Thc ptpcr urcd ln thir publicrtion meetsthc minimurn rcquircmcntrof \-Y/l-r-iltt Amcrican National Standard for Information Sciences- pcrmancncc of Gestureand Papcr for Printed Library Materids, exsr 239.4E-1984. the Dynamic Dimension orfr Essaysin honor of DavidMcNeill

Editedby SusanD. Duncan

JustineCassell Northwestern University

ElenaT. Levy Library of CongrcssCrtaloging-in-Publicatlon Dete University of Connecticut - Stamford Gcstureend the dynamic dimensionof tanguagc:cxays in honor of David McNeiIl / edit€dby SusanD. Duncan,lustine Casscll,Elena Lwy. p. cm. -- (Gesturestudics, rssx 1874-6829 ; v. r) Includesbibliographical references and index, r, Gesturc.L McNcill,David, IL Duncan,Susrn D. III. Cassell,tustine, 196o- Mcvy, ElenaTcrry, r95z- PrrZ.Gr6Sj 2oo7 8oE.5--dczz zoo7orr245 rsBN978 go zTzz8'4tr (Hb;alk. paper)

@2oo7 - fohn BenjaminsB.V No part of this bookmay be reproducedin anyform, by print, photoprint,microfilm, or any othermeans, without wrinen permissionfrom the publisher.

fohn BenjaminsPublishing co. . P.o. Box 36224' lo2o MEAmsrerdarn . The Netherlands fohnBenjamins North America. P.O. Box 27519 . philadelphia nr rgug-o5r9.usr fohn BenjaminsPublishing Company Amsterdam/ Philadelphia Face'to-faceDialogue rs a Micro-smial Contert The Exampleof Motor Mimicry'

JanetBavelas Univcrsityof Victoris

Fsce'lo-frcc dirl ue procccdsmorr|Gr by t|EErcnt,ts O|c Frticipdnts clnst&trly and prcciscly rsspondio

A diversegroup of scholan has proposedthat face-to-facedialogue is the basic or fundamentalsite of la.nguagcuse (e.g., Bavelas, 1990; Bavelas,

' I am plesscdto acknowledgeboth the manycollaboralon cited herEinand researchgrrnL9 ftom theSociol Sci€nccs and Hurnanilies R€scrrch Grant ofCaneda. I28 JANETBAVELAS

Hutchinson.Kenwood. & Matheson,1997; Clalk, 1996;Fillmore, l98l; Garrod & Pickering,2004;Goodwin, I98l; Levinson,1983; Linell, 1982,2005). Face-to- face dialoguemust havebeen the first languageof the earliesthumans; it is the infant's first languagedevelopmentally; and it remainsthe languageof everyday interactions.A corollary assumption,which I share with most of the above authors,is that, unlike written forms of languagcuse, face-to-facedialogue includesboth audibleacts (words and their prosody)and visibleones (such as co- occurring hand and other bodily ,facial displays,and gaze) that arc complementaryto or evenmomentarily replace words (Bavelas & Chovil' 2006). One importantfeature of face-to-facedialogue is its affordanceof micro' social inleraclion, that is, a high level of reciprocityand mutual influence. It representsone end ofa continuumofthe probabilityand speedofa responsefrom the other person.In publishedtext, for exampl€,therc is a low probability and high latencyof response;if the readersrespond to the writ€r at all, it is long after the act of writing. Lenen or especiallyemail are faster,and both are more likely (but not certain)to receivea reply. In a telephonedialogue, exchanges can occur rapidly, and even a momentaryfailure to respondis noticeable.In face-to-face dialogue, responsesare highly probable and extremely fast; frame-by-frame microanalysisreveals that addresseesoften providesin rltaneousfeedback lo ttLe speaker(e.g., Bavelas, Coates, & Johnson,2000).They nod or say "Mhrn" and constantlyconvey information to the speakerby their ever-changingfacial displaysof anentiveness,confision, alarm, or amusement,among many others. Theseresponses often overlapthe speaker'stum (Goodwin, 1986),but they are not considered interruptions. Indeed, such reciprocal responses are demonstrativslyessential to the speaker,whose narrativefahers when they are absent(Bavelas €t al., 2000). Thus, becauseof its reciprocity and precision, language use in face-to-facedialogue is not simply absractly social. A participant'scontribution does nol originateautonomously in his or her mind (or from "language"as an abstraction)and doesnot evaporateinto a socialvacuum. Rather, each contribution is part of a social interactionat the micro-level; fortunately, with video technology,these essentialdetails are also directly observableat thatlevel. There is accumulatingexperimental evidence for micro-socialeffects on both verbal and nonverbalbehaviours. For example,Clark and Wilkes-Gibbs (1986) showedthat speakersoften usedverbal references ihat the addresseehad helpedto shape.Schober and Clark (1989) then demonstratedthat tbeseterms were not as clear to overhearerswho did noa participatein their moment-by- moment creation.Other experimentshave shown that when speakersbecame aware that they did or did not have commonground with their addressee,they immediatelyadjusted their verbal descriptions(Isaacs & Clark, 1987)or hand gestures(Gerwing & Bavelas,2004).Evidence from McNeill's lab has revealed other micro-socialeffects on handgestures, which were formedin relationto the other person's spatial perspective(Ozytirek, 2000; 2002) or previous (Furuyama,2000). ln Clark and Krych's (2004) experiment,addressees often MICRGSOCIALCONTEXTI29

began to gesture relevant actiotlf during the speaker's verbal instructions; speaken intemrpted or changedtheir insEuctiofls,even in mid-s€nlence,in order to confirm (or correct) the addressee'sproposed action. Some gestureshave interactiverather than referentialfunctions (Bavelas, Chovil, Lawrie, & Wade, 1992) and are specializedfor the coordinationof face-to-facedialogue, where they haveprcdictable micro-effects on the addrcssee'sncxt act (Bavelas,Chovil, Coates,& Roe, 1995).Facial displays can also have micro-socialfirnctions. For example, Brunner's (1979) microanalysisconfirmed that addresseesoften preciselytimed their smilesto speakers'ufierancrs so that thcscsmiles served as back-channelresponses, just aswords 8nd nods do. In this chapt€r,I will illustratethe urility and importanceof a micro-social perspeclivewith a programof rcsearchon lhe historicalphenomenon of 'motor mimicry', which is Allport's (1968) term for a reactionby an observerthat is appropriale to the situation of the person bcing observedbut noaappropriate to the observer'sown immediale situation (e.g., $,incing at someoneelse's injury). Allpon had pointedout tbat "the linl€ understoodt€nd€ncy to elem€ntarymotor mimicry" (p.29)was still a "riddle in socialpsychology" (p.30). Previous theories had cast motor mimicry as "basicaUya perceptualmotor re€ction"(p.32), that is, within an individual framework.It tumed out that a solutionto Allport's riddle was to place motor mimicry in its micro.socialcontext. B€forc describingthis ,thougb, I will inuoduce the secondarytheme of this chaptcr, which is the apparentdifficulty tbat many researchershave in noticing the micro-social context of languageuse or in remembring it even affer it has b€en demoostratedas a solution.The next sectionillustrat€s this sub+hemeby analogy.

