviors (e.g., McNeill, 1985; Sanders, 1987; Scherer, 1980). Bavelas and Chovil (1991) proposed an Integrated Message Model (IMM) for understanding communication in face-to-face interaction. They extended the concept of language to include nonverbal acts such as facial displays, gestures and intonation that occur commonly along with spoken utterances. Both verbal and nonverbal acts are viewed as symbolic messages that are directed to others. These acts can be used by themselves to convey messages or integrated with other acts to convey more complex meanings. A second important proposition of the IMM model is that language is a social, interactive process. Verbal and nonverbal messages are used to decided to use the term "facial display" becauseour researchis concernedwith the use of facial displays as communicative messagesto others. We make no assumptionsabout the relationshipof thesedisplaysto underlyingemotion.
143 NICOLECHOVIL communicate with others and therefore will be shaped by the communicative characteristics of the situation.
Sociality There are several theories of facial expression that have accorded importance to social factors in the production of displays. These theories are divided as to whether social factors inhibit or enhance facial display. Ekman and Friesen (1969a; 1969b) proposed that spontaneous facial expressions of emotion are regulated by social factors and that individuals are likely to manage their expressions in the presence of others. Izard (1971) also postulated that individuals are socialized in childhood to inhibit various expressions of emotion: "As the individual moves into later childhood, peers and parents generally discourage the ready display of strong emotions" (p. 192). He discussed how children are taught to control display of various emotions (e.g., laughing, shame, fear, anger, as weil as distress). In contrast to the "social management" position, there is an emerging second view in which facial displays are assumed to be communicative acts that serve social motives (Bavelas, Black, Lemery, Mullett, 1986; Fridlund, 1991; in press); and because they are primarily social acts, they should be facilitated by the presence of others. A number of studies have provided empirical support for this position in that displays are more likely to occur in the presence of others than when alone (Brightman, Segal, Werther, & Steiner, 1975); smiles are directed to others (Kraut & Johnson, 1979; Jones & Raag, 1989; Jones, Collins, & Hong, 1991); and motor mimicry displays are affected by communicative variables such as eye contact (Bavelas, Black, Lemery, & Mullett, 1986). In addition, smiling is potentiated by implicit sociality and social imagery (Fridlund, Sabini, Hedlund, Schaut, Shenker, & Knauer, 1990; Fridlund, 1991).
Components of Sociality In facial display research, empirical definitions of "social" has varied from the use of a social stimulus (e.g., photograph of a person) to the investigation of specific social variables such as orientation or eye contact. The numerous ways sociality has been defined suggest that it is composed of several aspects. Indeed, social situations can differ both qualitatively and quantitatively in their level of sociality. For the purposes of the present study, sociality was defined as "the
144 IOURNALOF NONVERBALBEHAvlOR extent to which individuals can fully interact with each other through the auditory and visual channels of language." Sociality, in this sense can be broken down into a number of specific components, each of which might affect facial displays. Some of the aspects of sociality are: (a) presence of a social stimulus, (b) reciprocity of communication, and (c) accessibility of the visual channel of communication. These components can also be used to place situations on a continuum ranging from Iow sociality to high sociality. Low social situations would be ones in which the other person was not actually present such as reading a letter, listening to a tape-recording, or watching a videotape of another person. Examples of moderate sociality situations are nonvisual interactions with others such as telephone conversations. High social situations are ones in which indidviduals can interact face-to-face using both visual and auditory channels. To investigate the frequency of displays across different levels of sociality, four situations ranging from Iow sociality to high sociality were included in the present study: listening to a tape-recording of another person, interacting with another person in the same room but separated by partition, interacting over the telephone, and interacting face-to-face with another person.
Motor Mimicry Displays Although sociality has been found to affect facial displays, most of the findings were limited to smiles. Another type of display that occurs in social interactions is the motor mimicry display. Individuals often use facial displays to convey messages of "empathic understanding" when hearing of another's misfortune. In the past, motor mimicry was seen as a reflex action that was set oft by the perception of another's experience (for a historical review of motor mimicry see Bavelas, Black, Lemery, & Mullett, 1987). Although a social stimulus played a key role in the production of these displays, theorists such as Hoffman (1982) and Mehrabian and Epstein (1972) hypothesized that they resulted from one's own vicarious emotional experience. Bavelas, Black, Lemery, and Mullett (1986) showed that these displays are communicative messages in that they were strongly affected by probability of eye contact with the "victim", and there was high agreement among individuals as to the message conveyed by these displays. These displays have also been found to occur in conversations about "close-call" or "near miss" experiences (Chovil, 1989).
