Professional Documents
Culture Documents
Nonverbal
Communication
Valerie Manusov
The SAGE
Handbook of
Nonverbal
Communication
FM-Manusov.qxd 6/30/2006 5:25 PM Page ii
The SAGE
Handbook of
Nonverbal
Communication
Editors
Valerie Manusov R Miles L. Patterson
University of Washington University of Missouri, St. Louis
FM-Manusov.qxd 7/13/2006 4:44 PM Page iv
All rights reserved. No part of this book may be reproduced or utilized in any form or by any
means, electronic or mechanical, including photocopying, recording, or by any information storage
and retrieval system, without permission in writing from the publisher.
For information:
The SAGE handbook of nonverbal communication / [edited by] Valerie Manusov, Miles L. Patterson.
p. cm.
Includes bibliographical references and index.
ISBN 1-4129-0404-8 (cloth)
1. Nonverbal communication. I. Title: Handbook of nonverbal communication.
II. Manusov, Valerie Lynn. III. Patterson, Miles L.
BF637.N66S24 2006
302.2′22—dc22
2006004826
06 07 08 09 10 10 9 8 7 6 5 4 3 2 1
CONTENTS
Preface xi
Valerie Manusov and Miles L. Patterson
Acknowledgments ix
PART I: FOUNDATIONS
1. An Historical Overview of Nonverbal Research 3
Mark L. Knapp
8. An Evolutionary Approach to
Understanding Nonverbal Communication 139
Kory Floyd
ACKNOWLEDGMENTS
T his volume would not exist if not for the encouragement and
direction of Todd Armstrong, Editor from Sage Publications.
Along with help from Deya Saoud and Camille Herrera, we had an easy
time bringing the The SAGE Handbook of Nonverbal Communication
into being. We are grateful for the careful copy editing work of Linda
Gray and to the always diligent Astrid Virding for guiding the manuscript
through production. We also thank our wonderful authors for their
extensive work on these chapters. Authors provided several drafts of
their chapters, responding wonderfully to our often demanding feed
back. Because there is such demand for their expertise in contributing
to a wide variety of publications, we appreciate their efforts even more.
Valerie and Miles would like to thank our respective departments—
Communication at the University of Washington and Psychology at the
University of Missouri, St. Louis—for their support throughout this
process. More important, we would like to thank our respective fami
lies for their constant support and patience: Chuck and Cameron
McSween and Dianne and Kevin Patterson.
SAGE Publications gratefully thanks the following reviewers: Mark
L. Knapp, Jesse H. Jones Centennial Professor in Communication and
UT Distinguished Teaching Professor, University of Texas at Austin;
Susanne M. Jones, University of Minnesota, Twin Cities; Ross Buck,
University of Connecticut; and Peter A. Andersen, San Diego State
University.
◆ ix
FM-Manusov.qxd 6/30/2006 5:25 PM Page x
FM-Manusov.qxd 6/30/2006 5:25 PM Page xi
PREFACE
� Valerie Manusov
University of Washington
� Miles L. Patterson
University of Missouri, St. Louis
◆ xi
FM-Manusov.qxd 6/30/2006 5:25 PM Page xii
their academic life to understanding better communication works and how it has been
the processes involved in the give-and-take studied. Our second section, “factors of
of nonverbal communication. This helped influence,” brings together work on the
us in our goal for the Handbook to provide myriad forces that help shape our use of
a path to understanding the subtleties of nonverbal communication. This section
our social interactions and our relation emphasizes the importance of biology
ships with one another. The chapters in the (Buck & Renfro Powers), evolution
Handbook emphasize the primacy of non (Floyd), personality (Gifford), age (Feldman
verbal channels in facilitating interpersonal & Tyler), sex and gender (Hall), culture
contact and regulating our social worlds. (Matsumoto), and the media (Manusov &
Unlike other Handbooks, the current vol Jaworski). Each of these chapters argues for
ume’s chapters are not meant to be exhaus the ways in which the particular factors
tive of the research in the area. Rather, work to shape the practice and meaning of
authors were given the charge of making an nonverbal communication.
argument for what is important in their The third section of the handbook,
respective areas. Thus, for example, Fridlund “functions,” follows the premise that non
and Russell call for a move away from think verbal communication serves a variety of
ing about emotions as the primary function different purposes. That is, nonverbal com
of facial displays. Robinson, in his chapter munication facilitates short-term and long-
on physician-patient nonverbal interaction term ends in our social world. These
argues for the importance of a situated, functions include sending relational mes
focused microanalytic assessment of the cues sages of intimacy (Andersen, Guerrero,
that occur in such interactions. Walther & Jones) and dominance (Burgoon &
argues that computer- mediated communica Dunbar), expressing intentions and, to a
tion is not devoid of nonverbal cues as is lesser degree, emotions (Fridlund &
often asserted but, rather, that chronemics Russell), creating and managing impressions
have always been a source of message value (Keating), deceiving others or helping us
for people communicating online. As readers detect deception (Vrij), regulating interac
work through this Handbook, they will see a tion (Cappella & Schreiber), and building
range of expertise and perspective that and reflecting rapport (Tickle-Degnen).
reflects the amazing sophistication of current These chapters discuss the complexity of
scholarship on nonverbal communication. these communicative functions and suggest
To organize the large and diverse set the importance of nonverbal cues for the
of arguments about nonverbal communica communication of fundamental human
tion, we placed the chapters into four endeavors.
primary categories. This first section, An awareness of the importance of non
“foundations,” provides an array of issues verbal cues is reflected again in our fourth
that underlie all conceptualizations of and section, “contexts and consequences.” In
research into nonverbal communication. this set of chapters, the authors work to
Specifically, chapters include the broad his reveal the ways in which particular contexts
tory of nonverbal communication (Knapp), shape and make salient certain nonverbal
parallel processes in nonverbal commu processes. They also discuss the very real
nication (Patterson), methods (Gray & implications of nonverbal behavior within
Ambady), cognitive bases (Lakin), skills these contexts. The contexts we have
(Riggio), and coordination with language focused on for this Handbook are close rela
(Bavelas & Chovil). These issues are essen tionships (Noller), education (McCroskey,
tial for understanding how nonverbal Richmond, & McCroskey), physician-patient
FM-Manusov.qxd 6/30/2006 5:25 PM Page xiii
Preface–––◆–––xiii
INTRODUCTION
The Ubiquity and Social Meaningfulness
of Nonverbal Communication
� Howard Giles
University of California, Santa Barbara
� Beth A. Le Poire
California Lutheran University
◆ xv
FM-Manusov.qxd 6/30/2006 5:25 PM Page xvi
Introduction–––◆–––xvii
health care settings, and it can also be seen 2002). Additionally, because nonverbal
to have an effect in interpersonal relation behavior deficits are central to autism, non
ships. Specifically, parents can affect mental verbal behaviors are often used as an early
and physical health outcomes in their diagnostic tool for detecting this condition
children through their nonverbal behavior in children (Bristol-Power & Spinella,
(e.g., Miller, Benson, & Galbraith, 2001). 1999). Partners’ use of nonverbal commu
In addition, marital partners can also affect nication has also been linked to better
each other’s physical and mental health recidivism outcomes for substance abusers
through their nonverbal interaction behav (e.g., Le Poire, Dailey, & Duggan, 2002;
iors (e.g., Kiecolt-Glaser, McGuire, Robles, Le Poire, Duggan, & Dailey, 2001).
& Glaser, 2002). Nonverbal communication from care
In addition to health, however, we focus givers toward their children has also been
in this section on nonverbal communication associated with better health outcomes for
and development, making the argument children. Specifically, as up to half of all
that nonverbal communication from parents are not likely to talk directly with
parents to children and teachers to children their children about sex (Jaccard, Dittus, &
also affects the development of children Gordon, 2000), it is reassuring to realize
physically, socially, and intellectually (see that parenting style can reduce the inci
also, Feldman & Tyler, this volume). Also, dence of sexually risky behavior. A sum
we look at some of the ample research mary of more than 20 studies indicates that
focusing on the role of nonverbal commu parent-child closeness (which is enhanced
nication in close relationships, centering on through nonverbal expressions of warmth)
the impact of nonverbal communication is associated with reduced adolescent preg
with relational satisfaction. Finally, whereas nancy risk through sexual abstinence, post
these important outcomes are all housed ponement of intercourse, having fewer
within important relationships, nonverbal sexual partners, and using contraception
communication can have important ramifi consistently (see Miller et al., 2001, for a
cations in nonpersonal relationships as well review). Thus, authoritative parents seem
(see also Dovidio, Hebl, Richeson, & to provide the right mix of nonverbal
Shelton, this volume; Remland, this vol warmth and parental control necessary to
ume). We focus here primarily on the ways provide their children with the tools neces
in which nonverbal cues affect people’s sary to reduce their risk of pregnancy,
judgments of one another. regardless of whether or not they are talk
ing directly with their adolescent children
about sex and the potential outcomes of
HEALTH OUTCOMES risky sex.
This relationship of greater warmth
The role of nonverbal communication in and better health outcomes holds for sub
health care has been studied extensively stance use as well. For instance, adoles
within physician-patient interactions (e.g., cents’ substance use is associated with both
Aruguete & Roberts, 2002; Rosenthal, family affection and parental control
2002; see Robinson, this volume), and (Hall, Henggeler, Ferreira, & East, 1992).
physical health outcomes of older patients Consistently, moderate amounts of parental
have been linked to the nonverbal com control and parental support are related to
munication of physical therapists (e.g., decreased illicit drug use, whereas higher
Ambady, Koo, Rosenthal, & Winograd, amounts of control and support were both
FM-Manusov.qxd 6/30/2006 5:25 PM Page xviii
predictive of decreased alcohol use (Stice, on vocalics and touch as a basis for forming
Barrera, & Chassin, 1993). These results secure attachments to caregivers. Specifically,
provide evidence for the contention that the infants’ sensitivities to vocal cues appear to
nonverbal communication of warmth, in be rooted in the need for security in that
combination with higher amounts ofcon infants perceive up and down glides in pitch
trol, is associated with better health out as providing important information about
comes among adolescents. affect and security (Papousek, Bornstein,
Nonverbal expressions of negative emo Nuzzo, Papousek, & Symmes, 1990). Soon
tions within marriages may actually be thereafter, infants use vocalizations to estab
related to mental and physical health out lish communication with their caregivers. By
comes as well. Supportive nonverbal and 8 to 12 weeks, they can coo, and by around
verbal communication is one of the resources 6 months, infants can babble (Oller, 1986).
associated with close personal relationships. These vocalizations seem to be aimed specif
Such supportive communication diminishes ically at others. Infants actually vocalize
the expression of negative emotions and more when their parents are around than
enhances health in part through its positive when they are alone (Masataka, 1993).
impact on immune and endocrine regulation Moreover, when parents respond to these
(e.g., Kiecolt-Glaser et al., 2002). If spouses vocalizations, infants engage in even more
evoke greater amounts of negative emotional vocalizations (Legerstee, 1991).
expression, their actions can stimulate In addition, infants’ distress vocaliza
immune dysregulation, which may be one of tions occur in the first 4 to 5 months (Stark,
the core mechanisms underlying conditions Rose, & McLagen, 1975) and are highly
such as cardiovascular disease, osteroporo potent and arousing signals for caregivers
sis, arthritis, Type 2 diabetes, and certain (Van Egeren & Barratt, 2004). In this way,
cancers (Kiecolt-Glaser et al., 2002). distress cries may be the most adaptive
Unhappy marriages, then, may be rife with form of communication that infants pos
negative emotional expression, which can sess. Work comparing mothers with
explain the poorer mental and physical fathers, women with men, and parents with
health of unhappily married individuals. nonparents show that, across groups,
Supportive communication may facilitate people can interpret the earliest distress
greater physical health through its ability to cries correctly (Papousek, 1989). Mothers
ameliorate negative emotions. Alternatively, are better at distinguishing distress cries
less supportive communication and the for food as opposed to those of discom
resulting negative emotions may actually be fort, however (Stallings, Fleming, Corter,
a detriment to physical health. Worthman, & Steiner, 2001). In general,
parents (regardless of culture) respond to
distress cries by holding, rocking/bouncing,
NONVERBAL COMMUNICATION singing, or talking in melodic rhythms
AND DEVELOPMENTAL (Keller et al., 1996; Papousek & Papousek,
OUTCOMES 1991). If the child is out of a parent’s reach
momentarily, the parent will begin rapid-
In addition to specific health outcomes, fire, high-pitched verbalizations with a
the influence of nonverbal communication pitch that falls by the end (Papousek,
between parents and children also reaches to Papousek, & Bornstein, 1985).
better developmental outcomes for children. Besides vocalizations, infants rely on
For instance, infants rely almost exclusively other forms of nonverbal communication.
FM-Manusov.qxd 6/30/2006 5:25 PM Page xix
Introduction–––◆–––xix
Increased gaze between mothers and infants Louiselle, Misukanis, & Mueller, 1988)
is related to more frequent vocalizations by and self-actualization (Dominguez &
both, but especially by mothers (Stevenson, Carton, 1997).
Ver Hoeve, Roach, & Leavitt, 1986). Children are likewise influenced by the
Because infants often avoid eye contact nonverbal communication of their teachers
when they are distressed, mothers of dis (see McCroskey, Richmond, & McCroskey,
tressed infants usually attempt to reestablish this volume). In a now famous study,
eye contact through increased touch, smil Rosenthal and Jacobsen (1968) compared
ing, and social play (Beebe & Stern, 1977). the expectancy effects of teachers who had
In addition, infants who use gestures more been given high expectations for students to
frequently may actually acquire language teachers who had not been afforded them.
more quickly (Van Egeren & Barratt, 2004). In this study, and the 400 or so follow-up
Likewise, children who use their hands to studies (for a review, see Rosenthal &
point relatively early (some as early as 3 Rubin, 1978), teachers communicated their
months) use more gestures overall and have expectancies to students in the classrooms
better speech comprehension (Butterworth through a whole host of nonverbal and ver
& Morissette, 1996) than children who bal behaviors. Most relevant to nonverbal
point later in their development. communication, however, teachers commu
Although we would caution against nicated in ways that foster a more positive
elevating nonverbal communication to pri climate with students for whom they have
mary or even causal status in what follows, high expectations. In a follow-up meta
nonverbal cues are, arguably, central to analytic study examining interpersonal
the disciplinary styles that parents enact. expectancy effects in the classroom across
Specifically, parents who exhibit greater 31 meta-analyses, Harris and Rosenthal
warmth and responsivity toward their (1985) found significant effects for nonver
children (in combination with moderate to bal behaviors communicating expectancies.
high control attempts) have children who Among them, eye contact, wait time, ges
generally exhibit higher achievement (e.g., tures, distance, smiles, duration of interac
authoritative parenting; Baumrind, 1996). tions, and speech rate were all significant in
This communicative manifestation of predicting positive outcomes. Of the non
warmth is composed primarily of nonver verbal behaviors included in the analyses on
bal signals (see Andersen, Guerrero, & which Harris and Rosenthal’s analysis was
Jones, this volume). Furthermore, nonver based, only touch and lean were not related
bal warmth is related to better outcomes to outcomes.
in children. In one study, for example,
children of authoritative parents (i.e., those
showing high warmth and high demanding MARITAL NONVERBAL
ness) were the most cognitively motivated, PROCESSES AND OUTCOMES
competent, and achievement oriented
(Baumrind, 1991). In another study, they Nonverbal communication does not,
were also the most intrinsically motivated however, relate only to the successful devel
(Ginsburg & Bronstein, 1993). Further, opment of children. The valence of nonver
they attained the highest math and verbal bal communication has also been associated
achievement (Baumrind, 1991). Parents with marital satisfaction and marital stability.
with authoritative styles of parenting also Gottman and Levinson (1992) argue, based
had children with higher self-esteem (Buri, on numerous behavioral investigations, that
FM-Manusov.qxd 6/30/2006 5:25 PM Page xx
successful marriages have a 5:1 ratio in toward a partner and can often include psy
terms of positive to negative behaviors. chological abuse and intentional insults.
They found that couples who displayed Nonverbal indicators of contempt include
more positivity than negativity when they eye rolling, disgust facial expressions,
spoke to each other were more satisfied, less sneers, and hostile humor. It is likely that
likely to have thought about divorce, and the intentional use of insults within conflict
less likely to have actually separated. This in marital relationships will result in dis
ratio is especially important to marriages in tress as hurtful messages from romantic
which one or both partners are distressed partners and other family members elicit
(i.e., less satisfied). Individuals in distressed greater negative feelings than those from
relationships tend to display more negative other people (Vangelisti & Crumley, 1998).
and less positive affect, and they are more Such hurt, as part of a larger pattern of neg
likely to reciprocate negative affect (e.g., ative communication, can be destructive to
Noller, 1984, this volume). Negative behav family relationships in that negative com
iors may be the most predictive of marital munication (e.g., attacking the other, defen
satisfaction (e.g., Gottman & Levinson, siveness, crying, ignoring the message),
1986; Huston & Vangelisti, 1991), with greater distancing behavior, and lower
negative behaviors being more predictive of relationship satisfaction can ensue (e.g.,
marital satisfaction than positive ones (e.g., Vangelisti, 1994, 2001; Vangelisti &
Broderick & O’Leary, 1986). This is the Crumley, 1998; Vangelisti & Young,
case, despite the fact that happier partners 2000). Stonewalling, on the other hand,
display more positive behaviors than do includes responding to an onslaught of neg
their unhappier counterparts (e.g., Cutrona, ative affect with withdrawal including flat
1996). facial affect. Its use implies that the issue is
Negative nonverbal communication not worth addressing (strategically or not)
behaviors may also predict divorce. Gottman and is not worthy of an emotional response.
(1994) argues that the consistent use of Gottman (1994) notes that men are more
what he calls the “Four Horsemen of the likely to respond by stonewalling, because
Apocalypse” can portend the destruction of they tend to be more physiologically reac
the marriage and tends to mark distressed tive during conflict and thus may feel more
couples (i.e., those with low amounts of of an intense pressure to withdraw from
marital satisfaction and high amounts of conflict situations (Gottman & Levinson,
marital instability). Gottman and his 1988). Stonewalling is one mechanism used
colleagues found that couples in distress to withdraw.
display greater expressions of criticism, In tandem with this work is research on
contempt, defensiveness, and stonewalling demand-withdrawal patterns during mari
consistently. Whereas all these displays tal conflict (e.g., Caughlin & Vangelisti,
include combinations of verbal and non 1999). A large amount of research illustrates
verbal behavior, contempt and stonewalling, that when wives want changes in a relation
in particular, are typically potent and nega ship, they are likely to make demands that
tive nonverbal messages. According to are followed by the husbands’ tendency to
Gottman, contempt and stonewalling, withdraw verbally and nonverbally (e.g.,
including their nonverbal manifestations, Baucom, Notarius, Burnett, & Haefner,
are particularly predictive of divorce. 1990; Christensen & Shenk, 1991;
To be more specific, contempt includes Gottman & Levinson, 1988). This pattern
expressions of extremely negative affect of conflict is destructive in that dissatisfied
FM-Manusov.qxd 6/30/2006 5:25 PM Page xxi
Introduction–––◆–––xxi
marriages often evidence the demand- responses to others who do not nonverbally
withdraw pattern of conflict, and these position themselves “correctly” when the
marriages frequently end in divorce (e.g., national anthem is being performed when
Heavey, Christensen, & Malamuth, 1995; pledging allegiance to the flag.
Noller, Feeney, Bonnell, & Callan, 1994; Immediacy displays may also reflect
Schaap, Buunk, & Kerkstra, 1988). One of social distance and avoidance (Burgoon &
the reasons why these marriages may end is Hale, 1984). For example, backward lean
that this particular destructive conflict pat and frowning, are inferentially rich to the
tern is not easily alterable (e.g., Jacobson, extent they are mined by us to attribute the
Follette, & Pagle, 1986). actor as likeable, respected, committed,
We have shown some means through credible, persuasive, dynamic, and so on
which nonverbal behaviors are an integral (e.g., see Zhou et al., 2002). So important is
component of the communication and the use of nonverbal immediacy to convey
meaning acquisition processes through the closeness (or distance) that expressions of
child-rearing process on to marital dynam intimacy and closeness differentiate between
ics and even divorce. These highly personal marriages that are simply enduring and
communication situations highlight some those that are enduring and the ideal happy
of the many meaningful ways in which non marriage (e.g., Cuber & Haroff, 1965;
verbal communication influences personal Gottman & Levinson, 1988). Attributional
relationship and developmental outcomes. work like this can be just as easily invoked
Nonverbal communication can also be watching soap operas or TV news inter
highly influential across more nonpersonal views as it is in face-to-face interactions (see
relationships, however, and it is to this Manusov & Jaworski, this volume).
impact that we now turn our attention. Nonverbal cues also have their optimal
levels and latitudes of acceptance, however,
and, as such, can be overaccommodated as
THE NATURE OF NONVERBAL in the case where some older people become
COMMUNICATION IN NON the unwanted recipients of patronizing talk
INTERPERSONAL RELATIONSHIPS through exaggerated intonations, oversmil
ing, and slowed-down speech rates (Ryan,
One of the most frequently studied para Hummert, & Boich, 1995). Being the recip
meters of nonverbal communication (as we ient of such communications with patroniz
have just discussed and as can be seen ers representing different people in different
throughout this Handbook) is immediacy, contexts over time is not easily discountable.
defined as those communication behaviors, It can eventually imply a debilitating mes
some visual others vocal, that “enhance sage for the receiver that he or she is cogni
closeness to and nonverbal interaction tively and communicatively now much older
with another” (Mehrabian, 1969, p. 213). and “past it.” In other words, not only can
Signaling (consciously or unconsciously) nonverbal cues engender social support and
feelings of, or the intent to become, rela relational harmony, they can also be part
tionally closer to another comes in various and parcel of the social construction of
guises and sizes, including gaze and eye aging and even demise (Giles, 1999).
contact, smiling, forward body, and vocal The social meanings of these and other
warmth (Andersen, 1979). Such stances can immediacy cues (e.g., the timing of events,
also be processed vis-à-vis inanimate touch, proximity, body movements) vary in
objects, as witnessed in some people’s their interpretive potential rather dramatically
FM-Manusov.qxd 6/30/2006 5:25 PM Page xxii
between cultures (Burgoon, 1995; see also, The use of spectacles or even sunglasses in
Matsumoto, this volume) and often appear some contexts, watches, designer-labeled
as essential ingredients of cross-cultural clothing, jewelry, and men’s gold neck
training programs and etiquette books. chains; the plethora of possible hairstyles
Indeed, the absence of such cues, such as (buzz-cuts, braids, bob); and a ball-cap
a lack of nonverbal expressiveness or turned backward all have their unique social
neutrality, can influence important social meanings and, again, sometimes stimulate
decisions by teachers, medical personnel, law passionate reactions, deep-seated emotions,
enforcement, and so forth. In his discussion and intense comment from third parties.
of the performative nature of preaching, for Facial rings, lipstick colors, eye makeup, hair
example, Robinson (1980) remarked that “a styles, body size, gang graffiti and gang signs,
pastor’s words may insist, ‘This is impor and cosmetics are but a few of the myriad
tant,’ but if his voice sounds flat and expres ways we can send messages out (again more
sionless and his body stands limp, the or less strategically) about our ideologies,
congregation will not believe him” (cited in aspirations, and group memberships, be they
Mikkelson & Floyd, 2005, p. 194). Often, ethnic, sexual orientation, political, or what
reactions to the presence or the absence of ever (see Harwood & Giles, 2005).
nonverbal cues are stereotypic, leading to Given the recent fascination with physical
inaccurate presumptions and even irrespon beauty as evidenced by current American
sible actions. television programs focusing on cosmetic
There are, of course, many other non surgery (e.g., the Swan, Makeover), and the
verbal cues including accent that people use current emphasis on physical beauty and
to categorize others into social groups, youth evidenced by such trends as Botox and
often triggering allied trait attributions and breast augmentation, it is not that surprising
their associated affect (Giles & Billings, that nonverbal communication emphasizing
2004). Indeed, not a day goes by for one of physical attraction has received a great
us (HG) when his non-North American amount of research attention. Finally, the
accent is not commented upon by strangers. use of other possessions transported or in the
Speech rate, pitch, timing, rhythm, and workplace and home—certain kinds of cell
intonation are (paralinguistic) nonverbal phones; books, newspapers, or magazines
cues people can manipulate to manage read and or displayed; suitcase style; music
impressions or in forming impressions and movies played; new technological arti
of others (Pittam, 2000; see also Keating, facts; cars; residence location and architec
this volume). Paramount among these are ture; remodels; decorations and artwork;
historically sensitive dress styles and fash and so on—can communicate status and
ion, which sometimes elicit evaluative com trait attributions (cultured, pretentious, life
ments among (and subsequent affirmation span crisis, macho, cheap, etc.). They can
from) older males about certain young also influence encounters, even those with
women’s dress modes, such as “I’m glad I strangers, through the type of messages they
don’t have a daughter these days!” Wearing convey.
a tie, jacket, or suit can signal formality and
professionalism in some work and leisure
contexts yet be construed as irreverently
♦ Conclusions
Western in some Muslim contexts. Further
more, being under- or overdressed can Nonverbal communication is of consider
cause observational concern or lament. able consequence in many aspects of
FM-Manusov.qxd 6/30/2006 5:25 PM Page xxiii
Introduction–––◆–––xxiii
social life. Numerous investigations place communicators, and the ways in which the
nonverbal communication as central to context influences nonverbal meaning and
meaning acquisition (Birdwhistell, 1970), process and creates relevant, and often
especially in communication interactions problematic, consequences, which a more
where relational communication (Burgoon complete understanding of nonverbal com
& Le Poire, 1999), emotional expressions munication can help to address.
(Boone & Buck, 2003), and impression
management (Xin, 2004) are concerned.
Moreover, nonverbal communication can ♦ References
be influential in the relationships surround
ing health care, and they can also have a
significant impact on important familial Ambady, N., Koo, J., Rosenthal, R., &
relationships. Children and their nurturers Winograd, C. H. (2002). Physical thera
display nonverbal behavior that facilitates pists’ nonverbal communication predicts
the safety and development of infants, geriatric patients’ health outcomes.
Psychology and Aging, 17, 443–452.
children, and adolescents. The use of non
Andersen, J. F. (1979). Teacher immediacy as a
verbal warmth (in combination with mod
predictor of teaching effectiveness. In D.
erate to high control attempts), for Nimmo (Ed.), Communication yearbook 3
example, appears to ensure better health (pp. 543–559). New Brunswick, NJ:
outcomes for adolescents by reducing risky Prentice Hall.
sexual behavior and substance use and Argyle, M., Alkema, F., & Gilmour, R. (1971).
abuse. In addition, teachers’ use of nonver The communication of friendly and hostile
bal behavior communicates important attitudes by verbal and nonverbal signals.
expectancies to students and affects the European Journal of Social Psychology, 2,
student’s performance on a variety of acad 385–402.
emic indicators. Moreover, nonverbal com Argyle, M., Salter, V., Nicholson, H., Williams,
munication facilitates the “health” of M., & Burgess, P. (1970). The communica
tion of inferior and superior attitudes by
marriages as well in that more positive
verbal and nonverbal signals. British
nonverbal communication and less negative
Journal of Social and Clinical Psychology,
destructive nonverbal communication 9, 222–231.
(e.g., contempt, stonewalling, withdrawal, Aruguete, M. S., & Roberts, C. A. (2002).
expression of negative emotions) predict Participants’ ratings of male physicians who
greater marital stability and satisfaction vary in race and communication style.
and even better physical health outcomes. Psychological Reports, 91, 793–806.
Clearly, nonverbal communication func Baucom, D. H., Notarius, C. I., Burnett, C. K.,
tions in important socially meaningful ways & Haefner, P. (1990). Gender differences
across a variety of relationships. With this and sex-role identity in marriage. In F. D.
backdrop accessed, we trust readers will Fincham & T. N. Bradbury (Eds.), The psy
have a more informed springboard with chology of marriage (pp. 150–171). New
York: Guilford.
which to access specific chapters and the
Baumrind, D. (1991). The influence of parenting
various theoretical models discussed in
style on adolescent competence and sub
them. Along with the Handbook’s editors, stance use. Journal of Early Adolescence,
we hope that readers will see the complexi 11, 56–95.
ties of this communication system, the many Baumrind, D. (1996). Parenting: The discipline
foundational issues into which it is embed controversy revisited. Family Relations, 45,
ded, the myriad functions that it serves for 405–414.
FM-Manusov.qxd 6/30/2006 5:25 PM Page xxiv
Introduction–––◆–––xxv
Mikkelson, A. C., & Floyd, K. (2005, February). Rosenthal, R., & Jacobsen, L. (1968).
Effective preaching: How nonverbal imme Pygmalion in the classroom. New York:
diacy influences motivation, affective learn Holt, Rinehart, & Winston.
ing, and perceptions of credibility. Paper Rosenthal, R., & Rubin, D. B. (1978). Interpersonal
presented at the Western States Com expectancy effects: The first 345 studies.
munication Association Annual Meeting, Behavioral and Brain Sciences, 3, 377–386.
San Francisco. Ryan, E. B., Hummert, M. L., & Boich, L.
Miller, B. C., Benson, B., & Galbraith, K. A. (1995). Communication predicaments of
(2001). Family relationships and adolescent aging: Patronizing behavior toward older
pregnancy risk: A research synthesis. adults. Journal of Language and Social
Developmental Review, 21, 1–38. Psychology, 14, 144–166.
Noller, P. (1984). Nonverbal communication and Schaap, C., Buunk, B., & Kerkstra, A. (1988).
marital interaction. Oxford, UK: Penguin. Marital conflict resolution. In P. Noller
Noller, P., Feeney, J. A., Bonnell, D., & Callan, & M. A. Fitzpatrick (Eds.), Perspectives
V. (1994). A longitudinal study of conflict on marital interaction (pp. 245–270).
in early marriage. Journal of Social and Clevedon, UK: Multilingual Matters.
Personal Relationships, 11, 233–253. Stallings, J., Fleming, A. S., Corter, C.,
Oller, D. K. (1986). Metaphonology and infant Worthman, C., & Steiner, M. (2001). The
vocalizations. In R. A. B. Lindblom (Ed.), effects of infant cries and odors on sympa
Precursors of early speech (pp. 21–35). thy, cortisol, and autonomic responses in
New York: Stockton. new mothers and nonpostpartum women.
Papousek, M. (1989). Determinants of respon Parenting: Science and Practice, 1, 71–100.
siveness to infant vocal expression of Stark, R. E., Rose, S. N., & McLagen, M.
emotional state. Infant Behavior and (1975). Features of infant sounds: The first
Development, 12, 507–524. eight weeks of life. Journal of Child
Papousek, M., Bornstein, M. H., Nuzzo, C., Language, 2, 205–221.
Papousek, H., & Symmes, D. (1990). Infant Stevenson, M. B., Ver Hoeve, J. N., Roach, M.
responses to prototypical melodic contours A., & Leavitt, L. A. (1986). The beginning
in parental speech. Infant Behavior and of conversation: Early patterns of mother-
Development, 13, 539–545. infant vocal responsiveness. Infant Behavior
Papousek, M., & Papousek, H. (1991). Early ver and Development, 9, 423–440.
balizations as precursors of language devel Stice, E., Barrera, M., & Chassin, L. (1993).
opment. In M. E. Lamb & H. Keller (Eds.), Relation of parental support and control to
Infant development: Perspectives from adolescents’ externalizing symptomatology
German speaking countries (pp. 299–328). and substance use: A longitudinal examina
Hillsdale, NJ: Lawrence Erlbaum. tion of curvilinear effects. Journal of
Papousek, M., Papousek, H., & Bornstein, M. Abnormal Child Psychology, 21, 609–629.
H. (1985). The naturalistic vocal environ Van Egeren, L. A., & Barratt, M. S. (2004). The
ment of young infants: On the significance developmental origins of communication:
of homogeneity and variability in parent Interactional systems in infancy. In A.
speech. In T. Field & N. Fox (Eds.), Social Vangelisti (Ed.), Handbook of family com
perception in infants (pp. 269–297). munication (pp. 287–310). Mahwah, NJ:
Norwood, NJ: Ablex. Lawrence Erlbaum.
Pittam, J. (2000). Voice and social identity. van Swol, L. M. (2003). The effects of nonverbal
Thousand Oaks, CA: Sage. mirroring on perceived persuasiveness, agree
Robinson, H. (1980). Biblical preaching: The ment with an imitator, and reciprocity in a
development and delivery of expository group discussion. Communication Research,
messages. Grand Rapids, MI: Baker Books. 30, 461–480.
Rosenthal, R. (2002). Covert communication in Vangelisti, A. L. (1994). Messages that hurt. In
classrooms, clinics, courtrooms, and cubi W. R Cupach & B. H. Spitzberg (Eds.), The
cles. American Psychologist, 57, 839–849. dark side of interpersonal communication
FM-Manusov.qxd 6/30/2006 5:25 PM Page xxvii
Introduction–––◆–––xxvii
(pp. 53–82). Hillsdale, NJ: Lawrence Journal of Social and Personal Relation
Erlbaum. ships, 17, 393–424.
Vangelisti, A. L. (2001). Making sense of hurtful Xin, K. R. (2004). Asian American managers:
interactions in close relationships: When An impression gap?: An investigation of
hurt feelings create distance. In V. Manusov impression management and supervisor-
& J. H. Harvey (Eds.), Attributions, com subordinate relationships. Journal of
munication behavior, and close relation Applied Behavioral Science, 40, 160–181.
ships (pp. 38–58). Cambridge, UK: Zhou, Q., Eisenberg, N., Losoya, S. H., Fabes,
Cambridge University Press. R. A., Reiser, M., Guthrie, I. K., Murphy,
Vangelisti, A. L., & Crumley, L. P. (1998). B. C., Cumberland, A. J., & Shepard, S. A.
Reactions to messages that hurt: The influ (2002). The relations of parental warmth
ence of relational contexts. Communication and positive expressiveness to children’s
Monographs, 65, 173–196. empathy-related responding and social
Vangelisti, A. L., & Young, S. L. (2000). When functioning: A longitudinal study. Child
words hurt: The effects of perceived Development, 73, 893–915.
intentionality on interpersonal relationships.
FM-Manusov.qxd 6/30/2006 5:25 PM Page xxviii
01-Manusov.qxd 6/30/2006 3:58 PM Page 1
PART I
FOUNDATIONS
01-Manusov.qxd 6/30/2006 3:58 PM Page 2
01-Manusov.qxd 6/30/2006 3:58 PM Page 3
1
AN HISTORICAL OVERVIEW
OF NONVERBAL RESEARCH
� Mark L. Knapp
University of Texas at Austin
◆ 3
01-Manusov.qxd 6/30/2006 3:58 PM Page 4
4–––◆–––Foundations
The gestures of which I have thus far normally with important advances in rhetor
spoken are such as naturally proceed ical thought. But the elocutionists were
from us simultaneously with our words. keenly interested in body movements and
But there are others which indicate things vocalizations, and they developed detailed
by means of mimicry. For example, you lists of the many ways a body and voice can
may suggest a sick man by mimicking be used to deliver written speeches and liter
the gesture of a doctor feeling the pulse, ary works. In his effort to improve British
or a harpist by a movement of the hands education, Irish actor Thomas Sheridan
as though they were plucking the strings. sought to restore the stature of delivery in
(chap. III, 88) rhetorical study. His published lectures on
the subject from 1756 to 1762 discuss
An early attempt to ascribe meaning to
various nonverbal signals is also found in what is now standard speech-text
Quintilian’s Institutio Oratoria. In this text, material on oral interpretation, vocal
for example, throwing the head back was expressiveness, and gestures. Words,
a behavior that was believed to assist in Sheridan argues, are not the only con
expressing arrogance and the eyebrows stituent of language. Expressions and
were depicted as showing “anger by con gestures also communicate. Indeed, they
traction . . . cheerfulness by expansion” are more primitive than words, more
(chap. III, 69, 80). natural where words are artificial, more
Attention to nonverbal behavior associ universal where words are national, and
ated typically with the delivery canon virtu more expressive of emotion than the
ally disappeared during the Middle Ages sophisticated language of words. (Bizzell
and the Renaissance when a concern for & Herzberg, 1990, p. 650)
style predominated. There was, however,
a brief treatment of delivery in Thomas Gilbert Austin (1753–1837), another
Wilson’s The Arte of Rhetorique in 1553. well-known elocutionist, developed an
In this book, Wilson discusses “how to elaborate notation system with symbols for
speak in a pleasing tone, how to gesture recording voice speed, force, variety, paus
well, and how to pronounce correctly” ing, over 50 foot movements, over 100 arm
(cited in Bizzell & Herzberg, 1990, p. 587). positions, and thousands of hand positions,
In the mid-to-late 1600s, both Wilkins and not to mention arm elevation and motion,
Fenelon complained about the quality of among other behaviors (Bizzell & Herzberg,
pulpit oratory, whereas “others offered 1990). His illustrations from Chironomia
advice on delivery for preachers and (1806) are still reproduced widely (see
lawyers, with discussions of acting, facial Figure 1.1). The exceedingly detailed nota
expression, posture, movement, gesture, tions represent a significant early contribu
projection, tone, pace, and modulation” tion toward recording and analyzing
(Bizzell & Herzberg, 1990, p. 649). nonverbal behavior. But Austin’s approach
But it was the elocution movement was not widely accepted as an effective
that comprised the next major milestone method for learning how to perform while
linking nonverbal behavior and the rhetori delivering a speech because of the stilted and
cal canon of delivery. This movement, which unnatural behavior it effected.
began around 1750 and extended into Elocutionist François Delsarte (1811–
the early 20th century, is not associated 1871) also created a system of oratory that
01-Manusov.qxd 6/30/2006 3:58 PM Page 6
6–––◆–––Foundations
8–––◆–––Foundations
versus posed facial expression of emotion, people live and the cultural context with
vocal expression, and gestures” (p. 6). which they identify. This idea was credited
Anthropologist Eliot Chapple (1940, 1949), with rebutting Hitler and others who wanted
for example, began using a mechanical to persecute people for styles of body move
device, called an interaction chronograph, ment and gesture that were believed to be
which produced an ongoing graphic record inferior and unchangeable because of their
of who talked, when, and for how long. race, culture, or ethnic group. It also served,
Among other uses, data from this device however, to delay the search for behaviors
provided an early glimpse at response that may be common to human beings
matching and interaction synchrony. throughout the world (Ekman, 1998b).
In 1930, anthropologist Franz Boas was Franz Boas, Efron’s advisor and mentor to
perhaps the first social scientist to use the many of the anthropologists who made sem
motion picture camera to generate data in inal contributions to the study of nonverbal
natural settings with the goal of studying behavior in the 1950s, went so far as to say
human gestures, motor habits, and dance. this in the preface to Efron’s (1941) book:
Davis (1979) attributes the first microscopic “The environment has such a fundamental
frame-by-frame analysis of filmed movement influence that in larger groups, particularly
patterns to Halverson (1931), a child psy in sub-divisions of the White race, the genetic
chologist. During the 1920s, an extremely element may be ruled out entirely or almost
detailed system for annotating the move entirely as a determining factor” (p. 20).
ments of dancers quantitatively and qualita The belief that one might learn a lot
tively was devised by Rudolf Laban (1926; about a person’s internal states by observ
Hutchinson, 1970). Although nonverbal ing his or her behavior was another early
researchers have not embraced Labanotation 20th-century conception that fits the
(or any other comprehensive whole-body study of nonverbal behavior well. Allport
coding system) as a way of recording ongo and Vernon (1933), for example, pro
ing human behavior, coding and notational vided experimental evidence that certain
methods are common in nonverbal research. movement patterns were congruent with
Given the preceding trends, it is not certain personality traits. German expres
surprising that David Efron’s (1941) land sion psychology during the 1930s was also
mark study of gestures employed a variety studying movement patterns and their link
of innovative methods, including (1) direct age to personality as well as various aspects
observation of natural interaction; (2) of facial expressions of emotion (e.g.,
filmed interactions, which he studied and Asendorpf, 1982; Wallbott, 1982). But
from which he elicited perceptions of naive because this research was sometimes trans
observers; (3) frequency counts, graphs, formed into Nazi racist ideology (Asendorpf
and charts; and (4) sketches of interactant & Wallbott, 1982) and was not associated
gesturing made by a professional painter. with rigorous research standards (Ellgring,
As Ekman noted in his preface to the 1981), few German emigrants to the United
reissue of Efron’s (1972) book, “Efron’s States pursued this work.
methods are unique for his time and exem The relationship of more static features
plary for ours. Rarely has such a diversity of body shape and appearance to personal
of investigatory techniques been utilized in ity traits was also of interest during this
a single study of body movement” (p. 8). time. Kretschmer (1925), a psychiatrist,
Efron’s (1941) study is also significant coded the dimensions of many body parts
as an illustration of how nonverbal behavior in an effort to identify human form with
is a product of the environment in which certain abnormal personality characteristics
01-Manusov.qxd 6/30/2006 3:58 PM Page 9
and disorders. Sheldon (1940) also believed Trager, and Hall and they, like Goffman,
that body measurements would tell us a lot Scheflen, and Ruesch, believed an under
about a person’s temperament, intelligence, standing of the structure and organization
moral worth, and future achievement. of interaction depended on observations of
Much of his work used three broad classifi co-occurring verbal and nonverbal behavior
cations of body types: the thin ectomorph, (see also Bavelas & Chovil, this volume;
the muscular mesomorph, and the fatty Robinson, this volume). The tendency of
endomorph. He gathered data for his cata researchers to develop a program of research
log of body types by photographing fresh around a single nonverbal behavior, which
man students in the nude at various colleges occurred with some frequency during the
and universities. Subsequent analysis of his 1960s and 1970s, was probably not what
personal notes showed that he drew racist these “structuralists” had in mind.
conclusions from his work (Rosenbaum, The study of nonverbal behavior as a
1995). Although studies of characteristics part of the overall structure and organiza
associated with body shape stereotypes per tion of human interaction gained consider
sisted in nonverbal research, the validity of able momentum as a result of a 1955
these perceptions has not been established. collaboration at the Center for Advanced
Behavioral Studies at Stanford University
(see reviews by Kendon, 1990; Leeds-
♦ The 1950s: Laying Hurwitz, 1987). Among others, this group
included psychiatrists (e.g., Fromm-
a Foundation
Reichman, Brosin), anthropological linguists
(e.g., McQuown, Hockett), and anthropolo
For nonverbal behavior studies, the 1950s gists (e.g., G. Bateson, Birdwhistell). Using
was a decade of significant events and schol a brief film of “Doris,” one of Gregory
arship. The works of scholars like Trager Bateson’s patients in family therapy, efforts
(1958), Birdwhistell (1952, 1970), and were made to develop a detailed frame-by
E. T. Hall (1959) are considered fundamen frame analysis of both verbal and nonverbal
tal to those who later specialized in para- behavior. Ironically, the film showed a lot of
language, body movement/posture, and Doris’ behavior and little of Bateson’s, and
space. Ironically, some of these pioneers the analysis of mutual influence, so critical
were not as much interested in understanding to the structural approach, was restricted
nonverbal behavior per se as they were in severely. The results of this collaboration
understanding the structure and organiza were never published, but the approach and
tion of interaction as a whole. Kendon and its application can be found in the work
Sigman (1996) point out, “Birdwhistell of Pittinger, Hockett, and Danehy (1960),
used to say that the study of nonverbal Kendon (1982a), Birdwhistell (1970,
communication would be like the study of pp. 227–250), and in what Scheflen (1973)
non-cardiac physiology” (p. 231). Lyons called “context analysis.”
(1972) is more direct: “The fact that there The use of the terms kinesics (Birdwhistell,
is such a complete and intimate interpene 1952) and proxemics (Hall, 1959) for body
tration of language and non-language movement/posture and spatial relations,
should always be borne in mind in consid respectively, illustrated the belief that non
ering the relationship between verbal and verbal codes had a structure similar to a
non-verbal communication” (p. 54). linguistic code. About 1968, Birdwhistell
Language was very much a part of modified his approach, going beyond a
the academic background of Birdwhistell, purely linguistic analysis and maintaining
01-Manusov.qxd 6/30/2006 3:58 PM Page 10
10–––◆–––Foundations
that some, but not all, nonverbal behavior Goldman-Eisler (1968) were exploring
had an organization like language (Kendon & spontaneous speech, pauses, hesitations,
Sigman, 1996, p. 16). Birdwhistell’s place in and speech disturbances and their relation
the history of nonverbal studies lies not so to anxiety, emotion, and cognition.1 Frank
much with whether he was right or wrong (1957) published a seminal monograph on
about the way nonverbal behavior is coded, the role of touching in human interaction in
but in the fact that his work energized schol which he said, “in many interpersonal rela
ars in several disciplines to examine body tions, tactile ‘language’ functions most effec
movement and posture systematically from tively and communicates more fully than
various perspectives. vocal language” (p. 214). At about the same
Birdwhistell’s ideas also seem to have res time, Harlow’s (1958) work was receiving
onated with Erving Goffman, who took one widespread attention because it suggested
of his courses at the University of Toronto that touching was so powerful that it may
in the 1940s. Later, Goffman (1963, 1967, sometimes be preferred to nursing for young
1971) made it clear that his observations primates. A cornerstone for later studies of
associated with “interaction order” involved how environments affect human interaction
the dynamic interplay of the totality of was established in the “beautiful and ugly
behavior, including eye gaze, body move room” studies (Maslow & Mintz, 1956).
ment, gestures, positioning, verbalizations, It was also during the mid-1950s that
and the like. Goffman identified his approach Rosenthal (2002) felt he had unwittingly
to studying interaction with the way etholo directed the participants in his doctoral
gists studied animal behavior even though dissertation to behave in accord with his
he had little use for their Darwinian inter hypotheses. After establishing the existence
pretations (Kendon, 1988). of these expectancy effects in various arenas
Although the animal studies of etholo of human behavior, he then examined the
gists Konrad Lorenz and Niko Tinbergen way these expectancies were communica
were well known in the 1950s, it was not ted by subtle (out of awareness) nonverbal
until the 1960s that human ethology and signals (Rosenthal, 1966; Rosenthal &
the work of scholars like Eibl-Eibesfeldt Jacobson, 1968). Ekman, whose later work
(1970, pp. 442–534) found their place exerted a powerful influence on the modern
within the nonverbal literature and reinvig study of nonverbal communication, pub
orated the discussion of biological bases for lished his first article on the subject in 1957:
human behavior. For many years after World A proposal for a method of sampling and
War II, human behavior was commonly recording nonverbal behavior.
believed to be almost totally malleable by cul An early reference to the label, “nonver
ture, and any data that suggested otherwise bal communication,” is also found in the
were considered “politically loaded” and work of interpersonal psychiatrist Jurgen
fodder for racist actions. For example, Ruesch (1953) when he refers to its multi-
Ekman (1998b) reports the stiff opposi meaning potential:
tion he received in the late 1960s from
Birdwhistell, Gregory Bateson, and Margaret Silent actions . . . always have a potentially
Mead when he reported research supporting twofold function: they are an implementa
a pan-cultural morphology for certain facial tion in their own right and they may stand
expressions (Ekman & Friesen, 1971; for something else, or both. This double
Ekman, Sorenson, & Friesen, 1969). meaning of actions introduces great diffi
The 1950s were also a fertile time for culties into the evaluation of nonverbal
psychological approaches. Mahl (1987) and communication inasmuch as a perceiver
01-Manusov.qxd 6/30/2006 3:58 PM Page 11
can never be quite sure when action is (Henley, 1977; and later, Hall, 1984 and
intended to convey a message and when it this volume; Mayo & Henley, 1981).
is intended for other purposes. (p. 233) In 1969, a thorough review of nonverbal
research (Duncan, 1969) and what was to
Ruesch’s later collaboration with pho become a highly influential theoretical trea
tographer Weldon Kees may be the first tise on the origins, usage, coding, and cate
book with “nonverbal communication” in gorization of nonverbal behavior (Ekman
the title (Ruesch & Kees, 1956). & Friesen, 1969; updated in Ekman, 1999)
appeared. This was a time when theoretical
issues like awareness, intent, coding, mean
♦ The Tipping Point ing, classification, and units of measurement
were very much on the academic table. Even
questions about whether the term nonverbal
As an outgrowth of the pioneering work
described the emerging field of research ade
put forth in the 1950s, scholars with inter
quately (Harrison & Knapp, 1972; Kendon,
ests in paralanguage, kinesics, and prox
1981) and questions about when nonverbal
emics continued their research. But the late
behavior should be considered “commu
1960s and early 1970s was the period of
nicative” and when it should not were
emergence for a broadly visible, res
pectable, and sustainable field of nonverbal debated (e.g., Wiener, Devoe, Rubinow, &
studies. One cannot fully understand the Geller, 1972; and later, Russell &
emergence of nonverbal studies during this Fernández-Dols, 1997).
period without considering the climate of In 1970, a freelance journalist, Julius
the times. It was an era that favored a con Fast, wrote a best-selling book that went a
cern for personal relationships and the long way toward making the subject of
kind of revelations about informal inte “body language” a topic of national interest
raction that arose out of “consciousness and recognition. Other popular books
raising” and “sensitivity” groups. Davis followed (Montagu, 1971; Morris, 1971;
(1971) said, 1977, 1985). Some of the less academic
popular accounts portrayed the “reading”
The enormous public interest in non of body language in misleading ways
verbal communications seems to be part (Koivumaki, 1975), but these books contin
of the spirit of the times, the need that ued to arouse the public’s interest in the
many people feel to get back in touch subject. By 1970, courses in nonverbal
with their own emotions—the search for communication were being offered at
the emotional truth that perhaps gets Purdue, Michigan State, the University of
expressed nonverbally. (p. 2) California, and a few other universities (see
McCroskey, Richmond, & McCroskey,
Larger societal issues and trends also this volume). Textbooks for nonverbal
affected specific areas of nonverbal study. courses also appeared suddenly in social
For example, worries about world overpop psychology (e.g., Argyle, 1975; Mehrabian,
ulation (Erlich, 1971), and the fear that 1971; Weitz, 1974) and in communication
crowded humans might act like the aberrant (Eisenberg & Smith, 1971; Harrison, 1974;
rats in Calhoun’s (1962) widely publicized Knapp, 1972).
studies spurred the interest in space, density, On top of this growing interest in the
and territory. Soon after, the women’s study of nonverbal behavior was the
movement raised questions about nonverbal increasing availability of affordable video
signals as subtle manifestations of power tape recorders. In 1972, Harrison, Cohen,
01-Manusov.qxd 6/30/2006 3:58 PM Page 12
12–––◆–––Foundations
Crouch, Genova, and Steinberg summa behavior were published during the last
rized this period by saying, four decades of the 20th century. The
breadth and depth of the literature today is
Sharp changes have taken place in the too massive for the kind of broad-based
nonverbal communication literature in literature reviews that characterized the
the past decade, and in particular, in the research done prior to 1980 (e.g., Burgoon,
last two years. A decade ago, few books 1980; Knapp, Wiemann, & Daly, 1978).2
existed; and the early works tended to be But two general approaches can be said to
speculative, anecdotal, and tentative. . . . have characterized the research during this
Major works are now emerging which, period: The structural and ethological
on the one hand, organize and synthesize researchers tended to emphasize descrip
the existing data from a variety of fields. tions of how interactions are organized,
Research programs extending over a whereas others emphasized the manipula
number of years are now culminating tion of psychological variables and/or
and the results are becoming available. nonverbal behavior to observe the effects.
Theoretical issues have become clarified, Within each of these two broad approaches,
and a range of active theories vie for sup some scholars focused on a particular body
port. Finally, methodological problems part, some examined multiple nonverbal
are being examined and, frequently, they signals, and some studies went beyond
are being solved. . . . The amount of body movements per se and focused on
knowledge has now reached a “critical “related non-word phenomena” like physi
mass”—and a general availability—so cal appearance, environmental factors, and
that even more exciting things may be the use of space.
ahead. (pp. 473–474) In the tradition of those who laid the
foundation for scrutinizing the way interac
More exciting things were ahead. In tants structure and organize their behavior,
1979, Davis said, “In the past five years Kendon (1977) showed how interactants
alone more books have been published on use their faces, eyes, and spatial formations
body language than in the preceding 95. It for various interaction functions, includ
is an idea that has found its time” (p. 51). ing kissing and greeting rituals. Kendon’s
The first issue of the Journal of Environ (2004) analysis of gestures may be the most
mental Psychology and Nonverbal Behavior thorough work available on the subject. In
appeared in 1976, and in 1980 the name like manner, the detailed analyses of facial
was changed to the Journal of Nonverbal expressions of emotion by Ekman and
Behavior. As the output of research has his colleagues are unmatched (Ekman &
grown, articles and books devoted to the Friesen, 1975; Ekman, Friesen, & Hager,
special demands of conducting nonverbal 2002). Duncan and Fiske (1977) and
research and measuring nonverbal phenom Goodwin (1981) illustrated how speakers
ena have also been published (e.g., Kulp, and listeners coordinate their verbal and
Cornetto, & Knapp, 2005; Manusov, nonverbal behavior during the exchange
2005; Scherer & Ekman, 1982). of speaking turns, whereas Condon and
Ogston (1971) described a form of interde
♦ An Outpouring of Research pendent behavior between interactants as
synchrony. Subsequent studies examined
a variety of ways interactants coordinate
Myriad studies bearing on the structure, their behavior through matching and syn
organization, and effects of nonverbal chrony (Bernieri & Rosenthal, 1999; see
01-Manusov.qxd 6/30/2006 3:58 PM Page 13
Tickle-Degnen, this volume). Eibl-Eibesfeldt 1972; see also Andersen, Guerrero, &
(1970) compared the facial expressions of Jones, this volume). Multivariable studies
children born blind and deaf with those of are also common in studies that give special
others born with hearing and sight, whereas attention to context (Feldman, 1982;
other ethologists described patterns of eye Philippot, Feldman, & Coats, 1999), such
gaze, body movement, and paralinguistic as those detailed later in this volume.
phenomena (Hinde, 1972; von Cranach &
Vine, 1978).
In the “variable analytic” tradition, some ♦ Moving On
research programs have focused on specific
body areas, such as eye gaze and mutual gaze
(Argyle & Cook, 1976; Fehr & Exline, At different points in the 2,500-year history
1987), pupil dilation (Hess, 1975), vocal of nonverbal research, studies have varied
signals (Davitz, 1964; Scherer, 1986), and in terms of subject matter, methods, and
touch (Field, 1995). Four decades of work frequency. But the disciplinary breadth of
on the face by Ekman and his colleagues are interest and the number of studies conducted
the basis of a substantial body of literature during the past 40 years have provided
on the production and perception of expres unparalleled insights. Of course, not all sub
sions of emotion (Ekman et al., 1972; Ekman areas of nonverbal research flourish at the
& Rosenberg, 1997; see Matsumoto, this same time. Areas of study like the human use
volume). Even though most studies have of space and territory, pupil dilation, and
examined the face and voice as spontaneous some others popular in the 1960s and 1970s,
emotional expressions, the work of some are currently at low ebb, whereas the study of
scholars (e.g., Fridlund, 1994) views these gestures, as one example, is currently at high
expressions as social displays that are pur tide. The recent interest in and proliferation
posefully enacted to communicate. As noted of gesture research necessitated the establish
earlier, research in “related” areas like phys ment of a separate academic journal called
ical appearance (Berscheid & Walster, 1974; Gesture in 2001. The extent to which com
Rhodes & Zebrowitz, 2002), space (Burgoon, municators can encode and decode nonver
1978; Sommer, 1969, 2002), and the envi bal behavior accurately is another area of
ronment (Mehrabian, 1976) has been suffi considerable interest (Hall & Bernieri, 2001;
cient to establish an academic niche within see Riggio, this volume). Ambady, LaPlante,
the larger field of nonverbal studies. and Johnson (2001), for example, have
Typical of multivariable approaches shown that decoders often use very brief
are studies that focus on (1) interaction excerpts of another person’s ongoing nonver
processes like the reciprocity and compen bal behavior to make judgments. These “thin
sation of nonverbal behavior (Argyle & slices” have also been shown to be reliable
Dean, 1965; Patterson, 1984) and incon predictors of behavior. Other scholars are
gruent verbal and nonverbal signals exploring the conditions under which non
(Bavelas, Black, Chovil, & Mullett 1990; verbal behavior is “automatically” (with little
Volkmar & Siegel, 1982) and (2) interac or no awareness and/or intent) encoded and
tion outcomes or goals like deception decoded (Patterson, 1999; Spitz, 1997;
(Ekman, 2001; see also Vrij, this volume), Wegner & Bargh, 1998; see also Lakin, this
power and dominance (Ellyson & Dovidio, volume).
1985; see also Burgoon & Dunbar, this The application of nonverbal research
volume), expectancies (Blanck, 1993), and to particular settings like classrooms, court
immediacy (Andersen, 1985; Mehrabian, rooms, marriages, political speeches, medical
01-Manusov.qxd 6/30/2006 3:58 PM Page 14
14–––◆–––Foundations
interviews, marketing, and cross-cultural that identifies a starting point for describing
encounters is also considered a worthy target the total interactive situation. Patterson
for current nonverbal research (Riggio & (1983), Knapp (1984), and others have
Feldman, 2005). argued for multisignal, multimeaning, inter
Coexisting with this current tendency to active, and processual approaches to non
value things on the basis of their utility in verbal research. These are all elements that
important arenas of everyday life is a wide characterize natural, human interaction. To
spread interest in the management of close the extent that future nonverbal research
relationships. This trend has set the stage assumes these features, the study of nonver
for more studies of nonverbal communi bal signals may become what Birdwhistell,
cation among friends, family, and lovers Scheflen, Goffman, and others envisioned
(Feeney, Noller, Sheehan, & Peterson, in the mid-20th century.
1999; Knapp, 1983; Manusov, 1995;
Noller, 1984 and this volume), even though
the preponderance of nonverbal research to ♦ Notes
date has been conducted with interactants
who are strangers or acquaintances.
1. The Mahl and Goldman-Eisler volumes
Despite a current surge of interest in the
cited here summarize their research programs,
biological foundations of human behavior both of which were initiated in the early 1950s.
in all the behavioral sciences, we seem to be 2. This review is similarly constrained. Not
circumspect in attributing any given behav all important research contributions could be
ior exclusively to either culture or biology. noted. See Knapp and Hall (2005) for a more
Unlike the past, there appears to be a complete review of nonverbal research.
greater acceptance that both may play a
role in the manifestation of nonverbal
behavior (Segerstråle & Molnár, 1997). ♦ References
Paradoxically, nonverbal behavior seems to
be thought of and studied as both a subarea
that complements the study of human inter Allport, G. W., & Vernon, P. E. (1933). Studies in
expressive movement. New York: Macmillan.
action and a way of approaching the study
Ambady, N., LaPlante, D., & Johnson, E.
of human interaction per se. Those who
(2001). Thin-slice judgments as a measure
study gestures, for example, view the inter of interpersonal sensitivity. In J. A. Hall &
dependence of verbal behavior and gestures F. J. Bernieri (Eds.), Interpersonal sensitiv
as natural and integral to their studies (e.g., ity: Theory and measurement (pp. 89–101).
Feyereisen & de Lannoy, 1991; Kendon, Mahwah, NJ: Erlbaum.
2004; McNeill, 1992, 2000; see also, Andersen, P. A. (1985). Nonverbal immediacy
Bavelas & Chovil, this volume). Other in interpersonal communication. In A. W.
nonverbal scholars not only examine the Siegman & S. Feldstein (Eds.), Multichannel
interrelationship of verbal and nonverbal integrations of nonverbal behavior
behavior but also embrace the role of (pp. 1–36). Hillsdale, NJ: Erlbaum.
appearance, environmental factors, space, Argyle, M. (1975). Bodily communication. New
York: International Universities Press.
and time as parts of the nonverbal domain.
Argyle, M., & Cook, M. (1976). Gaze and
In short, nonverbal studies leave no element
mutual gaze. New York: Cambridge
of social interaction untouched. University Press.
In one sense, therefore, the term nonver Argyle, M., & Dean, J. (1965). Eye contact,
bal suggests a separate piece of the interac distance and affiliation. Sociometry, 28,
tion puzzle, but in another sense it is a term 289–304.
01-Manusov.qxd 6/30/2006 3:58 PM Page 15
16–––◆–––Foundations
The handbook of social psychology (Vol. 2, Ekman, P., & Friesen, W. V. (1975). Unmasking
4th ed.). New York: McGraw-Hill. the face. Englewood Cliffs, NJ: Prentice Hall.
Duncan, S., Jr. (1969). Nonverbal communica Ekman, P., Friesen, W. V., & Ellsworth, P. (1972).
tion. Psychological Bulletin, 72, 118–137. Emotion in the human face. Elmsford, NY:
Duncan, S., Jr., & Fiske, D. W. (1977). Face-to Pergamon Press.
face interaction: Research, methods, and Ekman, P., Friesen, W. V., & Hager, J. C.
theory. Hillsdale, NJ: Erlbaum. (2002). The facial action coding system
Efron, D. (1941). Gesture and environment. New (2nd ed.). Salt Lake City, UT: Research
York: King’s Crown Press. (Republished as Nexus eBook.
Gesture, race and culture, 1972. The Hague: Ekman, P., & Rosenberg, E. (Eds.). (1997).
Mouton) What the face reveals: Basic and applied
Efron, D. (1972). Gesture, race and culture. The studies of spontaneous expression using the
Hague: Mouton. facial action coding system (FACS). New
Eibl-Eibesfeldt, I. (1970). Ethology: The biology York: Oxford University Press.
of behavior (2nd ed.). New York: Holt, Ekman, P., Sorenson, E. R., & Friesen, W. V.
Rinehart & Winston. (1969). Pan-cultural elements of facial dis
Eisenberg, A. M., & Smith, R. R. (1971). Non plays of emotions. Science, 164, 86–88.
verbal communication. New York: Bobbs- Ellgring, H. (1981). Nonverbal communication:
Merrill. A review of research in Germany. German
Ekman, P. (1957). A methodological study of Journal of Psychology, 5, 59–84.
nonverbal behavior. Journal of Psychology, Ellyson, S. L., & Dovidio, J. F. (1985). Power,
43, 141–149. dominance and nonverbal behavior. New
Ekman, P. (Ed.). (1973). Darwin and facial York: Springer.
expression: A century of research in review. Erlich, P. R. (1971). The population bomb. New
New York: Academic Press. York: Ballantine Books.
Ekman, P. (1998a). Introduction to the third edi Fast, J. (1970). Body language. New York: Evans.
tion. In C. Darwin, The expression Feeney, J. A., Noller, P., Sheehan, G., & Peterson,
of the emotions of man and animals C. (1999). Conflict issues and conflict strate
(3rd ed., pp. xxi–xxxvi). New York: Oxford gies as contexts for nonverbal behavior in
University Press. close relationships. In P. Philippot, R. S.
Ekman, P. (1998b). Afterword: Universality of Feldman, & E. J. Coats (Eds.), The social
emotional expression? A personal history of context of nonverbal behavior (pp. 348–
the dispute. In C. Darwin, The expression 371). New York: Cambridge University
of the emotions in man and animals (3rd Press.
ed., pp. 363–393). New York: Oxford Fehr, B. J., & Exline, R. V. (1987). Social visual
University Press. interaction: A conceptual and literature
Ekman, P. (1999). Emotional and conversational review. In A. W. Siegman & S. Feldstein
nonverbal signals. In L. Messing & (Eds.), Nonverbal behavior and communica
R. Campbell (Eds.), Gesture, speech and sign tion (pp. 225–326). Hillsdale, NJ: Erlbaum.
(pp. 45–55). New York: Oxford University Feldman, R. S. (Ed.). (1982). Development of
Press. nonverbal behavior in children. New York:
Ekman, P. (2001). Telling lies. New York: Springer.
Norton. Feyereisen, P., & de Lannoy, J-D. (1991).
Ekman, P., & Friesen, W. V. (1969). The reper Gestures and speech: Psychological investi
toire of nonverbal behavior: Categories, gations. New York: Cambridge University
origins, usage, and coding. Semiotica, 1, Press.
49–98. Field, T. M. (Ed.). (1995). Touch in early devel
Ekman, P., & Friesen, W. V. (1971). Constants opment. Mahwah, NJ: Erlbaum.
across cultures in the face and emotion. Frank, L. K. (1957). Tactile communication.
Journal of Personality and Social Psychology, Genetic Psychology Monographs, 56,
17, 124–129. 209–255.
01-Manusov.qxd 6/30/2006 3:58 PM Page 17
Fridlund, A. J. (1994). Human facial expression: Hutchinson, A. (1970). Labanotation: The system
An evolutionary view. San Diego, CA: of analyzing and recording movement. New
Academic Press. York: Theater Arts Books.
Goffman, E. (1963). Behavior in public places: de Jorio, A. (1832). La mimica degli antichi
Notes on the social organization of gather investigata nel gestire napolitano [Gestural
ings. New York: Free Press. expression of the ancients in the light
Goffman, E. (1967). Interaction ritual: Essays on of Neapolitan gesturing]. Naples, Italy:
face-to-face behavior. New York: Doubleday Stamperia del Fibreno.
Anchor. Kendon, A. (1977). Studies in the behavior of
Goffman, E. (1971). Relations in public. New social interaction. Bloomington: University
York: Basic Books. of Indiana Press.
Goldman-Eisler, F. (1968). Psycholinguistics: Kendon, A. (1981). Introduction: Current issues
Experiments in spontaneous speech. New in the study of “nonverbal communi
York: Academic Press. cation.” In A. Kendon (Ed.), Nonverbal
Goodwin, C. (1981). Conversational organiza communication, interaction, and gesture
tion. New York: Academic Press. (pp. 1–53). New York: Mouton.
Gronbeck, B. E., McKerrow, R. E., Ehninger, D., Kendon, A. (1982a). The organization of behavior
& Monroe, A. H. (1990). Principles and in face-to-face interaction: Observations
types of speech communication. Glenview, on the development of a methodology. In
IL: Scott, Foresman/Little Brown. K. R. Scherer & P. Ekman (Eds.), Handbook
Hall, E. T. (1959). The silent language. Garden of methods in nonverbal research (pp. 440–
City, NY: Doubleday. 505). New York: Cambridge University Press.