L2 Self-imposedLimits on Observation

The microsocial perspectiveseems elusive in the study of languageuse, evenand perhapsespecially in socialpsychology (Cla*, 1985;Bavelas,2005). I 'nine-dot proposethat its fate is analogousto the classic problem' in perceptual ,which startswith a 3-by-3matrix ofpoints: ooo aaa oao

The problemis to draw a line throughthe middle of eacbof the dots only once,using four sbaightlines that are all connected.In otherwords, use four lines to connect all of the dots without lifting the pen or pencil from thc paper and without retracingthrough any of the dots.(The solution can bc foundon lhe wcb or in a reviewanicle suchas Kershaw& Ohlsson,2004.) 130 JANEIBAVELAS

To connect dre nine dots as required, one must think and act outside the appaFentsquarE or box that their configuration suggests.Tbe outer dots are not a borderor limit, but mostofus initially imposeone ourselvesand try to solvethe puzzle within a self-limited space,rather than using the spaceourcide it. Similarly, evenwhen ostrensibly studying language use, many researchers still operatewithin thc borders of the individual, with a self-imposed limit that excludes the micro- socialcontext of which the individualis part at any given moment.The restof this chapterwill be devotedto th€ possibility oflooking ouside rhe conceptuallimit of the isolatedindividual. There is an additional reasonto use the nine-dot pmblem as an analogy bere, namely,its remarkablercsistance to insightful solution.Chronicle, Omerod, and MacCregor(2001) have shown that verbaland visual hints raisethe successmte only slightly. Moreover,even when individualshave once seen or drawn the solution themselves,they may not be able to solve the pmblem later. (The reader who had previouslyknown the solutionmay havehad the sameexperience here.) I propose that the micro-social perspectiveon human behaviour faces a similar recalcitrance.Even when evidencereveals it as a solution,it seemsto be quickly forgotten,and the focuson the individualtakes over again.In brief, the nine-dot problemis rcally two problems:At first, it's hardto seethe solution. Then.even once you've seenthe solution,ir doesn't srick. We will see how this analogy works for the micro-socialaspects ofan illustrativecase.

2, Motor Mimicry is I Micro-socirl Act

2.I Backgroundand definition

When . our researchgroup approached Allport's riddle in the early l9g0's,we had the considerableadvantages of experimentalmethods and video technology (Bavelas, Black,Lemery, Maclnnis, & Mullett, l9g6a)so rhatwe werenot Iimired to descriptionsor still photographsof motor mimicry. Becauseof rhe capacityto MrcRo-soctALcoMExr l3l

observeclosely and repeatedly,we began to recognizethat the individral's reaction dependedon the micro-social context, io that ir pas nnre likcly lo occw whenanother person would see rr We thenhypothesized that motor mimicry was a communicativeact, skilfirlly and efliciently conveyingunderstanding of the other'ssituation.

2.2 Ffustexperiments

Our first firll experiment(Bavelas, Black, Lemery, & Mullen, 1986b)tested the communicativehypothesis by varying whetheror not anotherperson would see the panicipans' motor mimicry, that is, whether there was a receiver for any mimetic display the participant might make. The experimentwas explicitly designedto challengethe existingindividual explanations:

lf moto. mimicry is communic.tive, then ... dle prcbability of ih bcing sccn by a receivershould affect the sender'sdisplay offacial mimicry... . lf, on the othcr hand, such mimcsis reflccts c*scntially a pivatc cxgcricnce rhar happcns to rGsuh io ovcrt nonverbal bchavioa, thcn ... thc pnesencco. abscnce ofa rccciver should have no effect.(1986b:323).

In a highly controlled +second interval (described in detail below), the experimenter(El) apparentlyinjured his finger and then either did or did not make eye contact with the obscrving participant.This independentvariable required a carefully rehearsedsequence lhat began with the participant waiting while El and his assistant(E2) carrieda heavyTV monitor into the room to a 's table nearthe participant,where it slippedand droppedon E I finger.Over the next few seconds,the two exp€rimentalcooditions began to diverge:either the probabilityof eye contaclbetwcen E I and the participantincr€ased until it finally occurredor theprobability decreased and it did not occurat all. We took grcat care that the only difference beiweenconditions in this brief periodwas eye contact(or not). Becausethe experimenter'sface would ultimately be visible only in the eye-contactcondition, it was essentialthat his expression was not the cue to which the peaiciPantwas rcsponding(which would be a fatal confound in the design; p.325). Therefore,El's face initially beganto show visible pain in both conditions,and he also indicatedhis injury by a sharpintake of breath and body-acring.Then, al the point in the sequencewhen the pafiicipant in the no-eye-contactcondition would oot seeEl's face,his expressionwas blank in the eye-contactcondition. At the maximurnpoint of eye contact'El displayed no pain facially. Thus, the only differencebetween conditiors was whetherEl and the participantmade eye contact,that is, whether motor mimicry by the paaicipantwould be seenby a receiver(Et). Here is the proceduraldescription from the publishedarticle (1986b:324): I32 JANETBAVELAS

Evccontact condirion No cvc conBct condition

Injury ard iraat. of bcat!; hcc Iajuly ind inlakc ofbre.ft; face beginsto showPain. beginElo show pain. Bringshead up andglaDces at Hunchesdown over TV, with obscrvcrwilh defocus€deycs facc visiblc .o obEerverin as headrolls back. Fofilc.

Two sccordsaffcr thc sbn ofthc iljury: 's E2 litu TV oft EI's hand. E2 lifls Tv off E | hand. El pivotsfully rcwaFdobserver, El pivos tully towud E2, in scmi{roucll holdinghis in scmi-crouch.holding his h8nd.Looks at ha!d, lhen hand. t ooks at hard, then directlyat obs€rverfor I second directly at E for I ssccond- wi$ "blank"fac€.

Our dep€ndentmeasure was any expr€ssionof pain by the obsewer (the experimentalparricipano; it did not needto be tit€rallywhar El haddone, as long as it was appropriateto his situation.Microanalysis of the videotapesrevealed a clearresult: If the participantshad eye contact with E I , they typicaltywinced and also made a kaleidoscopicsequence of other pain displays in resPonseto his injury (e.g., knitted eyebrolvs,sharp intake of breath, vocalization such as "ouch"). At the point of maximumeye contact,when El looked fully at them, they madea very distinctdisplay. The participantswho saw the sameinjury (and the sameinitial facial displayofpain) typicallyeither displayed no motormimicry at all or startedto rcspondand quickly ceased-At their minimally social point, when El tumed to int€ractwith E2, few displayedany motor mimicry. The statisticallysignificant difference between experimental conditions in participants' motormimicry wasa micro-socialeffect of eyecontact. These results eliminated other, individualistic explanationsfor motor mimicry such as imitation and vicariousemotion, b€cause neither of tbesecaD accountfor the differencebetwe€n conditions, that is, when motor mimicry did no, occur. lt is importantto empbasizethat we did not deny that emotion may have occuned, only that it did not produce the motor mirnicry. In ov Wallet- process theory (Bavelas et al, 1986b:328),the intemal and communicative processesare indep€ndent,and it is the micro-social,communicstive process that determines the overt trhaviour. Thaa is, observers may or may not have experiencedvicarious pain in ieaction to the injury they saw, but this hypothesizedintemal state did not automatically r€sult in overt mimesis, because obsewerswho witnessedthe same injury without cye contact were virtually impassive.The evidencewas clear: The display of motor mimicry dependedon the presenceof a recipient;it was meantto be seen-Only when we includedlhe micro-socia[dimension-in this case,the possibility of immediatecommunication to anotherperson--{id th€ motormimicry becomeintelligible. In the secondpart of lhe experiment,another group of panicipants rated the videotapedreactions of the previous participants.These raters received a MICRGSOCIALCONTEXT I33

