Nonverbal Behavior and the Theory of Emotion: The Facial ... · Nonverbal Behavior and the Theory of Emotion: The Facial Feedback Hypothesis Ross Buck University of Connecticut The

Journal of Personality and Social Psychology1980, Vol. 38, No. 5, 811-824

Nonverbal Behavior and the Theory of Emotion: The FacialFeedback Hypothesis

Ross BuckUniversity of Connecticut

The facial feedback hypothesis, that skeletal muscle feedback from facial expres-sions plays a causal role in regulating emotional experience and behavior, is animportant part of several comtemporary theories of emotion. A review of relevantresearch indicates that studies reporting support for this hypothesis have, with-out exception, used within-subjects designs and that therefore only a restrictedversion of the hypothesis has been tested. Also, the results of some of thesestudies must be questioned due to demand characteristics and other problems.It is suggested that visceral feedback may make a more direct contribution toemotional processes than facial feedback does and that the "readout" functionsof facial expressions are more important than any feedback functions.

Peripheral theories of motivation assumethat important aspects of motivational statesdepend on peripheral signals and cues: drynessof mouth for thirst, stomach contractions forhunger, and so forth. These approaches havebeen challenged by research demonstratingthat peripheral cues are neither necessary norsufficient for the occurrence of motivated be-havior. At most, such cues appear to functionto help inform the central nervous system ofthe state of the organism. More recent the-ories have stressed the role of the centralnervous system in directly monitoring bodilyneeds and affecting appropriate behavior(Cofer & Appley, 1964). At present, there isrelatively little evidence for the importance ofperipheral processes in directing and con-trolling motivational phenomena (Buck, 1976,p. 59).

However, the peripheral position in thetheory of emotion has recently enjoyed anupswing of interest, stimulated in large partby research in nonverbal behavior. The centerof the argument goes back to the controversy

The author thanks Reuben Baron and Robert E.Miller for their comments and suggestions on thisarticle.

Requests for reprints should be sent to Ross Buck,Dept. of Communication Sciences, U-85, Universityof Connecticut, Storrs, Connecticut 06268.

over the James-Lange theory of emotion. In1884, William James (1884/1968) arguedthat peripheral bodily changes are essential toadd an emotional quality to the perception ofan event:

Bodily changes follow directly the perception of theexciting fact, and . . . our feeling of these samechanges as they occur is the emotion. . . . Withoutthe bodily states following on the perception, thelatter would be purely cognitive in form, pale, color-less, destitute of emotional warmth, (p. 19; italics inthe original)

James (1884/1968) clearly felt that skeletalmuscle activity, including facial expression,was involved in these changes:

Can one fancy the state of rage and picture noebullition of it in the chest, no flushing of the face,no dilation of the nostrils, no clenching of the teeth,no impulse to vigorous action, but in their steadlimp muscles, calm breathing, and a placid face?(p. 23)

However, early discussions and criticisms ofthe theory focused on the role of visceral sen-sations mediated by the autonomic nervoussystem. Cannon (1932) based his famous cri-tique of the James-Lange theory in part onthe notions that (a) visceral sensation is toodiffuse and insensitive to account for the widerange of human emotional experience and (b)the autonomic nervous system acts too slowly

Copyright 1980 by the American Psychological Association, Inc. 0022-3514/80/3805-0811$00.75

811

812 ROSS BUCK

to account for the speed and lability of hu-man emotional experience.

In recent years, two kinds of major at-tempts have been made to answer these criti-cisms. One, suggested over SO years ago byBertrand Russell (1961), is that cognitionsarising from the responder's understanding ofthe situation producing the bodily sensationsaccount for the quality and lability of emo-tional experience. This notion is at the heartof Schachter's (1964) self-attribution theoryof emotion, which holds that emotion is a jointfunction of autonomic arousal and cognitiveattributions or "labels" for that arousal. Fromthis point of view, Schachter was able to inte-grate much of the research on the James-Lange theory of emotion (cf. Buck, 1976).More recently, it has been pointed out thatproprioceptive cues associated with skeletalmuscle activity may be involved in additionto automatic arousal in the bodily sensationsassociated with emotion. This has led to ex-tensions of Schachter's position in the generalcontext of self-attribution theory (Bern, 1967,1972; Heider, 1958; Kelly, 1967). The re-sponder may use the information from his ownfacial expressions and other expressive behav-iors, as well as visceral cues, to interpret andlabel his or her emotional experience. Forexample, Laird (1974) has suggested thatvisceral arousal leads to self-attributions ofthe intensity of emotion, whereas feedbackfrom one's expressive behavior leads to self-attributions of the quality of emotion.

The other answer to the criticisms of theJames-Lange theory is based on Darwin'scontention, expressed in The Expression ofthe Emotions in Man and Animals, that emo-tional processes are directly and intimatelyrelated to emotional expression.

Most of our emotions are so closely connected withtheir expression, that they hardly exist if the bodyremains passive. . . . The free expression of outwardsigns of an emotion intensifies it. On the other hand,the repression, as far as this is possible, of all out-ward signs, softens our emotions. (Darwin, 1872,quoted in Plutchik, 1962, p. 153)

From this has developed the notion that theskeletal muscle activity associated with emo-tion, and particularly facial expression, plays adirect role in regulating emotional processes.

Ekman, Friesen, and Ellsworth (1972) arguethat these muscles are sufficiently differenti-ated and respond sufficiently quickly to under-lie emotional experience: "The face might . . .fill the informational gap left by a solelyvisceral theory of emotion, distinguishing oneemotion from another, changing rapidly andproviding feedback about what is occurring tothe person" (p. 173). This neo-Darwinianposition has been adopted by a number ofrecent theories of emotion that are in otherways quite different from one another, includ-ing the theories of Plutchik (1962), Tomkins(1962, 1963), Gellhorn (1964), and Izard(1971, 1977).

The Facial Feedback Hypothesis

Collectively, these theories suggest whatmight be termed the facial feedback hypothe-sis: that facial expressions provide feedbackto the responder that is necessary or sufficientto affect significantly his or her emotionalexperience and behavior. However, none ofthese theories has dealt with this specifichypothesis in sufficient detail. This article willcritically examine the facial feedback hypothe-sis and suggest that two versions of the hy-pothesis—a between-subjects version and awithin-subjects version—can be constructedfrom the existing discussions. There is noevidence for the former version and consider-able evidence against it, whereas the severalstudies appearing to support the latter versionare flawed in certain respects and are uncon-vincing. We shall suggest that there is suffi-cient evidence of the importance of autonomicfeedback to support a cognitive-interpretiveself-attribution theory such as that ofSchachter's, but that there is as yet no con-vincing empirical evidence that facial feedbackcontributes to such a process. Instead, we shallargue that facial expressions serve a "read-out" function as a channel of social communi-cation and that their feedback function is atbest of limited importance.