Hall, J. A. (1984). Nonverbal sex differences. Kendon, A. (1982b). The study of gesture: Some
Baltimore: Johns Hopkins University Press. observations on its history. Recherches
Hall, J. A. & Bernieri, F. (Eds.). (2001). Sémiotiques/Semiotic Inquiry, 2, 45–62.
Interpersonal sensitivity: Theory and Kendon, A. (1988). Goffman’s approach to
measurement. Mahwah, NJ: Erlbaum. face-to-face interaction. In P. Drew &
Halverson, H. M. (1931). An experimental study A. Wootton (Eds.), Erving Goffman:
of prehension of infants by means of sys Exploring the interaction order (pp. 14–40).
tematic cinema records. Genetic Psychology Cambridge, UK: Polity Press.
Monographs, 10, 107–286. Kendon, A. (1990). Some context for context
Harlow, H. F. (1958). The nature of love. analysis: A view of the origins of structural
American Psychologist, 13, 678–685. studies of face-to-face interaction. In
Harrison, R. P. (1974). Beyond words. Englewood A. Kendon (Ed.), Conducting interaction:
Cliffs, NJ: Prentice Hall. Patterns of focused encounters (pp.
Harrison, R. P., Cohen, A. A., Crouch, W. W., 15–49). New York: Cambridge University
Genova, K. L., & Steinberg, M. (1972). Press.
The nonverbal communication literature. Kendon, A. (2000). Introduction: Andrea de Jorio
Journal of Communication, 22, 460–476. and his work on gesture. In A. de Jorio
Harrison, R. P., & Knapp, M. L. (1972). Toward (Ed.), Gesture in Naples and gesture in clas
an understanding of nonverbal communi sical antiquity (pp. xix–cvii). Bloomington:
cation systems. Journal of Communication, Indiana University Press.
22, 339–352. Kendon, A. (2004). Gesture: Visible action
Henley, N. M. (1977). Body politics: Power, sex, as utterance. New York: Cambridge
and nonverbal communication. Englewood University Press.
Cliffs, NJ: Prentice Hall. Kendon, A., & Sigman, S. J. (1996). Ray L.
Hess, E. H. (1975). The tell-tale eye. New York: Birdwhistell (1918–1994). Semiotica, 112,
Van Nostrand Reinhold. 231–261.
Hinde, R. A. (Ed.). (1972). Non-verbal commu Kennedy, G. (1963). The art of persuasion in
nication. New York: Cambridge University Greece. Princeton, NJ: Princeton University
Press. Press.
01-Manusov.qxd 6/30/2006 3:58 PM Page 18
18–––◆–––Foundations
Kennedy, G. (1972). The art of rhetoric in the Human Communication Research, 21,
Roman world. Princeton, NJ: Princeton 456–477.
University Press. Manusov, V. (Ed.). (2005). The sourcebook of
Knapp, M. L. (1972). Nonverbal communica nonverbal measures. Mahwah, NJ: Erlbaum.
tion in human interaction. New York: Holt, Maslow, A. H., & Mintz, N. L. (1956). Effects
Rinehart & Winston. of esthetic surroundings: I. Initial effects
Knapp, M. L. (1983). Dyadic relationship of three esthetic conditions upon perceiving
development. In J. M. Wiemann & “energy” and “well-being” in faces. Journal
R. P. Harrison (Eds.), Nonverbal interac of Psychology, 41, 247–254.
tion (pp. 179–207). Beverly Hills, CA: Sage. Mayo, C., & Henley, N. M. (Eds.). (1981). Gender
Knapp, M. L. (1984). The study of nonverbal and nonverbal behavior. New York: Springer.
behavior vis-á-vis human communication McNeill, D. (1992). Hand and mind: What
theory. In A. Wolfgang (Ed.), Nonverbal gestures reveal about thought. Chicago:
behavior: Perspectives, applications, inter University of Chicago Press.
cultural insights (pp. 15–40). New York: McNeill, D. (2000). Language and gesture. New
C. J. Hogrefe. York: Cambridge University Press.
Knapp, M. L., & Hall, J. A. (2005). Nonverbal Mehrabian, A. (1971). Silent messages. Belmont,
communication in human interaction (6th CA: Wadsworth.
ed.). Belmont, CA: Wadsworth. Mehrabian, A. (1972). Nonverbal communica
Knapp, M. L., Wiemann, J. M., & Daly, J. A. tion. Chicago: Aldine-Atherton.
(1978). Nonverbal communication: Issues Mehrabian, A. (1976). Public places and private
and appraisal. Human Communication spaces. New York: Basic Books.
Research, 4, 271–280. Montagu, M. F. A. (1971). Touching: The
Koivumaki, J. H. (1975). Body language taught human significance of the skin. New York:
here. Journal of Communication, 25, 26–30. Columbia University Press.
Kretschmer, E. (1925). Physique and character: Morris, D. (1971). Intimate behavior. New York:
An investigation of the nature of constitu Random House.
tion and the theory of temperament. New Morris, D. (1977). Manwatching: A field guide
York: Harcourt Brace Jovanovich. to human behavior. New York: Abrams.
Kulp, C., Cornetto, K., & Knapp, M. L. (2005). Morris, D. (1985). Bodywatching. New York:
The description of nonverbal behavior. In Crown.
U. Ammon, N. Dittmar, K. J. Mattheier, & Noller, P. (1984). Nonverbal communication and
P. Trudgill (Eds.), An international handbook marital interaction. New York: Pergamon
of the science of language and society (Vol. 2, Press.
pp. xx–xx). New York: Walter De Gruyter. Patterson, M. L. (1983). Nonverbal behavior:
Laban, R. (1926). Choreographic. Jena, Germany: A functional perspective. New York:
Eugen Diederichs. Springer.
Leeds-Hurwitz, W. (1987). The social history Patterson, M. L. (1984). Nonverbal exchange:
of The natural history of an interview: A Past, present, and future. Journal of Non
multidisciplinary investigation of social verbal Behavior, 8, 350–359.
communication. Research on Language Patterson, M. L. (1999). The evolution of a par
and Social Interaction, 20, 1–51. allel process model of nonverbal communi
Lyons, J. (1972). Human language. In R. A. Hinde cation. In P. Philippot, R. S. Feldman, &
(Ed.), Non-verbal communication (pp. 49– E. J. Coats (Eds.), The social context of
85). New York: Cambridge University Press. nonverbal behavior (pp. 317–347). New
Mahl, G. F. (1987). Explorations in nonverbal York: Cambridge University Press.
and vocal behavior. Hillside, NJ: Erlbaum. Philippot, P., Feldman, R. S., & Coats, E. J.
Manusov, V. (1995). Reacting to changes in (Eds.). (1999). The social context of non
nonverbal behavior: Relational satisfaction verbal behavior. New York: Cambridge
and adaptation patterns in romantic dyads. University Press.
01-Manusov.qxd 6/30/2006 3:58 PM Page 19
Pittinger, R. E., Hockett, C. F., & Danehy, J. J. Segerstråle, U., & Molnár, P. (Eds.). (1997).
(1960). The first five minutes: A sample Nonverbal communication: Where nature
of microscopic interview analysis. Ithaca, meets culture. Mahwah, NJ: Erlbaum.
NY: Paul Martineau. Shawn, T. (1954). Every little movement: A
Quintilian, M. F. (1922/ca. 90 CE). The institu book about Francois Delsarte. Pittsfield,
tio oratoria, book XI (H. E. Butler, Trans.). MA: Eagle Print & Binding.
Cambridge, MA: Harvard University Press. Sheldon, W. H. (1940). The varieties of human
Rhodes, G., & Zebrowitz, L. A. (Eds.). (2002). physique: An introduction to constitutional
Facial attractiveness: Evolutionary, cognitive, psychology. New York: Harper & Row.
and social perspectives. Westport, CT: Ablex. Sommer, R. (1969). Personal space. Englewood
Riggio, R. E., & Feldman, R. S. (Eds.). (2005). Cliffs, NJ: Prentice Hall.
Applications of nonverbal communication. Sommer, R. (2002). Personal space in a digital
Mahwah, NJ: Erlbaum. age. In R. B. Bechtel & A. Churchman
Rosenbaum, R. (1995, January 15). The posture (Eds.), Handbook of environmental psy
photo scandal. New York Times Magazine, chology (pp. 647–660). New York: Wiley.
pp. 26–31, 40, 46, 55–56. Spitz, H. H. (1997). Nonconscious move
Rosenthal, R. (1966). Experimenter effects in ments: From mystical messages to facilitated
behavioral research. New York: Appleton communication. Mahwah, NJ: Erlbaum.
Century-Crofts. Trager, G. L. (1958). Paralanguage: A first approx
Rosenthal, R. (2002). Covert communication in imation. Studies in Linguistics, 13, 1–12.
classrooms, clinics, courtrooms, and cubi Volkmar, F. R., & Siegel, A. E. (1982).
cles. American Psychologist, 57, 839–849. Responses to consistent and discrepant
Rosenthal, R., & Jacobson, L. (1968). Pygmalion social communications. In R. S. Feldman
in the classroom. New York: Holt, Rinehart (Ed.), Development of nonverbal behavior
& Winston. in children (pp. 231–255). New York:
Ruesch, J. (1953). Synopsis of the theory of Springer.
human communication. Psychiatry, 16, Wallbott, H. G. (1982). Contributions of
215–243. the German “Expression Psychology” to
Ruesch, J., & Kees, W. (1956). Nonverbal com nonverbal communication research: Part
munication: Notes on the visual perception III. Gait, gestures, and body movement.
of human relations. Los Angeles: University Journal of Nonverbal Behavior, 7, 20–32.
of California Press. Wegner, D. M., & Bargh, J. A. (1998).
Russell, J. A., & Fernández-Dols, J. M. (1997). Control and automaticity in social life. In
The psychology of facial expression. New D. T. Gilbert, S. T. Fiske, & G. Lindzey
York: Cambridge University Press. (Eds.), The handbook of social psychology
Scheflen, A. E. (1973). Communicational struc (Vol. 1, 4th ed., pp. 446–496). New York:
ture: Analysis of a psychotherapy transaction. McGraw-Hill.
Bloomington: University of Indiana Press. Weitz, S. (Ed.). (1974). Nonverbal communica
Scherer, K. R. (1986). Vocal affect expression: tion: Readings with commentary. New
A review and a model for future research. York: Oxford University Press.
Psychological Bulletin, 99, 143–165. Wiener, M., Devoe, S., Rubinow, S., & Geller, J.
Scherer, K. R., & Ekman, P. (Eds.). (1982). (1972). Nonverbal behavior and nonverbal
Handbook of nonverbal behavior research. communication. Psychological Review, 79,
New York: Cambridge University Press. 185–214.
01-Manusov.qxd 6/30/2006 3:58 PM Page 20
02-Manusov.qxd 6/30/2006 4:00 PM Page 21
2
THE EVOLUTION OF THEORIES
OF INTERACTIVE BEHAVIOR
� Miles L. Patterson
University of Missouri, St. Louis
◆ 21
02-Manusov.qxd 6/30/2006 4:00 PM Page 22
22–––◆–––Foundations
cognitive and behavioral adjustments in a Although these were all interesting issues,
wide range of face-to-face encounters? This this line of research provided little insight
is the central question underlying the theo into the dynamic relationships across non
ries discussed in this chapter. The purposes verbal behaviors as people interacted with
of this chapter are (1) to trace the develop one another. The landscape changed dra
ment of theories of interactive behavior and matically, however, with the publication of a
(2) to discuss, in some detail, a more com 1965 article by Argyle and Dean.
prehensive, parallel process model of non
verbal communication (see also Patterson,
2001). Although this chapter focuses spe EQUILIBRIUM THEORY
cifically on the behavioral give-and-take
between people, these theories have impor In their equilibrium theory, Argyle and
tant implications for a wide variety of topics Dean (1965) focused on how individuals
in nonverbal communication, including maintain a comfortable or appropriate level
emotions, deception, influence, impression of behavioral intimacy or involvement in
management, and intimacy. interactions. They proposed that a small
Even though this chapter highlights the set of behaviors, including distance, gaze,
parallel process model of nonverbal com smiling, and verbal intimacy (self-disclosure)
munication, it is important to appreciate determines the overall level of involvement in
the how and why of the changing theoreti an interaction. As the underlying intimacy in
cal landscape over time. Because newer the a relationship increased, for example, from
ories build necessarily on earlier theories initial strangers to acquaintances to good
and on the research stimulated by them, it is friends or lovers, the comfortable level of
useful to consider the course of these devel involvement also increased. Over the course
opments. As a participant in these efforts of any specific interaction, there was pres
over the last 30 years, I cannot claim an sure to maintain a balance, or equilibrium, in
absence of bias regarding the merits of par the level of involvement. For example, if a
ticular theories. Nevertheless, perhaps I can stranger approached too closely, one might
provide an insider’s perspective on the evo turn away and avoid eye contact. This kind
lution of theories of interactive behavior. of adjustment was termed compensation
because the reduction in gaze compensated
for the too close approach.
♦ Early Theories Equilibrium theory was especially impor
tant because it was the first attempt to
explain the momentary behavioral adjust
With few exceptions, the advent of systematic ments that people make over the course of
empirical research on nonverbal communica an interaction. Early empirical research not
tion was marked by a focus on one behavior only supported equilibrium theory, but it
or channel at a time (e.g., Exline, 1963; also expanded the range of relevant behav
Sommer, 1959). For example, researchers iors (for a review, see Patterson, 1973. In
studying spatial behavior did not generally addition to the behaviors Argyle and Dean
concern themselves with the simultaneous (1965) proposed, body orientation, lean,
changes in gaze or facial expressions. touch, posture, and expressiveness also con
Instead, investigators often examined how tributed to the overall involvement between
culture, gender, personality, or the specific interaction partners (see Andersen, Guerrero,
situation affected the behavior of interest. & Jones, this volume).
02-Manusov.qxd 6/30/2006 4:00 PM Page 23
Over time, however, two distinct limita explaining patterns of compensation and
tions to equilibrium theory became evident. reciprocation, important differences were
First, the results of a few studies directly also evident.
contradicted the predictions of equilibrium My own arousal-labeling theory pro
theory (e.g., Breed, 1972; Chapman, 1975). posed that the experience of arousal in
Instead of compensating for increased response to a change in the partner’s non
involvement, individuals in these studies verbal behavior precipitated a labeling or
increased, or reciprocated, the higher self-attribution process (Patterson, 1976).
involvement of a partner. Second, the This process was the mechanism at the core
majority of the empirical research support of Schachter and Singer’s (1962) two-factor
ing equilibrium theory involved confeder theory of emotions. The arousal-labeling
ates who acted in a relatively extreme theory predicted that if the partner’s change
fashion toward their unsuspecting partners in nonverbal involvement (e.g., a close
in settings where the research participants approach, touch, and a high level of gaze)
had little control over their immediate was sufficient to produce arousal, individu
environments. Examples of this research als initiated the labeling process. Next,
included studies of spatial invasion, staring, if the resulting emotional state was posi
or the initiation of unexpected touch. tive (e.g., liking, love, comfort), then the
Under these circumstances, it is not surpris individual would reciprocate the partner’s
ing that most people compensated. That is, increased involvement. Thus, a close
they left the setting, turned away, or approach, smile, and touch from a good
avoided gaze in response to the confeder friend would increase arousal, be labeled as
ate’s increased involvement. This kind of liking, and lead to reciprocating the friend’s
pattern might not be expected between high involvement. This reciprocation might
good friends interacting on their own terri take the form of smiling back at the friend
tories. In fact, reciprocation might be more and increasing gaze. If similar behavior
common in interactions between friends, was initiated unexpectedly by a stranger,
family member, or lovers. A different arousal would also be increased, but it
approach was needed to explain both com would be more often labeled as discomfort
pensation and reciprocation between and lead to compensation. Thus, the recipi
strangers and intimate partners. ent might turn away and avoid gaze in
attempting to reestablish some degree of
comfort and control in the setting.
AFFECT-BASED THEORIES Around the same time, Burgoon pro
posed an expectancy-violations model of
From the mid-1970s to the mid-1980s, a personal space (Burgoon, 1978; Burgoon &
number of different theories were advanced Jones, 1976). Although this theory focused
to explain compensation and reciproca originally on the effects of preferred inter
tion across a wide range of relationships. action distances on communication out
Because there was research linking arousal comes, such as communicator credibility
to increased levels of nonverbal involve and attractiveness (Hale & Burgoon,
ment (e.g., Gale, Lucas, Nissim, & Harpham, 1984), its extension to effects on nonverbal
1972; McBride, King, & James, 1965), involvement was fairly direct. Specifically,
arousal seemed a likely mediator of nonver when expectations about preferred levels
bal adjustments. Although several theories of involvement are violated, arousal is
shared a common emphasis on arousal in increased, and a labeling or interpretation
02-Manusov.qxd 6/30/2006 4:00 PM Page 24
24–––◆–––Foundations
affect-based theories. That is, the greater but the few studies actually employing phys
the positive affect that one experiences, the iological measures have not demonstrated
greater the reciprocation of the partner’s that arousal is a necessary mediator.
nonverbal involvement, and the greater the Assessing the role of cognitions in this
negative affect that one experiences, the kind of research is difficult because these
greater the compensation for the partner’s measures have to be taken after the interac
nonverbal involvement. tions. Thus, it is only after the interaction is
Although these four theories propose completed that individuals might rate what
distinct processes mediating behavioral they think about their partners, themselves,
adjustments, it is very difficult to structure a and the interaction. Alternatively, research
critical test of their relative merits (but see participants might list the specific thoughts
Andersen, Guerrero, Buller, & Jorgensen, they recall from the interaction, sometimes
1998). In most cases, the theories make sim prompted by a videotape replay of the
ilar predictions for a specific set of circum interaction (Ickes, Bissonnette, Garcia, &
stances. For example, each of the theories Stinson, 1990; Patterson, 1983, p. 170).
predicts that substantially increased involve Such measures can provide some insight
ment (close approach, sustained gaze, a into what people might have been thinking
smile, and touch) from a disliked other pre during the interaction, but it is not the same
cipitates compensation (turning away and as being able to assess those cognitions as
gaze avoidance). Each of the theories also they happen.
predicts that similar increased involvement Assessment is further complicated by the
from a well-liked other precipitates recipro fact that reported cognitions and attribu
cation (increased gaze, a smile, and touch). tions are often the product of the behavior,
Actually measuring the hypothesized not the cause of the behavior (Bem, 1972).
mediating processes (arousal change and Thus, positive ratings of a confederate after
cognitions) as they occur in interactions, an interaction do not necessarily mean that
however, is demanding. Although the mon positive cognitions mediated a reciproca
itoring of physiological arousal in structured tion pattern. Rather, behavioral adjust
interactions has been done occasionally ments could have happened for other
(e.g., Coutts, Schneider, & Montgomery, reasons and, in turn, precipitated the cogni
1980; Whitcher & Fisher, 1979), most of tions. That is, the reported evaluations
the research on interactive behavior does might not be present at the time of the
not include physiological measures. In two actual behavior, but when participants are
studies in our own laboratory that did queried, they can provide such judgments
employ physiological measures, we found based on their behavior. Such a sequence
only very limited support for the predicted would be inconsistent with the predic
increase in arousal following confederates’ tions of these early theories. Of course, the
increased nonverbal involvement toward sequencing issue would not apply to
a subject (Ickes, Patterson, Rajecki, & Cappella and Greene’s (1982) discrepancy-
Tanford, 1982, study 2; Patterson, Jordan, arousal theory, because it excludes the role
Hogan, & Frerker, 1981). In addition, the of cognitions in mediating adjustments
intrusive nature of physiological measures specifically.
also reduces the external validity of the Although these issues are important con
results. It is probably fair to say that arousal cerns regarding these early theories, there
can mediate nonverbal adjustments, espe were more basic limitations to all the theo
cially when a partner’s behavior is extreme, ries. First, all the theories were reactive in
02-Manusov.qxd 6/30/2006 4:00 PM Page 26
26–––◆–––Foundations
others (see chapters by Floyd, Hall, Gifford, assessment, individuals may reevaluate the
and Matsumoto, this volume). The pro purpose (perceived function) of the interac
posed linkages among the various determi tion as they are also making nonverbal
nants, mediating processes, and interaction adjustments. Over time, these covert and
outcomes can be seen in Figure 2.1. overt adjustments promote stability in the
Because this approach emphasizes the interaction; but if they do not work, an early
functions of interaction and recognizes that termination of the interaction is likely.
affect alone cannot determine particular The functional approach emphasizes the
patterns of nonverbal involvement, the utility of nonverbal communication in serv
focus moves away from simply predict ing several general functions including
ing either compensation or reciprocation. (1) providing information, (2) regulating
Although people sometimes make reactive interaction, (3) expressing intimacy, (4) exer
adjustments of compensation and recipro cising influence, and (5) managing impres
cation, in the functional approach, individ sions (Patterson, 1991). Furthermore, similar
uals are portrayed as more proactive in patterns of behavior may be driven by differ
initiating specific patterns of behavior in the ent functions. For example, the same close
service of different goals. Because it is inap approach, smile, and touch might reflect inti
propriate to characterize such goal-driven macy or simply be an attempt to manipulate
patterns as simply compensation or recipro the partner. Although the functional per
cation, a different kind of outcome metric spective captures the complex nature of non
was proposed for the functional model: the verbal communication better than the
stability of nonverbal exchange. affect-based theories, it does come at a cost.
When the perceived function of a given Specifically, the functional model does not
interaction is shared by the partners, interac attempt straightforward, directional predic
tions will tend to proceed in a relatively tions of behavioral adjustments, like those
stable and predictable manner. As partners’ made by the affect-based theories. As a
similarity in culture and personality increases, result, it falls short on an important quality
the probability that expectancies and behav of a good theory: being easily testable. On
ioral predispositions will be more compatible the other hand, because individuals can be
also increases. In turn, this increases the like proactive in meeting their goals and act inde
lihood that nonverbal exchange will be more pendently of their underlying feelings, the
stable and predictable. Of course, there task of framing specific predictions will nec
are exceptions to this generalization. For essarily be difficult. An interesting applica
example, individuals who are complementary tion and extension of the functional model
on the dominance-submissiveness dimension can be seen in the area of social stigma and
will typically have more stable interac intergroup interactions (Hebl & Dovidio,
tions than those who are similar (see 2005; see also Dovidio, Hebl, Richeson, &
Burgoon & Dunbar, this volume, for more Shelton, this volume).
on the dominance-submissiveness dimen
sion). When individuals have a sense of insta
bility in the interaction, the model proposes ♦ Interaction Adaptation Theory
that they are likely to experience arousal
change and initiate a cognitive-affective
assessment of the situation (see Figure 2.1, In an ambitious attempt to resolve the incon
right half). Depending on the level of sistencies between empirical results and
arousal change and the cognitive-affective various theoretical explanations, Burgoon
02-Manusov.qxd 6/30/2006 4:00 PM Page 28
28–––◆–––Foundations
Determinants Preinteraction
Remote Proximate Mediators Interaction Phase
Arousal
Change
Behavioral
Cognitive-
Predispositions
Affective
Biology Assessment
Culture
Arousal
Gender
Personality
Cognitive-
Affective
Perceived
Situation Expectancies No
Function
Relationship Stable?
Nonverbal Yes
Involvement
Continue to
Termination
and her colleagues (Burgoon et al., 1998; determine a person’s interaction position
Burgoon, Stern, & Dillman, 1995) pro (IP)—that is, the dominant behavioral pre
posed the interaction adaptation theory disposition likely for a given setting with a
(IAT). In this theory, several basic concepts particular partner. In other words, one’s IP
are proposed as the major determinants of is an estimate of the actor’s likely behavior
behavioral adjustments in interactions. The shaped by biology, experience, individual
first three are the required (R), expected characteristics, and expectancies about a
(E), and desired (D) levels of functionally partner.
driven behavior patterns. The R component The particular valence and level of
refers to biological needs and drives, often involvement represented in a person’s IP
operating outside of awareness, that influ are highly variable and dependent on the
ence interactive behavior. The E component weight of the contributing R-E-D compo
refers to social factors, including knowl nents. For example, if a particular interac
edge of the setting, social norms, and the tion has implications for a person’s safety
partner’s typical behavior in the setting, and welfare, R will influence the final IP. If
that combine to determine behavioral the setting and interaction are constrained
expectancies. The D component refers to a by social norms—for example, in a job
range of individual factors, including per interview—the effect of E will be primary in
sonality characteristics, attitudes, and determining IP. Likewise, if the situation is
moods unique to a particular individual. less structured and social norms are mini
These three factors, in turn, combine to mized, the personality characteristics and
02-Manusov.qxd 6/30/2006 4:00 PM Page 29
momentary affect will result in D being form of matching and reciprocity, which
more important than R and E. Predictions promote coordination and similarity across
about the course of interaction adaptation interactants.
are possible only when the partner’s actual
interactive behavior (A) is known and
compared to the actor’s IP. ♦ Parallel Process Model of
In general, IAT predicts that when A Nonverbal Communication
matches or is only slightly discrepant from
IP, an actor should match or reciprocate
the partner’s behavior. As the discrepancy SETTING THE CONTEXT
between A and IP gets larger, actors are
more likely to engage in cognitive assess The theories discussed thus far have
ment and behavioral adjustment. According focused primarily on how individuals
to IAT, there is pressure to minimize the dis behave in interactions. Specifically, they
crepancy between A and IP to stabilize the address how we can explain, and poten
interaction. The predicted behavioral adap tially predict, patterns of nonverbal involve
tation is toward the factor (either A or IP) ment in social settings. Early theories were
that is more positively valenced. For primarily reactive in nature and stressed the
example, suppose the actor expects a high importance of affect in precipitating non
level of involvement (IP) from a partner, but verbal adjustments. The functional model
the partner initiates a much lower level of and IAT recognized the necessity of trying
involvement (A). In this case, the actor to explain not only reactive adjustments but
should compensate by trying to enlist also behavior initiated by actors. Although
greater involvement from the partner and, individual actors engage necessarily in some
in the process, reduce the discrepancy cognitive activity in the process of manag
between A and IP. If the actor expects a ing nonverbal involvement, the focus in
lower level of involvement (IP) from the both the earlier and the later theories was
partner, but the partner initiates a much clearly on behavior—that is, the encoding
higher level of involvement (A), then the or sending of nonverbal communication.
actor should converge with or reciprocate Whereas the decoding or receiving of
the partner’s high involvement. In the latter nonverbal behavior had been generally
case, as in the former, the discrepancy neglected in theories of interactive behavior,
between A and IP is reduced with the actor’s the opposite was the case in developing
behavioral adjustment. research and theory in social cognition (see,
In an experiment involving interactions e.g., Fiske & Taylor, 1995; Kunda, 1999).
among same- or cross-culture dyads, mixed This work, conducted primarily by social
support was found for the predictions of psychologists who were part of the “cogni
interpersonal adaptation theory (Burgoon tive revolution” in psychology, provided a
et al., 1998). In general, partners adapted to new perspective on the old issues of person
one another as a function of their individual perception and social judgment. For
and cultural characteristics as predicted by example, information-processing theories
interpersonal adaptation theory and by (e.g., Brewer, 1988; Fiske & Neuberg, 1990)
other theories reviewed here. Perhaps the focused largely on how a perceiver might
most important contribution of interper attend to and process a person’s characteris
sonal adaptation theory, however, is its tics, appearance, and category membership
emphasis on the pervasive pressure for in forming an impression. These theories
behavioral adjustments, typically in the recognized that these processes sometimes
02-Manusov.qxd 6/30/2006 4:00 PM Page 30
30–––◆–––Foundations
4:00 PM
Partner Social
Page 31
Judgments
Interpersonal Attentional
Expectancies Focus
Affect
Biology Cognitive
Culture Resources
Gender
Personality
Goals
Cognitive
Effort
Dispositions
Action Actor
Setting Schemas Behavior
32–––◆–––Foundations
personality introduce increased variability context for interaction, but the cognitive-
in communication. For example, even affective mediators are the processes that
though there is some degree of universality guide the course of communication. Inter
in expressive reactions, differences across personal expectancies affect the social judg
culture are also evident (Elfenbein & ment and behavioral processes in nonverbal
Ambady, 2002; Russell, 1994; see also communication simultaneously. For example,
Matsumoto, this volume). Next, the effect expectancies can create a self-fulfilling
of gender might be seen as the joint product prophecy in which actors’ own behavior
of biology (the hardwired patterns) and cul facilitates the behavior expected of a partner,
ture (societal norms) in shaping patterns of without the actors’ awareness of their role in
nonverbal communication (see Hall, this the process (Rosenthal, 1974). Nevertheless,
volume). Finally, individual differences in a partner’s subtle appearance cues or behav
personality also contribute to contrasting ior can also signal underlying dispositions
styles of nonverbal communication (see (Jussim, 1991; Zebrowitz & Collins, 1997),
Gifford, this volume). Thus, the combined resulting in an accurate judgment and not a
effects of the determinants produce both self-fulfilling prophecy. Affect is a product of
basic commonalities and differences in non an individual’s momentary disposition and
verbal communication. goals, his or her relationship to the partner,
and the setting constraints. Affect can influ
Social Environment. Because the determi ence both the formation of social judgments
nants also affect our choices of social envi (e.g., Alloy & Abramson, 1988) and the pat
ronments, they have another, indirect terns of nonverbal involvement, as seen in
influence on nonverbal communication, as the early theories reviewed in this chapter.
seen in the second stage of the model. Dispositions refer to actor states precipitated
Interactions occur with specific partners in a specific social environment. The more
in particular social settings. Because we obvious dispositions are linked to actors’
interact differently with different people and personality characteristics (see Gifford, this
in different settings, the social environment volume). For example, the experience of
constrains our patterns of nonverbal social anxiety in a particular interaction can
communication. Just as we select settings, so lead to decreased involvement (larger inter
do settings select us: I like to play golf, but personal distances and decreased gaze) and
the exclusive country clubs in St. Louis have can affect social judgments adversely
little interest in having me as a member, (Patterson & Ritts, 1997). Goals may be the
even if I could afford to join them. The most important of the mediators because
combined effect of self- and setting-selection they are the cognitive representations of
processes results in greater homogeneity desired states for which people strive (Berger,
among people in most settings (Barker, Knowlton, & Abrahams, 1996). Further
1968; Wicker, 1979). In turn, this increased more, goal-directed behavior, and even the
similarity among people in specific settings goals themselves, can be activated automati
not only facilitates greater accuracy in mak cally (Bargh & Chartrand, 1999).
ing social judgments of others (Funder, The final mediator in the model, cogni
1987; Swann, 1984) but also facilitates tive resources, refers to the total cognitive
behavioral coordination in interactions. capacity available for managing our every
day activities. In social interactions, cogni
Cognitive-Affective Mediators. The determi tive resources may be focused on a wide
nants and the social environment set the variety of concerns. For example, people
02-Manusov.qxd 6/30/2006 4:00 PM Page 33
34–––◆–––Foundations
First, we tend to pursue our communication approach and avoidance that have undoubt
goals in an efficient manner, minimizing edly been selected over the course of evolu
effort. Fiske and Taylor (1995, chaps. 4–7) tion. These would include expressive
characterize perceivers as “cognitive misers” reactions that signal a person’s intended course
as they make judgments of others. Second, of action (Fridlund, 1994; see also Fridlund
most people employ the behavioral strate & Russell, this volume). Automatic patterns
gies that are appropriate and follow social of increased involvement (e.g., close approach,
norms and customs (Berger, 1997, chap. 2). gaze, and touch) might be activated in
Thus, people typically take the path of least response to increased attraction or a need for
resistance and avoid calling undue negative comforting and supporting another person.
attention to themselves. In contrast, decreased involvement or behav
ioral avoidance may be precipitated by dis
Social Judgments. In general, research indi like, fear, or embarrassment. Besides the
cates that most initial social judgments hap hardwired, affect-driven patterns, other
pen more or less automatically, often outside patterns become automatic over time as a
of conscious awareness (e.g., Bargh, 1994; function of experience and learning. For
Brewer, 1988; see also Lakin, this volume). example, most of us learn over time how to
When the information about others (i.e., “make a good impression” when there is a
appearance and behavior) is ambiguous or lot at stake. The cognitive representations of
inconsistent, considerable cognitive effort these automatic sequences may be described
might be engaged in resolving a final judg as action schemata (see Figure 2.2, bottom)
ment, but only if the perceiver is motivated and can be initiated with little or no cogni
to do so (Gilbert et al., 1988). Nevertheless, tive effort (Abelson, 1981; Vallacher &
more is not always better when it comes to Wegner, 1987).
cognitive effort in making judgments. Rapid Another way to conceptualize the
judgments from “thin slices of behavior” dynamics of behavioral processes is in terms
are, more often than not, accurate (Ambady of potentially competing response systems—
& Rosenthal, 1992), and increased cogni that is, automatic versus controlled. Metcalfe
tive effort can even lead to more errors in and Mischel (1999) proposed such an
judgment (Patterson & Stockbridge, 1998; approach in their “hot/cool system analysis”
Wilson & Schooler, 1991). When additional of the conflict involved in the delay of grat
reflection is needed in making a judgment, it ification. The “hot” response is the auto
is possible only when there are sufficient cog matic approach to immediate gratification,
nitive resources available to the individual. which is initially under stimulus control
If a person is distracted, worried, tired, or (Metcalfe & Mischel, 1999), like Bargh’s
investing considerable effort in managing (1997) automatic actions. The hot system
behavior, then corrections to an initial judg develops early and is simple, reflexive, and
ment are unlikely. Thus, the initial automatic emotional in nature, like the affect-driven
judgment will dominate. reactions discussed in the early theories in
this chapter. In contrast, the “cool”
Social Behavior. Sending nonverbal commu response is a product of self-control. The
nication, like receiving it, engages a variety cool system develops later and is more com
of processes (from relatively automatic to plex, reflective, and cognitive in nature.
more controlled). On the automatic end, Strack and Deutsch (2004) proposed a
our behavioral repertoire encompasses a model similar to the hot/cool system that
wide range of basic, hardwired patterns of engages both reflective and impulsive
02-Manusov.qxd 6/30/2006 4:00 PM Page 35
36–––◆–––Foundations
Bargh, J. A., & Chartrand, T. L. (1999). The Chapman, A. J. (1975). Eye contact, physical
unbearable automaticity of being. American proximity and laughter: A reexamination
Psychologist, 54, 462–479. of the equilibrium model of social intimacy.
Barker, R. G. (1968). Ecological psychology: Social Behavior and Personality, 3, 143–155.
Concepts and methods for studying the Coutts, L. M., Schneider, F. W., & Montgomery,
environment of human behavior. Stanford, S. (1980). An investigation of the arousal
CA: Stanford University Press. model of interpersonal intimacy. Journal
Bem, D. J. (1972). Self-perception theory. In of Experimental Social Psychology, 16,
L. Berkowitz (Ed.), Advances in experimen 545–561.
tal social psychology (Vol. 6, pp. 1–62). Elfenbein, H. A., & Ambady, N. (2002). On
New York: Academic Press. the universality and cultural specificity
Berger, C. B. (1997). Planning strategic interac of emotion recognition: A meta-analysis.
tion. Mahwah, NJ: Erlbaum. Psychological Bulletin, 128, 203–235.
Berger, C. B., Knowlton, S. W., & Abrahams, Exline, R. V. (1963). Explorations in the process
M. F. (1996). The hierarchy principle in of person perception: Visual interaction in
strategic communication. Communication relation to competition, sex, and need for
Theory, 6, 111–142. affiliation. Journal of Personality, 31, 1–20.
Breed, G. (1972). The effect of intimacy: Fiske, S. T. (1992). Thinking is for doing:
Reciprocity or retreat? British Journal Portraits of social cognition from daguerreo
of Social and Clinical Psychology, 11, type to laserphoto. Journal of Personality
135–142. and Social Psychology, 63, 877–889.
Brewer, M. B. (1988). A dual process model Fiske, S. T., & Neuberg, S. L. (1990). A contin
of impression formation. In T. K. Srull & uum of impression formation, from cate
R. S. Wyer Jr. (Eds.) Advances in social gory-based to individuating processes:
cognition (Vol. 1, pp. 1–36). Hillsdale, NJ: Influences of information and motiva
Erlbaum. tion on attention and interpretation. In
Burgoon, J. K. (1978). A communication model M. Zanna (Ed.), Advances in experimental
of personal space violations: Explication social psychology (Vol. 23, pp. 1–74). New
and an initial test. Human Communication York: Academic Press.
Research, 4, 129–142. Fiske, S. T., & Taylor, S. E. (1995). Social cog
Burgoon, J. K., Ebesu, A. S., White, C. H., Koch, nition (2nd ed.). New York: McGraw-Hill.
P., Alvaro, E. M., & Kikuchi, T. (1998). Fridlund, A. J. (1994). Human facial expression: An
The many faces of interaction adaptation. evolutionary view. San Diego: Academic Press.
In M. T. Palmer & G. A. Barnett (Eds.), Funder, D. C. (1987). Errors and mistakes:
Progress in communication sciences (Vol. Evaluating the accuracy of social judgment.
14, pp. 191–220). Stamford, CT: Ablex. Psychological Bulletin, 101, 75–90.
Burgoon, J. K., & Jones, S. B. (1976). Toward Gale, A., Lucas, B., Nissim, R., & Harpham, B.
a theory of personal space expectations (1972). Some EEG correlates of face-to
and their violations. Human Communica face contact. British Journal of Social and
tion Research, 2, 131–146. Clinical Psychology, 11, 326–332.
Burgoon, J. K., Stern, L. A., & Dillman, L. Gilbert, D. T., & Krull, D. S. (1988). Seeing less
(1995). Interpersonal adaptation: Dyadic and knowing more: The benefits of percep
interaction patterns. Cambridge, UK: tual ignorance. Journal of Personality and
Cambridge University Press. Social Psychology, 54, 193–202.
Cappella, J. N., & Greene, J. O. (1982). A Gilbert, D. T., Pelham, B. W., & Krull, D. S.
discrepancy-arousal explanation of mutual (1988). On cognitive busyness: When person
influence in expressive behavior for adult perceivers meet persons perceived. Journal
and infant-adult interaction. Communication of Personality and Social Psychology, 54,
Monographs, 49, 89–114. 733–740.
02-Manusov.qxd 6/30/2006 4:00 PM Page 38
38–––◆–––Foundations
Hale, J. L., & Burgoon, J. K. (1984). Models of Patterson, M. L. (1991). A functional approach
reactions to changes in nonverbal immediacy. to nonverbal exchange. In R. S. Feldman &
Journal of Nonverbal Behavior, 8, 287–314. B. Rime (Eds.), Fundamentals of nonverbal
Hebl, M. R., & Dovidio, J. F. (2005). Promoting behavior (pp. 458–495). Cambridge, UK:
the “social” in the examination of social Cambridge University Press.
stigmas. Personality and Social Psychology Patterson, M. L. (2001). Toward a comprehen
Review, 9, 156–182. sive model of nonverbal communication. In
Ickes, W., Bissonnette, V., Garcia, S., & Stinson, W. P. Robinson & H. Giles (Eds.), The new
L. L. (1990). Implementing and using the handbook of language and social psychol
dyadic interaction paradigm. In C. Hendrick ogy (pp. 159–176). Chichester, UK: Wiley.
& M. S. Clark (Eds.), Research methods in Patterson, M. L., Jordan, A., Hogan, M. B., &
personality and social psychology (pp. Frerker, D. (1981). Effects of nonverbal
16–44). Newbury Park, CA: Sage. intimacy on arousal and behavioral adjust
Ickes, W., Patterson, M. L., Rajecki, D. W., & ment. Journal of Nonverbal Behavior, 5,
Tanford, S. (1982). Behavioral and cogni 184–198.
tive consequences of reciprocal versus Patterson, M. L., & Ritts, V. (1997). Social and
compensatory responses to preinteraction communicative anxiety: A review and meta
expectancies. Social Cognition, 1, 160–190. analysis. In B. R. Burleson (Ed.), Com
James, W. (1983). The principles of psychology. munication yearbook 20 (pp. 262–303).
Cambridge, MA: Harvard University Press. Thousand Oaks, CA: Sage.
(Original work published 1890) Patterson, M. L., & Stockbridge, E. (1998). Effects
Jussim, L. (1991). Social perception and social of cognitive demand and judgment strategy
reality: A reflection-construction model. on person perception accuracy. Journal of
Psychological Review, 98, 54–73. Nonverbal Behavior, 22, 253–263.
Kunda, Z. (1999). Social cognition: Making sense Rosenthal, R. (1974). On the social psychology of
of people. Cambridge, MA: MIT Press. the self-fulfilling prophecy: Further evidence
McArthur, L. Z., & Baron, R. M. (1983). for Pygmalion effects and their mediating
Toward an ecological theory of social mechanisms. New York: M.S.S. Information
perception. Psychological Review, 90, Corporation Modular Publication.
215–238. Russell, J. A. (1994). Is there a universal recog
McBride, G., King, M. C., & James, J. W. nition of emotion from facial expression?
(1965). Social proximity effects of galvanic A review of the cross-cultural studies.
skin responses in adult humans. Journal Psychological Bulletin, 115, 102–141.
of Psychology, 61, 153–157. Schachter, S., & Singer, J. E. (1962). Cognitive,
Metcalfe, J., & Mischel, W. (1999). A hot/cool social, and physiological determinants of
system analysis of delay of gratification: emotional state. Psychological Review,
Dynamics of willpower. Psychological 69, 379–399.
Review, 106, 3–19. Sommer, R. (1959). Studies in personal space.
Patterson, M. L. (1973). Compensation in non Sociometry, 22, 247–260.
verbal immediacy behaviors: A review. Strack, F., & Deutsch, R. (2004). Reflective and
Sociometry, 36, 237–252. impulsive determinants of social behavior.
Patterson, M. L. (1976). An arousal model Personality and Social Psychology Review,
of interpersonal intimacy. Psychological 8, 220–247.
Review, 83, 235–245. Swann, W. B., Jr. (1984). Quest for accuracy in
Patterson, M. L. (1982). A sequential functional person perception: A matter of pragmatics.
model of nonverbal exchange. Psycholo Psychological Review, 91, 457–477.
gical Review, 89, 231–249. Vallacher, R. R., & Wegner, D. M. (1987).
Patterson, M. L. (1983). Nonverbal behavior: What do people think they’re doing?
A functional perspective. New York: Action identification and human behavior.
Springer-Verlag. Psychological Review, 94, 3–15.
02-Manusov.qxd 6/30/2006 4:00 PM Page 39
Whitcher, S. J., & Fisher, J. D. (1979). decisions. Journal of Personality and Social
Mulitdimensional reactions to therapeutic Psychology, 60, 181–192.
touch in a hospital setting. Journal of Zebrowitz, L. A. (1997). Reading faces: Window
Personality and Social Psychology, 37, 87–96. to the soul? Boulder, CO: Westview Press.
Wicker, A. W. (1979). An introduction to Zebrowitz, L. A., & Collins, M. A. (1997).
ecological psychology. Monterey, CA: Accurate social perception at zero acquain
Brooks/ Cole. tance: The affordances of a Gibsonian
Wilson, T. D., & Schooler, J. W. (1991). approach. Personality and Social Psychology
Thinking too much: Introspection can Review, 1, 204–223.
reduce the quality of preferences and
02-Manusov.qxd 6/30/2006 4:00 PM Page 40
03-Manusov.qxd 6/30/2006 4:01 PM Page 41
3
METHODS FOR THE STUDY OF
NONVERBAL COMMUNICATION
� Heather M. Gray
Harvard University
� Nalini Ambady
Tufts University
◆ 41
03-Manusov.qxd 6/30/2006 4:01 PM Page 42
42–––◆–––Foundations
independent variables by researchers who questions that can be addressed and high
want to understand the cognitive, affective, light some issues that await further explo
or behavioral consequences of nonverbal ration. We then discuss major design
displays. For instance, to understand how considerations and the relative merits and
first impressions regulate interpersonal limitations of each approach.
processes, researchers can present nonverbal
cues and measure subsequent attitudes and
decisions (e.g., Ambady & Rosenthal, 1993; ♦ Nonverbal Behavior as an
Harris & Rosenthal, 2005; Zebrowitz & Independent Variable
Rhodes, 2004). Nonverbal cues are viewed
as dependent variables when the cue is
manipulated to infer subsequent cognitive We begin this section with a summary of
or affective changes by observing nonver the ways in which nonverbal behavior can
bal behavior (e.g., Dimberg, Thunberg, & be used as an independent variable, with
Elmehed, 2000; Richeson & Shelton, a focus on its widespread application of
2005). At other times, nonverbal displays photographs, vocal clips, video clips, and
are studied as both independent and depen interaction paradigms to the study of inter
dent variables, such as when researchers personal communication.
assess emotional responses to emotional dis
plays or investigate behaviors co-occurring PHOTOGRAPHS
during interaction between people (see, e.g.,
Bavelas & Chovil, this volume). The survey Some Appropriate Uses. Still photographs
of methods that follows is organized accord can be used to study a wide range of
ing to these two primary design considera processes, including the perception of fleet
tions (i.e., nonverbal cues as independent ing states and enduring traits. One particu
and dependent variables). larly active line of research asks whether
Although many researchers who focus emotions are recognized universally in
on interaction would not define their work static faces. The first studies to explore
according to these criteria but, rather, this question revealed that the six “basic”
define nonverbal communication within a emotions (anger, disgust, fear, happiness,
larger set of interactional processes (e.g., sadness, and surprise) are recognized at
adaptation), we believe that the current above-chance levels across cultures (e.g.,
approach can encapsulate some of what Ekman, 1972); in these studies, participants
occurs in those studies as well. Following viewed posed expressions and were asked to
this contention, we first review some ways identify the emotion being expressed from
in which nonverbal behavior is used as a list of predetermined labels. Recognition
an independent variable; such methods are of the intended category at accuracy rates
useful in uncovering the perceptual, affec greater than that expected by chance pro
tive, and cognitive consequences of behav vided evidence for the universality of emo
ior communicated in a range of nonverbal tion recognition.
channels. Next, we describe the ways in Current work extends this initial inquiry
which nonverbal behavior is assessed as a by focusing on a number of questions that
dependent variable, organized according to may have implications for our understand
the channel of communication used to infer ing of the form and function of emotional
the underlying subjective experience. For displays: What are the social functions
each method within these larger areas, we of these displays (Marsh, Adams, & Kleck,
provide examples of the kinds of research 2004), and do more “self-conscious”
03-Manusov.qxd 6/30/2006 4:01 PM Page 43
44–––◆–––Foundations
from a wide range of naturalistic situations; there are at least three ways to define accu
however, they provide little control over the racy in this measurement context: (a) the
qualities being communicated, the encoders, degree of correspondence between a judg
and the quality of the stimuli. Another, more ment and a criterion, (b) interpersonal con
controllable, option involves inducing emo sensus, and (c) the degree of a judgment’s
tions in encoders. Like spontaneous displays, utility to the perceiver. Many researchers
induced displays have the advantage of being choose the first definition, using the display
naturalistic; their primary drawbacks include intended by the encoder (or the experi
their diminished intensity and the ethical menter) as the criterion against which
issues associated with induced emotion. responses are compared. Because multiple
Given these issues, the preferred method criteria tend to overcome the inherent
involves obtaining posed displays from pro weaknesses of any single criterion, one
fessional or lay actors. Encoders can be option is to use more than one criterion in
asked to relax and imagine the scenario a single study.
vividly before performing it; this tends to There are several methods for validating
increase the intensity of expressions (Banse stimuli for recognizability and authenticity
& Scherer, 1996). For an extended discus before presenting them to study partici
sion of these issues, see Archer, Akert, and pants. One commonly used procedure
Costanzo (1993) and Scherer (2003). involves presenting stimuli to an indepen
Emotional expressions can also be either dent group of raters and computing the
“pure” or “blended” (consisting of a mix of “effective” reliability of their judgments (of
emotions, such as happiness and surprise). emotion or any other construct) using the
Although most preexisting sets contain rela Spearman-Brown formula with modified
tively pure expressions, perceivers appear to notations, as described by Rosenthal and
process blended expressions more efficiently, Rosnow (1991). The formula is R = nr/(1 +
possibly because they are more likely to (n − 1)r), where R = the effective reliability,
encounter them in real life (LaPlante & n = the number of raters, and r = the mean
Ambady, 2000). It is now possible to create reliability of all the raters (i.e., mean of the
blended expressions artificially using soft correlations).
ware such as Adobe Photoshop (e.g., Marsh, More pragmatic decisions must also
Elfenbein, & Ambady, 2003). New digital be made regarding the manner in which
editing techniques allowed Jones, Little, responses are provided. The majority of stud
Burt, and Perrett (2004) to establish that ies in this area have used forced-choice or
the apparent health of the skin surface influ dimensional response formats. Alternatively,
ences perceived attractiveness independently participants can use a free-response format
of face shape. Digital editing software also and provide any label they choose. The rel
permitted Adams and Kleck (2003) to ative merits of these response formats have
manipulate the direction of eye gaze in tar been hotly debated (for a review, see
gets expressing a range of emotions, which Elfenbein, Mandal, Ambady, Harizuka, &
was essential in documenting the joint con Kumar, 2002). Briefly, although many
tribution of eye gaze and emotional cues in scholars believe that emotions are categori
the processing of facial displays. cal in nature, Russell (1993, 1994) and
These assessments are all meant to others argue that emotions should not
capture accuracy, but defining accuracy is be construed as mutually exclusive cate
itself a challenging process (see Riggio, this gories and that forced-choice formats
volume). As outlined by Kruglanski (1989), inflate agreement artificially. Nevertheless,
03-Manusov.qxd 6/30/2006 4:01 PM Page 45
forced-choice designs are more convenient, motion often plays a critically important
obviating the time-consuming and often role in social judgment (Ambady, Hallahan,
ambiguous task of analyzing open-ended & Connor, 1999; Heider & Simmel, 1944;
responses. In addition, modifying forced- Knight & Johnston, 1997). These limitations
choice designs by adding a “none-of-the should be taken into account when consider
above” option appears to reduce artifactual ing how results may generalize to other kinds
agreement (Frank & Stennett, 2001; Haidt of viewing conditions.
& Keltner, 1999).
Although some scholars believe that all
humans are motivated naturally to decode VOCAL CLIPS
nonverbal cues (McArthur & Baron, 1983),
steps can still be taken to increase the partic Some Appropriate Uses. When we hear
ipants’ desire to take the perceptual tasks seri someone speak, we discern meaning not only
ously (such as limiting the length of the study from the words they choose but also from
and making it as interesting as possible). how those words are spoken. Prosodic cues
Also, it is important to consider the relative such as volume, pitch, and speech rate can
advantages and disadvantages of this method play a more important role than speech con
of stimulus presentation. The hallmark tent in our social inferences, perhaps because
advantages of photographs are their ease we implicitly understand that it is difficult
of use and high degree of experimental con for a speaker to control these signals (Ekman
trol. Whereas control is achieved most & Friesen, 1969). As a result, vocal cues
prominently by the selection of a specific have powerful effects on social interaction
channel of nonverbal behavior to be dis (e.g., Hummert, Mazloff, & Henry, 1999;
played (e.g., the face, the torso, the entire Neumann & Strack, 2000; Noller, 2005).
body), it is also achieved in the choice of For instance, new research reveals that
encoders, the nature of the displays, and prosody (i.e., vocal qualities) may be as
the context in which the display arises. important as facial displays in communicat
Researchers should also consider how ing emotion (for a review, see Johnstone &
the gender, age, socioeconomic status, and Scherer, 2000). Several related issues await
ethnicity of their encoders and perceivers further investigation. As in the case of facial
might affect the generalizability of results. expressions, there is some evidence for the
At each step, the investigator can make universal recognition of emotions based on
decisions necessary for addressing lingering language-free voice samples (see Scherer,
questions and advancing theory. With this 2003). At the same time, there appear to be
high level of control comes an inevitable culture-specific patterns in vocal emotion
trade-off in ecological validity. As discussed expression, similar to the nonverbal “accents”
by Elfenbein, Marsh, and Ambady (2002), in facial display of emotion across cultures
the nonverbal displays encountered in every (Marsh et al., 2003; Scherer, 2003).
day life “are subtle, embedded in a particular Some scholars in this area have focused
context, spontaneous, dynamic, fleeting, and on the link between psychological well-being
exist in combination with other expressions, and the ability to understand subtle vocal
words and behaviors” (p. 44). Standardized cues to emotion (e.g., Baum & Nowicki,
stimuli generally fail to meet at least some of 1998; Rosenthal, Hall, DiMatteo, Rogers, &
these criteria, but photographs often fail to Archer, 1979). This research has been aided
meet all of them. Only static stimuli can be by standardized sets of vocal cues, particu
presented, a drawback given that perceived larly the Profile of Nonverbal Sensitivity
03-Manusov.qxd 6/30/2006 4:01 PM Page 46
46–––◆–––Foundations
Design Considerations. When the decision Some Appropriate Uses. Visual clips com
is made to employ auditory samples in bine the visual nature of still photographs
research, one option is to use a previously with the dynamic nature of vocal clips. In
developed set of stimuli, such as the PONS or doing so, visual clips have been vital in
the DANVA. If the decision is made to con documenting the remarkable accuracy with
struct a new set of stimuli, the resulting series which perceivers can infer enduring and
of pragmatic and theoretical decisions paral fleeting characteristics from brief glimpses
lel those used in photograph-based research. of behavior. In a fascinating line of early
A unique concern here involves remov research, for example, Johansson (1973)
ing the content of the spoken messages, affixed lights to the major joints of
which is necessary for isolating the proso actors and recorded high-contrast images of
dic features. One option involves editing the actors moving in space, such that per
spoken samples by content filtering (which ceivers could only see a field of point lights.
removes the higher frequencies on which Perceivers could only identify the encoders
word recognition depends) or by randomized in these displays as human when exposed to
splicing (which rearranges segments of the moving point-light displays; in static form,
voice in a random manner), as is done in the they appeared as a random series of dots.
PONS test. Both can now be accomplished Subsequent work revealed that age and
quite easily using sound editing software; gender are also apparent in point-light
however, they produce sound patterns that movements (Koslowski & Cutting, 1977).
are not normally encountered in everyday Current work is exploring the extent to
life. which perceivers can recover other socially
The other option is to provide encoders relevant information, such as “Big Five”
with standardized content such as the alpha factors of personality (Heberlein, Adolphs,
bet (Berry, 1991) or a basic neutral sentence Tranel, & Damasio, 2004), from these
(Baum & Nowicki, 1998). This strategy highly impoverished displays.
retains the verbal content but standardizes Visual displays can be impoverished in
it so that only differences in nonverbal fea several other ways. One approach involves
tures remain (see also Noller, 2005). With sampling very brief samples, or “thin slices,”
the advent of analog and digital recording of ongoing movement. Even the “thinnest”
and playback devices, it is a relatively simple of slices capture the rich information on
matter to record, manipulate, and present which social judgments are made (for a
vocal cues to study participants (for more review, see Ambady, Bernieri, & Richeson,
details, see Scherer, 2003). Another advan 2000). Ambady and colleagues (1999), for
tage of this approach is that unlike still pho instance, asked perceivers to identify the
tographs, vocal cues are inherently dynamic. sexual orientation of encoders on the basis
To explore the role of nonverbal features in of silent, very brief (1 or 10 seconds) video
03-Manusov.qxd 6/30/2006 4:01 PM Page 47
clips. Impressions of sexual orientation con useful for studying the effects of psycho
formed to encoders’ self-reports at greater logical or physical trauma on nonverbal
than-chance levels, suggesting that sexual sensitivity. They can also be used to relate
orientation is conveyed reliably by dynamic individual differences in nonverbal sensitiv
nonverbal behavior. ity to any number of skills, including work
An exciting frontier of research is cata place performance. Constructing a new set
loguing the mental processes we are able of video clips “in-house” can also be advan
to accomplish automatically, without inten tageous, particularly if one is interested in
tion, deliberation, and conscious aware defining accuracy as the ability to predict a
ness. In our own work, we studied the novel criterion, such as student evaluations
connection between induced mood and (Ambady & Rosenthal, 1993), therapeutic
the validity of social judgments based on outcomes (Ambady et al., 2002), or testos
thin slices of behavior. Induced sadness terone level (Dabbs, Bernieri, & Strong,
impaired accuracy consistently, perhaps 2001). A more detailed summary of the
because it evoked a more reasoned, careful issues involved in such research can be
analysis of available information (Ambady found in Ambady et al. (2000).
& Gray, 2002). This finding is consistent A notable advantage of the visual-clip
with work by Patterson and Stockbridge approach is its flexibility. Depending on
(1998), who found that under some condi one’s theoretical orientation, behavior can
tions, perceivers made more accurate social be sampled from any channel of expressive
inferences when they were prevented from communication. The central trade-off in
deliberating carefully. Together, these results ecological validity concerns the extent to
suggest that first impressions may result which impressions formed on the basis of
from a relatively automatic form of cogni video clips differ from those made in every
tive processing (Choi, Gray, & Ambady, day life. Because perceivers do not have con
2004; Lakin, this volume). tact with the encoders they are viewing, they
do not experience the interaction demands
Design Considerations. Several important ubiquitous to real encounters (e.g., Gilbert,
construction and design decisions should Pelham, & Krull, 1988; Swann, 1984).
seem familiar at this point, because they are Additionally, relatively artificial settings
common to research using photographs, tend to heighten participants’ awareness
vocal clips, and video clips. Options for that their judgments are under scrutiny and
using preexisting sets of stimuli include therefore increase the extent to which infer
the PONS, described earlier, and an easily ence processes are consciously monitored
administered measure called the interper (Forgas, 1999). These limitations should
sonal perception task (IPT; Costanzo & be kept in mind when considering how
Archer, 1989). The original IPT contains the results obtained in these studies would
30 short scenes and the shortened version generalize to other situations.
contains 15 of those scenes (Costanzo &
Archer, 1993). Both tap the ability to infer
INTERACTIONS
kinship, deception, competition, status, and
intimacy. These scenes are naturalistic Some Appropriate Uses. An alternative to
rather than posed, so there is an objectively presenting standardized nonverbal displays
correct answer to each question. Because is to allow them to emerge more naturally
the PONS and the IPT have already been by constructing brief interactions between
normed on healthy populations, they are study participants. Interaction paradigms
03-Manusov.qxd 6/30/2006 4:01 PM Page 48
48–––◆–––Foundations
have proved useful for furthering our variety of personality dimensions. These
understanding of complex interpersonal ratings are then averaged and correlated
processes, such as interpersonal adaptation with self-ratings of personality. Although
(i.e., the tendency for interaction partners the accuracy of such ratings improves with
to adjust spontaneously to or mutually acquaintanceship (e.g., Funder & Colvin,
influence one another; see Cappella & 1988), the convergence of strangers’ ratings
Schreiber, this volume). A long history of with self-ratings is surprisingly high (e.g.,
research in this area has demonstrated that Albright et al., 1988).
adaptation is a ubiquitous feature of human
social exchange (for a review, see Burgoon, Design Considerations. When photographs,
Stern, & Dillman, 1995). In some of this vocal clips, or visual clips are chosen, one
research, nonverbal signs of adaptation must accept a trade-off in realism in exchange
are manipulated to investigate their conse for high experimental control. With interac
quences for relationship quality (Burgoon tions, the trade-off is reversed: Realism is
et al., 1995). Alternatively, as will be dis achieved in all the ways it is limited when
cussed in later sections, these signals can be standardized stimuli are presented. Limiting
observed naturalistically to draw inferences the extent to which participants feel they are
about preexisting relationships. being watched and studied, using hidden
Interaction paradigms can also be used cameras and microphones or one-way mir
to better understand how people in rela rors, further heightens the realism of interac
tionships come to understand one another. tion situations. Of course, less precision is
For instance, Ickes and his colleagues achieved over the interpersonal processes
devised an interaction paradigm for under investigation in the interaction para
studying empathic accuracy, the ability to digm than occurs in controlled experiments.
read mental and emotional states (for a Precision concerns are especially salient in
review, see Ickes, 2001). In their unstruc research on social perception, given that
tured dyadic interaction paradigm, two forming impressions is a serial process; the
participants are left alone together for a encoder must display the relevant cues, these
brief time and their spontaneous conver cues must be made available to the perceiver,
sation is videotaped. Later, the partici the perceiver must detect these cues, and
pants view the videotapes independently finally, the perceiver must interpret these cues
and complete written reports detailing the in light of previously stored knowledge
thoughts and feelings they experienced (Funder, 1999). When standardized stimuli
during the interaction. They also attempt are presented, the experimenter can control
to infer the thoughts and feelings their the first three steps and study how informa
partners experienced during the interac tion is processed once it has been displayed,
tion. Greater empathic accuracy is evi made available, and detected. When interac
denced by greater correspondence between tion paradigms are used, however, all the
inferred and self-reported experience. stages are free to vary.
Even less exposure to another person can Interaction paradigms present some
be sufficient for drawing surprisingly accu idiosyncratic methodological concerns.
rate inferences of personality, however. For instance, should a confederate be used?
In the zero-acquaintance paradigm (e.g., Confederates are helpful in standardizing
Albright, Kenny, & Malloy, 1988), unac interaction patterns, although very tightly
quainted peers assemble in small groups controlling the conversation between a con
and rate themselves and each other on a federate and a participant will likely result
03-Manusov.qxd 6/30/2006 4:01 PM Page 49
50–––◆–––Foundations
addition, they necessitate that muscle move guilt, and increased mental load. In an
ments be visible to the human eye. This attempt to discern how these emotional and
problem is solved by EMG recording, in cognitive changes are reflected in the voice,
which detected muscle movements are fil Vrij, Edward, Roberts, and Bull (2000) ana
tered, amplified, and smoothed before being lyzed vocal cues and verbal behavior during
subjected to traditional data-analysis proce deception and truth telling. Two reliable
dures (see Tassinary & Cacioppo, 2000, for vocal indicators of deception emerged: (1) a
more details). Like other psychophysiologi longer latency period (the time between a
cal techniques, EMG recording is relatively question and its answer) and (2) more speech
expensive and requires some additional train disturbances (instances of saying “ah” or
ing and equipment. An additional drawback “mm” between words; for a more detailed
is that the application of facial sensors is discussion, see Vrij, this volume).
relatively intrusive. Vocal cues also index interpersonal qual
ities reliably. For instance, speech accom
modation theory proposes that a speaker’s
VOCAL CUES need for affiliation is reflected in a tendency
to match a conversation partner’s speech
Some Appropriate Uses. Like facial dis cues (Giles & Smith, 1979). According to
plays, the voice can be an excellent index this perspective, people match their part
of emotional experience. For instance, ner’s vocal cues spontaneously when they
faster speech rates are associated with more want to gain approval (for a review, see
pleasant emotions, whereas sadness is con Buller, 2005). Gregory and Webster (1996)
sistently associated with the tendency to tested this hypothesis using naturalistic
speak at a very slow rate and take longer samples obtained from conversations
pauses between words (see review by between a talk show host (Larry King) and
Siegman, 1987). Ellgring and Scherer (1996) several of his guests. Lower status guests
capitalized on this association in their tended to accommodate their voices to
attempt to obtain a behavior-based mea King’s, whereas King tended to accommo
sure of progress in therapy for major date his voice to match those of higher sta
depression. In the early course of therapy, tus guests (e.g., sitting presidents). Other
and again during remission, patients pro vocal forms of accommodation, such as
vided speech samples during standardized matching accent patterns, response laten
interviews. Remission from depression was cies, and utterance durations, have been
marked by an increase in speech rate and identified (Buller, 2005).
a decrease in pause duration. On a broader
note, the authors speculate that observa Design Considerations. Quantifying vocal
tions of nonverbal behavior may be a useful cues involves a number of steps, beginning
tool in charting therapeutic progress, par with the recording of voice samples and the
ticularly given the limitations of patients’ isolation of specific segments for analysis.
ability and willingness to communicate some For speech rate analysis, trained coders
aspects of experience verbally. can be asked to listen to each segment and
Vocal cues can be used to index a broad count the number of syllables spoken.
range of emotional and cognitive states, a Alternatively, coders can simply rate speak
point illustrated by their application to ers’ voices on a continuous scale from rela
the study of deception. When telling a lie, tively slow to relatively fast. Perceived
deceivers often experience fear, excitement, similarity in speech rate can be measured by
03-Manusov.qxd 6/30/2006 4:01 PM Page 52
52–––◆–––Foundations
54–––◆–––Foundations
three minutes of a marital conflict discussion. Ekman, P., & Friesen, W. V. (1969). Nonverbal
Family Process, 38, 293–301. leakage and cues to deception. Psychiatry,
Chartrand, T. L., & Bargh, J. A. (1999). The 32, 88–106.
chameleon effect: The perception-behavior Ekman, P., & Friesen, W. V. (1976). Pictures
link and social interaction. Journal of of facial affect. Palo Alto, CA: Consulting
Personality and Social Psychology, 76, Psychologists Press.
893–910. Ekman, P., & Friesen, W. V. (1984). Emotion
Chiao, J. Y., Bordeaux, A. R., & Ambady, N. Facial Action Coding System (EMFACS).
(2004). Mental representations of social San Francisco: University of California
status. Cognition, 93, B49–B57. Press.
Chiu, P., Ambady, N., & Deldin, P. (2004). Elfenbein, H. A., & Ambady, N. (2002). On
Contingent negative variation to emotional the universality and cultural specificity
in- and out-group stimuli differentiates high- of emotion recognition: A meta-analysis.
and low-prejudiced individuals. Journal of Psychological Bulletin, 128, 203–235.
Cognitive Neuroscience, 6, 1830–1839. Elfenbein, H. A., Mandal, M. K., Ambady, N.,
Choi, Y. S., Gray, H. M., & Ambady, N. Harizuka, S., & Kumar, S. (2002). Cross-
(2004). The glimpsed world: Unintended cultural patterns in emotion recognition:
communication and unintended perception. Highlighting design and analytical tech
In R. Hassin, J. Bargh, & J. Uleman (Eds.), niques. Emotion, 2, 75–84.
The new unconscious (pp. 309–333). New Elfenbein, H. A., Marsh, A. A., & Ambady, N.