descriptionof El's injury, but his imageon the splir-screenvideo was coveredso that they would be unawareof the eye-contactvariable. Their instructionswere to rate eachperson they saw on video for "the extent to which the face exprcssedthat the person knew how the experimenter felt . . . [and] cared about what bad happenedto the experimentei'(Bavelas er al., 1986b:327;emphasis original). The facesof participantsin the no-eye-contactcondition were rated as significantly less aware of and less caring about El's injury than those in the eye cootact condition.(Note that thesewere outsidem'ratings of the facial expressions;we haveno informationon how the participantsactually felt.) Takenlogether, these two studies suggestedthat motor mimicry was both encodedand decoded as a brief but importantinterpersonal message. The nexaset of experimentsfocused specifically on encodingand decoding, in ord€r to demonstratethat micro.social factors determinednot only the occunencebut the very form of motor mimicry. Bavelas,Black, Chovil, Lemery, and Mullett (1988) examinedclosely the motor mimicry that was most often describedby the classicalautho$ cited above, namely,an observerwatching someoneelse who is leaningor duckingto one side.If the observeralso leansor ducks,this would be consideredmotor mimicry regardlessof whetherit is to the right or the left. However,our micro-socialanalysis proposed that thedirection it which the observerleans is theoreticallyinformafive. To explain,I'm going to usearbitrary pronouns for the two people:The first person,who initiatesthe sequenceby leaningor duckingto oneside (e.g., to avoid or reachfor something),will be femalewhile the personwho mimics her action will be male. lf the first personleaned to her left and the observer,who is facing her, leanedto his left as well, he would be doing exactly what she had done, which fits individualistictheories of motor mimicry such as literal imitation, taking the role of the other, or vicarious action. From a social or dyadic perspective,however, this looksodd becauseits effect is that the two of themare moving in oppositecompass directions:

To appearto be moving wil, the first person,the one facin8her mustlean or duck in the oppositedirection. lf sheleaned left, andthen he would leanright, which is in the samecomoass direction: I34 JANETBAVELAS

We proposedthat it is easierto sec lhis direction as mimetic becauseit looks more "together--€ perceptualr€lationship that H€ider (1958:200-201)called a "unit relation." In our first experiment, each participant was facing the experimenter,who told a story in which sheillustrated ducking away to avoid someone'selbow. As predicted,vimrally all of the participantswho duckedmimetically did so in the same compass dircction, rather than in the opposite compass direction. The participantsencoded their motormimicry in the mostintelligible form by ducking 'with' the storyteller. Next, in order to test the effectsof th€ two forms of mimetic leaningon viewerq we createdseveral descriptions or photographsof situations in which one personwas leaningor ducking (e.g.,to avoid a squashracquet in the face or to reach a coffee mug on a side table) and the person facing was also leaning. There werealways two vemionsofthe sarnescene: the secondperson was either leaninS, in the same or the oppositecompass directions. krge samplesof participants were significantlymore likely to choosethe venion in which the observetwas leaning in the same compassdirection as the one in which the obsewer appearcd to be more "involved" and to be acting "togeth€r with" the other person. Altogether,the resultsin this article (Bavelaset al., 1988)led to the conclusion thal lhe lorm of this kind of motor mimicry was shapedby is micro-social context,that is, the mimicry took the form thatwas mostreadily intelligible to the other person.Even the b-nsitory actionsof ducking or leaningwere setrsitiveto lheir appearancein relationto the otherperson. This was anotherstep outside the boundaryof the individual,but our experimentsstill had linle to do with face-to- face dialogue.

2.3 Motor mimicry in dialogue

As shownin our first experim€nt(Bavelas et al., 1986b),many expressions of motor mimicry are facial,for example,the observermay winceor look snrtled whenthe otherp€rson is suddenlyinjured. Kraut andJohnsron (1979) had pointed out that, in over I00 years of studies of facial expressionsince Darwin (1872i1998),almost none had examined its social functions. Indeed, most experimentsstudied individuals alone or with a non-reactiveinterviewer, in order 'true' that social factorswould not obscuretheir emotionalexpressions. As with the nine-dotproblem, there had beena self-imposedlimit to studyingthe facesof isolatedindividuals-yet without socialdata, an altemativetheory coutd hardly arise.Kraut and Johnstonalso introduceda us€fu|theoretical distinction between studying a facial configuration(e.g., smiling) as an emotional 'expression'of somcpresumed intemal state (whicb may be independeotofsocial context)versus 'display' studyingit as a social directedat others(which may be independentof internal statc). In our lab, Chovil (1989; l9l; l99l-92) began to garher systematicsocial data on facialdisplays in dialogue. MICRO-SOCI,ALCONTEXTI35

One of Chovil's studies(19t9; l99l) testedwhether the addr€ssee'smoior mimicry in face-to-face dialogue is comrnunicative (versus a purely individual reaction) by manipulating whether it would be seen.There were four experimental conditionsin which an addresseelistened to a narralor'sr€El close-call story: (l) the narrator and addresseewere face-to-face;(2) the narrator and addresseewere interactingthrough a partition;(3) narratorand addrcsse€were on lhe telephone; or (4) the address€eswere alone, listening to a dramatic close-call story on an answering-machine.ln the face-to-facecondition, there was a great deal of facial motor mimicry (e.9., displays offear, pain, or alarm), and thesewere significantly lessfrequent in the othertluEe conditions. These results replicated in dialoguethe finding ofBavelas et al. (1986b)that motor mimicry requireda visuallyavailable receiver. (There is still very little experimentalresearch on non-emotional functionsof facial displaysin dialogue;cf. Bavelas& Chovil, 2006; Bavelas& Gerwing,in press.) Later, we borrowedChovil's (19E9, l9l) close-calltask in order to leam more about what'mere listeners'do in face-to-facedialogue (Bavelas, ct al., 2000). They made two different kinds of responses:As one would exp€ct, addresseesmade the familiar gmeric responses,such as nodding,"yeah," and "mhm." But they also maderesponses that wercspecific to the speaker'stopic at that moment:supplying words or phrasesand displayingprecisely frned facial or bodily gesturessuch as lookingalarmed, recoiling, or crouchingslightly-in oaher words, motor rnimicry. Listeners made these specific responsesat surprisingly high averagerates (2 to 4 per minute), skewed toward the dramatic ending of the story. We thereforeproposed that the truc home of motor mimicry is in facc-!o- face dialogue, where it constitutesimpoftant feedbackto the narrator. Two €xp€rimentstested this hypothcsisby usingan unrelatedcognitive t sk !o distract addresseesin one conditionfrom the narrative(Bavelas et al.,2000). The results showed that distracting the addrcsseesfrom the contenl of the story vinually eliminatedspecific responses such as moaor mimicry, which stronglysuggests that theseresponses require cognitive processing and are not automaticor reflexive. Moreover,the dist$cted addresseeshad a stsong€ffect on the narralor'sstory- telling, particularlyat the dramaticending, where specificresponses were morc likely to occur in the nonnal listening condition. When the addrcss€e wss distracted and made only generic r€sponsesor none at all, the narrator's story- ending fell flat aod was significantly morc poorly told than when the narrator walt talking to a responsiveaddressee. A good nanalorrcquires a goodaddressee, and a good addresseeis not diffirsely socially presentbut specificallyactive at the micro-sociallevel. The final evidence(so far) for the micro-socialnanrre of motor mimicry was an analysis of precisely how atrd when addresseestime their resPonsesto the narrator(Bavelas, Coates, & Johnson,2002). We examinedinteractions in which narretorstold their close-call stories to undistractedaddressees (the control group in Bavelas,et al., 2000, exp. 2) and found a consistentand highly collaborative 136 JANETB.AVELAS