Hypotheses Previous investigations of sociality and the possibility that facial displays might be affected by each specific component of sociality lead to several
]45
NICOLE CHOVlL
alternative predictions. Based on Fridlund's (1991) finding that facial displays varied monotonically with the sociality of the situation, the first prediction was a general increase in the frequency of facial display across the four commünicativ situations: Tape-recording, Telephone, Partition, and Face-to-face. A second prediction was made based on the interactive component of sociatity. It was hypothesized that facial displays might be more likely to be produced in social situations where there is the possibility of reciprocal communication, regardless of whether individuals can see the other. This would lead to a higher frequency of facial displays in the three interactive conditions in which the speaker can receive communications by the other person (Face-to-face, Partition, and Telephone) than in the condition in which the listener cannot interact back with the speaker (Tape-recording). The final prediction is based on the assumption that visual availability of a receiver is an important factor in determining the likelihood of facial displays. If facial displays are more likely to be exhibited when the other person is Iooking at the displayer, then displays should occur more frequently in the visual condition (Face-to-face) than in the three nonvisual conditions (Tape-recording, Partition, and Telephone). Given the theoretical or empirical support for all of the above predictions, all were considered viable hypotheses.
Method
Participants Ninety women participated in the study. All were University of Victoria undergraduates. Their ages ranged from approximately 17 to 50 years, although most were in their early twenties. They were recruited either through the Psychology Department Volunteer Subject Pool or by canvassing classes. The decision to use only women was based on the complexity and time required in the collection of dyadic data. 2 In addition, previous research suggested that adult males are less likely to produce motor mimicry displays than adult females (Bavelas, Black, Lemery, Maclnnis, & Mullett, 1986). 3 2 Considerable time and effort is required to organize, schedule, and run dyads in experiments; see Solano, 1989 for a discussion of how this may inhibit investigations of interactive processes. Gender effects consist of two kinds: possible sex differences in the frequency of displays produced by male and female same sex dyads, as well as the effect of an opposite sex dyadic partner on displays produced by either a male or female participants. It was not possible to include a large enough sample as to be able to test for these effects. Secondly, because this was an initial study of social factors affecting facial displays in conversations, we wanted to
146 I()URNAL OF NONVERBAL BEItAVI()R There were ten dyads in each of the three social conditions (Face-toface, Partition, and Telephone) and ten subjects in the Tape-recording condition. Dyads were randomly assignecl to the three social conditions in permutations of three. Subjects were assigned to the Tape-recording condition when only one subject could be booked or when only one of a dyad showed up for the study. Five dyads (one Face-to-face, two Partition, and two Telephone) and ten subjects in the Tape-recording condtion had to be replaced because of videotaping problems, not following instructions, or the tendency for subjects in the nonvisual conditions to Iook down which obscured their faces from the camera. 4 The final number of subjects was 10 pairs in each of the three dyadic conditions and 10 subjects in the Taperecording condition.
Conditions Tape-recording condition. The participant was seated at a table that held a tape recorder and a telephone. Telephone condition. Participants were seated in separate rooms and talked to each other over a telphone system hooked up between the two rooms. Partition condition. Participants were seated facing each other at two adjacent tables that were separated by a partition that was approximately live feet (150 cm) high. Face-to-face condition. Participants were seated across from each other at a table.
Stimulus for Tape-recording Condition The close-call story used in the Tape-recording condition was told by a woman who had participated in another study (Chovil, 1989). The ciose elicit as many displaysas we could in order to compare frequenciesacrossdifferent conditions ratherthan test for genderdifferences. 4It was interesting,althoughextremelyfrustrating,to find that subjectswho were aloneor unable to see the other person offen Iooked down at the fahle, therebyobscuringtbek face from camera view. This appearedto be a natural way to talk when there was no one to Iook at. Attempts to induce people not to Iook down (e.g., we told them that the microphones would not be abte to pic:k up their voiceswhen they were Iooking down) wereoffen unsuccessful as the participantsbecameengrossedin listeningand "forgot" about where they were Iooking.