York: Oxford University Press. (2002). Emotional intelligence and the
Costanzo, M., & Archer, D. (1989). Interpret recognition of emotion from facial expres
ing the expressive behavior of others: The sions. In L. F. Barrett & P. Salovey (Eds.),
Interpersonal Perception Task. Journal of The wisdom in feeling (pp. 37–59). New
Nonverbal Behavior, 13, 225–245. York: Guilford.
Costanzo, M., & Archer, D. (Producers). Ellgring, H., & Scherer, K. R. (1996). Vocal
(1993). Interpersonal Perception Task-15 indicators of mood change in depression.
[Videotape]. (Available from the University Journal of Nonverbal Behavior, 20,
of California, 2000 Center Street, Berkeley, 83–110.
CA 94704) Forgas, J. P. (1999). Feeling and speaking: Mood
Dabbs, J. M., Bernieri, F. J., & Strong, R. K. effects on verbal communication strategies.
(2001). Going on stage: Testosterone in Personality and Social Psychology Bulletin,
greetings and meetings. Journal of Research 25, 850–863.
in Personality, 35, 27–40. Frank, M. G., & Stennett, J. (2001). The forced-
DePaulo, B. M. (1992). Nonverbal behavior and choice paradigm and the perception of facial
self-presentation. Psychological Bulletin, expressions of emotion. Journal of Per
11, 203–243. sonality and Social Psychology, 80, 75–85.
Dimberg, U., Thunberg, M., & Elmehed, K. Funder, D. C. (1999). Personality judgment:
(2000). Unconscious facial reactions to A realistic approach to person perception.
emotional facial expressions. Psychological San Diego, CA: Academic Press.
Science, 11, 86–89. Funder, D. C., & Colvin, C. R. (1988). Friends
Ekman, P. (1972). Universals and cultural and strangers: Acquaintanceship, agree
differences in facial expressions of emotion. ment, and the accuracy of personality judg
In J. Cole (Ed.), Nebraska Symposium on ment. Journal of Personality and Social
Motivation (pp. 207–283). Lincoln: Psychology, 55, 149–158.
University of Nebraska Press. Gilbert, D. T., Pelham, B. W., & Krull, D. S.
Ekman, P. (1994). Strong evidence for universals (1988). On cognitive busyness: When person
in facial expressions: A reply to Russell’s perceivers meet persons perceived. Journal
mistaken critique. Psychological Bulletin, of Personality and Social Psychology, 54,
115, 268–287. 733–740.
03-Manusov.qxd 6/30/2006 4:01 PM Page 56
56–––◆–––Foundations
Giles, H., & Smith, P. M. (1979). Accommodation Heider, F., & Simmel, M. (1944). An experimen
theory: Optimal levels of convergence. In tal study of apparent behavior. American
H. Giles & R. St. Clair (Eds.), Language Journal of Psychology, 57, 43–259.
and social psychology (pp. 45–65). Hummert, M. L., Mazloff, D., & Henry, C.
Baltimore: University Park Press. (1999). Vocal characteristics of older adults
Goffman, E. (1963). Behavior in public places. and stereotyping. Journal of Nonverbal
New York: Free Press. Behavior, 23, 111–132.
Gotlib, I. H., Kasch, K. L., Trail, S., Joorman, J., Ickes, W. (2001). Measuring empathic accuracy.
Arnow, B., & Johnson, S. L. (2004). In J. A. Hall & F. J. Bernieri (Eds.), Inter
Coherence and specificity of information- personal sensitivity: Theory and measure
processing biases in depression and social ment (pp. 219–241). Mahwah, NJ: Erlbaum.
phobia. Journal of Abnormal Psychology, Izard, C. E. (1979). The Maximally Discriminative
113, 386–398. Facial Movement Coding System (MAX).
Gottman, J. M. (1979). Marital interaction: Newark: University of Delaware, Infor
Empirical investigations. New York: mation Technologies and University Media
Academic Press. Services.
Gottman, J. M. (1994). What predicts divorce? Johansson, G. (1973). Visual perception of
Hillsdale, NJ: Erlbaum. biological motion and a model for its analy
Greenwald, A. G., Nosek, B. A., & Banaji, sis. Perception and Psychophysics, 14,
M. R. (2003). Understanding and using the 201–211.
Implicit Association Test: I. An improved Johnstone, T., & Scherer, K. R. (2000). Vocal
scoring algorithm. Journal of Personality communication of emotion. In M. Lewis
and Social Psychology, 85, 197–216. & J. Haviland (Eds.), Handbook of emo
Gregory, S. W., & Webster, S. (1996). A non tion (2nd ed., pp. 220–235). New York:
verbal signal in voices of interview partners Guilford.
effectively predicts communication accom Jones, B. C., Little, A. C., Burt, D. M., &
modation and social status perceptions. Perrett, D. I. (2004). When facial attrac
Journal of Personality and Social Psycho tiveness is only skin deep. Perception, 33,
logy, 70, 1231–1240. 569–576.
Guerrero, L. K., & Le Poire, B. A. (2005). Knight, B., & Johnston, A. (1997). The role
Nonverbal research involving experimen of movement in face recognition. Visual
tal manipulations by confederates. In Cognition, 4, 265–273.
V. Manusov (Ed.), The sourcebook of non Kozlowski, L. T., & Cutting, J. E. (1977).
verbal measures; Going beyond words Recognizing the sex of a walker from a
(pp. 507–522). Mahwah, NJ: Erlbaum. dynamic point-light display. Perception and
Haidt, J., & Keltner, D. (1999). Culture and Psychophysics, 21, 575–580.
facial expression: Open-ended methods find Kruglanski, A. W. (1989). The psychology
more expressions and a gradient of recogni of being “right”: The problem of accuracy
tion. Cognition and Emotion, 13, 225–266. in social perception and cognition.
Harris, M., & Rosenthal, R. (2005). No more Psychological Bulletin, 106, 395–409.
teachers’ dirty looks: Effects of teacher Lakin, J. L., & Chartrand, T. L. (2003). Using
nonverbal behavior on student outcomes. nonconscious behavioral mimicry to create
In R. E. Riggio & R. S. Feldman (Eds.), affiliation and rapport. Psychological Science,
Applications of nonverbal communication 14, 334–339.
(pp. 157–192). Mahwah, NJ: Erlbaum. LaPlante, D., & Ambady, N. (2000). Multiple
Heberlein, A. S, Adolphs, R., Tranel, D., & messages: Facial recognition advantage
Damasio, H. (2004). Cortical regions for for compound expressions. Journal of Non
judgments of emotions and personality verbal Behavior, 24, 211–224.
traits from point-light walkers. Journal of Marsh, A. A., Adams, R. B., & Kleck, R. E.
Cognitive Neuroscience, 16, 1143–1158. (2004). Why do fear and anger look the
03-Manusov.qxd 6/30/2006 4:01 PM Page 57
way they do? Form and social function in Patterson, M. L., & Stockbridge, E. (1998). Effects
facial expression. Personality and Social of cognitive demand and judgment strategy
Psychology Bulletin, 31, 73–86. on person perception accuracy. Journal of
Marsh, A. A., Elfenbein, H. A., & Ambady, N. Nonverbal Behavior, 22, 253–263.
(2003). Nonverbal “accents”: Cultural dif Richeson, J. A., & Shelton, J. N. (2005). Thin
ferences in facial expressions of emotion. slices of racial bias. Journal of Nonverbal
Psychological Science, 14, 373–376. Behavior, 29, 75–86.
Matsumoto, D., & Ekman, P. (1988). Japanese Riggio, R. E., & Riggio, H. R. (2005). Self-
and Caucasian facial expressions of emo report measures of emotional and nonver
tion (JACFEE) [Slides]. San Francisco: San bal expressiveness. In V. Manusov (Ed.),
Francisco State University, Department of The sourcebook of nonverbal measures:
Psychology, Intercultural and Emotion Going beyond words (pp. 105–112).
Research Laboratory. Mahwah, NJ: Erlbaum.
McArthur, L. Z., & Baron, R. M. (1983). Roberts, L. J. (2005). Conflict, real life, and
Toward an ecological theory of social per videotape: Procedures for eliciting natural
ception. Psychological Review, 90, 215–238. istic couple interactions. In V. Manusov
Mogg, K., Millar, N., & Bradley, B. P. (2000). (Ed.), The sourcebook of nonverbal mea
Biases in eye movements to threatening sures: Going beyond words (pp. 471–481).
facial expressions in generalized anxiety Mahwah, NJ: Erlbaum.
disorder and depressive disorder. Journal of Rosenthal, R., Hall, J. A., DiMatteo, M. R.,
Abnormal Psychology, 109, 695–704. Rogers, P. L., & Archer, D. (1979).
Neumann, R., & Strack, F. (2000). “Mood con Sensitivity to nonverbal communication:
tagion”: The automatic transfer of mood The PONS Test. Baltimore: Johns Hopkins
between persons. Journal of Personality University Press.
and Social Psychology, 79, 211–223. Rosenthal, R., & Rosnow, R. L. (1991).
Nisbett, R., & Wilson, T. (1977). Telling more Essentials of behavioral research: Methods
than we can know: Verbal reports on and data analysis (2nd ed.). Boston:
mental processes. Psychological Review, McGraw-Hill.
84, 231–259. Rosip, J. C., & Hall, J. A. (2004). Knowledge
Noller, P. (2005). Standard content methodol of nonverbal cues, gender, and nonverbal
ogy: Controlling the verbal channel. In decoding accuracy. Journal of Nonverbal
V. Manusov (Ed.), The sourcebook of Behavior, 28, 264–286.
nonverbal measures: Going beyond words Rothman, A. D., & Nowicki, S. (2004). A mea
(pp. 417–430). Mahwah, NJ: Erlbaum. sure of the ability to identify emotion
Nosek, B. A., Banaji, M. R., & Greenwald, in children’s tone of voice. Journal of
A. G. (2002). E-research: Ethics, security, Nonverbal Behavior, 29, 67–92.
design, and control in psychological research Russell, J. A. (1993). Forced-choice response
on the Internet. Journal of Social Issues, format in the study of facial expression.
58, 161–176. Motivation and Emotion, 17, 41–51.
Nowicki, S., & Duke, M. (1994). Individual Russell, J. A. (1994). Is there universal recogni
differences in the nonverbal communica tion of emotion from facial expressions?
tion of affect: The Diagnostic Analysis A review of the cross-cultural studies.
of Nonverbal Accuracy Scale. Journal of Psychological Bulletin, 115, 102–141.
Nonverbal Behavior, 789–735. Scherer, K. R. (2003). Vocal communication of
Patterson, M. L. (2005). The passing encounters emotion: A review of research paradigms.
paradigm: Monitoring microinteractions Speech Communication, 40, 227–256.
between pedestrians. In V. Manusov (Ed.), Siegman, A. W. (1987). The telltale voice:
The sourcebook of nonverbal measures: Nonverbal messages of verbal communica
Going beyond words (pp. 431–440). tion. In W. Siegman & S. Feldstein (Eds.),
Mahwah, NJ: Erlbaum. Nonverbal behavior and communication
03-Manusov.qxd 6/30/2006 4:01 PM Page 58
58–––◆–––Foundations
(2nd ed., pp. 351–433). Hillsdale, NJ: Vrij, A., & Mann, S. (2005). Police use of non
Erlbaum. verbal behaviors as indicators of deception.
Swann, W. B., Jr. (1984). Quest for accuracy in In R. E. Riggio & R. S Feldman. (Eds.),
person perception: A matter of pragmatics. Applications of nonverbal communication
Psychological Review, 91, 457–477. (pp. 63–94). Mahwah, NJ: Erlbaum.
Tassinary, L. G., & Cacioppo, J. T. (2000). The White, C. H., & Sargent, J. (2005). Researcher
skeletomotor system: Surface electromyog choices and practices in the study of non
raphy. In J. T. Cacioppo, L. G. Tassinary, verbal communication. In V. Manusov
& G. Berntson (Eds.), Handbook of psy (Ed.), The sourcebook of nonverbal mea
chophysiology (2nd ed., pp. 163–199). sures: Going beyond words (pp. 3–21).
New York: Cambridge University Press. Mahwah, NJ: Erlbaum.
Tiedens, L. Z., & Fragale, A. R. (2003). Power Wilcox, B., & Clayton, F. (1968). Infant visual
moves: Complementarity in dominant and fixation on motion pictures of the human
submissive nonverbal behavior. Journal face. Journal of Experimental Child Psycho
of Personality and Social Psychology, 84, logy, 6, 22–32.
558–568. Zebrowitz, L.A. (1997). Reading faces: Window
Tracy, J. L., & Robins, R. W. (2004). Show to the soul? Boulder, CO: Westview Press.
your pride: Evidence for a discrete emo Zebrowitz, L. A., & Rhodes, G. (2004).
tion expression. Psychological Science, 15, Sensitivity to “bad genes” and the anom
194–197. alous face overgeneralization effect: Cue
Vrij, A., Edward, K., Roberts, K. P., & Bull, R. validity, cue utilization, and accuracy in
(2000). Detecting deceit via analysis of judging intelligence and health. Journal of
verbal and nonverbal behavior. Journal of Nonverbal Behavior, 28, 167–185.
Nonverbal Behavior, 24, 239–263.
04-Manusov.qxd 6/30/2006 7:23 PM Page 59
4
AUTOMATIC COGNITIVE
PROCESSES AND NONVERBAL
COMMUNICATION
� Jessica L. Lakin
Drew University
◆ 59
04-Manusov.qxd 6/30/2006 7:23 PM Page 60
60–––◆–––Foundations
automatic cognitive bases of nonverbal of will to start the process, whereas control
behaviors. To accomplish this objective, lability refers to the fact that an act of
I first discuss what it means for a cognitive will can stop the process once it has been
process to be (relatively) automatic or (rela started (i.e., it does not run to conclusion
tively) controlled. This is followed by specific autonomously). Finally, and despite the lim
examples of nonverbal communication that ited amount of cognitive resources that
appear to be relatively automatic. I then con people have and the already high demands
clude with a discussion of important issues on these resources, controlled processes
and future directions, including methodolog require a share of these limited attentional
ical considerations, the role of controlled resources (i.e., they require cognitive effort).
processes, and the importance of cognitive Automatic processes are more difficult
resources. to define. Because the four characteristics
associated with controlled processes do not
occur in an all-or-none fashion, it is not
♦ Automatic and correct to assume that automatic processes
are, by default, characterized by unaware
Controlled Processes
ness, unintentionality, and uncontrollability
and require no cognitive effort. Although a
In 1975, Posner and Snyder considered a process that has these four characteristics
basic question of human existence: How would certainly be considered automatic,
much control do people have over their processes that are characterized by one,
thoughts, behaviors, and decisions? Since two, or even three of these features have
they posed their question, researchers have also been referred to historically as auto
demonstrated that much of what we do cog matic (Bargh, 1994, 1996, 1997). It is also
nitively happens without intention, aware possible for various combinations of these
ness, or conscious control (Hassin et al., four basic characteristics to occur. For
2005). Although the history of automaticity example, experienced drivers intend to get
and automatic processes has been reviewed in a car and go somewhere, even if when
in extensive detail elsewhere (Bargh, 1994, they arrive, they have no conscious aware
1996, 1997; Bargh & Chartrand, 1999), ness of anything that occurred during the
a basic discussion of what it means for a trip. This example demonstrates that there
process to be more controlled or more auto are several types of automatic processes,
matic seems necessary before turning atten resulting in a continuum of automaticity
tion to specific examples of nonverbal (ranging from completely automatic to
communication and their relationship to completely controlled) rather than a simple
automatic cognitive processes. dichotomy (Bargh, 1996, 1997; Bargh &
Controlled processes are characterized Chartrand, 1999).
by awareness, intentionality, controllability, Specifically, preconscious automaticity
and cognitive effort (Bargh, 1994, 1996; represents the completely automatic end of
Posner & Snyder, 1975; Shiffrin & Schneider, the continuum, because it corresponds to
1977). Aspects of awareness include recog the initial unconscious processing of incom
nizing the cognitive process or stimulus ing environmental information. This analy
consciously, but they also involve an sis occurs without intention, control, or
acknowledged recognition of the influence awareness, and it is largely effortless. Goal-
that the process or stimulus is having. directed automaticity, however, represents
Intentionality refers to the necessity of an act a point somewhere in the middle of the
04-Manusov.qxd 6/30/2006 7:23 PM Page 61
62–––◆–––Foundations
that are temporarily accessible (i.e., primed; activating stereotypes for African Americans
as in the work by Higgins et al., 1977; Srull causes hostility (Bargh et al., 1996) and
& Wyer, 1979) and that salient physical decreased intellectual performance (Steele &
cues (e.g., sex, race) can automatically acti Aronson, 1995; Wheeler, Jarvis, & Petty,
vate stereotypes associated with particular 2000). As evidenced by the social inappro
groups of people (Brewer, 1988; Devine, priateness of many of these behavioral
1989; Fiske & Neuberg, 1990). Together, effects, participants are not aware that they
this work demonstrates that people make are occurring and are not being affected
inferences about the behaviors of others, intentionally, nor do they seem to be able to
without intention, control, effort, or aware control the effects that the activation of the
ness of having done so. Many of these various constructs is having.
behaviors are nonverbal cues.
ATTITUDES
BEHAVIOR
Just as impressions are formed automati
Behavior is also affected automatically by cally when traits are accessible, either chroni
external stimuli. In the first demonstration of cally or temporarily, evaluations of stimuli
this idea, Bargh et al. (1996) showed that are activated automatically when the stimuli
activating the trait “rude” caused people to are presented. This automatic evaluation
interrupt an experimenter who was ostensi work can be traced back to Zajonc (1980),
bly helping a confederate more quickly than who argued that the evaluation of a stimulus
did participants who had the trait “polite” is connected closely to the representation of
activated. This finding has been replicated the stimulus itself. That is, when the stimu
in a number of behavioral domains: Activating lus is presented, a positive or negative eval
“politicians” causes long-windedness (Dijk uation of that stimulus becomes activated
sterhuis & van Knippenberg, 2000), and acti without awareness, effort, or intention. In
vating “supermodels” causes poor performance support of this idea, research has demon
on a trivia test, whereas activating “profes strated that attitudes are activated automat
sors” causes better performance (Dijksterhuis ically when a stimulus is presented (Bargh,
& van Knippenberg, 1998). Chaiken, Govender, & Pratto, 1992; Fazio,
Likewise, activating traits associated Sanbonmatsu, Powell, & Kardes, 1986).
with conformity causes people to agree Using a procedure similar to the activation
more with a group of confederates (Epley & of impressions and behaviors, Kawakami,
Gilovich, 1999); activating helpfulness traits Dovidio, and Dijksterhuis (2003) have even
causes people to be more helpful (Macrae & found that when a stereotype is made acces
Johnston, 1998); activating aggression- sible, people report attitudes consistent
related ideas causes people to give longer with the stereotypical attitudes of that
“shocks” to another participant (Carver, group. For example, when the category of
Ganellen, Froming, & Chambers, 1983); “elderly” is accessible, people report more
activating the elderly stereotype causes conservative attitudes, and when the cate
people to walk more slowly (Bargh et al., gory “skinhead” is accessible, people report
1996), increases slowness on a lexical more prejudiced attitudes. These effects
decision task (Dijksterhuis, Spears, & occur even when the categories were made
Lepinasse, 2001), and promotes poor accessible to participants with a subliminal
memory (Dijksterhuis, Aarts, Bargh, & priming procedure, again suggesting that this
van Knippenberg, 2000; Levy, 1996); and effect occurs without conscious awareness.
04-Manusov.qxd 6/30/2006 7:23 PM Page 63
64–––◆–––Foundations
communication directly (for another review, interpersonal distance, and eye contact
see Choi, Gray, & Ambady, 2005). Very (Bernieri, 1988; Bernieri, Davis, Rosenthal,
little research on nonverbal communication, & Knee, 1994; Grahe & Bernieri, 1999; see
however, has been conducted in such a way Tickle-Degnen, this volume). Marital status
as to show definitive evidence that the effects can even be determined via nonverbal
are occurring automatically (i.e., by reduc means, given the surprising finding that
ing the likelihood of consciousness being the longer people have been married, the
involved, by using funneled-debriefing pro more similar they look (Zajonc, Adelmann,
cedures, by having awareness checks, etc.). Murphy, & Neidenthal, 1987).
Nevertheless, there are several programs of Is the encoding and decoding of this
research that demonstrate automatic non social relationship information occurring
verbal communication indirectly. A sensitive automatically? As the majority of this
reader will no doubt note that the subcate research has not been conducted with an
gories listed below correspond with some of eye toward answering this specific question,
the chapters in the Contexts and Functions the answer must be inferred from the
sections of this Handbook. A brief, non- methodologies that have been used, as well
comprehensive review of these topics is cov as some of the research findings. This infor
ered here to demonstrate that some of these mation suggests several reasons to accept
effects occur automatically. the argument that these processes are
occurring automatically. First, decoding of
social relationship information, as evi
SOCIAL RELATIONSHIPS denced by research using the IPT, often
occurs quickly; clips from the task are less
Information about social relationships is than one minute, suggesting that people’s
relatively easy to infer from people’s nonver decoding of this information must also
bal behaviors. In terms of social status, dom occur very quickly (reducing the likelihood
inant people tend to stand taller, use more of conscious involvement). Second, certain
personal space, interrupt more, and talk types of conscious instructions interfere
louder than less dominant people (Henley, with the accuracy of decoding relationship
1977; Burgoon & Dunbar, this volume; but information (e.g., Patterson & Stockbridge,
see Hall, 2005). Powerful people look at 1998; Patterson, this volume). When this
others the same amount when speaking as type of interference happens, it suggests
when listening, but less powerful people that people have developed automatic
look more when listening than when they are and efficient strategies for processing the
speaking (Fehr & Exline, 1987). Relationship information, and conscious attention inter
status can also be inferred from nonverbal feres with the use of the strategies on which
behaviors. Research using the Interpersonal people typically rely.
Perception Task (IPT) (Costanzo & Archer, Third, participants in studies such as the
1989; see Riggio, this volume), which con ones reviewed above can rarely identify the
tains video clips of less than a minute, has factors that influenced their judgments
demonstrated that one can accurately iden (e.g., Bernieri et al., 1994). One study even
tify when people are related and when they found that confederates who were
are in significant relationships (Smith, instructed to create liking with a partner
Archer, & Costanzo, 1991). Rapport could not identify accurately the behaviors
can also be determined from watching they used and how they used them, even
people interact, at least partially because rap though they were successful at their goal
port is related to interactional synchrony, (Palmer & Simmons, 1995). Research on
04-Manusov.qxd 6/30/2006 7:23 PM Page 65
66–––◆–––Foundations
re-create without truly being fearful (Ekman, suggesting that this decoding can occur
1985). Finally, a recent meta-analysis sug automatically.
gests that observers can detect accurately the
state and trait anxiety of others, although
the magnitude of these effects depends on PREJUDICE
the communication channel (Harrigan,
Wilson, & Rosenthal, 2004). Because anxi Prejudice, or negative feelings about
ety is a negative emotional state that most groups of people, can also be communi
people would not want to convey to others, cated automatically through nonverbal cues
this information is probably being encoded (see Dovidio, Hebl, Richeson, & Shelton,
without intention or awareness. this volume). Because categorical informa
Tone of voice may also convey informa tion (e.g., race, sex; Brewer, 1988; Fiske &
tion automatically about a speaker’s emo Neuberg, 1990) is processed automatically
tional state. Neumann and Strack (2000) in most cases, simply seeing a member of
had participants listen to a speech that was a stereotyped group can activate, without
delivered in either a slightly happy or a awareness, thoughts and feelings about that
slightly sad voice. After hearing the text, group (Blair, 2002; Devine, 1989; Fiske,
participants rated their own mood. People’s 1998). It has been argued that a dissocia
moods were affected by the tone in which tion often exists between people’s conscious
the passage was read and, consistent with and unconscious attitudes toward minority
emotional contagion, participants automat groups; therefore, researchers have been
ically encoded this emotional information. able to measure both and determine the
When they repeated the text that they effects that each has on social interactions.
heard, they imitated the tone of the original This work has shown that nonverbal
reading spontaneously. This encoding behaviors like gaze, blinking, body posture,
occurred despite the fact that participants and interpersonal distance can indicate preju
were not consciously aware of the original diced feelings toward interaction partners
tone of the message and were not given (Crosby, Bromley, & Saxe, 1980; McConnell
instructions to mimic the original tone. & Liebold, 2001; Towles-Schwen & Fazio,
The automaticity of encoding of emo 2003). In one demonstration, people who
tional expressions is demonstrated most indicated negative feelings toward Blacks on
clearly in research by Ulf Dimberg and an implicit measure of prejudice also blinked
colleagues (Dimberg 1982, 1997; Dimberg, more (a sign of uncomfortableness) and
Thunberg, & Elmehed, 2000). In his work, made less eye contact during an interaction
Dimberg had participants look at pictures with an African American partner (Dovidio,
of people displaying happy or angry facial Kawakami, Johnson, Johnson, & Howard,
expressions. He then measured their subtle 1997). Dovidio, Kawakami, and Gaertner
facial movements with electromyographic (2002) replicated this effect and showed that
(EMG) technology. The results showed that the less positive nonverbal behaviors of par
participants moved the muscles associated ticipants who had implicit racial biases led
with smiling when looking at the happy the African American partners of these par
pictures and the muscles associated with ticipants to feel that the interaction had been
frowning when looking at the angry pic less friendly.
tures. This effect occurred even when the Likewise, Vanman, Saltz, Nathan, and
facial expressions were presented outside Warren (2004) have also shown that facial
of conscious awareness (i.e., subliminally), EMG activity can be related to another
04-Manusov.qxd 6/30/2006 7:23 PM Page 67
person’s race; participants who had cheek instructions (Godfrey, Jones, & Lord,
EMG activity (i.e., activity that would be 1986). People copy the postures of interac
associated with smiling) when viewing tion partners more when they are told those
pictures of White targets were more likely partners might be helpful than when oppor
to nominate a White applicant for a tunities to get help from the partners are not
prestigious award than a Black applicant. available (La France, 1985). Nonverbal
Participants in these studies were not aware behaviors are also dependent on the people
of the negative nonverbal behaviors they for whom the display occurs (see DePaulo
encoded or the fact that these nonverbal & Friedman, 1998, for a review). Finally,
behaviors were affecting their interactions the ecological theory of perception suggests
negatively. This work therefore provides that people are able to gain, relatively auto
evidence that both encoding and decoding matically, information about other people
of prejudicial feelings can occur automati from their appearances and movements,
cally. In fact, as Dovidio et al. (2002) have information that Gibson (1979) calls affor
discussed, researchers first turned to non dances. This information is able to be gath
verbal behaviors as indicators of prejudice ered easily because appearance and behavior
because of their hypothesized “leakiness” express some personality characteristics
(i.e., uncontrollability) and spontaneity. accurately (Zebrowitz & Collins, 1997).
Nonverbal communication is therefore
sensitive to conscious goals and situational
IMPRESSION MANAGEMENT constraints. Although this strategic use of
AND PERSONALITY EXPRESSION nonverbal behaviors suggests that their cog
nitive basis is conscious, being strategic does
Although impression management and not demand consciousness. There are several
self-presentation can certainly be conscious lines of research that suggest that people
(see Keating, this volume), these commu alter their nonverbal behaviors automati
nicative functions can also occur relatively cally. Tice, Butler, Muraven, and Stillwell
effortlessly and without conscious guid (1995) hypothesized that people’s default
ance, particularly in cases where people self-presentational strategy with strangers is
are not particularly motivated to convey a to present a positive image, whereas people’s
desired identity. Moreover, it is also clear default self-presentational strategy with
that people may infer traits and make attri friends is to present a modest image.
butions from the nonverbal behaviors of Consistent with this proposition, when par
others automatically and that these judg ticipants behaved consistently with their
ments can be quite accurate. Together, this default tendencies, their self-presentations
literature suggests that people often express used few cognitive resources, which led to
their personality and decode personality increased memory of the details of the inter
information about others automatically action. Participants who used nondefault
through nonverbal behaviors. self-presentational strategies (e.g., modesty
People tend to use different self-presenta with strangers) were not able to pursue these
tional strategies in different situations and strategies automatically and used conscious
with different types of people. For example, cognitive resources to accomplish their goals
when people are instructed to make an (see also Patterson, Churchill, Farag, &
interaction partner like them, they nod, Borden, 1991/1992). Other research has
smile, and make more eye contact than also demonstrated that self-presenting in a
do people who did not receive these way that is inconsistent with personality is
04-Manusov.qxd 6/30/2006 7:23 PM Page 68
68–––◆–––Foundations
cognitively taxing and requires attentional from nonverbal presentations and that
resources (Pontari & Schlenker, 2000). length of observation time did not signifi
Attributional processes, whereby people cantly affect accuracy; accuracy was just as
infer traits and make dispositional or good when people observed behavior for
situational attributions, can also happen 30 seconds as it was when people observed
automatically. An extensive literature on behavior for five minutes (Ambady &
spontaneous trait inferences has demon Rosenthal, 1992). The fact that people
strated that simply presenting people with can be so accurate at decoding in such
descriptions of other people’s behaviors short time frames suggests that the process
results in inferring traits from them is likely not occurring with much conscious
(Uleman, 1999; Uleman et al., 1996). intention or guidance. Moreover, when
Participants in these studies were not conscious attention is devoted to this type
instructed to infer traits and typically even of task, decoding accuracy decreases (Gilbert
denied that they had done so (Winter & & Krull, 1988; Patterson & Stockbridge,
Uleman, 1984). Related research has found 1998).
that people might also make emotion
inferences when reading descriptions of
other’s behavior automatically (Gernsbacher, EXPECTANCIES
Goldsmith, & Robertson, 1992). Finally,
people make both dispositional and situa Just as people convey personality through
tional attributions without the use of limited their nonverbal behaviors, expectancies that a
cognitive resources or conscious intention person holds for others are also conveyed
(Gilbert, Pelham, & Krull, 1988; Krull, through nonverbal behaviors, seemingly with
1993). Thus, when people perceive the out intention or awareness (see Remland,
behaviors of others, they seem to decode this volume). Communication of these
this information effortlessly and uninten expectancies—in certain contexts and under
tionally, at least in most circumstances. certain conditions—results in a self-fulfilling
Automatic self-presentation, impression prophecy, whereby perceivers elicit behav
management, trait inferences, and attribu iors that are consistent with their original
tions are likely to be beneficial to most ideas and then conclude that their expectan
people in most situations but only to the cies were truthful, without realizing the role
extent that these relatively effortless catego that they have played in confirming them
rizations are correct. Ambady and her col (Rosenthal, 2003). For example, researchers’
leagues have demonstrated that people’s behaviors can affect the responses of their
automatic decodings of the behaviors of participants (Rosenthal, 1976), and the
others are remarkably accurate (Ambady, expectations of teachers have been found
Bernieri, & Richeson, 2000; Gray & to affect the behaviors of their students
Ambady, this volume). In one empirical (Rosenthal & Jacobson, 1968). That people
demonstration, participants watched video are unaware that they have elicited the
clips of teachers’ nonverbal behaviors, behavior they expected suggests that these
some as short as six seconds, and then expectancies are being conveyed automati
judged teacher effectiveness. Participants’ cally. This is particularly the case when nega
ratings were strongly correlated with tive expectancies are conveyed, which would
students’ end-of-semester ratings (Ambady likely be controlled if people were aware of
& Rosenthal, 1993). them (e.g., Word, Zanna, & Cooper, 1974).
A meta-analysis has revealed a medium to There is also evidence that people are
large effect size for accuracy of predictions affected automatically by the expectancies
04-Manusov.qxd 6/30/2006 7:23 PM Page 69
that others have for them (i.e., automatic conscious awareness (e.g., Bavelas, Black,
decoding of expectancies). Word et al. Lemery, & Mullett, 1986; Bernieri, 1988;
(1974) showed that when White partici Bernieri, Reznick, & Rosenthal, 1988;
pants in a second study were treated as La France, 1979, 1982; La France &
Black participants had been treated in a first Broadbent, 1976). Because this research
study (e.g., abrupt questioning, lack of eye does not typically include awareness checks
contact), the quality of the interaction or attempts to rule out conscious involve
decreased. The White participants stuttered ment, the automaticity of these effects is,
more and made less eye contact as a result of again, inferred from participants’ apparent
the expectancies for a less successful inter lack of awareness.
view that the interviewer was conveying. There is more definitive evidence that
Chen and Bargh (1997) found a similar behavioral mimicry can occur noncon
result: When stereotypes of African Americans sciously, however, and that this automatic
were activated outside of conscious aware mimicry results in smoother interactions
ness, participants treated an interaction and the development of liking. Chartrand
partner in a way that was consistent and Bargh (1999) found that participants
with this stereotype, eliciting more hostility. mimicked the nonverbal behaviors of a con
Participants were unaware that the stereo federate without conscious awareness. In a
type had been activated and were unaware second study, they found that mimicry
that their own behavior was influencing the leads to increased liking for the mimicker
behavior of their interaction partner. The and smoother interactions. Thus, mimick
results of this study are consistent with ing others and being mimicked can commu
many other research findings that stereotype nicate rapport automatically. This finding
activation results in stereotype-consistent is consistent with the work of Lakin and
behaviors (see Wheeler & Petty, 2001, for Chartrand (2003), who found that partici
a review). Again, these studies demonstrate pants who pursue an unconscious affilia
consistently that people are not aware that tion goal are more likely to mimic the
their behavior is affected or is being affected behaviors of an interaction partner than
by the expectancies of others. participants who do not have an active
affiliation goal (see also Lakin, Jefferis,
Cheng, & Chartrand, 2003).
RAPPORT A second study expanded on this finding
by giving participants who had a noncon
Not all behaviors that occur outside of sciously activated affiliation goal a success
awareness are problematic. For example, or failure experience (Lakin & Chartrand,
many people have been interested in the 2003). Participants either succeeded (i.e.,
relationship between posture sharing (either had a pleasant interaction with a confeder
mimicking or mirroring) and the develop ate) or failed at their goal (i.e., had a rela
ment and maintenance of rapport (for a tively less pleasant interaction with a
review, see Tickle-Degnen, this volume). confederate). In a subsequent interaction
The fact that people mimic the nonverbal with a different confederate, participants
behaviors of others (both significant others who still had affiliation goals (i.e., those who
and strangers) has long been established initially failed) mimicked the nonverbal
(Chartrand et al., 2005). Further work has behaviors of their interaction partner more
demonstrated that people tend to mimic than participants who had been successful.
others and thereby demonstrate or develop This effect occurred despite the fact that par
rapport without intention, control, or ticipants did not consciously know they were
04-Manusov.qxd 6/30/2006 7:23 PM Page 70
70–––◆–––Foundations
pursuing an affiliation goal and did not typically compares participants who are
consciously acknowledge the confederate’s mimicked with participants who are not
behaviors or how the confederate’s behav mimicked, these effects occur arguably as a
iors were affecting their own behaviors. direct result of the positive feelings that
A recent social exclusion experience may occur automatically with mimicry. The
also lead to increases in behavioral mimicry majority of this work includes awareness
(Lakin et al., 2005; Lakin & Chartrand, checks and other methodological strategies
2005). The need to belong is one of the that reduce the role of consciousness in
strongest core motivations, and it affects these effects (Bargh & Chartrand, 2000).
people’s thoughts, feelings, and behaviors
frequently (Williams, Forgas, & von Hippel,
2005). It is therefore not surprising that ♦ Discussion and
when people have been excluded from a Future Directions
social group, they will engage in behaviors
that help them to create liking and allow
them to reenter the group. Recent research As the literatures reviewed above demon
has shown that mimicking the nonverbal strate, nonverbal behaviors can communi
behaviors of group members may be one cate information automatically about social
such strategy (Lakin et al., 2005). When relationships, emotions, prejudice, person
participants were excluded from a comput ality, and expectations, as well as indicate
erized ball-tossing game and then interacted relationship status and rapport. As a whole,
with a confederate in a different context, this research compellingly argues that non
they mimicked the behaviors of the confed verbal communication has, at least in part,
erate more than when they had not been its basis in automatic processes. There are,
excluded during the ball-toss game. In other however, several important topics that
words, participants who were trying to cre deserve research attention.
ate liking and affiliation were able to pur
sue this goal through mimicking another METHODOLOGICAL CONCERNS
person. Together with the findings from
Lakin and Chartrand (2003), this work Whereas not all the research reviewed here
suggests that people can pursue a goal to has provided definitive evidence that the
develop rapport or liking by mimicking the processes are occurring without awareness,
behaviors of others automatically. intention, control, or cognitive effort, the
Finally, paradigms that involve having a several lines of research that do provide
confederate mimic the behaviors of partici this evidence demonstrate that nonverbal
pants have demonstrated that being mimic communication can and does occur auto
ked causes people to be more interdependent matically. Moreover, the research that does
(van Baaren, Maddux, Chartrand, de Bouter, not contain definitive evidence often has
& van Knippenberg, 2003), be more helpful methodological characteristics suggesting
(van Baaren, Holland, Kawakami, & van that people are not aware of what their non
Knippenberg, 2004), and provide bigger tips verbal behaviors are communicating or the
(van Baaren, Holland, Steenaert, & van extent to which their own thoughts, feelings,
Knippenberg, 2003; for an overview, see and behaviors are being affected by the non
Tickle-Degnen, this volume). Mimicry also verbal communications of others.
leads to greater persuasion (Bailenson & Although this suggests that a significant
Yee, 2005). Because work on mimicry component of nonverbal communication
04-Manusov.qxd 6/30/2006 7:23 PM Page 71
72–––◆–––Foundations
social cognition (2nd ed., pp. 1–40). Personality and Social Psychology Bulletin,
Hillsdale, NJ: Erlbaum. 20, 303–311.
Bargh, J. A. (1996). Automaticity in social Bernieri, F. J., Reznick, J. S., & Rosenthal,
psychology. In E. T. Higgins & A. W. R. (1988). Synchrony, pseudosynchrony,
Kruglanski (Eds.), Social psychology: and dissynchrony: Measuring the entertain
Handbook of basic principles (pp. 169– ment process in mother–infant interactions.
183). New York: Guilford Press. Journal of Personality and Social Psycho
Bargh, J. A. (1997). The automaticity of everyday logy, 54, 243–253.
life. In R. S. Wyer (Ed.), The automaticity of Blair, I. V. (2002). The malleability of automatic
everyday life: Advances in social cognition stereotypes and prejudice. Personality and
(Vol. 10, pp. 1–61). Mahwah, NJ: Erlbaum. Social Psychology Review, 6, 242–261.
Bargh, J. A., Chaiken, S., Govender, R., & Brewer, M. B. (1988). A dual process model of
Pratto, F. (1992). The generality of the impression formation. In T. K. Srull & R. S.
automatic attitude activation effect. Journal Wyer (Eds.), Advances in social cognition
of Personality and Social Psychology, 62, (Vol. 1, pp. 1–36). Hillsdale, NJ: Erlbaum
893–912. Associates.
Bargh, J. A., & Chartrand, T. L. (1999). The Buck, R. (1984). The communication of emo
unbearable automaticity of being. American tion. New York: Guilford.
Psychologist, 54, 462–479. Carver, C. S., Ganellen, R. J., Froming, W. J.,
Bargh, J. A., & Chartrand, T. L. (2000). The & Chambers, W. (1983). Modeling: An
mind in the middle: A practical guide to analysis in terms of category accessibility.
priming and automaticity research. In H. T. Journal of Experimental Social Psychology,
Reis & C. M. Judd (Eds.), Handbook of 19, 403–421.
research methods in social and personality Chartrand, T. L., & Bargh, J. A. (1996).
psychology (pp. 253–285). New York: Automatic activation of impression forma
Cambridge University Press. tion and memorization goals: Nonconscious
Bargh, J. A., Chen, M., & Burrows, L. (1996). goal priming reproduces effects of explicit
Automaticity of social behavior: Direct task instructions. Journal of Personality and
effects of trait construct and stereotype acti Social Psychology, 71, 464–478.
vation on action. Journal of Personality and Chartrand, T. L., & Bargh, J. A. (1999). The
Social Psychology, 71, 230–244. chameleon effect: The perception–behavior
Bargh, J. A., Gollwitzer, P. M., Lee-Chai, A., link and social interaction. Journal of
Barndollar, K., & Trötschel, R. (2001). The Personality and Social Psychology, 76,
automated will: Nonconscious activation 893–910.
and pursuit of behavioral goals. Journal of Chartrand, T. L., Maddux, W. W., & Lakin, J.
Personality and Social Psychology, 81, L. (2005). Beyond the perception–behavior
1014–1027. link: The ubiquitous utility and motiva
Bavelas, J. B., Black, A., Lemery, C. R., & tional moderators of nonconscious mim
Mullett, J. (1986). “I show how you feel”: icry. In R. R. Hassin, J. S. Uleman, &
Motor mimicry as a communicative act. J. A. Bargh (Eds.), The new unconscious
Journal of Personality and Social Psycho (pp. 334–361). New York: Oxford
logy, 50, 322–329. University Press.
Bernieri, F. J. (1988). Coordinated movement Chen, M., & Bargh, J. A. (1997). Nonconscious
and rapport in teacher-student interac behavioral confirmation processes: The self-
tions. Journal of Nonverbal Behavior, 12, fulfilling nature of automatically-activated
120–138. stereotypes. Journal of Experimental Social
Bernieri, F. J., Davis, J. M., Rosenthal, R., & Psychology, 33, 541–560.
Knee, C. R. (1994). Interactional synchrony Choi, Y. S., Gray, H. M., & Ambady, N. (2005).
and rapport: Measuring synchrony in dis The glimpsed world: Unintended communi
plays devoid of sound and facial affect. cation and unintended perception. In
04-Manusov.qxd 6/30/2006 7:23 PM Page 74
74–––◆–––Foundations
76–––◆–––Foundations
Personality and Social Psychology, 69, of Personality and Social Psychology, 84,
797–811. 1093–1102.
Tice, D. M., Butler, J. L., Muraven, M. B., & Vanman, E. J., Saltz, J. L., Nathan, L. R., &
Stillwell, A. M. (1995). When modesty Warren, J. A. (2004). Racial discrimination
prevails: Differential favorability of self- by low-prejudiced Whites facial movements
presentation to friends and strangers. Journal as implicit measures of attitudes related
of Personality and Social Psychology, 69, to behavior. Psychological Science, 15,
1120–1138. 711–714.
Tiedens, L. Z., & Fragale, A. R. (2003). Power Wheeler, S. C., Jarvis, W. B. G., & Petty, R. E.
moves: Complementarity in dominant and (2000). Think unto others: The self-destructive
submissive nonverbal behavior. Journal impact of negative racial stereotypes.
of Personality and Social Psychology, 84, Journal of Experimental Social Psychology,
558–568. 37, 173–180.
Towles-Schwen, T., & Fazio, R. H. (2003). Wheeler, S. C., & Petty, R. E. (2001). The effects
Choosing social situations: The relation of stereotype activation on behavior: A
between automatically activated racial atti review of possible mechanisms. Psychological
tudes and anticipated comfort interacting Bulletin, 127, 797–826.
with African Americans. Personality and Williams, K. D., Forgas, J. P., & von Hippel, W.
Social Psychology Bulletin, 29, 170–182. (Eds.). (2005). The social outcast: Ostracism,
Uleman, J. S. (1999). Spontaneous versus inten social exclusion, rejection, and bullying.
tional inferences in impression forma New York: Psychology Press.
tion. In S. Chaiken & Y. Trope (Eds.), Winter, L., & Uleman, J. S. (1984). When are
Dual-process theories in social psychology social judgments made? Evidence for the
(pp. 141–160). New York: Guilford Press. spontaneousness of trait inferences. Journal
Uleman, J. S., Newman, L. S., & Moskowitz, of Personality and Social Psychology, 47,
G. B. (1996). People as flexible interpreters: 237–252.
Evidence and issues from spontaneous trait Word, C. O., Zanna, M. P., & Cooper, J.
inference. In M. P. Zanna (Ed.), Advances in (1974). The nonverbal mediation of self-
Experimental Social Psychology (Vol. 28, fulfilling prophecies in interracial interac
pp. 179–211). San Diego, CA: Academic Press. tion. Journal of Experimental Social
van Baaren, R. B., Holland, R. W., Kawakami, K., Psychology, 10, 109–120.
& van Knippenberg, A. (2004). Mimicry and Zajonc, R. B. (1980). Feeling and thinking:
prosocial behavior. Psychological Science, Preferences need no inferences. American
15, 71–74. Psychologist, 35, 151–175.
van Baaren, R. B., Holland, R. W., Steenaert, B., Zajonc, R. B., Adelmann, P. K., Murphy, S. T.,
& van Knippenberg, A. (2003). Mimicry & Neidenthal, P. M. (1987). Convergence
for money: Behavioral consequences of in the physical appearance of spouses.
imitation. Journal of Experimental Social Motivation and Emotion, 11, 335–346.
Psychology, 39, 393–398. Zebrowitz, L. A., & Collins, M. A. (1997).
van Baaren, R. B., Maddux, W. W., Chartrand, Accurate social perception at zero acquain
T. L., de Bouter, C., & van Knippenberg, A. tance: The affordances of a Gibsonian
(2003). It takes two to mimic: Behavioral approach. Personality and Social Psychology
consequences of self-construals. Journal Review, 1, 204–223.
04-Manusov.qxd 6/30/2006 7:23 PM Page 78
05-Manusov.qxd 6/30/2006 7:24 PM Page 79
5
NONVERBAL SKILLS AND ABILITIES
� Ronald E. Riggio
Kravis Leadership Institute,
Claremont McKenna College
◆ 79
05-Manusov.qxd 6/30/2006 7:24 PM Page 80
80–––◆–––Foundations
concrete abilities. The third shifts the Inventory (EQI; Bar-On, 1997). Its measures
concept of nonverbal skill from a study of of emotional intelligence resemble self-report
inferred states to the study of process. A personality assessments typically.
clear example of this last shift is illustrated Not unlike emotional intelligence, non
by research on expectancy effects (Harris & verbal skills are important for success in var
Rosenthal, 2005; Rosenthal & Jacobson, ious aspects of social life. Specifically, there is
1968) that prompted scholars to try to evidence that nonverbal skills and abilities
understand the “process” of how positive are important in initiating and maintaining
(and negative) expectations are conveyed. social interaction, developing interpersonal
The work by researchers who were focus relationships, and managing impressions.
ing on individual differences in nonverbal Nonverbal skills and abilities are also linked
communication skill represented the ground to stress management and to success in
work for the highly popular concept of careers in various business settings (Riggio,
emotional intelligence (Goleman, 1995, 1992a, 2005). Importantly, nonverbal skills,
1998; Mayer & Salovey, 1997; Salovey & unlike personality traits, can be learned and
Mayer, 1990). Indeed, Friedman (1979) improved. For example, research on decep
used the analogy of an intelligence test, with tion detection suggests that this decoding
its “right” and “wrong” answers, when dis ability improves by providing feedback
cussing the measurement advantages of a concerning performance accuracy and with
nonverbal skill approach to personality over practice (Zuckerman, Koestner, & Alton,
traditional traitlike measures. The same 1984; Zuckerman, Koestner, & Colella,
measurement distinction divides the emo 1985), as well as by providing decoders
tional intelligence research community with information concerning more accurate non
its distinction between the “abilities model” verbal clues to deception (de Turck,
and the “mixed model” (Caruso, Mayer, & Harszlak, Bodhorn, & Texter, 1990; see also
Salovey, 2002; Mayer, Salovey, & Caruso, Patterson, Foster, & Bellmer, 2001). There is
2000) for explaining and describing emo also evidence that nonverbal expressiveness
tional intelligence. The abilities model of can be improved through training (Taylor,
emotional intelligence focuses on abilities to 2002; see also Vrij, this volume).
perceive, understand, use, and manage emo Following the belief in the importance of
tions, and it is represented by the Multifactor nonverbal skills for an array of outcomes,
Emotional Intelligence Scale (MEIS) and the this chapter reviews theory and research
Mayer, Salovey, Caruso, Emotional Intellig on the skill approach to nonverbal com
ence Test (MSCEIT) (Mayer, Salovey, & munication by focusing on three general
Caruso, 1997, 2002) measures. domains of nonverbal skills and abilities:
Both these instruments are performance- (1) nonverbal decoding skill, (2) nonverbal
based assessments of emotional abilities, encoding skill, and (3) skill in regulating
many of which are central to skill in nonver nonverbal communication. In addition to
bal communication and, like most measures defining these skill domains, the means of
of emotional intelligence ability, have correct measuring these nonverbal skills will be
and incorrect answers. The mixed model of reviewed, as will research on the impact
emotional intelligence combines emotional of each specific nonverbal skill domain on
skills and abilities with traitlike notions outcomes in human social interaction.
of personality, which are represented by Finally, the research and application poten
Goleman’s (1995, 1998) work and instru tials of the nonverbal skills and abilities
ments like the Bar-On Emotional Quotient perspective will be discussed.
05-Manusov.qxd 6/30/2006 7:24 PM Page 81
82–––◆–––Foundations
content-standard sentence (“I am going out aware than those receiving low scores (e.g.,
of the room now but I’ll be back later.”). In Ambady, Hallahan, & Rosenthal, 1995;
all cases, respondents choose the correct Hall & Carter, 1999). In addition, physi
emotional expression (“happy,” “sad,” cians receiving higher scores on the PONS
“angry,” or “fearful”). Scores are the had more satisfied patients than their col
number of correct choices on each subtest. leagues with low-sensitivity scores (DiMatteo,
There are both adult and child versions of Friedman, & Taranta, 1979).
the DANVA (Nowicki & Duke, 2001) as The PONS assesses a wide range of sen
well as a modified version of both measures sitivity to different channels of nonverbal
useful in the field (Duke & Nowicki, 2005). communication: audio versus visual, the
The DANVA assesses sensitivity to both relatively “rich” nonverbal cues provided
visual and auditory cues and increases the by facial expressions versus the more lim
range of nonverbal expressions by including ited range of cues emitted by the body
the ability to decode both posed and spon channel. It also involves nonverbal cues dis
taneous expressions of emotions. Research played in a number of enacted scenarios,
shows that scores on the DANVA correlate such as “helping a customer” or “talking
with greater overall social competence and about one’s divorce,” sometimes combining
better psychosocial adjustment (Nowicki & the nonverbal display of emotion with cues
Duke, 2001). of dominance-submission (e.g., “talking
Perhaps the most well-known measure of to a lost child” or “asking forgiveness”).
nonverbal decoding skill is the aforemen Other items involve reading cues that
tioned PONS (Rosenthal et al., 1979). The represent complex blendings of emotions
PONS consists of 220 brief, two-second and circumstances (e.g., “returning a faulty
audio and video clips of a woman enacting item to a store,” “talking about one’s wed
multiple, emotionally laden scenes (e.g., ding”), whereas some are seemingly devoid
expressing jealous rage, asking for a favor, of emotional content (e.g., “ordering food
talking about the death of a friend). By in a restaurant”). The breadth of the
masking the spoken words via electronic PONS’s sampling of enacted scenes and the
content filtering or randomized splicing of multiple channels of nonverbal cue displays
the audio track, only nonverbal cues are pre seem to capture the diversity of nonverbal
sented. Eleven different channels (e.g., face behavior, but this has led to low internal
only, body only, audio only, and all possible consistency (see Hall, 2001) and can be
combinations) are used to assess specific time-consuming to implement. Other limi
aspects of nonverbal decoding. After each tations of the PONS are its reliance on a
clip, respondents are presented with a multi- lone sender and the fact that the scenarios
ple-choice item and instructed to choose the are posed rather than genuine enactments
correct portrayal. A total accuracy score rep of emotions.
resents general sensitivity to nonverbal cues The Interpersonal Perception Task (IPT;
(Hall, 2001). There are also brief versions of Archer & Costanzo, 1988; Costanzo &
the PONS that use only the audio or visual Archer, 1989) is another measure of non
cues. The full-length PONS has been used verbal decoding skill but one that focuses
widely in research and has demonstrated more on reading nonverbal, verbal, and sit
good psychometric properties (Hall, 2001; uational cues to assess the ability to decode
Rosenthal, 1979). Not surprisingly, research the relationships among video interactants,
with the PONS suggests that persons receiv their status or dominance in relationship to
ing high scores are more interpersonally one another, their level of intimacy in the
05-Manusov.qxd 6/30/2006 7:24 PM Page 83
relationship, and whether they are truth communicator, the IPT has many segments
telling or deceiving. Scenes include trying to with two or more interactants, requiring a
determine a child’s parent (relationship), more sophisticated level of decoding skill,
identifying the person of higher status or the such as the ability to notice inconsistencies
winner of a sporting event (status or domi between verbal and nonverbal channels or to
nance), or trying to determine the length of use cues of dyadic rapport (Bernieri & Gillis,
a couple’s dating relationship (intimacy). 2001; Bernieri & Rosenthal, 1991). It is also
The IPT consists of 30 brief video (with full likely that correct decoding of IPT scenes
audio) scenes of one, two, or more stimulus requires not just recognition of specific cues,
persons, communicating with an inter as in decoding facial expressions of emotion,
viewer, communicating with another person but also sophisticated interpretation of
on the telephone, or in face-to-face commu complex verbal, nonverbal, and social cues.
nication. After watching the scene, respon For example, one study found that whereas
dents choose the correct interpretation via Japanese and American participants scored
a multiple-choice question. There is also a about the same in the visual-only condition
brief version of the IPT (IPT-15, Costanzo of the IPT-15, Japanese participants’ accu
& Archer, 1993) that removed some of the racy in relation to Americans dropped off
problematic items from the longer version considerably when the vocal and verbal cues
(e.g., items that were decoded at about were added (Iizuka, Patterson, & Matchen,
chance or less than chance accuracy). 2002). Furthermore, respondents who rely
The IPT goes beyond nonverbal decod on “common sense” heuristics tend to make
ing and presents individuals being assessed systematic errors on the IPT, such as assum
with both the audio and the video interac ing that the older interactant (or the male
tion between or among participants. Most interactant) has the higher status (Archer &
other measures of nonverbal decoding skill Costanzo, 1988). An individual who has
eliminate the verbal content to focus solely great skill in nonverbal and situational
on nonverbal (visual and paralinguistic) decoding should be more likely to avoid
cues. The authors of the IPT suggest that these errors.
it is more closely aligned with the construct A limitation of all the performance
of “social intelligence” than with emotional measures of nonverbal decoding skill is the
intelligence (Archer, Costanzo, & Akert, relatively small sampling of what is obvi
2001). Research with the IPT has been ously a vast domain of nonverbal behavior.
primarily focused on psychometric issues, For instance, the BART and DANVA focus
although there is evidence that persons exclusively on emotional communication.
scoring high on the IPT are more socially Although the PONS assesses skill in decod
aware and socially competent (Archer et al., ing an array of nonverbal messages involving
2001). In a study of college roommates, cues of emotion, status, and relationship, it is
high scorers on the IPT had higher-quality limited to one sender; the IPT has multiple
and more supportive relationships with one senders, but it is brief (15 or 30 items) and
another than did low IPT scorers (Hodgins suffers from low internal consistency (Hall,
& Zuckerman, 1990). 2001). The extremely low internal consis
The IPT has the advantage of using natu tency coefficients (KR20s of .38 and .52 for
ralistic interactions as opposed to the posed the brief and long versions of the IPT) sug
communications used in most other mea gest that the IPT may indeed be assessing a
sures of nonverbal decoding skill. More broad domain of interpersonal skills. Finally,
over, rather than presenting a single performance measures of nonverbal skill are
05-Manusov.qxd 6/30/2006 7:24 PM Page 84
84–––◆–––Foundations
time-consuming to develop and often diffi particularly empathy and being other-
cult to administer. oriented (Losoya & Eisenberg, 2001).
Another approach to measuring nonver Nonverbally sensitive individuals may also
bal skill involves the use of self-report be more prone to emotional contagion
methods. Self-report methods offer the effects, vicariously experiencing others’
advantages of sampling across a broad emotions (Hatfield, Cacioppo, & Rapson,
range of nonverbal skill-related areas, and 1994). Indeed, there is a .48 correlation
they are relatively easy to administer (Riggio between the emotional sensitivity scale
& Riggio, 2001). The earliest published of the SSI and a self-report measure of emo
attempt to assess nonverbal decoding skill tional contagion (Riggio & Carney, 2003).
via self-reports was by Zuckerman and This suggests that although nonverbal sen
Larrance (1979) with their Perceived sitivity is considered to be an important
Decoding Ability (PDA) and Perceived component of global nonverbal skill or
Encoding Ability (PEA) tests. As part of a competence, it is possible to be too nonver
larger “social skill” model, Riggio’s Social bally sensitive. Davis (1983, 1994), in his
Skills Inventory (SSI; Riggio, 1986, 1989, multidimensional model of empathy, con
2005; Riggio & Carney, 2003) uses self- ceptualizes one aspect of other-oriented
report techniques to measure two decoding sensitivity as taking another’s perspective or
abilities, emotional sensitivity (the ability to showing some empathic concern with
decode emotions and other nonverbal cues) others’ emotional states. A different form
and social sensitivity (the ability to decode of empathy, however, is labeled “personal
and understand social situations, social distress,” suggesting emotional contagion
roles, and social scripts). Unfortunately, effects. Taken to an extreme, nonverbal
measuring nonverbal decoding skill via self- sensitivity may cause people to experience
reports has demonstrated only limited suc some personal distress. It is important to
cess. For example, correlations between emphasize that when considering global
self-report measures of decoding skill and skill or competence in nonverbal and emo
performance measures have been positive tional communication, an individual needs
but low (typically below .20). The emo to also possess skills in decoding, encoding,
tional sensitivity subscale of the SSI, how and regulating communication and emo
ever, has had a slightly stronger relationship tional processes (see Riggio, 1986; Riggio
with decoding scores on the DANVA & Carney, 2003).
(decoding faces) and a similar decoding
facial expressions subtest of the Multifactor
Emotional Intelligence Test (correlations of DETECTION OF DECEPTION
.22 and .26, respectively; Riggio & Carney,
2003). In addition, self-reported emotional The ability to detect deception accu
sensitivity is correlated substantially with rately is a particular nonverbal decoding
relevant self-reported social behaviors, skill. Because of its complexity, the ability
including the size and closeness of social to detect deception is rare, with a very small
networks (Riggio, 1986; Riggio, Watring, percentage of the general population able
& Throckmorton, 1993). to detect deception much above chance
Skill in decoding nonverbal communica levels (Ekman, O’Sullivan, & Frank, 1999;
tion is considered a very important compo Malone & DePaulo, 2001; Vrij, this vol
nent of nonverbal ability. It is aligned closely ume). Yet some individuals seem to be par
with important personality characteristics, ticularly successful at detecting deception
05-Manusov.qxd 6/30/2006 7:24 PM Page 85
and are labeled “wizards” (O’Sullivan, detection, are correlated, although the
2005). These rare individuals are particu magnitude of relationships is modest. For
larly skilled at reading nonverbal cues, example, the ability to detect one kind of
although they also possess a number of lie is related to detecting the accuracy
other skills and qualities critical for detect of another form of lie (Frank & Ekman,
ing deception. These include the ability to 1997). In addition, the ability to detect
notice inconsistencies in and between non deception is slightly positively related to
verbal and verbal cues, a strong motivation the ability to decode nonverbal cues of
to both observe and analyze human behav emotion (Buller & Burgoon, 1994; Malone
ior, and an ability to avoid the systematic & DePaulo, 2001). Likewise, a review of
cognitive biases that hamper the ability to intercorrelations among standardized per
detect deception successfully in the general formance measures of decoding skill (e.g.,
population (O’Sullivan, 2005). PONS, IPT, CARAT) shows that there are
A great deal of research suggests why positive relationships among the tests, but
decoding of deception is such a difficult they are quite modest (Hall, 2001). This
skill. First, there is a trusting bias: a ten result suggests, perhaps, that nonverbal
dency to believe that others’ communica decoding skill is complex and multifaceted,
tions are more likely to be truthful than consisting of multiple abilities that are
deceptive (Ekman, 1985; Riggio, 1992b). somewhat related to, but considerably inde
This manifests itself in proportionately more pendent from, one another.
judgments of “truth” as opposed to “lie” in
most experimental investigations of decep
tion detection. Second, people may not be ♦ Skill in Nonverbal Encoding
able to hone their deception detection skills
if they receive inadequate feedback concern
ing whether someone was lying or telling the Nonverbal encoding skill, also referred to as
truth (i.e., in everyday life, we may never nonverbal expressiveness, involves the abil
find out for sure if a friend or relation was ity to send nonverbal messages to others
lying to us or not). People may also rely too accurately. Typically, performance mea
much on stereotypical cues of deception, sures of individual differences in nonverbal
such as an inability to maintain eye contact, encoding consist of videotaping participants
nervous fidgeting, and slow, staccato speech while they are sending emotional expres
patterns, cues that may be unrelated to sions spontaneously or while posing them
actual deception (DePaulo et al., 2003). on cue. For example, Buck’s (2005) slide-
Furthermore, because deceivers are also viewing technique was reviewed earlier as a
aware of these stereotypical deception cues, method of assessing nonverbal decoding
they may take steps to ensure that they mon skill. But the slide-viewing technique can
itor and control these “known” clues to also be used to measure individual differ
deception. For example, in one study, par ences in the spontaneous encoding of
ticipants engaged in significantly greater emotions: A measure of spontaneous encod
eye contact when deceiving as compared ing ability consists of the percentage of
with truth telling, presumably in an effort judges who can identify the emotion being
to look more honest (Riggio & Friedman, portrayed correctly via facial expressions
1983). while the sender is viewing or discussing the
There is some evidence that nonverbal emotion-evoking slide. More frequently,
decoding abilities, including deception however, participants are asked to pose an
05-Manusov.qxd 6/30/2006 7:24 PM Page 86
86–––◆–––Foundations
implicated in the ability to deceive success affect, such as the spontaneous expression
fully, but the relationship is complex. Non- of anger or dislike, can have equally nega
verbally expressive individuals, as measured tive outcomes (see Burgoon & Bacue,
by a posed emotional encoding task, were 2003). Therefore, it is important to con
somewhat more successful at deceiving sider the monitoring and regulation of
others, but much of their success was expressive displays when discussing compe
related to a “demeanor bias.” That is, they tence in nonverbal skills and abilities.
simply looked more honest and truthful
than nonexpressive persons regardless of ♦ Skill in Nonverbal
whether they were truth telling or lying.
Regulation and Control
Presumably, this was due to their tendency
to be animated and expressive and to avoid
“nervous” cues that are stereotypically asso Skill in nonverbal communication involves
ciated with deception (Riggio & Friedman, more than just encoding and decoding abil
1983). ities. The ability to regulate and control
Nonverbal or emotional expressiveness one’s nonverbal communication is also
has often been either equated with the per a key component of what it means to
sonality trait of extraversion or considered be “nonverbally skilled” (Riggio, 1986).
a component of it. Consistent evidence sug Specifically, the ability to regulate both the
gests, however, that although expressive experience of emotions and the expression
ness and extraversion are related, they of those emotions has received a great deal
are different constructs (Friedman, 1983; of attention (e.g., Eisenberg, Champion, &
Riggio & Riggio, 2002). The fact that non Ma, 2004; Gross, 1998). People learn to
verbal expressive ability is most often mea suppress the expression of emotions, which
sured via self-report instruments (with is likely responsible for the consistent
shared method variance inflating the corre differences in emotional expressiveness
lations between measures of expressiveness due to culture (Ekman & Friesen, 1969;
and measures of extraversion) further Matsumoto, 2001, this volume; Matsumoto
drives the misconception that expressive & Yoo, 2005; but see Fridlund, 1994, this
ness is “just extraversion.” volume, for an alternative view) and gender
The ability to convey nonverbal mes (Hall, Carter, & Horgan, 2000; Riggio &
sages to others, particularly the sending of Carney, 2003). Regulating or managing
emotional messages, is a critical skill for emotions is also one of the core elements in
social success, and a fundamental compo the abilities model of emotional intelligence
nent of the larger construct of communica (Mayer et al., 2000). Gross (2001) suggests
tion competence (see Greene & Burleson, that regulation of emotion can involve
2003). Performance-based measurement of two processes: reappraisal and suppression.
nonverbal encoding skill is a difficult task; Reappraisal involves altering both the expe
therefore, it has not received as much atten rience and the expression of an emotion,
tion as nonverbal decoding ability. Instead, whereas suppression involves the inhibition
researchers have relied on self-report mea of emotionally expressive behavior. In
sures of nonverbal expressiveness. Whereas either case, the nonverbal communication
evidence suggests that the ability to express of emotion can be controlled and can thus
positive emotions creates more favorable be considered an “ability.”
impressions and can positively influence Snyder’s (1974, 1987) research on self-
others’ moods, the expression of negative monitoring represents some of the earliest
05-Manusov.qxd 6/30/2006 7:24 PM Page 88
88–––◆–––Foundations
work attempting to measure control over Individuals scoring high on EC, for example,
nonverbal communication. By definition, are able to stifle the expression of felt emo
self-monitoring is, in part, the ability to tions or cover the display of the felt emo
monitor, regulate, and control one’s non tion with another emotional “mask.”
verbal displays. Yet self-monitoring also Tucker and Riggio (1988) found that indi
involves being attentive to others’ nonverbal viduals scoring high on EC were more diffi
cues and an ability to adjust one’s own cult to decode in a spontaneous emotional
expressive behavior to try to “fit in” to sending using the slide-viewing technique,
social situations. It has been argued, based suggesting that they monitor and stifle their
on factor analyses of Snyder’s (1974) self- emotional expressions. Research also sug
monitoring scale, that the SMS also mea gests that persons with high levels of EC
sures elements of nonverbal encoding and have greater comfort speaking before large
decoding skill (Riggio & Friedman, 1982). audiences, performed better in a discussion-
Indeed, a key validation study conducted oriented problem-solving group, and had
by Snyder (1974) involved correlating more formal acting experience than persons
SMS scores with posed emotional encoding lacking EC (Riggio, 1986; Riggio, Riggio,
and decoding tasks. Positive relationships Salinas, & Cole, 2003).
between SMS and emotional encoding and Another measure of the ability to control
decoding success suggest that self-monitoring nonverbal and emotional communication
could represent a measure of global social is the Emotion Regulation Questionnaire
skill or competence. Factor analyses of the (ERQ), a 10-item self-report instrument with
original SMS demonstrated separate factors subscales measuring the ability to regulate
that measured other-directedness, a ten emotions via reappraisal or suppression
dency to focus on others’ behavior, likely (Gross & John, 2003). Research using the
a prerequisite for nonverbal decoding ERQ suggests that there are psychologically
skill, and extraversion-acting, a factor that “healthy” and “unhealthy” means to regu
involves not only control or communication late and control the expression of emotion.
but also a form of social acting skill that For example, reappraisal-based EC or regu
would be related to posed nonverbal encod lation leads to more positive outcomes (e.g.,
ing skill (Briggs & Cheek, 1988; Briggs, positive emotions, well-being), whereas con
Cheek, & Buss, 1980). These results suggest trolling the expression of emotion via sup
that global nonverbal and social skills or pression leads to more negative outcomes.
competence is composed of encoding, Riggio and Zimmerman (1991) found that
decoding, and regulatory skills, inspiring a persons scoring high on SSI-EC reported
more general, multidimensional model of using social support strategies less to cope
nonverbal and social skills (Riggio, 1986, with stress, whereas persons scoring high on
1989; Riggio & Carney, 2003). emotional expressiveness and sensitivity
In this multidimensional nonverbal and reported having larger supportive social net
social skill model, emotional control (EC) is works and making greater use of social sup
one of three core elements of nonverbal skill port strategies to cope with stress.
that involves the ability to control and regu The ability to control strong emotions
late emotional and nonverbal displays. The can be an asset in formal social settings,
other two elements in this model, discussed such as in public speaking, in the work
earlier, are emotional or nonverbal encod place, and in positions of authority when
ing (labeled emotional expressivity) and under crisis or stress, where an individual
decoding (labeled emotional sensitivity). would not want to arouse others’ anxiety
05-Manusov.qxd 6/30/2006 7:24 PM Page 89
90–––◆–––Foundations
Ambady, N., LaPlante, D., & Johnson, E. Bonanno, G. A., Papa, A., Lalande, K., Wesphal,
(2001). Thin-slice judgments as a measure M., & Coifman, K. (2004). The importance
of interpersonal sensitivity. In J. A. Hall & of being flexible: The ability to both enhance
F. J. Bernieri (Eds.), Interpersonal sensitiv and suppress emotional expression predicts
ity: Theory and measurement (pp. 89–101). long-term adjustment. Psychological Science,
Mahwah, NJ: Erlbaum. 15, 482–487.