Two Versions of the Hypothesis

The facial feedback hypothesis occupies aprominent place in recent theories of emotion.However, it has never been stated clearly and

THE FACIAL FEEDBACK HYPOTHESIS 813

its implications have never been fully explored.The heart of the facial feedback hypothesis isthe causal assertion that feedback from facialexpressions affects emotional experience andbehavior. One of the central implications ofthis would seem to be that the degree of feed-back should be positively related to otherindices of emotion—that the greater the feed-back, the greater the emotional response. Sincevirtually by definition facial feedback must berelated to facial expression, such expressionshould likewise be positively related to otheremotional indices. These implications areclearly stated in the above quotation fromDarwin that emotions "hardly exist if thebody remains passive" and in Lanzetta, Cart-wright-Smith, and Kleck's (1976) statementthat "awareness of one's facial expressions isthe emotion" (p. 371).

Given that the facial feedback hypothesisimplies a "positive relationship" between fa-cial expression and other indices of emotion,one must then inquire about how that rela-tionship is assessed. Is it an intersubject rela-tionship, that is, do persons who show muchexpression have a greater emotional responseon other indices than other persons who showlittle expression? This seems to be implied inDarwin's statement. Or is it an intrasubjectrelationship, that is, will a given individualshow a greater emotional response on an occa-sion when he or she is expressive than whenhe or she is not? Or do both relationshipsapply?

Unfortunately, the existing discussions ofthe facial feedback hypothesis do not makethe answers to these questions clear, and onemay in fact construct from them two versionsof the hypothesis. One, which we shall de-scriptively term the "between-subjects ver-sion" of the hypothesis, holds that individualdifferences in the feedback from emotionalexpression are related to individual differenceson other emotional indices: that for a givenemotional stimulus, Person A who freely ex-presses his or her feelings will have a greaterresponse on other indices of emotion than willa different Person B who shows little expres-sion. The other "within-subjects version" ofthe facial feedback hypothesis states that Per-son A confronted by a given emotional stimu-lus will have a greater response on other af-

fective indices if he or she freely expresses anemotion than if that same Person A were toshow little expression. These two versions ofthe facial feedback hypothesis appear similarbecause both involve the relationship betweenfacial expression and other indices of affectiveresponse. However, the between-subjects ver-sion is a statement of an intersubject relation-ship, whereas the within-subjects version is astatement of an intrasubject relationship. Sta-tistically and logically, intersubject correlationdoes not predict intrasubject correlation andvice versa.

This is illustrated in Tables 1 and 2, whichshow hypothetical facial and physiological re-sponses of Persons A and B to two affectivestimuli. They also illustrate that if one makesthe common assumption that stronger emo-tional stimuli lead to larger responses on allindices of affect, the between-subjects versionimplies the within-subjects version, whereasthe opposite is not the case. Table 1 shows thebetween-subjects version of the hypothesis:Person A has a larger facial reaction thanPerson B, regardless of the nature of the af-fective stimulus, and therefore Person A'sphysiological responses are also larger. For thestronger emotional stimulus (Stimulus 2) ,both A and B show larger responses than theydo with Stimulus 1. The result is that bothintersubject and intrasubject relationshipsbetween the facial and physiological responsesare positive.

Table 2 shows a situation where the within-subjects version of the facial feedback hy-pothesis holds, but the between-subjects doesnot. Person A is more facially expressive buthas a smaller physiological response than Per-son B, resulting in a negative intersubjectrelationship between facial expression andphysiological response. However, both A andB have larger facial reactions and conse-quently larger physiological responses to Stim-ulus 2 than they do to Stimulus 1, resulting ina positive intrasubject relationship. Thus, thewithin-subjects version of the facial feedbackhypothesis does not imply the between-sub-jects version.

The following paragraphs will review theevidence relative to the between-subjects andwithin-subjects versions of the facial feedbackhypothesis.

814 ROSS BUCK

Table 1Between-Subjects Version of the Facial Feedback Hypothesis Showing a Positive InlrasubjectRelationship Between Facial and Physiological Response

Individual

Person AFacial activityPhysiological R

Person BFacial activityPhysiological R

Average showing intrasubjectrelationship

Facial activityPhysiological R

Stimulus 1

ModerateModerate

LowLow

Low-moderateLow-moderate

Stimulus 2

HighHigh

ModerateModerate

High-moderateHigh-moderate

Average showingintersubjectrelationship



The Between-Subjects VersionThe between-subjects version of the facial

feedback hypothesis implies that a lack ofovert expression somehow short-circuits thewhole emotional process. A nonexpressive per-son should be lower than an expressive personon all indices of emotion—behavioral, self-report, and physiological—when both are con-fronted by a given affective stimulus. To assessthe validity of this hypothesis, one wouldinvestigate differences in the emotional re-sponse of individuals who differ in facial feed-back. The most obvious and convenient, if notthe most powerful, study of this kind for usewith humans would be correlational in nature:Affective stimuli would be presented to ex-pressive and nonexpressive persons and theirreactions would be assessed, with the expecta-tion that nonexpressive persons would show,fewer indications of emotion on all indices.Experimental procedures could be approxi-mated in humans by studying persons withdifferent degrees of facial paralysis, with theexpectation that measures of emotional re-sponding would vary inversely with the degreeof paralysis. Experimental techniques couldalso be applied in animals, particularly pri-mates; the hypothesis would imply that drugor lesion-induced facial paralysis should inter-fere with the acquisition, if not the retention,of emotional behavior.

