SIGNIFICANCE OF POSTURE AND POSITION IN THE … · SIGNIFICANCE OF POSTURE AND POSITION ... the...

14
Psychological Bulletin 1969, Vol. 71, No. 5, 359-372 SIGNIFICANCE OF POSTURE AND POSITION IN THE COMMUNICATION OF ATTITUDE AND STATUS RELATIONSHIPS 1 ALBERT MEHRABIAN 2 University oj California, Los Angeles The present paper attempts to present a review of the experimental findings dealing with the posture and position of a communicator relative to his attitude and status to his addressee. More studies are available which exhibit the detailed functional relationships of posture and position variables to com- municator attitudes than to communicator-addressee relative status. Distance, eye contact, body orientation, arms-akimbo position, and trunk relaxation have been found most consistently to be indicators of communicator attitude toward an addressee. These variables along with the degree of arm openness of female communicators and degree of asymmetry in the arrangement of arms and legs have been found or hypothesized to be associated with status relationships with the addressee. The present paper is an attempt to sum- marize investigations of the significance of the posture, distance, and orientation of com- municators toward addressees as possible in- dicators of communicator attitude and status relative to the addressee. For the purposes of discussion, attitude is broadly denned as the degree of liking, positive evaluation, and/or preference of one person for another. Findings of such studies can be applied to infer attitudes in interviews or research set- tings where the researcher is interested in inferring the attitude of his subject, whose ex- plicit verbalizations of attitude are of doubt- ful validity. For instance, the independent assessment of attitudes as they are communi- cated by posture and position cues can allow the measurement of the degree of inconsis- tency in the attitude communications of an individual, such as between posturally com- municated attitudes and verbally communi- cated attitudes (e.g., Beakel & Mehrabian, 1969). Furthermore, since in a number of everyday communication situations the com- munication of attitudes in an overt manner is culturally discouraged, the use of nonverbal cues can be a valuable means of determining the attitude of a communicator. 1 This research was supported by United States Public Health Service Grant MH 13509. 2 Requests for reprints should be sent to Dr. Albert Mehrabian, Department of Psychology, Uni- versity of California, Los Angeles, Los Angeles, California 90024. It was in contexts where overt expressions of attitude were not possible that the sig- nificance of nonverbal cues in attitude com- munication was initially denoted by psycho- analysts. Posture was used as a source of information about clients' characteristics, feelings, and attitudes toward others and themselves. Informal writings based on case studies are exemplified by the work of Deutsch (1947, 1952), who noted that the posture of a client relates to his motivations, attitudes, and intentions, which may or may not be verbalized, and that characteristic postures are associated with the initiation and termination of speech. Deutsch and Mur- phy (19SS) provided a number of specific examples whereby feelings and attitudes of clients were inferred from their postures. Reich (1945) and Braatoy (1954) denoted postural rigidity or tension as an important indicator of the correlated degree of diffi- culty encountered in the manipulation of client characteristics. In agreement with Reich and Braatoy, Lowen (1958) suggested that a client's characteristics are closely in- tertwined with characteristic postures and gestures, and therefore the manipulation of client characteristics can be enhanced through the manipulation of the latter. Like Deutsch and Murphy (1955), Fromm-Reichmann (1950) employed the changing postures of her clients to infer their feelings. She imi- tated the postures of her clients in order to 359

Transcript of SIGNIFICANCE OF POSTURE AND POSITION IN THE … · SIGNIFICANCE OF POSTURE AND POSITION ... the...

Page 1: SIGNIFICANCE OF POSTURE AND POSITION IN THE … · SIGNIFICANCE OF POSTURE AND POSITION ... the degree of directness or immediacy of in- ... new point, (b) he is taking an ...Published

Psychological Bulletin1969, Vol. 71, No. 5, 359-372

SIGNIFICANCE OF POSTURE AND POSITIONIN THE COMMUNICATION OF ATTITUDE

AND STATUS RELATIONSHIPS1

ALBERT MEHRABIAN 2

University oj California, Los Angeles

The present paper attempts to present a review of the experimental findingsdealing with the posture and position of a communicator relative to hisattitude and status to his addressee. More studies are available which exhibitthe detailed functional relationships of posture and position variables to com-municator attitudes than to communicator-addressee relative status. Distance,eye contact, body orientation, arms-akimbo position, and trunk relaxationhave been found most consistently to be indicators of communicator attitudetoward an addressee. These variables along with the degree of arm opennessof female communicators and degree of asymmetry in the arrangement ofarms and legs have been found or hypothesized to be associated with statusrelationships with the addressee.

The present paper is an attempt to sum-marize investigations of the significance ofthe posture, distance, and orientation of com-municators toward addressees as possible in-dicators of communicator attitude and statusrelative to the addressee. For the purposes ofdiscussion, attitude is broadly denned as thedegree of liking, positive evaluation, and/orpreference of one person for another.

Findings of such studies can be applied toinfer attitudes in interviews or research set-tings where the researcher is interested ininferring the attitude of his subject, whose ex-plicit verbalizations of attitude are of doubt-ful validity. For instance, the independentassessment of attitudes as they are communi-cated by posture and position cues can allowthe measurement of the degree of inconsis-tency in the attitude communications of anindividual, such as between posturally com-municated attitudes and verbally communi-cated attitudes (e.g., Beakel & Mehrabian,1969). Furthermore, since in a number ofeveryday communication situations the com-munication of attitudes in an overt manner isculturally discouraged, the use of nonverbalcues can be a valuable means of determiningthe attitude of a communicator.

1 This research was supported by United StatesPublic Health Service Grant MH 13509.

2 Requests for reprints should be sent to Dr.Albert Mehrabian, Department of Psychology, Uni-versity of California, Los Angeles, Los Angeles,California 90024.

It was in contexts where overt expressionsof attitude were not possible that the sig-nificance of nonverbal cues in attitude com-munication was initially denoted by psycho-analysts. Posture was used as a source ofinformation about clients' characteristics,feelings, and attitudes toward others andthemselves. Informal writings based on casestudies are exemplified by the work ofDeutsch (1947, 1952), who noted that theposture of a client relates to his motivations,attitudes, and intentions, which may or maynot be verbalized, and that characteristicpostures are associated with the initiationand termination of speech. Deutsch and Mur-phy (19SS) provided a number of specificexamples whereby feelings and attitudes ofclients were inferred from their postures.Reich (1945) and Braatoy (1954) denotedpostural rigidity or tension as an importantindicator of the correlated degree of diffi-culty encountered in the manipulation ofclient characteristics. In agreement withReich and Braatoy, Lowen (1958) suggestedthat a client's characteristics are closely in-tertwined with characteristic postures andgestures, and therefore the manipulation ofclient characteristics can be enhanced throughthe manipulation of the latter. Like Deutschand Murphy (1955), Fromm-Reichmann(1950) employed the changing postures ofher clients to infer their feelings. She imi-tated the postures of her clients in order to

359

Page 2: SIGNIFICANCE OF POSTURE AND POSITION IN THE … · SIGNIFICANCE OF POSTURE AND POSITION ... the degree of directness or immediacy of in- ... new point, (b) he is taking an ...Published

360 ALBERT MEHRABIAN

facilitate her own intuitive inference of theirunverbalized attitudes. Unfortunately, for themost part, observations by psychoanalystsof client characteristics or attitudes whichwere based on postural variables remainedinformal. Thus, hypotheses which relatedpostural variables to communicator attitudesor feelings remained mostly implicit in suchwork.