pattem betweennarrelot and addrcss€e:The narrator would occasionally glance at the addrcssec,initiating eye contact. This is when addresseesmade their generic or specificresponses, during mutualgaze. Then the speakerlooked away again, which ended the eye contacr So the speakerstarted the sequenceby making eye contact and rhe addresseeresponded, which terminatcd the speaker's gaze. This pattemwas statisticellydifrerent fiom chancenot only for the sampleas a whole but for eachdyad. Their reciprocalcoordination was the samefor both generic and specific respoDses,that is, for motor mimicry too. Speakes and addresseesin face-to-facedialogue produced these responses together, as part ofa micro-social sequence.Note that this resultreplicates lhe original Bavelaset al. (1986b)eye- contact effec, this time with a dyad in sponlaneousdialogue. The five articles reviewed here contained I I studies all strongly suggesaing that on€ answerto Allport's ( 1968:30)"riddle" of motor mimicry is to seeit as 8 micro-socialprocess and not as an individual process.Motor mimicry is not an individual phenomonon. It depends on, is shapcd by, and in turn influences a panicular momentin dialogue,illustrating lhe fine-tunedreciprocity that is the essenceof the micro-socialcontext. However, as we will seenext, this is nol the directioniD which the lircraturehas gone.

2.4 Re-imposingthe Limits

Recall that the seconddifficulty of the nine-dor problem is that giving hins or directionsfor a soluliondoes not help much; the self-imposedlimits seemto refum on their own. The samehas been true for the aboveresearch on the micro- s@ial natureof motor mimicry. From the first experiment(Bavelas et al., 1986b) to the latestdialogue studies (Bavelas et al., 2000;2002),lhe resultshave clearly demonstraEdthat motor mimicry is part of a communicativesocial interaction at the micro-levelrather than an imitativeor emotionalresponse by an individual. Yet, as I will illustratewith a sampleofcications !o the first experiment,one finds quitea differentvenion in the subsequentliterature. Becauseof the surprisingoutcome of this surveyand becausea literatur€ reviewcannot, by definition,be anonymous,it is importantto emphasizeis very narrow scope.The sole purposeherc was to asccrtainhow the findings of one €xperimentwere represented in subsequen!research or theory,that is, to look for an overall panem in the literaure. For this reason,the focus was solely on each citing article's descriptionof the Bavelaset al- (l9E6b) cye-contactexp€riment (d€scribedin detail in section2.2), which hasbeen available long enoughto have accumulateda numberof citations.ln eacharticle rcviewed, this experimentwas typically only one citation among numcrousothers and was usually describedin a few lines; ii was never the main point of an anicle. Therefore,my comments cannotand do not apply to otherparts of any article,much lessto an entireanicle or its authors.Finally, beforejudging thosewho mis-cit€dth€ experiment,each of us should ask ourselveswhether we have ever read something through our own MICRO,SOCIALCONTEXTI37

pneconceptions.It is this all-too-familiartendency --€taying within one's self- imposed framework-lhat is my point here. The samplecame from Psyclnfo and Google Scholar,which in late 20O6 listed37 and 50 citations,respectively, ofBav€las et al. (l9E6b. I selectedthe 39 unique citations that were in refereedjoumals or book chaptersayailable in Englishand that cited this exp€rimentin the contextof a discussionof mimicry (excluding ab'strdcts,conference proceedings, articles and full books on uuelated topics,and, obviously,self-citadons). There were two clesr pattemsin these39 citations.Fi6t, there were no criticismsof the our actual experimentaldesign. Althougha handfulquestioned the generalizabilityofour conclusions,most of the citations wer€ positive rcferences to the experiment in support of the citing article'stheor€tical position. Second, only five of the 39 citations were wholly accurate in their descriptionsof the results (and the procedureif they includ€d it): Bertenthal, Longo,& Kosobud(2006:223), Fridlun d (1991:230& 237), Fridlund.Kenworthy, & laftey (1992:19l-192),Gibbs & Van Orden(203:5) and Parkinson(2005:2E2 & 297-298). Twelve other citations r€ported accurately that we had proposed a communicativeor signallingfunction of motormimicry (Bush,Ban, McHugo,& Lanzett^,1989i 49; Chartrand& Bargh, 1999:.901;Chartrand, Maddux, & Lakin, 2005:338& 350;DePaulo, 1992:221,229,&230; Dijkslerhuis & Bargh,200l:13; Hess, 2001:398;Hess & Blairy, 2O0l:129 & 139; Hess, Philippot, & Blairy, 1998:509-510;Krauss & Fussell,1996:664 & 690; bkin & Chartrand,2005:2E3 & 285; Manusov & Rodriguez,1989:16; Wilson & Knoblich; 2005:463)but includedinaccurate or inconsistentdescriptions of the variablesor procedure;see detailsbelow. The remainderwere inaccuratedescriptions of the variables, results, and/or procedure of cited experiment. Moreov€r, lhe pattem was not random: The vast majority described the experimeot as supporting one of three individual theoies of motor mimicry: automaticor nonconsciousimitation, emotional contagion, or expressionof an intemal, mental state. In effect, these theories recast the experimenteras a mere stimulus for imitation rathcr than a consciousbeing who was capable of perceiving the participant's responsein their micrGsocial interdction.ln the following, I will first cxplicate how these theoriescannot explainour resultsand thendescribe in more detail how this incongnrencesrme about. A theory of motor mimicty as automatically sctivated imitation of the other person'sfacial disptay does not explain the significanttylorver frequencyof mimicry in the no-eye-contactcondition. First, the procedural description reproduced in Section 2.2 shows tbat, in lhe initial two seconds, thc experimenler'sfacial display of pain was identicaland equally visible in both conditions, so both conditions should have automatically triggered a nesponse. Second,at the peak poinr of motor mimicry (during full eyc conlact),his facial expressionwas gone. There was none to imitate, yet this was the Point of maximum difference beueeo the two conditions.Third, behavioursthat we I38 JANETBAVELAS included as motor mimicry were much more varied thao a literally imitative facial display. ln keepingwith the historicaldefinition of motor mimicry, we counted bihaviours "indicative of pain" (1985b:324),including vocalizationsand certain"ny head movements. Fourth, the mean rcsponse time was 1.27 seconds, during which the participanastook into accountboth the injury and the probability ofeye contacqth€ir reaction was fast but not simpl€automaticity' Nor can the results be atrributed to emotional conlagion of the experimenter'spain, again becauseof the no-eye-contactcondition. Because the apparentpain ofthe injury was the same in both conditions, emotional contagion cannot explain why the participanb who witnessedthe same degreeof pain without eye conlactwere unlikely to display mimicry. Also, as reportedin the anicle ( 1986b:325),participants in the eye-contactcondition were sigoificantly more likely to include a smile mixed in with their pain displays, which is inconsistentwith feeling vicarious pain. We interpretedthose smiles as also communicative,that is, asresssuring or face-savingsmiles to the experimenter. For similar reasons,tbe intcrpretationof our participants' actions as expressionsof an intemal slate such as rapport, liking, empathy, or an emotion (presumably pain) also has no basis in the evidence. There was no self-IEPortor oth€rmeasurc of the participants'feelings. To infer the participants'intemal state iiom their motor mimicry and thenexplain the mimicry by this inferredinternal statewould be circular.The only ratingsobtained (Bavelas et al., 1986b,pan 2) were from third parties on very specific scales, not global ratings of rapport, liking, empathy,or any emotional stat€. To int€rpret motor mimicry as the expressionof an intemal state would also require us to assumethat the less reactive participants in the no-€ye-contactcondition did not feel anything when they witressed an equally painful injury. We had pointed out that the data excludedexplanations based on intrapsychicstatcs such as empathy(1986b:327) and offered insteadour parallel-processmodel, which separatesfeelings from ovcrt displays( 1986b:328). ln short, lhe evidence in the 1986bexperiment is incompatible with each of the three most common re-interpretations,primarily becauseof the significanr effect ofthe independenlvariable, which showedwheu motor mimicry would and would not occur. However, this incompatibilitydid not arise in most of the citations, mainly because6e rclevant elenFnb were abcentor transformed. This occurred in severaldiffer€ni ways: A large proporlion of the inaccuratecitarions simply did not mention an independentvariable (Aarts, Gollwitzer, & Hassin,2004:34;Bush et al., 1989:49; Fischer,Rotteveel, Evers, & Manstea420o4.-225: Garrod & Pickering 2004:10; Hess,2001:398 & zlfi); Hess& Blairy, 2OOl:129&139; Hesset al., 198:509- 510; Krauss & Fussell, 1996:664 & 590; Lundqvist, 2005:263; Marsh, Richardson,Baron, & Schmidt,2006;13; Pasupathi, Carstens€n, & Levenson, 1999:175-176;Richardson, Manh, & Schmidt, 2005:62; Russell, 2003:155; Sebanz, Knoblich, & Pilz, 20015:1235;Sonnby-BorgstrdrL 2002:433; Sonnby- Borgstrdm & Jdnsson,20O4:103; Sonnby-Borgstrom, J6nsson, & Svensson, MICRO-SOCI,ALCONTEXTI39