Studies analogous to these have appearedduring the controversy over the James-Langetheory, but they have been directed towardthe evaluation of the affective contribution of

visceral rather than facial feedback. For ex-ample, Hohmann (Note 1) studied the re-ported emotional experiences of patients withspinal cord injuries and demonstrated thatpatients with higher spinal lesions (and thusa greater loss of bodily sensation) reported amore cognitive kind of emotional experience.("It's a mental kind of anger." "It doesn'thave the heat to it that it used to.") Simi-larly, Delgado (1969) observed that a patientwho had undergone a unilateral sympathec-tomy reported that he could no longer bethrilled by music on the sympathectomizedside of his body, whereas his response on theother side was unchanged. Also, animal studieshave shown that bilateral sympathectomyretards the acquisition of emotional behavior,although it has no effects on the retention oflearned emotional responses (Wynne & Solo-mon, 1955). These studies have demonstratedconvincingly that visceral feedback normallyplays an important role in emotional processes,although it is apparently neither necessary norsufficient for all kinds of emotional experienceor behavior (Buck, 1976, pp. 42-49).

Similar kinds of studies could evaluate thevalidity of the between-subjects version of thefacial feedback hypothesis in explaining emo-tional experience and behavior. However, noneof the recent studies purporting to supportthe facial feedback hypothesis has used thebetween-subjects paradigm necessary to evalu-ate the between-subjects form of the hypothe-sis. In fact, the available evidence bearing onthe between-subjects form of the facial feed-


Table 2Within-Subjects Version of the Facial Feedback Hypothesis Showing a Negative IntersubjectRelationship Between Facial and Physiological Responses

Individual

Person AFacial activityPhysiological R

Person BFacial activityPhysiological R

Average showing intrasubjectrelationship

Facial activityPhysiological R

Stimulus 1

ModerateLow

LowModerate


Stimulus 2

HighModerate

ModerateHigh


Average showingintersubjectrelationship

High-moderateLow-moderate

Low-moderateHigh-moderate

back hypothesis appears to be negative (cf.Buck, 1979). Correlational studies on theintersubject relationship between facial/ges-tural expressiveness and physiological respond-ing indicate that facially nonexpressive per-sons tend to have larger skin conductance andheart rate responses to emotional stimuli thando expressive persons (Buck, 1977; Buck,Miller, & Caul, 1974; Buck, Savin, Miller, &Caul, 1969, 1972; Lanzetta & Kleck, 1970;Miller, 1974). Also, Waid (1976) has sum-marized a number of studies suggesting thatskin conductance responding is positivelyrelated to the degree of socialization, and itseems unlikely that highly socialized personsare more overtly emotional than less socializedpersons. In fact, one might expect the oppositeto be the case.

The studies on the relationship betweenovert emotional expressiveness and physiologi-cal responding that have used between-sub-jects designs have tended to support H. E.Jones' (1950) distinction between "external-izing," where a person is outwardly expressivebut has small skin conductance responses,versus "internalizing," where one is less out-wardly expressive but has large skin con-ductance responses. Conceptually, the exter-nalizing-internalizing phenomenon seems in-consistent with the facial feedback hypothesisinsofar as the latter implies that increasedexpression should cause an increased responseon other emotional indices. Izard (1977) hasnoted that facial feedback may occur withoutovert expression—he cites, for example, the

Schwartz (Schwartz, Fair, Greenberg, Freed-man, & Klerman, 1974; Schwartz, Fair, Salt,Mandel, & Klerman, 1976) demonstrations of"covert expression" showing that electromyo-graph recordings can reveal the occurrence offacial muscle movement not visible on theface. However, this does not alter the implica-tion that the degree of expression should ingeneral be positively related to other emo-tional indices.

The internalizing-externalizing distinctionhas been discussed in detail elsewhere (seeBuck, 1976, pp. 259-274; 1979). The re-mainder of this article will discuss the within-subjects version of the facial feedback hypoth-esis.

The Within-Subjects Version

The within-subjects version of the facialfeedback hypothesis states that a given indi-vidual will have a greater emotional responseon other indices if he or she is expressive thanif he or she is not. At first glance, this ap-pears similar to the between-subjects versionof the hypothesis, and it is all too easy to dis-cuss the "relationship" between emotionalexpression and other affective indices withoutmaking clear whether one is considering theintersubject or intrasubject relationship. How-ever, the two in fact are quite different: Onedoes not predict the form of the other eithermathematically or conceptually, they involvedifferent assumptions about the mechanismsinvolved, they have different practical implica-

816 ROSS BUCK

tions, and they require different kinds ofresearch design and analysis to test their va-lidity. Their apparent similarity has thuscaused considerable confusion.1

The difference between the between-subjectsand within-subjects versions of the facial feed-back hypothesis was demonstrated in Table 2by the fact that it is possible simultaneouslyto have negative intersubject and positive in-trasubject relationships between emotionalexpression and other affective indices. Buck etal. (1974) in fact found this pattern of re-sults: Both a negative intersubject correlationbetween expressiveness measures and skin con-ductance responding (average r = —.50, z —2.92, p < .005), which is contrary to thebetween-subjects version of the facial feed-back hypothesis, and a positive intrasubjectcorrelation between rated expressiveness andskin conductance responding (average r —.15, z = 3.75, p < .001), which is consistentwith the within-subjects version of the hy-pothesis. In other words, Person A apparentlyhas a larger skin conductance response whenhe or she is expressive than when he or she isnot, but Person A who in general is expressivetends to have a smaller skin conductance re-sponse than Person B who in general is notexpressive.

A number of studies have appeared in recentyears that have been cited as supporting thefacial feedback hypothesis. None of these dif-ferentiates between the between-subjects andwithin-subjects versions of the hypothesis.Some investigators, in making the generalstatement that their studies involve the rela-tionship between the expression of emotionand other indices of emotional state, implythat their findings should be applicable to thebetween-subjects version of the facial feed-back hypothesis. However, these studies with-out exception used within-subjects designs,and their findings cannot be used to evaluatea hypothesis concerning a between-subjectsphenomenon.

Once the confusion about the between-sub-jects and within-subjects versions of the facialfeedback hypothesis has been cleared, it be-comes possible to examine the explanatorypower and usefulness of the remaining within-subjects version of the hypothesis that hasbeen the subject of considerable experimenta-

tion. One can ask, for example, how impor-tant facial feedback normally is in the emo-tional life of the individual. The evidencebearing on the between-subjects version ofthe hypothesis shows that nonexpressive per-sons have strong skin conductance and heartrate reactions to affective stimuli, implyingthat facial feedback may not be necessary foremotion to occur. It remains to be demon-strated whether facial feedback is as effectiveas visceral feedback has been shown to be inaffecting emotional experience. It is still possi-ble that facial feedback may be sufficient toalter, or even create, emotional experience andbehavior. The succeeding paragraphs will re-view the studies relevant to the within-sub-jects version of the facial feedback hypothe-sis in an attempt to address these issues.