A general theory which would relate pos-ture, orientation, and distance cues to eitherthe communication of attitudes or status re-lationships is lacking. However, there are afew concepts which have occasionally beenused to relate these nonverbal variables toattitudes. For example, Hall (1963) sub-sumed a number of the variables under hisconcept of proxemics. Some proxemic varia-bles are distance between a communicatorand his addressee, the degree of orientationof a communicator toward his addressee (i.e.,the degree to which a communicator's bodyis turned in, versus away from, the directionof his addressee), and the presence or ab-sence of touching or eye contact betweencommunicator and addressee. Thus, the con-cept of proxemics subsumes variations in pos-ture and distance variables which relate tothe degree of directness or immediacy of in-teraction between a communicator and hisaddressee (Mehrabian, 1967; Wiener & Meh-rabian, 1968). In addition to eye contact anddirectness of body orientation, Machotka(196S) also denoted accessibility of a com-municator's body to the addressee (such asopenness of the arrangement of arms) as arelevant attitude- or affect-communicatingvariable. The latter can also be construed asa proxemic variable. Hall's (1963) interestin the introduction of the concept of prox-emics seems to have been its function in thedelineation of those aspects of posture anddistance which are acceptable and representimplicit standards within a given social orsubcultural group. The concept of immediacyproposed by Mehrabian (1967) and Wienerand Mehrabian (1968) was used primarilyfor the characterization of the role of thesepostural variables in the determination ofattitudes between communicators. Greaterimmediacy or, in Hall's terms, greater prox-emic relationships were hypothesized to cor-

respond to more positive attitudes. However,very immediate postures or positions werehypothesized to lead to negative attitude in-ferences if the immediacy exceeded that al-lowed by the implicit social norms of a sub-culture.

The differential role of nonverbal cues inthe communication of gross affect (e.g., atti-tude) versus specific affect (e.g., fear) hasalso been considered in broad terms. Ekmanand Friesen (1967) presented a reformula-tion of Ekman's (1964) findings relating tothe relative roles of bodily and facial cuesin the communication of affect. They dis-tinguished stationary positions from move-ments of body and face and suggested thatmovements are more likely to communicatespecific emotions, whereas postures and sta-tionary facial expressions are more likely tocommunicate gross affect. They also sug-gested that the intensity of specific emotionscan be inferred from bodily or facial cueswhich are either stationary or moving. How-ever, since facial movements are more fre-quent than bodily movements (i.e., changesin posture), it would follow that facial cuescan be more informative about the intensityof emotions of a communicator than can hisbodily cues. It, thus, seems useful to be ableto delineate some of the significant variablesof body position, rather than of body move-ments, for the inference of communicator at-titude, since the former can serve as onebasis for attitude inference when a communi-cator does not, or cannot, express his emo-tions in the more readily recognized verbaland facial channels.

Early studies are exemplified by Allportand Vernon's (1933) investigation of therelation of postural and gestural styles topersonality characteristics and James' (1932)study dealing with the significance of postureas communicating feeling or attitude. Sincethe present paper deals mainly with attitudecommunication, the study by James seems tobe one of the most directly relevant of theearlier investigations.

James (1932) used photographs of one maskedmale model as stimuli. He asked his subjects aboutthe attitude being expressed by each posture and theportions of the posture which were most significant.

Page 3: SIGNIFICANCE OF POSTURE AND POSITION IN THE … · SIGNIFICANCE OF POSTURE AND POSITION ... the degree of directness or immediacy of in- ... new point, (b) he is taking an ...Published

POSTURE AND POSITION IN COMMUNICATION OF ATTITUDE 361

He used 347 photographs in which the positions ofhead, trunk, feet, knees, and arms were systemati-cally varied—certain combinations being eliminateddue to their awkward quality. He selected 30 ofthese photographs on the basis of the highest agree-ment among his three 5s' judgments of the atti-tude being communicated. Two additional experi-ments in which 5s interpreted the set of 30selected postures yielded the following four posturalcategories: (a) approach, an attentive posture com-municated by a forward lean of the body; (6)withdrawal, a negative, refusing, or repulsed pos-ture communicated by drawing back or turningaway; (c) expansion, a proud, conceited, arrogant,or disdainful posture communicated by an expandedchest, erect or backward-leaning trunk, erect head,and raised shoulders; (d) contraction, a depressed,downcast, or dejected posture communicated by aforward-leaning trunk, a bowed head, droopingshoulders, and a sunken chest. For each of thesefour generic categories, the head and trunk posi-tion (s) were found to be the most important indi-cators. However, specific discriminations within eachcategory were determined by the position of handsand arms. In his third experiment, James (1932)found that postures were generally interpreted inthe same way whether S viewed the posture andinterpreted it or whether S viewed and imitatedthe posture and then interpreted it. Finally, Jamesfound that a decoder-S's response could be affectedby the situation in which the posture occurred[Mehrabian, 1968a, pp. 296-297].

James' categories of approach and with-drawal relate to the backward lean of thetorso and directness of orientation variablesconsidered subsequently. As indicated previ-ously, his findings suggest that a forward leancommunicates a relatively positive attitude,whereas a backward lean or turning awaycommunicates a more negative attitude.James' expansion category seems relevant tothe communication of status differences,whereas his contraction category does notyield information about a communicator'sattitude toward an addressee.

In considering the previous findings byJames, and other findings reported subse-quently, it is useful to distinguish betweenencoding and decoding methodologies instudies of communication. A decoding meth-odology is one in which the subjects of theexperiment are presented with prepared stim-uli, as in James' study, and are instructed toinfer feelings and attitudes from the stimuli.In contrast, in an encoding methodology, thesubjects are placed in experimental situations

which elicit different kinds of attitudes fromthem, and the observations of their postureand positions are used to study the con-comitants of the induced attitude. Typicalencoding methodologies employ role playing,in which a subject is requested to assume acertain type of role or attitude toward hisaddressee (e.g., Rosenfeld, 1966). Occasion-ally there are studies which take advantageof existing likes and dislikes among subjectsand studies which actually induce like or dis-like in a subject toward the addressee (e.g.,Exline & Winters, 196S).

In the following sections, the attitude- andstatus-communicating significance of variouscues are discussed separately. Before pro-ceeding to a discussion of experimental find-ings relating to attitude and status communi-cation, it is important to note that many ofthe variables to be considered can also serveother functions in communication, functionswhich the author does not deal with. Forexample, the work of Condon and Ogston(1966) has dealt with synchronous relationsbetween verbal and nonverbal cues emittedby a communicator or between the communi-cator's verbal and addressee's nonverbal be-haviors. One implication of their work is thatthere is a coactive regulation of communi-cator-addressee behaviors which is an intrin-sic part of social interaction and which iscertainly not exhausted through a considera-tion of the verbal portion of communications.The latter concept of regulation in communi-cation has been the primary focus of thework of Scheflen (1964, 196S). According toScheflen, changes in posture, eye contact, orposition may be employed by a communicatorto indicate that (a) he is about to make anew point, (b) he is taking an attitude rela-tive to several points being made by himselfor by his addressee, and (c) he is tempo-rarily removing himself from the communi-cation situation, as would be the case if hewere to select a large distance from the ad-dressee or to orient so that his back werefacing the addressee. There are many inter-esting aspects of the regulative function ofnonverbal cues which have heretofore beendealt with only in some informal descriptionsof interpersonal interactions.