2QO3:4&16:' Van Swol (2003:462;Vreeke & van der Mark, 2ffi3:183; Williams, 2002:449:Wilson & Knoblich; 2ffi5:460 & t163).Most of theseciations were in support of demonstratingthat mimicry could be experimentallyelicited but, lacking an independentvariable, lhey could not describe the main purposeof the experiment,which wasto showwhen it would nol occur. Severalarticles did describean independentvariable. ln one case,only the timing of the €ye contact was incoEect (Depaulo, 1992:221). More often, a differcnt independent variable was described (Jakobs, Manstead, & Fischer, 2001:52;Lakin & Chartrand,2003334: Tiedens& Fragale,2003:559). In three cases,a small changetransformcd the msnipulationfrom one that made the panicipont's expression visiblc to tbe experimenter into one thar made the experimenter'sfacial expressionmore clearly visible as a stimulus to th€ participant (Chart-and et al., 2005:338; Dijksterhuis & Bargh, 2001:10; Niedenthal,Barsalou, Winkielman, Krauth-Gruber, & Ric, 2005:192).Recall that theexperimenter's facial expressionwas equally visible or not in bothconditions. Therewere also changesin the dependentvariables alalysed (Bush,et al., 1989:49;Chartrand et a1.,2005:350; Fischer et al.,200/.1225i DePaulo, 1992:221 & 229; Hess,2001:u100;Hess & Blairy, 7AOll29 &139t Lakin & Chartrand, 2005:283& 285; Marsh et al.,2005:13;Pasupathi et al., 1999:176;Russell, 20o3:155i Van Swol, 20O3:462;Williams, 2002:449; Wilson & Knoblich; 2005:463).The most common enor was describingvariables that went well beyond what we had measured,usually internal statessuch as rapport, empathy, affiliation,or emotionalcontagion, but therewere also a fcw instancesofdifferent mimicry (e.g., yawning or sadness).In other cases,it was not possibleto t6ll whether the error was the independent or depend€nt variable (Lundqvist, 2006:263; Sonnby-Borgstrom, 2N2:433: Sonnby-Borgstrom & Jdnsson, 2003:143;Sonnby-Borgsfi

the aniclesincluded the subsequentexp€riments that demonstsatedmotor mimicry in dialogue.)It is importantto emphasizeagain that lhis was a nanow searchthat doesnot impugnanyone's scholarship, much lesssuggest that anyone deliberately distoned our experimental results. I believe thal the authors simply read our experimentas fifting a familiar or exp€ctd panem. In the metaPhorof this chapter, it seemsthat the micro-social coDtextin the experiment was as irtelevant as thewhite spaceoutside the ninedots and was similarly ignorcd.

3, Why Don't We Soethe Micro-social?

I propose that it is primarily the bndency to s€e the individual as a natural unit of analysisthat preventsour seeingthe micro-socialcontext that surrounds and profoundlyinfluences each individual. Focusing on the individual createsan implicit border,as in the ninedot ploblem,which seemsto limit our perceptualor conceptualfield to mental rather than social processes.Even when researchers brielly stepoutside and noticewhat is happeningaround and in interactionwith the individual,the focus soon retreatsto the isolatedindividual. Both language and socialinteraction have been predominantly attributed to mentalprocesses, not just in the broaderdomains of linguisticsand psychology,but even in thosethat one might expect to be especiallyinterested in face-to-facedialogue, such as psycholinguisticsand socialpsychology. Because this choiceof the individualas the unit of analysis seems to b€ a relativ€ly unexaminedone, it is worth consideringbriefly its causesand efrects,focusing on the disciplineI know best, psychology. In an insightful analysis,Darziger (1990) has shown how the person historicallycalled "the subjecf'has beensocially construcred in psychology. He describcda

RobinsonCruso€ myh [which]made il seemcmincntly reasonable to ignore thc seningsthar had produccd thc human behsvior ... aDdto rcanributeit asa DroD€r8of individualsin isolation. (1990: lt6)

Danzigerwent on to point our that psychologistsdid not invent the Robinson Crusoemyth (althoughth€y continue lo contributeto it); westemculture is highly individualistic,placing a high valueon individualactions and attribut€s.I would add that the circumscribedindividual is appealingbecause he or she is an obvious biological unit (even though nor a viable one if mly isolated).This physical boundaryalso defineseach of us as a separateentity in lhe law and in societyat large. According to Danziger (1990), the expcrimental methods to which psychologyaspired even in its formativeyears served to reinforcethe Robinson Crusoemyth in ways that systematicallyrcmoved what I am calling the micro- socialcontext: MlcRo-soclALcoNTExTl4l

Experimental methods isolated individuols ftom the social context of their exislence and soughtto establishtimeless laws ofindividual behaviorby analogywith t]re laws ofnatural science. Shared social meanings and rclations werc automatically broken up into th€ prop€rties of sepaEte individuals []€/srJl fearur€s of an eovironmenl that was €xternal to each of them. ... Anything social becsme a matter of €xtemal influencerhat did not aff€ctthe idenrityofthe individualunder study, (l9m:18G87)