147
NICOLE CHOVlL
call incident involved a skiiing accident in which the woman fell and disIocated her hip. The story was chosen specifically because it had been weil told, was concerned with a potentially very serious accident, and had elicited three motor mimicry displays at the time it was originally told.
Procedure A potential participant was first asked on the telephone if she would be willing to participate in a study on conversations involving a couple of topics. (This study was tun as part of a larger study; the data analyzed were for the second of the two topics.) The researcher gare a general overview of the research over the telephone and if the person agreed to participate, a time was arranged for her to come to the lab. The study was conducted in the Human Interaction Lab of the Psychology Department at the University of Victoria. When the participants arrived, they were seated in the room and told that they would be doing a couple of tasks. (The first task involved deciding whether the participants would be willing to include various foods in dinner menu.) After the first task was completed, participants in the dyad conditions were asked to teil each other about a "close call" or "near miss" experience. Specifically, they were instructed to teil each other in some detail about a situation involving a close-call experience in which something bad almost happened or the experience was not as bad as it could have been. In the Telephone condition, the instructions differed only in that the participants were told at the beginning that they would be having a conversation over the telephone. After the experimenter had given the instructions, one participant was taken to the other room. Participants in the Tape-recording condition were told there would not be another person joining them but that they could still participate in the study if they wanted (none refused). The experimenter explained that the tasks involved leaving some telephone messages and the participants were given the same first task as the dyads. For the second task they were told that they would first listen to a short tape-recorded story and afterwards they were to describe a close call they had experienced. They were told to imagine that they were leaving this message on a friend's answering machine. (For the purposes of the study, only reactions to the tape-recorded story were analyzed.) The rest of the instructions were the same as those given in the other three conditions. When the participants were finished the experimenter came back into the room and explained that the study was over. The participants were shown the videotape, given an explanation as to the purpose of the study
148 IOURNALOF NONVERBALBEHAVlOR and asked to sign a permission form on the use of the videotape for analysis.
Scoring All tapes were scored for the frequency of motor mimicry displays. Only the first story was scored for the dyad data. The scoring procedure used in the present study was adapted from the method used in Bavelas et al. i1986b) study of motor mimicry. However, because Bavelas et al. only scored one type of motor mimicry, it was necessary to modify the system so that other types of motor mimicry displays that might occur in response to the varied themes of the close call stories could be scored. The two main criteria used in the Baveläs et al. method were used to define motor mimicry displays: (a) any display that was an appropriate reaction to the negative event being described and (b) the display was one that might be made by someone in that situation. This meant that motor mimicry displays were listener displays that took the form of displays that might have been made by the person when she was in the close call situation, for example, a listener wincing at a descrilätion of the storyteller's being hurt. Facial displays had to be noticeabte in real time and not just "micro" actions. For the most part, mimetic displays were quite stylized and immediately followed a relevant statement made by the storyteller. When the display was less clearly motor mimicry, it was included if it was an appropriate respõnse in the situation being describecl and it foliowed a relevant statement. Extremely ambiguous cases such as slight eyebrow raises, were not included as scorable motor mimicry displays. The following are two examples of listener displays that wem scored as motor mimicry displays. Square brackets [ ] indicate when the display occurred.
Example 1: "grimace" display. (eyebrows up, eyes squinted, mouth corners drawn back). A: " . . . and I was brushing oft her leg and I was leaning down, sitting down and I had my face like rig[ht by her foot right?] And it was her hind f o o t . . . " Example 2: "fear" display. (eyebrows raised, eyes widened). B: "He drove, he drove too close and we had one front wheel off the cliff, [my sicle and the truck was going over the edge.]" All tapes were scored by the author and a second trained scorer (who was unfamilar with the theoretical delineation of sociality and predictions described earlier). It was impossible to keep scorers blind to conditions
149
NICOLE CHOVlL
due to the use of a telephone in one condition and the absence of attempts by subjects to make eye contact in the Tape-recording and Partition conditions. Frequencies of scored displays per story were correlated for the two scorers; r = .94. Examination of the two sets of scores revealed that the second scorer had recorded a higher mean number of displays than the author had. The videotapes were then reviewed, and differences were resolved. To avoid the possibility of bias on the part of the author, this resolution tended more offen to include rather than exclude additional displays identified by the second scorer.