Ambady, N., & Rosenthal, R. (1992). Thin slices Briggs, S. R., & Cheek, J. M. (1988). On the
of expressive behavior as predictors of inter nature of self-monitoring: Problems with
personal consequences: A meta-analysis. assessment, problems with validity. Journal
Psychological Bulletin, 111, 256–274. of Personality & Social Psychology, 54,
Archer, D., & Costanzo, M. (1988). The 663–678.
Interpersonal Perception Task. Berkeley, Briggs, S. R., Cheek, J. M., & Buss, A. H.
CA: University of California Extension (1980). An analysis of the self-monitoring
Media Center. scale. Journal of Personality & Social
Archer, D., Costanzo, M., & Akert, R. (2001). Psychology, 38, 679–686.
The Interpersonal Perception Task (IPT): Brunswik, E. (1956). Perception and the repre
Alternative approaches to problems of sentative design of psychological experi
theory and design. In J. A. Hall & F. J. ments (2nd ed.). Berkeley, CA: University of
Bernieri (Eds.), Interpersonal sensitivity: California Press.
Theory and measurement (pp. 161–182). Buck, R. (1976). A test of nonverbal receiving
Mahwah, NJ: Erlbaum. ability: Preliminary studies. Human Commu
Argyle, M. (1999). Nonverbal vocalizations. nication Research, 2, 162–171.
In L. K. Guerrero, J. A. DeVito, & M. L. Buck, R. (1978). The slide-viewing technique for
Hecht (Eds.). The nonverbal communica measuring nonverbal sending accuracy: A
tion reader: Classic and contemporary read guide for replication. Catalog of Selected
ings (pp. 135–148). Prospect Heights, IL: Documents in Psychology, 8, 62.
Waveland Press. Buck, R. (2005). Measuring emotional experi
Bar-On, R. (1997). Bar-On Emotional Quotient ence, expression, and communication: The
Inventory: Technical manual. Toronto, slide-viewing technique. In V. Manusov
Canada: Multi-Health Systems. (Ed.), The sourcebook of nonverbal mea
Bernieri, F. J. (2001). Toward a taxonomy of sures: Going beyond words (pp. 457–470).
interpersonal sensitivity. In J. A. Hall & Mahwah, NJ: Erlbaum.
F. J. Bernieri (Eds.), Interpersonal sensitiv Buck, R., Miller, R. E., & Caul, W. F. (1974).
ity: Theory and measurement (pp. 3–20). Sex, personality and physiological variables
Mahwah, NJ: Erlbaum. in the communication of emotion via facial
Bernieri, F. J., & Gillis, J. S. (2001). Judging rap expression. Journal of Personality and
port: Employing Brunswik’s lens model to Social Psychology, 30, 587–596.
study interpersonal sensitivity. In J. A. Hall Buck, R., Savin, V. J., Miller, R. E., & Caul, W.
& F. J. Bernieri (Eds.), Interpersonal sensi F. (1972). Nonverbal communication of
tivity: Theory and measurement (pp. 67– affect in humans. Journal of Personality
88). Mahwah, NJ: Erlbaum. and Social Psychology, 23, 362–371.
Bernieri, F. J, & Rosenthal, R. (1991). Coordinated Buller, D. B., & Burgoon, J. K. (1994).
movement in human interaction. In R. S. Deception: Strategic and nonstrategic com
Feldman & B. Rime (Eds.), Fundamentals of munication. In J. A. Daly & J. M. Wiemann
nonverbal behavior (pp. 401–432). New (Eds.), Strategic interpersonal communi
York: Cambridge University Press. cation (pp. 191–224). Hillsdale, NJ:
Birdwhistell, R. L. (1970). Kinesics and context: Erlbaum.
Essays on body motion communication. Burgoon, J. K., & Bacue, A. E. (2003). Nonverbal
Philadelphia: University of Pennsylvania communication skills. In J. O. Greene &
Press. B. R. Burleson (Eds.), Handbook of
05-Manusov.qxd 6/30/2006 7:24 PM Page 92
92–––◆–––Foundations
The Affective Communication Test. Journal Hall, J. A., Carter, J. D., & Horgan, T. G.
of Personality and Social Psychology, 39, (2000). Gender differences in non
333–351. verbal communication of emotion. In
Friedman, H. S., & Riggio, R. E. (1981). Effect A. H. Fischer (Ed.), Gender and emotion:
of individual differences in nonverbal Social psychological perspectives (pp. 97–
expressiveness on transmission of emotion. 117). New York: Cambridge University
Journal of Nonverbal Behavior, 6, 96–104. Press.
Friedman, H. S., & Riggio, R. E. (1999). Harris, M. J., & Rosenthal, R. (2005). No more
Individual differences in ability to encode teachers’ dirty looks: Effects of teacher
complex affects. Personality and Individual nonverbal behavior on student outcomes.
Differences, 27, 181–194. In R. E. Riggio & R. S. Feldman (Eds.),
Friedman, H. S., Riggio, R. E., & Casella, D. F. Applications of nonverbal communication
(1988). Nonverbal skill, personal charisma, (pp. 157–192). Mahwah, NJ: Erlbaum.
and initial attraction. Personality and Social Hatfield, E., Cacioppo, J. T., & Rapson, R.
Psychology Bulletin, 14, 203–211. (1994). Emotional contagion. Cambridge,
Goleman, D. (1995). Emotional intelligence: England: Cambridge University Press.
Why it can matter more than IQ. New Henley, N. (1977). Body politics: Power, sex,
York: Bantam. and nonverbal communication. Englewood
Goleman, D. (1998). Working with emotional Cliffs, NJ: Prentice-Hall.
intelligence. New York: Bantam. Hodgins, H., & Zuckerman, M. (1990). The
Greene, J. O., & Burleson, B. R. (Eds.) (2003). effect of nonverbal sensitivity on social
Handbook of communication and interaction. Journal of Nonverbal Behavior,
social interaction skills. Mahwah, NJ: 24, 155–170.
Erlbaum. Howard, A. (1997). A reassessment of assess
Gross, J. J. (1998). The emerging field of emotion ment centers: Challenges for the 21st
regulation: An integrative review. Review of century. Journal of Social Behavior and
General Psychology, 2, 271–299. Personality, 12(5), 13–52.
Gross, J. J. (2001). Emotion regulation in Iizuka, Y., Patterson, M. L., & Matchen, J. C.
adulthood: Timing is everything. Current (2002). Accuracy and confidence on the
Directions in Psychological Science, 10, Interpersonal Perception Task: A Japanese-
214–219. American comparison. Journal of Non
Gross, J. J., & John, O. P. (1995). Facets of verbal Behavior, 26, 159–174.
emotional expressivity: Three self-report Kring, A. M., Smith, D. A., & Neale, J. M.
factors and their correlates. Personality and (1994). Individual differences in disposi
Individual Differences, 19, 555–568. tional expressiveness: Development and vali
Gross, J. J., & John, O. P. (2003). Individual dation of the emotional expressivity scale.
differences in two emotion regulation Journal of Personality and Social Psychology,
processes: Implications for affect, relation 66, 934–949.
ships, and well-being. Journal of Personality Losoya, S. H., & Eisenberg, N. (2001). Affective
and Social Psychology, 85, 348–362. empathy. In J. A. Hall & F. J. Bernieri
Hall, J. A. (2001). The PONS test and the (Eds.), Interpersonal sensitivity: Theory and
psychometric approach to measuring inter measurement (pp. 21–43). Mahwah, NJ:
personal sensitivity. In J. A. Hall & Erlbaum.
F. J. Bernieri (Eds.), Interpersonal sensitiv Malone, B. E., & DePaulo, B. M. (2001).
ity: Theory and measurement (pp. 143– Measuring sensitivity to deception. In J. A.
160). Mahwah, NJ: Erlbaum. Hall & F. J. Bernieri (Eds.), Interpersonal
Hall, J. A., & Carter, J. D. (1999). Gender- sensitivity: Theory and measurement
stereotype accuracy as an individual differ (pp. 103–124). Mahwah, NJ: Erlbaum.
ence. Journal of Personality and Social Mast, M. (2002). Dominance as expressed
Psychology, 77, 350–359. and inferred through speaking time: A
05-Manusov.qxd 6/30/2006 7:24 PM Page 94
94–––◆–––Foundations
Riggio, R. E., & Friedman, H. S. (1982). The Rosenthal, R., & Jacobson, L. (1968). Pygmalion
interrelationships of self-monitoring factors, in the classroom. New York: Holt, Rinehart,
personality traits, and nonverbal social and Winston.
skills. Journal of Nonverbal Behavior, 7, Salovey, P., & Mayer, J. D. (1990). Emotional
33–45. intelligence. Imagination, Cognition, and
Riggio, R. E., & Friedman, H. S. (1983). Personality, 9, 185–211.
Individual differences and cues to deception. Snyder, M. (1974). The self-monitoring of
Journal of Personality and Social Psychology, expressive behavior. Journal of Personality
45, 899–915. and Social Psychology, 30, 526–537.
Riggio, R. E., & Friedman, H. S. (1986). Snyder, M. (1987). Public appearances/private
Impression formation: The role of expres realities: The psychology of self-monitoring.
sive behavior. Journal of Personality and New York: W.H. Freeman.
Social Psychology, 50, 421–427. Taylor, S. J. (2002). Effects of a nonverbal skill
Riggio, R. E., & Riggio, H. R. (2001). Self- training program on perceptions of
report measurement of interpersonal sensi personal charisma. Unpublished doctoral
tivity. In J. A. Hall & F. J. Bernieri (Eds.), dissertation, University of California at
Interpersonal sensitivity: Theory and mea Riverside.
surement (pp. 127–142). Mahwah, NJ: Thornton, G. C. (1992). Assessment centers in
Erlbaum. human resource management. Reading,
Riggio, R. E., & Riggio, H. R. (2005). Self-report MA: Addison-Wesley.
measures of emotional and nonverbal Tucker, J. S., & Riggio, R. E. (1988). The role
expressiveness. In V. Manusov (Ed.), The of social skills in encoding posed and spon
sourcebook of nonverbal measures: Going taneous facial expressions. Journal of
beyond words (pp. 105–111). Mahwah, NJ: Nonverbal Behavior, 12, 87–97.
Erlbaum. Zaidel, S. F., & Mehrabian, A. (1969). The abil
Riggio, R. E., Riggio, H. R., Salinas, C., & Cole, ity to communicate and infer positive and
E. J. (2003). The role of social and emo negative attitudes facially and vocally.
tional communication skills in leader emer Journal of Experimental Research in
gence and effectiveness. Group Dynamics: Personality, 3, 233–241.
Theory, Research, and Practice, 7, 83–103. Zuckerman, M., Koestner, R., & Alton, A. O.
Riggio, R. E., Watring, K., & Throckmorton, B. (1984). Learning to detect deception. Journal
(1993). Social skills, social support, and of Personality and Social Psychology, 46,
psychosocial adjustment. Personality and 519–528.
Individual Differences, 15, 275–280. Zuckerman, M., Koestner, R., & Colella, M. J.
Riggio, R. E., & Zimmerman, J. A. (1991). Social (1985). Learning to detect deception from
skills and interpersonal relationships: Influ three communication channels. Journal of
ences on social support and support seeking. Nonverbal Behavior, 9, 188–194.
In W. H. Jones & D. Perlman (Eds.), Zuckerman, M., & Larrance, D. (1979). Individ
Advances in personal relationships (Vol. 2, ual differences in perceived encoding and
pp. 133–155). London: Jessica Kingsley decoding abilities. In R. Rosenthal (Ed.), Skill
Press. in nonverbal communication (pp. 171–203).
Rosenthal, R. (Ed.). (1979). Skill in nonverbal Cambridge, MA: Oelgeschlager, Gunn, &
communication. Cambridge, MA: Oelge Hain.
schlager, Gunn, & Hain. Zuckerman, M., Lipets, M. S., Koivumaki, J. A.,
Rosenthal, R., Hall, J. A., DiMatteo, M. R., & Rosenthal, R. (1975). Encoding and
Rogers, P. L., & Archer, D. (1979). decoding nonverbal cues of emotion. Journal
Sensitivity to nonverbal communications: of Personality and Social Psychology, 32,
The PONS test. Baltimore, MD: Johns 1068–1076.
Hopkins University Press.
05-Manusov.qxd 6/30/2006 7:24 PM Page 96
06-Manusov.qxd 7/6/2006 4:53 PM Page 97
6
NONVERBAL AND VERBAL
COMMUNICATION
Hand Gestures and Facial Displays as Part
of Language Use in Face-to-Face Dialogue1
� Nicole Chovil
Independent Researcher and Education Consultant
◆ 97
06-Manusov.qxd 6/30/2006 7:26 PM Page 98
98–––◆–––Foundations
100–––◆–––Foundations
does not necessarily depend on the micro- polysemous, yet each meaning is ordinarily
context in which they occur. unambiguous when it occurs in a particular
The context that determines the mean conversational context (i.e., from a wide
ing of both audible and visible acts of turn to wide awake to nationwide).
meaning is multilayered and includes who Similarly, in the following gestural
the participants are, why they are talking, example, the speaker is describing a picture
how they have been using the word or act of a dress with an unusually wide hipline,
so far, the particular topic in that phase of which extends about a meter on either side of
the conversation, the precise point in the the waist (cf. Bavelas, Kenwood, Johnson, &
utterance, and the simultaneous other ele Phillips, 2002, Figure 1). The underlining
ments of the integrated message of which indicates where gestures occurred in relation
it is a part at that moment. As an example to the words; the brackets contain an itali
of the importance of all of these layers cized description of each gesture; and S =
of context for the meaning of words, speaker, A = addressee. (For readers who
the adjective wide has a couple dozen do not usually watch conversational actions
meanings in the Random House Unabridged frame by frame, the best way to understand a
Dictionary (1993). Like most words, it is written example is to act it out oneself.)
Example 1.
S: “OK. Ah, like a huge skirt that goes out like this?” ____________
[both hands move from waist to full out] [holds width]
A: “Like one of the round ones?” ____________
[hands curve out from waist] [holds width]
By moving her hands out and especially is usually inseparable from its linguistic and
by holding them in place at the farthest micro-social contexts.
extent, the speaker indicated, among other Contextual specificity does not apply
things, that the skirt was “wide.” The only to hand gestures. In the first systematic
addressee confirmed his understanding by description of a conversational facial dis
replicating both movements. play, Ekman (1979) showed that the same
In different conversational moments, physical eyebrow actions can have several
her gesture for a wide skirt could have indi different meanings (e.g., as a baton or a
cated the length of a fish that the speaker question mark), depending on conversa
caught, the metaphorical amount of work tional context. Chovil (1989) found that
the speaker has left to do, or the begin even stereotypic expressions can have varied
ing of a tree-hugging gesture. Yet at the meanings. The classic nose-wrinkle of dis
moment it occurred in its particular conver gust can also convey rejection of other kinds,
sational context—even though there was no which have nothing to do with smell (e.g.,
reference to “wide” in the words of either a disliked movie, an unpleasant chore), and
person—it was unambiguous. As Goodwin an angry expression may not indicate con
(2000) illustrated through a detailed micro comitant anger. In the following example
analysis of a hand gesture, simply “locating from Chovil’s data, the speaker was humor
the lexical affiliate of a gesture does not ously describing a past argument with her
constitute establishing its meaning” (p. 92) sister about whether she should cut her
because the meaning of any word or gesture hair.
06-Manusov.qxd 6/30/2006 7:26 PM Page 102
102–––◆–––Foundations
Example 2.
Exactly as she was describing her own and Gerrig (1990), and Streeck and Knapp
part in the argument, she configured her eyes, (1992) have also noted that hand gestures
brows, and mouth in a classic anger display. can depict, demonstrate, or reenact. There is
As soon as the relevant phrase was over, she a difference between a hand action, which
smiled along with the addressee at her own has a practical function in the material world
humor. She was not angry when she made (e.g., turning on a light switch or holding a
the display; indeed, she may not have been telephone) and a hand gesture, which has a
nearly as angry at the time of the argument communicative function in the social world
as the current display indicated. She was (e.g., as part of telling someone to switch on
exaggerating for effect, as confirmed by her the light or that you will call them later).
immediate smile, which the addressee shared. Practical and material considerations shape
We propose that, just as with words, the the hand action, but social and communica
addressees seldom have difficulty selecting tive considerations shape the hand gesture.
the correct meaning, largely because of the Because of these considerations, the hand
contextual redundancy that supports it. action and gesture should look different
in predictable ways (Gerwing & Bavelas,
2004). Very few studies have even recorded
♦ II. Symbolic Acts the difference between hand actions and
hand gestures; an exception is LeBaron and
In proposing that a hand gesture or facial Streeck’s (2000) comparison of instrumental
display in dialogue is a symbol, we intend actions to later gestures for the same actions.
the simplest sense of the term: Symbols have Several recent experiments have demon
referents; they are something that stands for strated a key part of our proposal, namely,
something else (Quine, 1987, p. 763). Put if social and communicative factors shape
in other terms, symbols are encoded acts, hand gestures, then these factors should
although the encoding is ordinarily ana cause variation in gestures for the same ref
logic or iconic (see Bavelas & Chovil, erent. Ozyurek (2000, 2002) showed that
2000). The case for the proposition that speakers made a gesture depicting the same
some nonverbal acts are symbolic involves motion differently depending on their spatial
somewhat different issues for hand gestures relationship to their addressees. Other exper
and for facial displays, which will be iments have also demonstrated that the ref
treated separately in the following. erent is not the sole determinant of the form
of a gesture. Rather, linguistic principles
HAND GESTURES AS SYMBOLS unique to dialogue can influence the shape
of gestures: When the participants shared
McNeill (1992) pointed out that “ges common ground about an object, they made
tures are not just movements and can never sketchier gestures to depict it than when
be fully explained in purely kinesic terms. the information was new to one of them
They are not just the arms waving in the air (Gerwing & Bavelas, 2004). Similarly,
but symbols that exhibit meaning in their within a dialogue, later gestures for familiar
own right” (p. 105). Kendon (1985), Clark (“given”) information were shorter than
06-Manusov.qxd 6/30/2006 7:26 PM Page 103
those for new information (Woods, 2005). Because of the vastly greater research
These effects are identical to the effects of interest in emotional expression, there is
given-versus-new information on verbal remarkably little scholarship on the use of
communication, such as the length of verbal the face for communication (for a more
reference (e.g., Fowler, 1988). The results general discussion of emotional communi
also accord with Grice’s (1967/1989) max cation, however, see Fridlund & Russell,
ims of manner and quantity: The sketchier this volume). We know of three systematic
or shorter gestures were sufficient for their descriptions: Ekman’s (1979) above-
purpose but no more than that. mentioned description of eyebrow move
ments, Brunner’s (1979) analysis of smiles
FACIAL DISPLAYS AS SYMBOLIC as back-channels, and Chovil’s systematic
identification of conversational facial dis
The distinction between symbol and plays other than smiles (1989, 1991/1992).
referent is even more subtle and important The latter study documented the wide
for faces, because there is a strong tendency variety of syntactic and semantic functions
to equate a facial expression with an emo of participants’ facial displays in sponta
tional expression. In this view, facial neous face-to-face dialogue. For example,
expressions of emotion are nonsymbolic, speakers facially portrayed themselves as
involuntary acts that reveal information they might have appeared at another time,
about the individual’s intrapsychic state. in another situation (see Example 2, earlier
Ekman, the pioneer in the study of face and in this chapter); they also portrayed others’
emotion, anticipated other functions of the reactions (e.g., a disapproving relative), and
face in his early work, however (e.g., they marked syntactic emphasis, questions,
Ekman, 1979). More recently, he also made and other narrative features, usually with
several distinctions between facial expres eyebrow movements.
sions of emotion and facial actions that are One limitation of the research just dis
conversational signals: cussed is that it has been almost entirely
descriptive, documenting the nonemotional
Most importantly, the conversational sig role of facial displays in face-to-face dia
nals [italics added] are part of the struc logue but not offering an alternative theo
ture of the conversation, part of the flow retical conception of them. We (Bavelas &
of talk, and governed by the rules which Chovil, 1997) found a promising theory in
govern the production of speech. While Clark’s (1996; Clark & Gerrig, 1990) con
facial expressions of emotion [italics cept of demonstration as a distinct method
added] often occur during conversation, of signaling (adapted from Peirce, cited in
their location in the speech flow is related Buchler, 1940). Clark and Gerrig (1990)
not to the structure of talk but to the proposed that many conversational actions,
semantics, revealing an emotional reac such as quoting what someone else said,
tion to what is being said or not being are demonstrations rather than descriptions
said. (Ekman, 1997, p. 340) or indications (the other two methods of
signaling). In addition, “people can demon
Thus, Ekman’s first criterion, synchrony strate a cough, the rhythm of a part of a
with speech, is the same as ours. Kraut and Chopin prelude, the sound of a car engine,
Johnston (1979) also proposed a distinction . . . or the appearance of a chimpanzee”
between the facial expression of emotion (Clark & Gerrig, 1990, pp. 766–767).
and a socially oriented facial display (which The speaker need not actually have a cough
is the term we use in this chapter). or be playing a Chopin prelude (and is
06-Manusov.qxd 6/30/2006 7:26 PM Page 104
104–––◆–––Foundations
certainly not a car engine or a chimpanzee), synchronized with words in both timing and
nor is he or she making such a claim, meaning. In this section, we propose that
because a demonstration is not literal (it is these two synchronies of timing and meaning
“non-serious”; Goffman, 1974). The prin work together to produce an integrated but
ciple of demonstration means that speakers often complex whole. Most of the available
do not necessarily or even usually use their research relevant to this proposal has focused
facial displays to portray the way they feel on gestures; at present, we have to rely on
at that moment; rather, the speaker is illus anecdotal observation for facial displays.
trating some aspect of the conversational What may be the best evidence of the pre
topic of the moment. cise integration of audible and visible acts is
When demonstrating, the speaker does easy to demonstrate: Speakers usually coor
not simply reproduce the literal expression; dinate their hand gestures and facial displays
a demonstration is selective, deleting irrele to verbal syntax. McNeill (1985) found that
vant features and retaining or even exagger “gestures synchronize with parallel linguistic
ating the relevant ones. For example, units [and] almost never cross clause bound
squinting one’s eyes to indicate skepticism aries” (pp. 160–161). Ekman (1997) made
or disbelief may demonstrate looking more the same general point about facial displays:
closely at something, but it is likely to be a
highly stylized and different in form (e.g., Take for example, a person who says he
quicker) than literally squinting to read fine had been afraid of what he would learn
print. In support of this, Gilbert, Fridlund, from a biopsy report, and was so relieved
and Sabini (1987) showed that individuals when it turned out to be negative. When
who were demonstrating facial displays to the word “afraid” is said, the person
various odors produced facial configura stretches back his lips horizontally, refer
tions that were clearer to observers than ring facially to fear. (p. 340)
when they were actually smelling the odor Ekman went on to point out that the
and having the same reaction sponta above facial action, used to “refer to a fear
neously. Arguably, these results illustrate not felt now,” would not only be a trans
the selective nature of demonstration. We formed version of the emotional expression
proposed at the beginning of this section of fear but “would be likely to be made very
that if conversational hand gestures and quickly, much more quickly than the actual
facial displays are symbols, then social expression of emotion would be” (p. 340).
and communicative considerations would Presumably, the display would be quicker
shape them. What we know about the prin in order to synchronize with the word
ciples of this selective process is encourag “afraid.” One important facility of the facial
ing but far too little; the determinants of the muscles is that they can track the speed of
form of symbolic nonverbal acts are an words or phrases. In Example 2, presented
important area for further research. earlier in this chapter, the speaker’s face
changed rapidly from an angry expression
♦ III. Integrated (but Not to a smile exactly when her phrase ended.
Necessarily Redundant) We can illustrate the precise integration
of all three elements (hand gestures, facial
With Words
displays, and words) with a brief example
from our data (Bavelas, 2000). The speaker
One of our defining criteria for nonverbal was telling the addressee about a close call
acts that are part of language use in face-to he once had, when he fell into a river and
face dialogue was that they must be tightly nearly drowned:
06-Manusov.qxd 6/30/2006 7:26 PM Page 105
Example 3.
S: “So, my-my-my head is in the water like this,
[head back, eyes shut, impassive face]
and basically it’s, water’s going over my head.
[head to vertical] [hands sweep beside head]
And it’s str–I grew really, really calm.”
[puzzled ] [serious face, looking at A]
During each underlined phrase, the puzzling. Each phrase of this example
speaker depicted some aspect of his dilemma demonstrates precise coordination of
gesturally or facially. In the first two lines, he words, hand (and head) gestures, and
demonstrated that “like this” meant a par facial displays, all serving the immediate
ticular position of his head in the water and narrative purpose.
also his closed eyes and impassive facial dis Coordination seems to be an important
play; then he returned his head to vertical factor in creating the meaning of these
and demonstrated that “water’s going over speech-related nonverbal acts. Engle (2000;
my head” meant that the water (represented see also Engle, 1998; Engle & Clark, 1995)
by his hands) was sweeping past the sides of conducted an intensive analysis of multi-
his head. Altogether, this first sentence said modal signals (speech, gestures, diagrams,
and showed that he was on his back with the and object demonstrations), which yielded
water flowing around, but not over, his face. several lines of evidence for the temporal
At the same time, he illustrated his helpless and linguistic integration of iconic and
ness facially, with his eyes closed and an indexical conversational gestures with
impassive expression, both also synchronous speech. For example, the gesture and the
with the verbal description of his dilemma. immediately accompanying speech segment
Accomplishing these depictions required a were co-expressive, referring to the same
high degree of coordination and integration. underlying referent:
For example, in order to show how “water’s
going over my head,” he had to return his For all but one of the 108 [communica
head to vertical and lift his hands up to the tive] nonverbal signals, a co-expressive
sides of his head (the preparatory phase) speech segment could be found within
before he started to say the phrase. [a] two intonation unit time window.
His impassive expression foreshadowed . . . In stark contrast to communicative
the latter part of his next sentence (“I grew nonverbal signals, in 14 of [the] 17
really, really calm”), but first he interrupted non-communicative cases, no co
his narration to insert a metacommunicative expressive speech was present. (Engle,
comment on his own reaction: He said a 1998, pp. 323–324)
shortened version of “And it’s strange,”
while making a very brief but clear facial One implication of Engle’s findings is
display of puzzlement, as if still unable to that timing is a metacommunicative tool
understand what he was describing as that speakers use to signal what is in the
his strange calmness in the situation. same integrated unit of meaning (Engle,
He then returned to the main narrative 2000). Bavelas, Holt, and Allison (2000)
line by depicting, verbally, prosodically, analyzed over 1,700 gestures to learn how
and facially, the calmness he now found they were connected to co-occurring speech.
06-Manusov.qxd 6/30/2006 7:26 PM Page 106
106–––◆–––Foundations
The data revealed that, whereas speakers received a single unitary interpretation
sometimes used linguistic markers (e.g., a distinct from the interpretation of either
deictic expression or a dummy noun phrase), constituent, not a preference for one
the most common link was simply timing rather than the other of two discrete
(70% of the gestures). Unfortunately, we do messages. (Sanders, 1987, p. 142; italics
not have comparable information for facial added)
displays. There are to our knowledge no sys
tematic studies of the precise temporal and These interpretations are consistent with
linguistic relationship of facial displays to Engle’s (1998) proposal that, following
words, although advances in digital analysis Grice’s (1967/1989) “cooperative princi
make such frame-by-frame analysis possi ple,” both speaker and addressee assume
ble, albeit still labor-intensive. that “all signals in a particular composite
signal are intended to be treated from the
start as contributing to a single, unified
REDUNDANCY AND interpretation” (Engle, 1998, p. 321).
NONREDUNDANCY But what is the internal nature of a mul
timodal message that produces a single,
Engle’s (2000) data also confirmed that, unified interpretation? We propose that, pri
although the gestures were virtually always marily because of synchronous timing, mul
consistent with the co-expressive speech, timodal elements can range from completely
they were sometimes complementary rather redundant to highly nonredundant and still
than duplicating the speech. This observa remain unified. Our research group has
tion contradicts the possibility that hand examined the degree or rate of redundancy
gestures are simply a redundant mode of in hand gestures with different functions.
expression. Examples 1 and 3 each illus Bavelas, Chovil, Lawrie, and Wade (1992),
trated that gestures can convey important for example, examined the degree of redun
information that is not in the speaker’s dancy of a gesture with its accompanying
words (e.g., the width of the dress and the phonemic clause and found that, across
way his head was in the water). Sometimes several different descriptive tasks, gestures
the simultaneous audible and visible ele depicting features of the task topic were
ments of a message, taken separately, might much more redundant with the words than
appear to contradict each other. As Sanders were gestures that referred to the interlocu
(1987) pointed out, however, receivers inte tor or to the interaction itself. The latter
grate these apparent contradictions at the (which we called interactive gestures) were
level of overall meaning (rather than at the usually completely nonredundant, although
level of components or physical source). they depended on and contributed to the
For example, Bugental, Kaswan, and Love meaning of the clause.
(1970; cited in Sanders, 1987) found An example from our data (Bavelas,
Sutton, Gerwing, & Johnson, 2002) illus
positively valued utterances paired with a trates a nonredundant interactive gesture.
negatively valued facial expression and At the beginning of their getting-acquainted
vocal qualities were judged by respon conversation, one participant had answered
dents to be sarcastic. Negatively valued the other’s inquiry by saying that he was a
utterances paired with positively valued Political Science major. A minute later, after
nonverbal displays were judged to involve they had been discussing another topic, the
joking. Thus, these inconsistent pairs same speaker returned to his academic
of utterances and nonverbal displays standing:
06-Manusov.qxd 6/30/2006 7:26 PM Page 107
Example 4.
S: “This is my last term, and, ah, Political Science. I was a double major . . .”
[flicks hand to A]
Because his addressee already knew portraying a past or present personal reac
what his major was, naming it again was tion) were redundant with speech; the other
not new but given (i.e., shared) informa 162 semantic displays by speakers were
tion. The speaker’s words (“and, ah, nonredundant. The 315 syntactic facial dis
Political Science”) were cryptic and did not plays by speakers (e.g., grammatical mark
fit the syntax of his sentence; they also con ers such as emphasis or question markers)
tained no reference to the addressee’s prior were virtually always nonredundant with
knowledge. In our view, it was the hand words, although not necessarily with
flick at the addressee that made the socially prosody. Finally, the 160 facial displays by
necessary reference; we interpret this ges listeners were, by definition, nonredundant
ture to mean “as you already know,” that with speech, because the listener was the
is, as citing or acknowledging that the person who was not speaking at the
addressee obviously still remembered what moment. Thus, over 70% of all displays
the speaker’s major was. The effective sen conveyed information that was not in the
tence would be, in words, “This is my last words. We speculate that the smiles by either
term and, as you know, I’m in Political speaker or listener, which were not analyzed
Science. I was a double major . . .” in this study, would follow a similar pattern.
One limitation of our analysis in Bavelas In sum, the third defining criterion of the
et al. (1992) was that it did not distinguish nonverbal acts of interest in this chapter is
among different experimental conditions that they form an integrated whole, with
and therefore included some conditions in words and each other. Integration does not
which there was no addressee or in which necessarily or even usually mean duplication,
the speaker and addressee were interacting as there is at least some evidence that the var
through a partition. More recently, Bavelas, ious modalities can convey different (nonre
Gerwing, Sutton, and Prevost (2002, 2005) dundant) information from each other.
examined gestural redundancy as a function When and how these diverse but unified
of the presence and visual availability of the elements operate is an important question for
addressee. When speaker and addressee were future research.
face to face, fewer than 20% of the speaker’s
gestures conveyed only information that was
also in their words; over 80% also included ♦ IV. Communication
some nonredundant information. In con
in Dialogue
trast, when the speakers were on the tele
phone or talking into a tape recorder to no
one, their gestures were significantly more This final section examines evidence that the
redundant; almost 60% of their gestures speakers and addressees use hand gestures
were entirely redundant with their words. and facial displays to communicate. To do
Likewise, Chovil (1989, 1991/1992; so, it is first necessary to discuss methodol
see also Bavelas & Chovil, 1997) reported ogy, because there are three different meth
redundancy data on 880 conversational ods for examining these issues. The first two
facial displays. She found that 243 of the focus on speaker and addressee separately:
405 semantic displays by speakers (e.g., An encoding design tests the conditions
06-Manusov.qxd 6/30/2006 7:26 PM Page 108
108–––◆–––Foundations
under which speakers do or do not produce Johnston, 1979; Schneider & Josephs,
hand gestures or facial displays; they should 1991). Only Chovil’s (1989, 1991) experi
be more likely to do so in face-to-face dia ment on addressees’ facial displays involved
logue than in other conditions. A decoding a conversational dialogue.
design seeks evidence that those who see In contrast, and for obvious reasons,
such acts also understand their meaning. We encoding studies of conversational gestures
have reviewed most of these studies in other have involved conversations, but virtually
places (Bavelas & Chovil, 2000; Bavelas, none of them have been spontaneous
Gerwing, Sutton, et al., 2005; Chovil, 1997; dialogues. In most of these gesture studies,
see also Kendon, 1994; Gray & Ambady, the addressees were nonreactive confeder
this volume) and will only summarize the ates or the experimenter, or even an imag
pattern here. The vast majority of studies ined other (e.g., Alibali, Heath, & Myers,
show evidence for communication, in that 2001; Bavelas, Kenwood, et al., 2002;
(1) individuals tend to produce more ges Beattie & Aboudan, 1994; Cohen, 1977;
tures or facial displays when someone Cohen & Harrison, 1973; Emmorey &
would see them and (2) observers can garner Casey, 2001; Krauss, Dushay, Chen, &
accurate information from these acts. Rauscher, 1995). Only three studies, to
Indeed, these studies have been so successful our knowledge, involved spontaneous dia
that we can now begin to see their limita logues between two participants (Bavelas
tions. Therefore, we will point out here et al., 1992; Bavelas, Gerwing, Sutton,
what, in retrospect, appear as deficiencies et al., 2005; Rimé, 1982).
in studies of isolated individuals (including As shown in Bavelas and Chovil (2000),
some of our own experiments) and will there are still far fewer decoding than
suggest a third method, one that examines encoding studies, and virtually all of them
the speaker and addressee together, within involve a similar design, one that tests
their interaction. whether outsiders to the interaction who
later view the gesture or facial display can
ascertain or at least agree on its meaning
ENCODING AND (e.g., Bavelas et al., 1986; Bavelas et al.,
DECODING DESIGNS 1990; Rosenfeld, Shea, & Greenbaum,
1979; Shea & Rosenfeld, 1976).3 In one
The typical encoding design varies study, Graham and Argyle (1975) showed
whether or not the speaker has a visually that the addressees were more accurate at
available recipient, for example, whether drawing figures when the speaker who
the speaker is alone or in the presence of described them had been able to gesture,
another person. The main limitation of the but there was little or no interaction invol
existing encoding designs is that they sel ved between speakers and addressees.
dom include a spontaneous face-to-face Although these studies are encouraging
dialogue, which should be the baseline con for demonstrating some decoding of gestures
dition. For example, most studies of facial or facial displays, the viewers were rarely
displays have instead used mere presence, the original addressees, a methodological
eye contact, or social context instead of choice that raises at least two problems.
face-to-face dialogue (e.g., Bavelas, Black, First, most decoder studies do not present
Lemery, & Mullett, 1986; Fernandez-Dols the entire conversation, so that the decod
& Ruis-Belda, 1995; Fridlund, 1991b; ing outsider has a fraction of the context
Jones, Collins, & Hong, 1991; Jones & that the addressee had. Our unpublished
Hong, 2001; Jones & Raag, 1989; Kraut & pilot studies with these designs showed
06-Manusov.qxd 6/30/2006 7:26 PM Page 109
110–––◆–––Foundations
includes a few sections adapted from that article, expression (pp. 334–346). Cambridge:
with the permission of the Journal of Language Cambridge University Press.
and Social Psychology and Sage Publications. Bavelas, J. B., & Chovil, N. (2000). Visible acts
2. Clark (1996, chap. 1) has outlined in of meaning. An integrated message model
fuller detail the characteristics of face-to-face of language use in face-to-face dialogue.
conversation as a fundamental setting for Journal of Language and Social Psychology,
language use. 19, 163–194.
3. We are not including any method that Bavelas, J. B., Chovil, N., Coates, L., & Roe, L.
isolates gestures from their verbal context (e.g., (1995). Gestures specialized for dialogue.
by using only the video without a sound track), Personality and Social Psychology Bulletin,
because such procedures treat conversational 21, 394–405.
gestures as if they were emblems. Bavelas, J. B., Chovil, N., Lawrie, D. A., & Wade,
A. (1992). Interactive gestures. Discourse
Processes, 15, 469–489.
♦ References Bavelas, J. B., Coates, L., & Johnson, T. (2002).
Listener responses as a collaborative process:
The role of gaze. Journal of Communication,
Alibali, M. W., Heath, D. C., & Myers, H. J. 52, 566–580.
(2001). Effects of visibility between speaker Bavelas, J. B., Gerwing, J., Allison, M., & Sutton,
and listener on gesture production: Some C. (2005, June). Evidence for grounding with
gestures are meant to be seen. Journal of non-redundant gestures: Co-constructing vir
Memory and Language, 44, 169–188. tual spaces. Paper presented at the second
Bangerter, A. (2004). Using pointing and conference of the International Society for
describing to achieve joint focus of atten Gesture Studies, Lyon, France.
tion in dialogue. Psychological Science, 15, Bavelas, J. B., Gerwing, J., Sutton, C., & Prevost,
415–419. D. (2002, June). Gestures in face-to-face,
Bavelas, J. B. (1990). Nonverbal and social telephone, and tape recorder conditions.
aspects of discourse in face-to-face interac Paper presented at the first conference of the
tion. Text, 10, 508. International Society for Gesture Studies,
Bavelas, J. B. (2000). Nonverbal aspects of flu Austin, TX.
ency. In H. Riggenbach (Ed.), Perspectives Bavelas, J. B., Gerwing, J., Sutton, C., &
on fluency (pp. 91–101). Ann Arbor: Prevost, D. (2005). Gesturing on the tele
University of Michigan Press. phone: Independent effects of dialogue and
Bavelas, J. B. (in press). The micro-social dimen visibility. Unpublished manuscript.
sion of face-to-face dialogue. In S. Duncan Bavelas, J. B., Holt, T., & Allison, M. (2000).
& E. Levy (Eds.), Papers in honor of Links between gestures and words.
David McNeill. Washington, DC: Unpublished research report.
American Psychological Association. Bavelas, J. B., Hutchinson, S., Kenwood, C., &
Bavelas, J. B., Black, A., Chovil, N., & Mullett, J. Matheson, D. H. (1997). Using face-to-face
(1990). Equivocal communication. Newbury dialogue as a standard for other communi
Park, CA: Sage. cation systems. Canadian Journal of Com
Bavelas, J. B., Black, A., Lemery, C. R., & munication, 22, 5–24.
Mullett, J. (1986). “I show you how you Bavelas, J. B., Kenwood, C., Johnson, T., &
feel.” Motor mimicry as a communicative Phillips, B. (2002c). An experimental study
act. Journal of Personality and Social of when and how speakers use gestures to
Psychology, 50, 322–329. communicate. Gesture, 2, 1–17.
Bavelas, J. B., & Chovil, N. (1997). Faces in dia Bavelas, J. B., Sutton, C., Gerwing, J., & Johnson,
logue. In J. A. Russell & J. M. Fernandez- T. (2002d). Analysis of gestures in dialogue
Dols (Eds.), The psychology of facial [Training CD], Department of Psychology,
06-Manusov.qxd 6/30/2006 7:26 PM Page 112
112–––◆–––Foundations
University of Victoria, Victoria, B.C., Clark, H. H., & Gerrig, R. J. (1990). Quotations
Canada. as demonstrations. Language, 66, 764–805.
Beattie, G., & Aboudan, R. (1994). Gestures, Clark, H. H., & Krych, M. A. (2004). Speaking
pauses and speech: An experimental while monitoring addressees for understand
investigation of the effects of changing ing. Journal of Memory and Language, 50,
social context on their precise temporal 62–81.
relationships. Semiotica, 99, 239–272. Cohen, A. A. (1977). The communicative
Birdwhistell, R. L. (1966). Some relationships functions of hand illustrators. Journal of
between American kinesics and spoken Communication, 27, 54–63.
American language. In A. Smith (Ed.), Com Cohen, A. A., & Harrison, R. P. (1973).
munication and culture (pp. 182–189). Intentionality in the use of hand illustrators
New York: Holt, Rinehart & Winston. in face-to-face communication situations.
Blurton-Jones, N. G. (1972). Criteria for use in Journal of Personality and Social Psychology,
describing facial expressions of children. In 28, 276–279.
N. G. Blurton-Jones (Ed.), Ethological stud Duncan, S., & Fiske, D. (1977). Face-to-face
ies of child behavior (pp. 365–413). interaction: Research, methods, and theory.
Cambridge: Cambridge University Press. New York: Erlbaum.
Brunner, L. J. (1979). Smiles can be back Ekman, P. (1979). About brows: Emotional and
channels. Journal of Personality and Social conversational signals. In J. Aschoof, M. von
Psychology, 37, 728–734. Cranach, K. Foppa, W. Lepenies, & D. Ploog
Buchler, J. (Ed.). (1940). Philosophical writings of (Eds.), Human ethology (pp. 169–202).
Peirce. London: Routledge and Kegan Paul. Cambridge, UK: Cambridge University Press.
Bugental, D. E., Kaswan, J. W., & Love, L. R. Ekman, P. (1993). Facial expression and emo
(1970). Perception of contradictory mean tion. American Psychologist, 48, 384–392.
ings conveyed by verbal and nonverbal Ekman, P. (1997). Should we call it expression
channels. Journal of Personality and Social or communication? European Journal of
Psychology, 16, 647–655. Social Sciences, 10, 333–359.
Camras, L. A. (1977). Facial expressions used Ekman, P., & Friesen, W. V. (1969). The reper
by children in a conflict situation. Child toire of nonverbal behavior. Categories, ori
Development, 48, 1431–1435. gins, usage, and coding. Semiotica, 1, 49–98.
Chovil, N. (1989). Communicative functions of Emmorey, K., & Casey, S. (2001). Gesture,
facial displays in conversation. Unpublished thought and spatial language. Gesture, 1,
doctoral dissertation, Department of Psycho 35–50.
logy, University of Victoria, Victoria, British Engle, R. A. (1998). Not channels but composite
Columbia, Canada. signals: Speech, gesture, diagrams and object
Chovil, N. (1991). Social determinants of facial demonstrations are integrated in multimodal
displays. Journal of Nonverbal Behavior, explanations. In M. A. Gernsbacher & S. J.
15, 141–154. Derry (Eds.), Proceedings of the Twentieth
Chovil, N. (1991/1992). Discourse-oriented Annual Conference of the Cognitive Science
facial displays in conversation. Research Society (pp. 321–326). Mahwah, NJ:
on Language and Social Interaction, 25, Erlbaum.
163–194. Engle, R. A. (2000). Toward a theory of multi-
Chovil, N. (1997). Facing others: A social com modal communication combining speech,
municative perspective on facial displays. gestures, diagrams, and demonstrations
In J. A. Russell & J.-M. Fernandez-Dols in instructional explanations. Unpublished
(Eds.), The psychology of facial expression doctoral dissertation, School of Education,
(pp. 321–333). Cambridge: Cambridge Stanford University.
University Press. Engle, R. A., & Clark, H. H. (1995, March).
Clark, H. H. (1996). Using language. Using composites of speech, gestures, dia
Cambridge: Cambridge University Press. grams and demonstrations in explanations
06-Manusov.qxd 6/30/2006 7:26 PM Page 113
114–––◆–––Foundations
LeBaron, C., & Streeck, J. (2000). Gestures, the mind (pp. 763–765). Oxford: Oxford
knowledge, and the world. In D. McNeill University Press.
(Ed.), Language and gesture (pp. 99–117). Random house unabridged dictionary (2nd ed.).
Cambridge: Cambridge University Press. (1993). New York: Random House.
Leeds-Hurwitz, W. (1987). The social history Rimé, B. (1982). The elimination of visible
of The History of an Interview: A behaviour from social interactions: Effects
multidisciplinary investigation of social on verbal, nonverbal and interpersonal
communication. Research on Language variables. European Journal of Social
and Social Interaction, 20, 1–51. Psychology, 12, 113–129.
Leeds-Hurwitz, W. (1989). Communication Roberts, G. L., & Bavelas, J. B. (1996). The
in everyday life: A social interpretation. communicative dictionary: A collaborative
Parkside, WI: University of Wisconsin- theory of meaning. In J. Stewart (Ed.),
Parkside Press. Beyond the symbol model. Reflections on
Levinson, S. (1983). Pragmatics. Cambridge: the nature of language (pp. 139–164).
Cambridge University Press. Albany: SUNY Press.
Linell, P. (1982). The written language bias in Rosenfeld, N. M., Shea, M., & Greenbaum,
linguistics. Linkoping, Sweden: University of P. (1979). Facial emblems of “right” and
Linkoping, Department of Communication. “wrong”: Topographical analysis and
McNeill, D. (1985). So you think gestures are derivation of a recognition test. Semiotica,
nonverbal? Psychological Bulletin, 92, 26, 15–34.
350–371. Sanders, R. E. (1987). The interconnection of
McNeill, D. (1992). Hand and mind: What utterances and nonverbal displays. Research
gestures reveal about thought. Chicago: on Language and Social Interaction, 20,
University of Chicago Press. 141–170.
McQuowan, N. (Ed.). (1971). The natural history Scheflen, A. (1968). Human communication:
of an interview. Microfilm collection of man Behavioral programs and their integration
uscripts on cultural anthropology. Chicago: in interaction. Behavioral Science, 13,
University of Chicago, Joseph Regenstein 44–55.
Library, Department of Photoduplicating. Scherer, K. R. (1980). The functions of nonver
Ozyurek, A. (2000). The influence of addressee bal signs in conversation. In R. N. St. Clair
location on spatial language and representa & H. Giles (Eds.), The social and psycho
tional gestures of direction. In D. McNeill logical contexts of language (pp. 225–244).
(Ed.), Language and gesture (pp. 99–117). Hillsdale, NJ: Erlbaum.
Cambridge, UK: Cambridge University Press. Schneider, K., & Josephs, I. (1991). The
Ozyurek, A. (2002). Do speakers design their expressive and communicative functions of
cospeech gestures for their addressees? The preschool children’s smiles in an achieve
effects of addressee location on representa ment-situation. Journal of Nonverbal
tional gestures. Journal of Memory and Behavior, 15, 185–198.
Language, 46, 665–875. Schober, M. F., & Clark, H. H. (1989).
Pike, K. L. (1972). Towards a theory of the struc Understanding by addressees and over-
ture of human behavior. In R. M. Brend hearers. Cognitive Psychology, 21,
(Ed.), Selected writings: To commemorate 211–232.
the 60th birthday of Kenneth Lee Pike Shea, M., & Rosenfeld, H. M. (1976).
(pp. 106–116). The Hague: Mouton. Functional employment of nonverbal social
Poyatos, F. (1980). Interactive functions and limi reinforcers in dyadic learning. Journal of
tations of verbal and nonverbal behaviors in Personality and Social Psychology, 34,
natural conversation. Semiotica, 30, 211–244. 228–239.
Quine, W. V. O. (1987). Symbols. In R. L. Slama-Cazacu, T. (1976). Nonverbal com
Gregory (Ed.), The Oxford companion to ponents in message sequence: “Mixed
06-Manusov.qxd 6/30/2006 7:26 PM Page 115
syntax.” In W. C. McCormack & S. A. Weiner, M., Devoe, S., Rubinow, S., & Geller, J.
Wurm (Eds.), Language and man: Anthro (1972). Nonverbal behavior and nonverbal
pological issues (pp. 217–227). The communication. Psychological Review, 79,
Hague: Mouton. 185–214.
Streeck, J., & Knapp, M. L. (1992). The inter Woods, J. (2005). New vs. given information: Do
action of visual and verbal features in gestures dance to the same tune as words?
human communication. In F. Poyatos (Ed.), Unpublished honours thesis, Department of
Advances in nonverbal communication Psychology, University of Victoria, Victoria,
(pp. 3–24). Amsterdam: Benjamins. B.C., Canada.
06-Manusov.qxd 6/30/2006 7:26 PM Page 116
07-Manusov.qxd 6/30/2006 4:17 PM Page 117
PART II
FACTORS OF INFLUENCE
07-Manusov.qxd 6/30/2006 4:17 PM Page 118
07-Manusov.qxd 6/30/2006 4:17 PM Page 119
7
THE BIOLOGICAL FOUNDATIONS
OF SOCIAL ORGANIZATION
The Dynamic Emergence of Social Structure
Through Nonverbal Communication
◆ 119
07-Manusov.qxd 6/30/2006 4:17 PM Page 120
120–––◆–––Factors of Influence
between elements—that is, between at the same time they reflect the primordial
individuals. In this chapter, we present a imperatives seen in simple creatures: arousal
view of social structure as an emergent, self- and quiescence, approach and avoidance,
organizing dynamic system, and we seek dominance and submission, cooperation
to demonstrate the critical role of nonver and competition, courtship and sex.
bal communication in this emergence. We
review evidence relevant to this thesis,
arguing that even simple organisms show ♦ Selfishness, Altruism, and
evidence of motivational-emotional systems Communicative Genes
underlying basic behavior and sociality.
These systems exist not only within the
A fundamental issue in any discussion of
individual organism but also “out there” in
communication and sociality is the question
the communicative relationship associated
of cooperation and competition. Are living
with genetically structured displays in
creatures responding to genetic influences
senders and preattunements in receivers.
that are fundamentally selfish? Is altruism
The result is a spontaneous communicative
a cultural ideal with no biological under
interplay from which, over time, social struc
pinning? Or is cooperation—and attach
ture emerges. In human beings, this emer
ment and love—anchored in fundamental
gence proceeds so naturally and effortlessly
biological imperatives? The answers to
that it is often taken for granted, but many of
these questions turn on the technical issue
the deepest meanings conveyed by everyday
of the unit of selection in evolution, and
nonverbal communication reflect these
communication theory has important con
unseen, primordial biological imperatives.
tributions to make in this domain.
This chapter considers how, through
genetically structured displays and preat
tunements, spontaneous communicative SELFISH GENES
interplay creates the basis for the social
organization of species. The chapter moves The “selfish gene” theory of Richard
from the biochemical bases of motivation Dawkins (1989) and others was based on
and emotion that humans share with sim the interpretation of natural selection as
pler species to the particular ways in which operating at the level of the gene (i.e., that
social and moral emotions emerge from this individuals are motivated to maximize the
foundation. It examines the special role of number of genes contributed to subsequent
language in human social organization and generations; Hauser, 1996). A corollary of
the interaction between spontaneous non this gene selectionist account of evolution
verbal communication and linguistically was that the gene is the ultimate unit of evo
organized communication. Finally, it pre lutionary selection: The gene is programmed
sents human interaction as a self-organizing only to make copies of itself. Dawkins
dynamic system involving the direct com (1982) argued that the unit of evolutionary
munication of social and moral emotions, selection must be self-replicating across the
including pride and arrogance, guilt and span of evolutionary time. He termed this
shame, envy and jealousy, pity and scorn, unit an active, germ-line replicator. His
and, potentially, moral emotions, such as candidate for replicator was the gene, which
triumph, humiliation, resentment, and con is maintained virtually intact from genera
tempt. These social and moral emotions are tion to generation across evolutionary
the foundation of human interaction, and timescales. Fitness in the evolutionary sense
07-Manusov.qxd 6/30/2006 4:17 PM Page 121
is based not on the survival of the individ B’s general sending accuracy, A’s general
ual but on the survival of the individual’s receiving ability, and the unique ability of
genes, termed inclusive fitness. A to read B. The latter quantity reflects
One of the implications of this position the unique relationship of A and B and is
is that genuine altruism—sacrificing one’s termed the “relationship effect.”
own genetic fitness to preserve the genetic
fitness of others—should not exist at the COMMa to b = (SENDINGa + RECEIVINGb +
122–––◆–––Factors of Influence
124–––◆–––Factors of Influence
the deception. The result is an evolutionary noted, honest expressivity comes at a poten
“arms race” in which senders evolve pro tial cost to the sender—those who show
gressively more persuasive and manip their feelings accurately put themselves at
ulative “sales pitches,” and receivers risk of exploitation—and costly signals are
become increasingly sales resistant (Krebs more reliable. Thus, the costly nature of the
& Dawkins, 1984). Zahavi (1977) moder display is consistent with its representing an
ated this view by suggesting that displays accurate readout of the sender’s motivational-
can be honest but if and only if they are emotional state. Second, there is evidence
costly to the sender: In the absence of costs, that spontaneous emotional expressivity is in
the display becomes unreliable because it is fact valued by others. Sabatelli and Rubin
too easily imitated. (1986) found that emotional expressivity is
The question of honest versus mani positively related to judgments of interper
pulative communication can be approached sonal attraction, independent of physical
from the viewpoint of economic game attractiveness. Third, studies demonstrating
theory. Frank (1988, 2001, 2004) argued that personality judgments based on brief
that cooperation can succeed only in the “thin slices” of nonverbal behavior have
presence of markers that distinguish indi accurately predicted interpersonal outcomes
viduals who are willing to cooperate—that (Ambady & Rosenthal, 1992) show that
is, individuals who are “trustworthy”— emotional expressivity has an immediate
from those who will not. Such markers impact, which could allow fast and accu
must involve nonverbal behaviors that rate evaluation of the trustworthiness of
occur during the communication between strangers.
players (Boone & Buck, 2003). For For nonverbal expressiveness to function
example, Frank, Gilovich, and Regan as an accurate “sincerity detector,” sponta
(1993) allowed partners to meet and inter neous displays must differ from intentional
act before a one-shot Prisoner’s Dilemma pseudospontaneous behaviors. Indeed,
(PD) game but not discuss the game itself. spontaneous expressive behavior does
Cooperation was enhanced overall, but differ from posed or intentionally enacted
more significantly, the participants could behavior, so that there is a signal basis for
apparently predict the decision of partners distinguishing between honest and manipu
to cooperate or compete, so that dyads lative displays. Cohn and Schmidt (2004)
tended to show either mutual cooperation compared spontaneous smiles elicited by
or mutual defection, with relatively few viewing a comedy film with deliberate
cases of one partner taking advantage of the smiles, using automatic feature tracking to
other. Apparently, the intent to cooperate extract and represent dynamic features of
or compete was somehow discernable in the behavior as well as its morphology.
the partners’ behavior during the face-to They found that spontaneous and deliber
face interaction. ate smiles could be distinguished: There
Boone and Buck (2003) combined con were differences in the timing of smile
siderations about whether displays are onsets, and there was a tight coupling
honest or manipulative with evidence about between duration and amplitude in sponta
superficially self-defeating cooperation in neous smiles consistent with the contention
social dilemmas and suggested that a key that the spontaneous smile is automatic and
marker of trustworthiness is spontaneous involuntary. Also, there is initial empirical
nonverbal expressivity. This argument was evidence that expressiveness predicts coop
based on several considerations. First, as eration in the PD game. Rauh, Polonsky,
07-Manusov.qxd 6/30/2006 4:17 PM Page 125
126–––◆–––Factors of Influence
the area have reached a critical mass, and the life cycle of the slime mold Dictyo
thereby to coordinate collective behavior. stelium discoideum illustrates the implica
It is used by certain pathogenic bacteria to tions of cell communication for larger
trick the immune systems of their hosts. The issues of overall social organization: the
bacteria are quiescent—not activating a organization of multicelled organisms.
full immune system response—until their Slime molds are abundant (a spoonful of
numbers are sufficient to overwhelm the rich garden soil contains millions), their
victim. When a quorum is attained, the reg principal food is rotting organic material,
ulation of bacterial genes is changed so that and they are responsible for the slimy mate
toxic “virulence factors” are produced. For rial found on rotting wood. At one stage in
example, in cystic fibrosis, when the bacte their life cycle, D. discoideum are single-
ria reach a certain concentration, they pro celled amoebae that live independently and
duce a “biofilm,” a tough shell that protects display negative chemotaxis. That is, they
them from attack from the immune system produce pheromones that repel similar
of the victim or from antibiotics. The bacte amoebae (Newell, 1981), a function analo
ria can then reproduce, produce toxins, and gous to territorial/threat displays in more
damage tissues in relative safety (Riedel & complex species, that tend to spread indi
Eberl, 2002). viduals evenly over the available environ
The molecules that bacteria use to ment. When food in the environment
communicate have structures similar to becomes scarce, the negative chemotaxis
those used by humans and many other crea ends, and a positive chemotaxis begins,
tures. The signals or displays used by bacte attracting rather than repelling the amoe
ria typically involve amino acids or peptides bae. In D. discoideum the molecule on
functioning as pheromones (Gallio, Sturgill, which the positive chemotaxis is based is
Rather, & Kylsten, 2002). Peptides are cri cyclic adenosine monophosphate, a ubiqui
tical signaling molecules in human beings, tous molecule involved in metabolism and
including the endorphins, oxytocin (OXY), in the functioning of many neurotrans
vasopressin (VP), cholecystokinin, diazepam- mitters in human beings. The result is an
binding inhibitor, GnRH, adrenocortico aggregation center where the single-celled
tropic hormone, and insulin. Many of these amoebae come together and form a multi-
molecules exist virtually unchanged in celled form termed a “grex,” which may be
simple creatures. Moreover, these mole made up of millions of individuals.
cules are also associated with subjectively This journey culminates in the differenti
experienced feelings and desires in human ation of the cells into a “fruiting body.”
beings, including attachment, euphoria, Cells at the front end die and become a cel
pain, erotic feelings, panic, anxiety, and lulose stalk anchored to a secure footing,
stress. Gallio and colleagues (2002) con and a mass of cells from the rear then liter
cluded that peptide-mediated signaling is ally climb over their fallen colleagues to
a “widely conserved mechanism for signal the top of the stalk and form spores that
release . . . not only the signal releasing become released into the environment.
mechanism seems conserved, but also the Given favorable conditions, the spores
molecule that carries it out . . . (suggesting become individual amoebae and begin the
that) . . . these molecules share a common life cycle anew (Waters & Bassler, 2005). In
ancestry” (pp. 12212–12213). this process, the individual amoebae that
form the stalk give up any possibility for
Slime Molds. Whereas quorum sensing passing on their genes, whereas the individ
demonstrates that bacteria communicate, uals that climb the stalk carry on the
07-Manusov.qxd 6/30/2006 4:17 PM Page 127
species. As noted, giving up the potential to are polygamous. Prairie voles (Microtus
pass on one’s own genes while enhancing ochrogaster) show monogamy as evidenced
the genetic potential of other individuals by selective and lasting partner preferences
is the biological definition of altruism (pair bonds) that are activated by mating
(Wilson, 1975). Thus, D. discoideum in its (Wang & Aragona, 2004). The bonds are
life cycle demonstrates what might be lasting: If the bonded partner dies, a surviv
termed proto-threat behavior in its negative ing prairie vole will live alone rather than
chemotaxis, proto-attachment behavior in take a new mate. Both male and female
its positive chemotaxis, and proto-altruistic provide extensive care for their pups, with
behavior in its stalk formation. the male helping to build the nest and
There are many other examples of spending almost as much time as the female
paleosociality in simple creatures. In insect with the young. If separated from their
societies, for example, social organization is parents, the pups become agitated and dis
largely “hard wired”—that is, it is under play ultrasonic distress calls and high stress
close genetic control and has little to do with evidenced by increases in cortisol. In con
learning or flexibility. The genes accomplish trast, closely related species—the meadow
social organization via spontaneous com vole (Microtus pennsylvanicus) and mon
munication: genetically based displays and tane vole (Microtus montanus)—are non-
preattunements in different individual organ monogamous. The latter breed promiscuously
isms interacting with one another over time. and nest independently, and the males play
no parenting role. Even the females aban
don their pups soon after birth, and the
SOCIALITY IN MAMMALS pups do not appear to be distressed by
abandonment (Carter, Lederhendler, &
In mammals, there is an additional con Kirkpatrick, 1997).
sideration for understanding social organi There is evidence that attachment in
zation: Unlike most reptiles, the mammal is voles is related to the peptides OXY and
helpless at birth. The brain and spinal cord VP. Brain areas associated with OXY are
do not come preprogrammed: They are much larger in the prairie vole than in the
programmed by the animal’s experience montane vole (Insel, Young, & Wang,
over the course of development in the terri 1997), and the distribution of VP receptors
torial and social environment. This experi in the male brain is different, with the
ence requires communication mechanisms prairie vole and other monogamous species
involving bonding, nurturance, and protec having a relatively high density of VP recep
tion (i.e., attachment), to ensure that the tors (termed V1aR) in the ventral forebrain,
young animal receives adequate parental in an area of the brain also associated with
attention until it is physically and socially dopamine-mediated reward (Lim et al.,
mature enough to function on its own. An 2004; Young, Lim, Gingrich, & Insel,
examination of voles can help demonstrate 2001). Moreover, OXY and VP are critical
these communicative mechanisms. to the establishment of the pair bonds of
prairie voles. In the natural setting, mating
Polygamous and Monogamous Voles. The is associated with OXY release, which
vole is a particularly interesting case illus apparently functions to cement partner
trating the biology of attachment. The vole preferences in that the individuals prefer
is a mouse-size rodent, some species of who they are with when OXY increase
which are monogamous, forming lasting occurs (Carter, DeVries, & Getz, 1995).
female-male bonds, whereas other species Thus, OXY is both necessary and sufficient
07-Manusov.qxd 6/30/2006 4:17 PM Page 128
128–––◆–––Factors of Influence
play relationships with peers. This sets plays a similar role, but now in interaction
the stage for the heterosexual affectional with language, a new source of social orga
system, which is the basis for adult monkey nization missing in other animals.
communicative relationships and social
organization. Harlow suggested that these
are analogous to infancy, childhood, and ♦ Language, Nonverbal
adolescence in humans. Communication, and
Harlow’s work also demonstrated that Human Social Organization
early social deprivation can have devastat
ing effects on an individual animal’s abilities
to participate in social behavior. Contact The human brain shares many features
comfort—contact with a soft skinlike with the brains of other animals, particu
surface—was found to be sufficient to larly in the phylogenetically older subcorti
instill a basic sense of trust, but trusting cal and limbic regions, which are associated
individuals raised separately from peers with motivational and emotional process
showed significant deficits in courting and ing (Buck, 1999). In those regions, similar
sexual behavior after puberty. Animals neurochemicals appear to function in ways
reared in wire cages were extremely fearful analogous to those in other animals, and
and aggressive as adults (Harlow, 1971). they do seem to have the ability in some
There is evidence that these social deficits cases to overwhelm more “rational” linguis
are related to an inability to communicate. tic control mechanisms. Addiction provides
Specifically, Miller, Caul, and Mirsky many tragic examples. Indeed, it can be
(1967) demonstrated that monkeys isolated argued that these motivational-emotional
in infancy were unable as adults either to systems set the agenda for human commu
send or to receive accurately in Miller’s nication and underlie the human equiva
cooperative conditioning paradigm measur lents of the primordial characteristics
ing communication in monkeys (Miller, found in all social structures—arousal
Banks, & Ogawa, 1962, 1963). quiescence, approach-avoidance, dominance-
submission, cooperation-competition, and
courting-sexuality—all of which are commu
SUMMARY nicated nonverbally.