Page 4: SIGNIFICANCE OF POSTURE AND POSITION IN THE … · SIGNIFICANCE OF POSTURE AND POSITION ... the degree of directness or immediacy of in- ... new point, (b) he is taking an ...Published

362 ALBERT MEHRAB1AN

Distance

In his informal discussion of the signifi-cance of distance between communicators,Hall (1959, 1963) noted the presence of im-plicit norms within any culture or subcultureregarding the permissible ranges of distancebetween two speakers. If the distance betweentwo speakers exceeds or is less than the limitswhich are implicitly allowed, then negativeattitudes are elicited or inferred. Hall (1964)noted that in the United States, distancesof from 6 to 18 inches are typical for inti-mate interpersonal situations, distances offrom 30 to 48 inches are typical of casual-personal interaction, distances of from 7 to12 feet are characteristic of social-consulta-tive situations, and distances of 30 feet andmore are characteristic of public interactionsituations.

One implication of the norms provided byHall is that if a communicator exceeds thedistance which is appropriate to a given so-cial situation or tries to maintain a smallerdistance than is appropriate, then a negativeattitude can be inferred by his addressee.Hall (1959) provided several examples ofinteractions among communicators from dif-ferent cultures whose implicitly acquirednorms for such distances were different andthus led to misunderstandings about atti-tudes. Studies by Garfinkel (1964) and Felipeand Sommer (1966) support Hall's observa-tions. Garfinkel (1964) found that the vio-lation of the implicit norms regarding allow-able distances led to the bewilderment andembarrassment of an addressee and to hissubsequent avoidance of the communicator.Again, Felipe and Sommer (1966) found thatwhen a communicator assumed an inappro-priately close position to another person, thatperson left earlier than he otherwise wouldhave.

For those variations in distance which oc-cur within the culturally acceptable limits, anumber of experimental studies have yieldedsystematic findings relating the distance tothe attitude between a communicator and hisaddressee. Sommer (1967) reviewed some ofthe studies relating attitudes of communica-tors to distances which they maintain vis-a-vis their addressees. Leipold (1963) used an

encoding method in which subjects were in-terviewed by an experimenter who they ex-pected would evaluate them positively ornegatively, based on information provided tothe subjects by a confederate of the experi-menter prior to the interview. Subjects whoexpected a negative evaluation selected chairswhich were farther away from the experi-menter during the interview than did sub-jects who expected a positive evaluation.Also, Little (1965) used line drawings, sil-houettes, and live actresses in experimentsin which the subject selected (encoded) ap-propriate distances between them to conveyattitudes. He found that smaller distanceswere selected by subjects for closer relation-ships between communicators.

Rosenfeld (1965) instructed his subjectsto role play (encode) an approval-seeking incontrast to an approval-avoiding attitudetoward another "subject" (a confederate inthe experiment). Rosenfeld found that underthe approval-seeking instructions, subjectssat closer to the confederate than they didunder the approval-avoiding instructions.

When Golding (1967) used semantic dif-ferential ratings of line drawings involvinghuman figures, he found that closer distanceswere interpreted (decoded) as being accept-ing and responsive, whereas greater distancesbetween communicators were interpreted inopposite terms.

Mehrabian (1968a) used both encodingand decoding methods to investigate the re-lation of distance to attitude. In his decodingmethod, subjects were requested to infer thedegree to which another person liked or dis-liked them on the basis of the distance thathe stood from them. In the encoding method,subjects were required to imagine liked ver-sus disliked addressees and to assume a stand-ing position characteristic of their interac-tions with such people. He found that whena communicator stood close (i.e., 3, as op-posed to 7, feet) to his addressee, a morepositive attitude was both inferred and com-municated. Mehrabian (1968b) also used anencoding method in which the subject wasrequired to role play five degrees of attitudetoward the addressee. It was found that dis-tance linearly decreased as positive attitudetoward the addressee increased.

Page 5: SIGNIFICANCE OF POSTURE AND POSITION IN THE … · SIGNIFICANCE OF POSTURE AND POSITION ... the degree of directness or immediacy of in- ... new point, (b) he is taking an ...Published

POSTURE AND POSITION IN COMMUNICATION OF ATTITUDE 363

Although Hall (1959, 1966) informally dis-cussed possible systematic differences in thedistances which communicators in differentnational groups assume vis-a-vis each other,experimental studies of the problem havebeen lacking. A recent study by Little (1968)requested subjects of various national groupsto position dolls relative to one another fora variety of social situations. His findingsindicate that, on the average, over the dif-ferent social situations, Greeks, Americans,Italians, Swedes, and Scots, in that order,assigned increasing distances between com-municators. He also investigated the effectsof affect and familiarity on the distances atwhich the dolls were placed from each other.He found that "friends are seen as interact-ing closer together than acquaintances, andacquaintances closer than strangers. In re-gard to affect, however, although pleasanttopics clearly produced the closest placementof the figures, the neutral and unpleasanttopic situations were not significantly differ-ent ... [Little, 1968, p. 5]." AlthoughLittle's data provided a basis for assessingthe effect on distance of status differences,his analyses of the data do not yield suchinformation. He did indicate, however, that"women see interactions of women with au-thority figures or superiors taking place ata greater distance than men view similartransactions of male figures [1968, p. 5]."

In sum, then, the findings from a largenumber of studies corroborate one anotherand indicate that communicator-addresseedistance is correlated with the degree of nega-tive attitude communicated to and inferredby the addressee. In addition, studies carriedout by sociologists and anthropologists indi-cate that distances which are too close, thatis, inappropriate for a given interpersonalsituation, can elicit negative attitudes whenthe communicator-addressee relationship isnot an intimate personal one.

Lott and Sommer (1967) found that peopleof equal status sat closer than people of un-equal status. Mehrabian (1968a) used anencoding method with seated subjects anddid not find any relationship between thestatus of a communicator relative to his ad-dressee and the distance of the former to thelatter. The latter finding is not inconsistent

with the findings of the Lott and Sommer(1967) study, however, because the experi-mental conditions which were explored inMehrabian's (1968a) study always involvedan unequal status between the communicators(i.e., the addressee was either of a higher ora lower status than the communicator, or thediscrepancy in status was the same in allconditions), and the studies yielded no sig-nificant difference in the distance that a com-municator placed himself from a higher statusaddressee compared to a lower status ad-dressee. The evidence from the two studiestherefore suggests that the distance betweentwo communicators is positively correlatedwith their status discrepancy.

Sommer's (1967) review of status relation-ships and spatial arrangements suggested thatperhaps the body orientation of communica-tors, rather than the distances between them,is a more important variable for the commu-nication of status relationships. Findings re-lating orientation are reviewed subsequentlyin a separate section.

Eye Contact

Hall (1963) included eye contact, alongwith distance, as another index of proxemics.In other words, he conceptualized that thedegree of eye contact of a communicator withhis addressee could also serve as an index ofthe attitude of the communicator toward theaddressee. Kendon (1967) provided a reviewof part of the literature dealing with the sig-nificance of eye contact in the communicationof attitudes. He distinguished between twomajor functions served by eye contact—theregulatory function and the expressive func-tion. Thus, for example, eye contact is re-garded as regulating the initiation andtermination of verbal interchanges. (Inciden-tally the same function may be associatedwith the remaining posture and position vari-ables which are considered in the presentreview, for example, Scheflen, 1964, 196S.)However, the more interesting aspect of eyecontact in the present paper is its expressivefunction, that is, the attitude-communicatingsignificance of varying degrees of eye contactbetween communicators.