Notic€ that drawing a conceptualcircle that includesonly the individual inevitably meansthat what the other person in an interactionis doing at any momentbecomes external or environmentaland thus,at most,a stimulusfor the actions of the individual. Because face-to-facedialogue is intimately and constantlyreciprocal, treating the otherperson as extemalto the individualmakes it impossibleto seeor studymicro.social phenomena. (Th€ useofconfederates in so€ialpsychology is an aftemptto controlor eliminatethe influenceof the other person,although it is much more likely to producean artificial interactionwith significanteffects on behaviour;e.g., Beattie & Aboudan,1994). As discussedelsewhere (Bavelas, 2005), another reinforcer of the notion of the individual as a naturalunit of study has beena mistakenapplication of the principle of reductionism,bonowed from natural science. It is true that reductionismadvocates reducing complex phenomena to separatesimpler parts (Reber, 1985:623),but it does not mean always studyingthe smallestor most molecularpart. As Luria (1987:675)emphasized, it is essentialto preserveall of the "basicfeatures" ofthe phenomenonof interest.Luria pointedout thatchemists cannotstudy water by separateresearch into hydrogenand oxygen,because those elementsare below the level of the phenomenonof interest;they do not presewe its basic features.Studying isolated individuals is equallyunlikely to Predictthe propertiesof their observablemicro-social interactions. The implicit assumption of reductionismis that studyingindividuals who havebeen conceptually or even physicallyisolated from eachother will naturallylead to an understandingof their micro-socialinteractions, albeit at some unspecifiedpoint in the future-an outcomethat hasnot yet, to my knowledge,been demonstrated' What many of us havediscovered is that Permittingdyadic interaction does not, as feared, preclude true experiments,nor does conductingexPeriments necessarilydestroy social interaction.As shown in the examplesat the end of Section| .l as well as in Section2.3, it is Possibleto seta task thatwill createthe desiredvariable and then let the padiciPantsinteract fre€ly within it. Ancient

interest,the analysiscan be highly reliable,if time-consuming.We attributethis high inter-analystreliability to the fact that the analyststhemselves participate I42 JANETBAVELAS constantlyin face-to-facedialogue in their everydaylives, so they are natural experts in understanding iB nuances. In the exPerimenb described above and miny others, a focus on the micro-social dimension has been, not just Possible, but richly rewarding.

4. Epllogue

The ideasproposed here may seem far from the work of David McNeill' but thercis a dircct andcontinuing debt. His l9E5 articleon gesturelcd to an ongoing program of research on hand gestures that has belped shape the micro-social perspective. Equatty important, it led me to be able to see s€vctal kinds of so- called nonverbalcommunication in more linguistic ways; to do meaning-based analysis insteadofphysical description; and to do experimentalresearch with real dyads in real conversations.In addition, his intellectual generosity and tolerance ofideas thatare often quitedifferent from his own is an exampleto us all.

References

Asrts, H., Gollwitzq, P. M., & Hassirl R- R- (2004).C,od contagion:P.rcciving is for puisuing. Jounol ofPeoonaliry and SocialPqchologt, 87,23-37. Allporr,G, W. (196E).The historical background of modemsocial psychology. In G. Lindzey& E.Arcnsoo(E.),ftehondbo*oJnciolpsychologtPrdcd.).MenloPs*,CA:AddisorF Wesley. Bargh,J. A., & Fcrguson,M. J.(20m), Bcyond bcbaviorism: On the sutomaticity ofhigher meDtsl proc€ss. Psy'chologicalBu e,in, I 26, 925-945. Bevcl&s,J. B. (1990).Noflvcrbal aod socialaspccts of discourlc in fscc-lo f.ce intcraction.farr. t0,5-8. Bavelas.J. B-(2005). The two solitud€.:Reconciling and Lsnguagc .nd SociEl Inldaction. ln K. Fitch & R. Srndcrs(Eds.\, Handbook of langaqe ord s&ial ihteretion (pp.l?9-200).Mahwah, NJ: Erlboum. Bavelas,J.8., Blac( A., Chovil,N., Lemery,C. R-,& Mullett,J. (19t8).Form and function in motor rdmicry. Topographicevidcoce fiar Ihc primsry frmctionis cornmunicative.,tr,Dat CommunicationRetearch, I 1, 27t299. Bsvelas,J.8., Black.A., Lemery,C. R, Maclmis, S., & Mulletq J. (1986a).Experimenral mcthodsfor sndying "el€m€ntarymotor mieicry.' Jounal of NonverbalBehaior 10, to2-lt9- Bavclas,J,8., Black,A., Lcmcry,C. R., & Mulletr,J. (1986b)."l showbow you feel":Moror mimicry as a communicafivercl Journal of Perconalityand Sociol Psychologt,50, 322- 329. Bavelas,J.8., Blach A., Lcmcry,C. R., & MulleG J. (19t7). Moror mimicryas primitive €mpsthy.ln N. Eisenberg& J. Stayet (Eds.\,Enpatlry and its developnent(ID.!17-138). Cambridge:Cambridgc Univcrsity Press. Brvclas,J. 8., & Chovil,N. (197)- Facesin dialogue.In J. A. Rusrell& J.-M.Femandez-Dolg @e.), fhe psycholog/ of Iacial apr6tion (pp.3l4-3it6). Cambddgc,UK: Crmhidge UnivcrsityPrcss, Bavclai' J. B-, & Chovil, N. (2006).Hrnd g..turts and frid displaysas pen of languageusc in facc-to-facedialoguc. In V. Msnusov& M. Panerson(Eds,l, Handbookof nonverbal communication(pp.97- I | 5). Thous.ndO8t$ CA: Srgc. Bavelas,J. 8., Chovil, N., Conr€s,L., & R.oc,L. (1995). cesturcs sp.sialized for disloguc. PerEonalityand SocialPsrrhologt Bu etin, 2r, 394-405. MICRO-SOCIALCONTEXT I43