Language introduces in human beings
Examples of social organization from a source of behavioral control and social
simple bacteria to mammals including organization that is lacking in animals,
primates demonstrate common principles creating a real discontinuity with other
involving an unfolding of biological poten species (Buck, 1988). As complex proposi
tial in the course of development. At all tions become expressible and recordable in
these levels, social organization emerges out language, questions of linguistic organiza
of communicative interactions as a self- tion that are distinct from biological fac
organizing dynamic system. In mammals, tors become important determinants of
attachment systems involving bonding, human social organization. On the other
nurturance, and protection of the young, hand, language itself is based on biological
are critical and demand communication mechanisms, associated particularly with
systems through which attachment can func the left hemisphere (LH) of the brain in
tion. We now turn to human social organi almost all humans. For example, damage
zation, where nonverbal communication to the LH has long been associated with
07-Manusov.qxd 6/30/2006 4:17 PM Page 130
130–––◆–––Factors of Influence
Fundamental Social Motives. The motiva jealousy, and scorn on the other relates to
tion arising from attachment has, arguably, whether the responder is securely or inse
two aspects: the need to be loved and the curely attached in that particular comparison
need to meet or exceed expectations (Buck, situation. A secure person may experience
1988). People are strongly motivated to be pride, guilt, envy, and pity in situations where
esteemed and loved and to do that which is an insecure person may experience arro
expected and indeed to exceed expectations. gance, shame, jealousy, and scorn.
Individuals can attain or fail to attain these
two social goals themselves, and they can The Dynamics of Social Emotions. These
compare themselves with other persons who fundamental interpersonal contingencies
attain or fail to attain these goals. The rela do not elicit social emotions in isolation
tive strength of these two social motives from one another. Rather, because success
should be related to the security of the and failure are relative, the social emotions
attachment of the person to reference are dynamically related one to the other.
persons evaluating them. Persons who are If P is successful and experiences pride-
insecurely attached are likely to need reas arrogance, P will tend to pity-scorn the less
surance that they are loved, and that over successful O. On the other hand, O will
whelming need might overshadow needs to tend to be envious-jealous of P and will
meet or exceed expectations. Secure individ experience guilt-shame in comparison to P.
uals, in contrast, should be relatively assured The resulting dynamics of the social emo
of love, and they will instead be able to focus tions are summarized in Figure 7.1a.
on meeting and exceeding expectations. To test these predictions, American and
Japanese students were asked to read a
Fundamental Interpersonal Contingencies. series of six simple scenarios, imagining
The comparisons just described constitute they had gone home for a high school
fundamental interpersonal contingencies, reunion and heard the scenario about an old
which yield eight primary social emotions, acquaintance “O” (Buck, Nakamura, Vieira,
including pride-arrogance when one is & Polonsky, 2005). The scenarios involved
doing comparatively well, guilt-shame positive or negative outcomes for O that
when one is not doing well, envy-jealousy were or were not deserved: for example,
when the other is doing well, and pity- O won the lottery, was hit by lightning,
scorn when the other is not doing well invented a new computer chip, and was
(Buck, 2004). More specifically, a person P jailed for selling drugs to children.
who succeeds experiences what could be Participants were asked to rate how they
labeled in English pride or arrogance; if P would feel about O and how O would feel
fails, P experiences guilt or shame; if com about them, using the list of the primary
parison person O succeeds, P tends to expe social emotions in English or Japanese as
rience envy or jealousy; if O fails, P tends to appropriate. The results relevant to the pre
experience pity or scorn. These comparisons dictions about emotion dynamics are sum
involve relative dominance, which, as we marized in Table 7.1. Briefly, across both
have seen, is a fundamental characteristic of nations, in all cases save one, the average
social structure and organization where one correlations were in the predicted direction
is in a position of relative strength and the and statistically significant. Thus, the predic
other in a position of relative weakness. The tions about the dynamics of social emotions
difference between pride, shame, envy, and in Figure 7.1a were strongly supported in
pity on the one hand and arrogance, shame, both America and Japan.
07-Manusov.qxd 6/30/2006 4:17 PM Page 132
132–––◆–––Factors of Influence
PITY-SCORN
PRIDE- GUILT-
ARROGANCE SHAME
ENVY-
JEALOUSY
a.
CONTEMPT
Pity-Scorn
TRIUMPH HUMILIATION
Pride- Guilt-
Arrogance Shame
Envy-
Jealousy
b. RESENTMENT
ADMIRATION-
SYMPATHY
ADMIRATION-
c. SYMPATHY
Table 7.1 Hypotheses About the Dynamics of Primary Social Emotions and Average
Correlations From America and Japan
America Japan
NOTE: Pride is given as an example. Data represent the average correlations computed across all relevant
social emotions. English pronunciation of equivalent words in Japanese: Hokori, pride; goumannsa,
arrogance; zaiakukann, guilt; hazukashisa, shame; urayami, envy; sitto, jealousy; awaremi, pity; keibetsu,
scorn.
***p < .001. **p < .01. *p < .05.
07-Manusov.qxd 6/30/2006 4:17 PM Page 134
134–––◆–––Factors of Influence
This result supports the key proposition moral emotions. We propose that like social
that social emotions emerge as dynamic emotions, moral emotions involve issues of
systems from biologically based prosocial relative success or failure on the part of O:
attachment, in interaction with fundamental If we are winning and perceive that we
interpersonal contingencies involving social deserve it, we respond with feelings of tri
comparison, over the course of individual umph; if we lose and deserve it, we experi
development. Although the language used ence humiliation. If O wins and deserves it,
to label the social emotions is cultural, we experience admiration even as we may
attachment and the eight unions of interper simultaneously experience envy-jealousy.
sonal contingencies are fundamentally similar But if we believe that O does not deserve
in two nations with different values and lan success, our envy-jealousy is mixed with bit
guages. The implication is that this emer ter resentment. If O loses and deserves it, we
gence is a pancultural phenomenon, common feel contempt, but if O’s loss is not deserved,
to all cultures and to all historical times. we feel sympathy. Morality is a double-
The social emotions are “on,” and the edged sword, with the capacity to amelio
dynamic pattern of social emotions is rate potential conflict arising from inequities
present, in every human communicative or greatly exacerbate such conflict.
exchange. The interchange depicted in
Figure 7.1a is rather competitive, with the Moral Indignation. When unfairness is per
proud-arrogant individual clearly express ceived, moral considerations can greatly
ing dominance to the partner and the exacerbate the conflict shown in Figure
partner responding with a downcast, sub 7.1a, evoking strikingly strong negative
missionlike display. Such feelings can be emotion: moral indignation, as illustrated
raised even in trivial encounters, as in in Figure 7.1b. P is triumphant and shows
passing on the street. Generalizing from scorn toward O, who responds with humil
Ambady and Rosenthal (1992), we can be iation and resentment: a powerful, poiso
exquisitely sensitive to thin slices suggesting nous combination fraught with potential
even a hint of pride-arrogance: It can set off conflict and aggression. Resentment can
alarm bells that one is being put down, lead to increased conflict that can become
snubbed, and disrespected. Perhaps such deadly, with murderous assaults on O being
signals—a smirk, a quick turning away of perceived to be morally justified and even
the eyes, a subtle raise of the chin—account gratifying, eliciting schadenfreude: a delight
for the ability of people in the Frank et al. in the discomfort and even agony of the
(1993) study to calculate that their partner other. Such perceptions can become socially
is likely to compete. Conversely, we quickly shared, particularly through encourage
grasp weakness and submission in others— ment via media propaganda, and lead to
responding involuntarily with pity or scorn— long-term conflicts between peoples that
and we can directly perceive honest and can turn into war, ethnic cleansing, and
friendly trustworthiness. genocide.
is following the rules fairly, then each can ♦ Summary and Conclusions
view the other with admiration and sym
pathy. Moreover, they themselves feel a
sense of dignity and trust in the relation Signs of basic sociality—arousal-quiescence,
ship. Prosocial emotions can occur, such as approach-avoidance, dominance-submission,
gratitude and appreciation, when the other cooperation-competition, courting-sexual
follows the rules and does the expected. ity—are rooted in the deepest and most
These and other positive moral feelings ancient regions and molecules of the brain.
constitute the emotional bases of civil Their influence is almost unnoticed—they
exchange (see Figure 7.1c). Even bitter ene function largely below the radar of
mies may engage in civil interaction, albeit consciousness—but it is profound and per
grudgingly, if they can mutually acknowl vasive. These signs constitute the mecha
edge that each is following the rules and nism by which social structure emerges and
acting with dignity. Seemingly trivial ritu organizes itself, whether in bacteria and
als of greeting, making brief eye contact, slime molds or in the rough-and-tumble
and saying “good day,” function to remind play of children. The mechanism of the
interactants that each is following the rules emergence is spontaneous nonverbal com
and acting with dignity. It is possibly the munication—unconscious, intuitive, and
nonverbal expression of civility that under effortless—involving the direct perception of
lies the perception of mutual trustworthi and preprogrammed responses to displays.
ness that enabled participants in Frank et Spontaneous communication links us one to
al. (1993) to guess correctly that their another in a primordial biological dance.
partner would cooperate—that is, that We have seen that informal human
their partner was trustworthy (Boone & communication involves social comparisons
Buck, 2003). that evoke a dynamic system of social emo
Moral emotions are based on the same tions including pride and arrogance, guilt
elements as social emotions: an affective and shame, envy and jealousy, and pity and
caring based biologically on attachment, scorn, and potentially also moral emotions,
interacting with fundamental interpersonal such as triumph, humiliation, resentment,
contingencies involving comparative suc and contempt. These potentially disruptive
cess and failure. But, with moral emotions emotions are often finessed by common and
there is the additional element of fairness virtually unconscious verbal and nonverbal
and justice, equity, or following the rules. rituals of politeness that establish the social
Moral emotions emerge together with lubricant of civility, evoking moral emotions
social emotions as dynamic systems, intu of admiration, sympathy, trust, respect, and
itively and effortlessly, during socioemo gratitude. They are leavened by other basic
tional development. They emerge, as Piaget signals, including subtle signs of attraction
(1971) suggested, from experiences in that hint at courtship. These signs may be
social interactions, particularly play. They used to deceive and manipulate, but the very
have animal analogies in the pack structure fact of deception implies their potential to
that emerges spontaneously from the inter reveal underlying truth.
actions of young wolves raised together by Human social behavior retains a founda
Woolpy and Ginsburg (1967) and in the tion in primordial sociality as evidenced
need for peer interactions in young mon by emergent systems of social and moral
keys demonstrated in the work of Harlow emotions that naturally, effortlessly, and
(1971). largely unconsciously guide each and every
07-Manusov.qxd 6/30/2006 4:17 PM Page 136
136–––◆–––Factors of Influence
human communicative exchange, medi Buck, R., Nakamura, M., Vieira, E. T., Jr., &
ated by nuances of nonverbal behavior that, Polonsky, M. (2005, November). Dynamics
although subtle, can engender deep and pow of higher level social emotions: A cross-
erful emotions ranging from enraged humili national comparison of America and Japan.
Paper presented at the annual conference of
ation and resentment to delight and triumph,
the National Communication Association,
tender nurturance and protectiveness, pas
Boston.
sionate desire, and compassionate love.
Buck, R., & VanLear, C. A. (2002). Verbal
and nonverbal communication: Distin
guishing symbolic, spontaneous, and
♦ References pseudo-spontaneous nonverbal behavior.
Journal of Communication, 52, 522–541.
Burgdorf, J., & Panksepp, J. (2001). Tickling
Ambady, N., & Rosenthal, R. (1992). Thin induces reward in adolescent rats. Physiology
slices of behavior as predictors of inter & Behavior, 72, 167–173.
personal consequences: A meta-analysis. Carter, C. S., DeVries, A. C., & Getz, L. L.
Psychological Bulletin, 2, 256–274. (1995). Physiological substrates of mam
Boone, R. T., & Buck, R. (2003). Emotional malian monogamy: The prairie vole model.
expressivity and trustworthiness: The role Neuroscience & Biobehavioral Reviews,
of nonverbal behavior in the evolution 19, 303–314.
of cooperation. Journal of Nonverbal Carter, C. S., Lederhendler, I. I., & Kirkpatrick,
Behavior, 27, 163–182. B. (Eds.). (1997). The integrative neurobiol
Buck, R. (1984). The communication of emo ogy of affiliation. Annals of the New York
tion. New York: Guilford Press. Academy of Sciences, 807, 260–272.
Buck, R. (1988). Human motivation and emo Cohn, D. F., & Schmidt, K. L. (2004). The tim
tion (2nd ed.). New York: Wiley. ing of facial motion in posed and sponta
Buck, R. (1999). The biological affects: A typol neous smiles. International Journal of
ogy. Psychological Review, 106, 301–336. Wavelets, Multiresolution, and Information
Buck, R. (2002). The genetics and biology of Processing, 2, 1–12.
true love: Prosocial biological affects and Darwin, C. (1998). The expression of the emo
the left hemisphere. Psychological Review, tions in man and animals (3rd ed.). New
109, 739–744. York: Oxford University Press. (Original
Buck, R. (2004). The gratitude of exchange and work published 1872)
the gratitude of caring: A developmental Dawkins, R. (1982). The extended phenotype:
interactionist perspective of moral emotion. The gene as the unit of selection. Oxford,
In R. A. Emmons & M. McCullough (Eds.), UK: Oxford University Press.
The psychology of gratitude (pp. 100–122). Dawkins, R. (1989). The selfish gene: New edi
New York: Oxford University Press. tion. New York: Oxford University Press.
Buck, R., & Duffy, R. (1980). Nonverbal com Dawkins, R., & Krebs, J. R. (1978). Animal
munication of affect in brain damaged signals: Information or manipulation?
patients. Cortex, 16, 351–362. In J. R. Krebs & N. B. Davies (Eds.), Behav
Buck, R., & Ginsburg, B. (1991). Emotional com ioural ecology: An evolutionary approach
munication and altruism: The communi (pp. 282–309). Oxford, UK: Blackwell.
cative gene hypothesis. In M. Clark (Ed.), Frank, R. (1988). Passions within reason: The
Altruism: Review of personality and social strategic role of the emotions. New York:
psychology (Vol. 11, pp. 149–175). Newbury Norton.
Park, CA: Sage. Frank, R. (2001). Cooperation through emo
Buck, R., & Ginsburg, B. (1997). Communicative tional commitment. In R. M. Hesse (Ed.),
genes and the evolution of empathy. Evolution and the capacity for commit
In W. Ickes (Ed.), Empathic accuracy ment (pp. 57–76). New York: Russell Sage
(pp. 17–43). New York: Guilford Press. Foundation.
07-Manusov.qxd 6/30/2006 4:17 PM Page 137
138–––◆–––Factors of Influence
8
AN EVOLUTIONARY APPROACH
TO UNDERSTANDING
NONVERBAL COMMUNICATION
� Kory Floyd
Arizona State University
◆ 139
08-Manusov.qxd 6/30/2006 7:26 PM Page 140
140–––◆–––Factors of Influence
142–––◆–––Factors of Influence
A proximal cause is the condition or set of fell in love with each other, couldn’t imagine
conditions that appears to facilitate the their lives without each other, and wanted to
behavior in the specific time, place, and spend the rest of their lives together. Few are
manner in which it occurred. In response likely to say that they got married in order to
to the question, “Why did you eat dinner pass their genetic materials on to future gen
tonight?” one might answer by identifying erations. The theory of natural selection
a proximal cause: “Because I felt hungry.” posits that this is, indeed, the ultimate cause
By contrast, an ultimate cause is the condi of pair bonding, irrespective of people’s con
tion or set of conditions that represents the scious awareness. That is, the evolutionary
original or higher-order cause of a behavior approach does not require individuals to be
(and often dictates the connection between aware of the ultimate causes of their behav
the behavior and its proximal causes). iors in order for those ultimate causes to be
Thus, an ultimate answer to the question, operative.
“Why did you eat dinner tonight?” might
be, “Because I must eat in order to survive, Adaptations need not be adaptive for
and so I have evolved the sensation of modern living. When people think about
hunger as a way of motivating me to eat on their modern environments, it can be diffi
a regular basis.” This example illustrates cult to understand how particular adap
how a given nonverbal behavior (e.g., eat tations are beneficial. Why, for instance, do
ing) can be considered to be caused both by humans have a preference for sweet, fatty,
a proximal agent (hunger) and by an ulti and salty foods? How could such a prefer
mate agent (human need for nutrients). ence possibly be adaptive, when overindul
Although both are valid causes, an evolu gence in these types of foods can lead to
tionary approach is concerned primarily obesity, high cholesterol, heart disease, and
with identifying ultimate causes that even death? Evolutionary theories do not
provide some clue as to the environmen try to explain human adaptations with ref
tal challenges that adaptations evolved erence to modern living. Rather, such theo
to meet. In many cases, however, there is an ries focus on physical and psychological
unmistakable link between ultimate and traits that would have been adaptive in the
proximal causes. societies of our ancestors. Evolution usually
operates slowly, and modern civilization is
Adaptations need not operate at a conscious remarkably young when considered on an
level. When asked to account for their own evolutionary timescale. Agriculture was
behaviors, people commonly identify proxi invented only about 10,000 years ago, and
mal causes with little or no regard for ulti civilization (let alone, modern civilization)
mate, higher-order causes. This is often is an even more recent phenomenon.
because they are simply unaware of what the Humankind has spent more than 99% of
ultimate causes might be or how they might its history living in hunter-gatherer soci
be operating through more proximal causes eties, and as Morris (2001) pointed out, it is
(Kenrick & Simpson, 1997). Evolutionary not likely that natural selection has made
theories acknowledge that many ultimate any noticeable modifications in the human
causes operate outside of individuals’ con brain in the short period of time repre
scious awareness of them, and they contend sented by modernity. As a result, some
that this is not problematic. For example, if traits that were adaptive for our hunter-
people are asked to explain why they got gatherer ancestors may be useless or even
married, they will likely say it is because they maladaptive in modern times.
08-Manusov.qxd 6/30/2006 7:26 PM Page 143
Adaptations need not be adaptive for every members, and all group members are
person or in every instance. It can be diffi equally cared for. In such a situation,
cult for an individual to see how particular anything that benefits the group as a whole
traits might be adaptive if, for whatever also benefits each member individually,
reason, those traits do not produce the because of the communal sharing of
adaptive outcomes for that individual. The resources. Suppose, however, that a group
human sex drive can provide a useful member were to find a large sum of cash
example. The fact that sex is physically plea that the other group members did not
surable for (most) humans can be consid know about. Would it benefit this member
ered to be adaptive in the sense that it to contribute the money to the group? Yes,
motivates humans to engage in intercourse, because an economic benefit to the group
which is necessary for reproduction of the would benefit all members. It would bene
species.3 To be an adaptation, a trait must fit this individual more to keep the money
prove advantageous in solving an evolution and not to disclose its existence to the
ary challenge (in this case, the challenge of group, however, because this member
procreation). This does not mean, however, would then be over-benefited in relation to
that the trait must produce its adaptive his or her fellow members.4 In a grave
result for every person. The sex drive is financial crisis, this member might even be
adaptive because the challenge of reproduc able to survive, whereas others in the
tion is met more effectively with it than group—or the group as a whole—perished.
without it. Moreover, an adaptation need Adaptations work in much the same way,
not produce its adaptive result in every to advantage the individual rather than
instance. Of course, very few instances of any group to which he or she belongs.
sexual intercourse result in pregnancy, rela Therefore, when it comes to situations in
tive to the number of times humans engage which an individual’s priorities conflict
in intercourse overall. Again, however, this with a group’s, evolved adaptations tend to
does not mean that the sex drive is not adap privilege the success of the individual over
tive (in fact, it would be quite maladaptive if that of the group.
every instance of intercourse did result in
pregnancy). Adaptations need only provide Summary. These principles illustrate the
advantages relative to their alternatives. application of TNS and related theories to
various aspects of human behavior. Indeed,
Adaptations operate at the individual level, the breadth of behaviors evolutionary theo
not the group or species level. Humans ries can explain is certainly one of their
belong to a number of important groups, strengths (see Floyd & Haynes, 2005).
including their families, social networks, Whereas many otherwise excellent theories
and professional networks that, in various used commonly in nonverbal communi
ways, help ensure their survival. In many cation research are limited to explaining
cases, therefore, what is beneficial to the specific behaviors or tendencies, such
group is beneficial to the individual as expectancy violations (Burgoon & Hale,
member, and vice versa. Often, however, 1988), adaptation (Andersen, 1985), or
individual and group priorities are in con arousal (Cappella & Greene, 1982), TNS
flict. Consider, for example, a communal and evolutionary psychology can be applied
living situation in which each person’s to multiple aspects of nonverbal behavior.
money and possessions are considered to In the second section of this chapter, I will
be the collective property of all group discuss the application of evolutionary
08-Manusov.qxd 6/30/2006 7:26 PM Page 144
144–––◆–––Factors of Influence
146–––◆–––Factors of Influence
This is just one example of how the logic primary emotions there are, and what they
behind TNS and evolutionary psychology are (Cornelius, 1996; Ekman, 1972; Izard,
could be used to link an emotion display 1977; Tomkins, 1962, 1963). Included on
to viability, fertility, or both. Of course, nearly every list of primary emotions, how
increases in vigilance and energy in the face ever, are fear, sadness, surprise, anger, and
of fear may seem to have a direct relation happiness or joy. Second, nonverbal com
ship with survival ability if the threat evok munication scholars have debated the
ing the fear is physical (e.g., being chased by extent to which these forms of evidence
an animal, being stalked by a criminal), but provide logical support for Darwin’s propos
their effect on survival may seem to be more itions (e.g., Ekman, 1992; Russell, 1994).
dubious if the threat is psychological (e.g., Some, such as Fridlund (1994) and Fridlund
facing impending exams or a visit to the and Russell (this volume), have offered
dentist). This may well be the case. To the alternative theoretic accounts. Although
extent that dealing successfully even with an analysis of these competing viewpoints
psychological stressors can contribute to is beyond the scope of this chapter, it
viability, however, even minutely or indi is important to note that scholars have
rectly (e.g., conquering one’s fear enough to not always agreed about the merits of
do well in an exam, which helps secure a the Darwinian perspective or the extent to
good grade, which contributes to the attain which the existing evidence supports it.
ment of gainful employment and access to
resources), the mechanisms that foster such Cross-Cultural Consistency in Emotion
abilities can be “selected for” through the Displays. As Ekman (1973) explained,
process of natural selection. Because nat Darwin’s (1872) proposition that emotion
ural selection operates on an exceedingly displays are innate suggests that they are
slow time scale (e.g., producing changes not the product of cultural socialization
over millennia), characteristics that confer and should, therefore, show more con
even seemingly inconsequential advantages sistency than variability from culture to cul
with respect to viability or fertility, and that ture.5 An impressive body of research has
are at least partially heritable, are likely to accumulated that addresses this hypothe
be selected for. sis with respect to facial emotion displays
With respect to the link between emo (arguably the most communicative of all
tion displays and their underlying emo nonverbal emotion displays; Knapp, 1978).
tional experiences, Darwin (1872) argued Some early investigations conducted by
that displays of the primary emotions, at researchers attempting to identify cultural
least, had evolved as a product of evolu variation in facial emotion displays, instead
tionary pressures and were, therefore, as provided evidence of beyond-chance consis
innate as the emotional experiences them tency in how people from different cultures
selves (Ekman, 1973). Several forms of interpret them. For example, Triandis and
evidence can be examined as sources of Lambert (1958) showed photos of an actress
support for Darwin’s (1872) proposition. expressing different emotions to American
In this section, I identify three such forms college students, Greek college students,
of evidence. Two preliminary observations and Greek villagers from the Island of
are warranted, the first of which is that Corfu; their comparisons indicated that
Darwin’s proposition referred to the pri “Greek subjects, even when they come from
mary or “chief” emotions, and contem very different populations, rate emotional
porary scholars disagree on how many expressions in the same way as American
08-Manusov.qxd 6/30/2006 7:26 PM Page 147
college students” (p. 323). Other investiga are relatively uninhibited in their expressions
tions have also found substantial cultural of emotions such as happiness, sadness, or
consistency, in cultures ranging from fear, whereas older individuals have a
American, Mexican, Japanese, Chinese, and greater capacity to control their emotion
Turkish (e.g., Cüceloglu, 1970; Dickey & displays in the service of politeness or other
Knower, 1941; Vinacke, 1949; Vinacke & social norms. If Darwin’s (1872) proposi
Fong, 1955; Matsumoto, this volume). tion regarding the innateness of emotion dis
Although these early studies provided plays is accurate, however, one should find
preliminary evidence for cross-cultural consistency in the forms of emotion display
consistency in facial displays of emotion, from infancy onward through the life cycle.
their contributions were constrained by To address this question, Darwin began
a number of methodological limitations. chronicling expressive behavior in his own
Work by a range of scholars has remedied child, and his observations were offered in
many of these shortcomings and demon his 1872 book and in a later essay titled
strated cross-cultural consistency in facial A Biographical Sketch of an Infant (1877).
displays of emotion across both literate His observations, although methodologi
and preliterate cultures (e.g., Eibl- cally compromised, indicated the presence
Eibesfeldt, 1972; Ekman, 1968; Ekman of at least seven distinctive facial emotion
& Friesen, 1971; Ekman, Friesen, & displays beginning as early as 45 days.
Ellsworth, 1972; Ekman, Sorenson, & Systematic research on the ontology of
Friesen, 1969; Izard, 1971). Their collective emotion displays was relatively scarce in
finding is one of consistency across a range the century following Darwin’s publica
of cultures with respect to how emotions tions, but several more contemporary stud
are conveyed (at least, facially). Although ies have addressed the enactment of facial
some studies have questioned the degree of emotion displays during infancy. Research
cross-cultural consistency (e.g., Elfenbein & on smiling, for instance, has indicated that
Ambady, 2002), and others have ques externally elicited smiles begin to appear
tioned its very existence (e.g., Fernández- around the end of the first month (Sroufe,
Dols & Ruis-Belda, 1995), much of the 1984), and that by the age of 12 weeks,
evidence supports the hypothesis of cross- infants have begun to smile selectively to
cultural consistency. newly mastered activities and familiar
persons (Lewis, Sullivan, & Brooks-Gunn,
Consistency in Emotion Displays of Infants 1985). Fox and Davidson (1988) also
and Noninfants. If the means of conveying found that 10-month-olds produced more
emotions are learned principally through Duchenne smiles (accompanied by contrac
socialization and enculturation, then it fol tion of the lateral orbicularis oculi) with
lows that newborn infants, being possessed their mothers than with strangers and that
neither of the cognitive capacity nor the only Duchenne smiles were associated
experience to be functionally affected by with left frontal EEG activity (hypothesized
enculturation messages, should display their to be an approach-related pattern). As
emotions differently than children, adoles Charlesworth and Kreutzer (1973) noted,
cents, and adults, the latter groups having there is little evidence of a one-to-one rela
been subjected to the influences of socializa tionship between smiling and a specific
tion. Of course, differences between infants class of external stimuli; rather, infant smil
and noninfants are evident in terms of the ing can be elicited via a variety of auditory,
ability to control emotion displays; infants visual, tactile, and kinesthetic stimuli. Most
08-Manusov.qxd 6/30/2006 7:26 PM Page 148
148–––◆–––Factors of Influence
relevant to Darwin’s proposition is the age hypothesis, which posits that merely engag
at which facial emotion displays begin to be ing in a facial display of emotion will initi
observed in infants; the earlier the displays, ate physiological changes consistent with
the less likely they are products of social that emotion (e.g., smiling makes one feel
ization or enculturation. happier).
Several emotion displays, in addition In an experimental test of the facial feed
to smiling, have been studied with infants, back hypothesis, Ekman, Levenson, and
including displays of interest (Sullivan & Friesen (1983) instructed scientists and pro
Lewis, 1989), surprise (Camras, 1988), fessional actors to enact facial displays of
disgust (Rosenstein & Oster, 1988), fear six emotions while their autonomic nervous
(Schwartz, Izard, & Ansul, 1982), and system arousal was measured. Importantly,
anger (Stenberg, Campos, & Emde, 1983). participants were told how to configure their
Also instructive is research on congenitally faces physically in particular ways. Rather
blind children, whose condition eliminates than being instructed to “display anger,” for
their ability to imitate visually observed instance, participants were asked to pull
expressive behavior in others. Early work their eyebrows down and together, press
by Freedman (1965) indicated that congen their lips together, raise their upper eyelids,
itally blind infants smiled at expected times etc. Each expression was held for 10 sec
and (by 6 months of age) for expected dura onds. Facial displays of anger, fear, sadness,
tions, leading Freedman to conclude that happiness, and surprise were accompanied
smiling is an innate, rather than learned, by increases in heart rate, whereas displays
behavior. Later work by Eibl-Eibesfeldt of disgust elicited a small decrease in heart
(1970) and Charlesworth (1970) further rate. Displays of anger, sadness, and happi
supported the parallels in expressions of ness also elicited increases in skin tempera
blind and sighted infants, adding credence ture, whereas displays of fear, surprise, and
to Darwin’s (1872) assertion regarding the disgust were accompanied by decreased skin
evolution of emotion displays. temperature.
These physiological responses to display
Physiological Correlates of Emotion ing emotion closely resemble the correlat
Display. A third form of evidence buttress es of experiencing emotion (see Levenson,
ing Darwin’s proposition derives from 1992). Hess, Kappas, McHugo, Lanzetta,
research showing that engaging in facial and Kleck (1992) also found that displaying
displays of emotion elicits physiological anger and happiness increased participants’
changes that mimic those associated with heart rates, whereas displaying peaceful
the experience of those emotions. As ness decreased it. Later experiments by
Levenson (1992) detailed, evidence from Levenson, Carstensen, Friesen, and Ekman
several laboratories has indicated distinc (1991) and Levenson, Ekman, and Friesen
tions between emotional experiences with (1990) replicated the findings of Ekman et
respect to their effects on autonomic ner al. (1983). Ekman (1989) also replicated
vous system arousal. If Darwin’s conjecture those results with a non-American culture,
regarding the evolution of emotion displays the Minangkabau of Sumatra, Indonesia (a
is correct, one would also expect to find fundamentalist Muslim, matrilineal society).
that emotion displays correspond to physi The findings that emotion displays are
ological profiles similar to those character judged consistently across cultures, begin to
izing their associated emotions. This is appear early in life, and are associated with
the principle behind the facial feedback specific physiological reactions indicative of
08-Manusov.qxd 6/30/2006 7:26 PM Page 149
the underlying emotions do not unequivo physical attractiveness with a high degree of
cally prove the validity of the evolutionary consistency. If beauty really were in the eye
explanation, but these forms of evidence of the beholder (a notion that dates back at
make it more difficult to argue that emotion least to the third century, BC; Rubenstein,
displays are merely social or cultural con Langlois, & Roggman, 2002), then evaluat
structions. A second active area of nonver ing attractiveness as a signal of reproduc
bal communication research in which tive potential would not be a reliable
evolutionary principles can be applied is that strategy. The second assumption is that the
of interpersonal attraction. links between markers of attractiveness and
reproductive potential are real, not socially
or culturally contrived. Just as the evolu
ATTRACTION tionary perspective focuses attention on
how facial displays of emotion might serve
It is difficult to overstate the importance the evolutionary end of survival (as in the
of attraction in human social interaction. example of the fear expression detailed
Although a number of qualities may lead above), it also focuses on the ways in which
one person to be attracted to another as a markers of physical attractiveness are hon
friend or potential romantic partner (e.g., est indicators of fitness, genetic quality,
personality, similarity, propinquity), a great health, fertility, and reproductive value. As
many markers of physical attractiveness are above, I will offer in this section several
nonverbal, having to do with personal phys forms of evidence that support and are
ical appearance (see Andersen, Guerrero, illustrative of the evolutionary perspective
& Jones, this volume). Thus, for nonverbal on attractiveness.
communication researchers, these compo
nents of attractiveness are of clear relevance. Intra- and Intercultural Consistency in
From an evolutionary perspective, phys Judgments of Attractiveness. The contention
ical attraction is important for one funda that “beauty is in the eye of the beholder”
mental reason: It is often the instigating suggests both within-culture and, especially,
force behind the development of reproduc between-culture variation in what people
tive relationships. Without attraction to find physically attractive in others. Such
motivate initial interaction, many poten variation does exist, particularly between
tially significant relationships might simply cultures; however, research on human
be forgone. The evolutionary approach to attraction has also identified a number of
understanding attractiveness therefore aspects of physical attraction that are
focuses on what aspects of physical appear remarkably consistent within and across cul
ance are judged to be attractive, and how tures. In a series of studies, for example,
attraction to these aspects of physical Cunningham, Roberts, Wu, Barbee, and
appearance is consequential with respect to Druen (1995) asked people of different races
reproductive success. This perspective pro to judge the attractiveness of Caucasian,
poses that humans (and many other species) Asian, Hispanic, and Black women and
are adaptively attuned to markers of physi found a correlation of .93 between racial
cal attractiveness as signals of fertility and groups in their attractiveness ratings. In their
reproductive potential. second study, the researchers found that
The causal links proffered by the evolu Taiwanese participants agreed strongly with
tionary approach rest on at least two impor the ratings of the other groups (r = .91) and
tant assumptions. First, people evaluate that the degree of exposure to Western
08-Manusov.qxd 6/30/2006 7:26 PM Page 150
150–––◆–––Factors of Influence
media did not moderate this effect. In their development, genetic problems such as
third study, Cunningham et al. found that recessive genes and homozygosity, and
Blacks and Caucasians agreed with each environmental stressors such as parasites,
other strongly on attractiveness ratings pollutants, or extreme temperatures, cause
(r = .94). Other studies have documented the individual to exhibit fluctuating asym
consensus in ratings of attractiveness between metry (FA), or deviations from exact sym
North Americans and South Africans, among metry on bodily features that tend to be
Chinese, Indians, and the English, between symmetrical (Møller & Pomiankowski,
Black and Caucasian Americans, and between 1993). FA is assessed by taking precise mea
Americans, Russians, and Ache Indians (see sures of physical features such as the length
Cross & Cross, 1971; Jackson, 1992; Jones, of the ear, the breadth of the elbow, or the
1996; Morse, Gruzen, & Reis, 1976; width of the feet on one side of the body
Thakerar & Iwawaki, 1979). and comparing them to the same measures
In a meta-analysis of the attractiveness taken from the opposite side. Larger dis
literature, Langlois et al. (2000) examined crepancies indicate higher FA. Only the
919 effect sizes from a sample of over 1,800 most genetically fit individuals can main
empirical studies focusing on the assess tain symmetry under conditions of develop
ment of physical attractiveness. Their prin mental stress; low FA therefore serves as a
cipal findings all supported the notions of marker of genetic quality (Møller, 1997).
intra- and intercultural consistency in how A number of studies verify direct rela
people rate others’ attractiveness. Specifically, tionships between symmetry and percep
Langlois et al. found within-culture agree tions of attractiveness. Grammer and
ments of .90 for ratings of adults’ attrac Thornhill (1994), for example, measured
tiveness and .85 for ratings of children’s the symmetry of adults’ faces and found
attractiveness. Cross-ethnic agreement was strong linear relationships between symme
.88 and cross-cultural agreement was .94. try and perceptions of the attractiveness
These findings contradict the “eye of the of those faces for both men and women.
beholder” hypothesis and support the Similarly, Hume and Montgomerie (2001)
contention that many signals of physical examined the symmetry-attractiveness rela
attractiveness are judged consistently both tionship by manipulating the amount of
within and between cultures. symmetry in facial photographs. They
found the same direct relationship between
Relationship of Attractiveness Markers to symmetry and attractiveness ratings (see
Reproductive Success. The evolutionary also Perrett et al., 1999).
explanation assumes not only that people Importantly, symmetry has also demon
evaluate attractiveness with relative consis strated direct effects on reproductive suc
tency but also that markers of attractiveness cess. Thornhill and Gangestad (1994)
have real connections with reproductive found, for instance, that symmetry was lin
success. In this section, I examine two phys early related to the number of sexual part
ical characteristics that (1) are associated ners that male and female college students
strongly and consistently with attractive reporting having had. This effect held after
ness and (2) have direct relationships with controlling for the confounding effect of
reproductive success. A first predictor of age (see also Gangestad & Thornhill, 1997).
facial and bodily attractiveness is symmetry, These results suggest that more symmetri
or the extent to which two sides of a body cal people have more sexual opportunities
mirror each other. During an individual’s than less symmetrical people. Moreover,
08-Manusov.qxd 6/30/2006 7:26 PM Page 151
Thornhill, Gangestad, and Comer (1995) exclusively during pregnancy and lactation
found a direct relationship between the man’s (Björntorp, 1987) and because a WHR of
symmetry and his partner’s likelihood of .70 or below corresponds to the absence of
achieving orgasm. This is significant for major obesity-related disorders, such as dia
reproductive success because female orgasm betes, carcinomas, or heart disease (Barbieri,
increases the proportion of her partner’s 1990). Indeed, research indicates that higher
sperm that is retained in her reproductive WHRs are associated with women’s diffi
tract, boosting the chances of pregnancy culties in becoming pregnant (Kaye, Folsom,
(see Baxter & Bellis, 1993). Importantly, Prineas, Potter, & Gapstur, 1990). In a
no other variables measured in the study study of 500 women attending a fertility
(including the man’s height, income, and clinic, for instance, Zaadstra, Seidell, van
sexual experience, and the couple’s ratings Noord, te Velde, Habbema, Vrieswijk, and
of their mutual love) predicted the woman’s Karbaat (1993) reported that women who
likelihood of achieving orgasm. had a WHR under .80 were more than twice
A second example, relevant for women, is as likely to get pregnant following 12
their waist-to-hip ratio (WHR), or the ratio rounds of artificial insemination as were
of waist width to hip width. Across cultures, women with ratios exceeding .80.
and even across time periods in Western cul Symmetry and WHR are but two
tures, a female WHR of approximately .70 examples of physical characteristics that
has been seen as maximally attractive (i.e., show both strong, reliable associations with
when the waist is about 70% as wide as attractiveness and direct relationships with
the hips; see Singh & Young, 1995). For reproductive success (others may include
example, Singh (1993) found that every sin physical stature, skin clarity, and natural
gle Miss America winner from 1923 to 1987 body odor). From the evolutionary perspec
had a WHR between .69 and .72 and that tive, therefore, one would contend that
the WHRs of Playboy centerfolds also these characteristics are attractive to people
ranged only from .68 to .71. Singh and Luis because attraction to these characteristics is
(1995) demonstrated that the preference for reproductively beneficial. I have argued
a .70 WHR was consistent across cultures. elsewhere (Floyd & Haynes, 2005) that
As with symmetry, WHR also has direct one of the great strengths of the evolution
effects on reproductive success, and research ary approach is its ability to explain multi
indicates that a WHR of .70 is the WHR ple aspects of behavior and to do so at a
of maximum female fertility. After puberty, level of explanation that often transcends
estrogens lead to the accumulation of body the social, cultural, or political influ
fat in the gluteofemoral region (thighs and ences that are often the focus of social
buttocks), whereas androgens (such as science research. Such a perspective allows
testosterone) promote fat accumulation researchers the breadth to investigate how
in the abdominal region. WHR therefore human communicative behaviors are
serves as a reliable index of the distribution related to pervasive, engrained motivations
of fat between the upper and lower body that are often independent of the effects of
and also of the relative proportions of gender, race, class, sexual orientation, or a
intra-abdominal and extra-abdominal fat host of other proximal influences. In this
(Deprés, Prudhomme, Pouloit, Tremblay, manner, human communicative behavior
& Bouchard, 1991). can be understood within a broad context
This is significant for fertility, both that is relatively unbound by time, locale,
because gluteofemoral fat is used nearly or policy, and is often even unrestricted to
08-Manusov.qxd 6/30/2006 7:26 PM Page 152
152–––◆–––Factors of Influence
humans. The evolutionary perspective may be the same. For example, Floyd and
allows one to understand how nonverbal Morman (2001) compared biological fathers
communicative behaviors fit within a larger and stepfathers in terms of the amount of
and broader picture than many other non affection they communicated to their sons.
verbal theories point to. The hypothesis that fathers give more affec
tion to biological sons than to stepsons can
easily be derived on the basis of inclusive
♦ Using TNS in Nonverbal fitness theory. One might arrive at the same
prediction, however, by reasoning that men
Communication Research
feel closer to their biological sons than to
stepsons because they have longer histories
To conclude this chapter, I offer three recom with biological sons.
mendations for researchers wishing to capital To test an evolutionary hypothesis fairly,
ize on the advantages of evolutionary theories the researcher must design the test in such a
in their study of nonverbal behavior. The first way that one explanation can be ruled out. If
is to conceptualize research questions in terms it is a difference in closeness that causes
of ultimate causality. That is, researchers fathers to be more affectionate with biologi
should consider the relationship that specific cal sons than stepsons, for instance, then the
communicative behaviors might have either to difference in affectionate communication
survival or procreation (or both). The links should fail to manifest (or be diminished) if
for some behaviors might be fairly obvious; the level of closeness is covaried out. In fact,
one can certainly identify how a behavior like Floyd and Morman (2001) found that the
flirting is related (eventually) to procreation, mean difference in affectionate communica
or how the fight-or-flight response to fear is tion was maintained even when closeness
related to survival. The links for other behav and other relational characteristics were con
iors may be less apparent but are there trolled for. Certainly, ruling out one rival
nonetheless (e.g., the reproductive benefits of hypothesis does not provide unequivocal evi
being attracted to symmetrical sexual part dence for another, but it strengthens one’s
ners). Researchers could apply evolutionary claim. By crafting research designs carefully
principles to the study of such diverse behav to rule out rival causes, researchers can test
iors as immediacy, deception, touch avoid explanations derived from evolutionary the
ance, and cosmetic use to identify their ories while simultaneously ruling out expla
implications for survival and procreation. nations offered by other perspectives.
Conceptualizing research questions in terms Finally, researchers must consider con
of ultimate causality requires researchers to text so as to avoid oversimplified hypothe
consider the survival and procreative purposes ses. Although the evolutionary paradigm
that a given behavior might serve. focuses on ultimate, rather than proximal,
Equally important is that researchers causes, the characteristics evolution selects
formulate hypothesis tests to rule out for in individuals are often manifested only
rival explanations. This is a fundamental in particular contexts. Some may be more
principle of the scientific design, but it is genetically prone than others to reciprocate
especially important for researchers study nonverbal aggression, for instance, but this
ing communication from an evolutionary doesn’t necessarily mean that such people
perspective because one can often deduce have higher mean levels of aggressive
the same prediction using evolutionary ness than others, because an aggressive
and nonevolutionary theories. Although context is required to bring the tendency
the explanations differ, the basic prediction to fruition. Careful consideration of the
08-Manusov.qxd 6/30/2006 7:26 PM Page 153
environment and social context will help integrations of nonverbal behavior (pp. 1–36).
scholars avoid oversimplified hypotheses. Hillsdale, NJ: Erlbaum.
Researchers must bear in mind that evolu Andersen, P. A., & Guerrero, L. K. (Eds.) (1998).
tionary adaptiveness matters, in terms of Handbook of communication and emotion:
Research, theory, applications, and contexts.
predicting behavior, but it is not the only
San Diego, CA: Academic Press.
thing that matters. Contextual influences,
Barbieri, R. L. (1990). The role of adipose tissue
such as the history of a relationship or the
and hyperinsulinemia in the develop
constraints of a social context must also be ment of hyperandrogenism in women. In
considered in order to avoid the fallacy R. E. Frisch (Ed.), Adipose tissue and reproduc
that evolutionary pressures are the only tion (pp. 42–57). Basel, Switzerland: Karger.
pressures that exert influence on interper Baxter, R. R., & Bellis, M. A. (1993). Human
sonal behavior. sperm competition: Ejaculate manipulation
by females and a function for the female
orgasm. Animal Behavior, 46, 887–909.
♦ Notes Bjorklund, D. F., & Pellegrini, A. D. (2002). The
origins of human nature: Evolutionary
developmental psychology. Washington,
1. This is true, at least, for sexually reproduc DC: American Psychological Association.
ing species; see Bjorklund and Pellegrini (2002). Björntorp, P. (1987). Fat cell distribution and
2. Not all characteristics are inherited; if a metabolism. In R. J. Wurtman & J. J.
woman loses her hearing because of an injury, Wurtman (Eds.), Human obesity (pp. 66–72).
this will not make her more likely to have babies New York: New York Academy of Sciences.
who are deaf. Burgoon, J. K., & Hale, J. L. (1988).
3. Of course, advances in reproductive tech Nonverbal expectancy violations: Model
nology, such as in vitro fertilization, have elimi elaboration and application to immediacy
nated intercourse as an absolute precursor to behaviors. Communication Monographs,
reproduction. Such technology was not available 55, 58–79.
to our hunter-gatherer ancestors, for whom the Buss, D. M. (1994a). The evolution of desire:
adaptive nature of the human sex drive would Strategies of human mating. New York:
have been selected. Basic Books.
4. Such a benefit does require that the others Buss, D. M. (1994b). The strategies of human
in the group be unaware of this member’s good mating. American Scientist, 82, 238–249.
fortune. As Cosmides and Tooby (1989, 1992) Buss, D. M. (1999). Evolutionary psychology:
have suggested, humans have evolved a type of The new science of the mind. Boston: Allyn
cheating detection mechanism by which they & Bacon.
notice, remember, and respond to perceived Camras, L. A. (1988, May). Darwin revisited: An
cheating on the part of others, so as to protect infant’s first emotional facial expressions.
their self-interests. Paper presented at International Conference
5. For brevity’s sake, I will use the term on Infant Studies, Washington, DC.
emotion displays in this section to refer to expres Cannon, W. (1929). Bodily changes in pain,
sions of primary emotions, including fear, sad hunger, fear, and rage: Research into the
ness, surprise, anger, and happiness. function of emotional excitement. New
York: Harper & Row.
Cappella, J. N., & Greene, J. O. (1982). A dis
♦ References crepancy-arousal explanation of mutual
influence in expressive behavior for adult and
infant-adult interaction. Communication
Andersen, P. A. (1985). Nonverbal immediacy Monographs, 49, 89–114.
in interpersonal communication. In A. W. Charlesworth, W. R. (1970). Surprise reactions
Siegman & S. Feldstein (Eds.), Multichannel in congenitally blind and sighted children.
08-Manusov.qxd 6/30/2006 7:26 PM Page 154
154–––◆–––Factors of Influence
156–––◆–––Factors of Influence
British Journal of Developmental Psycho Singh, D., & Luis, S. (1995). Ethnic and gender
logy, 3, 307–316. consensus for the role of waist-to-hip ratio on
Møller, A. P. (1997). Developmental stability judgment of women’s attractiveness. Human
and fitness: A review. American Naturalist, Nature, 6, 51–65.
149, 916–942. Singh, D., & Young, R. K. (1995). Body weight,
Møller, A. P., & Pomiankowski, A. (1993). waist-to-hip ratio, breasts, and hips: Role
Fluctuating asymmetry and sexual selec in judgments of female attractiveness and
tion. Genetica, 89, 267–279. desirability for relationships. Ethology and
Morris, R. (2001). The evolutionists: The strug Sociobiology, 16, 483–507.
gle for Darwin’s soul. New York: W. H. Sroufe, L. A. (1984). The organization of emo
Freeman. tional development. In K. R. Scherer &
Morse, S. T., Gruzen, J., & Reis, H. (1976). The P. Ekman (Eds.), Approaches to emotion
“eye of the beholder”: A neglected variable (pp. 109–128). Hillsdale, NJ: Erlbaum.
in the study of physical attractiveness. Stenberg, C., Campos, J., & Emde, R. (1983).
Journal of Personality, 44, 209–225. The facial expression of anger in seven
Nesse, R. (1990). Evolutionary explanations of month old infants. Child Development, 54,
emotions. Human Nature, 1, 261–289. 178–184.
Perrett, D. I., Burt, D. M., Penton-Voak, I. S., Sullivan, M., & Lewis, M. (1989). Emotion and
Lee, K. J., Rowland, D. A., & Edwards, R. cognition in infancy: Facial expressions
(1999). Symmetry and human facial attrac during contingency learning. International
tiveness. Evolution and Human Behavior, Journal of Behavioral Development, 12,
20, 295–307. 221–237.
Ploog, D. (1986). Biological foundations of Thakerar, J. N., & Iwawaki, S. (1979). Cross-
the vocal expressions of emotions. In cultural comparisons in interpersonal
R. Plutchik & H. Kellerman (Eds.), attraction of females toward males. Journal
Emotion: Theory, research, and experience of Social Psychology, 108, 121–122.
(Vol. 3, pp. 173–197). New York: Thornhill, R., & Gangestad, S. W. (1994).
Academic Press. Human fluctuating asymmetry and
Rosenstein, D., & Oster, H. (1988). Differential sexual behavior. Psychological Science, 5,
facial response to four basic tastes in new 297–302.
borns. Child Development, 59, 1555–1568. Thornhill, R., Gangestad, S. W., & Comer, R.
Rubenstein, A. J., Langlois, J. H., & Roggman, (1995). Human female orgasm and mate
L. A. (2002). What makes a face attractive fluctuating asymmetry. Animal Behavior,
and why: The role of averageness in defin 50, 1601–1615.
ing facial beauty. In G. Rhodes & L. A. Tomkins, S. S. (1962). Affect, imagery, con
Zebrowitz (Eds.), Facial attractiveness: sciousness: Vol. 1. The positive affects.
Evolutionary, cognitive, and social perspec New York: Springer.
tives (pp. 1–33). Westport, CT: Ablex. Tomkins, S. S. (1963). Affect, imagery, con
Russell, J. A. (1994). Is there universal recogni sciousness: Vol. 2. The negative affects.
tion of emotion from facial expression? New York: Springer.
Psychological Bulletin, 115, 102–141. Tooby, J., & Cosmides, L. (1990a). On the
Schwartz, G., Izard, C. E., & Ansul, S. (1982, universality of human nature and the uni
May). Heart rate and facial response to queness of the individual: The role of genet
novelty in 7- and 13-month-old infants. ics and adaptation. Journal of Personality,
Paper presented at International Conference 58, 17–68.
on Infant Studies, Austin, TX. Tooby, J., & Cosmides, L. (1990b). The past
Singh, D. (1993). Adaptive significance of explains the present: Emotional adapta
waist-to-hip ratio and female physical tions and the structure of ancestral envi
attractiveness. Journal of Personality and ronments. Ethology and Sociobiology, 11,
Social Psychology, 65, 293–307. 375–424.
08-Manusov.qxd 6/30/2006 7:26 PM Page 157
Triandis, H. C., & Lambert, W. W. (1958). national-race groups in Hawaii: II. Oriental
A restatement and test of Schlosberg’s faces. Journal of Social Psychology, 41,
theory of emotion with two kinds of 185–195.
subjects from Greece. Journal of Abnormal Wallace, A. R. (1858). On the tendency of
and Social Psychology, 56, 321–328. varieties to depart indefinitely from the
Trivers, R. L. (1971). The evolution of recipro original type. Journal of the Proceedings
cal altruism. Quarterly Review of Biology, of the Linnean Society (Zoology), 3,
46, 35–37. 53–62.
Vinacke, W. E. (1949). The judgment of facial Zaadstra, B. M., Seidell, J. C., van Noord, P. A.
expressions by three national-race groups H., te Velde, E. G., Habbema, J. D. F.,
in Hawaii: I. Caucasian faces. Journal Vrieswijk, B., et al. (1993). Fat and female
of Personality, 17, 407–429. fecundity: Prospective study of effect of
Vinacke, W. E., & Fong, R. W. (1955). The body fat distribution on conception rates.
judgment of facial expressions by three British Medical Journal, 306, 484–487.
08-Manusov.qxd 6/30/2006 7:26 PM Page 158
09-Manusov.qxd 7/10/2006 3:23 PM Page 159
9
PERSONALITY AND
NONVERBAL BEHAVIOR
A Complex Conundrum
� Robert Gifford
University of Victoria
◆ 159
09-Manusov.qxd 7/10/2006 3:23 PM Page 160
160–––◆–––Factors of Influence
used, with permission, from the above Web site for future research that incorporates the
of Ross Woodrow.) complexities I detailed, and discuss an
exemplar study that reflects this paradigm.
It is possible to infer character I begin, however, with a brief history of
from features, if it is granted that scholarship connecting nonverbal cues
the body and the soul are changed to personality.
together by the natural affections.
that she was a liar, stubborn, a gifted Nevertheless, the Allport-Vernon book
abstract thinker with an IQ above average, did not stimulate much new published
greedy, an unreliable friend, and a sneak; research over the next three decades; only a
that she would punish her enemies by any few scattered studies may be found from
means; and that she needed at least 9 hours the 1930s until the early 1960s. Perhaps the
of sleep each night! Unlike a painting, how first modern study that would fit comfort
ever, human faces are dynamic, and they ably in this chapter was Ralph Exline’s
convey different dispositional impressions (1963) investigation of visual interaction in
depending on the person’s state. For groups of men and groups of women who
example, when different emotions are had been categorized in terms of their need
experienced by an actor, different disposi for affiliation. He found that need for affil
tions are inferred by observers (Montepare iation was related to mutual glances, but it
& Dobish, 2003). The first important worked differently for men and women.
scientific study relevant to the present Exline’s study has the further distinction of
chapter was summarized in Allport and recognizing that nonverbal behavior should
Vernon’s (1933) groundbreaking mono be examined within interacting groups,
graph. Allport and Vernon did not attempt rather than assuming implicitly that people
to relate nonverbal behavior to personality express themselves nonverbally without
dispositions. Rather, they sought to find reference to others—that is, always in the
unity (or something close to it) among the same way (see the physiognomists).
expressive movements of their subjects. In the four decades since the mid-1960s,
Their hypothesis appears to have been a number of scientific studies (although per
rooted in the Aristotelian proposition, haps fewer than one might expect) have
quoted at the start of the chapter, that one’s been conducted on the relations between
whole body and personality are a kind of nonverbal behavior and personality, but
unity in which every aspect is mirrored in progress has been slow and fitful. Twenty
every other aspect. This view was champi years after Exline’s (1963) study, reviewers
oned by German psychologists such as were forced to conclude that the expression
William Stern, who influenced Allport and of personality in nonverbal behavior “can
Vernon profoundly. The purpose of the not be said to be strongly supported by the
1933 monograph was to demonstrate the evidence” (Bull, 1983, p. 113) and that “in
existence of consistency among a person’s general, much of the research on personal
expressive movements, which would, in the ity correlates has shown . . . relatively
Aristotle-Stern sense, support the very con weak relationships to nonverbal behavior”
struct and existence of personality. When (Heslin & Patterson, 1982, p. 131).
Allport and Vernon were writing, personal Recognizing the complexity of nonver
ity was not a widely respected construct in bal behavior, however, Patterson (1995,
psychology, and it was therefore in need of this volume) developed his parallel process
empirical support. Allport and Vernon’s model, which posits that nonverbal behav
results showed promise, in that two clusters ior is a balance of behavior and social cog
of expressive movements, one “general” nitive judgment processes, often automatic
and one “specific,” were found, albeit with or overlearned, but sometimes under con
lower than desirable reliability. Their book trol and in the service of a particular goal.
was the basis for some optimism that per Plausible as it is, the theory also is difficult
sonality and nonverbal behavior could be to test empirically, as Patterson (this
studied profitably. volume) attests. This chapter, then, while
09-Manusov.qxd 6/30/2006 7:49 PM Page 162
162–––◆–––Factors of Influence
acknowledging the value of the parallel Although this problem is beyond the scope of
process model, focuses on the more modest, this chapter, suffice it to say there are prob
but more easily researched, social judgment lems with self-report measures as well as
aspect of communication. As will be seen, problems with ratings by significant others
there are plenty of complexities even within (see Funder, 2003; Kenny, 1994). The second
this limited portion of the whole. main problem involving accuracy is this:
Thus, at the beginning of the 21st How well do decoders detect the (true) level
century, considerable research remains to of a person’s dispositions from the others’
be done before the complex connections nonverbal behavior? In the typical study,
among personality, nonverbal behavior, and decoding accuracy is measured as the dis
inferences about personality made by crepancy between the decoder’s assessment
observers of others’ nonverbal behavior can of the disposition and the actor’s self-rating
be understood. This chapter summarizes (or the ratings of the actor’s significant
recent work and offers a paradigm that others). Given the uncertainty of self-ratings,
might accelerate progress in the social judg and even those of significant others (Kenny,
ment portion of the problem. The paradigm 1994), these assessments should not be
might be considered a subset of the ambi granted the status of truth or treated as the
tious framework proposed by Patterson criterion against which observers’ ratings are
(1995) for understanding nonverbal com measured. Nevertheless, they certainly have
munication in general. some face validity. They are, after all, the
views of the actors by the actors or by those
who know them well.
♦ Key Distinctions: If the premise that even these assessments
are fallible is accepted, it becomes unrea
Encoding, Decoding,
sonable to tarnish any lack of agreement
Accuracy, and Agreement
with these assessments on the part of other
observers as a lack of achievement or error.
Encoding is the outward, objective, visi The observer’s assessment has its own
ble manifestation of personal dispositions inherent value as the view of another
in nonverbal behavior. The fundamental person after watching the actor’s nonverbal
hypothesis of researchers in this area is that behavior, with the advantage of some per
valid encoding does occur. A crucial, and spective, detachment, and often, objectivity.
still largely unanswered, question remains, Thus, the two assessments of the disposi
however. How much encoding occurs and tion should be granted equal ontological
for which dispositions? Decoding is the use status, and it is preferable to characterize
by observers of nonverbal behavior to infer any difference between the two assessments
personal dispositions in others. The inter as a discrepancy rather than as observer
esting question in this area concerns the error. In sum, neither the self’s nor the
accuracy of that decoding by different kinds observer’s rating is valid necessarily, and
of observers, for different dispositions, under the neutral term agreement should be used
different conditions. The wild card in this rather than accuracy.
endeavor, however, is accuracy. How valid The decoding of personality is itself
are our inferences of personality from non fraught with problems. Assessments of a
verbal behavior? person’s dispositions vary with the type of
Accuracy, in this sense, is itself problem information given to the judge. For example,
atic in two important ways. The first issue when judges in an interview study either
concerns the validity of personality measures. saw a silent videotape of an interview with
09-Manusov.qxd 6/30/2006 7:49 PM Page 163
a manager (i.e., only the nonverbal behav that does not guarantee decoding validity
ior) or read a transcript of the same inter (Kenny, 1991), as exemplified by (inaccu
view (thus, no nonverbal behavior), the rate) stereotyping. The second complexity
correlations between their assessments of concerns the relevance of the disposition’s
the managers’ extraversion and conscien domain to the context in which the nonver
tiousness were r = .27 and r = .30, respec bal behavior occurs (e.g., examining inter
tively; that is, they shared about 9% of their personal traits in interpersonal contexts). A
variance (Motowidlo, Burnett, Maczynski, researcher should not expect conscientious
& Witkowski, 1996). Given that the over ness, for example, to be encoded strongly
lap between the assessments was quite low, in a casual conversation, but the same
they cannot both be accurate. Which assess researcher could expect extraversion to be
ment was more accurate, that based on encoded strongly in a conversation.
the manager’s words, as written, or that The third complexity involves the poten
based on the manager’s wordless, nonver tial interference with encoding of situa
bal behavior? tional factors. One such factor is the mutual
influence of interacting participants. Some
writers have assumed that the presence of
TEN COMPLEXITIES others will always affect the encoding for
any individual (e.g., Kanki, 1985), but the
With these issues in mind, a summary of effect of others can be tested with intraclass
current knowledge may be essayed. All the correlation analyses (e.g., Gifford, 1994).
following findings are subject to a blanket Sometimes the effect of others is found,
cautionary note, however, because most empirically, to be minimal. The fourth
studies have not dealt with all the complex complexity relates to another aspect of the
ities of encoding and decoding research. situation, the nature of the activity or inter
Specifically, I argue that researchers must action during which encoding is investi
manage (at least) 10 complexities in their gated. For example, individuals may not
research designs and analyses (i.e., Type I merely enact a given behavior (e.g., smiling)
or Type II errors in the relations between more or less frequently depending on
nonverbal behavior and personality that context (e.g., at a party vs. a funeral), but
may occur if a researcher relies solely on a their encoding (the correlation with a given
simple Pearson correlation or fails to take disposition, regardless of the frequency of
into account one or more of the accuracy or the act) may differ with the situation. In
agreement issues). Some complexities are a study that demonstrated this, disposi
familiar and some less so; some are easier to tional public self-consciousness was encoded
manage than others. in hand movements differently when partic
The first complexity is perhaps the most ipants were lying than when they were
obvious, but published studies that over telling the truth (Vrij, Akehurst, & Morris,
look it have appeared. All the measures 1997).
must have adequate internal consistency The fifth complexity is a matter of clear
and interrater reliability. Ordinarily, this reporting more than a problem in itself. For
should be at least .75, but .80 or better is example, in one study, neuroticism was cor
very desirable. Adjustments for error and related with touching the self more, fewer
apologies for lower levels of reliability expressive gestures, and more gaze aversion
sometimes are made, but they are not con (Campbell & Rushton, 1978). The first two
vincing. Although interrater reliability is encodings were based on a teacher’s rating of
a prerequisite for decoding validity, even the person’s neuroticism, however, whereas
09-Manusov.qxd 6/30/2006 7:49 PM Page 164
164–––◆–––Factors of Influence
the third encoding was based on a self-report not only are there cultural differences in the
measure of neuroticism. That is, the different amount or frequency of nonverbal behav
assessments of neuroticism related to differ iors (e.g., Hall, 1966; see Matsumoto, this
ent nonverbal behaviors. Such results should volume), but nonverbal behavior may also
be reported as based on different measures of differentially encode (correlate with) disposi
the disposition. In a study that illustrated the tions in different cultures (e.g., Andersen &
problem of considering self-ratings and rat Guerrero, 1998).
ings by others to be equivalent, self-report
measures of emotional expressiveness yielded Summary. To learn whether nonverbal
different relations to a disposition (neuroti behavior truly encodes personality disposi
cism again) than did rated behavioral assess tions, researchers must navigate at least 10
ments of emotional expressiveness (Riggio & design and analysis complexities: (1) True
Riggio, 2002). encoding can be obscured through the use of
A sixth complexity is that relations unreliable measures; (2) encoding should be
between personality and nonverbal behav studied in a context in which the disposition
ior can differ with different combinations is salient; (3) others in the interaction might
of traits. For example, individuals who are influence an individual’s encoding; (4) encod
shy and sociable avert their gaze more and ing might occur differently when a person is
engage in more self-manipulation than engaged in different activities or purposes; (5)
others (Cheek & Buss, 1981), but this is not encoding may depend on who (e.g., self or
true of the other combinations (e.g., shy but significant others) assesses the disposition;
not sociable persons). The value of the (6) nonverbal behavior may encode combi
Cheek and Buss study lies in its demonstra nations of dispositions without encoding
tion that combinations of dispositions that combination’s constituent dispositions;
sometimes reveal more about encoding (7) combinations of nonverbal behaviors may
than individual dispositions do. A seventh encode dispositions without the individual
complexity is that dispositions can be behaviors doing so; (8) encoding may depend
encoded by a group of behaviors without on the gender composition of the group; (9)
any particular behavior doing so (Aries, encoding sometimes differs for male and
Gold, & Weigel, 1983). Sometimes a pat female individuals; and (10) cultural groups
tern or profile of nonverbal behaviors must vary in their encoding patterns.
be measured before significant encoding
(in their case, dominance) can be detected.
Eighth, in the same study, the encoding ♦ Encoding and
of dominance occurred only in same-sex
Decoding Personality
groups, not in mixed-sex groups. Thus,
encoding may be different depending on the
sex composition of an interacting group. Keeping these complexities in mind, I turn
The ninth form of complexity is also related now toward the research focused on encod
to sex: A given encoding relation may be true ing and decoding personality. An example
for one sex but not for the other. For example, of relatively straightforward encoding
extraversion correlates strongly with the use of results comes from a study of interacting
broad gestures among women but not among female dyads (Berry & Hansen, 2000). In
men (Lippa, 1998). The tenth complexity is Big Five personality terms, more agreeable
that personality is encoded by nonverbal women gestured more, used more open
behavior differently across cultures. That is, body postures, visually attended to their
09-Manusov.qxd 6/30/2006 7:49 PM Page 165
interaction partner more, used fewer visual mouth as shy, and those who smile less as
dominance behaviors, and displayed fewer too serious (Ferrari & Swinkels, 1996).
negative facial expressions than did less Are these accurate assessments of a tar
agreeable women. Women who were more get person’s personality or mere “decoding
open to experience visually attended to errors” (Bull, 1983)? Observers appear to
their interaction partners more than those decode confidently and with greater con
who were less open to experience. More sensus (Gifford, 1994; Lippa & Dietz,
extraverted persons seem to use more ani 2000), but the evidence that they do so
mated, expressive, and animated gestures— accurately is mixed or even discouraging,
that is, faster and more energetic gestures as shown, for example, by the Cleeton and
using the hands farther from the body Knight study. On the positive side, some
(Lippa, 1998)—than more introverted research shows that if one is willing to define
persons. Children with more internal, accuracy as observer agreement with target
rather than external, locus of control ten self-assessments, removing nonverbal behav
dencies smile more and engage in fewer off- ior from a job interview (by conducting
task activities (Carton & Carton, 1998). it by telephone, as opposed to in person)
Individuals with avoidant attachment styles reduces accuracy (Blackman, 2002). Thus,
tend to choose larger interpersonal distances nonverbal behavior certainly can contribute
(Kaitz, Bar-Haim, Lehrer, & Grossman, to accurate judgments.
2004), as do those with greater trait anxiety A variant on the study of decoding is the
(e.g., Patterson, 1973) and weaker affiliative study of decoding ability, sometimes called
tendencies (e.g., Mehrabian & Diamond, nonverbal sensitivity (e.g., Rosenthal, 1979;
1971). see Riggio, this volume). Decoding as a skill
Whether or not dispositions are encoded related to the judge’s own experience and
in nonverbal behavior, decoders believe background is often applied to constructs
they are. In an early study, personnel man other than personality (i.e., emotion; Mullins
agers were quite confident that job applica & Duke, 2004). More intelligent judges are
tion photographs revealed the applicants’ also more accurate (Lippa & Dietz, 2000),
character (Viteles & Smith, 1932). Observers’ at least for some dispositions: More intelli
ratings may be reliable, which suggests gent university-student judges assessed
accuracy, but they often do not correlate dispositional extraversion and an omnibus
with any of the targets’ physical features (across dispositions) measure more accu
(e.g., Cleeton & Knight, 1924). Researchers rately than less intelligent university-
still investigate alleged nonverbal “power student judges. On the other side of the
codes” (Schwartz, Tesser, & Powell, 1982) lens, which dispositions are easiest to
and the “shared meaning” of postures decode from nonverbal behavior? Several
(Kudoh & Matsumoto, 1985). “High studies (e.g., Ambady, Hallahan, & Rosenthal,
persuasive” nonverbal behavior patterns in 1995; Borkenau & Liebler, 1992; Gifford,
actors (direct gaze, more gestures, fewer 1994; Lippa & Dietz, 2000) report that
self-touches) are judged to be more sociability or extraversion is the most legi
assertive, forceful, powerful, and intelligent ble or accurately discernable disposition.