Reece and Whitman (1962) studied theeffect of an investigator's warmth and cold-

Page 6: SIGNIFICANCE OF POSTURE AND POSITION IN THE … · SIGNIFICANCE OF POSTURE AND POSITION ... the degree of directness or immediacy of in- ... new point, (b) he is taking an ...Published

364 ALBERT MEHRABIAN

ness upon the amount of verbal output in asituation where the subject was free associat-ing. They defined warmth in terms of morefrequent smiling, the absence of finger-tap-ping movements, greater degree of eye con-tact with the subject, and greater degree offorward bodily lean (in a chair) toward thesubject. They found that the nonverbal vari-ables assumed to indicate experimenterwarmth or positive attitude toward the sub-ject did significantly affect the total numberof words produced by the subject, that is,more words were produced by the subjectwhen the experimenter nonverbally indicateda more positive attitude. The findings byReece and Whitman do not permit an assess-ment of the individual roles played by moreforward bodily lean, more frequent smiling,greater eye contact, or the lack of nervoushand movements in the attributing of warmthto the experimenter. The following experi-ments, however, do indicate that eye contactis an indicator of variations in attitude to-ward an addressee.

Studies by Exline and his colleagues (e.g.,Exline, 1963; Exline & Eldridge, 1967; Ex-line, Gray, & Schuette, 196S; Exline & Win-ter, 196S) or Nachshon and Wapner (1967)can be interpreted as indicating that a higherpercentage of eye contact between communi-cators is typically associated with more posi-tive attitudes between the communicators.Exline et al. (196S) found that female sub-jects, who tend to be more affiliative thanmale subjects, have more eye contact with anexperimenter than do male subjects. Further-more, they found that when the experimenterquestioned the subjects about potentially em-barrassing contents, eye contact was less thanwhen he questioned them about innocuouscontents. Both sets of findings can be inter-preted to indicate that greater degrees of eyecontact are associated with more positive atti-tudes toward the interviewer. In the case ofthe difference between the eye contact of malesand females, the attitudes of females, who aremore affiliative, can be assumed to be morepositive toward people in general or to theinterviewer in particular. In the case of theeffects due to topic, the experimenter whoasked relatively innocuous questions can beconsidered to be more positively received than

the one who asked personal and possibly em-barrassing questions.

Exline and Winters (1965) reported thatsubjects avoided the eyes of an interviewer,and disliked him, after he had commentedunfavorably about their performance. Again,Exline and Eldridge (1967) used a decodingmethod and found that the same verbal com-munication was decoded as being more fa-vorable by a subject when it was associatedwith more eye contact than when it was pre-sented along with less eye contact.

Argyle and Dean (1965) found that for agiven degree of communicator attitude to-ward an addressee, the degree of eye contactdecreased as closeness increased. Fischer(1968) reported similar findings: "Whenthree profiles are freely placed, the metricdistance between figures facing each other islarger than the distance to the third figure,thus compensating for the latter's lesser per-ceived social closeness by greater metriccloseness [p. 13]." The latter findings sug-gest that both eye contact and closeness ad-ditively reflect degree of communicator atti-tude toward or intimacy with the addressee,and therefore that increases (or decreases)in the former are associated with compensa-tory decreases (or increases) in the latterwhen the attitude is constant.

For communicators who were standing,Mehrabian (1968a) found that whereas maleshad significantly more eye contact with likedaddressees than with disliked addressees, fe-males did not. Mehrabian and Friar's (1969)study indicated that in a sitting position,male communicators, irrespective of the ad-dressee sex, had less eye contact with dislikedaddressees than with liked addressees; fur-ther, female communicators had significantlyless eye contact with disliked male addresseesthan with any of the other three addresseegroups, that is, liked males, liked females,and disliked females. Mehrabian (1968b) in-vestigated the eye contact of seated commu-nicators as a function of five degrees of com-municator attitude toward the addressee. Hefound that the amount of eye contact withan addressee was a parabolic function of at-titude toward that addressee, such that eyecontact was minimal for a disliked addressee,approached a maximum value for addressees

Page 7: SIGNIFICANCE OF POSTURE AND POSITION IN THE … · SIGNIFICANCE OF POSTURE AND POSITION ... the degree of directness or immediacy of in- ... new point, (b) he is taking an ...Published

POSTURE AND POSITION IN COMMUNICATION OF ATTITUDE 365

toward whom the attitude was neutral, andslightly diminished for addressees who wereliked very much. Thus, as in the case of thestudies of Argyle and Dean (196S) or Fischer(1968), compensatory decreases in eye con-tact were obtained when communicators as-sumed small distances to addressees who wereliked very much.

In sum, findings which relate degree ofeye contact to attitudes in nonthreatening in-terpersonal situations suggest that malesshow greater variability in their eye contactwith their addressees relative to females andthat they more consistently exhibit greaterdegrees of eye contact with liked than dis-liked addressees. Further, experimental stud-ies provide consistent support for the assumedcorrespondence of the decoded significanceof eye contact by an addressee and that en-coded by a communicator.

It is interesting to extrapolate the hypothe-sis which relates degree of eye contact toattitudes to obtain a hypothesis for the visualbehavior of approval-seeking or dependentindividuals relative to that of more indepen-dent or dominant persons. To a dependent orapproval-seeking individual, others are al-most by definition more important sources ofgratification (i.e., reinforcers) than to an in-dependent person. Thus, a dependent personwould be expected to communicate more posi-tive attitudes nonverbally, as with relativelyhigher levels of eye contact. In support ofthe latter, Efran and Broughton (1966) re-ported a significant correlation between scoreson the Crowne and Marlowe (1960) socialdesirability scale and the extent of eye con-tact the subjects had with their addressees.However, in a subsequent study, Efran(1968) failed to replicate that finding. Ad-ditional findings by Exline and Messick(1967) suggest that eye contact may be usedby dependent individuals not only to com-municate more positive attitudes, but alsoto elicit such attitudes when they are notforthcoming: They found that dependentmales had more eye contact with a listenerwho provided them with few as compared tomany social reinforcers, whereas dominantmales decreased their eye contact with lessreinforcing listeners.

Studies have also indicated that eye con-

tact is significantly related to the status re-lationship between communicators. Sommer(1967) reviewed some of the literature deal-ing with the spatial arrangement of subordi-nates and leaders. For example, findings byHearn (19S7) imply that eye contact with anaddressee is a parabolic function of the statusof that addressee, provided distance andother variables are held constant. Thus, eyecontact is moderate with a very high-statusaddressee, at a maximum with a moderatelyhigh-status addressee, and at a minimumwith a very low-status addressee.

There is some available evidence relevantto that part of the previous interpretation ofHearn's (19S7) findings which deals withmoderately high-status versus low-status ad-dressees. Efran (1968) investigated the eyecontact of freshmen with seniors and fresh-men who were the addressees. The results in-dicated that, of a senior-freshman pair whowere simultaneously being addressed by afreshman, the senior received more eye con-tact from the subject than did the lowerstatus freshman.

Among standing communicators, Mehra-bian (1968a) found that both males and fe-males had significantly more eye contactwith high-status addressees than with low-status addressees. The difference, however,was greater for male than for female com-municators. Mehrabian and Friar (1969)found that, regardless of a seated communi-cator's sex, there was significantly less eyecontact with low-status male addressees thanwith any of the other three addressee groups,that is, low-status females, high-status males,and high-status females. In addition, thestatus and attitude factors interacted in de-termining degree of eye contact of a seatedcommunicator with the addressee. For malecommunicators, there was more eye contactwith liked high-status addressees than withdisliked high-status, liked low-status, or dis-liked low-status addressees—the latter threemeans not differing significantly among them-selves. For female communicators, however,the significant difference was less eye con-tact with disliked low-status addressees thanwith disliked high-status, liked high-status, orliked low-status addressees—the latter threemeans not differing significantly among them-

Page 8: SIGNIFICANCE OF POSTURE AND POSITION IN THE … · SIGNIFICANCE OF POSTURE AND POSITION ... the degree of directness or immediacy of in- ... new point, (b) he is taking an ...Published

366 ALBERT MEHRAB1AN

selves. In other words, for seated male com-municators, it was the liked high-status ad-dressees who elicited a discriminably greaterdegree of eye contact, whereas for seated fe-male communicators, it was the disliked ̂ low-status addressees who elicited a discriminablysmaller degree of eye contact.