Experimental Evidence

The first experiments relevant to the within-subjects version of the facial feedback hy-pothesis stemmed from self-attribution ap-proaches. Bandler, Madaras, and Bern (1968)demonstrated that subjects who "observedthemselves" escaping from shocks rated those

1 It might be noted that the within-subjects versionof the hypothesis may be stated in a strong versionthat holds across situations and a weak version thatpertains to only one situation. The former statesthat Person A will have a greater response on otheraffect indices in a situation where A is overtly ex-pressive than in a different situation where A is notexpressive. For example, if a given male is overtlyexpressive in an aggressive situation but not in anemotional situation (i.e., viewing affect-laden colorslides), one would expect that he would have greaterphysiological responses in the aggressive situationthan he would in the emotional. The latter versionstates that within a given kind of affective situation,Person A will have a greater response on otheraffect indices when A is expressive than when A isnot. For example, if a male is overtly expressive toone affect-loaded color slide but not another, heshould also have a greater physiological response tothe former slide than to the latter. There does notappear to be any evidence for or against the stronger"across situation" version of the within-subjectshypothesis, as no study has yet investigated overtexpressiveness and other affect indices in the sameperson in qualitatively different situations: All haveinvolved only one kind of affective situation. Thisarticle therefore will not consider it further (cf.Buck, 1979).


shocks as more uncomfortable than physicallyidentical shocks that they endured withoutescaping. Kopel and Arkowitz (1974) founda similar result in an experiment in whichsubjects role played being upset or calm whengiven shocks. The subjects who role played anupset reaction showed a subsequent decreasein their pain and shock tolerance thresholdsrelative to their performance before roleplaying, whereas subjects who role played acalm reaction showed higher thresholds to theshocks (rm = .296 and .515, respectively).2

Physiological responses (skin conductance inHandler et al., 1968; heart rate in Kopel &Arkowitz, 1974) were not significantly af-fected by the role playing manipulations inthese experiments.

The first studies directed specifically at eval-uating the facial feedback hypothesis werereported by Laird (1974), who (like Ihe aboveinvestigators) couched his discussion in theterms of self-attribution theory. However, healso pointed out that the neo-Darwinian the-ories such as those of Tomkins, Izard, andGellhorn would make similar predictions.Laird attempted to manipulate facial expres-sions without manipulating cognitive aware-ness of the expressions or their relevance tothe experiment. He did this by telling sub-jects to contract certain facial muscles thatwould result in facial expressions without spe-cifically asking them to make those expres-sions. Thus for a "frowning" face, subjectswere asked to bring their eyebrows down andtogether and to contract the jaws; for a"smiling" expression they were to draw thecorners of the mouth back and up. The sub-jects watched and rated their affective re-sponses to slides designed to elicit mild posi-tive and negative reactions (slides of childrenplaying vs. slides of Ku Klux Klan members)while "frowning" or "smiling." As expected(Laird, 1974), subjects reported feeling moreangry when "frowning" and more happy when"smiling," although the effect of the expres-sions was weaker than the effects of the twotypes of pictures: "The differences betweenthe means of the two picture conditions were21 to 7 times as large as the differences be-tween expression condition means, despite thefact that the pictures were not by any meansthe most emotionally evocative." (p. 481; no

rm could be calculated from the data reported.)In a second study, subjects rated cartoonsthey viewed while "smiling" as funnier thancartoons viewed while "frowning." These re-sults were interpreted as supporting the con-tention that the self-observation of one's ex-pressive behavior is one determinant of thequality of one's emotional experience.

Laird showed a particular sensitivity topossible experimenter bias and other demandcharacteristics in these experiments. In thefirst study, he provided a control group ofsubjects who received the instructions with theexperimental subjects but did not make theactual expressions while viewing the slides.The ratings of these control subjects were notsignificantly affected by the instructions. Al-though Laird acknowledged that this proce-dure was not foolproof, it did provide someevidence that the results were not due toexperimenter bias. Also, the experimental sub-jects in both of Laird's studies were given aformal postexperimental questionnaire to as-sure that they were not aware of the actualpurpose of the manipulation of their faces. Aconsiderable number of subjects were judgedto be aware and were therefore excluded fromthe analysis, including 16% of the experi-mental subjects in the first study and 19% ofthe subjects in the second.3

- Effect size estimates, based on rm, are given whenpossible (Cohen, 1977; Friedman, 1968; Rosenthal,1976).

3 More recently, Laird has been studying indi-vidual differences in subjects' response to the ex-pression manipulation, expecting such differences torelate to styles of self-attribution. Duncan andLaird (1977) studied the relationship between sub-jects' response to the expression manipulation andtheir response to a forced-compliance manipulation.They found that responses on the two tasks wererelated and interpreted the data in terms of indi-vidual differences in self-attribution styles. Also,Laird and Crosby (1974) found intersubject rela-tionships between the response to an expressionmanipulation and questionnaire items assumed toreflect the use of situational versus self-producedcues. These studies are not directly relevant to theevaluation of the facial feedback hypothesis: Essen-tially they imply that facial feedback may be usedby some persons but not by others. However, bothof these studies suffered from high attrition ratesdue to awareness of the purposes of the experiment—35% in the former study and 19% in the latter—sothat any results must be viewed with caution.

818 ROSS BUCK

Lanzetta, Kleck, and their colleagues haveconducted a series of experiments on the fa-cial feedback hypothesis that Ekman andOster (1979) consider to have shown the"strongest evidence for a positive link betweenvoluntary facial expression and emotional ex-perience" (p. 546). Unlike the self-attributiontheorists, Lanzetta, Cartwright-Smith, andKleck (1976) hold the neo-Darwinian posi-tion that there is a "direct link between fa-cial expressive behavior and the processesresponsible for autonomic arousal and sub-jective experience" and imply that the "aware-ness of one's facial expression is the emotion"(p. 369; p. 371). Laird (1974) and Lanzettaet al. (1976) have discussed the relative mer-its of the direct versus self-attribution posi-tions: This topic will not be treated here.