(Hart & Morry, 1997). Individuals who This, however, may be a function of the
speak in a tight-lipped manner or who turn second complexity mentioned previously.
their heads while speaking may be judged Most studies use conversations as the activ
as “uptight,” those who speak with a hand ity, and extraversion is particularly salient
over their mouths or smile with a closed for conversations.
09-Manusov.qxd 6/30/2006 7:49 PM Page 166
166–––◆–––Factors of Influence
Head Behaviors
Trunk Behaviors
Self-Rated Observer-Rated
Disposition Disposition
Arm and
Hand Behaviors
Leg Behaviors
Ecological Validity Cue Utilization
Achievement
and decoding occur. Few studies have done nomothetic goals (“How do people
what seems most productive, however: (1) decode?”), the study combines the ratings
to investigate all three processes; (2) to of “everyone”—parents, supervisors,
assess the relative strengths of encoding, friends, partners, peers, or members of
decoding, and agreement; and (3) to take other cultures—as the observers of interest.
the context into account to provide some Of course, if observers in general or from a
understanding of how nonverbal behavior particular group use nonverbal cues idio
communicates (and miscommunicates) per syncratically, the interrater reliability of
sonality. Some other notable exceptions their target disposition ratings will be low,
that focus on nonverbal behavior and dis and it will be inappropriate to correlate
positions include those by Borkenau and their ratings with the nonverbal behavior
Liebler (1992) and Lippa (1998). scores (decoding correlations) or with the
The paradigm is employed, in part, to targets’ self-ratings (agreement correla
understand the cue utilization policies of tions). Thus, studies with any sort of nomo
observers, individually or in aggregate. Some thetic goals depend on the idea, and
early studies focused on individual abilities, therefore must hypothesize, that
such as those of clinicians (e.g., Hoffman, a group of observers will reliably agree on
1960), and found that their judgments, as actors’ dispositions. If a specified group of
revealed through their use of cues, did observers do not agree, then conclusions
not match well with their own impressions about their cue-utilization policies cannot
of how they use those cues. A more be stated, probably because members of
recent individual-level focus has been on that group do not use the same cues.
the ability or sensitivity of individual The earliest study that used this para
observers (e.g., Rosenthal, 1979). When digm was one by Brunswik himself in 1945
a researcher has more aggregate, (but not reported until later; Brunswik,
09-Manusov.qxd 6/30/2006 7:49 PM Page 168
168–––◆–––Factors of Influence
1956, pp. 26–29). The central cues he because the actors in the study were
employed were, perhaps appropriately engaged in a conversation; other disposi
enough for a first and early study, physiog tions (e.g., conscientiousness, openness to
nomic: actor’s height of forehead, length of experience, and emotional stability, the
nose, and so on. A decidedly verbal study other three dispositions in the Big Five)
using the paradigm to study extraversion in were not included because they would
relation to vocal behavior was published not have been examined in a context that
two decades later (Scherer, 1978). In one should have made them particularly salient.
study that fulfilled most of the goals of the
proposed paradigm, behavioral cues were
examined as mediators of the encoding- ♦ Interlude: Potential Outcomes
decoding process (Borkenau & Liebler,
of Encoding-Decoding Studies
1992). The same judges served as raters of
the physical cues and as decoders, however,
which compromised the independence of Before describing the exemplar study in detail,
the behavior scores and trait ratings. it may be useful to discuss the generic
Perhaps the first study that examined potential outcomes of studies that use this
nonverbal behavioral mediators and used paradigm. The first assumption is that all
behavior scorers who were independent of the judgments (e.g., self-ratings, behavior
both targets and observers was conducted by scoring, and observer ratings) are reliable;
Gifford, Ng, and Wilkinson (1985). That if some are not, they cannot be used with
study identified nonverbal cues exhibited by any pretence of validity. In general, encod
job applicants that mediated (and failed to ing, decoding, and agreement may be weak
mediate) agreement between job applicant or strong for any disposition, and the
and personnel officer assessments of the pattern of results may be different for each
applicant’s social skill and motivation to disposition.
work. That study, however, did not employ The first type of potential outcome
personality dispositions as the psychological occurs when, for a given disposition, encod
constructs. A subsequent study, with a coor ing, decoding, and agreement all are weak.
dinated set of independent targets, raters, In this case, (1) personality is not reflected
and observers, that did investigate nonverbal consistently in nonverbal behavior (at least
behaviors and how they encode dispositions not in the behaviors studied), (2) observers
as well as which of these same nonverbal do not use this set of behavior cues to arrive
behaviors are employed by observers to infer at their inferences, and (3) observer infer
targets’ personality finally appeared just over ences do not agree with the self or knowl
a decade ago (Gifford, 1994). edgeable other assessments of actors.
Gifford (1994) will be described in some Second, if decoding is strong but encoding
detail as an exemplar of the proposed para is weak, observers apparently are using
digm. It examined the eight dispositions invalid stereotypes. (One suspects, without
that comprise the interpersonal aspect of the benefit of data, that this was the case
personality and form a circumplex (Wiggins, with Lavater and his fellow physiogno
1979). The primary axes of the circumplex mists.) Agreement should be weak in such a
(dominance and warmth) are interpreted by case, because there are no true relations
some theorists as two of the Big Five per between personality and nonverbal behav
sonality domains (McCrae & Costa, 1989; ior for observers to decode legitimately.
Peabody & Goldberg, 1989). Interpersonal Third, if strong encoding but weak
dispositions were selected deliberately decoding is found, observers are unable to
09-Manusov.qxd 6/30/2006 7:49 PM Page 169
deduce correctly which nonverbal cues same objective (visible) aspects of reality as
reflect the actors’ personality. The potential encoding. If so, the centuries-old conviction
for strong agreement is present, but it is that dispositions truly are “legible” would
unrealized. Fourth, if agreement is strong receive convincing support.
but both encoding and decoding are weak,
observers must be using nonverbal behav
iors for decoding that the researcher has not ♦ An Exemplar
measured. Some nonverbal cue or other
must have been providing valid information
about the actor’s personality, or agreement The following study is presented partly for
would not be possible. The researcher must its results, which illustrate many specific
explore the impression formation process, encoding and decoding results with per
perhaps through interviews with observers, sonality and nonverbal behavior. But it is
to learn which unstudied nonverbal cues the also presented as a way of introducing
observers might have been using to succeed many of the intricacies of conducting
in matching the assessments of the actors. encoding-decoding research, including pro
Fifth, if weak encoding and high agree posed solutions to problems that arise in
ment are found, observers again must be the course of analyzing the data in such
using valid but unmeasured nonverbal cues, studies. Based on the available literature,
unless the unlikely case that the observers the hypotheses of the exemplar study
are clairvoyant holds true (Reichenbach, (Gifford, 1994) were that encoding would
1938). As Wiggins (1973) wryly notes, be weak to moderate but that observers
“Such a possibility is assigned rather low would have a strong tendency to decode.
priority as a contemporary scientific expla Agreement, based on recent research that
nation” (p. 159). This is a case in which indicates dispositions are communicated
researchers must rethink their choice of cues, to different degrees (Gifford et al., 1985;
seeking others that do encode the disposi John, 1990), was expected to vary across
tion. One way to accomplish this might be dispositions. For dispositions with low
to interview the judges, asking them to agreement, self-observer encoding-decoding
reflect on their inferences. discrepancies were expected to be high.
Finally, if strong encoding, strong For dispositions with high agreement, self-
decoding, and strong agreement are found, observer discrepancies were expected
the researcher may conclude that the whole to be low.
process is working as researchers in this The target participants were 60 under
area dream it does, and they may be able to graduates drawn from a psychology depart
supply a satisfying account of this assess ment participant pool. Ten all-male and 10
ment process. A sober second thought, all-female triads were formed into conver
however, is that actors (or their intimates) sational groups, and one group at a time
and observers could be agreeing on an was filmed as it conversed. The participants
inaccurate view of the actors’ personality, were given a list of suggested topics, but
something akin to a folie à deux. A more they were encouraged to converse on any
likely interpretation is that the strong medi topic they chose. A week or so prior to the
ation of objective nonverbal behaviors, conversation, the participants were given
assessed reliably by independent observers, Wiggins’ (1979) Interpersonal Adjective
would be substantial evidence that the Scales Inventory (IAS). The IAS covers two
observers’ decoding is valid, given that of the Big Five domains of personality, but
they have been demonstrated to rely on the it was chosen to maximize the relevance of
09-Manusov.qxd 6/30/2006 7:49 PM Page 170
170–––◆–––Factors of Influence
selected dispositions to the context. From between a nonverbal behavior cue and a
the top of the circumplex, these scales are disposition, however, necessarily is a valid
ambitious-dominant, gregarious-extraverted, encoding link.
warm-agreeable, unassuming-ingenuous, lazy- Three specific threats to the validity or
submissive, aloof-introverted, cold-quarrel generalizability of an encoding link may
some, and arrogant-calculating (see Gifford & be identified. First, the correlation could be
O’Connor, 1987). influenced by the actions of others in the
The videotapes were then scored using conversation; a valid encoding link should
the Seated Kinesic Activity Notation System be empirically attributable to an individual,
(SKANS IV; Gifford, 1986), in which 38 uncontaminated by group influence, if it
kinesic and facial behaviors are measured in is to be considered a valid personality-
one of three ways: frequency, duration, or nonverbal behavior link. Second, correla
time sampling. In a second sample of partic tions may be due to chance; to be valid, an
ipants, 21 unacquainted peers of the targets encoding link should have reasonable
were shown 5-minute selections from the strength and be part of an ordered pattern
middle of the conversations over several ses of correlations around the interpersonal
sions with the audio track turned off. Each circle. If a behavior is truly relevant to inter
time the tape was played, each observer was personal behavior, it should not merely
asked to focus on only 1 of the 3 partici correlate with one disposition on the circle.
pants shown in the tape. The tape was then Its correlations should rise and fall around
replayed and the observers watched another the interpersonal circle in an ordered man
participant. Thus, all 21 observers viewed ner (Gifford, 1991). Third, the possibility
all 60 participants. After each tape was of sex differences raises the issue of general
shown, each observer completed a 40-item izability of a given putative encoding link
short version of the IAS about one target. to both sexes. For example, a valid link
The observers’ task was demanding, so they between a disposition and a nonverbal
made their ratings over several sessions. behavior for women may not be valid for
They were paid $50 for their efforts and men, or vice versa. For example, using most
offered a prize of $50 for being the most of one’s body when gesturing validly signals
accurate (defined as coming the closest to extraversion for women, but it does not
the self-ratings of the 60 target individuals— work as a valid signal for men (Lippa, 1998).
really, as noted, a measure of agreement). Each of these threats was considered in
Most measures (self-assessments, observer preliminary analyses (for details, see
assessments, and SKANS IV measures) Gifford, 1994). In all, because of significant
were adequately reliable. Some behaviors, group influence or failure to conclusively
however, occurred infrequently; they were map onto the interpersonal circle, 19 of the
difficult to score owing to camera place 27 remaining nonverbal behaviors were
ment; or interjudge agreement was low. rejected as not demonstrably valid encoders of
Others were combined because they were interpersonal dispositions. The eight nonver
highly correlated. Thus, the remaining bal behaviors identified as valid encoders of
analyses were based on 27 nonverbal interpersonal circle traits are head orienta
behaviors. Pearson correlation coefficients tion, nods, arm wrap, gestures, object
between the actors’ self-assessed disposi manipulation, left leg lean, leg movement,
tions and their nonverbal behaviors repre and leg extension. Their significant links
sent the left, or encoding, half of the lens (p < .05) with the eight dispositions of the
diagram. Not every significant correlation interpersonal circle are displayed in one
09-Manusov.qxd 6/30/2006 7:49 PM Page 171
Gregarious-Extraverted
Ecological Validity: Cue Utilization:
Multiple R = .41 Multiple R = .80
R2 = .17 R2 = .65
Head Orientation
.39
Head Recline
–.40***
Nods
*
.28
.27*
Head Shakes .52***
Smiles .32**
.26*
Trunk Recline
Extraverted– –.30**
–.46*** Extraverted–
Arm Wrap
Gregarious .41*** .62*** Gregarious
(Self) Gestures –.24* (Others)
Left Hand Vertical .39**
Hand Extension .35**
.27*
Right Leg Orientation .23*
Leg Movement
Achievement = .45
172–––◆–––Factors of Influence
PARTICULAR ENCODING
GROUP VERSUS
AND DECODING LINKS
INDIVIDUAL DECODING
For readers who are interested, consider
Despite these findings, however, decod
able information about particular relations
ing actually may not be much stronger than
between interpersonal dispositions and
encoding. As noted, observer decoding rat
nonverbal behaviors is available in Figures
ings are based on 21 raters. Multiple raters
3 to 10 of Gifford (1994). Here, only one
almost necessarily increase the reliability
lens model is presented, as an example, in
of ratings. When ratings are more reliable,
the interest of saving space.
correlations involving them are stronger
because less error is involved. Stronger cor
relations are more likely to be statistically AGREEMENT AND NONVERBAL
significant and therefore to be included in COMMUNICATION
the lens diagrams. Analyses that corrected
for attenuation and estimated the reliability Agreement is measured as the correla
of single judges (see Gifford, 1994, for tion between typical self-assessments and
details) showed that one typical decoding typical assessments by observers, and it
link shrank from r = −.58 to r = −.35. The is represented in Figures 9.1 and 9.2 by
matched encoding link for this decoding the curved line at the bottom. The use
link was r = −.29, not much less than of correlations overcomes several of the
r = −.35. Thus, observers as a group decode classic Cronbach (1955) criticisms of accu
strongly, but researchers who wish to racy research. Across the eight dispositions
09-Manusov.qxd 6/30/2006 7:49 PM Page 173
in the full study, agreement averaged .27 necessary. As noted earlier, however, the
(r− to Z-transformed), which is significant greater number and magnitude of decoding
(p < .02), if moderate in magnitude. links is due partly to psychometric considera
Agreement ranged from .18 (NS) for both tions—that is, the superior reliability of
lazy-submissive and cold-quarrelsome to decoding. Agreement is higher in general
.45 (p < .001) for gregarious-extraverted when there are more matched links. The exis
and .41 (p < .001) for aloof-introverted. tence of matched links, with their lines going
The relations between encoding and from the disposition to a behavior and from
decoding fall into two categories, each with the behavior to the observer’s assessment,
two forms. First, matched links may be clearly suggests that agreement increases when
identified. One form of matched link occurs information “flows” via such matched links.
when a nonverbal behavior significantly Conversely, agreement is lower when many
encodes self-assessments and is also used to mismatched links occur. The same trend was
a significant degree by observers to decode also demonstrated earlier in a personnel selec
or infer that self-assessment. Across the 8 tion context by Gifford et al. (1985).
dispositions, 14 matched links of this form When information does not flow, either
were found. Matched links are underlined encoding has not occurred (no behaviors
in Figure 9.2. Another form of matched link measured encode the disposition) or the
occurs when a link is significant on neither observer has used cues other than those
side of the lens: Observers are saying that that the encoding analysis suggests are
a given behavior does not encode a given valid indicators of a disposition. The com
trait, and based on the self-assessments, it munication of self-assessed personality was
does not. In this study, 105 such matches quite good (i.e., agreement was relatively
occurred. Second, mismatched links may be high) for some dispositions. Considering
identified. One form of mismatched link that observers saw only 5 minutes of a
occurs when a nonverbal behavior does soundless conversation among individuals
encode a self-assessed disposition, but they had never met, their decoding of gre
observers do not use that cue. For example, garious-extraverted, r = .45, and aloof-
more lazy-submissive persons manipulate introverted, r = .41, for example, is quite
objects (e.g., their clothing, pen, paper) an “achievement.” This study demon
more than others do, but observers do not strates that (1) the encoding of interper
use object manipulation as a cue to lazy- sonal dispositions in nonverbal behavior is
submissive. In this study, six mismatches of moderate (median multiple R = .41), (2)
this kind occurred across the eight interper the decoding of the same dispositions is
sonal dispositions. The other form of mis moderate by individual observers and
matched link occurs when observers use a strong by groups of observers (median
particular nonverbal cue to form their multiple R = .80), and (3) agreement is low
impression but that cue does not encode to moderate (mean r = .27), yet significant
that disposition. For example (see Figure (p < .02). Each of these results will be dis
9.2), observers believe that more gregari cussed in turn. Apart from the specific
ous-extraverted persons orient their heads magnitudes of these links, the combined
toward their companions more, but (based findings show exactly how information
on self-assessments) this is not so. appears to flow from the actor to the
In this study, 83 mismatches of this kind observer—that is, how personality and non
were observed across the eight dispositions. verbal behavior are connected, and how
Decoders appear to use many more cues to observers infer (and misinfer) personality
infer self-assessed dispositions than were by watching other people.
09-Manusov.qxd 6/30/2006 7:49 PM Page 174
174–––◆–––Factors of Influence
176–––◆–––Factors of Influence
Table 9.1 A Process Model for the Encoding and Decoding of Nonverbal Behavior in
Personality
toward a fuller understanding, which future Borkenau, P., & Liebler, A. (1992). Trait infer
researchers will achieve, of both social judg ences: Sources of validity at zero acquain
ment and the delicate behavioral dance tance. Journal of Personality and Social
involved in nonverbal behavior (Patterson, Psychology, 62, 645–657.
Brunswik, E. (1956). Perception and the represen
this volume). Table 9.1 is an effort to create
tative design of psychological experiments.
a process model of the problem, one that is
Berkeley, CA: University of California Press.
adapted from Craik (1968), who used such
Bull, P. (1983). Body movement and interper
a model for another area of research, envi sonal communication. New York: Wiley.
ronmental perception. Even this does not Campbell, A., & Rushton, J. P. (1978). Bodily
encompass the range of considerations that communication and personality. British
should be taken into account in any serious Journal of Social and Clinical Psychology,
scientific study of personality and nonverbal 17, 31–36.
behavior, but perhaps other observations Carton, J. S., & Carton, E. E. R. (1998).
made in this chapter will contribute to that Nonverbal maternal warmth and children’s
goal. locus of control of reinforcement. Journal
of Nonverbal Behavior, 22, 77–86.
Cheek, J. M., & Buss, A. H. (1981). Shyness and
♦ References sociability. Journal of Personality and Social
Psychology, 41, 330–339.
Cleeton, G. U., & Knight, F. B. (1924). Validity
Allport, G., & Vernon, P. (1933). Studies of character judgments based on external
in expressive movement. New York: criteria. Journal of Applied Psychology, 8,
Macmillan. 215–231.
Ambady, N., Hallahan, M., & Rosenthal, R. Craik, K. H. (1968). The comprehension of the
(1995). On judging and being judged accu everyday physical environment. Journal
rately in zero-acquaintance situations. Journal of the American Institute of Planners, 34,
of Personality and Social Psychology, 69, 29–37.
518–529. Cronbach, L. J. (1955). Processes affecting scores
Andersen, P. A., & Guerrero, L. K. (1998). The on “understanding of others” and “assumed
bright side of relational communication: similarity.” Psychological Bulletin, 52,
Interpersonal warmth as a social emotion. 177–193.
In P. A. Andersen & L. K. Guerrero (Eds.), Duncan, S. F., Jr., & Fiske, D. W. (1985).
Handbook of communication and emotion: Interaction structure and strategy. New
Research, theory, applications, and contexts York: Cambridge University Press.
(pp. 303–329). San Diego: Academic Press. Exline, R. V. (1963). Explorations in the process
Aries, E. J., Gold, C., & Weigel, R. H. (1983). of person perception: Visual interaction in
Dispositional and situational influences relation to competition, sex, and need for
on dominance behavior in small groups. affiliation. Journal of Personality, 31, 1–20.
Journal of Personality and Social Psychol Exline, R. V., Ellyson, S. L., & Long, B. (1975).
ogy, 44, 779–786. Visual behavior exhibited by males differ
Berry, D. S., & Hansen, J. S. (2000). Personality, ing as to interpersonal control orientation
nonverbal behavior, and interaction quality in one- and two-way communication
in female dyads. Personality and Social Psy systems. In P. Pliner, L. Krames, & T.
chology Bulletin, 26, 278–292. Alloway (Eds.), Nonverbal communication
Blackman, M. C. (2002). The employment inter of aggression (Vol. 2, pp. 21–51). New
view via the telephone: Are we sacrificing York: Plenum.
accurate personality judgments for cost effi Ferrari, J. R., & Swinkels, A. (1996). Classic
ciency? Journal of Research in Personality, cover-ups and misguided messages: Exam
36, 208–233. ining face-trait associations in stereotyped
09-Manusov.qxd 6/30/2006 7:49 PM Page 178
178–––◆–––Factors of Influence
Motowidlo, S. J., Burnett, J. R., Maczynski, J., & Scherer, K. R. (1978). Personality inference from
Witkowski, S. (1996). Predicting managerial voice quality: The loud voice of extroversion.
job performance from personality ratings European Journal of Social Psychology, 8,
based on a structured interview: An interna 467–487.
tional replication. Polish Psychological Schwartz, B., Tesser, A., & Powell, E. (1982).
Bulletin, 27, 139–151. Dominance cues in nonverbal behavior. Social
Mullins, D. T., & Duke, M. P. (2004). Effects of Psychology Quarterly, 45(2), 114–120.
social anxiety on nonverbal accuracy and Viteles, M. S., & Smith, K. R. (1932). The pre
response time I: Facial expressions. Journal diction of vocational aptitude and success
of Nonverbal Behavior, 28, 3–33. from photographs. Journal of Experimental
Patterson, M. L. (1973). Stability of nonverbal Psychology, 15, 615–629.
immediacy behaviors. Journal of Experimen Vrij, A., Akehurst, L., & Morris, P. (1997).
tal Social Psychology, 9, 97–109. Individual differences in hand movements
Patterson, M. L. (1995). A parallel process model during deception. Journal of Nonverbal
of nonverbal communication. Journal of Behavior, 21(2), 87–102.
Nonverbal Behavior, 19, 3–29. Watson, D. (1989). Strangers’ ratings of the five
Peabody, D., & Goldberg, L. R. (1989). Some robust personality factors: Evidence of a
determinants of factor structures from surprising convergence with self-report.
personality-trait descriptors. Journal of Journal of Personality and Social Psychology,
Personality and Social Psychology, 57, 57, 120–128.
552–567. Wiggins, J. S. (1973). Personality and predic
Reichenbach, H. (1938). Experience and predic tion: Principles of personality assessment.
tion. Chicago: University of Chicago Press. Reading, MA: Addison-Wesley.
Riggio, H. R., & Riggio, R. E. (2002). Wiggins, J. S. (1979). A psychological taxonomy
Emotional expressiveness, extraversion, of trait-descriptive terms: The interpersonal
and neuroticism: A meta-analysis. Journal domain. Journal of Personality and Social
of Nonverbal Behavior, 26(4), 195–218. Psychology, 37, 395–412.
Rosenthal, R. (Ed.). (1979). Skill in nonver Young, L. (1993). The naked face: The essential guide
bal communication: Individual differences. to reading faces. New York: St. Martins Press.
Cambridge, MA: Oelgeschlager, Gunn &
Hain.
09-Manusov.qxd 6/30/2006 7:49 PM Page 180
10-Manusov.qxd 6/30/2006 5:00 PM Page 181
10
FACTORING IN AGE
Nonverbal Communication
Across the Life Span
� Robert S. Feldman
University of Massachusetts at Amherst
� James M. Tyler
Purdue University
◆ 181
10-Manusov.qxd 6/30/2006 5:00 PM Page 182
182–––◆–––Factors of Influence
behaviors accurately (DePaulo & Rosenthal, Indeed, beginning with infancy and contin
1982). In comparable fashion, encoding uing throughout adulthood, the ability to
nonverbal behaviors involves the expres differentiate, interpret, and understand
sion of internal experiences in a manner facial expressions is a central component
that others can decode accurately, conse in developing and maintaining relation
quently enabling one to achieve one’s ships. Evidence shows typically that the
desired goals (Burgoon, 1994). Importantly ability to identify facial expressions accu
for this chapter, decoding and encoding rately improves with age, although even
ability improves generally from infancy young infants distinguish among and appro
through early adulthood, although both priately respond to a variety of expressions
may begin to deteriorate at more advanced (Walker-Andrews & Dickson, 1997).
ages (Malatesta, 1981). It is difficult to assess how infants process
The primary goal of the present chapter facial displays, with most studies necessarily
is to examine the development and progres using visual preference or habituation to
sion of nonverbal communication from measure the distinctions that infants make
infancy throughout late adulthood, focus between expressions. In broad terms, how
ing attention on two of the most commonly ever, as early as the first few days of life,
researched channels: facial expressions infants appear to possess some instinctive
and body movements. The first section, capacity for nonverbal communication
Decoding Nonverbal Communication, (i.e., they imitate some facial expressions
involves the ability to decode facial expres and gestures; Field, Woodson, Greenberg,
sions, first in infants and children and sub & Cohen, 1982). Moreover, numerous
sequently in younger and older adults, and features of infants’ perceptual development
focuses on the ability to decode body move indicate sensitivity to visual, acoustic, and
ments, again focusing on infants and motion information, connoting that infants
children and then on adults. The second are responsive to environmental stimuli
section, Encoding Nonverbal Communica (Bertenthal, Proffitt, Spetner, & Thomas,
tion, concerns the capacity to encode non 1985). This early sensitivity may allow for
verbal communication, following the same the eventual recognition and discrimination
outline as the previous section. In the final of different facial expressions.
section, we describe potential directions for To determine infants’ ability to distin
future research. guish facial expressions, researchers typically
expose them to numerous stimuli, including
a variety of posed faces. The results of such
♦ Decoding Nonverbal research indicate that within hours of birth,
Communication neonates look longer at and visually track
stimuli that appear more face-like (Johnson,
Dziurawiec, Ellis, & Morton, 1991). Infants
DECODING FACIAL EXPRESSIONS:
INFANTS AND CHILDREN also appear to differentiate between happy
and surprised expressions and between
Facial expressions are characterized happy and fearful expressions, but only if
typically as a fundamental aspect of non the happy expressions occur first (Young-
verbal communication, with the ability to Browne, Rosenfeld, & Horowitz, 1977).
decode such expressions playing an impor Moreover, infants discriminate between
tant role in the successful navigation of frowning and smiling, especially if posed by
ongoing interactions (Burgoon, 1994). their mother, and they tend to look longer at
10-Manusov.qxd 6/30/2006 5:00 PM Page 183
Factoring in Age–––◆–––183
joy expressions compared with neutral or animals (Feinman & Lewis, 1983). For
angry ones (Barrera & Maurer, 1981). In instance, both 12- and 18-month-olds pre
short, this work suggests that infants may sented with a novel toy remained closer
possess an early sensitivity to recognize and to their mother when she posed a fearful
differentiate facial displays. expression, moved to a middistance for
Because neonates can discern only blurry neutral expressions, and wandered farthest
faces (i.e., they distinguish hairline, eyes, when mothers portrayed happy expressions
nose, and mouth), it is probable that they (Klinnert, 1984). Moreover, 12-month-olds
discriminate facial expressions primarily played less with a toy if their mothers
based on feature information (Ludemann, displayed negative expressions toward that
1991). By 6 months, however, visual sharp toy (Hornik, Risenhoover, & Gunnar,
ness improves, and infants’ sensitivity to 1987). By preschool, children match facial
contrasts enables them to make finer- expressions of emotion to narrated stories
grained affective distinctions between (Borke, 1971) and label facial displays with
expressions (Gwiazda, Bauer, & Held, basic emotions (i.e., happiness, sadness,
1989). Thus, infants as young as 3 to 4 anger, and fear) at better than chance
months old appear to discriminate between accuracy (Denham & Couchoud, 1990;
facial expressions, based initially on feature Philippot & Feldman, 1990).
differences and later on affectively relevant In broad terms, positive emotions are
information (Kestenbaum & Nelson, 1990). recognized earlier and more accurately than
As infants approach 1 year of age, their negative ones (Camras & Allison, 1985).
gaze tends to focus increasingly on the Specifically, 3- to 6-year-old-children can
facial area, and they start to process infor generally identify happiness more accu
mation conveyed by facial expressions in rately than they can identify sadness and
a more cognitively complex way (McClure, anger, and they identify sadness and anger
2000). Specifically, infants move beyond better than fear, disgust, surprise, and
simply distinguishing among expressions shame (Harrigan, 1990). Moreover, when a
to assessing events and regulating behavior facial expression conflicts with a situation
in accord with information derived from (e.g., receiving a gift but displaying a sad
the facial expressions of adults. This abil face), preschoolers attend more to facial
ity—social referencing—is explored empiri expressions than to circumstances to iden
cally by placing an infant in situations that tify the emotion (Wiggers & Van Lieshout,
involve conflicting outcomes and having 1985). For instance, 3- to 5-year-old children
their caretakers react with predetermined use facial cues to interpret whether people
cues when the infant seeks her guidance actually like a drink, particularly when the
(Saarni, Mumme, & Campos, 1998). expressions are highly exaggerated (Eskritt
In a classic study of social referencing, & Lee, 2003). In other words, the degree to
Sorce, Emde, Campos, and Klinnert (1985) which nonverbal cues are salient influences
observed that the majority of 1-year-old preschool children’s interpretation of facial
infants crossed a visual cliff when their expressions.
mothers posed happy expressions, whereas Although the speed that children
no infants crossed the cliff when their process facial expressions improves with
mothers posed fearful expressions. Similar age overall, it varies with the specific emo
examples of social referencing have been tion. For example, happiness is detected
reported using various stimuli to trigger more quickly than anger, fear, and sad
emotion, including toys, strangers, and ness (Boyatzis, Chazan, & Ting, 1993).
10-Manusov.qxd 6/30/2006 5:00 PM Page 184
184–––◆–––Factors of Influence
Moreover, even though young children when judging facial expressions, particu
identify negative facial expressions half as larly when perceiving negative emotions
slow as adults, their accuracy improves (Levinson, Carstensen, Friesen, & Ekman,
with age. This finding is reflected most 1991). Older adults’ accuracy does improve,
strongly in sadness expressions, with however, when identifying expressions of
children’s error rates decreasing from 17% people their own age (Malatesta, Izard,
at age 7 to 8% at age 10 (DeSonneville Culver, & Nicolich, 1987). In their research,
et al., 2002). Beyond the age of 10, com Malatesta et al. found that, across three age
parison with adults makes it clear that groups (young to old), adults performed
children’s processing speed and accuracy best when decoding facial expressions of
increases significantly. Indeed, by adoles adults their own age.
cence, most can organize facial expressions In more precise terms, older compared
into six distinct categories (i.e., happiness, with younger adults rate sad, but not
sadness, surprise, anger, fear, and disgust) happy, facial expressions as significantly
and differentiate the intensity within each, more intense, and are generally less accu
thus allowing for a more detailed assess rate at identifying negative and neutral
ment of facial displays (Ley & Strauss, facial expressions in comparison with
1986). Even at very low levels of intensity, positive ones (McDowell, Harrison, &
nearly 100% of older teens can recognize Demaree, 1994). In fact, older adults are
joy expressions, consistent with the “happy less likely to attend to negative compared
face advantage,” which shows higher over with neutral or positive expressions, thus
all accuracy rates for decoding happy faces decreasing the overall accuracy with which
(Hess, Blairy, & Kleck, 1997). they decode negative displays (Mather &
In short, the ability to decode facial Carstensen, 2003). Moreover, compared
expressions accurately improves with age, with younger adults, older adults’ arousal
with preschool children showing better and amygdala activation diminishes when
than chance accuracy at matching facial viewing negative facial expressions, sug
displays with emotions. Beyond the age of gesting that older adults’ encoding of
10, children’s accuracy is nearly compara negative emotions may be somewhat
ble with adults’. Thus, as children get older, constrained (Mather et al., 2004).
they are able to discriminate and under Other research has examined this poten
stand increasingly complex facial displays tial age-related bias toward negative expres
across a wide range of social interactions. sions more closely. For instance, in one
study, participants viewed a pair of faces
briefly presented on a computer screen, and
DECODING FACIAL EXPRESSIONS: then one face was subsequently replaced
YOUNGER AND OLDER ADULTS with a dot (Mather & Carstensen, 2003).
Participants were instructed to respond as
Although decoding capacity appears to quickly as possible to indicate which side of
improve from infancy through early adult the screen the dot was on. Younger adults’
hood, few studies have examined closely the reaction times remained the same for both
interpretation and decoding of facial expres emotional (positive and negative) and neu
sions in the context of adult developmental tral faces, whereas older adults responded
change over time. The evidence available sug significantly faster when the dot was
gests that older adults commit more decoding located where a happy rather than neutral
errors than younger and middle-aged adults face had previously been. Conversely, older
10-Manusov.qxd 6/30/2006 5:00 PM Page 185
Factoring in Age–––◆–––185
adults’ reaction times were significantly degraded in some respects compared with
slower when the dot was located where an younger adults.
angry or sad face had been in comparison
with a neutral face. In brief, older adults
showed a bias to attend to neutral rather DECODING BODY MOVEMENTS:
than to negative facial expressions, and to INFANTS AND CHILDREN
positive rather than neutral ones, whereas
younger adults did not exhibit these biases. Although body movement has received
In addition, older adults decoded positive considerably less research attention than
faces compared with negative ones more has facial expression, the evidence available
accurately and, in a recognition test, exhib indicates that people decode body move
ited better recall for previously seen posi ments relatively less accurately than they
tive rather than negative faces (Mather & decode facial expressions (Ambady &
Carstensen, 2003). Moreover, similar results Rosenthal, 1992). Nonverbal communica
emerged using a forced choice memory test tion has been investigated using various
in which participants were shown pairs of kinesic forms, including gait, geometric
faces matched for emotional expression and figures, and expressive movement perfor
asked which face they had seen before. mances. The research reveals that various
Specifically, senior adults were most accurate body movements communicate important
at discriminating between happy-new and personal and social information.
happy-old faces, again also indicating that A variety of evidence suggests that body
they recognize positive facial expressions movements communicate emotions reli
better than negative ones; in contrast, type ably, whether they are posed, presented
of emotional expression did not significantly through dance, or elicited during scenes por
affect younger adult’s recognition. traying interpersonal behavior (DeMeijer,
In some respects, then, older adults tend 1989). In general, emotion-based body
not to fare as well as younger adults in movement cues are characterized by varia
decoding the facial expressions of others. tions in form, tempo, force, and direction
This is not to say that older adults are (Montepare, Koff, Zaitchik, & Albert,
unable to interpret facial displays success 1999). For instance, movements character
fully. Indeed, one study examining the ized by extreme muscle exertion in the
facial expressions of physical therapists absence of an actual physical barrier may
found that older adults decoded both posi lead to inferences of fear or anger (Wallbott,
tive and negative facial displays accurately 1980). Similarly, if an individual with no
(Ambady, Koo, Rosenthal, & Winograd, apparent physical reason unexpectedly
2002). Specifically, older adults judged steps backward, it may imply surprise or
therapists accurately as warm, caring, amazement. In short, body movements can
concerned, and empathic when therapists be analyzed as communications that impart
displayed positive expressions and as indif gross categories of emotions (e.g., like-dis
ferent, distant, and cold when they pre like), with postural changes potentially
sented negative expressions. Whereas this reflecting corresponding changes in under
study does not address age-related differ lying interpersonal emotion structures
ences, it suggests that older adults certainly (Ekman, 1965).
have the capacity to interpret and decode With respect to development, children
facial expressions accurately, although as are generally more skilled at decoding
other research documents, their skills are compared with encoding body movements,
10-Manusov.qxd 6/30/2006 5:00 PM Page 186
186–––◆–––Factors of Influence
which may suggest that decoding ability using point light displays presented 4- and
develops first (Kumin & Lazar, 1974). 6-month-old infants with a human running
Specifically, older children are more accu in place and a foil (the same form but con
rate than younger children at decoding verted 180°). Using the observer format as
emotional meaning in body movements and in the previous study, the data revealed that
at using movement cues to judge emotional both infant groups were significantly above
intensity (Michael & Willis, 1968). For chance and that infants preferred the target
example, 6-year-olds who viewed a film rather than the foil. In short, evidence
with a body movement only and a head appears to indicate that sensitivity to body
movement only (masked faced) condition movement may manifest in infants between
displayed better accuracy in the body 4 and 6 months of age.
compared with the head-only condition, With respect to older children, in one
whereas 4-year-olds were less accurate in study, 5-, 8-, 10-, and 12-year-olds made
both conditions (Pendleton & Snyder, emotion attribution judgments, both in nar
1982). Other studies indicate that even very rative form and by choosing one of four
young children—4- and 5-years-olds— emotions that best matched a particular
when asked to choose a movement (from a body movement (Van Meel, Verburgh, &
video) that corresponds to a gestural mes deMeijer, 1993). In both cases (narrative/
sage of a character depicted in a vignette, forced choice), 5-year-olds’ ability to iden
display a 60% decoding accuracy, suggest tify emotion appropriately was significantly
ing that by age 4, children possess sufficient below older children. Although the study
capacity to understand some gestural mean did not involve accuracy scores specifically,
ings (Boyatzis & Satyaprasad, 1994). the results suggest that 5-year-olds failed
Other researchers suggest that sensitivity to decode emotion meaning beyond chance
to body movement manifests as early as 4 to level. Another study involving children
6 months of age (Fox & McDaniel, 1982). aged 4, 5, and 8 years and adults, however,
Using Johansson’s (1973) classic point light found that 5-year-olds did exceed chance
technique (i.e., points of lights on the main identification of happiness, sadness, and fear,
limbs and joints represent a moving person although not anger (Boone & Cunningham,
as small luminous dots moving against a 1998). Even 4-year-olds identified some
black background), Fox and McDaniel emotional meaning in body movement by
presented targets to infants (2, 4, and 6 exceeding better than chance identification
months) that consisted of biological body of sadness. Moreover, although 5-year-olds
movement patterns and foils (i.e., moving failed to identify anger beyond chance, they
dots that did not meet criteria for biological used cues that are associated with adult
movement). Independent observers made attributions of anger to make ratings of
target location judgments based on observ greater emotional intensity.
ing only the infants and not the stimuli. Relative to 4-year-olds, 5-year-olds
Results indicated that the judgment of appear to show increased skill in the ability
observers associated with 2-month-old to recognize emotional expression via body
infants did not differ from chance. Those movement. The most significant growth in
associated with the 4- and 6-month-olds, this ability appears to occur between the
however, made judgments significantly ages of 5 and 8, with 8-year-olds performing
better than chance, suggesting that sensitiv better than chance and showing little differ
ity to body movement may appear by 4 ence compared with adults in their capacity
months of age. The same experimenters also to recognize emotions in body movements.
10-Manusov.qxd 6/30/2006 5:00 PM Page 187
Factoring in Age–––◆–––187
Sadness was the earliest emotion to be 5-year-olds, but not 4-year-olds, were
recognized via body movement, in contrast similar to adults.
to facial expression research, which sug
gests that happiness is the first recognized
(Cunningham & Sterling, 1988). DECODING BODY MOVEMENTS:
Another classic study also suggests that YOUNGER AND OLDER ADULTS
attributions pertaining to social events may,
in part, be grounded in body movement pat Although variations in body movements
terns (Heider & Simmel, 1944). When adults convey important personal information con
describe an animated film in which a large cerning people’s emotions, research investi
triangle, a small triangle, and a circle move at gating the differences in adults’ ability to
various speeds along different trajectories, decode nonverbal expressions, particularly
they virtually never speak in purely geomet in older adults, is rather limited (Aronoff,
ric terms. Instead, they describe a series of Woike, & Hyman, 1992). Moreover, the
interpersonal events that converge on a studies that have examined these differences
common theme. Specifically, they usually tend to focus primarily on decoding facial
describe a scenario in which a woman (cir expressions. Consequently, with respect to
cle) is chased and trapped by a male bully age-related differences, relatively little is
(large triangle) until rescued by a second known concerning the extent to which body
helpful male (small triangle). This descrip movements provide cues to decode nonver
tion style suggests patterns of movement that bal communication accurately.
when displayed by people, reveal states such The evidence available, however, sug
as fear, aggression, and affiliation. gests that certain features involving trunk
The tendency to attribute human charac movement, degree of openness, vertical
teristics to the figures depicted in such direction, force, and speed serve as cues
displays has been shown to appear at to the recognition of various emotions
an early age (Kassin & Baron, 1986). For (Harrigan & Rosenthal, 1983). For
example, after viewing the film, adults example, trunk and head position and
and preschoolers produced open-ended degree of body openness are associated with
anthropomorphic descriptions, although sadness, open arm movements are linked
preschoolers did so less than adults (Berry with warmth and empathy, and leaning
& Springer, 1993). Specifically, 75% of forward and tilting one’s head are tied to
adults described the film anthropomorphi interpersonal positiveness (Matsumoto &
cally (i.e., attribute human states and rela Kudoh, 1987). Moreover, adults who were
tions, e.g., fear and protectiveness, to the shown Heider and Simmel’s (1944) film
geometric figures), whereas only about half judge violence as associated with rapid
of the children did so. The open-ended ver movements, gentleness with slower move
balization task, however, may have under ments, and hesitation with sudden reduc
estimated younger children’s abilities. A tions in speed (Berry, Misovich, Kean, &
second experiment using a fixed-choice Baron, 1992). Another early study that used
question format revealed clear distinctions films of adults playing a ring-toss game also
between the responses of 3- and 4-year revealed that people tend to readily attribute
olds and that of 5-year-olds and adults. psychological qualities to body movements
Specifically, adults and 5-year-olds had (Wolff, 1943). Specifically, adults were
higher canonical patterns of responses than characterized as weak, introverted, and pes
3- and 4-year-olds, and the attributions of simistic if they exhibited a stooped posture,
10-Manusov.qxd 6/30/2006 5:00 PM Page 188
188–––◆–––Factors of Influence
shuffled their feet, or failed to bend their sadness. In contrast, few age-related differ
knees while walking. In contrast, adults ences emerge with respect to the accurate
were viewed as carefree and easygoing if decoding of happy and neutral emotions.
they walked with a sauntering gait and as Another interesting line of research using
happy and optimistic if they briskly lifted expressive dance movements provides evi
their feet while walking. dence that adults use body movements
Recent research using point light displays to decode particular emotions (Boone &
also supports the idea that adults frequently Cunningham, 1998). Specifically, dancers
identify emotion via body movements and enacted trunk and arm movements previ
suggests further that older and younger ously identified as emotion cues, and
adults use similar cues to decode such move decoders rated each movement for 12 dif
ments (Montepare & Zebrowitz-McArthur, ferent emotions. In broad terms, adults used
1988). For instance, angry gaits are associ six cues—upward arm movement, muscle
ated with heavy footedness, sad gaits with tension, forward leaning, directional
less arm swing, and proud and angry gaits changes in face and torso, tempo changes,
with greater stride length. More precisely, and how long arms are kept close to the
anger may be characterized by variations in body—with better than 90% accuracy to
velocity, force, abrupt tempo changes, and distinguish among happiness, sadness,
body angularity, sadness with contracted, anger, and fear. Moreover, as the number of
soft, and smooth movements, and happiness movement cues specifying an emotion
with expanded, action-filled, loose, fast, and increased, the frequency with which adults
somewhat jerkier movements (DeMeijer, decoded the target emotion accurately
1989). In contrast, neutral movements are increased.
related to minimal and contracted actions, Boone and Cunningham (1998) also
in addition to loose, slow, soft, and smooth found similarity between children and
movements. adults in their ability to decode body move
Evidence also suggests some differences ments accurately. Specifically, adults and
in the ease of recognizing certain body children identified anger by increased face
movement-based emotions based on age and torso movement, happiness by increased
and development. For example, younger upward arm movements, sadness by down
adults are more adept at decoding anger ward head movement and less muscle ten
compared with happiness or sadness on the sion, and fear by rigid body posture with
basis of cues provided by a person’s gait head positioned up. More interestingly,
(Montepare, Goldstein, & Clausen, 1987). adults and 8-year-olds displayed better than
With respect to age differences, older adults chance accuracy on all four emotions, and
make fewer accurate identifications than minimal differences arose between adults’
do younger adults, although both groups and 8-year-olds’ ability to identify the dif
decode emotions from body movement at ferent emotions accurately.
above chance levels (Malatesta et al., In summary, children’s capacity to dis
1987). Specifically, older adults are less criminate, identify, and understand the
skilled than younger adults at decoding complex facial expressions of others show
negative compared with positive emotion significant age-related improvement. Like
displays. This is particularly true for anger wise, as children get older, they become
and sadness, which older adults frequently increasingly more aware that people’s
judge as neutral, although they are more body movements may reveal interper
accurate at identifying anger rather than sonal events, relationships, and personal
10-Manusov.qxd 6/30/2006 5:00 PM Page 189
Factoring in Age–––◆–––189
attributes. With respect to older adults, expression of joy, sour taste evokes expres
although they possess the capacity to sions of sadness, and a jack-in-the-box, arm
decode facial expressions accurately, they restraint, and masked stranger situations
commit more decoding errors than younger produce surprise and, in some cases, joy
adults. Similarly, although older adults ful expressions. Moreover, infants’ facial
identify emotions via body movements at movements display temporal patterns
a better than chance rate, they do so less involving smiling, brow knitting, and pout
accurately than younger adults, especially ing, and all but one of the facial muscle
when decoding negative emotions. actions visible in adults can also be identi
fied in infants (Oster, 1978).
Infants as early as 2 months of age can not
♦ Encoding Nonverbal only distinguish among and imitate happy,
Communication sad, and surprised facial expressions but also
convey facial displays of interest, smiling,
anger, and disgust (Izard & Malatesta,
ENCODING FACIAL EXPRESSIONS: 1987). For instance, around 3 to 4 weeks of
INFANTS AND CHILDREN age, infants exhibit social smiling, coalescing
into Duchenne smiles around the end of
An essential aspect underlying people’s the first month, and becoming gradually
capacity to adapt to the social world stronger over the course of the next 5 months
involves knowing how to encode and con (Messinger & Fogel, 1998). In addition,
trol nonverbal emotion displays. From a after gazing at another’s face, 1- to 3-month
developmental perspective, for instance, old infants often smile spontaneously,
instructing children to follow rules or mod although around the fourth month smiling
ulate their emotional behavior often repre becomes reserved increasingly for the infant’s
sents more a request to control facial caregivers (Oster, 1978). Specifically, infants
expressions rather than the direct control often respond as a function of their mothers’
of emotions (Ceschi & Scherer, 2003). expressions of joy, sadness, and anger, and in
Examples of these expression control behav part, they exhibit an age-related increase in
iors can be noted in young children, gradu the ability to match their mothers’ facial
ally increasing in frequency throughout expressions (Izard, Fantauzzo, Castle, &
adulthood (Saarni & von Salisch, 1993). Haynes, 1995). For example, 2.5-month-old
Theorists have even argued that very young infants distinguish between and respond
infants possess some basic signaling capaci accurately to positive and negative conditions
ties in which facial expressions comprise a based, respectively, on their mother’s expres
major component (Trevarthen, 1985). sions of interest and joy or sadness, anger,
Although encoding facial displays has and withdrawal (Haviland & Lelwica, 1987).
been examined across a variety of contexts, Notwithstanding this capacity to interpret
systematic investigation of infants’ facial and match their mothers’ expressions, how
expressions is still somewhat limited (Fox ever, young infants are still prone to exhibit
& Davidson, 1986). Evidence does suggest, positive rather than negative expressions to
however, that young infants exhibit a their mothers’ negative facial displays
variety of distinct facial expressions in (Matias & Cohn, 1993).
response to diverse stimuli (Campos, Moreover, infants 2 to 8 months of age
Campos, & Barrett, 1989). For example, respond with facial expressions of frustra
the most common response to tickling is the tion, predominantly anger, when expected
10-Manusov.qxd 6/30/2006 5:00 PM Page 190
190–––◆–––Factors of Influence
Factoring in Age–––◆–––191
Some research suggests, however, that older which may increase the likelihood that older
adults are less successful than younger adults’ expressions are misinterpreted at a
adults at posing facial expressions accu higher rate than younger adults’ expressions.
rately (Malatesta, 1981). In contrast, other More precisely, as late adulthood
studies investigating spontaneous facial unfolds, physiological characteristics of
expressions or relived emotions show no emotional experience, including somatic
age-related differences in the accuracy of activity and skin conductance, tend to
facial displays and, in some cases, even indi diminish (Levinson et al., 1991). In addi
cate a more accurate expressivity in older tion, other physical changes including sur
adults (Levinson et al., 1991). For example, face musculature and wrinkling also make
accuracy ratings for facial expressions gen it more difficult to interpret older adults’
erated when adults intentionally recalled an facial expressions accurately (Malatesta et al.,
extreme emotional occurrence revealed no 1987). In short, although getting older does
age-related effects (Malatesta et al., 1987). not necessarily result in decreased affectiv
Levinson et al.’s (1991) study shows that ity, understanding older adults’ facial
older adults’ relived emotions are experi expressions may be obscured by wrinkles
enced just as intensely, are just as likely to and other physiological changes. These
elicit facial expressions, and just as fre changes create a situation in which it may
quently elicit emotion-specific autonomic be harder for other people to encode the
nervous system (ANS) activity as are younger older adults’ facial displays successfully
adults’ relived emotions. The capacity of (Malatesta, 1981).
older adults’ facial expressions to recruit Not only may older adults’ facial char
emotion-specific ANS activity, however, acteristics make it difficult to encode
does show clear and significant age-related expressions, but stereotypic beliefs may
diminishment. also affect negatively their capacity to do
Notwithstanding the conflicting evidence so accurately. For example, Matheson
regarding age-related differences in the capac (1997) found that judges’ ratings of pain
ity to encode facial expressions accurately, (facial displays) were more accurate when
researchers agree that as people get older their assessing older adults than younger adults.
expressiveness gradually declines, which may Although judges estimated more pain in
lead to increased difficulty in interpreting older adults’ expressions across all response
their facial expressions. Research supporting types (i.e., genuine, posed, and baseline),
this perspective shows that older adults com older adults did not, however, report any
pared with younger adults have more closed more pain than younger adults. This sug
and less expressive facial expressions, gests that judges may have been biased to
although some self-report studies indicate note pain in older adult’s facial expres
that older adults inhibit affective displays less sions, undermining the impression that rat
often than do younger adults (Malatesta, ings of older adults’ pain was more
1981). Studies using more objective measures, accurate. The stereotypic idea that increased
however, indicate that older adults’ facial pain is a “normal process of aging” may
expressions are more difficult to decode than explain, in part, why higher pain ratings
younger adults and are perceived as sadder were ascribed to older adult’s expressions,
(Malatesta et al., 1987). In short, researchers even though they did not necessarily report
agree that older adults’ attempts to encode more pain. This finding adds to the idea
facial expressions may be complicated by that older adults’ success at encoding facial
facial changes associated with increased age, expressions may be affected negatively by
10-Manusov.qxd 6/30/2006 5:00 PM Page 192
192–––◆–––Factors of Influence
Factoring in Age–––◆–––193
194–––◆–––Factors of Influence
Factoring in Age–––◆–––195
196–––◆–––Factors of Influence
of structure and motion on perceptions of Campos, J. J., Campos, R. G., & Barrett, K. C.
social causality, Personality and Social (1989). Emergent themes in the study of emo
Psychology Bulletin, 18, 237–244. tional development and emotion regulation.
Berry, D. S., & Springer, K. (1993). Structure, Developmental Psychology, 25, 394–402.
motion, and preschoolers’ perceptions of Camras, L. A. (1992). Expressive development
social causality. Ecological Psychology, 5, and basic emotions. Cognition & Emotion,
273–283. 6, 269–283.
Bertenthal, B. I., Proffitt, D. R., Spetner, N. B., Camras, L. A., & Allison, K. (1985). Children’s
& Thomas, M. A. (1985). The development understanding of emotional facial expres
of infant sensitivity to biomechanical sions and verbal labels. Journal of Non
motions. Child Development, 56, 531–543. verbal Behavior, 9, 84–94.
Blake, J., McConnell, S., Horton, G., & Camras, L. A., & Sachs, V. B. (1991). Social ref
Benson, N. (1992). The gestural repertoire erencing and caretaker expressive behavior
and its evolution over the second year. in a day care setting. Infant Behavior and
Early Development and Parenting, 1(3), Development, 14, 27–36.
127–136. Carstensen, L. L., Gottman, J. M., & Levinson,
Boone, R. T., & Cunningham, J. G. (1998). R. W. (1995). Emotional behavior in
Children’s decoding of emotion in expres long-term marriage. Psychology and Aging,
sive body movement: Development of cue 10, 140–149.
attunement. Developmental Psychology, Ceschi, G., & Scherer, K. R. (2003). Children’s
34, 1007–1016. ability to control the facial expression
Boone, R. T., & Cunningham, J. G. (2001). of laughter and smiling: Knowledge and
Children’s expression of emotional meaning behavior. Cognition & Emotion, 17,
in music through expressive body movement. 385–411.
Journal of Nonverbal Behavior, 25, 21–41. Cole, P. M. (1986). Children’s spontaneous
Borke, H. (1971). Interpersonal perception of control of facial expression. Child Develop
young children: Egocentrism or empathy? ment, 57, 1309–1321.
Developmental Psychology, 5, 263–269. Cunningham, J. G., & Sterling, R. S. (1988).
Boyatzis, C. J., Chazan, E., & Ting, C. Z. Developmental change in the understanding
(1993). Preschool children’s decoding of of affective meaning in music. Motivation
facial emotions. Journal of Genetic Psycho and Emotion, 12, 399–413.
logy, 154, 375–382. DeMeijer, M. (1989). The contribution of gen
Boyatzis, C. J., & Satyaprasad, C. (1994). eral features of body movement to the attri
Children’s facial and gestural decoding and butions of emotions. Journal of Nonverbal
encoding: Relations between skills and with Behavior 13, 247–268.
popularity. Journal of Nonverbal Behavior, Denham, S. A., & Couchoud, E. A. (1990).
18, 37–55. Young preschoolers’ ability to identify emo
Boyatzis, C. J., & Watson, M. W. (1993). tions in equivocal situations. Child Study
Preschool children’s symbolic representa Journal, 20, 153–169.
tion of objects through gestures. Child DePaulo, B. M., & Rosenthal, R. (1982).
Development, 64, 729–735. Measuring the development of nonverbal
Burgoon, J. K. (1994). Nonverbal signals. In sensitivity. In C. E. Izard (Ed.), Measuring
M. L. Knapp & G. R. Miller (Eds.), emotions in infants and children. New
Handbook of interpersonal communication York: Cambridge University Press.
(2nd ed., pp. 229–285). Thousand Oaks, DeSonneville, L. M. J., Verschoor, C. A.,
CA: Sage. Njiokiktjien, C., Op het Veld, V.,
Buss, K. A., & Kiel, E. J. (2004). Comparison of Toorenaar, N., & Vranken, M. (2002).
sadness, anger, and fear facial expressions Facial identity and facial emotions: Speed,
when toddlers look at their mothers. Child accuracy, and processing strategies in
Development, 75, 1761–1773. children and adults. Journal of Clinical
10-Manusov.qxd 6/30/2006 5:00 PM Page 197
Factoring in Age–––◆–––197
and Experimental Neuropsychology, 24, Goodwyn, S. W., Acredolo, L. P., & Brown, C. A.
200–213. (2000). Impact of symbolic gesturing on
Efron, D. (1972). Gesture, race, and culture. early language development. Journal of
The Hague: Mouton. Nonverbal Behavior, 24, 81–103.
Ekman, P. (1965). Differential communication Gwiazda, J., Bauer, J., & Held, R. (1989). From
of affect by head and body cues. Journal visual acuity to hyperacuity: A 10-year
of Personality and Social Psychology, 2, update. Canadian Journal of Psychology,
726–735. 43, 109–120.
Eskritt, M., & Lee, K. (2003). Do actions speak Harrigan, J. A. (1990). The effects of task order
louder than words? Preschool children’s use on children’s identification of facial
of the verbal-nonverbal consistency princi expressions. Motivation and Emotion, 8,
ple during inconsistent communications. 157–169.
Journal of Nonverbal Behavior, 27, 25–41. Harrigan, J. A., & Rosenthal, R. (1983).
Feinman, S., & Lewis, M. (1983). Social refer Physicians’ head and body positions as
encing at ten months: A second-order effect determinants of perceived rapport. Journal
on infants’ responses to strangers. Child of Applied Social Psychology, 13, 496–509.
Development, 54, 878–887. Haslett, B. B., & Samter, W. (1997). Children
Feldman, R. S., Jenkins, L., & Popoola, O. communicating: The first 5 years. Hillsdale,
(1979). Detection of deception in adults NJ: Lawrence Erlbaum.
and children via facial expressions. Child Haviland, J. M., & Lelwica, M. (1987). The
Development, 50, 350–355. induced affect response: 10-week-old infants’
Feldman, R. S., Tomasian, J., & Coats, E. J. responses to three emotion expressions.
(1999). Adolescents’ social competence and Developmental Psychology, 23, 97–104.
nonverbal deception abilities: Adolescents Heider, F., & Simmel, M. (1944). An experimen
with higher social skills are better liars. tal study of apparent behavior. American
Journal of Nonverbal Behavior, 23, Journal of Psychology, 57, 243–259.
237–249. Hess, U., Blairy, S., & Kleck, R. E. (1997). The
Field, J. (1976). The adjustment of reaching intensity of emotional facial expressions
behavior to object distance in early infancy. and decoding accuracy. Journal of Non
Child Development, 47, 304–308. verbal Behavior, 21, 241–257.
Field, T. M., Woodson, R., Greenberg, R., & Hodgins, H. S., & Koestner, R. (1993). The ori
Cohen, D. (1982). Discrimination and imi gins of nonverbal sensitivity. Personality
tation of facial expressions by neonates. and Social Psychology Bulletin, 19,
Science, 218, 179–181. 466–473.
Fox, N. A., & Davidson, R. J. (1986). Hornik, R., Risenhoover, N., & Gunnar, M.
Psychophysiological measures of emotion: (1987). The effects of maternal positive,
New directions in developmental research. neutral, and negative affective communica
In C. E. Izard & P. B. Read (Eds.), tions on infant responses to new toys. Child
Measuring emotions in infants and children Development, 58, 937–944.
(Vol. 2, pp. 13–47). Boston: Cambridge Izard, C. E., Fantauzzo, C. A., Castle, J. M., &
University Press. Haynes, O. M. (1995). The ontogeny and
Fox, R., & McDaniel, C. (1982). The perception significance of infants’ facial expressions in
of biological motion by human infants. the first 9 months of life. Developmental
Science, 218, 486–487. Psychology, 31, 997–1013.
Fridlund, A. J., Ekman, P., & Oster, H. (1987). Izard, C. E., & Malatesta, C. Z. (1987).
Facial expressions of emotion. In A. W. Perspectives on emotional development I:
Siegman & S. Feldstein (Eds.), Nonverbal Differential emotions theory of early
behavior and communication (2nd ed., emotional development. In J. D. Osofsky
pp. 143–223). Hillsdale, NJ: Lawrence (Ed.), Handbook of infant development
Erlbaum. (pp. 494–554). New York: Wiley.
10-Manusov.qxd 6/30/2006 5:00 PM Page 198
198–––◆–––Factors of Influence
Johansson, G. (1973). Visual perception of bio older women. Psychology and Aging, 2,
logical motion and a model for its analysis. 193–203.
Perception and Psychophysics, 14, 201–211. Masur, E. F. (1983). Gestural development,
Johnson, M. H., Dziurawiec, S., Ellis, H., & dual-directional signaling, and the transi
Morton, J. (1991). Newborns’ preferential tion to words. Journal of Psycholinguistic
tracking of face-like stimuli and its subse Research, 12, 93–109.
quent decline. Cognition, 40, 1–19. Mather, M., Canli, T., English, T., Whitfield, S.,
Kassin, S. M., & Baron, R. M. (1986). On the Wais, P., Ochsner, K., et al. (2004).
basicity of social perception cues: Develop Amygdala responses to emotionally val
mental evidence for adult processes? Social enced stimuli in older and younger adults.
Cognition, 4, 180–200. Psychological Science, 15, 259–263.
Kastner, M. P., & Crowder, R. G. (1990). Mather, M., & Carstensen, L. L. (2003). Aging
Perception of the major/minor distinction: and attentional biases for emotional faces.
IV. Emotional connotations in young Psychological Science, 14, 409–415.
children. Music Perception, 8, 189–201. Matheson, D. H. (1997). The painful truth:
Kestenbaum, R., & Nelson, C. A. (1990). The Interpretation of facial expressions of pain
recognition and categorization of upright in older adults. Journal of Nonverbal
and inverted emotional expression by Behavior, 21, 223–238.
infants. Infant Behavior and Development, Matias, R., & Cohn, J. F. (1993). Are max-
13, 497–511. specified infant facial expressions during
Klinnert, M. D. (1984). The regulation of infant face-to-face interaction consistent with
behavior by maternal facial expression. Infant differential emotions theory? Developmen
Behavior and Development, 7, 447–465. tal Psychology, 29, 524–531.
Kumin, L., & Lazar, M. (1974). Gestural Matsumoto, D., & Kudoh, T. (1987). Cultural
communication in preschool children. similarities and differences in the semantic
Perceptual and Motor Skills, 38, 708–710. dimensions of body postures. Journal of
Legerstee, M., Corter, C., & Kienapple, K. Nonverbal Behavior, 11, 166–179.
(1990). Hand, arm, and facial actions McClure, E. B. (2000). A meta-analytic review
of young infants to a social and nonso of sex differences in facial expression pro
cial stimulus. Child Development, 61, cessing and their development in infants,
774–784. children, and adolescents. Psychological
Levinson, R. W., Carstensen, L. L., Friesen, W. Bulletin, 126, 424–453.
V., & Ekman, P. (1991). Emotion, physiol McDowell, C., Harrison, D., & Demaree, H.
ogy, and expression in old age. Psychology (1994). Is right hemisphere decline in the
and Aging, 6, 28–35. perception of emotion a function of aging?
Ley, R. G., & Strauss, E. (1986). Hemispheric International Journal of Neuroscience, 79,
asymmetries in the perception of facial 1–11.
expression by normals. In R. Bruyer (Ed.), McKenzie, B. E., Skouteris, H., Day, R. H., &
The neuropsychology of face perception Hartman, B. (1993). Effective action by
and facial expression (pp. 269–290). infants to contact objects by reaching and
Hillsdale, NJ: Erlbaum. leaning. Child Development, 64, 415–429.
Ludemann, P. M. (1991). Generalized discrimi Messinger, D. S., & Fogel, A. (1998). Give and
nation of positive facial expressions by take: The development of conventional
seven- and ten-month-old infants. Child infant gestures. Merrill-Palmer Quarterly,
Development, 62, 55–67. 44, 566–590.
Malatesta, C. Z. (1981). Affective development Michael, G., & Willis, F. N. J. (1968). The devel
over the lifespan: Involution or growth? opment of gestures as a function of social
Merrill-Palmer Quarterly, 27, 145–173. class, education, and sex. Psychological
Malatesta, C. Z., Izard, C. E., Culver, C., & Record, 18, 515–519.
Nicolich, M. (1987). Emotion communi Montepare, J., Goldstein, S., Clausen, A. (1987).
cation skills in young, middle-aged, and The identification of emotions from gait
10-Manusov.qxd 6/30/2006 5:00 PM Page 199
Factoring in Age–––◆–––199
11
WOMEN’S AND MEN’S
NONVERBAL COMMUNICATION
Similarities, Differences,
Stereotypes, and Origins
� Judith A. Hall
Northeastern University
◆ 201
11-Manusov.qxd 6/30/2006 5:03 PM Page 202
202–––◆–––Factors of Influence
men’s and women’s nonverbal behavior and previous research were upheld, and many
I work to provide explanations for differ other differences emerged as well.
ences in those behaviors. But, lest the reader Table 11.1 shows the behaviors for
expect too much, let me emphasize that des which Briton and Hall (1995a) found
cription is much easier than explanation. significant differences in the ratings of men
Explaining effects as complex as those asso and women. The reader will probably find
ciated with gender is not done easily. these results to be unsurprising because
In the present chapter, I summarize stereo the beliefs are robust and probably widely
types (i.e., everyday beliefs) about nonverbal shared. Only two behaviors did not show
gender differences, the actual differences as a significant perceived gender difference:
revealed by observational research, and some “frowns at others” and “interacts facing
possible theoretical frameworks within directly.” The stereotypes suggest that
which we might understand the differences. women are seen as more expressive, involved,
The nonverbal differences to be discussed warm, fluent, and skilled in their nonverbal
include specific behaviors such as smiling and communication than are men.
gazing and also accuracy in expression via These beliefs also coincide with well-
nonverbal cues (both deliberate and sponta documented stereotypes that depict women
neous) and accuracy in receiving nonverbal as being more emotionally expressive and
cues (both noticing or recalling and interpret sensitive than men (e.g., Bem, 1974; Brody
ing cues). My overall goal is to summarize a & Hall, 2000; Johnson & Shulman, 1988;
large literature in a small space and in a way LaFrance & Banaji, 1992; Spence &
that is fair to different perspectives. Helmreich, 1978), and they also coincide
with the ways men and women describe
themselves. Fischer and Manstead (2000)
♦ Stereotypes About Nonverbal found that in all the 37 countries in which
they gathered data, women rated themselves
Gender Differences
higher on the nonverbal expression of
emotion than men rated themselves. In the
Briton and Hall (1995a) asked over 400 United States, studies have shown pre
college students for their opinions about 19 dictable differences in men’s and women’s
nonverbal behaviors and skills in relation to self-reports of emotional expressiveness (e.g.,
gender. Each student made a separate rating Gross & John, 1998), though sometimes the
for men and for women on a 1–10 scale. For magnitude of the difference varies with the
some behaviors, it was easy to predict the specific emotions being rated (Brody & Hall,
beliefs that would emerge, based on previ 2000). Thus, stereotypic beliefs about the
ous stereotype or self-rating studies: People nonverbal behavior of men and women are
would believe men to be louder, to have less well developed and widely shared.
variable voices, to smile less, to gaze less, to
be less emotionally expressive, to use hand
and arm gestures less, and to be less skilled ♦ What Are the Actual
in encoding (sending) and decoding (judg Nonverbal Differences
ing) nonverbal cues (Korzenny, Korzenny, Between Men and Women?