Finally, as in the case of cross-culturalstudies of distance, studies of eye contactand other posture and position cues couldyield consistent differences in, for instance,national groups varying in the degree towhich they are socially stratified. It may behypothesized that variations in eye contactas a function of variations in status of anaddressee may be more clearly defined in themore authoritarian than the more democrati-cally oriented cultures. In addition to eyecontact, the various cues of body relaxation,such as asymmetry of limb placement or side-ways and reclining angles of seated communi-cators, may also exhibit less ambiguous re-lationships as a function of the status of theaddressee in authoritarian than in democrati-cally oriented cultures.

Body Orientation

Body orientation (i.e., the degree to whicha communicator's shoulders and legs areturned in the direction of, rather than awayfrom, his addressee) can also serve as an in-dicator of communicator attitude or status.However, investigations which have employedbody orientation as a dependent measurehave yielded findings which are not as un-ambiguous as those obtained in relation todistance or eye contact (e.g., Argyle & Ken-don, 1967). For example, Rosenfeld (1965)did not find a significant difference in thebody orientation of his subjects toward theaddressee in an approval-seeking, in contrastto an approval-avoiding, situation.

It should be noted that, in a number ofstudies where body orientation has been avariable of interest, the effects of body orien-tation and eye contact have been confounded.Greater degrees of eye contact with an ad-dressee tend to be associated with a moredirect orientation of the head, shoulders, andlegs of a communicator toward his addressee.For example, Mehrabian (1968a), using anencoding method, found that for communi-

cators who are in a standing position, shoul-der orientation (i.e., the number of degreesthat a plane perpendicular to the plane ofthe subject's shoulders is turned away fromthe median plane of his addressee) correlated— .41 with eye contact. In other words, in astanding position, the greater the directnessof orientation toward the addressee, thegreater was the eye contact with the ad-dressee. Mehrabian (1968b) used indices ofhead, shoulder, and leg orientation, in addi-tion to the eye contact measure, based on en-coded communications of seated subjects.Average figures from the latter and the Meh-rabian and Friar (1969) study indicated in-tercorrelations among the head, shoulder, andleg orientation measures in excess of .80.Furthermore, eye contact correlated —.51with head orientation; —.40 with shoulderorientation; and —.34 with leg orientation.The data relating the various orientationmeasures to each other and to eye contact forseated and standing communicators indicatethat there is considerable consistency amongthe various orientation measures. Therefore,shoulder orientation alone can be used as asummary index of the body orientation ofstanding or seated communicators. Further-more, the correlations of eye contact withthe various orientation indices are low enoughthat body orientation and eye contact can betreated as separate indices of attitude andstatus.

In the following experiment by Mehrabian(1967) the effects of eye contact and bodyorientation were experimentally separated.The experimenter, while presenting a brieftalk to her subjects about a topic which wasnot related to attitude communication, sys-tematically varied her posture vis-a-vis thetwo subjects who were simultaneously pres-ent. For half of the pairs of subjects, theexperimenter had eye contact and direct body(i.e., shoulder and leg) orientation 90% ofthe time with one subject in the pair and hadeye contact and direct body orientation 10%of the time with the other subject in the pair.For the other half of the paired subjects, oneof each pair received 90% eye contact and10% body orientation, whereas the other sub-ject in the pair received 10% eye contactand 90% body orientation. Thus, the effects

Page 9: SIGNIFICANCE OF POSTURE AND POSITION IN THE … · SIGNIFICANCE OF POSTURE AND POSITION ... the degree of directness or immediacy of in- ... new point, (b) he is taking an ...Published

POSTURE AND POSITION IN COMMUNICATION OF ATTITUDE 367

of eye contact and body orientation wereseparated in the design, and the female sub-jects' judgments (decodings) of how muchthey thought the experimenter liked themserved as the dependent measure. More eyecontact communicatd a more positive atti-tude. Further, when eye contact was present,less direct body orientation of the communi-cator was interpreted as an indicator of lesspositive attitude than when there was moredirect body orientation. This effect due tobody orientation was not found when therewas an absence of eye contact.

Mehrabian (1968a) did not find any sig-nificant relationship between the body orien-tation of standing communicators and theirencoded attitudes toward their addressees.However, in the case of seated communica-tors, Mehrabian (1968b) found the followingrelationships between shoulder orientationand increasing degrees of positive communi-cator attitude (corresponding to intense dis-like, moderate dislike, neutral, moderate lik-ing, and intense liking). Shoulder orientationwas not a discriminator of varying degrees ofmale communicator attitude except when theaddressee was liked very much, in which casethe shoulder orientation was less direct. Theshoulder orientation of female communicatorsfunctioned in a similar manner to their eyecontact: It was a parabolic function of in-creasing degrees of attitude and was mostdirect for neutral addressees, least direct forintensely disliked addressees, and moderatelydirect for intensely liked addressees.

Shoulder orientation of a communicator to-ward the addressee has also been found to berelated to the relative status of a communi-cator to the addressee. For communicators ina standing position, Mehrabian (1968a)found that shoulder orientation was more di-rect with a high-status than with a low-statusaddressee, regardless of the attitude towardthe addressee.

Additional decoding experiments in whichthe effects of body orientation and eye con-tact can be separated are required to clarifythe relationships of directness of shoulderorientation to attitudes and status relation-ships. The evidence which is presently availa-ble suggests that males use a less direct bodyorientation when the addressee is liked very

much and that females use very indirect bodyorientation with intensely disliked addressees,least indirect orientation with neutral ad-dressees, and moderately indirect orientationwith liked addressees. The latter findings arefor communicators who are seated. For stand-ing communicators, body orientation has notbeen found to be different when the addresseeis extremely liked versus extremely disliked.Finally, body orientation is more direct to ahigher status than to a lower status addressee.

Accessibility of Body—Openness oj Armsand Legs

Machotka's (1965) observations suggestedthat subjects were more drawn to a nudefigure in a moderately open-armed positionthan to figures which had a very open ora closed-arm position. The subjects in thestudy rated line drawings displaying differentdegrees of the openness of arm arrangementof nude female figures. Compared to the fig-ures with moderate or very open arm posi-tions, the figures with closed-arm positionswere judged as being cold, rejecting, shy, andpassive. Thus, Machotka's study suggestedthat more accessible communicator posturescommunicated more positive attitudes towardthe addressee. Mehrabian (1968a) used twoindices of the accessibility of a communica-tor's body, namely, the degree of openness ofthe arrangement of the arms and legs. Nosignificant correlation between the opennessof the arrangement of arms and legs of thestanding communicators was obtained. Fur-thermore, the male communicators did notassume a more open arm or leg position withthe liked than disliked addressees. However,females assumed a more open arrangementof the arms while communicating to likedmale than to disliked male addressees; therewas no corresponding difference when the ad-dressees were female. In a study in which fivedegrees of communicator attitude were ex-plored, Mehrabian (1968b) found no signifi-cant relationship between openness of thearrangement of arms or legs of seated com-municators and their attitude toward theaddressees.