Bavelas,J. 8., Chovil, N., Lowrie, D. A., & Wad!, A (1992). Inrqrcrivc gEshrres.Dircorrrr€ Procetsg.15.469489. Bavelas,J. 8., Coates,L., & Johnson,T. (2000),Listcners as co-nsratoJs, Jowtul of PeBonolity and SocialPsycholog', 79, 941-952. Bavelas,J. B.,Coarcs, L., & Jghnson,T. (2002).Listcncr rcsponscs a9 . collaborativep.occss: The .ol. of gaz.c.Joumdl o/Comt runicdioa, 52,56580- BayelaEJ. 8., & Ccrwin& J. (in pr€ss).Convcrs.tioorl hsndgestut6 rnd facirl di+laF in f8cc- to-facc dialogtc- In K- Fiedler (Et), fuial communicotior,(pp. 2E3-308).Ncl, Yo*: PsychologyPrcss- Bavelas,J.8., Hutchinson,S., Kcnwood,C., & Marheson,D. H, (197). Using face-to-face diafoguc as a stand8rd for oth€r communication systcms, Canodian Joumol ol Comfiunicot ion. 22. 5.24- Beanic,C., & Aboudan,R. ( I 994).G€sturEs, pauscs and spcech:An expcrimcotalinvcstigstiol of thc effectsofchanging socirl conrexroo lheir pGciserernporal Elrrio',d.rips. &niotico, 99, 239-272. Benenthal,B, I., Longo,M. R., & Kosobu4A, (2006).lmitarivc rrsponsc aendencies follo$/inB observationof intransitiveactiors. Joumal o! ExperimentalPsychologt: H.rrnan Perceprio, and Perhnnance,J2, 210-225. Brunner,L. J. ( 1979).Smiles csn bc backchanncls-Jountal oJPqto,rality orrd$cial PqEholoAL 37.728-734. Bush,L. K., Barr,C, L-, McHugo,G. J., & t"anzcuaJ. T. ( 1989).Th€ ctfec{s offacial controlard faciafmimicry on subjectivcrtsctions to com.d,t totrtines.Moliwtion aad Enotion, I J,31. <7 Chattmnd,T. L., & Bargh J. A. (1999).The chamclcon effect: The -beh.vior linl and sociafintrrsction. Jo,./nal of Penonalityand SociolPsycholqt, 76,893-910. Channo4 T. L., Maddur, w. W., & t-.litr, ,. L. (2005).Beyond thc pcrE€prion+chaviorlink: The ubiquitousutility and motivadooalmoder.tors of looconsciousmimicry, ln R. R, Hassin,J. S. Ufcma+ & J. A. BargD(Eds.), fre nen'uncontcious (pq'-334-36 | ). Ncw Yo*: OxfordUniveFity Pr.ss. Chovil, N. (f989). Communicalirehnctions oJ lacial displas in contgaation- Unpublishcd docteraldissertation, Dep8fii€nt of Psychology,UniveNity of vidorie, victoria, 8.C., Canads. Chovif, N- (1991).Sociaf d€termiosrts of facbl diqlays. Jounnl of llonverbol Behoviorl5, l4l- 154. Chovif, N. (f99f-92). Discounc-oricntcdfacial displaysin conveisation.Reseorch on Lanpoge and Sociallnteroction, 25, t63-194. Chroniclc,E. P.,Ornemd, T. C., & Mac€regor,J. N. (2001).Whcn insight ju$ won'tcome: The failure of visual cues in the oine-dot problem. Suarle y Jownal o! Eryterinental Psycholog,5a. 1-919. Cladq H. H. (t985). Languateusc and larguagGusr6. ln G. Lindzey& E" Aronson(Eds'), Ire hondbot olsrml psyclologr (3rd ed, Vol. 2, pP.l79-23l). New York: Harpctaad Row. ClarfqH. H. (19961.Using language. Cambridge, UK. Crnbridge Univctsity Pte$s clad<, H. H., & Krych, M. A, (2004).speaking whilc monitoriogsddrr$sees for undcFlanding. Journal of Memo,,^ and Language,50, 6241 . Cla& H. H., A Wikes-Gibbs"D. ( l9t6)- Rafering .s a collaborativeProc€ss. Cogt ition, 22' l' 39. Danziger, K (lgqD. Constructitrg,he flbjnr. Historical origins oI ps1'cMogical research- Cambridgc,UK: CambridgcUniversity Pt6s Darwiq, C. (6gq. ne expres-siono1 tt e enoriov in mon ond animals (yt ed) Lo oni Harp€rcollins.(O.i8inal wo* published1872). DePaufo,b. M. (1992).Nonvsrbal bchavior aod sclf-Pres€nt.tion.Pslcholo4ical Bulletin, III' 203-243. Dijkst€rhuis,A. (2005).Why we ate socialaniEats: The high roadlo imitatior as socislgluc ln S' Hurfcy& N- ChaterlEds), Pertpcrivs on initotion: Fmm neuroscieneto tociol science: I44 JANET BAVELAS

Yol.2: Imitation, human develoPmenl, and cuhrre OP,207'2201. Cambridgc, MA: MIT Prcss. Dijksrerhuis,A., & Bsrgh J. A (2001).Thc PercePtion-bchaviot cxpressway: Aulometic eff'Glsof social p€rceplionoo social behavior.In M- P. Z.nna (Ed-)' Atwnees in Experinentol SocialPslcho logt, 31, l 4O. Fillmorc,C. J. (1981). and thc descripiooof discouse.In P. Cole (Ed.)' Radtcal pmg'r,orics(W. | 43-| 66). Nctv Yott: AcadcrnicPress. Fische;.A. H.. Rottevecl,M., Evc6. C., & Ma$tea4 A. S. R. (2004).Emotional assimilation; How wc are intlucnccdby othcrs' cmolioos.Cahien & Psyclmlogiecogt itive/Currart Psycholog of Cognition,22,22t-245 Fridlund,A. J. (l9l). Socidity of solilary smiling: Potcnliatiouby at| implicir audic,i,ceJoumal d Petsonalityand SocialPsychologt, 60,229-240. Frid.lund,A. J., Kenworthy,K. G.,& Jaffey,A. K. (1992)Audiencc cffec|s in affectiveimagcry: Repficalionand cxr€nsionto dysphoricinagay. Jwmol of NonvetbolBehavton 16, l9l- 2t2. Furuyam!, N. (2000). Gestursl inter.ction betwc€nthc irrslructorsnd the leamer in origami instruction.In D- McNeilt (Ed.), Languageond gesrure(Pp99-ll7). c.mb.id8ei CrmbridgeUnivcrsity Press. Gano4 S., & Pickerirg,M. J. (2004).Why i! convcrs.tionso €asy?Trcnds in Cogni,iw Sciences, 8,t-l L Ger*ing, J., & Bavelas,J. B. (2(M). Linguisricinflucnccs on gesorc's tomr. Cestute,4. 157-195. Gibb6,R.w., Jr. & vaD ordcn, G. C. (2003).A.r€ emotioncl oxpr€ssions intentional? A self- organizationalapproach. Conscio sness& Emotion,4. l-16. Goodwin, C. (1981). Conversationalorganizatiorr: lrrter&tion betweenspeaken ond hearers. NewYork AcadcmicPrcss. Goodwin, C- (19E6).Bctwecr 8nd widrio: Altemrrive s.qu€ntial treatrn€ntsof coBtinuctsatd asscssmcnts.Hlraaz ,S, udies, 9, 205-21'l. Herder,F. (1958).I/re psychologtofintetpersoLal rela,iont-Nev York: JohnWiley. Hess,U, (200f). Th€ corhmunicationof emorion.ln A. Kaszriak (Ed.r, Enotions, qualia, and consciousnest(pp.397409). Sing"lporc: World Scientifi c Publishing. Hcss, U., & Bbiry, S. (2001). Facial mimicry r||d €nlotionalcdlagion !o dymmic emotiol|al faciaf cxpr€ssionsand thet influcnce on decoding acctJtacy.h emationol Journal of Psyehopltysiolog,10, 129-l4l . Hcss,U., Philippot,P., & Blairy, S. (1998).Faci8l rcactions ro emorionalfacial cxprcssions: Afllct or cognirion?Cog"rition and Emotion, 12,509-531. lsaacs,E. A.. & Cla*, H. H. (1987). Rcfcrenccsin convesatioosherwc€n cxpcrts and novic4s. Journal of ExperinentalPslchologt: Gezeral,I16,2G31. Jakobs,E., Manstea4 A. S-R., & Fischr,A. H. (2001).Social conrcxt €ffccts on facialacaivity in a negAtiveemotional s€ftin9 Enotion,1,51.69. KershEw,T- C,, & OhlrsoaS. (2004).Mulriplc causer of difficultyin iDsight:The case of rh€ ninc-dotprobfem. Jo,,rza, of ExperimmrdlPsycholo&,: Ieomiag, Memory,dnd Cogflition, 30,3-t3. Krauss, R. M., & Furs€ll, S. R. (1996). Social psychohgicrl modcls of interF6ooal communication.ln E. T. Higgins& A. W. Kruglanski(E&.),.tociarps.rahologt: Handbook o/baticprinciples Gtp.655-701)- Ncw York:cuilford. Ktaul R. E.,& Johnston,R. E, (1979).Social Bnd cmotional meisag€s of smiling:An ethologicll tpproach.Journol ofPenonolitt, ondSociol Ps]Ehologt,37, 1539-1553, Lakin, J. L., & CbrrEand, T. L. (2001). Using ronconsciousbehavioral mirnicry to crEate affi liation andrapport. Psychological Science, I 4, 33+339. Lakin,J. L., & Charrrar4T- L. (2005).Exclusioo.nd noncoisciou.s bcluvioral mimicry. ln K. D. Wiffiams, J- P. Forgas,& W, von Hipp€t (Eds.), The social o cat!: Ostrocisnt,socnl exclusion,rcjection, ahd bu ying Gtp.279-295).Nelr York: PsychologyPress. Lcvinson,S. C. (1983)-Pragnotics. Cambridge, UK: Crmbrid8eUnivcrsiry h€ss. MICRO-SOCI,ALCONTEXT I45