& Sanchez de Rota, 1985; Kramer, 1977;
Rosenkrantz, Vogel, Bee, Broverman, &
Broverman, 1968; Zuckerman & Larrance, In 1984, I published a meta-analysis of
1979).2 Indeed, all the predictions based on nonverbal gender differences that covered
11-Manusov.qxd 6/30/2006 5:03 PM Page 203
Women Rated Higher Than Men Men Rated Higher Than Women
specific behaviors as well as communication on the Hall (1984) review. Children are
skills. Effects for both the gender of the defined as 2 through 12 years, and adults
person whose nonverbal behavior is mea are defined as college-age and older.
sured and the gender of the interaction part Adolescents’ behavior is mentioned when
ner were reviewed (though the latter had possible, but there is much less research on
fewer results than the former), based on this age group. Because the present review
studies published in English up through is necessarily a simplification, the reader is
1983. This was the first quantitative treat referred to Hall and the other cited works
ment of this subject and it has stood the for discussion of many additional method
test of time rather well even though much ological issues and qualifying points.
research has been published since (its conclu
sions also concurred well with a qualitative
review done by Vrugt & Kerkstra, 1984). NONVERBAL BEHAVIORS
In the following sections, the main con
clusions of the 1984 review (omitting those Smiling. The 1984 analysis found that
studies that included infants) are summa when adolescents and adults were inter
rized and updated as space allows, under acting with others, there was clear evidence
the two main headings of “nonverbal based on 23 studies that women smiled
behaviors” and “nonverbal skills.” The more than men. Gender differences in smiling
number of studies available for different were negligible, however, when people were
behaviors and skills varies dramati observed alone. Based on 20 studies among
cally; because of space limitations, the pre children, even when observed in a social
sent chapter focuses on the most well- situation, there was also little difference.
substantiated results. Results that are not There is also a well-documented gender dif
given a citation to a specific study are based ference in posed photographs. In one such
11-Manusov.qxd 6/30/2006 5:03 PM Page 204
204–––◆–––Factors of Influence
study, Dodd, Russell, and Jenkins (1999) find this but found the exact opposite
examined over 16,000 school yearbook (Halberstadt, Hayes, & Pike, 1988).
photographs and found, as did Hall (1984), Another measure of apparent sincerity is
that the younger ages (kindergarten through the Duchenne smile (Ekman, Davidson, &
Grade 3) showed little difference, but the Friesen, 1990), the term used to describe
gender difference was pronounced after the combination of upturned mouth (zygo
that (including among college students and maticus muscle) and crinkles around the
faculty or staff). The generality of the effect eyes (orbicularis oculi muscle). This con
is evident in a study that found male and figuration has been shown in a number of
female physicians interacting with their studies to reflect felt positive affect more
patients to show a marked difference in than does the upturned mouth alone. Hecht
smiling (Briton & Hall, 1995b). and LaFrance (1998) measured Duchenne
LaFrance, Hecht, and Levy Paluck’s (“enjoyment”) smiles and non-Duchenne
(2003) recent meta-analysis included over (“social”) smiles separately and found that
400 male-female comparisons. They con women displayed more of both kinds than
cluded, consistent with Hall (1984), that men. Thus, the idea that women’s expres
women smiled more than men, that there sions are phony is not well supported by the
was no gender difference in the absence of available research.3
social interaction, that the difference was Hall and Halberstadt (1986) examined
greatest in same-gender pairs, and that social moderating factors for Hall’s (1984) adult
tension was associated with a bigger differ studies and found that when the circum
ence. They also found numerous other mod stances of observation were rated as more
erators, including the following: The gender anxiety provoking, the gender difference was
difference was especially large when people larger. Whether women increased their smil
were aware of being observed, were under ing under such circumstances, or men reduced
explicit instructions to get acquainted, were theirs, could not be ascertained. Finally, Hall
engaged in self-disclosure, and were adoles (1984) also found evidence (as did Hinsz &
cents. The latter result is especially interest Tomhave, 1991) that women received more
ing considering the overall lack of a gender smiles from others than men did.
difference for children. It seems that adoles
cence, a time of acute awareness of gender Facial Expressiveness. Research on children’s
norms, is also the time when boys’ and girls’ facial behavior is scanty and inconclusive,
smiling is most discrepant. but five out of six published studies
One potentially important moderator— reviewed by Hall (1984) that measured
the apparent sincerity of the smile—has been objectively the extent or frequency of facial
examined in only very limited fashion, but movements in adults found that women’s
it is important because of the stereotype faces were significantly more expressive
that women use smiles in an emotionally than men’s. It is not clear whether expres
false way. Bugental, Love, and Gianetto siveness means that more emotions are
(1971), in a much-cited study, found that being shown or that there is simply more
the positivity of women’s facial expressions movement overall.
was less concordant with the positivity of
their spoken words than was the case for Gazing. Twenty-five studies of children and
men. On the basis of this finding, the authors 61 studies of adults were located that mea
called women’s smiles “perfidious.” A sured overall gazing during conversation
larger, more recent, study not only did not (frequency or duration). Females in both
11-Manusov.qxd 6/30/2006 5:03 PM Page 205
age groups engaged in higher levels of the dominance or power construct (a topic to
gazing than males did, with the differ be taken up in a later section).
ence being somewhat larger for adults.
Subsequent research continues to find more Head, Hand, and Arm Gestures. Hand
gazing by women (e.g., McCormick & gestures, as discussed here, are the fluid
Jones, 1989, who observed couples in bars). hand movements that accompany speech,
Hall’s review (1984) also found that women rather than the discrete “emblematic” hand
received more gaze than men did, making movements that can substitute for words
for a pattern also seen for smiling: The least (such as the A-okay sign or the slit-throat
gaze occurred when men interacted with sign; Knapp & Hall, 2005). In terms of
men, and the most gaze occurred when speech-dependent hand movements and
women interacted with women. other movements of the head and arms,
A conspicuous moderator of women’s Hall (1984) concluded, based on 15 studies,
elevated gaze tendencies was uncovered in that women engage in more expressive
a series of studies by Aiello (e.g., Aiello, movements during conversation than men
1977). As the physical distance between do. This is in line with women’s greater
conversing interactants increased, men’s facial expressivity noted above.
gazing increased, as predicted by theories
of intimacy compensation (Argyle & Dean, Body Movements and Positions. The ways
1965). Women’s gazing fell off abruptly, people sit, stand, and move their bodies
however, after a separation of about six feet, have often been noted to distinguish men
creating a reversal in the gender difference. from women. Men were observed to be
It seems that women’s heightened gazing is more restless (e.g., foot and leg move
tied closely to the higher level of involve ments, body shifts, fidgeting; 14 results),
ment produced by closer interactions. more expansive (e.g., wide knees, legs, or
The “visual dominance ratio” is a particu arms; six results), more relaxed (e.g., trunk
lar pattern of gazing during interaction, lean, reclining, feet on table; six results),
which refers specifically to the ratio of the and less involved (e.g., nods, forward lean;
percentage of gaze while speaking to the per 18 results) than were women (Hall, 1984).
centage of gaze while listening. A higher ratio Finally, based on 11 results, Hall (1984)
indicates that the person is relatively more concluded that women engaged in more
likely to gaze while speaking compared to lis self-touching during interaction than men
tening. The visual dominance ratio is higher did, a result also found by McCormick and
in more dominant or powerful persons (hence Jones (1989) for couples observed in bars.
its name), as documented in numerous stud Interesting, for this behavior, Briton and
ies (reviewed by Dovidio & Ellyson, 1985). Hall’s (1995a) stereotype data were wrong.
Dovidio, Ellyson, Keating, Heltman, and
Brown (1988) measured the visual domi Interpersonal Touch. Hall (1984) and Stier
nance ratio in cross-gender interactions in and Hall (1984) concluded that, in general,
which neither person had a power or exper women were more likely to touch others
tise advantage over the other. In these inter than men were, based on 20 studies of
actions, women displayed a lower visual children and 18 studies of adults. This
dominance ratio than did men. Although it is result concurs with Willis and Rinck’s
not clear whether this is an effect of the (1983) study, in which college students
actor’s or the target’s gender, it is nevertheless kept logs of touches immediately after they
interesting because of its theoretical links to occurred, and also with later studies of
11-Manusov.qxd 6/30/2006 5:03 PM Page 206
206–––◆–––Factors of Influence
adults in public places (McCormick & are more likely to initiate nonhand touches
Jones, 1989; Willis & Dodds, 1998). In an with men (e.g., DiBiase & Gunnoe, 2004;
additional study that combined several age Hall & Veccia, 1990). Notably, all this
groups, Berkowitz (1971) observed touches research has been on friendly or at least
in public in six areas of the world. The innocuous touches. For more aggressive
overall tendency of women to touch more touches, the gender differences are unknown,
than men was significant and evident in 23 though one might surmise that they vary with
of the 24 groups observed (four age groups the gender of the target and the specific kind
across six areas of the world). of aggressive touch (Archer, 2002).
Interpersonal touch is, however, a com
plex topic both methodologically and theo Interpersonal Distance and Facing Orienta
retically, and several moderating variables tion. Hall (1984) summarized 28 studies of
are evident in its use. First, the preponder children and 59 studies of adults in which
ance of findings showing more female- interpersonal distance was observed in rela
initiated touch may be especially dependent tively nonreactive ways (such as unobtru
on the high levels of touch typically shown sive observation in public). In both age
between females. Stier and Hall (1984) groups, males established larger interper
and, later, Montemayor and Flannery sonal distances than females did, with the
(1989) noted that the most unambiguous effects being more pronounced in adults
differences were between female-female and than in children. Separate analyses of two
male-male dyads. Because of the cultural other methodologies—staged (where par
norms and values prevailing in some coun ticipants are asked to set comfortable dis
tries, perhaps especially the United States, tances to others in the laboratory) and
men may be particularly reluctant to touch projective (where participants position
other men in most public settings. humanoid figures on a board or in a
Apart from the gender composition of the picture)—show similar though weaker
dyad, several other moderators have been results; specifically, little gender difference
discussed. Sports settings show a disinhibi among children and larger distances for
tion of male-male touch. The type of touch men among adults. For the nonreactive
matters as well, with handshaking being studies, it was also possible to look at the
much more notable between men than gender of target effects. In five studies of
between women (Hall, 1984). Evidence also children and 20 studies of adults, it was very
suggests that when touch in cross-gender evident that males were approached less
dyads is observed, with the question being closely than were females. As was the case
whether the man touches the woman more or for smiling and gazing, the separate actor
vice versa, both the type of touch and the type gender and target gender effects combined
of relationship matter. In younger, less com such that distance was smallest in female-
mitted couples, there is a stronger tendency female dyads and largest in male-male dyads.
for the male to touch the female (as in Regarding directness of orientation, five
Henley, 1973), whereas in older and more studies of children and 18 studies of adults
established couples the balance of touching is indicated that females oriented their heads
likely to be shifted to the female (e.g., Hall & and bodies more directly toward a con
Veccia, 1990; Willis & Dodds, 1998). Several versation partner than males did. Several
studies also suggest that men are more likely studies that used naturalistic observation or
to touch women with their hand, but women confederates who invaded someone’s space
11-Manusov.qxd 6/30/2006 5:03 PM Page 207
208–––◆–––Factors of Influence
similarity in nonverbal skill, but even areas I used the correlation metric because it
in which men’s accuracy exceeds women’s. has the broadest comparability across
different domains of research (Lipsey &
Wilson, 2001; Rosenthal, 1991).4
♦ How Accurate To assess accuracy of the gender differ
Are the Stereotypes? ence beliefs, Briton and Hall (1995a) corre
lated the effect sizes for the belief data
shown in Table 11.1, separately for male
Thus far, I have worked to document and female perceivers, with the effect sizes
beliefs about, and actual differences in, for the actual differences in Hall (1984). The
men’s and women’s nonverbal behavior resulting correlations were .74 for the aver
and skill. A comparison between the stereo aged group of female perceivers and .68
types and the actual gender differences for the averaged group of male perceivers,
indicates there can be no doubt that the meaning that both men and women held
stereotypes are overwhelmingly correct in very accurate beliefs about the patterning of
substance. There are at least two ways to the nonverbal gender differences. Hall and
appraise the accuracy of people’s beliefs in Carter (1999) also gathered stereotype data
more precise quantitative terms. One way is about nonverbal gender differences and
to make a direct comparison (e.g., to gather were able to duplicate this group accuracy
people’s beliefs about the rate of hand ges result almost exactly. In that study, accuracy
turing by men and women during conversa was significantly higher for female than for
tion and then to compare this with the male perceivers. Thus, women possessed
actual rate). In most research, however, more accurate knowledge than men about
people are asked to state their beliefs on an how the genders differ in their nonverbal
arbitrary rating scale that cannot be com behavior.
pared directly with measured behaviors. A
second way is to correlate the beliefs with
the actual differences, a method that does APPRAISING THE SIZE
not require identical metrics for both beliefs OF THE DIFFERENCES
and actual differences. This method, called
sensitivity correlation (Judd & Park, 1993) How big are nonverbal gender differ
or profile correlation (Vogt & Colvin, ences? This is an important question that
2003), captures the covariation between can be addressed in several different ways.
believed and actual differences across One way is to ask whether the differences
behaviors. are big enough to be noticeable in everyday
To apply this method, the actual differ life. According to Cohen (1988), an effect
ences between males and females must be size correlation (r) of about .25 is large
quantified on a common metric. In my enough to be visible to the naked eye, assum
meta-analysis (Hall, 1984), I used the ing repeated exposure. The average effect
point-biserial Pearson correlation between size of the nonverbal gender differences is in
gender (coded male = 0, female = 1) and this range (.20 and above) for expression
the behavior in question, although other accuracy, smiling, facial expressiveness,
indices of effect size are available, notably decoding skill, recall of appearance, gazing
Cohen’s d, defined as the difference (both gazing and being gazed at), interper
between two means divided by their within- sonal distance (both initiated and received),
group standard deviation (Cohen, 1988). body restlessness, bodily expansiveness,
11-Manusov.qxd 6/30/2006 5:03 PM Page 210
210–––◆–––Factors of Influence
bodily expressiveness (mainly hand ges of over 300 meta-analyses on the effects of
tures), self-touching, speech errors, and interventions in education and psychology,
filled pauses. The fact that the observed dif finding an overall effect of r = .24. The non
ferences are this large may help to explain verbal gender differences are comparable
why stereotypes about the differences are with these overall effects.
so accurate. Richard et al.’s (2003) analysis of gender
But there are other ways to ask the mag effects (for a long list of psychological
nitude question. One is in terms of the vari variables) revealed an average correlation
ance explained by the gender differences. of .12, consistent with a similar quantita
In other words, in an average study, how tive summary of gender difference meta
much of the variation in the observed analyses done by Hall (2006), suggesting
behavior is accounted for by gender? The that the nonverbal gender differences are
answer is “not much.” Correlations in the larger on average than other gender differ
range of .20 to .40 account for only small ences reported in the social-personality lit
amounts of variance (4%–16%). This erature. Finally, Hall (2006) compiled a
means that in absolute terms, very little of list of other correlates (besides gender) for
the variation in nonverbal behavior and nonverbal sensitivity and smiling and found
skill is due to gender. Another way to that gender correlated with nonverbal sen
express these differences is to examine the sitivity as strongly (on average) as other
distributions for men and women. With variables did, and it correlated with smiling
effects in the range we find in this area of only slightly more weakly than other vari
research, the male and female distributions ables did. Thus, within a relative frame
will overlap a great deal. To be more exact, work, gender differences in nonverbal
only about 30% of their joint distributions communication are within the range typical
will not be overlapped (Cohen, 1988). This of social-personality psychology in general,
means that many men will exhibit behavior larger than other gender differences on
within the range of “women’s behavior” average, and not very different from other
and vice versa. For example, many men will correlates of the nonverbal behaviors.
smile more than the average woman.
Caution must be taken with the “magni
tude” approach, however. Before we con WHERE DO THE
clude that gender is only a trivial predictor DIFFERENCES COME FROM?
of nonverbal behavior, we must ask a
broader and more relativistic question: This most complex question has no
How do the nonverbal gender differences simple answer, and before starting, several
compare to other social psychological important observations must be made. First,
effects, other gender differences, and other because nonverbal behavior is often ambigu
correlates of the same nonverbal behaviors? ous in meaning (Knapp & Hall, 2005), and
Within these frames of reference, the non studies reporting gender differences hardly
verbal gender differences fare rather well. ever ascertain what the behaviors mean to
Richard, Bond, and Stokes-Zoota (2003) the people observed, it is very difficult to
performed a summary of over 450 meta know what interpretations to place on the
analyses of social and personality phenom gender differences in terms of intentions,
ena (involving over 25,000 studies and 8 motivations, correlated states, and so forth.
million participants). The overall effect was Second, it is not possible to do controlled
an average correlation of .21. Lipsey and experiments that could lead to confident
Wilson (1993) did a nonoverlapping review inferences about cause, given that we
11-Manusov.qxd 6/30/2006 5:03 PM Page 211
cannot assign people randomly to be either strong and high-status people behave.
male or female. Third, because gender is a Women’s behaviors, therefore, are tied
potent social variable whose impact on more to lower social status than to being
behavior begins at birth, it is likely not pos female per se. Furthermore, the subtle and
sible to pinpoint “the” cause of a gender nonconscious nature of nonverbal commu
difference. Indeed, considered in a wider nication provides an effective means not
historical perspective, social expectations, only to display power, status, and domi
modeling, and force of habit may keep a nance but also to enforce and maintain
gender difference alive even if the circum domination over women. Henley’s argu
stances that were the original “cause” are ment had a great deal of influence because
no longer operative. Fourth, there is no it was consonant with the feminist psychol
reason to assume that one unified theory is ogy zeitgeist and it unified many findings
required to explain the plethora of nonver under one parsimonious explanation (Vrugt,
bal gender differences that exist. Although 1987). As a consequence, it has been pre
a single grand explanation may succeed sented—often as fact rather than hypothe
eventually in explaining all or most of the sis—in social psychology and psychology of
differences, it is also possible that there are women textbooks up to the present day.
many different causes that either coexist in Furthermore, Henley’s (1977, 1995) ideas
producing a given gender difference or were more provocative than systematic, in
apply to different behaviors. Below, I offer that she did not claim to put forth a fully
a quick summary of several theoretical developed theory that could account for all
positions concerning possible causes for behaviors, all moderating factors, and all
gender differences in nonverbal behavior. definitions of the power construct. And, of
Andersen (1998) proposed a biologically course, at the time of her seminal work
based explanation for women’s advantage in there was much less empirical research
decoding nonverbal cues. Women may, for available on which to base a comprehensive
example, have evolved to be more sensitive theory. From the perspective of some 30
to nonverbal cues than men because of years later, it is now possible to fill in a cru
advantages in terms of survival of offspring. cial empirical gap.
Some nonverbal differences (e.g., gazing) do Henley’s argument rested on three inter
appear in early infancy, which may support locking claims: that men have more power,
an inborn mechanism; on the other hand, dominance, and status than women; that
social learning may occur even within the men and women differ in their nonverbal
first year of life, and furthermore, some dif communication; and that people high ver
ferences only appear years later (e.g., smil sus low in power, dominance, and status
ing). Perhaps because it is difficult to prove differ in their nonverbal communication in
the biological case, or because sociocultural the same way that men and women differ.
explanations are more compelling to social The first two of these claims are, broadly
researchers, most writers have pursued an speaking, not in dispute. Interestingly, how
approach based on social factors. ever, the third claim was assumed to be true
Henley (1977, 1995) made the most without much empirical substantiation, and
ambitious theoretical proposition when she authors since Henley have not often ques
argued that nonverbal gender differences tioned this assumption. Hall, Halberstadt,
are rooted in gender differences in power, and O’Brien (1997) and Hall, Coats, and
status, and dominance. Women, she argued, Smith LeBeau (2005) undertook an eval
behave the way weak and low-status people uation of this third claim by conducting
behave, whereas men behave the way meta-analyses of the relations of nonverbal
11-Manusov.qxd 6/30/2006 5:03 PM Page 212
212–––◆–––Factors of Influence
behavior to power, dominance, and status produced those social norms in the first
(for simplicity, I will just say “power”). place. Thus, power would be a real but
If Henley’s (1977, 1995) power-based more distant cause. The difficulty of proving
theory has viability, the gender differences this “root cause” argument reduces its
should parallel the power differences.5 Such value, however.
parallelism is evident for the visual domi Setting the power hypothesis aside for
nance ratio, bodily openness, loudness of the moment, what explanations remain? As
voice, interruptions, and back-channel just noted, societal norms, roles, and expec
responses. Some nonverbal behaviors that tations are clearly consistent with the
are more prevalent in persons with higher observed gender differences. To review the
power, however, are actually more charac literature on gender roles is clearly beyond
teristic of women (more facial expressive the scope of this chapter, but it is clear that
ness, smaller interpersonal distances, greater women in many cultures are expected to
decoding skill, greater encoding skill),6,7 be competent in the social domain, indeed
and some nonverbal behaviors that are responsible for the positive outcomes of
more characteristic of women do not show social interaction, and there are well-
overall power effects (smiling, gazing, nod established gender norms on dimensions
ding, gesturing, and direct orientation). such as communion, relationship orienta
Thus, the parallelism required by the tion, concern with feelings, social division of
power-based theory is very inconsistent. labor, and the experience and expression of
Though as an all-encompassing theory affect (especially positive affect; Alexander
Henley’s (1977, 1995) argument appears & Wood, 2000; Clancy & Dollinger, 1993;
not to be viable, future research may still Cross & Madson, 1997; Eagly, 1987; Jansz,
succeed in identifying circumstances and 2000). Nonverbal cues and skills are relev
behaviors in which power underlies gender ant to such norms and expectations. The
differences. Schmid Mast and Hall (2004b), existence of norms and expectations creates a
for example, found that smiling was not chicken-and-egg relation with observed dif
related overall to either manipulated power ferences, however: The one brings about and
or trait dominance in a laboratory study, reinforces the other in a cycle.
but that smiling was in fact elevated in one In the most detailed theoretical state
circumstance: among women (not men) ment to date, Hall, Carter, and Horgan
who wanted, and received, assignment to (2000) offered a conceptual analysis of
the lower power role. This suggests that how various factors may impinge on
future studies could profit from looking women’s smiling, expression skill, and
carefully at moderating factors. interpersonal sensitivity. To illustrate with
If power is not the immediate cause of the the case of smiling, we identified numerous
gender differences, what is? One possibility potential factors that could serve to elevate
is that power is the root cause of whatever is women’s smiling. At the heart of the analy
the more proximal cause. For example, sis is the experience of positive affect and
although social norms that encourage the facial feedback process whereby smil
women to be expressive, warm, approach ing itself produces positive affect (Strack,
able, interpersonally involved, sensitive, and Martin, & Stepper, 1988). The fact that
the like may be the important direct influ women are smiled at more than men could
ences in the lifetime of a given woman (with produce more smiling through the reci
the complementary norms having direct procity process (Hinsz & Tomhave, 1991),
influences on a man), one might still argue which then produces positive affect
that it was the power differential that through facial feedback, more smiling, and
11-Manusov.qxd 6/30/2006 5:03 PM Page 213
214–––◆–––Factors of Influence
empirical investigations of sex and gender Clancy, S. M., & Dollinger, S. J. (1993).
in interaction (pp. 83–100). Mahwah, NJ: Photographic depictions of the self: Gender
Erlbaum. and age differences in social connectedness.
Archer, J. (2002). Sex differences in physically Sex Roles, 29, 477–495.
aggressive acts between heterosexual part Cohen, J. (1988). Statistical power analysis for
ners: A meta-analytic review. Aggression the behavioral sciences (2nd ed.). Mahwah,
and Violent Behavior, 7, 313–351. NJ: Erlbaum.
Argyle, M., & Dean, J. (1965). Eye-contact, Cross, S. E., & Madson, L. (1997). Models
distance and affiliation. Sociometry, 28, of the self: Self-construals and gender.
289–304. Psychological Bulletin, 122, 5–37.
Baron-Cohen, S., Wheelwright, S., Hill, J., Raste, Cunningham, M. R. (1977). Personality and
Y., & Plumb, I. (2001). The “Reading the the structure of the nonverbal communica
Mind in the Eyes” Test Revised Version: tion of emotion. Journal of Personality,
A study with normal adults, and adults with 45, 564–584.
Asperger syndrome or high-functioning DePaulo, B. M. (1992). Nonverbal behavior
autism. Journal of Child Psychology and and self-presentation. Psychological Bulletin,
Psychiatry, 42, 241–251. 111, 203–243.
Bem, S. L. (1974). The measurement of DiBiase, R., & Gunnoe, J. (2004). Gender
psychological androgyny. Journal of and culture differences in touching behav
Consulting and Clinical Psychology, 42, ior. Journal of Social Psychology, 144,
155–162. 49–62.
Berkowitz, W. R. (1971). A cross-national Dodd, D. K., Russell, B. L., & Jenkins, C.
comparison of some social patterns of (1999). Smiling in school yearbook photos:
urban pedestrians. Journal of Cross- Gender differences from kindergarten to
Cultural Psychology, 2, 129–144. adulthood. The Psychological Record, 49,
Biehl, M., Matsumoto, D., Ekman, P., Hearn, 543–554.
V., Heider, K., Kudoh, T., et al. (1997). Dovidio, J. F., & Ellyson, S. L. (1982). Decoding
Matsumoto’s and Ekman’s Japanese and visual dominance behavior: Attributions of
Caucasian Facial Expressions of Emotion power based on the relative percentages of
(JACFEE): Reliability data and cross- looking while speaking and looking while
national differences. Journal of Nonverbal listening. Social Psychology Quarterly, 45,
Behavior, 21, 3–21. 106–113.
Briton, N. J., & Hall, J. A. (1995a). Beliefs Dovidio, J. F., & Ellyson, S. L. (1985). Patterns
about female and male nonverbal commu of visual dominance behavior in humans.
nication. Sex Roles, 32, 79–90. In J. F. Dovidio & S. L. Ellyson (Eds.),
Briton, N. J., & Hall, J. A. (1995b). Gender- Power, dominance, and nonverbal behavior
based expectancies and observer judgments (pp. 128–149). New York: Springer.
of smiling. Journal of Nonverbal Behavior, Dovidio, J. F., Ellyson, S. L., Keating, C. F.,
19, 49–65. Heltman, K., & Brown, C. E. (1988). The
Brody, L. R., & Hall, J. A. (2000). Gender, emo relationship of social power to visual dis
tion, and expression. In M. Lewis & J. M. plays of dominance between men and
Haviland-Jones (Eds.), Handbook of emo women. Journal of Personality and Social
tions (2nd ed., pp. 338–349). New York: Psychology, 54, 233–242.
Guilford. Eagly, A. H. (1987). Sex differences in social
Brunner, L. J. (1979). Smiles can be back chan behavior: A social role interpretation.
nels. Journal of Personality and Social Hillsdale, NJ: Erlbaum.
Psychology, 37, 728–734. Ekman, P., Davidson, R. J., & Friesen, W. V.
Bugental, D. E., Love, L. R., & Gianetto, R. M. (1990). The Duchenne smile: Emotional
(1971). Perfidious feminine faces. Journal expression and brain physiology: II. Journal
of Personality and Social Psychology, 17, of Personality and Social Psychology, 58,
314–318. 342–353.
11-Manusov.qxd 6/30/2006 5:03 PM Page 216
216–––◆–––Factors of Influence
Horgan, T. G., Schmid Mast, M., Hall, J. A., & McClure, E. B. (2000). A meta-analytic review
Carter, J. D. (2004). Gender differences in of sex differences in facial expression pro
memory for the appearance of others. cessing and their development in infants,
Personality and Social Psychology Bulletin, children, and adolescents. Psychological
30, 185–196. Bulletin, 126, 424–453.
Jansz, J. (2000). Masculine identity and restric McCormick, N. B., & Jones, A. J. (1989).
tive emotionality. In A. H. Fischer (Ed.), Gender differences in nonverbal flirtation.
Gender and emotion: Social psychological Journal of Sex Education & Therapy, 15,
perspectives (pp. 166–186). Paris: Cambridge 271–282.
University Press. Montemayor, R., & Flannery, D. J. (1989). A
Johnson, J., & Shulman, G. (1988). More alike naturalistic study of the involvement of
than meets the eye: Perceived gender differ children and adolescents with their mothers
ences in subjective experience and its dis and friends: Developmental differences in
play. Sex Roles, 19, 67–79. expressive behavior. Journal of Adolescent
Judd, C. M., & Park, B. (1993). Definition Research, 4, 3–14.
and assessment of accuracy in social Murphy, N. A., Hall, J. A., & Colvin, C. R.
stereotypes. Psychological Review, 100, (2003). Accurate intelligence assess
109–128. ments in social interaction: Mediators and
Knapp, M. L., & Hall, J. A. (2005). Nonverbal gender effects. Journal of Personality, 71,
communication in human interaction (6th 465–493.
ed.). Belmont, CA: Wadsworth. Noller, P., & Gallois, C. (1986). Sending emo
Korzenny, B. A. G., Korzenny, F., & Sanchez de tional messages in marriage: Non-verbal
Rota, G. (1985). Women’s communication behaviour, sex and communication clarity.
in Mexican organizations. Sex Roles, 12, British Journal of Social Psychology, 25,
867–876. 287–297.
Kramer, C. (1977). Perceptions of male and Richard, F. D., Bond, C. F., Jr., & Stokes-
female speech. Language and Speech, 20, Zoota, J. J. (2003). One hundred years
151–161. of social psychology quantitatively
LaFrance, M., & Banaji, M. (1992). Toward a described. Review of General Psychology,
reconsideration of the gender-emotion rela 7, 331–363.
tionship. In M. S. Clark (Ed.), Emotion and Rosenkrantz, P., Vogel, S., Bee, H., Broverman,
social behavior (pp. 178–201). Newbury I., & Broverman, D. M. (1968). Sex-role
Park, CA: Sage. stereotypes and self-concepts in college
LaFrance, M., Hecht, M. A., & Levy Paluck, E. students. Journal of Consulting and Clinical
(2003). The contingent smile: A meta Psychology, 32, 287–295.
analysis of sex differences in smiling. Rosenthal, R. (1991). Meta-analytic procedures
Psychological Bulletin, 129, 305–334. for social research (rev. ed.). Newbury
Lippa, R. A., & Dietz, J. K. (1998). The relation Park, CA: Sage.
of gender, personality, and intelligence to Rosenthal, R., Hall, J. A., DiMatteo, M. R.,
judges’ accuracy in judging strangers’ per Rogers, P. L., & Archer, D. (1979).
sonality from brief video segments. Journal Sensitivity to nonverbal communication:
of Nonverbal Behavior, 24, 25–43. The PONS test. Baltimore: The Johns
Lipsey, M. W., & Wilson, D. B. (1993). The effi Hopkins University Press.
cacy of psychological, educational, and Rosip, J. C., & Hall, J. A. (2004). Knowledge
behavioral treatment: Confirmation from of nonverbal cues, gender, and nonverbal
meta-analysis. American Psychologist, 48, decoding accuracy. Journal of Nonverbal
1181–1209. Behavior, 28, 267–286.
Lipsey, M. W., & Wilson, D. B. (2001). Scherer, K. R., Banse, R., & Wallbott, H. G.
Practical meta-analysis. Thousand Oaks, (2001). Emotion inferences from vocal
CA: Sage. expression correlate across languages and
11-Manusov.qxd 6/30/2006 5:03 PM Page 218
218–––◆–––Factors of Influence
12
CULTURE AND
NONVERBAL BEHAVIOR
� David Matsumoto
San Francisco State University
♦ Defining Culture
Over the history of time, people have had to solve a host of distinct
social problems in order to adapt and thus achieve reproductive success,
including negotiating complex status hierarchies, forming successful
work and social groups, attracting mates, fighting off potential rivals of
food and sexual partners, giving birth and raising children, and battling
nature (Buss, 1991, 2001). Universal biological imperatives are associ
ated with a universal set of psychological problems that people need to
solve in order to survive; thus, all individuals and groups of individuals
must create ways to deal with these universal problems. The ways that
each group develops then become their culture.
In my view, culture is the product of the interaction between univer
sal biological needs and functions, universal social problems created to
Author’s Note: I thank Marija Drezgic, Devon McCabe, and Joanna Schug
for their aid in conducting the literature review; Seung Hee Yoo for her com
ments on a previous version of this chapter; and Sanae Nakagawa, Andres
Olide, and Akiko Terao for their aid in the functioning of my laboratory.
◆ 219
12-Manusov.qxd 6/30/2006 5:04 PM Page 220
220–––◆–––Factors of Influence
address those needs, and the contexts in emotion, division of labor by sex, revenge
which people live. Culture is created as and retaliation, mate selection and sexual
people adapt to their environments in order jealousy, self-enhancement, and personality
to survive, and it results from the process can be traced to the core aspect of a univer
of individuals’ attempts to adapt to their sal human nature based on biological imper
contexts in addressing the universal social atives and universal social problems of
problems and biological needs. Although adaptation and living.
many different definitions of culture exist But many mental and behavioral
(e.g., Berry, Poortinga, Segall, & Dasen, processes are also culture-specific. Different
1992; Jahoda, 1984; Kroeber & Kluckholn, cultures develop different ways of dealing
1963; Linton, 1936; Rohner, 1984; with the biological imperatives and univer
Triandis, 1972), I define culture as a shared sal social problems based on their contexts.
system of socially transmitted behavior that Language is an example of a very culture-
describes, defines, and guides people’s ways specific behavior. Each culture has its
of life, communicated from one generation own language, with its own vocabulary,
to the next. syntax, grammar, phonology, and pragmat
Because people must deal with the same ics (Barnlund & Araki, 1985; Barnlund &
set of biological needs and functions and Yoshioka, 1990; Chen, 1995; Gudykunst
universal social problems, it is very possible & Mody, 2001; Kim et al., 1996; Minami
and in many cases very likely that the ways & McCabe, 1995; Nomura & Barnlund,
in which they are addressed are the same. 1983). The need to have language may be a
That is, universal biological needs and social pancultural universal problem; and having a
problems can lead to similar solutions language may be a universal solution to this
across cultures, especially over time in our problem. But the specific way in which each
evolutionary history. Thus, many aspects of culture solves this problem—that is, devel
our mental processes and behaviors can be ops its own language—is different in every
considered universal. For example, all culture.
humans appear to have some degree of spe
cific fears, such as to snakes, spiders,
heights, and darkness, because these types of ♦ The Role of Culture
fears have led in our evolutionary history to
in the Nonverbal
greater probability of survival (Seligman &
Communication Process
Hager, 1972). As well, people have a ten
dency to perceive their own ingroup as het
erogeneous, fully recognizing the individual As with verbal communication, culture
differences that exist in that group, whereas influences nonverbal behaviors in profound
they perceive other groups as more homoge ways. By far the largest research literature
neous, assuming less diversity within the on this topic is related to facial expressions
group (Linville & Jones, 1980; Triandis, of emotion, which I review later in this
McCusker, & Hui, 1990). People also seem chapter. In this section, I highlight briefly
to have a natural proclivity to fears of the role of culture on other types of
strangers and outgroup members, which nonverbal behaviors before turning to the
may be a universal basis for ethnocentrism, larger discussion of culture and emotional
prejudice, aggression, and even war (Buss, expressions.
2001; see also Dovidio & colleagues, this
volume). Other universal processes, such as Culture and Gestures. The study of culture
incest avoidance, facial expressions of and gestures has its roots in the study by
12-Manusov.qxd 6/30/2006 5:04 PM Page 221
David Efron (Boas & Efron, 1936; Efron, engaged in more gazing and had more
1941), who examined the gestures of Sicilian direct orientations when interacting with
and Lithuanian Jewish immigrants in New others, less interpersonal distance, and
York City. Efron found that there were dis more touching. Within the United States,
tinct gestures among traditional Jews and there are also differences in gaze and visual
Italians but that the traditional gestures behavior between different ethnic groups
disappeared as people were more assimilated (Exline, Jones, & Maciorowski, 1977;
into the larger American culture. This work LaFrance & Mayo, 1976).
was followed initially by that of Ekman and
his colleagues (Ekman, 1976; Friesen, Ekman, Culture and Interpersonal Space. Hall
& Wallbott, 1979), who documented cultural (1966, 1973) specified four different levels of
differences in emblematic gestures between interpersonal space use depending on social
Japanese, Americans, and New Guineans. relationship type: intimate, personal, social,
Morris and his colleagues (Morris, Collett, and public. Whereas people of all cultures
Marsh, & O’Shaughnessy, 1980) have also seem to make these distinctions, they differ
well documented many cultural differences in in the spaces they attribute to them. Arab
gestures. The American A-OK sign, for males, for example, tend to sit closer to each
example, is an obscene gesture in many cul other than American males, with more
tures of Europe, having sexual implications. direct, confrontational types of body orien
Placing both hands at the side of one’s head tations (Watson & Graves, 1966). They also
and pointing upward with the forefingers sig were found to use greater eye contact and to
nals one is angry in some cultures; in others, speak in louder voices. Arabs, at least in the
however, it means that one wants sex. past, learned to interact with others at dis
tances close enough to feel the other person’s
Culture and Gaze. Research on humans breath (Hall, 1963). Furthermore, Latin
and nonhuman primates has shown that Americans tend to interact more closely than
gaze is associated with dominance, power, do students of European backgrounds
or aggression (Fehr & Exline, 1987) and (Forston & Larson, 1968), and Indonesians
affiliation and nurturance (Argyle & Cook, tend to sit closer than Australians
1976). Fehr and Exline suggested that the (Noesjirwan, 1977, 1978). Italians interact
affiliative aspects of gazing begin in infancy, more closely than either Germans or
as infants attend to adults as their source Americans (Shuter, 1977), and Colombians
of care and protection. Cultures create rules were found to interact at closer distances
concerning gazing and visual attention, than did Costa Ricans (Shuter, 1976).
however, because both aggression and affil
iation are behavioral tendencies that are Culture and Other Nonverbal Behaviors.
important for group stability and mainte Other studies have documented cultural
nance. Cross-cultural research has docu differences in other nonverbal behaviors
mented differences in these rules. Arabs, for as well, such as in the semantic meanings
example, have been found to gaze much attributed to body postures (Kudoh &
longer and more directly at their partners Matsumoto, 1985; Matsumoto & Kudoh,
than do Americans (Hall, 1963; Watson & 1987) and vocal characteristics and hand
Graves, 1966). Watson (1970), who classi and arm movements (Vrij & Winkel, 1991,
fied 30 countries as either a “contact” cul 1992). Collectively, the evidence provides
ture (those that facilitated physical touch or more than ample support for the contention
contact during interaction) or a “noncon that culture plays a large role in molding
tact” culture, found that contact cultures our nonverbal behaviors, which comprise
12-Manusov.qxd 6/30/2006 5:04 PM Page 222
222–––◆–––Factors of Influence
Measurement
Citation Participants Eliciting Stimuli System Emotionsa
(Continued)
12-Manusov.qxd 6/30/2006 5:04 PM Page 224
224–––◆–––Factors of Influence
Measurement
Citation Participants Eliciting Stimuli System Emotionsa
NOTE: FACS, Facial action coding system; EMFACS, emotion facial action coding system; JACFEE, Japanese
and Caucasian facial expressions of emotions; WISC-R, Weschler Intelligence Scale for Children—Revised.
a. Corresponding to facial muscle configurations coded in the face that match those in JACFEE.
instance, studies have shown that the uni 1997). This similarity exists in people of
versal facial expressions of emotion occur in as widely divergent cultures as the United
congenitally blind individuals (Charlesworth States and the Minangkabau of West
& Kreutzer, 1973). Research on nonhuman Sumatra, Indonesia. In addition, there is uni
primates has also demonstrated that the versality in the antecedents that bring about
expressions that are universal to humans emotion (Scherer, 1997a, 1997b).
also occur in animals, and that animals
have many different yet stable signals of
emotion (Chevalier-Skolnikoff, 1973; Geen, CULTURAL DIFFERENCES
1992; Hauser, 1993; Snowdon, 2003). IN EXPRESSING EMOTION:
Likewise, the emotions portrayed in the uni CULTURAL DISPLAY RULES
versal facial expressions correspond to emo
tion taxonomies in different languages Despite the existence of universal facial
around the world (Romney, Boyd, Moore, expressions of emotion, people around the
Batchelder, & Brazill, 1996; Romney, world do express emotions differently. The
Moore, & Rusch, 1997; Shaver, Murdaya, first evidence for cultural differences in
& Fraley, 2001; Shaver, Wu, & Schwartz, expression was Friesen’s (1972) study, in
1992). which the spontaneous expressions of
There is also cross-cultural similarity in Americans and Japanese were examined as
the physiological responses to emotion when they viewed highly stressful films in two
these facial expressions are used as markers, conditions, first alone and then a second
in both the autonomic nervous system and time in the presence of an older, male
brain activity (Davidson, 2003; Ekman, experimenter. In the first condition, the
Levenson, & Friesen, 1983; Levenson, American and Japanese participants
Ekman, & Friesen, 1990; Levenson, Ekman, were similar in their expressions of dis
Heider, & Friesen, 1992; Tsai & Levenson, gust, sadness, fear, and anger; in the second
12-Manusov.qxd 6/30/2006 5:04 PM Page 225
226–––◆–––Factors of Influence
emotions as they feel them with no modifi would do if they felt each emotion in four
cation. But individuals can also amplify social situations: with family members, close
(exaggerate) or deamplify (minimize) their friends, colleagues, and strangers. These
expressions; for instance, feelings of sadness categories were chosen because they repre
may be intensified (amplification) at funerals sent a broad range of social categories within
or minimized (deamplification) at weddings. which people interact, and because previous
People can mask or conceal their emotions research has demonstrated considerable vari
by expressing something other than what ability in cultural values and attitudes across
they feel, as when nurses or physicians hide these social situations (Brewer & Kramer,
their emotions when speaking with patients 1985; Tajfel, 1982).
with terminal illness, or when employees in In our first study using the DRAI
service industries (e.g., flight attendants) (Matsumoto, Takeuchi, Andayani, Kouznet
interact with customers. Individuals may sova, & Krupp, 1998), participants from the
also learn to neutralize their expressions, United States, Japan, South Korea, and
expressing nothing, such as when playing Russia completed the DRAI along with an
poker (poker face) and to qualify their feel individual-level measure of individualism-
ings by expressing emotions in combination, collectivism. Our results showed that
such as when feelings of sadness are mixed Russians exerted the highest control over
with a smile, with the smile commenting on their expressions, followed by South Koreans
the sadness, saying “I’ll be OK.” All these and Japanese; Americans had the lowest
behavioral responses have been found to scores. Significant sex differences were also
occur when spontaneous expressive behav found, with females exerting more control on
iors have been studied (Cole, 1986; Ekman anger, contempt, disgust, and across all emo
& Rosenberg, 1998). tions when with family members, and males
Recently, my colleagues and I created exerting more control on fear and surprise.
the Display Rule Assessment Inventory Our most recent study involving the
(DRAI), in which participants choose a DRAI (Matsumoto, Yoo, Hirayama, &
behavioral response when they experi Petrova, 2005) provided evidence for its
ence different emotions in different social internal and temporal reliability and for
situations (Matsumoto, Takeuchi, Andayani, its content, convergent (with measures of
Kouznetsova, & Krupp, 1998; Matsumoto, emotion regulation), discriminant (correla
Choi, Hirayama, Domae, & Yamaguchi, tions with personality controlling for emo
2005). The emotions were those that previ tion regulation), external, and concurrent
ous research has shown to be universally predictive validity (with personality). The
expressed and recognized: anger, contempt, findings also indicated that expression reg
disgust, fear, happiness, sadness, and sur ulation occurs in the various ways discussed
prise; these were selected because universal earlier, and not on a simple expression-
ity served as a basis by which to examine suppression dimension. Additionally, there
display rules initially and by which compar were consistent and predictable cultural
isons across cultures would be meaningful. differences among American, Russian,
To build internal consistency, a synonym for and Japanese participants. For instance,
each emotion label was also included in the Americans and Russians both expressed
initial DRAI—hostility, defiance, aversion, anger and contempt more than Japanese.
worry, joy, gloom, and shock, respectively— Americans expressed fear and disgust more
resulting in a total of 14 emotions terms. than Russians, and Americans expressed
Participants are asked to consider what they happiness more than did Russians and
12-Manusov.qxd 6/30/2006 5:04 PM Page 227
228–––◆–––Factors of Influence
found that Americans rated external dis were interpreted as occurring as a function
play more intensely than the Japanese, of cultural display rules, none actually mea
but that the Japanese rated internal experi sured display rules and linked them to the
ence more intensely than Americans. judgments. A recent study from our labora
Within-country analyses indicated no sig tory, however, has closed this loop. In this
nificant differences between the two ratings study, American and Japanese participants
for the Japanese; the Americans, however, completed the DRAI and viewed a series of
rated external displays more intensely than facial expressions of emotion portrayed at
they rated subjective experience. high and low intensities (Matsumoto, Choi,
These findings were extended by et al., 2005). They made three judgments for
Matsumoto and colleagues (2002) by hav each face: a categorical judgment of which
ing American and Japanese observers rate emotion was portrayed, and intensity rat
expressions expressed at 0%, 50%, 100%, ings of the strength of the external display
and 125% intensities. The data for the and the presumed subjective experience of
100% and 125% expressions replicated the the expressor. American and Japanese
previous findings: Americans rated external judges thought that the expressors of high
display significantly higher than internal intensity expressions displayed the emotions
experience, whereas there were no differ more strongly than they felt them. When
ences for the Japanese. Also, there were no judging the low intensity expressions,
differences between external and internal Americans and Japanese also rated the
ratings for either Americans or Japanese on expressor’s internal experience higher than
0% expressions, which were expected. On they did the external display, but the effect
50% expressions, however, the findings was significantly larger for the Japanese. All
were intriguing. Whereas there was no these differences were mediated by display
difference between external and internal rules as assessed by the DRAI, suggesting
ratings for the Americans, the Japanese that one’s own rules for expression manage
rated internal experience higher than exter ment influences one’s judgments of expres
nal display. We interpreted these findings as sion management in others.
suggesting that for weaker expressions,
Japanese may assume that a display rule is
operating, and may thus infer more emo A POSSIBLE INGROUP ADVANTAGE
tion being felt than is actually displayed. IN RECOGNIZING EMOTIONS?
When Americans see a weak expression,
however, there need not be any such One type of cultural difference in judg
assumption; thus, they interpret the same ment that has recently received attention
amount of emotion felt as expressed. For concerns the possibility of an ingroup advan
strong expressions, Japanese may assume tage in emotion recognition (Elfenbein &
that the context was such that the expres Ambady, 2002). This is defined as the ten
sion was justified; thus, they infer a level dency for members of a cultural group to be
of emotion felt that is commensurate with more accurate in recognizing the emotions
what is shown. When Americans see a strong of members of their own cultural group
expression, however, they know that there than of other, relatively more disparate
is a display rule to exaggerate one’s feelings; groups. Although previous research testing
thus, they compensate for this display rule this hypothesis (Boucher & Carlson, 1980;
by inferring less emotion felt. Kilbride & Yarczower, 1983; Markham &
One limitation of all the studies cited in Wang, 1996) provided mixed results,
this section was that, although the findings Elfenbein and her colleagues have recently
12-Manusov.qxd 6/30/2006 5:04 PM Page 229
reported a number of studies in support of Four of these were associated with non
it (Elfenbein & Ambady, 2002, 2003a, significant interaction Fs that test the
2003b; Elfenbein, Mandal, Ambady, & ingroup effect. Two involved studies of
Harizuka, 2002). facial expressions (Kilbride & Yarczower,
Elsewhere, I have suggested that studies 1983; Mehta et al., 1992), and both these
must meet two methodological require involved facial action coding system
ments to test the ingroup hypothesis ade (FACS) coding of the facial muscles in the
quately (Matsumoto, 2002). First, studies expressions. The FACS codes were equiva
should employ balanced designs in which lent but not exactly the same across the
all judge cultures view expressions por expressor ethnicities as they are in the
trayed by members of all the other cul Japanese and Caucasian facial expressions
tures in the study. Second, because balanced of emotion (JACFEE), thus allowing for
studies include stimuli expressed by people minor cultural differences in the expres
of multiple cultures, it is necessary to sions to exist (perhaps, corresponding to
ensure that the stimuli are equivalent Elfenbein and Ambady’s, 2002, 2003a;
across the cultural groups in terms of their Elfenbein et al., 2002, “emotion dialects”).
physical signaling properties related to When balanced studies employ expres
emotion. Given both of these concerns, sions that are equivalent in their physical
Matsumoto (2002) concluded that Elfenbein signaling properties (the JACFEE), there
and Ambady’s (2002) original meta-analysis is no support for the ingroup hypothesis
could not support the ingroup hypothesis (Matsumoto, 2002; Matsumoto & Choi,
because they did not review the studies 2004). This is the case whether the expres
as to whether or not they met these two sions being judged are full-face, high
requirements. intensity expressions, or low intensity
When balanced studies are examined as expressions where signal clarity is weaker
to whether or not they employed stimuli (Matsumoto & Choi, 2004). Future studies
that were equivalent in their physical sig will need to isolate differences in expres
naling properties or not, the data are clear: sions across encoder cultures while holding
All the studies reported by Elfenbein and constant nonmorphological features of the
colleagues to date supporting the ingroup face that may contribute to emotion signal
hypothesis have used stimuli that were ing. There are many aspects of the face that
not equivalent across the cultural groups may contribute to emotion signaling,
(Elfenbein & Ambady, 2003a, 2003b; including facial physiognomy, cosmetics,
Elfenbein et al., 2002; Elfenbein, Mandal, and hairstyle, in addition to the actual
Ambady, Harizuka, & Kumar, 2004). expressions themselves (Ekman, 1979;
Furthermore, a close examination of the Matsumoto & Choi, 2004). Research is yet
balanced studies they reviewed in Table 4 to test the possible contributory roles of
of their original meta-analysis (Elfenbein these aspects of the face to emotion signal
& Ambady, 2002) shows that only five ing, which is a possible rich source of infor
studies provide evidence that the physical mation in the future.
signaling properties of the expressions used
as stimuli were equivalent across the
expressor ethnicities (Albas, McCluskey, ♦ Conclusion
& Albas, 1976; Kilbride & Yarczower,
1983; McCluskey, Albas, Niemi, Cuevas,
& Ferrer, 1975; McCluskey & Albas, In considering cultural influences on
1981; Mehta, Ward, & Strongman, 1992). nonverbal behavior, it is first important to
12-Manusov.qxd 6/30/2006 5:04 PM Page 230
230–––◆–––Factors of Influence
expression (pp. 91–168). New York: Sight, sound, and sense (pp. 124–156).
Academic Press. Bloomington: Indiana University Press.
Chen, G. M. (1995). Differences in self- Ekman, P., & Friesen, W. (1969). The repertoire
disclosure patterns among Americans ver of nonverbal behavior: Categories, origins,
sus Chinese: A comparative study. Journal usage, and coding. Semiotica, 1, 49–98.
of Cross-Cultural Psychology, 26, 84–91. Ekman, P., & Friesen, W. (1971). Constants across
Chesney, M. A., Ekman, P., Friesen, W. V., culture in the face and emotion. Journal of
Black, G. W., et al. (1990). Type a behavior Personality and Social Psychology, 17,
pattern: Facial behavior and speech com 124–129.
ponents. Psychosomatic Medicine, 52, Ekman, P., & Friesen, W. V. (1975). Unmasking
307–319. the face; a guide to recognizing emotions
Chevalier-Skolnikoff, S. (1973). Facial expres from facial clues. Englewood Cliffs, NJ:
sion of emotion in nonhuman primates. Prentice Hall.
In P. Ekman (Ed.), Darwin and facial Ekman, P., Friesen, W. V., & Ellsworth, P.
expression (pp. 11–89). New York: (1972). Emotion in the human face: Guide
Academic Press. lines for research and an integration of find
Cole, P. M. (1986). Children’s spontaneous con ings. New York: Pergamon Press.
trol of facial expression. Child Development, Ekman, P., Friesen, W. V., O’Sullivan, M.,
57, 1309–1321. Chan, A., Diacoyanni-Tarlatzis, I., Heider,
Darwin, C. (1998). The expression of emotion K., et al. (1987). Universals and cultural dif
in man and animals. New York: Oxford ferences in the judgments of facial expres
University Press. (Original work published sions of emotion. Journal of Personality
1872) and Social Psychology, 53, 712–717.
Davidson, R. J. (2003). Parsing the subcompo Ekman, P., Levenson, R. W., & Friesen, W. V.
nents of emotion and disorders of emotion: (1983). Autonomic nervous system activity
Perspectives from affective neuroscience. In distinguishes among emotions. Science,
R. J. Davidson, K. R. Scherer, & H. H. 221, 1208–1210.
Goldsmith (Eds.), Handbook of affective Ekman, P., Matsumoto, D., & Friesen, W. (1997).
sciences (pp. 8–24). New York: Oxford Facial expression in affective disorders. In P.
University Press. Ekman & E. L. Rosenberg (Eds.), What the
Edelmann, R. J., Asendorpf, J., Contarello, A., face reveals: Basic and applied studies of
Georgas, J., Villanueva, C., & Zammuner, spontaneous expression using the facial
V. (1987). Self-reported verbal and non action coding system (FACs) (pp. 331–341).
verbal strategies for coping with embarrass New York: Oxford University Press.
ment in five European cultures. Social Ekman, P., & Rosenberg, E. L. (Eds.). (1998).
Science Information, 26, 869–883. What the face reveals: Basic and applied
Efron, D. (1941). Gesture and environment. studies of spontaneous expression using the
Oxford, England: King’s Crown Press. facial action coding system (FACs). New
Ekman, P. (1972). Universal and cultural differ York: Oxford University Press.
ences in facial expression of emotion. In J. Ekman, P., Sorenson, E. R., & Friesen, W. V.
R. Cole (Ed.), Nebraska symposium on (1969). Pancultural elements in facial dis
motivation, 1971 (pp. 207–283). Lincoln: plays of emotion. Science, 164, 86–88.
Nebraska University Press. Elfenbein, H. A., & Ambady, N. (2002). On the
Ekman, P. (1973). Darwin and facial expres universality and cultural specificity of
sion; A century of research in review. New emotion recognition: A meta-analysis.
York: Academic Press. Psychological Bulletin, 128, 205–235.
Ekman, P. (1976). Movements with precise mean Elfenbein, H. A., & Ambady, N. (2003a).
ings. Journal of Communication, 26(3), Cultural similarity’s consequences: A dis
14–26. tance perspective on cross-cultural differen
Ekman, P. (1979). Facial signs: Facts, fantasies, ces in emotion recognition. Journal of
and possibilities. In T. Sebeok (Ed.), Cross-Cultural Psychology, 34, 92–110.
12-Manusov.qxd 6/30/2006 5:04 PM Page 232
232–––◆–––Factors of Influence
Elfenbein, H. A., & Ambady, N. (2003b). When programs for expressive behavior. Journal of
familiarity breeds accuracy: Cultural expo Research in Personality, 26, 273–280.
sure and facial emotion recognition. Jour Gosselin, P., Kirouac, G., & Dore, F. (1995).
nal of Personality and Social Psychology, Components and recognition of facial
85, 276–290. expression in the communication of emo
Elfenbein, H. A., Mandal, M. K., Ambady, N., tion by actors. Journal of Personality and
& Harizuka, S. (2002). Cross-cultural Social Psychology, 68, 83–96.
patterns in emotion recognition: Highlighting Gudykunst, W. B., & Mody, B. (Eds.). (2001).
design and analytic techniques. Emotion, 2, Handbook of international and intercul
75–84. tural communication. Newbury Park, CA:
Elfenbein, H. A., Mandal, M. K., Ambady, N., Sage.
Harizuka, S., & Kumar, S. (2004). Gudykunst, W. B., & Nishida, T. (1984).
Hemifacial differences in the in-group Individual and cultural influences on uncer
advantage in emotion recognition. Cognition tainty reduction. Communication Mono
and Emotion, 18, 613–629. graphs, 51, 23–36.
Ellgring, H. (1986). Nonverbal expression of Gudykunst, W. B., & Ting-Toomey, S. (1988).
psychological states in psychiatric patients. Culture and affective communication.
European Archives of Psychiatry and American Behavioral Scientist, 31, 384–400.
Neurological Sciences, 236, 31–34. Hall, E. T. (1963). A system for the nota
Exline, R. V., Jones, P., & Maciorowski, K. tion of proxemic behaviors. American
(1977, August). Race, affiliative-conflict Anthropologist, 65, 1003–1026.
theory and mutual visual attention during Hall, E. T. (1966). The hidden dimension. New
conversation. Paper presented at the York: Doubleday.
American Psychological Association Annual Hall, E. T. (1973). The silent language. New
Convention, San Francisco. York: Anchor.
Fehr, B. J., & Exline, R. V. (1987). Social visual Hauser, M. (1993). Right hemisphere domi
interactions: A conceptual and literature nance for the production of facial expres
review. In A. W. Siegman & S. Feldstein sion in monkeys. Science, 261, 475–477.
(Eds.), Nonverbal behavior and communi Heller, M., & Haynal, V. (1994). Depression
cation (Vol. 2, pp. 225–326). Hillsdale, NJ: and suicide faces. Cahiers Psychiatriques
Erlbaum. Genevois, 16, 107–117.
Forston, R. F., & Larson, C. U. (1968). The Hofstede, G. H. (1980). Culture’s consequences:
dynamics of space: An experimental study in International differences in work-related
proxemic behavior among Latin Americans values. Beverly Hills: Sage.
and North Americans. Journal of Commu Izard, C. E. (1971). The face of emotion. East
nication, 18, 109–116. Norwalk, CT: Appleton-Century-Crofts.
Frank, M. G., Ekman, P., & Friesen, W. V. Jahoda, G. (1984). Do we need a concept of cul
(1993). Behavioral markers and recogniz ture? Journal of Cross-Cultural Psychology,
ability of the smile of enjoyment. Journal 15, 139–151.
of Personality and Social Psychology, Keltner, D., Moffitt, T., & Stouthamer-Loeber,
64, 83–93. M. (1995). Facial expressions of emotion
Friesen, W. V. (1972). Cultural differences in and psychopathology in adolescent boys.
facial expressions in a social situation: An Journal of Abnormal Psychology, 104,
experimental test of the concept of display 644–652.
rules. Unpublished doctoral dissertation, Kilbride, J. E., & Yarczower, M. (1983). Ethnic
University of California, San Francisco. bias in the recognition of facial expressions.
Friesen, W. V., Ekman, P., & Wallbott, H. Journal of Nonverbal Behavior, 8, 27–41.
(1979). Measuring hand movements. Journal Kim, M. S., Hunter, J. E., Miyahara, A., Horvath,
of Nonverbal Behavior, 4, 97–112. A. M., Bresnahan, M., & Yoon, H. J. (1996).
Geen, T. (1992). Facial expressions in socially iso Individual vs. culture-level dimensions of
lated nonhuman primates: Open and closed individualism and collectivism: Effects on
12-Manusov.qxd 6/30/2006 5:04 PM Page 233
234–––◆–––Factors of Influence
personality and social psychology (Vol. 13, American and European American dating
pp. 175–212). Thousand Oaks, CA: Sage. couples during interpersonal conflict.
Shuter, R. (1976). Proxemics and tactility in Journal of Cross-Cultural Psychology, 28,
Latin America. Journal of Communication, 600–625.
26, 46–52. Vrij, A., & Winkel, F. W. (1991). Cultural
Shuter, R. (1977). A field study of nonverbal com patterns in Dutch and Surinam nonverbal
munication in Germany, Italy, and the United behavior: An analysis of simulated police/
States. Communication Monographs, 44, citizen encounters. Journal of Nonverbal
298–305. Behavior, 15, 169–184.
Snowdon, C. T. (2003). Expression of emotion Vrij, A., & Winkel, F. W. (1992). Cross-cultural
in nonhuman animals. In R. J. Davidson, police-citizen interactions: The influence of
K. Scherer, & H. H. Goldsmith (Eds.), race, beliefs, and nonverbal communication
Handbook of affective sciences (pp. 457– on impression formation. Journal of Applied
480). New York: Oxford University Press. Social Psychology, 22, 1546–1559.
Tajfel, H. (1982). Social psychology of inter Watson, O. M. (1970). Proxemic behavior:
group relations. Annual Review of Psychol A cross-cultural study. The Hague, The
ogy, 33, 1–39. Netherlands: Mouton.
Triandis, H. C. (1972). The analysis of subjec Watson, O. M., & Graves, T. D. (1966).
tive culture. New York: Wiley. Quantitative research in proxemic behavior.
Triandis, H. C., McCusker, C., & Hui, C. H. American Anthropologist, 68, 971–985.
(1990). Multimethod probes of individual Waxer, P. H. (1985). Video ethology: Television
ism and collectivism. Journal of Personality as a data base for cross-cultural studies in
and Social Psychology, 59, 1006–1020. nonverbal displays. Journal of Nonverbal
Tsai, J. L., & Levenson, R. W. (1997). Cultural Behavior, 9, 111–120.
influences of emotional responding: Chinese
12-Manusov.qxd 6/30/2006 5:04 PM Page 236
13-Manusov.qxd 6/30/2006 5:05 PM Page 237
13
CASTING NONVERBAL
BEHAVIOR IN THE MEDIA
Representations and Responses
� Valerie Manusov
University of Washington
� Adam Jaworski
Cardiff University
◆ 237
13-Manusov.qxd 6/30/2006 5:05 PM Page 238
238–––◆–––Factors of Influence
a review of online communication, see include when not to talk (i.e., to use silence;
Walther, this volume). We argue that non Jaworski, 1997), how to gesture (Morris,
verbal cues as they occur within mediated Collett, Marsh, & O’Shaughnessy, 1980),
coverage1—and the talk about those cues and rules for the use of time (Levine, 1997).
within the coverage—affect and reflect Media play a large part in teaching the non
many deeply held beliefs about the nature, verbal code, at least within some societies.
importance, and meaning of nonverbal
behavior. To make this argument, we try to
answer four questions: (1) How are nonver AFFECTIVE SOCIALIZATION
bal codes learned? (2) What is the content of
those codes? (3) What codes exist in media An important part of being socialized
talk about nonverbal cues? and (4) What into a particular communication commu
consequences do such codes have for media nity, and one that is particularly important
users? Our goal in answering these ques for understanding nonverbal behavior and
tions is to reveal the processes—and the media exposure, involves learning the ways
implications—of the media as an influenc in which emotions ought to be expressed
ing factor for our understanding and use of and experienced. Nonverbal cues are con
nonverbal cues. We begin with how nonver nected commonly with emotional expres
bal cues may be learned. sion (Ekman, 1972; but see Fridlund &
Russell, this volume), and research has
found that media perform a number of
♦ Learning the Nonverbal roles in the development and understanding
of affect. These roles include the normative
Code: Socialization
occurrence of nonverbal expressions of
emotion (i.e., how often such emotions
Socialization can be defined as a process of occur in everyday life) and the type of emo
acquiring the cultural (including verbal and tion presented as most common in interac
nonverbal) knowledge and skills needed tion. Research has also examined media
to become competent members of particu exposure and the development of particular
lar communities. Communicative compe affective skills. In addition, researchers have
tencies are included within the larger investigated the connection between non
socialization process (Hymes, 1972) and verbal images in media and peoples’ affec
involve “socialization through language tive reactions to those presentations. This
and socialization to use language” (Ochs & section reviews some arguments regarding
Schieffelin, 1986, p. 2) and learned nonver the connection between media and affective
bal communication (see Feldman & Tyler, understanding and response.
this volume). Saville-Troike (2003), in One of the most common areas of
examining the socialization processes research connecting nonverbal media dis
involved in nonverbal communication, dis plays and affective socialization revolves
cusses how, apart from universal nonverbal around what children may learn about
patterns, such as some of the facial expres emotional expression from media exposure,
sions of emotion identified by Darwin particularly from television. Coats and
(1872) (see also Ekman, 1972), children Feldman (1995) (see also Coats, Feldman, &
in different societies are exposed to— Philippot, 1999) argue that television is a pri
or taught explicitly—appropriate age-, mary means through which children (and
gender-, class-, and other-related nonverbal others) learn how to encode and decode
behavior. These learned behavioral patterns emotions and to discover when specific
13-Manusov.qxd 6/30/2006 5:05 PM Page 239
emotional expressions are appropriate (see viewing are less easy to predict. For
also Saarni, 1985). More specifically, Wilson example, Coats and Feldman (1995)
and Smith (1998) noted that many “young assessed some of the potential outcomes
children encounter certain emotions and of exposure to television on the use of
affective situations on television long before nonverbal emotional expressions and
they experience those same phenomena in found that children who were frequent
real life” (p. 533). Some children’s shows TV watchers were better encoders of
even focus on teaching viewers to recognize those emotions expressed commonly on
and label feelings (Wilson & Smith, 1998). television shows than were infrequent
Researchers have speculated on the watchers; they were also better at encod
reason for television’s primacy in teaching ing spontaneous (but poorer at encoding
young viewers about affect. For instance, posed) emotional displays. The authors
Collins (1983) argues that young children interpreted the findings about overall
are particularly affected by television encoding skill cautiously, suggesting that
portrayals of nonverbal expressions children who watch TV frequently are less
because, in part, their cognitive develop able to regulate their emotional expres
ment may not yet allow them to focus as sions than are children less often exposed
fully on verbal content; thus, nonverbal ele to television. An additional study found
ments are more likely to capture their atten that frequent viewers could better decode
tion. Researchers suggest, however, some others’ facial expressions, but they “had a
problems with such learning. For Houle less differentiated, more simplistic view
and Feldman (1991), “Television presents of the consequences of nonverbal self-
salient, engaging, even prestigious mod presentation” (Feldman, Coats, & Spielman,
els of emotionality and emotional expres 1996, p. 1718).
sion . . . [But] it appears that the models Other researchers have focused on what
television presents . . . are different from is being learned from media: the affective
[the] models presented in children’s social content and responses to affect. Weiss and
environments” (p. 329). Specifically, (1) Wilson (cited in Wilson & Smith, 1998),
televised emotional expressions appear, in for instance, found that subplots in a
general, much more often than they do in television show that were humorous led
everyday interactions and (2) certain emo children to have less negative evaluations
tions (e.g., happiness, sadness, and anger) of the story’s main plot (i.e., an earth
occur more commonly than others and in quake). Specifically, children who saw
different proportions than they are likely to shows with a humorous subplot rated the
do in nonmediated contexts (Houle & main character’s mood more positively
Feldman, 1991). Furthermore, several stud and, perhaps more important, judged the
ies have documented that families on televi occurrence of earthquakes in real-life as
sion are more likely to deal with simple, less severe than did those who did not
rather than complex, emotions (e.g., have the positively affective subplot. Media
Larson, 1993), which may be one explana images have also been found to induce feel
tion for why certain simple skills, but not ings of horror: Wilson and Smith (1998)
complex skills, are facilitated by increased argued that children tend to be frightened
television viewing. by horrific images, with younger children
As would be expected, the amount of more likely to be scared by visually fright
exposure appears to play a role in affec ening characters and older children fright
tive socialization. But the specific findings ened by seeing a televised victim’s terror
regarding affective learning and television reactions.