Results
Manipulation Check To determine the relative level of sociality for each of the four conditions, a group of participants from a first year psychology class (N = 65) were asked to rank order the four situations from "most social" (1) to "least social" (4). Social was defined as "how close people would feel in the situation and how easily the people would find it to converse with each other." The average ranking was calculated for each situation and transformed so that the values would reflect the increasing sociality associated with the different situations. The four situations were given the following rank ordering: Tape-recording, .14; Partition, 1.28; Telephone, 1.81; Face-to-face, 2.81.
Description of Stories The time taken by each participant to teil about her close call story ranged from one minute and five seconds to five minutes and thirteen seconds. Mean length of stories was two minutes and nine seconds in the Face-to-face condition, two minutes and fifteen seconds in the Partition condition, two minutes and twenty-four seconds in the Telephone condition, and one minute and thirty-three seconds in the Tape-recording condition. There was no significant overall difference between length of stories across the four conditions (F (3,36) = 1.70, p = .18). The themes of the close-call stories were quite varied and included, among others, incidences of car or bicycle accidents (Face-to-face, Partition, and Telephone), incidents of almost being attacked in a park, walking home, and while riding a bicycle at night (Partition), physical injuries such as being kicked by a horse (Face-to-face), skiing accidents (Tape-recording
150 tOURNAL OF NONVERBALBEHAVlOR FIGURE 1 Mean Number of Listener Motor Mimicry Displays Found in the Four Communicative Conditions. Line Graph Shows Mean Sociality Rankings for Each Condition 3.5
1.5 3 W
2.5
e3
==
O. e3
t~ LL 0 ..J
2
~ llg
1.5
""
0.5
0
0.5 0
Tape-recording
Displays Sociality
0.11 0.14
Partltion
Telephone
Face-to-face
0.32 1.28
0.71 1.81
1.14 2.81
Condition
and Telephone), falling down stairs (Telephone), almost drowning while swimming (Partition), and being shot at by a sniper in the Middle East (Telephone).
Statistical Analyses All quantitative results are based on the final set of displays after discussion. Statistical analyses were also performed using the second scorer's set of scores and the results did not differ. Due to pronounced heteroscedasticity, the data were first transformed using a log transformation [Ioglo (X + 1)]. to meet the assumptions of parametric statistics. The transformed data were analyzed by a one-way ANOVA, yielding a significant main effect, F(1,36) = 5.86, p < .002. Figure | shows the mean number of motor mimicry displays that occurred in each of the four conditions. The average sociality ranking for each condition is also shown. A linear contrast, applied to the four conditions, was found to signifi-
151 NICOLE CHOVlL
cant F(1,36) = 12.43, p = .003. The frequency of facial displays increased montonically across the four conditions in the following order: Tape-recording, Partition, Telephone, Face-to-face. These results parallel the sociality rankings for each of the situations and provides support for the hypothesis that level of sociality mediates facial displays. A second contrast was used to test the hypothesis that the three interactive conditions (Face-to-face, Telephone, and Partition) would have a significantly higher frequency of displays than the non-interactive condition (Tape-recording). This hypothesis was supported F(1,36) = 7.93, p = .008. A third contrast was used to test the hypothesis that visual availability is an important factor in determining the likelihood of facial displays. A significant difference was found between the Face-to-face condition and the three nonvisual conditions (Telephone, Partition, Tape-recording) F(1,36) = 12.43, p = .001, and therefore this hypothesis was supported.