In an experiment in which subjects inferred(decoded) the attitude of a seated communi-cator, accessibility of posture was found to

Page 10: SIGNIFICANCE OF POSTURE AND POSITION IN THE … · SIGNIFICANCE OF POSTURE AND POSITION ... the degree of directness or immediacy of in- ... new point, (b) he is taking an ...Published

368 ALBERT MEHRABIAN

relate to inferred attitudes. Mehrabian(1968a) found that whereas an open postureof male encoders did not communicate a morepositive attitude than a closed posture, anopen posture of female encoders sometimesdid. For the open posture, encoders wereasked to sit with their arms resting on theirlaps with hands not touching and both feetresting on the floor in a normal legs-uncrossedposition. For the closed posture, they wereasked to sit with their arms folded and theirlegs crossed. The communicators were mod-erately relaxed in both postural positions. Intwo experiments, relative to the decoder, thecommunicators were either discriminablyolder and more formally dressed, or of thesame age, or in the age range from 10 to 13.For a relaxed posture, which is more typicalof everyday communication situations, a moreaccessible posture of males was not decodedby males or females as being a significantdeterminer of attitude, whereas a relativelyaccessible posture of females (younger orolder but not of the same age as the decoder)was decoded as communicating a more posi-tive attitude.

Whereas the body accessibility of a com-municator does not seem to be a consistentcorrelate of the communicator's attitude to-ward the addressee, some findings suggestthat it is systematically related to status rela-tive to the addressee. Standing female com-municators assumed more open arm positionsin the presence of higher status addressees,whereas there was no similar difference inthe case of male communicators (Mehrabian,1968a). In Mehrabian and Friar's (1969)study, seated female communicators who wereaddressing a high-status addressee assumeda significantly less open arm position thanfemale communicators who were addressing alow-status addressee. Thus, only in the caseof female communicators has an open armposition been found to indicate status rela-tionships. While the person was standinggreater degrees of openness of arm position(i.e., arms tending to be hanging rather thancrossing in front of the chest) occurred witha high-status addressee, whereas while theperson was seated less arm openness occurredwith a high-status addressee than with a low-status addressee.

Recent data, discussed in the following sec-tion, together with the preceding relation-ships of the openness of arm arrangement andstatus of an addressee suggests that an openarm position of seated communicators maymore appropriately be considered an index ofrelaxation, with relatively more open posi-tions indicating greater relaxation. In con-trast, for standing communicators, a foldedarm position may be more relaxed than onewith the arms hanging. Given these con-siderations, then, the experimental data needto be reexamined in terms of the degree ofrelaxation which can be inferred from arm orleg arrangement, rather than the accessibilitythey provide to the body. Thus, for instance,although the folded arm position of seatedfemales may be a more "proper" and tenseposition, that same closed arm position whilestanding may be considered a more relaxedposition and may thus occur more with lowerstatus addressees.

Arms-Akimbo Position

Mehrabian (1968a) found that the use ofthe arms-akimbo position by a standing com-municator was indicative both of the atti-tude and the status of the communicatorrelative to his addressee. Specifically, hefound that there was a greater tendency fora communicator to use the arms-akimbo posi-tion with disliked addressees than with likedaddressees. Furthermore, there was a greatertendency to use an arms-akimbo position withlow-status than with high-status addressees.

Relaxation Measures

Schlosberg (1954) denoted the dimensionof sleep-tension as being relevant to the ex-pression of emotions. Dittmann, Parloff, andBoomer (1965), in selecting stimuli for anexperiment in which they were concernedwith the pleasantness of bodily and facialcues, indicated that it was more difficult tocharacterize the pleasantness of bodily thanof facial cues. However, in selecting pleasantbody positions they looked for relaxed pos-tures with little movements, while in select-ing unpleasant body positions they looked forthose involving obvious muscle tension orfidgety and nervous activity. In their study,which was a decoding one, they found that

Page 11: SIGNIFICANCE OF POSTURE AND POSITION IN THE … · SIGNIFICANCE OF POSTURE AND POSITION ... the degree of directness or immediacy of in- ... new point, (b) he is taking an ...Published

POSTURE AND POSITION IN COMMUNICATION OF ATTITUDE 369

dancers were more attuned to bodily cuesthan psychotherapists who attempted to makemost of their judgments on the basis offacial pleasantness. However, the investiga-tors noted that both the psychotherapists andthe dancers were influenced more by thefacial than by the bodily cues in making theirjudgments of pleasantness and unpleasant-ness.

In a number of studies, separate indicesof the degree of relaxation of a communica-tor's hands, legs, and trunk have been ob-tained. It will be recalled that Reece andWhitman (1962) assumed that a warm ex-perimenter attitude was communicated to thesubject if the experimenter leaned forwardin his chair, smiled, kept his hands still, andhad more eye contact with the subject. How-ever, as has already been noted, their findingsdid not clearly indicate which of these pos-tural and facial variables influenced the num-ber of words produced by a subject who wasfree associating. A number of recent findingsindicate that a smaller reclining angle of acommunicator while seated, and therefore asmaller degree of trunk relaxation, communi-cates a more positive attitude. For example,Mehrabian (1968a) found that both maleand female addressees inferred a more nega-tive attitude when the communicator wasleaning backward and away from them thanwhen he was in a forward leaning position.Further, Mehrabian (1968b) found that inattempting to communicate variations in atti-tude, communicators' reclining angles withdisliked or neutral addressees were not sig-nificantly different, but that the angles de-creased for increasing positive attitudes to-ward the addressees. Finally Gottheil, Corey,and Paredes (1968) reported a significantcorrelation between the reclining angle whicha person assumed while talking to an inter-viewer and the distances which he used torepresent his typical distances with signifi-cant others in social situations. That is tosay, persons who tended to communicate lesspositive feelings by reclining more also tendedto perceive themselves as assuming greaterdistances from addressees in general.

In addition to reclining angle, a second in-dex of trunk relaxation is the degree of side-ways lean of a seated communicator. Meh-

rabian's (1968b) study which related thedegree of sideways lean to attitude towardthe addressee yielded different results formale and female communicators. Male com-municators exhibited less sideways lean andgenerally less body relaxation with intenselydisliked males, whereas females exhibitedtheir largest degree of sideways lean with in-tensely disliked male or female addressees.For the remaining four degrees of attitude,the sideways-lean index of relaxation ex-hibited similar patterns for male and femalecommunicators. Thus, sideways lean wasmoderately high for disliked addressees, waslowest for neutral addressees, and was mod-erately high for liked and intensely likedaddressees.

In contrast to indices of body relaxation,those relating to arm and leg relaxation havefailed to yield consistent relationships withattitude (e.g., Mehrabian, 1968b). Earlierstudies employed rather global judgments ofthe degree of relaxation in the arms or legsof communicators. In several of the author'srecent unpublished experiments, the degreeof asymmetry in the arrangements of thearms and legs was employed as a possiblemeasure of relaxation in the limbs. Factor-analytic results from such experiments sug-gested that the following postural and move-ment variables define relaxation: sidewayslean while seated, reclining angle, leg-posi-tion and arm-position asymmetry, arm open-ness, and higher rates of gesticulation androcking and lower rates of lateral swivel whileseated on a desk chair. Thus, the heretoforeneglected variables of symmetry in the ar-rangement of arms and legs, which have beenfound to be valid indices of relaxation-ten-sion, seem promising variables for exploringattitudes and status relations between com-municators.