Like the Handler et al. (1968) and Kopeland Arkowitz (1974) studies, the experimentsof Lanzetta and Kleck's group have usedelectric shocks as affective stimuli. Lanzettaet al. (1976) discussed three separate experi-ments. In the first two, subjects were given a"baseline" series of shocks varying in inten-sity. Each shock was preceded by a signalthat indicated its intensity. The skin con-ductance response to the signal and the sub-jects' self-reports of shock painfulness wereassessed. In the first study, following thebaseline shocks, subjects were asked to "hide"any overt response to the signal, and anotherseries of shocks was administered. In the sec-ond study, subjects were instructed to "hide"or "reveal" their overt response to the signalin a complex, although not counterbalanced,sequence. The skin conductance and self-re-port responses to the anticipation and recep-tion of shocks under these instruction condi-tions were compared to the previous responseto the baseline series. Unfortunately, the re-sults of these two studies are impossible tointerpret because of the noncounterbalancednature of their designs. In both cases, respond-ing during a baseline phase was compared toresponding during a subsequent manipulationphase, so that habituation and other possibletime-series effects were confounded with theeffects of the manipulations.

In the third study, the instructions to thesubjects were presented in a random order,allowing comparisons between the instruction

conditions. The subjects were told to poseanticipating and receiving one of two levels ofshock, either no shock at all or "extremelyintense, almost unbearable shock." Actually,three levels of shock were delivered, definedas 33%, 66%, and 99% of the subject's previ-ously established tolerance limit. A color slidesignaled the subject regarding both the levelof shock and the posing condition that he orshe was to use on a given trial: A red 1 indi-cated that a low shock was to be given butthat an intense shock should be posed, a green2 indicated a moderate shock and a pose of noshock, and so forth.

The results indicated that when the subjectsposed intense shocks, they rated the shocks asmore intense (rm = .769) and had larger skinconductance responses both to the signals andto the shocks themselves (rm = .716 and .866,respectively). Lanzetta et al. concluded thatthe results were strongly supportive of thefacial feedback hypothesis. Subsequent experi-ments by Kleck et al. (1976) using male sub-jects further supported these conclusions,showing a decrease in skin conductance re-sponding (rm = .462 in Study 1, .429 in Study2) and self-reported pain (rm — .506 in Study1, .726 in Study 2) when the attenuation ofexpressive behavior was caused by social cues—the instruction that the subject was beingobserved while receiving shocks—rather thanby direct instructions to modify expressivebehavior.

In a recent experiment, Colby, Lanzetta,and Kleck (Note 2) had male subjects endurea pain tolerance test in which shock intensitywas increased to the subjects' tolerance limit,whereupon he pushed a button to escape. Sub-jects were asked to pose being in high, mod-erate, or no pain during these trials, with theexpectation that a lack of overt expressionwould be associated with lower skin conduc-tance responding and a higher pain tolerance.Only the first of these hypotheses was con-firmed: Skin conductance was highest whenthe subjects posed high pain (rm = .548), butpain tolerance scores were not significantlyaffected by the posing conditions.

Detailed reports of other studies relevantto the within-subjects version of the facialfeedback hypothesis have not yet appeared inthe experimental literature. For example, Ar-


gyle (1975) reports Shimoda's finding that ifa person adopts a facial expression duringinteraction, his or her mood tends to changeaccordingly, and Izard, Kidd, and Kotsch(Note 3) have reported that subjects in-structed to hold their faces in positions char-acteristic of given emotions tended to ratetheir subjective experience accordingly. Also,Kleck and Lanzetta (Note 4) reported thatLanzetta and Johnson altered the mood ratingsof subjects via a "biofeedback" procedurethat induced them to smile or frown. On theother hand, Ekman and Oster (1979, p. 546)report that Tourangeau and Ellsworth, usingfacial manipulations, failed to show significanteffects on self-reported emotion and foundambiguous effects on physiological responses.Although this result is not directly related toemotion, it might be noted that Cartwright-Smith (Note 5) has recently made the in-teresting finding that an exaggerated facialdisplay of effort (grimace) may increase thestrength of the grip in the forearm.

Evaluation

In evaluating whether these studies offerconvincing support for the within-subjectsversion of the facial feedback hypothesis,several problems stand out. First, all of thesestudies used highly artificial experimental situ-ations and employed deception to some extent.One of the major dependent variables ofthese experiments—self-reported emotional ex-perience—is particularly susceptible to experi-menter bias and other demand characteristicsthat are difficult to control in such studies.Only the experiment by Laird (1974) dem-onstrated an open concern with this possibil-ity, and the measures he took to guard againstit led to a high rate of subject loss that wasnot found in the other experiments. One mustquestion, therefore, the degree to which de-mand characteristics may have affected theself-reports in these studies. In adopting agiven expression, a subject might consciouslyor unconsciously conclude that a correspondingemotion is desired by the experimenter andmight respond accordingly. This explanationcannot, however, account for the physiologicalfindings.

Another problem with these studies is thatalthough they all generally supported thewithin-subjects version of the facial feedbackhypothesis, their results were not entirelyconsistent with one another. For example,Lanzetta and Kleck's group has consistentlyreported that expression manipulations affectphysiological responding, whereas the studiesof Handler et al. (1968) and Kopel and Arko-witz (1974) did not. Also, the finding byColby et al. (Note 2) that an expression ma-nipulation did not influence pain tolerance isnot consistent with the studies showing thatpain and shock tolerance thresholds and re-ports of shock painfulness are affected by suchmanipulations. The explanations for these in-consistencies are not clear.

A third problem concerns the studies em-ploying experimenter-induced facial move-ment, such as those of Laird and his col-leagues. Izard (Note 6) has had difficulty us-ing this technique with emotions other thanhappiness and anger, and he reports thatKotsch found an increase in anger ratingsusing non-emotion-related facial contractions.Izard suggests from his experience that onlywhen "subjects initiate the facial patterns forreasons of their own, the reasons consonantwith the desired effect," will the facial musclemovements play a role in emotion regulation.A related potential problem with experimenter-induced feedback, which is acknowledged byLaird (1974), is that such expressions mustbe held for an unnaturally long period of time.The feedback from such unnatural expressionscould be discounted by the subject, thus work-ing against the hypothesis (Ekman & Oster,1979). Alternately, such extended feedbackmight be stronger and more obvious to thesubject than that from natural expression, thusleading to an inaccurate confirmation of thehypothesis.

A final problem with these studies is thatall of those employing "subject generatedmovements" have used electric shock as anaffective stimulus. Some would argue that theaffective state induced by shock is not clearlyemotional, and in any case such a reliance ona single kind of eliciting stimulus is hazardous.