Discussion The results of the present study suggest that the degree to which individuals can interact with each other affects the extent to which individuals exhibit motor mimicry displays in communicative situations and that the displays are affected by specific components that make up sociality. First of all, sociality appears to potentiate facial displays generally; communicative situations ranked more social were associated with higher frequencies of displays. These findings corroborate and extend the sociality findings reported by Fridlund (1991 ). The results also suggest that at least two components of sociality affect facial displays: the interactive nature of the communicative situation, and the visual availability of the receiver. Facial displays occurred frequently more in the three interactive conditions than in the non° interactive condition. Listeners who could see the story-teller also exhibited more facial displays than listeners who did not have visual access to the story-teller. The importance of visual availability underscores the findings of Bavelas, Black, Lemery, and Mullett (1986) and extends these findings to more naturalistic situations and to other kinds of motor mimicry displays. The results also study echo the findings of Brightman, Segal, Werther, and Steiner (1975) and Kraut and Johnson (1979). In all three studies, the visual presence of another increased the likelihood of display. That some facial displays were exhibited in the two nonvisual conditions raises the question of why we would make facial displays in situations where the other person cannot see us. One explanation is that dis-
152 IOURNALOF NONVERBALBEHAVIOR plays are more likely to be exhibited as the communicative situation approximates full face-to-face interaction. As Fridlund (in press) and Fridlund, Sabini, Hedlund, Schaut, Shenker, and Knauer (1990) noted, there are times when we imagine others are present andlor behave as though they were present. Thus, in the present study, some of the participants in the nonvisual interactive conditions might have treated the interaction as though it were face-to-face. The use of a social stimulus (tape-recording of a person.) produced almost no displays, even though the same story did when told face-to-face. This suggests that mere presence of a social stimulus is not the primary elicitor of motor mimicry displays, rather, being able to communicate with another is the critical variable. It is interesting to note that the results do not support the social inhibition model of facial displays. This model would have predicted an opposite pattern, namely that participants would be more likely to exhibit facial displays when alone (or when the other person could not see them). It is possible though, that inhibition may occur with only certain types of displays (e.g., pain or discomfort displays), or under certain lypes of social situations, such as being observed (Kleck, Vaughan, Cartwright-Smith, Vaughan, Colby, and Lanzetta, 1976). Nonetheless, the results from the present study suggest that presence of others does not diminish all types of facial displays. It is important to note that the findings are limited by the fact that only female subjects were used. Although we can only speculate at this time, studies that include male participants may find a lower overall rate of facial displays. Whether men would respond differently than women across the four social conditions remains to be determined. In addition to the facial display findings, informal observation of the tapes suggested that individuals tailor their messages to the communicative requirements of the situation and will alter the mode of conveyance depending upon the channels that are available. Although the listeners in the two nonvisual conditions (Telephone and Partition) did not exhibit as many motor mimicry displays in response to hearing about the close-call situations, this does not mean that they were less responsive to the stories. A number of participants in these conditions did respond with comments such as "Oh, that's terrible" or "I can't believe that happened". Subjects in the Tape-recording condition however, seldom made facial displays or vocalized. The few remarks that were made were typically said very softly, as if to themselves. This made sense given that, there was no actual person there to hear the remark. This observation suggests that the participants were very responsive to the specific dynamics of the situation. Unfor-
153 NICOLE CHOVlL
tunately, the conversations were limited in terms of the amount of dialogue that took place between participants and thus modifications are necessary in order to follow-up on this observation adequately. Nevertheless, the findings from this study suggest that facial displays are more likely to be exhibited in social interactions and illustrate their important role in conveying messages to others in face-to-face communication.