The findings relating to relaxation can be sum-marized by noting that there is a curvilinear rela-tionship between the attitude of a communicatortoward his addressee and the degree of relaxationmanifested by the communicator. Degree of relaxa-tion is either very high or very low for a dislikedaddressee and is moderate for a liked addressee[Mehrabian & Friar, 1969].

Further, the obtained sex differences provideinformation regarding the preferred modes of

Page 12: SIGNIFICANCE OF POSTURE AND POSITION IN THE … · SIGNIFICANCE OF POSTURE AND POSITION ... the degree of directness or immediacy of in- ... new point, (b) he is taking an ...Published

370 ALBERT MEHRABIAN

expressing dislike through variations in re-laxation on the part of males and females.

It seems that a very disliked male addressee whocould be threatening to another male elicits a rela-tively tense posture from that communicator, where-as a disliked female addressee elicits an extremelyrelaxed posture from a male communicator. Femalecommunicators, on the other hand, do not exhibitdifferences in their body relaxation with male andfemale addresses and in both instances communicatedislike with greater degrees of relaxation.

The findings relating to sideways lean and shoul-der orientation suggest a consistent pattern of dif-ferences for the nonverbal behaviors of male andfemale communicators vis-a-vis male addressees whoare disliked intensely. Relative to females, malecommunicators exhibit less body relaxation (as in-dicated by less sideways lean) and a greater degreeof vigilance (as indicated by more direct shoulderorientation and a similar trend approaching sig-nificance for eye contact) toward intensely dislikedmales. Otherwise, for the remaining four degrees ofattitude, the communication behaviors of males andfemales exhibit similar patterns. The differences ob-tained for intensely disliked male addressees can beinterpreted by suggesting that these addressees arepotentially a greater physical threat to a male thana female communicator and therefore elicit a greaterdegree of body tension and vigilance from the formerthan from the latter [Mehrabian, 1968b, p. 29].

In connection with the latter discussion ofthe significance of vigilance which is exhib-ited by less sideways lean, more direct shoul-der orientation, and more eye contact, it isinteresting to note the following findings byEllsworth and Carlsmith (1968): "If thetopic of conversation is neutral to generallypositive, subjects like the interviewer signifi-cantly more when she looks them in the eye.. . . But in a conversation which is indirectlybut persistently critical of the subject, thisrelationship is reversed [p. 18]." The find-ings from the two latter studies indicate thata threatening person elicits more vigilanceand eye contact and that threatening com-ments, when associated with more eye con-tact, are more offensive. In short, in the con-text of a hostile relationship, higher levels ofeye contact are elicited and communicatemore negative feelings.

There are additional findings which relatevarious indices of relaxation and communica-tor-addressee relative status. Goffman (1961)noted that in psychiatric staff meetings, lowerstatus participants (interns) were less re-laxed than the higher status participants

(psychiatrists). Findings obtained from com-municators who were in a standing positioncorroborated some of the observations byGoffman. Mehrabian (1968a) found thathand as well as leg relaxation was greaterwhile communicating to a lower status ad-dressee. Further, Mehrabian and Friar (1969)found that communicators exhibited moresideways lean with low-status than with high-status addressees.

In sum, then, in contrast to its relationshipto attitude, relaxation seems to exhibit alinear relationship with status, as follows:There is a high degree of relaxation with alow-status addressee, a moderate degree of re-laxation with a high-status addressee, and anintermediate degree of relaxation with peers.

SUMMARY

An overview of the findings which relatethe posture of a communicator to his atti-tude toward his addressee suggest the fol-lowing consistent patterns. The distance be-tween a communicator and his addressee is adecreasing linear function of the degree ofliking of the addressee. Eye contact is mini-mal for disliked addressees, approaches amaximum value for addressees toward whomthe attitude is neutral, and slightly dimin-ishes for addressees who are liked very much.For female communicators, body orientationtoward the addressee, like their eye contact,is a parabolic function of attitude toward theaddressee, such that the least direct orienta-tion occurs for intensely disliked addressees,the highest degree occurs for addressees whoare regarded as neutral, and, again, a rela-tively high degree occurs for addressees whoare liked very much. For male communicators,the only significant difference occurs with in-tensely liked addressees, who receive a lessdirect body orientation. The slight decreasein the directness of orientation of males orfemales toward intensely liked addresseesmay be understood in terms of the tendencyof communicators to assume a side-by-sideand very close position when they communi-cate to such addressees.

Several variables relating to the position-ing of the arms and legs of a communicatorhave been explored. These include degree ofasymmetry, and openness of the arrangement

Page 13: SIGNIFICANCE OF POSTURE AND POSITION IN THE … · SIGNIFICANCE OF POSTURE AND POSITION ... the degree of directness or immediacy of in- ... new point, (b) he is taking an ...Published

POSTURE AND POSITION IN COMMUNICATION OF ATTITUDE 371

of arms and legs, as well as uses of the arms-akimbo position. An open arrangement of thearms of seated females has been found to com-municate a more positive attitude to older oryounger addressees, but not to addressees ofthe same age. An arms-akimbo position tendsto be used by communicators who wish toindicate a more negative feeling to their ad-dressees. There are no studies of the relationbetween an asymmetrical arrangement of thearms or legs and attitudes, however, thesehave been found to be indices of communica-tor relaxation. Two additional indices of re-laxation, reclining and sideways-leaning an-gles while seated, have yielded significantrelations with attitude. A larger reclining an-gle communicates a more negative attitude,and whereas moderate values of sidewayslean convey positive attitudes, very large orsmall reclining angles communicate neutralor negative attitudes.

Findings which relate the posture and posi-tion of a communicator to his status relativeto his addressee suggest the following con-sistent pattern. The distance between twocommunicators is positively correlated withtheir status discrepancy. General considera-tions suggest that eye contact is moderatewith high-status, is at a maximum with mod-erately high-status, and is at a minimum withlow-status addressees, with some supportingevidence which deals with moderately high-status- versus low-status-addressee communi-cation situations. Further studies require de-signs which allow the assessment of the sepa-rate effects of body orientation and eye con-tact in relation to status difference betweencommunicators.

The use of an arms-akimbo position is lessprobable when the addressee is of a higherstatus. Relationships between indices of re-laxation of hands, legs, and trunk and thestatus of an addressee indicate that when thecommunicator is in a standing position, handas well as leg relaxation is greater while com-municating to a low-status than to a high-status addressee. When the communicator isin a seated position, the sideways-lean indexof trunk relaxation has been found to begreater with low-status than with high-statusaddressees.

REFERENCES

ALLPORT, G., & VERNON, P. Studies in expressivemovement. New York: MacMillan, 1933.

ARGYLE, M., & DEAN, J. Eye contact, distance andaffiliation. Sociometry, 1965, 28, 289-304.

ARGYLE, M., & KENDON, A. The experimental analy-sis of social performance. In L. Berkowitz (Ed.),Advances in experimental social psychology. NewYork: Academic Press, 1967.

BEAKEL, N. G., & MEHRABIAN, A. Inconsistent com-munications and psychopathology. Journal of Ab-normal Psychology, 1969, 74, 126-130.

BRAATOY, T. F. Fundamentals of psychoanalytictechnique. New York: Wiley, 1954.

CONDON, W. S., & OGSTON, W. B. Sound film analy-sis of normal and pathological behavior patterns.Journal of Nervous and Mental Disease, 1966,143, 338-347.