All in all, it could be argued that thesestudies do not convincingly demonstrate thatfacial feedback is a sufficiently important phe-

820 ROSS BUCK

nomenon in the emotional life of the indi-vidual to "fill the information gap left by asolely visceral theory of emotion." Neverthe-less, they do suggest that the within-subjectsversion of the hypothesis is valid under someconditions. The evidence that facial expressioncan apparently affect skin conductance re-sponding under some conditions is particularlyinteresting. Kleck et al. (1976) suggest thatthis may be due to the conditioning experi-ences of the individual over his or her life-time. They argue that both expressive behav-ior and autonomic nervous system responsesare unconditioned responses to affective stim-uli and that the former typically precede theslow-acting physiological responses. The in-ternal cues associated with the expressionscould come to function as conditioned stimulifor the autonomic nervous system responses,so that they could come to evoke those re-sponses in the absence of the affective uncon-ditioned stimulus.

However, there is another factor that mustbe considered by any theory that regards fa-cial feedback as playing a causal role in emo-tional phenomena. This, as Ekman and Friesen(1969, 1975) have shown, is that much facialbehavior is not associated with external af-fective stimuli but is instead related to socialand situational demands. Our long experienceof "lying with our faces" in conformance withsuch demands must lead us to disregard ourown facial feedback as a reliable source ofinformation to some extent. In the terminol-ogy of conditioning theory, our facial expres-sions must occur so often in the absence ofthe unconditioned stimulus that a considerableamount of extinction must take place. Ekmanand Friesen's (1969) finding that facial move-ments are more controlled than other bodymovements would suggest that such extinctionwould be particularly common with facial ex-pressions. However, the organism might learnto discriminate between the internal cuesassociated with spontaneous facial expressions,which tend to be reliably linked to actualaffective stimuli, and those associated withintentional facial expressions, which do not.This may account for the fact, according toKleck and Lan/etta (Note 4), that posedfacial expression in the absence of any affectivestimulus fails to cause physiological arousal.

Implications for the Theory of Emotion

The results of the studies on the facialfeedback hypothesis do not appear to compelthe acceptance of a peripheral theory of emo-tion: There is nothing to suggest that thecentral nervous system is not responsible forthe overall control of the affective process.However, it is clear from the studies of theeffects of visceral feedback that the affectiveprocess is normally accompanied and enrichedby peripheral feedback. The veteran with aspinal cord injury and the patient with a uni-lateral sympathectomy report that their expe-rience of emotion is incomplete. Peripheralfactors, including facial and other bodily feed-back, normally contribute to the emotionalstate, and under the right circumstances theymay alter the emotional state.

Peripheral Feedback and Conditioning

It may well be that proprioceptive and in-teroceptive conditioning are central to theprocess by which peripheral events becomeable to alter affective states. We carry aroundwithin us a large variety of internal visceraland proprioceptive stimuli associated withdigestion, circulation, posture, facial expres-sion, sexual functioning, and so forth, that canbecome conditioned stimuli and responses asso-ciated both with each other and with externalstimuli. For example, the internal stimuliassociated with eating and digestion can be-come conditioned stimuli producing emotionalresponses similar to those generally producedby the eating situation. Thus a child mightconsistently receive affection while eating,which is an unconditioned stimulus for com-fortable affect. Since the internal stimuli re-lated to digestion are consistently associatedwith affection, these internal stimuli maybecome conditioned stimuli producing com-fortable affect, and the individual may come toderive unusual comfort from eating. Similarly,if a child's parents consistently fight at thetable, the same kinds of internal stimuli couldbecome conditioned stimuli for fearful andangry emotions. An interesting and importantaspect of this analysis is that the latter childcould not escape from these internal stimuliin the same way that he or she could escapefrom other external stimuli that became con-


ditioned to cause negative affects (cf. Buck,1976, pp. 115-122).

Razran (1961) points out that these inter-nal stimuli, based as they are on recurringbodily functions, are always available for con-ditioning and are ever present afterward toreflect the results of that conditioning. More-over, due to the relative paucity of sensoryfibers from the internal organs, they are gen-erally not consciously reportable. As a result,a complex, unique, and largely unconscioussystem of conditioned internal reactions mustbe built up within the life of each individual.One might expect that "facial feedback"would participate in this conditioning process,with the stipulation that intentional facialdisplays may be less reliably linked to otherstimuli than are spontaneous displays.

One of the interesting features of this con-ditioning approach is that it has been advancedto explain both the positive correlation be-tween facial expression and skin conductanceresponding found in studies using within-subjects designs and the negative relationshipbetween these variables found in studies usingbetween-subjects designs. As noted above,Kleck et al. (1976) suggested that expressiveresponses may come to function as conditionedstimuli for autonomic responses because oftheir common associations with the uncondi-tioned affective stimuli and the time differencebetween them. The result would be a positiverelationship between expression and autonomicresponding when considered within the condi-tioning history of a single individual. Simi-larly, Buck et al. (1974) invoked a condition-ing explanation for the negative relationshipbetween facial expressiveness and skin con-ductance responding. It was reasoned that thesocial learning experiences associated withinhibiting an overt affective response may tendto involve stressful social rebukes that areunconditioned stimuli eliciting autonomicarousal. The inhibitory responses may becomeconditioned stimuli eliciting similar autonomicresponses because of their association withthese experiences. If a given individual en-dures many of these unpleasant social learningexperiences, he or she will come to mask fa-cial expression but will display large au-tonomic responding in affective situations.Another person who endures fewer inhibitory

experiences will have more facial expressionand less autonomic responding. The resultwould be a negative relationship between ex-pressiveness and autonomic responding be-tween subjects. Thus a consideration of pro-prioceptive and interoceptive conditioning maysuggest the conditions under which facial andother kinds of peripheral feedback may cometo alter other indices of affect.

Facial Behavior as a Readout Device

The conclusion that facial feedback mayplay a secondary role in affective processesdoes not imply that facial expression is anepiphenomenon with little importance to thestudy of emotion. It could be argued that thereal importance of facial expressions and othernonverbal behaviors lies not in their role as afeedback device but in their role as a readoutdevice. Our facial expressions reflect centralprocesses, not the reverse. That, some etholo-gists have argued, is what they are there for.For example, Andrew (196S) has suggestedthat the use of facial expression and otherkinds of communicative displays is a functionof the extent of socialization of the givenspecies and the communication demands of theenvironment, with the consequent need to co-ordinate social behavior. Social species showa much greater range of facial expressionsthan do species that lead solitary lives, andspecies that live in settings requiring muchcommunication (e.g., the plains-dwelling ba-boon) show more sophisticated repertories ofnonverbal displays than do similar specieswith other lifestyles (e.g., the forest-dwellingmandrill and drill baboon).