References Bavelas, J. 8., Black, A., Lemery, C. R., Maclnnis, S., & Mullett, J. (1986a). Experimental methods for studying "elementary motor mimicry". Journal of Nonverbal Behavior, I0, 102-119. Bavelas, J. B., Black, A., Lemery, C. R., & Mullett, J. (1986b). "1 show how you feel." Motor mimicry as a communicative act. Ioumal of Personality and Social Psychology, 50, 322-329. Bavelas, J. B., Black, A., Lemery, C. R., & Mullett, J. (1987). Motor mimicry as primitive empathy. In N. Eisenberg and J. Strayer (Eds.), Empathy and its development (pp. 317-338). Cambridge: Cambridge University Press. Bavelas, J. B., & Chovil, N. (1990). So you think language is verbal? An integrated message mode/of verbal and nonverbal communication. Manuscript in preparation. Brightman, V., Segal, A., Werther, P., & Steiner, J. (1975). Ethological study of facial expression in response to taste stimuli. Ioumal of Dental Research, 54, L141. Chovil, N. (1989) Communicative functions of facial displays in conversation. Unpublished doctoral dissertation. University of Victoria, Victoria, British Columbia, Canada. Darwin, C. (1965). The expression of emotion in man and animals. Chicago: The University of Chicago Press. (Originally published 1872) Ekman, P. (1972). Universals and cultural differences in facial expressions of emotion. In J. Cole (Ed.), Nebraska Symposium on Motivation: Vol. 19 (pp. 207-283). Lincoln: University of Nebraska Press. Ekman, P., & Friesen, W. V. (1969a). Nonverbal leakage and clues to deception. Psychiatry, 32, 88-105. Ekman, P., & Friesen, W. V. (1969b). The repertoire of nonverbal behavior--Categories, origins, usage, and coding. Semiotica, 1, 49-98. Ekman, P. (Ed.), (1982). Emotion in the human face (2nd ed.). Cambridge: Cambridge University Press. Emde, R. N. (1984). Levels of meaning for infant emotions: A bisocial view. In tn K. Scherer & P. Ekman (Eds.), Approaches to emotion (pp. 129-157). Hillsdale, NJ: Erlbaurn. Emde, R. N., Gaensbauer, T. I., & Harmon, R. J. (1976). Emotional expression in infancy. New York: International Universities Press. Fridlund, A. J. (in press). Evolution and facial action in reflex, social motive, and paralanguage. Biological Psychology. Fridlund, A. J. (1991). The sociality of solitary smiling: Potentiation by an implicit audience. Ioumal of Personality and Social Psychology, 60, 229-240. Fridlund, A. J., Sabini, J. P., Hedlund, L. E., Schaut, J. A., Shenker, J. t., & Knauer, M. J. (1990). Audience effects on solitary faces during imagery: Displaying to the people in your head. Journal of Nonverbal Behavior, 14, 113-137. Fridlund, A. J., Ekman, P., & Oster, H. (1987). Facial expressions of emotion: Review of literature, 1970-1983. In A. W. Siegman & S. Feldstein (Eds.), Nonverbal behavior and communication (2nd ed., pp. 143-224). Hillsdale, NJ.: Lawrence Erlbaum Associates.
154 IOURNAL OF NONVERBAL BEHAVIOR Hoffman, M. L. (1982). The measurement of empathy. In C. E. Izard (Ed.), Emotions in infants and children. New York: Cambridge University Press. Izard, C. E. (1971). The face of emotion. New York: Appleton-Century-Crofts. Kleck, R. E., Vaughan, R. C., Cartwright-Smith, J., Vaughan, K. B., Colby, C. Z., & Lanzetta, J. T. (1976). Effects of being observed on expressive, subjective, and physiological responses to painful stimuli. Joumal of Personality and Social Psychology, 34, 121 t-1218. Kraut, R. E., & Johnson, R. E. (1979). Social and emotional messagesof smiling: An ethological approach. Joumal of Persona!ity and Social Psychology, 42, 853-863. Malatesta, C. Z. (1985). Human infant: Emotional expression development. In G. Zivin (Ed.), The development of expressive behavior (pp. 183.219). Orlando, FL: Academic Press. Mehrabian, A., & Epstein, N. (1972). A measure of emotional empathy. Journat of Personality, 40, 525-543. Solano, C. H. (1989). The state of social psychology: Issues, themes, and controversies. Newbury Park, CA: Sage. Sorce, J., & Emde, R. N. (1981). Mother's presence is not enough: Effecl of emotional availability on infant exploration. Developmental Psychology, 17, 737-745. Tomkins, S. (1982). Affect theory. In P. Ekman (Ed.), Emotion in the human face (pp. 353-395). Cambridge: Cambridge University Press. Trevarthen, C. (1984). Emotions in infancy: Regulatorsof contact and relationships with persons. In K. Scherer & P. Ekman (Eds.), Approaches to emotion Ipp. 12%157). Hillsdale~ NJ: Erlbaum.