Linclf, P. (19t2). The writren ldnguoge bio6 in linguittid. Dcp.rtmenl of Conmunicadon Sfudies, Universityof Linkoping, Swcdcn. LincU. P. (2005). The wrilen longuoge btat in lirrgultrics: hs nalure, origlns and lransJormation . London:Routledge. Lundqvist"L.-O. (2006).A Swedishadspt tion of the EmotionalContagion Sculc: Factor structu€ and psychometric prop€.ties, korrdinaeiaD Joumdl oIPslchologt, 47, 263-212. Luri4 A- R, (19E7). Rcducrionism in psychology. lD R. L. Cregory (Eti.I". Ihe qord corrrponion to the n ind (pp. 675-576).Oxfo.d, UK: Oxfod Univcrsityhess. Manusov, V., & Rodrigrrez, J- S- (1989). lnlcotionality bchind nolverbol m€ssag6: A pc.ceivcr's p€Gpcctive- Jounnl of Noaverbol Behovior, 13, 15-24. Marsh, K. L., Richordson,M- J., Bsron, R. M., & Schmidt,R. C. (2006). Contnsting approrchcs to perceiving and acting wiah orhets. E(orogica, Pqvcholog', lE, l-38- McNeill, D. (1985).So you lhiok gesturesarc nonvctbal?PslchologicolBu etin,92. ]50-371. Niedenthel, P. M,, Barsalou, L. W., Winkielma4 P., Krauth-Cruber, S-. & Ric, F. (2005). Embodirient in atritudes. social perceprion. aDd emorion. Personolity ond Social Psychologt Revie--,9,lE4.2l L Ozyiirek, A. (2000). The influence of addrcsseelocarion on spalial lan8uage and repltscntational g€sturcs ofdir€ction. ln D. McNcill (Ed.), longuoEe ond gesture (1,p.8t83). Cambridge, UK: CambridgeUniversity Press. 64nt*- A. QN2r. Do sp€ake.s dcsign thcir co-spoechg6tures for 6cir addressccs?Tlrc elleqs of addrcssee focation on representational gesrtnes.Jo rnal of Memon, an l Laagxage, 46, 688-704. Parkinson, B. (2mJ). Do faciol firovcmcnts .xp.css cnrotions or commuDicatc nrotives? Personalin and Sociol Ps)cholog, R?l.ierr, 9, 218-31|, Pasupathi,M., Carstensen,L. L,, Levenson.R. W, , & Gonman, J, M, (1999). Responsivc fistening in fong-manied couples: A psycholinguislic pctrp€ctivc. Jounol of Nonverbal Behavior.23, 113-19!. Rebcr, A, S. (1985). The Penguit, dic,ior@t ,oJ psycholos,^. London, UK: Penguin. Richardson,M. J., Marsh, K. L.. & Schmidt,R, C. (2005).Effccts ofvisual and verbal inlciaction on unintentional inlerpe$onal cooadinarion,Jotnrnl of Expe menral Psycholoq': Hu an Percapnon and Per.Iormorrce.J l. 62-79 Rusself,J. A- (2003). Core alTectand thc psychologicalconstruclion of cmotion. Psychological Reviev.I 10.145-172. Schober, M. F-, & Clarh H. H. (19E9). Under$anding by addressccsand overfEarffs. CoStritir€ Ps)choloL,^, I l,2l |-232. Sebcnz N.. Knoblich, C., & Prinz, W (2005). How t$'o sharea rask: Corcpt€s€tttingttimulus- rcsponse m|ppings. Jonrnol of Etperimentol Pry,cholv\,^: Humon PercePtio ond Peformance, 3 l, 1234-1245, Smirh, A. (1966). The theo\, ol morol senri,nentt. London. UK: A. Miller' (Original wo.k ?ublished1759.) Sonnby'Borgstrdm,M. (2002).Auomatic mimicry rcactionsas relstedto differcncesin lmotional ernrx,lhy. Scandinavion Joumol of Psl,chologt. 4J,433443 Sonnby-Borgs(tom,M., & Jdnsson,P. (2003). Mod€ls'of'self and dodels-of-othcrsas n laled to i:rciaf-musclc rcactioDs at difrlEnt levcls of cogni0ve co,Ithol. Scandinavion Joumal of P 'chologl,,44, l4l'151. Sonnby-Borgsrrdn,M., & Jtinsson,P. (2004). Dismissing-avoid&t psltem of arachment atrd minticry reacrions .t different levcls of infonnarion proccssing' Scoadinavian Joumal of Psvcholog),4 5. 103- | | 3. Sonnby-Bor8striim,M.. Junsson.P., & Svcnsson.O (2003). Emotionll empathy cs Llated to mimiiry reactions at diffetcnl lcvcls of informarion proces.tinS. Joumal of Noaverbal Behavtor. / /. J-aJ, Ti€dens, L. 2., & Fragale, A. R, (2003) Power moves: complement rity in dominant and submissive nonverbrl bch^uiot. 'loumal of Pe$ornlity ond Seiol Psyclalogt' dl' 558-568' 146 JANETBAVELAS v8n Swol, L. M. (2(x)3). Thc cft.rs of mnvqtal mfuroringon Dercciyld Frso6ivcncss, sgrEemeD3rrilh rs imitsior, {d recipnccity in s group di5,f;vsaiotr.Cofirraraicotion Rqqnh, 30,45148O. Vrcd(C, c. J., & van der MaL l. L- (2003). Erngty, sn ir cllrrivc &L New lde6 in Psycholog, 2I, 177-2O7. Wi[itms, A. C. d€ C. (2002).Frcirl qplssiotr ofpain: An cvoluriomry rccotnL Eehqviotd ord Brain Sci.nces,25, 4391t8. Wilsoo, M., & Knoblicb G, (2m5). ThGcssc for moror involvencnt h pcrcciviogconspccifics. PsyMogicd Eullain, I J 1, ffi473.