Due to the importance of the affective read-out in the coordination of social behavior,affective responses that differ in their "visibil-ity"—the degree to which they are normallyapparent to the responder and to observers—must undergo both different evolutionary his-tories within the species and different sociallearning histories within the individual. Re-sponses that are relatively "invisible" to oth-ers, including most autonomic responses,should be relatively unaffected by selectionpressures favoring an affective readout, whilemore visible responses may be magnified inthe service of social coordination. Since Dar-

822 ROSS BUCK

win (1872), it has been suggested that affec-tive displays evolve by the elaboration of somevisible response that tends to occur naturallyin a given affective situation. This processof "ritualization" has been described by anumber of writers (Blest, 1961; Tinbergen,1959). For example, the use of the eyebrowflash in human displays has been held to bethe result of the ritualization of the eyebrowlift associated with the widening of the eyesthat naturally accompanies an increase invisual attention (Eibl-Eibesfelt, 1972).

Affective responses that differ in visibilitymust also undergo different conditioning his-tories within the life of an individual (Buck,1971). External agents of socialization areable to subject visible displays, including fa-cial expressions and other nonverbal behav-iors, to rewards and punishments that maycause them to be intensified or masked. Also,an individual has the opportunity to observethe visible responses of others to serve asmodels for his or her own behavior. As aresult, visible affective responses come underthe influence of a variety of "display rules"(Ekman & Friesen, 1975). This is not truewith the less visible autonomic responses. Theautonomic response in a given affective situa-tion is normally a consequence of the ex-istence of stimuli that evoke conditioned orunconditioned autonomic responses. The re-sponder and others in the situation are notordinarily aware of such responses and cannotcontrol them unless special equipment (i.e.,biofeedback devices) is employed.

Because of the difference in the visibility ofexpressive nonverbal behavior and autonomicresponses and the consequent difference intheir social learning histories, it is perhapsnot surprising that they seem to be related toone another in complex ways. Also, it can beseen that facial expressions and other non-verbal behaviors not only give others a moreinformative readout of the affective state ofthe individual but also allow that individualgreater control over such a readout. We trynot to show our happiness, surprise, or angerwhen it is not appropriate to do so. Such acontrolled display of affect is, and perhaps haslong been, necessary for the smooth function-ing of primate social organizations. Such dis-plays may have also resulted in more complex

facial feedback to the responder, but thisseems to have been secondary to their readoutfunctions.

Conclusions

This article argues that the evidence fortheories positing that facial feedback has amajor causal role in emotional processing isunconvincing. Considerable confusion has re-sulted from the failure of these theories toexamine the implications of the facial feed-back hypothesis fully, particularly the differ-ence between what are here termed the be-tween-subjects and the within-subjects ver-sions of the hypothesis. There is evidence thatin within-subjects designs, posing differentemotional expressions leads to correspondingchanges on other affective indices, but manyof the findings, particularly those involvingsubjective report, may be due to demand char-acteristics. At present there is insufficient evi-dence to conclude that facial feedback is eithernecessary or sufficient for the occurrence ofemotion, and the evidence for any contribu-tion of facial feedback to emotional experi-ence is less convincing than the evidence forvisceral feedback. It is argued instead thatfacial expression has evolved in humans as ameans of affective communication and thatfacial expressions and other nonverbal behav-iors provide a controlled "readout" of centralaffective processes. The basic central affectiveprocesses have perhaps not changed greatly

• through mammalian evolution, but the abilityto express them has.

Reference Notes

1. Hohmann, G. W. The effect of dysfunctions of theautonomic nervous system on experienced feelingsand emotions. Paper presented at the Conferenceon Emotions and Feelings, New School of SocialResearch, New York, 1962.

2. Colby, C. A., Lanzetta, J. T., & Kleck, R. E.Effects of the expression of pain on autonomic andpain tolerance responses to subject-controlledshock. Paper presented at the meeting of theEastern Psychological Association, Boston, April1977.

3. Izard, C. E., Kidd, R. F., & Kotsch, W. E. Thefacial expression of emotion as an activator of thesubjective experience of emotion. Unpublishedmanuscript, Vanderbilt University, 1974.


4. Kleck, R., & Lanzetta, J. The role of expressivebehavior in emotional processes. Paper presentedat the National Conference on Body Language,New York, 1977.

5. Cartwright-Smith, J. The grimace: Facilitatingforearm strength by exaggerating the nonverbalfacial display of effort. Paper presented at themeeting of the Eastern Psychological Association,Philadelphia, Pa., April 1979.

6. C. E. Izard, Personal communication, December12, 1977.

References

Andrew, R. J. The origins of facial expressions. Sci-entific American, 1965, 213(4), 88-94.

Argyle, M. Bodily communication. New York: Inter-national Universities Press, 197S.

Handler, R., Madaras, G. R., & Bern, D. J. Self-ob-servation as a source of pain perception. Journalof Personality and Social Psychology, 1968, 9,205-209.

Bern, D. J. Self perception: An alternative interpre-tation of cognitive dissonance phenomena. Psycho-logical Review, 1967, 74, 183-200.

Bern, D. J. Self-perception theory. In L. Berkowitz(Ed.), Advances in experimental social psychology(Vol. 6). New York: Academic Press, 1972.

Blest, A. D. The concept of ritualization. In W. H.Thorpe (Ed.), Current problems in animal be-haviour. Cambridge, England: Cambridge Univer-sity Press, 1961, 102-124.

Buck, R. Differences in social learning underlyingovert-behavioral, self-report, and physiological re-sponses to emotion. Research in Education, 1971,6, 17. (Abstract)

Buck, R. Human motivation and emotion. NewYork: Wiley, 1976.

Buck, R. Nonverbal communication accuracy in pre-school children: Relationships with personality andskin conductance. Journal of Personality and So-cial Psychology, 1977, 33, 225-236.

Buck, R. Individual differences in nonverbal sendingaccuracy and electrodermal responding: The ex-ternalizing-internalizing dimension. In R. Rosen-thai (Ed.), Skill in nonverbal communication:Individual differences. Cambridge, Mass.: Oelge-schlager, Gunn & Hain, 1979.