CROWNE, D. P., & MARLOWE, D. A new scale of so-cial desirability independent of psychopathology.Journal of Consulting Psychology, 1960, 24, 349-354.

DEUTSCH, F. Analysis of postural behavior. Psycho-analytic Quarterly, 1947, 16, 195-213.

DEUTSCH, F. Analytic posturology. PsychoanalyticQuarterly, 1952, 21, 196-214.

DEUTSCH, F., & MURPHY, W. F. The clinical inter-view. New York: International University Press,1955. 2 vols.

DITTMANN, A. T., PARLOFF, M. X., & BOOMER, D. S.Facial and bodily expression: A study of recep-tivity of emotional cues. Psychiatry, 1965, 28,239-244.

EFRAN, J. S. Looking for approval: Effects on visualbehavior of approbation from persons differingin importance. Journal of Personality and SocialPsychology, 1968, 10, 21-25.

EFRAN, J, S., & BROUGHTON, A. Effect of expectan-cies for social approval on visual behavior. Jour-nal of Personality and Social Psychology, 1966,4, 103-107.

EKMAN, P. Body position, facial expression andverbal behavior during interviews. Journal of Ab-normal and Social Psychology, 1964, 68, 295-301.

EKMAN, P., & FRrESEN, W. V. Head and body cuesin the judgment of emotions: A reformulation. Per-ceptual and Motor Skills, 1967, 24, 711-724.

ELLSWORTH, P. C., & CARLSMITH, J. M. Effects ofeye contact and verbal content on affective re-sponse to a dyadic interaction. Journal of Per-sonality and Social Psychology, 1968, 10, 15-20.

EXLINE, R. V. Explorations in the process of personperception: Visual interaction in relation to com-petition, sex, and need for affiliation. Journal ofPersonality, 1963, 31, 1-20.

EXLINE, R. V., & ELDRIDOE, C. Effects of two pat-terns of a speaker's visual behavior upon the per-ception of the authenticity of his verbal message.Paper presented at the meeting of the EasternPsychological Association, Boston, April 1967.

EXLINE, R. V., GRAY, D., & SCHUETTE, D. Visualbehavior in a dyad as affected by interview con-

Page 14: SIGNIFICANCE OF POSTURE AND POSITION IN THE … · SIGNIFICANCE OF POSTURE AND POSITION ... the degree of directness or immediacy of in- ... new point, (b) he is taking an ...Published

372 ALBERT MEHRABIAN

tent and sex of respondent. Journal of Personalityand Social Psychology, 1965, 1, 201-209.

EXLINE, R. V., & MESSICK, D. The effects of depen-dency and social reinforcement upon visual be-havior during an interview. British Journal ofSocial and Clinical Psychology, 1967, 6, 256-266.

EXLINE, R. V., & WINTERS, L. C. Affective relationsand mutual glances in dyads. In S. Tomkins &C. Izzard (Eds.), Affect, cognition and personality.New York: Springer, 1965.

FELIPE, N. J., & SOMMER, R. Invasions of personalspace. Social Problems, 1966, 14, 206-214.

FISCHER, C. T. Social schemas: Response sets or per-ceptual meanings? Journal of Personality and So-cial Psychology, 1968, 10, 8-14.

FROMM-REICHMANN, F. Psychoanalysis and psycho-therapy. Chicago: University of Chicago Press,1950.

GARFINKEL, H. Studies of the routine grounds ofeveryday activities. Social Problems, 1964, 11,225-250.

GOFFMAN, E. Encounters. Indianapolis: Bobbs-Mer-rill, 1961.

GOLDINO, P. Role of distance and posture in theevaluation of interactions. Proceedings of the 75thAnnual Convention of the American PsychologicalAssociation, 1967, 2, 243-244.

GOTTHEIL, E., COREY, J., & PAREDES, A. Psychologi-cal and physical dimensions of personal space.Journal of Psychology, 1968, 69, 7-9.

HALL, E. T. The silent language. New York: Faw-cett, 1959.

HALL, E. T. A system for the notation of proxemicbehavior. American Anthropologist, 1963, 65,1003-1026.

HALL, E. T. Silent assumptions in social communica-tion. Disorders of Communication, 1964, 42, 41-55.

HALL, E. T. The hidden dimension. New York:Doubleday, 1966.

HEAEN, G. Leadership and the spatial factor in smallgroups. Journal of Abnormal and Social Psychol-ogy, 1957, 54, 269-272.

JAMES, W. T. A study of the expression of bodilyposture. Journal of General Psychology, 1932, 7,405-437.

KENDON, A. Some functions of gaze-direction in so-cial interaction. Acta Psychologica, 1967, 26, 22-63.

LEIPOLD, W. D. Psychological distance in a dyadicinterview. Unpublished doctoral dissertation, Uni-versity of North Dakota, 1963.

LITTLE, K. B. Personal space. Journal of Experi-mental Social Psychology, 1965, 1, 237-247.

LITTLE, K. B. Cultural variations in social schemata.Journal of Personality and Social Psychology,1968, 10, 1-7.

LOTT, D. F., & SOMMER, R. Seating arrangementsand status. Journal of Personality and Social Psy-chology, 1967, 7, 90-95.

LOWEN, A. Physical dynamics of character structure:Body form and movement in analytical therapy.New York: Grune & Stratton, 1958.

MACHOTKA, P. Body movement as communication.Dialogues: Behavioral Science Research, 1965, 2,33-65.

MEHRABIAN, A. Orientation behaviors and nonverbalattitude communication. Journal of Communica-tion, 1967, 17, 324-332.

MEHRABIAN, A. Inference of attitude from the pos-ture, orientation, and distance of a communicator.Journal of Consulting and Clinical Psychology,1968, 32, 296-308. (a)

MEHRABIAN, A. Relationship of attitude to seatedposture, orientation, and distance. Journal of Per-sonality and Social Psychology, 1968, 10, 26-30.(b)

MEHRABIAN, A., & FRIAR, J. T. Encoding of attitudeby a seated communicator via posture and posi-tion cues. Journal of Consulting and Clinical Psy-chology, 1969, 33, in press.

NACHSON, I., & WAPNER, S. Effect of eye contactand physiognomy on perceived location of otherperson. Journal of Personality and Social Psy-chology, 1967, 7, 82-89.

REECE, M. M., & WHITMAN, R. N. Expressive move-ments, warmth and verbal reinforcement. Journalof Abnormal and Social Psychology, 1962, 64,234-236.

REICH, W. Character analysis. (Trans, by T. P.Wolfe) New York: Orgone Institute Press, 1945.

ROSENFELD, H. M. Effect of an approval-seekinginduction on interpersonal proximity. Psychologi-cal Reports, 1965, 17, 120-122.

ROSENFELD, H. M. Approval-seeking and approval-inducing functions of verbal and nonverbal re-sponses in a dyad. Journal of Personality and So-cial Psychology, 1966, 4, 597-605.

SCHEFLEN, A. E. The significance of posture in com-munication systems. Psychiatry, 1964, 27, 316-331.

SCHEFLEN, A. E. Stream and structure of communi-cational behavior. Eastern Pennsylvania Psychia-tric Institute, Behavioral Studies Monograph No.1, 1965.

SCHLOSBERO, H. Three dimensions of emotion. Psy-chological Review, 1954, 61, 81-88.

SOMMER, R. Small group ecology. Psychological Bul-letin, 1967, 67, 145-151.

WIENER, M., & MEHRABIAN, A. Language withinlanguage: Immediacy, a channel in verbal com-munication. New York: Appleton-Century-Crofts,1968.

(Received July 10, 1968)