Buck, R., Miller, R. E., & Caul, W. F. Sex, person-ality, and physiological variables in the communi-cation of affect via facial expression. Journal ofPersonality and Social Psychology, 1974, 30, 587-596.

Buck, R., Savin, V. J., Miller, R. E., & Caul, W. F.Nonverbal communication of affect in humans.Proceedings of the 77th Annual Convention of theAmerican Psychologkal Association, 1969, 4, 367-368. (Summary)

Buck, R., Savin, V. J., Miller, R. E., & Caul, W. F.Communication of affect through facial expressionin humans. Journal of Personality and Social Psy-chology, 1972, 23, 362-371.

Cannon, W. B. The wisdom of the body. New York:Norton, 1932.

Cofer, C. N., & Appley, M. H. Motivation: Theoryand research. New York: Wiley, 1964.

Cohen, J. Statistical power analysis for the behav-ioral sciences (Rev. ed.). New York: AcademicPress, 1977.

Darwin, C. The expression of emotions in man andanimals. London: John Murray, 1872.

Delgado, J. M. R. Physical control of the mind.New York: Harper & Row, 1969.

Duncan, J., & Laird, J. Cross-modality consistenciesin individual differences in self-attribution.Journal of Personality, 1977, 45, 191-206.

Eibl-Eibesfeldt, I. Similarities and differences be-tween cultures in expressive movements. In R. A.Hinde (Ed.), Nonverbal communication. Cam-bridge, England: Cambridge University Press, 1972.

Ekman, P., & Friesen, W. V. Nonverbal leakage andclues to deception. Psychiatry, 1969, 32, 88-105.

Ekman, P., & Friesen, W. V. Unmasking the face.Englewood Cliffs, N.J.: Prentice-Hall, 1975.

Ekman, P., Friesen, W. V., & Ellsworth, P. Emotionin the human face. Pergamon Press, New York.1972.

Ekman, P., & Oster, H. Facial expressions of emo-tion. Annual Review of Psychology, 1979, 30, 527-554.

Friedman, H. Magnitude of experimental effect anda table for its rapid estimation. Psychological Bul-letin, 1968, 70, 245-251.

Gellhorn, E. Motion and emotion: The role ofproprioception in the physiology and pathology ofemotions. Psychological Review, 1964, 71, 457-472.

Heider, F. The psychology of interpersonal relations.New York: Wiley, 1958.

Izard, C. E. The face of emotion. New York: Apple-ton-Century-Crofts, 1971.

Izard, C. E. Human emotions. New York: PlenumPress, 1977.

James, W. What is an emotion? In M. Arnold (Ed.),The nature of emotion. Baltimore, Md.: Penguin,1968. (Originally published, 1884.)

Jones, H. E. The study of patterns of emotionalexpression. In M. Reymert (Ed.), Feelings andemotions. New York: McGraw-Hill, 1950.

Kelley, H. H. Attribution theory in social psychol-ogy. In D. Levine (Ed.), Nebraska Symposium onMotivation (Vol. 15). Lincoln: University of Ne-braska Press, 1967.

Kleck, R. E., et al. Effects of being observed on ex-pressive, subjective, and physiological reactions topainful stimuli. Journal of Personality and SocialPsychology, 1976, 34, 1211-1218.

Kopel, S. A., & Arkowitz, H. S. Role playing as asource of self-observation and behavior change.Journal of Personality and Social Psychology, 1974,29, 677-686.

Laird, J. D. Self-attribution of emotion: The effectsof expressive behavior on the quality of emotionalexperience. Journal of Personality and Social Psy-chology, 1974, 20, 475-486.

824 ROSS BUCK

Laird, J., & Crosby, M. Individual differences inself-attribution of emotion. In H. London & R.Nisbett (Eds.), Cognitive alteration of feelingstates. Chicago: Aldine-Atherton, 1974.

Lanzetta, J. T., Cartwright-Smith, J., & Kleck, R. E.Effects of nonverbal dissimulation on emotionalexperience and autonomic arousal. Journal ofPersonality and Social Psychology, 1976, 33, 354-370.

Lanzetta, J. T., & Kleck, R. E. Encoding and decod-ing of nonverbal affect in humans. Journal ofPersonality and Social Psychology, 1970, 16, 12-19.

Miller, R. E. Social and pharmacological influenceson nonverbal communication in monkeys and man.In L. Krames, T. Alloway, & P. Pliner (Eds.),Nonverbal communication. New York: PlenumPress, 1974.

Plutchik, R. The emotions: Facts, theories and anew model. New York: Random House, 1962.

Razran, G. The observable unconscious and theinferable conscious in current Soviet psycho-physiology. Psychological Review, 1961, 68, 81-147.

Rosenthal, R. Experimenter effects in behavioralresearch (Enlarged ed.). New York: Irvington,1976.

Russell, B. An outline of philosophy. Cleveland,Ohio: World, 1961.

Schachter, S. The interaction of cognitive and physio-

logical determinants of emotional state. In L.Berkowitz (Ed.), Advances in experimental socialpsychology (Vol. 1). New York: Academic Press,1964.

Schwartz, G. E., Fair, P. L., Greenberg, P. S., Freed-man, M., & Klerman, J. L. Facial electromyogra-phy in the assessment of emotion. Psychophysiol-ogy, 1974, 11, 237.

Schwartz, G. E., Fair, P. L., Salt, P. S., Mandel, M.R., & Klerman, J. L. Facial muscle patterning toaffective imagery in depressed and non-depressedsubjects. Science, 1976, 192, 489-491.

Tinbergen, N. Comparative studies of the behavior ofgulls (Laridae): A progress report. Behavior, 1959,15, 1-70.

Tomkins, S. Affect, imagery, and consciousness: Vol.1. The positive affects. New York: Springer, 1962.

Tomkins, S. Affect, imagery, and consciousness: Vol.2. The negative affects. New York: Springer, 1963.

Waid, W. N. Skin conductance response to bothsignaled and unsignaled noxious stimulation pre-dicts level of socialization. Journal oj Personalityand Social Psychology, 1976, 34, 923-929.

Wynne, L. C., & Solomon, R. L. Traumatic avoid-ance learning: Acquisition and extinction in dogsdeprived of normal autonomic function. GeneticPsychology Monographs, 1955, 52, 241-284.

Received June 4, 1979 •

Documents

Nonverbal Behavior and the Theory of Emotion: The Facial ... · Nonverbal Behavior and the Theory of Emotion: The Facial Feedback Hypothesis Ross Buck University of Connecticut The