13-Manusov.qxd 6/30/2006 5:05 PM Page 240
240–––◆–––Factors of Influence
242–––◆–––Factors of Influence
including scholarship whose aim is to soap operas and discerned that female
critique the code’s content. characters tend to reflect only a few body
types, whereas males are portrayed with
more variation (although televised body
UNCOVERING THE CODE types are, overall, thinner than in the actual
population). Also, characters with larger
Considerable research has focused on bodies tended to be cast in fewer “roman
media sources to discover the content of tic” contexts and are outfitted less reveal
certain types of coverage, the relative fre ingly than thinner characters. Larger bodied
quency of such coverage, and possible links characters also tend to be older than their
between the coverage and other important thinner counterparts (White et al., 1999).
variables. Most commonly, researchers
have investigated the coverage of particular
ideologies and their potential associations CRITIQUING THE CODE
with viewers’ reactions. Content analyses
have revealed, for instance, that televi Although research has not investigated
sion emphasizes physical attractiveness for all aspects of the cultural code presented
women (Furnham & Mak, 1999), although nonverbally in media sources, it has
recent trends suggest that attractiveness is revealed an overall tendency to present ide
emphasized increasingly for males as well alized images (young, attractive, thin con
(Bordo, 1999; Coupland, 2005). sumers). In response to this apparent code,
The focus on appearance appears to a number of sociologic and discourse
reflect other ideologies. Coupland (2003) analytic scholars have critiqued the presen
argues that advertisements, epitomized in tation of these ideologies (e.g., Bancroft,
skin-care and other beauty-product market 1998; Jaworski, 2003; Shaw, 1998), with a
ing, display an ageist ideology, representing particular concern with those ideologies
getting older solely in physical terms and surrounding the body. Their studies share a
as fundamentally negative. In an analysis common premise that in late-modern, capi
of the texts in British magazine skin-care talist societies, personal identities have
advertisements, Coupland (2003) found become more fluid and changeable than
that discourses are “scientized” and “mar they used to be in traditional, premodern
ketized” and employed to provide desirable societies. In the contemporary world, schol
solutions to alleviate “premature ageing,” ars argue that identity has taken the form of
particularly for women. She also argues a highly reflexive “project of self” linked to
that the ethos of body management and lifestyle choices and stances, consumption
modification for the sake of maintaining of goods and services, and mass media
a youthful appearance is sanctioned by mediation of personal experience (i.e., a
the normalization of plastic surgery proce persistent but contingent and ongoing,
dures, to which cosmetics have become often aspirational “program” that provides
an “inexpensive” precursor or temporary a focus for our daily activities, narratives,
replacement (Coupland, 2005). and encounters with others; c.f. Bauman,
Other research has focused on the repre 1998; Coupland, Nussbaum, & Grossman,
sentation of different physical types in the 1993; Giddens, 1991).
media, presenting an ideology regarding a More central to nonverbal cues, Feather-
preference for thinness. For example, White, stone (1991) and Shilling (1993) provide
Brown, and Ginsburg (1999) content evidence that the body has become a project
analyzed prime-time shows and afternoon site, fueled by media images of the “ideal”
13-Manusov.qxd 6/30/2006 5:05 PM Page 243
body, leading some to strive for “perfec birth, who says she would rather have a
tion” and leaving the others frustratingly nose job than her hearing. Needless to
“inadequate.” Moreover, the theme of the say, there is nothing wrong with her nose
ideal and “deviant” bodies is the dominant at all. (Edwards, 1998, p. 29)
one in the literature dealing with media
representations of the body and the viewers’ The concept of deviant bodies can be
or readers’ reactions to these images. related to Goffman’s presentation of stigma
For example, a clear pattern emerged in as “an undesired differentness from what
Jaworski’s (2003) discourse analytic study we had anticipated” (Goffman, 1997,
of the interviews with “ordinary” members p. 74). The fear of the deviant body implies
of the public in the BBC four-part series and can only be made meaningful in relation
Naked, broadcast in the United Kingdom in to the ideal, desirable, and normal body. Of
November and December 1998. Specifically, course, neither is the ideal or idealized image
Jaworski observed that the four, roughly of the body a given, nor is it the same and
identified, age groups of the interviewees— unchanging for all people and at all times.
middle aged, young adults, teenagers, and Just as individuals’ identities, self-percep
elderly—positioned talk about their bodies tions, other representations, and so on are
in terms of an ideal of youthful looks, good flexible, changeable, and multiplex, so are
health, and sexual attractiveness and fulfill the representations of the ideal body. At any
ment. In contrast, the groups associated the given time in a specific community (social,
deviant body image with prepuberty, old age, gender, professional, or any other group),
illness, and loss of sexual drive or partners. however, there are certain well-recognized
The findings of this work are consistent and accepted principles on which the idea of
with Urla and Terry’s scholarship (1995) an ideal body is based. This is, for example,
distinguishing between normal and deviant evident from the different representations of
bodies (where normal does not equate with the “ideal” body in the history of (Western)
natural). Urla and Terry argued, like art, whether the representation is Leonardo
Jaworski (2003), that the power constraints da Vinci’s drawing of the classically perfect
imposed on body representations have led to Vitruvian Man (ca. 1490), Jacques-Louis
the widespread acceptance of a binary split David’s painting, Napoleon in His Study
between acceptable and unacceptable body (1812), depicting the Emperor as fusing
images: “normal versus deviant” or “healthy the monarchichal body politic with revolu
versus pathological.” This is problematic tionary symbolism, or the practices of “nor
because “deviance” is not so much linked to malizing” whiteness through the colonial
a particular set of body traits as to their per photography of the 19th century represent
ceived connotations imposed by a hegemonic ing the imperfect body of the racial “other”
social order and value system. In the case (Mirzoeff, 1995).
of the Naked documentary, the problem is The discursive or visual representations
exemplified in the comment by The Sunday of such ideal bodies are constituted by and
Times TV critic, Edwards (1998), who sums are constitutive of the power relations
up the program’s impact as follows: operating within the community. With
regard to the issue of social preoccupation
[Whereas] some people might tune in to with the “ideal versus deviant” bodies,
Naked because of the promise of nudity, Douglas (1966) argues that because the
what keeps you watching is the stories, body can be used as a symbol of any
not the pictures, most dramatically that bounded system, such as society, negotiat
of Louise, a 15-year-old girl, deaf since ing the margins of bodily acceptability can
13-Manusov.qxd 6/30/2006 5:05 PM Page 244
244–––◆–––Factors of Influence
be indicative of negotiating the margins of subvert and challenge these ideologies. This
society, and the shifting of these margins reflexivity of the interviewees is consistent
may have consequences for the centers of with other forms of reflexivity associated
power. This preoccupation can explain with the late-modern era (Beck, 1992,
our fear of, and at the same time fascina 1994; Giddens, 1990, 1991), allowing indi
tion with, “freaks,” “monsters,” “cyborgs,” viduals to tackle their anxieties and uncer
“aliens,” and so on (Eubanks, 1996). tainties concerning the changing beliefs and
Yet our fascination with such patholo value systems and their own shifting identi
gized and stigmatized bodies also has more ties as seen and experienced through their
of a personal dimension. Shildrick (2002) bodies (Lupton & Tulloch, 1998).
argues that “monsters” (i.e., the excluded
bodies that fail to conform to any corpo
real norm) may sometimes turn up in our SUMMARY
own self-perceptions.
Instead of remaining at the outer regions In this section, we discussed scholarship
of our embodied selves, “monsters” may at that has investigated and critiqued the con
times reflect aspects of our own subjectivi tent of media portrayals of at least some
ties, creating uncertainties and anxieties of nonverbal cues (i.e., those focusing primar
our self-perception and self-identification ily on appearance cues or the body). This
(Stafford, 1991). For example, as far as work reflects several ideologies common to
our humanity is defined in contrast to non media representations, helping to reveal the
human creatures (e.g., monsters), we often larger cultural code reflected in these repre
project a “monstrous” image onto our own sentations, which may be, at once, both
bodies through negative self-perception utilitarian and problematic for the overall
(the extreme case being dysmorphophobia) culture. In some cases, these implied ideolo
or, in an instance of carnivalesque play gies are derived from the nonverbal images.
with identities, by putting on a mask of a In other cases, they are apparent in how
Halloween monster. As Shaw (1998) media talk about the nonverbal cues. The
observes, however, the women she inter latter, an analysis of metacommunication, is
viewed on the role of media images of the at the core of studying ideological stances in
body, their self-perception, and identity media texts (as well as in face-to-face inter
action), because it is at this level that
are not deluded “cultural dopes.” . . . we can gain insight into how social groups
Rather, they are an active, interpreting, value and orient to language and communi
knowledgeable and diverse audience, cation (varieties, processes, effects) through
who attribute meaning to cultural images the study of folk beliefs about commu
of female “beauty,” and negotiate their nication, people’s attitudes, and awareness
relationship with their own bodies about their verbal and nonverbal conduct.
within the constraints of the “fashion We turn now more directly to the metacom
beauty complex.” (p. 22) municative role of the media.
assess what media sources (e.g., talk media. In a qualitative discourse analytic
show guests) presume about various study, for example, Jaworski and Galasiński
nonverbal cues (e.g., Greenberg, Sherry, (2002) examined U.K. press reports of
Busselle, Hnilo, & Smith, 1997; Shaw, President Clinton’s testimony to the grand
1998). Additional work has concerned jury in the Clinton-Lewinsky affair as seen
itself with the media’s own portrayals of in the video released on September 21,
specific nonverbal acts (e.g., Featherstone, 1998. This study focuses on the papers’
1991). These discourses or “stories” form choices of the descriptors and interpreta
part of what we term here verbal and visual tions of Clinton’s nonverbal behavior in
metadiscourses of nonverbal behavior (i.e., the video, particularly in the ways they
talk or visual presentation that comments combined written text and still pictures
on the meaning of the nonverbal cues). In from the video in reporting how Clinton
our own work, which we review in this “looked” and “behaved” during the
section, we argue that metadiscursive repre testimony.
sentations may enter the public conscious The relative indeterminacy and immedi
ness and come to constitute structured acy of nonverbal behavior (e.g., facial
understandings, perhaps even “common expressions, gestures, body posture) allow
sense” understandings of how communica the media as well as the social actors
tion works, what it is usually like, what in noninstitutional, face-to-face contexts
certain ways of speaking and nonverbal to use metapragmatic manipulation of
behavior connote and imply, and what they nonverbal behavior (i.e., strategic [re-]
ought to be like. That is, metadiscourse interpretation) or glossing of a particular
works at an ideological level and influences nonverbal expression. Jaworski and
people’s actions and priorities in a wide Galasiński’s analysis revealed that the
range of ways, some clearly visible and metapragmatic nature of the reports was
others much less so. not so much aimed at the “accurate”
reporting of what happened regarding
Clinton’s facial expressions, gestures, tone
JAWORSKI AND GALASIŃSKI of voice, and certain physiological reactions
(e.g., perspiration). Rather, they aimed at
Through metadiscourse commentary constructing a particular version of reality,
(remarking on communicative perfor a version that is ideologically compatible
mance, style, rhetorical function, or silence; with the dominant ideologies subscribed to
Jaworski, 1997), people can influence and by the newspapers.
negotiate how an utterance is or should Overall, the comparison of the eight
have been heard or try to modify the values newspapers’ accounts of Clinton’s nonver
attributed to it (e.g., “What I meant to say bal behavior demonstrates a rather clear
was . . .”). At the metalevel of communica split between broadsheet and tabloid
tion, we can mark our personal or group newspapers. The former tended to be posi
identities, display expertise, claim incompe tive (The Guardian, Financial Times, The
tence, and do many other sorts of “personal Independent, The Daily Telegraph) or at
identification work” or “social relationship least nonjudgmental (The Western Mail).
work” (Jaworski, Coupland, & Galasiński, Of the broadsheets, only The Daily
2004). Therefore, as an important site of Telegraph in its editorial was strongly criti
“ideological work,” metadiscursive studies cal of Clinton, which is not very surprising
have been especially fruitful in unraveling given the newspaper’s conservative bias.
the underlying code (or codes) held by the The tabloids, on the other hand, tended to
13-Manusov.qxd 6/30/2006 5:05 PM Page 246
246–––◆–––Factors of Influence
be mildly critical (The Daily Mail), at a event. Across 218 texts, derived from an
minimum, or, more commonly, sensation array of print, radio, and television discus
ally negative and disparaging (The Sun, The sions of the handshake, the authors argued
Mirror). With regard to the last three that three potentially competing views on
papers, despite its conservatism, The Daily nonverbal behavior emerged. These por
Mail’s position was relatively uncritical trayed nonverbal cues
of the Democratic President. The other
two tabloids, despite their opposing politi (1) as informative (i.e., reflecting emo
cal leanings (The Sun to the right and The tions, statements of the relationship
Mirror to the left), seemed to adopt a unan between the two men and their cultures,
imous stance with regard to the coverage and indicators of personal characteris
of this story by overriding their political tics), (2) as performative (i.e., as strate
differences with the overarching ethos of gic, symbolic, and performed for a larger
sensationalism and focusing on human inter audience), and (3) as transformative
est stories in tabloid publishing (Allan, 1999). (having the potential to alter the nature
of interaction and future events). (p. 2;
italics ours)
MANUSOV AND COLLEAGUES
The same behaviors (the handshake and
As critical discourse analysts, Jaworski surrounding nonverbal cues, such as facial
and Galasiński’s (2002) work involves expressions and other arm movements) were
assessing the performative value of meta legitimately given a range of interpretations,
discursive commentary. Manusov and her reflecting very different conceptions of how
colleagues (Manusov & Bixler, 2003; nonverbal cues work in communication.
Manusov & Milstein, 2005), on the other These findings comment on the “relative
hand, focused on metadiscourse as a way to indeterminacy of nonverbal behavior” as
learn about how the media talk about non argued by Jaworski and Galasiński (2002),
verbal cues and, in doing so, reflect a way and they are consistent with Patterson
of thinking about the nature and meaning (1983) and Manusov (1990) in showing that
of nonverbal cues. To explore media reflec the same behaviors allow for diverse inter
tions regarding the way nonverbal cues may pretations, even of their very origin.
work and what they may mean, Manusov Manusov and Milstein (2005) found a
and her colleagues accessed media texts similar variability in frames for the hand
that commented on the 1993 handshake shake and its accompanying cues in their
between Yitzhak Rabin and Yasser Arafat. analysis of Israeli and Palestinian press cov
The handshake event, while not representa erage of the same event. Somewhat similar to
tive of everyday interaction, is recognized Manusov and Bixler (2003), Manusov and
as a highly significant and public nonverbal Milstein (2005) found two primary frames
cue. It has also been talked about in media for presenting the meaning of nonverbal cues
sources for the past 13 years. in this press, which they labeled representa
Manusov and Bixler (2003) set out to tions (a combination of nonverbal cues’
look at U.S. metadiscourse specifically to informative and performative natures) and
uncover the media’s implied portrayal of transformations. Representations involved
the nature of nonverbal behavior (i.e., the conceptualization of nonverbal cues as
whether it was a set of natural signs, strate a “stand-in” for a larger process, issue, or
gic symbols, etc.) in its coverage of the state, often something intangible or abstract
13-Manusov.qxd 6/30/2006 5:05 PM Page 247
and based in the larger event or context in from particular kinds of advertising
which the behaviors occurred. The more strategies. Specifically, Chaudhuri and Buck
specific meanings for the nonverbal cues (1995) argued that there is an important
framed as representations were peace link between an advertisement model’s
hope-optimism, violence, betrayal, anguish, facial expression and the likelihood that an
authority-legitimacy, agreement-promise, audience will share the emotion vicariously
and dislike. Finally, transformative meanings (for a more general discussion of emotional
focused on the potential of nonverbal cues to contagion, see Tickle-Degnen, this volume).
bring about a new state or process or to serve Furthermore, the authors argued that, “The
as a time marker from which events could be consumer comes to associate the brand with
measured. The more specific transformations the emotion generated (happiness, or relief
included gaining legitimacy, increasing status, from fear and anxiety) and sees the brand as
working as a curse, and moving backward. a status instrument that obtains rewards
and stays punishment” (Chaudhuri & Buck,
1995, p. 425). In their study, advertising
SUMMARY strategies designed to enhance mood were
associated positively with actual affective
Together, the work on metadiscourse response. Attempts to portray status were
provides an indirect, or even ironic, means linked positively with audience arousal.
for understanding the role of the media in
nonverbal communication by focusing on
the “talk” about the nonverbal cues. Thus, SELF-IMAGE
unlike much of the work described earlier,
our work focusing on metadiscourse allows The work locating a link between media
the larger nonverbal code to be understood portrayals of ideal attractiveness and the
by investigating the way it is discussed or development of attractiveness schemata have
presented in the media, with an assumption also found important connections to the
that such presentations may work to pro effects or consequences of such schemata.
mote particular beliefs and ideologies rather Many of these consequences revolve around
than others. We turn now toward other one’s own image and esteem. In particular,
implications of the media code for non activating appearance schemata prior to
verbal cues. watching appearance-related commercials
has led female viewers to feel more anger,
less confidence, and greater overall body
♦ The Consequences dissatisfaction (Hargreaves & Tiggemann,
of the Code 2002). Body dissatisfaction is often
associated with such outcomes as lowered
self-esteem (Henderson-King & Henderson-
Many of the studies mentioned in the previ King, 1997; Richards, Casper, & Larson,
ous sections concern not only what is cov 1990) and eating disorders (Harrison, 1997;
ered in media portrayals of nonverbal cues. Twamley & Davis, 1999). Activation of
They are also concerned deeply with the appearance schemata may also lead to dis
consequences or outcomes of the portrayals, crimination against larger body types (White
especially if viewed over time and over et al., 1999).
contexts. For instance, immediate affective The way people talk about the body,
responses have been investigated as resulting as revealed by Shaw (1998) and Jaworski
13-Manusov.qxd 6/30/2006 5:05 PM Page 248
248–––◆–––Factors of Influence
Those rated more favorably were more about Ronald Reagan than when discussing
likely to win a mock election by both Walter Mondale) also had an audience that
university students and a sample from the was significantly more likely to vote for
general population. Although not stated by Reagan than the audiences of CBS or NBC.
the researchers, the candidates in the favor In the study, the authors provide a case for
able conditions tended to be smiling and the role of facial expressions, rather than
had good posture, which, arguably, is what audience selectivity, as the likely causal
affected the election outcome. variable in voting behavior.
An affective consequence of televised Such bias may occur when talking to, in
nonverbal behaviors was found in a study addition to talking about, politicians. In
of political leaders’ facial displays. In their three studies of Israeli television interview
analysis of the emotional responses viewers ers, Babad (1999) asked U.S. participants
had to videotaped excerpts of the candidates (who did not understand the language in the
in the 1984 U.S. presidential election, interviews) to judge the interviewers’ non
Sullivan and Masters (1988) found that for verbal behaviors. Despite Israeli law that
most of the candidates, neutral and happy demands broadcasters act neutrally to can
or reassuring facial displays evoked (self didates of all parties, Babad’s first study
reported) positive emotions in audience reflected that all six interviewers studied
members. These reactions were more intense used different nonverbal behaviors with dif
in displays that came closest to the election, ferent interviewees, although some of the
particularly for Ronald Reagan (the pattern interviewers were more consistent than were
was reversed for Reagan’s primary Demo others. In the second study, analysis of one
cratic rival, Walter Mondale). Importantly, interviewer’s behaviors when talking to two
the authors found that emotional responses candidates for prime minister in 1996
predicted attitudes toward the candidates, (Shimon Peres and Benjamen Netanyahu)
such that audience members viewed more showed that the interviewer’s nonverbal
favorably those candidates (such as Gary cues were judged as much more favorable
Hart) whose facial expressions evoked the when talking to one candidate than to the
most positive self-reported emotional other. The third study included a more
responses (Sullivan & Masters, 1988). detailed analysis of the first set of data. In it,
Newscasters’ nonverbal behaviors when Babad found that the positive impressions
discussing a candidate also appear to play a emitted by the interviewers were based on
role in audiences’ judgments about political more smiling and nodding, a relaxed face,
candidates. An initial analysis of 227 seg and more round hand movements. Negative
ments of the 1976 U.S. presidential election impressions were tied to “beating” hand
coverage documented that newscasters may movements, nonverbal regulation of the
use more positive facial expressions when interview, forward lean, head thrusting,
talking about one candidate or another blinking, and appearing sarcastic. Overall,
(Friedman, Mertz, & DiMatteo, 1980). more “aggressive” interviewers were those
Later research (e.g., Mullen et al., 1986) likely to use nonverbal cues more differen
found evidence that such facial “bias” may tially across interviewees (Babad, 1999).
affect viewers’ attitudes toward the candi But it is not just interviewers who may
dates: The only U.S. station (ABC) whose act or react differently and affect audience’s
broadcaster showed facial favoritism views. Other research has looked at the
toward a candidate in 1984 (Peter Jennings nonverbal reactions of other parties in and
used more positive facial cues when talking watching an interaction. Seiter (2001), for
13-Manusov.qxd 6/30/2006 5:05 PM Page 250
250–––◆–––Factors of Influence
by—people in interaction, the media are terms of how we act, how we look (or
also, for many, an important factor of influ aspire to look), or who we find desirable,
ence. We argue in this chapter that they powerful, or trustworthy, we need to
work to shape their audience to have cer continue to examine critically the portray
tain expectations, which are often at odds als and the (meta)commentaries of non
with what everyday life would lead audi verbal displays in the media.
ences to expect. Those expectations involve,
among other things, the frequency and type
of emotional expressions, standards of ♦ Note
appearance, and gender roles. The affective
and cognitive responses that audiences have
1. In this chapter, we discuss research
to, and take away from, media are part of a
primarily on media use in developed (First World)
larger communicative code, which includes
nations.
judgments of others, expected behaviors,
and interpretations of actions, that the
media work to affect and reflect. The code,
as revealed in media talk and visual and ♦ References
aural content, also appears to bring about
certain kinds of behaviors, such as particular Allan, S. (1999). News culture. Buckingham:
consumption and voting patterns. Looking Open University Press.
at the media as a “factor of influence” creat Axsom, D., Yates, S., & Chaiken, S. (1987).
ing and re-creating a larger code for its users Audience response as a heuristic cue in per
necessitates a larger critique of the code. In suasion. Journal of Personality and Social
this chapter, we reviewed some of the pub Psychology, 53, 30–40.
lished critiques that others have done. We Babad, E. (1999). Preferential treatment in tele
offer here one additional comment. vision interviewing: Evidence from non
As Ron Scollon (1998) has argued, all verbal behavior. Political Communication,
16, 337–358.
mediated communication is, at some level,
Bancroft, A. (1998). The model of a man:
produced as face-to-face, interpersonal
Masculinity and body image in men’s
communication. In this sense, nonverbal lifestyle magazines. In J. Richardson &
cues displayed in the media, although more A. Shaw (Eds.), The body in qualitative
self-conscious, scripted, rehearsed, and research (pp. 26–38). Aldershot, UK:
aided by image consultants and makeup Ashgate.
artists, are inevitably subject to the same Barthes, R. (1977). Image–music–text. London:
rules of production and interpretation as Fontana.
nonverbal behavior in other, unmediated Bauman, Z. (1998). Globalization: The human
contexts. What is important, however, is consequences. Cambridge, UK: Polity Press.
the immense influence the media have on Beck, U. (1992). Risk society: Towards a new
their audiences through the propagation of modernity. London: Sage.
Beck, U. (1994). The reinvention of politics:
certain modes of behavior and types of
Towards a theory of reflexive moderniza
body image while suppressing and devalu
tion. In U. Beck, A. Giddens, & S. Lash
ing others. Some of these processes are not (Eds.), Reflexive modernization: Politics,
new. After all, looking at ancient Greek tradition and aesthetics in the modern
statues or 17th-century European portraits, social order (pp. 1–55). Cambridge, UK:
we tend to see mostly idealized images of Polity Press.
men and women. As the media wield a huge Bordo, S. (1999). The male body. New York:
amount of power in steering our lives in Farrar, Straus and Giroux.
13-Manusov.qxd 6/30/2006 5:05 PM Page 252
252–––◆–––Factors of Influence
Greenberg, B. S., Sherry, J. L., Busselle, R. W., Jaworski, A., Coupland, N., & Galasiński, D.
Hnilo, L. R., & Smith, S. L. (1997). (Eds.). (2004). Metalanguage: Social and
Daytime television talk shows: Guests, ideological perspectives. Berlin, Germany:
content, and interactions. Journal of Mouton de Gruyter.
Broadcasting & Electronic Media, 41, Jaworski, A., & Galasiński, D. (2002). The ver
412–426. bal construction of non-verbal behaviour:
Gregory, S. W., Jr., & Gallagher, T. J. (2002). British press reports of President Clinton’s
Spectral analysis of candidates’ nonverbal grand jury testimony video. Discourse &
vocal communication: Predicting U.S. presi Society, 13, 629–649.
dential election outcomes. Social Psychology Klassen, M. L., Jasper, C. M., & Schwartz, A.
Quarterly, 65, 298–308. M. (1993). Men and women: Images of
Halberstadt, A. (1986). Family socialization of their relationships in magazine advertise
emotional expression and nonverbal com ments. Journal of Advertising Research,
munication of styles and skills. Journal of 22, 30–39.
Personality and Social Psychology, 51, Kolbe, R. H. (1990). Gender roles in children’s
827–836. advertising: A longitudinal content analysis.
Hargreaves, D., & Tiggemann, M. (2002). The In J. H. Leigh & C. R. Martin (Eds.),
effect of television commercials on mood Current issues and research in advertising
and body dissatisfaction: The role of (pp. 197–206). Ann Arbor: University of
the appearance–schema activation. Journal Michigan, Graduate School of Business
of Social and Clinical Psychology, 21, Administration.
287–308. Larson, M. S. (1993). Family communication on
Harrison, K. (1997). Does interpersonal attrac prime-time television. Journal of Broadcast
tion to thin media personalities promote ing and Electronic Media, 37, 349–357.
eating disorders? Journal of Broadcasting Levine, R. (1997). A geography of time. New
& Electronic Media, 41, 478–500. York: Basic Books.
Henderson-King, E., & Henderson-King, D. Lupton, D. (2000). The social construction of
(1997). Media effects on women’s body medicine and the body. In G. L. Albert,
esteem: Social and individual difference fac R. Fitzpatrick, & S. C. Scrimshaw (Eds.),
tors. Journal of Applied Social Psychology, Handbook of social studies in health and
27, 167–173. medicine (pp. 50–63). London: Sage.
Houle, R., & Feldman, R. S. (1991). Emotional Lupton, D., & Tulloch, J. (1998). The adolescent
displays in children’s television program “unfinished body,” reflexivity and HIV/
ming. Journal of Nonverbal Behavior, 15, AIDS risk. Discourse & Society, 4, 19–34.
261–271. Manusov, V. (1990). An application of attribu
Hymes, D. (1972). On communicative com tion principles to nonverbal messages in
petence. In J. Pride & J. Holmes (Eds.), romantic dyads. Communication Mono
Sociolinguistics (pp. 269–293). Harmonds graphs, 57, 104–118.
worth, UK: Penguin. Manusov, V., & Bixler, N. R. (2003,
Jaworski, A. (1997). “White and white”: November). The polysemous nature of non
Metacommunicative and metaphorical verbal behavior: Variety in media framing
silences. In A. Jaworski (Ed.), Silence: Inter of the Rabin-Arafat handshake. Paper
disciplinary perspectives (pp. 381–401). presented to the National Communication
Berlin, Germany: Mouton de Gruyter. Association, Miami.
Jaworski, A. (2003). Talking bodies: Representa Manusov, V., & Milstein, T. (2005). Inter
tions of norm and deviance in the BBC preting nonverbal behavior: Representation
Naked programme. In J. Coupland & and transformation frames in Israeli and
R. Gwyn (Eds.), Discourse, the body and Palestinian coverage of the 1993 Rabin-
identity (pp. 151–176). Basingstoke, UK: Arafat handshake. Western Journal of
Palgrave Macmillan. Communication, 69, 183–201.
13-Manusov.qxd 6/30/2006 5:05 PM Page 254
254–––◆–––Factors of Influence
McNeal, J. U. (1987). Children are consumers: Roy, A. (2005). The “male gaze” in Indian
Insights and implications. Lexington, MA: television commercials: A rhetorical analy
Lexington Books. sis. In T. Carilli & J. Campbell (Eds.),
Mirzoeff, N. (1995). Bodyscape: Art, Modernity Women and the media: National and global
and the Ideal Figure. London: Routledge. perspectives (pp. 3–18). Lanham, MD:
Morris, D., Collett, P., Marsh, P., & University Press of America.
O’Shaughnessy, M. (1980). Gestures: Their Saarni, C. (1985). Indirect processes in affect
origins and distribution. New York: socialization. In M. Lewis & C. Saarni
Scarborough. (Eds.), The socialization of emotions
Mullen, B., Futrell, D., Stairs, D., Tice, D. M., (pp. 187–212). New York: Plenum.
Baumeister, R. F., Dawson, K. E., et al. Saville-Troike, M. (2003). The ethnography of
(1986). Newscasters’ facial expressions and communication: An introduction (3rd ed.).
voting behavior of viewers: Can a smile Oxford: Blackwell.
elect a president? Journal of Personality and Scollon, R. (1998). Mediated discourse as social
Social Psychology, 51, 291–295. interaction. London: Longman.
Myers, P. N., & Biocca, F. A. (1999).The effect Seiter, E. (1993). Sold separately: Children
of television advertising and program and parents in consumer culture. New
ming on body image distortions in young Brunswick, NJ: Rutgers University Press.
women. In L. K. Guerrero, J. A. Devito, & Seiter, J. S. (2001). Silent derogation and
M. L. Hecht (Eds.), The nonverbal commu perceptions of deceptiveness: Does com
nication reader: Classic and contemporary municating nonverbal disbelief during an
readings (2nd ed., pp. 92–100). Prospect opponent’s speech affect perceptions
Heights, IL: Waveland Press. of a debater’s veracity? Communication
Nabi, R. L., & Hendricks, A. (2003). The Research Reports, 18, 334–344.
persuasive effect of host and audience reac Shaw, A. (1998). Images of the female body:
tion shots in television talk shows. Journal Women’s identities and the media. In
of Communication, 53, 527–543. J. Richardson & A. Shaw (Eds.), The body
Ochs, E., & Schieffelin, B. (1986). Language in qualitative research (pp. 7–25).
socialization across cultures. New York: Aldershot, UK: Ashgate.
Cambridge University Press. Shildrick, M. (2002). Embodying the monster:
Patterson, M. L. (1983). Nonverbal behavior: Encounters with the vulnerable self.
A functional perspective. New York: Springer. London: Sage.
Patterson, M. L., Churchill, M. E., Burger, Shilling, C. (1993). The body and social theory.
G. K., & Powell, J. L. (1992). Verbal and London: Sage.
nonverbal modality effects on impressions Stafford, B. M. (1991). Body criticism: Imaging
of political candidates: Analysis from the the unseen in enlightenment art and medi
1984 presidential debates. Communication cine. Cambridge, MA: MIT Press.
Monographs, 59, 231–242. Sullivan, D. G., & Masters, R. D. (1988).
Richards, M. H., Casper, R. C., & Larson, “Happy warriors”: Leaders’ facial displays,
R. W. (1990). Weight and eating disorders viewers’ emotions, and political support.
among pre- and young adolescent boys and American Journal of Political Science, 32,
girls. Journal of Adolescent Health Care, 345–368.
11, 85–89. Synnott, A. (1993). The body social: Symbolism,
Rosenberg, S. W., Bohan, L., McCafferty, P., & self and society. London: Routledge.
Harris, K. (1986). The image and the vote: Twamley, E. W., & Davis, M. C. (1999). The
The effect of candidate presentation on sociocultural model of eating disturbance in
voter preference. American Journal of young women: The effects of personal
Political Science, 30, 108–127. attributes and family environment. Journal
13-Manusov.qxd 6/30/2006 5:05 PM Page 255
PART III
FUNCTIONS
14-Manusov.qxd 6/30/2006 5:06 PM Page 258
14-Manusov.qxd 6/30/2006 5:06 PM Page 259
14
NONVERBAL BEHAVIOR IN
INTIMATE INTERACTIONS AND
INTIMATE RELATIONSHIPS
� Peter A. Andersen
San Diego State University
� Laura K. Guerrero
Arizona State University
� Susanne M. Jones
University of Minnesota, Twin Cities
◆ 259
14-Manusov.qxd 6/30/2006 5:06 PM Page 260
260–––◆–––Functions
Relational Context
Relationship Enhancement
type of interaction, a focus that has two we argue that nonverbal communication is
primary foundations: (1) that intimate inter the sine qua non of intimacy. Although
actions are necessary to develop and main intimacy can be created by talk, we con
tain intimate relationships and (2) that tend that nonverbal communication is
emotions and behaviors associated with the intimacy’s primary vehicle.
experience of intimacy are displayed within As a framework for exploring the role
the context of intimate interaction. that nonverbal behavior plays in the
Although intimacy can be concept intimacy process, we present an interaction-
ualized as an experience consisting of felt centered model of intimacy processes related
emotions and perceptions of understand to nonverbal behavior (see Figure 14.1).
ing, or as a relationship that is character This model modifies and extends Prager’s
ized by affection and trust, ultimately (1995, 2000; Prager & Roberts, 2004) con
intimacy is located in interaction (Prager, ceptualization of intimacy. Positive non
2000). Intimate interaction is the vehicle verbal involvement cues occupy a central
through which people exchange intimate position in the model. Although individuals
actions, thoughts, and feelings. Relation typically manifest positive involvement when
ships are creations of interaction, with they experience intimate thoughts and feel
partners labeling relationships as intimate ings, it is the expression of intimacy (through
(or nonintimate) based on the communica positive involvement cues) that sustains
tion patterns that have occurred between and enhances those intimate thoughts and
them. Thus, we argue that intimacy is feelings. Ultimately, both the experience and
experienced and expressed in interaction. expression of intimacy lead to relationship
Although verbal factors are an important enhancement; relational partners who
component of intimate interaction, we engage in intimate interaction frequently
shall demonstrate in this chapter that and routinely are more likely to be satisfied
nonverbal behaviors play a critical role in and committed as well as to manage conflict
creating and sustaining intimate interac effectively (Prager, 2000). Variables associ
tions and relationships. More pointedly, ated with relational context, such as the
14-Manusov.qxd 6/30/2006 5:06 PM Page 261
type and stage of a relationship, frame the 1995). In his triangular theory of love,
experience and expression of intimacy and Sternberg (1986) conceptualized intimacy as
provide guidelines for what constitutes appro the experience of warm, affectionate feel
priate levels of intimacy within interactions. ings that occur during interaction with close
With our interaction-centered model friends and others we hold dear. Similarly,
serving as a guide, this chapter is organized Clarke, Allen, and Dickson (1985) defined
around three issues. First, we distinguish warmth as a positive emotion that charac
between the experience and expression terizes close, intimate relationships, and
of intimacy. Second, we examine nonverbal Andersen and Guerrero (1998) argued that
cues of positive involvement and argue that warmth is a “pleasant, contented, intimate
these cues are the basic building blocks of feeling that occurs during positive inter
intimate interaction. Third, we examine actions” with others (p. 306). As these
positive involvement behaviors in the con conceptualizations suggest, intimate feel
text of intimate relationships, showing that ings are sustained, enhanced, and created
these behaviors are associated with relation through intimate interaction with others.
ship maintenance and enhancement. This illustrates the complexity of intimacy.
Intimacy is experienced typically during
interaction in close relationships in the pres
THE EXPERIENCE VERSUS THE ence of positive verbal and nonverbal
EXPRESSION OF INTIMACY behavior that reflects and creates feelings of
warmth. The experience of interpersonal
Scholars have distinguished between
warmth can lead people to engage in more
the experience and expression of intimacy,
positive behavior, just as positive behavior
sometimes referring to these components as
can trigger feelings of warmth and intimacy.
latent versus manifest intimacy, respectively
(e.g., Sternberg, 1986). The experience of
intimacy is located in internal processes NONVERBAL INVOLVEMENT CUES
related to perceptions and felt emotions. At
the perceptual level, Prager and Roberts The expression of intimacy involves
(2004) argued that intimacy is experienced verbal and nonverbal behavior (Prager,
through shared knowledge. This knowl 2000). Verbal expressions of intimacy
edge, which is gained through intimate include self-disclosure, verbal responsive
interaction, is stored in cognitive schemas. ness, and intimate words (Andersen, 1998).
The extent to which these schemas are Self-revealing statements that express vul
accurate and reflect shared meaning defines nerable emotions are especially conducive
the quality of intimate relationships. The to intimacy (Prager & Roberts, 2004).
perception of understanding is also impor Statements expressing agreement and vali
tant. Prager (2000) noted that perceiving dation also foster intimacy (Andersen &
oneself to be “liked, accepted, understood, Guerrero, 1998). Nonverbal expressions of
cared for, or loved” is an essential part of intimacy include a wide range of behaviors
the intimacy experience (p. 231; see also, that reflect both positive affect and involve
Reis & Shaver, 1988). ment, such as gaze, smiling, forward lean,
At the emotional level, people experience and affirming head nods. According to
intimacy as an affective state characterized Prager (2000), positive involvement behav
by subjective feelings of warmth and affec iors contribute “substantially to people’s
tion (Andersen & Guerrero, 1998; Prager, intimate experiences” (p. 233) and are a
14-Manusov.qxd 6/30/2006 5:06 PM Page 262
262–––◆–––Functions
High Involvement
Intimacy/ Dominance/
Affiliation Aggression
Social Avoidance/
Politeness Withdrawal
Low Involvement
264–––◆–––Functions
partners would be more familiar with one (Marston, Hecht, & Robers, 1987; Shaver,
another’s communication style. Burgoon Schwartz, Kirson, & O’Connor, 1987).
and Newton (1991) added a sixth dimen Liking is communicated through eye contact,
sion, affect, as relevant but not essential to smiling, facial and gestural animation, and
nonverbal involvement. According to this head nodding (Floyd & Ray, 2003; Palmer
perspective, the simultaneous manifestation & Simmons, 1995). Finally, joy is associated
of involvement behaviors and positive affect with being physically energetic, smiling,
cues (such as smiling and vocal warmth) laughing, approaching others, and sound
helps create or sustain intimacy. In contrast, ing enthusiastic (Shaver et al., 1987). When
when involvement behaviors are paired these emotions are exchanged, an intimate
with negative affect cues (e.g., scowling and interaction is created that helps produce and
a loud, angry voice), dominance or aggres sustain an intimate relationship.
sion is communicated. Instead of using the term positive
Dillard, Solomon, and Palmer (1999) involvement, some scholars use the term
argued that involvement is conceptualized immediacy to describe a set of behaviors
by the level of intensity or engagement that communicates both involvement and
present within an interaction. Involvement positive affect (Andersen, 1985). Mehrabian
cues are present in interactions characterized (1967, 1969b) coined the term immediacy
as either affiliative-intimate or dominant- to describe approach behaviors that reflect
aggressive, depending on the type of affect the intensity level of interaction. Later,
present in the interaction (see also, Mehrabian (1981) argued that immediacy
Cappella, 1983; Guerrero, 2004; Prager, behaviors also signal attentiveness, heighten
1995; see Figure 14.2). Thus, nonverbal sensory stimulation, and communicate lik
involvement and displays of positive affect ing, with individuals moving toward
combine to produce intimate interactions. people and things they like, and away from
Studies demonstrate that both participants people and things they dislike. On the basis
and observers perceive people to be more of these characteristics, Mehrabian’s con
intimate when they use involvement behav ceptualization of immediacy seems to cap
iors such as eye contact and forward leans, ture intensity (or involvement) and liking
coupled with positive affect cues such as (or warm feelings). Andersen (1985) added
smiling (e.g., Burgoon, Buller, Hale, & that immediacy behaviors increase physi
deTurck, 1984; Burgoon & Le Poire, 1999; cal and psychological closeness, signal
Burgoon & Newton, 1991). availability for interaction, are physiolog
Researchers have also identified a host ically arousing, and communicate positive
of behaviors associated with positive affect affect.
and interpersonal warmth (see Andersen & Mehrabian’s (1967, 1969b) work
Guerrero, 1998, for a review). Interpersonal focused on five immediacy behaviors:
warmth is part of a cluster of social emotions interpersonal distance, touch, gaze, body
related to intimacy and affection that orientation, and lean. Andersen (1985)
includes love, liking, and joy. These emo expanded the domain of immediacy behav
tions are expressed via positive involvement iors to include a wide variety of kinesic,
(Guerrero & Andersen, 2000; Guerrero & vocalic, and chronemic cues (e.g., smiling,
Floyd, 2006). Love is communicated non- warm vocal tone, and time spent together).
verbally by physical closeness, positive As noted earlier, some scholars view imme
touch, smiling, mutual gaze, spending time diacy as one of several dimensions under
together, warm vocal tones, and giving gifts the broader construct of involvement
14-Manusov.qxd 6/30/2006 5:06 PM Page 265
(Burgoon & Newton, 1991; Dillard et al., chronemics. When discussing the behaviors
1999). For these scholars, immediacy is associated with each of these subcodes, we
limited to involvement behaviors reflect also comment on relational context vari
ing physical and psychological closeness ables (such as type and stage of relationship)
between two people during interaction, that influence the display and interpretation
with Mehrabian’s (1969b) original list of of various positive involvement cues.
immediacy behaviors—distance, touch,
gaze, body orientation, and lean—providing Proxemics. At least four types of proxemic
the best exemplars. Regardless of the behavior are related to the experience or
approach taken, however, scholars study communication of intimacy: interpersonal
ing involvement and immediacy appear to distance, lean, body orientation, and the
agree that behaviors representing the inter physical plane (Andersen, 1999). Interper
section of involvement and positive affect sonal distance refers to the physical space
are the building blocks of intimate inter between two people. Hall’s (1966) classic
action. Like Prager (1995), we refer to these work on conversational distances suggests
behaviors as positive involvement cues. that in North American culture, the dis
tance ranging from 0 to 18 inches is the
“intimate” zone, reserved generally for inti
Specific Behaviors Reflecting mate interaction with close relational part
Positive Involvement
ners. Mehrabian (1969a) contended that
Next, we identify specific nonverbal close interpersonal distances are related to
behaviors that have been categorized as liking under most circumstances, and early
positive involvement cues. This discussion is empirical research showed that close dis
organized by examining various subcodes tances are related to positive interpersonal
of nonverbal communication, such as prox attitudes (Mehrabian & Ksionsky, 1970),
emics and haptics. Our review focuses on liking, and friendship (Priest & Sawyer,
dynamic nonverbal behaviors rather than 1967).
nonverbal cues that are typically static, Mehrabian (1969a) also suggested,
such as appearance and the environment. however, that close distances can lead to
Although cues related to appearance and less liking and intimacy under certain cir
the environment undoubtedly shape percep cumstances. Burgoon and her colleagues
tions and intimate experiences, we believe (e.g., Burgoon & Hale, 1988; Burgoon,
static cues have less potential for influencing Manusov, Mineo, & Hale, 1985) have
the ongoing process of intimate interaction shown that close distance only produces
than do dynamic cues such as touch and increased liking and positive perceptions
gesturing, which frequently change through if a person is perceived to be rewarding
out the course of interaction (Guerrero & (e.g., attractive, high status). When some
Floyd, 2006). This perspective is consistent one judged as nonrewarding gets close,
with work by scholars studying involve people tend to evaluate that person more
ment and immediacy; although these schol negatively. Research shows that relation
ars have not differentiated between dynamic ships also matter. Morton (1977), for
and static nonverbal cues, the behaviors example, verified that people are less com
they have identified as constituting involve fortable standing or sitting close to
ment and immediacy have tended to fall a stranger than a friend. Guerrero (1997)
under the subcodes of proxemics, haptics, revealed that people tend to sit closer
kinesics, vocalics, and to a lesser extent, to romantic partners than to friends.
14-Manusov.qxd 6/30/2006 5:06 PM Page 266
266–––◆–––Functions
Together, these studies suggest that inter vulnerable body parts such as the face or
personal distance reflects the intimacy level torso “always . . . experienced as intimate”
of relationships and that relational context (Prager, 2000, p. 233).
helps determine how people interpret and Despite the consistency with which touch
respond to positive involvement behavior. is perceived as intimate, touch may function
Lean, body orientation, and the physical to express intimacy differentially depending
plane decrease the vertical and horizontal on the stage of a relationship. In field stud
distance between people. Forward leans ies conducted at airports, zoos, and theaters,
communicate immediacy, involvement, and for example, more touch was associated
affection and tend to lead to perceptions with feelings of greater intimacy during rela
of greater intimacy (e.g., Burgoon, 1991; tionship escalation (Guerrero & Andersen,
Palmer, Cappella, Patterson, & Churchill, 1991; Heslin & Boss, 1980; McDaniel &
1990; Ray & Floyd, 2000). People comfort Andersen, 1998). In these studies, touch was
able with intimacy expression are more observed unobtrusively and correlated posi
likely to lean forward than are those fearful tively to participants’ or observers’ ratings
of intimacy (Guerrero, 1996). Face-to-face of intimacy in developing relationships. In
body orientation is related to intimacy long-term relationships, touch tends to level
(Andersen, 1999). In one study, couples who off or decline. Likewise, in a study of dating
expressed support for each other during dis and married couples, Emmers and Dindia
agreements (with support potentially being a (1995) found a nonlinear relationship
reflection of intimacy) were more likely to between touch and relational intimacy in
use direct body orientation and close dis self-reported private touch, with touch
tancing (Newton & Burgoon, 1990). Studies peaking and then leveling off or decreasing
suggest that women friends are especially slightly at the highest reported levels of
likely to use direct body orientation to intimacy. These studies suggest that touch is
express intimacy (e.g., Guerrero, 1997). more than a reflection of intimate experi
Communicating on the same physical plane ence. Instead, touch may be an essential part
also reduces height differentials, leading to of developing and escalating intimate rela
more intimate interaction (Andersen, 1999). tionships. Once the experience of intimacy is
stable in longer-term relationships, touch
Haptics. Touch is vital to human develop may become less necessary.
ment; loving physical contact, for example, Touch also appears to be important
enables children to reach full social and in communicating intimacy across a variety
intellectual potential and helps them become of relationship types. Monsour (1992) arg
comfortable with intimacy (Guerrero, 2000; ued that physical contact is essential to
Montagu, 1978). Some scholars have even perceptions of intimate interaction for
suggested that intimacy is impossible in the friends. Similarly, Marston, Hecht, Manke,
absence of touch (e.g., Morris, 1971). McDaniel, and Reeder (1998) found
Research has confirmed that across many that tactile behavior (e.g., embracing, hug
contexts, including friendships, romances, ging, or kissing) is the primary way people
family relationships, and therapeutic and communicate intimacy in romantic rela
medical treatments, touch is associated tionships. The criss-cross hug may be per
with intimacy. Indeed, Prager (1995, 2000) ceived as especially intimate across various
considers touch to be a fundamental types of relationships (Floyd, 1999).
component of intimate interaction. As she Studies comparing other types of non
put it, “Touch further intensifies the experi sexual touch suggest that face touching is
ence of intimacy” with welcome touch on seen as particularly intimate, presumably
14-Manusov.qxd 6/30/2006 5:06 PM Page 267
because the face is a sensitive and vulnera behavior, body movements, and gestures
ble part of the body (Burgoon, 1991; Lee & (Burgoon et al., 1996). Intimate kinesic
Guerrero, 2001), which comports with expressions include smiling, facial pleasant
Prager’s (1995, 2000) argument that inti ness, increased eye contact, and gestures that
macy requires one to become vulnerable. In connote immediacy, affection, closeness, and
Burgoon et al.’s (1984) study, observers warmth (Kleinke, 1986). The face is consid
who viewed videotaped tactile interactions ered the primary and most trusted source of
rated touch conditions as more intimate emotional information (Ekman & Friesen,
than nontouch conditions. Because touch is 1975; Knapp & Hall, 2006; Planalp et al.,
considered an intimate behavior, however, 1996). Therefore, the face carries important
interactants must be sure that touch is messages of positive affect that help create
welcome (Andersen, 1998; Prager, 2000). intimate interaction. Burgoon and Newton
Studies show that some people are highly (1991) found that both facial and gestural
touch avoidant and do not like intimate animation predict relational intimacy.
touch, will not volunteer for touch experi Additionally, when couples are intimate,
ments, stand and sit out of reach of other they also tend to synchronize their gestures,
interactants, and dislike a situation if they body movements, and facial expressions
are touched (see Andersen, 2005). Touch (Tickle-Degnen & Rosenthal, 1990; see also
avoidant people will adjust to greater levels Tickle-Degnen, this volume).
of touch, however, depending on rela Eye behavior typically is crucial to
tional stage and their partner’s tactile pref the experience of intimacy. The primary
erences. In short, the relationship trumps oculesic behaviors are gaze (i.e., a person
the trait when the two conflict (Guerrero looks at another person) and eye contact
& Andersen, 1991, 1994), illustrating the (i.e., simultaneous gaze). The latter is par
importance of relational context. ticularly important for initiating social
Of course, touch is inappropriate in some interactions because it serves as an invita
relationships. Interpretations of excessively tion to communicate and is vital to attribu
intimate behaviors, such as touch, can lead tions of intimacy (Andersen, 1985). Argyle
to employees’ reports of sexual harassment. (1972) and Breed (1972) found mutually
Lee and Guerrero (2001) found that among causal relationships between eye contact
ambiguous touches, a gentle facial touch and intimacy. Eye contact is essential for
or an arm around a coworker’s waist were communicating positive involvement and
rated as the most intimate and, therefore, fostering the experience of intimacy in face
inappropriate for coworkers. In another to-face contexts (Andersen & Andersen,
study, soft touches were viewed as less harm 1984; Mehrabian, 1981). Experiments by
ful and harassing than were hard touches Burgoon, Coker, and Coker (1986) showed
(Black & Gold, 2003). Whether touch is that less eye contact had a negative effect on
perceived as sexually harassing is not just a perceived intimacy compared with normal
function of the area touched and the type of or greater levels of eye contact.
touch, but also of the physical attractiveness
of the transgressor (Black & Gold, 2003; Vocalics. Most of the research focusing on
Cartar, Hickes, & Slane, 1996). Generally, vocal cues has not addressed their poten
the more attractive the transgressor, the tial for communicating intimacy. Rather,
more acceptable the tactile behavior is rated. research has focused on the related
constructs of affect, liking, closeness, imme
Kinesics. Kinesics encompass nonverbal diacy, and affection. Next to the face,
behaviors that include facial expressions, eye the vocal channel is the key medium for
14-Manusov.qxd 6/30/2006 5:06 PM Page 268
268–––◆–––Functions
desperation (Wertin & Andersen, 1996). and how much shared understanding one
Similarly, dissynchronous behaviors such perceives initially, is likely to exert a strong
as arriving late for a meeting or social influence on how people react to positive
engagement communicate less friendliness involvement behavior. The second condi
and sociability (Baxter & Ward, 1975). tion revolves around the concept of reci
procity. If one person engages in positive
involvement behavior and the other person
Interpreting Positive Involvement
Cues as a Reflection of Intimacy responds by compensating (e.g., backing
away) rather than reciprocating (e.g., smil
Intensifications of positive involvement ing and increasing eye contact), intimate
behavior usually have a direct, positive rela interaction has not occurred. It is our
tionship with experienced intimacy. This position that intimate interaction requires
is the position of the direct effects model participation by both partners rather than
(Andersen, 1985, 1999) and a related per the simple display of positive involvement
spective, the social meaning model (Burgoon behavior by one partner.
et al., 1986), both of which have been widely Our position regarding reciprocity is
supported. The direct effects model posits supported by scholarship on intimacy as
that although immediacy (or positive well as theories of nonverbal communica
involvement behavior) is moderated by tion. In Prager’s (2000) model of intimacy,
situational, relational, cultural, and person partners are viewed as highly interdepen
ality factors, in most circumstances greater dent. Prager described the interdependent
nonverbal immediacy inherently produces processes underlying intimate interaction
positive person perceptions and greater rela this way: “The experiences (feelings and per
tional intimacy (Andersen, 1998, 1999). ceptions) that Partner A’s behavior elicits in
The social meaning model is based on the Partner B prompt Partner B to behave inti
principle that there are consensually recog mately. The intimate behaviors performed
nized meanings for nonverbal communi by Partner B then shape the experiences of
cation within social communities or the Partner A and so on” (p. 230). As this sce
broader society (Burgoon & Newton, 1991; nario illustrates, the creation of intimate
Burgoon et al., 1985). Research on the interaction requires individuals to express
social meaning model has confirmed that their intimate feelings to each other, as well
people interpret positive involvement cues as to respond positively to one another’s
as a reflection of intimacy (Burgoon & expressions of intimacy. Theories focusing
Le Poire, 1999; Burgoon & Newton, 1991). on patterns of nonverbal communication
Positive involvement behavior is, how come to similar conclusions. For example,
ever, most likely to lead to intimate inte according to Andersen’s (1985, 1998) cog
ractions and intimate relationships when nitive valence theory, positive reactions to
two conditions are met. The first of these immediacy behavior lead to reciprocity and
conditions—that the behavior is welcome— increased intimacy. Similarly, in Cappella
was alluded to earlier in this chapter. and Greene’s (1982) discrepancy arousal
Behaviors such as touch and close proxemic theory, individuals who feel positive emo
distancing, in particular, can be interpreted tion in response to a partner’s increase in
as threatening or aggressive if uninvited and expressive, warm behavior are theorized to
unwanted, even if positive affect cues are reciprocate by engaging in similarly expres
present. The experience of intimacy, in sive and warm behavior that creates inti
terms of how much positive affect one feels mate interaction. Research stemming from
14-Manusov.qxd 6/30/2006 5:06 PM Page 270
270–––◆–––Functions
272–––◆–––Functions
274–––◆–––Functions
276–––◆–––Functions
Patterson, M. L. (1988). Functions of nonverbal Reis, H. T., & Shaver, P. (1988). Intimacy as
behavior in close relationships. In S. Duck an interpersonal process. In S. Duck (Ed.),
(Ed.), Handbook of personal relationships Handbook of personal relationships
(pp. 41–56). New York: Wiley. (pp. 367–389). New York: Wiley.
Planalp, S., DeFrancisco, V. I., & Rutherford, D. Shaver, P. R., Schwartz, J., Kirson, D., &
(1996). Varieties of cues to emotion in nat O’Connor, C. (1987). Emotion knowl
urally occurring situations. Cognition and edge: Further explorations of a prototype
Emotion, 10, 137–153. approach. Journal of Personality and Social
Prager, K. J. (1995). The psychology of inti Psychology, 52, 1061–1086.
macy. New York: Guilford. Simpson, J. A., Orina, M., & Ickes, W. (2003).
Prager, K. J. (2000). Intimacy in personal rela When accuracy hurts and when it helps: A
tionships. In C. Hendrick & S. S. Hendrick test of the empathic accuracy model in mar
(Eds.), Close relationships: A sourcebook ital interactions. Journal of Personality and
(pp. 229–242). Thousand Oaks, CA: Sage. Social Psychology, 85, 881–893.
Prager, K. J., & Roberts, L. J. (2004). Deep Stern, D. (1980). The first relationship: Mother
intimate connection: Self and intimacy in and infant. Cambridge, MA: Harvard
couple relationships. In D. J. Mashek & University Press.
A. P. Aron (Eds.), Handbook of closeness Sternberg, R. J. (1986). A triangular theory of
and intimacy (pp. 43–60). Mahwah, NJ: love. Psychological Review, 93, 119–135.
Erlbaum. Tickle-Degnen, L., & Rosenthal, R. (1990). The
Priest, R. F., & Sawyer, J. (1967). Proximity and nature of rapport and nonverbal correlates.
peership: Bases of balance in interpersonal Psychology Inquiry, 1, 285–293.
attraction. American Journal of Sociology, Wertin, L., & Andersen, P. A. (1996, February).
72, 633–649. Cognitive schema and the perceptions of
Ray, G. B., & Floyd, K. (2000, May). Nonverbal sexual harassment: A test of cognitive
expressions of liking and disliking in initial valence theory. Paper presented at the
interaction: Encoding and decoding per annual meeting of the Western States
spectives. Paper presented at the meeting Communication Association, Pasadena, CA.
of the Eastern States Communication Asso Wilmot, W. W. (1995). Relational communica
ciation, Pittsburgh, PA. tion. New York: McGraw-Hill.
14-Manusov.qxd 6/30/2006 5:06 PM Page 278
15-Manusov.qxd 6/30/2006 5:06 PM Page 279
15
NONVERBAL EXPRESSIONS OF
DOMINANCE AND POWER IN
HUMAN RELATIONSHIPS
� Judee K. Burgoon
University of Arizona
� Norah E. Dunbar
California State University, Long Beach
◆ 279
15-Manusov.qxd 6/30/2006 5:06 PM Page 280
280–––◆–––Functions
and maintaining social order. And, as with Koch, 1998). This chapter is meant to
so many other aspects of social behavior, further that discussion.
much of this work is done nonverbally. It is
the nonverbal signals of dominance-
submission and power-powerlessness that ♦ Conceptualizing Power,
constitute the focus of this chapter.
Dominance, and Status
The social significance of nonverbal
expressions of dominance and power is
underscored by the panoply of scientific Decades of scholarly debate on what con
research and commentary it has attrac stitutes power have produced a consensual
ted. Ellyson and Dovidio (1985) note such view of power as the capacity to produce
diverse tributaries as Charles Darwin’s intended effects, and in particular, the
The Expression of Emotion in Man and ability to influence the behavior of another
Animals (1872/1965), Sigmund Freud’s person (Berger, 1994; Burgoon et al.,
psychoanalytic treatises, works by person 1998; Dunbar, 2004; French & Raven,
ality and social psychologists in the early 1959). Power takes many forms, and
1900s, and anthropological works by its multidimensional nature is reflected in
Ray Birdwhistell (1970) and Edward Hall the classification of power into three
(1959, 1966). A confluence of research on domains: power bases, power processes,
such interrelated constructs as status, and power outcomes (Olson & Cromwell,
authority, rank, control, influence, exper 1975). Power bases refer to resources such
tise, leadership, domineeringness, assertive as rewards or knowledge that form the
ness, and aggressiveness has galvanized foundation for control over others (French
interest further in this fundamental social & Raven, 1959). Several of the nonverbal
dimension. Though these various con display patterns to be discussed in this
structs are not synonymous, they all fall chapter are linked to these bases of power.
under what Edinger and Patterson (1983) Power processes, on the other hand, are
referred to as the social control aspects of the specific strategies (often nonverbal)
interaction and what Hall, Coats, and used to exert power in interactions. Power
Smith LeBeau (2005) labeled the vertical outcomes are the compliance, conformity,
dimension of human relationships. cooperation, or obedience that one
In this chapter, we conceptualize domi secures; that is, they are the actual influ
nance and power as incorporating not ence that is achieved over others’ beliefs
only reflexive, fixed action patterns that and actions (Wheeless, Barraclough, &
are under the control of external stimuli Stewart, 1983).
but also deliberate, adaptive, and change The conceptualization of dominance has
able ones that are under the control of varied according to disciplinary perspective.
actors themselves (consistent with some For personality psychologists, dominance is
ethologists and behavioral ecologists, e.g., considered an enduring individual trait that
Bernstein, 1980; Fridlund, 1991a, 1991b; designates one’s characteristic temperament
Liska, 1992). Greater emphasis on the and behavioral predispositions (e.g., Cattell,
strategic aspects of dominance squares Eber, & Tatsuoka, 1970; Ridgeway, 1987).
with our writings on interpersonal domi Social skills are part of this equation, as the
nance as social, interactional, situational, ability to be forceful, to take initiative, and
intentional, and dynamic (Burgoon & to be expressive yet relaxed and poised
Dunbar, 2000; Burgoon, Johnson, & are all facets of dominance displays that
15-Manusov.qxd 6/30/2006 5:06 PM Page 281
282–––◆–––Functions
284–––◆–––Functions
act to maximize their interpersonal rewards actions, and perceived by their partners as
and minimize their interpersonal costs nondominant.
(Rusbult & Arriaga, 1997; Thibaut &
Kelley, 1959). Social exchange theories
view power as a characteristic of a relation GENDER POLITICS
ship, not an individual. Power is achieved
dyadically when a person is valued as an Another social model of power is what
exchange partner and there are few alterna has become known as the “gender politics
tives; people depend more on partners who hypothesis” (Henley, 1977, 1995; LaFrance
hold high exchange value (Emerson, 1962). & Henley, 1993; also see Hall, this volume).
This power-dependence relationship may In this model, nonverbal behaviors are a
be expressed nonverbally through influence primary means by which those in positions
strategies related to making oneself appear of power (usually men) exercise social con
more attractive as a partner or signaling trol and interpersonal dominance. Henley
one’s interest or disinterest in an exchange argued that many behaviors that may seem
relationship (Burgoon, Dunbar, & Segrin, unimportant are actually reflections of soci
2002). etal biases founded in power differences. For
One social exchange theory that focuses example, she claimed that women are more
on nonverbal cues specifically is dyadic likely to exhibit circumspect demeanor,
power theory (DPT) (Dunbar, 2004; tense posture, gaze aversion or vigilant
Rollins & Bahr, 1976). It proposes that watching, more smiling, touch avoidance,
perceptions of legitimate authority and and greater emotional expressivity. Henley
access to resources increase individuals’ per drew a parallel between the behaviors asso
ceptions of power compared with interac ciated with status and the behaviors associ
tion partners. Perceptions of power, in turn, ated with sex: “The same behaviors
influence the use of dominant communica exhibited by superior to subordinate are
tion to control the interaction, which those exhibited by men to women; and
results in greater influence over decisions. women exhibit to men the behaviors typical
In DPT, the relationship between power of subordinate to superior” (Henley, 1977,
and dominance is theorized to be cur p. 180). Undergirding the gender politics
vilinear, because power is sometimes latent hypothesis are three key premises: (1) that
rather than overt (Komter, 1989). For nonverbal differences between men and
example, powerless individuals may remain women are substantial; (2) that observed
silent if they fear retaliation or termination nonverbal patterns reflect social disparities;
of the relationship from their more power that is, nonverbal patterns are systematically
ful partner (Cloven & Roloff, 1993; Leung, correlated with degree of power or domi
1988). On the other hand, extremely pow nance; and (3) those who occupy the sub
erful individuals may maintain control ordinate role (usually women) are more
without ever having to initiate any control socially perceptive and vigilant by virtue
attempts (Bugental & Shennum, 2002; of their position of greater weakness and
Christensen & Heavey, 1990; Dunbar, vulnerability.
2004). In a test of DPT, Dunbar and Research has documented large sex or
Burgoon (2005a) found that the most pow gender differences in nonverbal behavior.
erful individuals (i.e., with the most influ For example, several reviews by Hall (1998,
ence) were the most facially and gesturally this volume; Hall et al., 2005) and others
expressive, the least controlled in their body (e.g., Andersen, 1998; Burgoon & Dillman,
15-Manusov.qxd 6/30/2006 5:06 PM Page 285
1995; LaFrance, Hecht, & Levy Paluck, around expectations that establish a “power
2003; Riggio, this volume) have found and prestige order” (Berger, Conner, &
that, compared with men, women generally Fisek, 1974; Ridgeway & Berger, 1986;
smile, gaze, nod, and gesture more; use Ridgeway & Walker, 1995). According to
more direct body orientation; receive but do this theory, group members develop expec
not give more touch; are less relaxed postur tations about others’ likely contributions to
ally but also exhibit less shifting of body and the task based on status characteristics, and
feet; have more speech errors, and give more these performance expectations confer an
back-channel responses. Women are also “expectation advantage or disadvantage,”
more skilled decoders and encoders of non depending on whether the individual is
verbal and emotional communication and expected to contribute favorably or unfa
exhibit nonverbal behaviors indicative of vorably to successful task completion.
docility and openness to others. Status characteristics within this model
Nevertheless, many of the empirical find involve any quality of actors around which
ings do not comport with the gender politics evaluations of and beliefs about them come
hypothesis. For example, the meta-analysis to be organized (e.g., age, sex, race, ethnicity,
by Hall et al. (2005) found only four behav education, occupation, physical attractive
iors that were associated with higher actual ness, and intelligence; Berger, Rosenholtz, &
verticality (their term for dominance, status, Zelditch, 1980). Furthermore, expectation
and power): closer physical distances to states theory differentiates between specific
others, using more open body postures, and diffuse status characteristics. Specific
interrupting more, and speaking more status characteristics are socially valued
loudly (see also, Hall, Rosip, Smith LeBeau, skills, expertise, or social accomplishments
Horgan, & Carter, 2006; Snodgrass, Hecht, that imply a specific and bounded range of
& Ploutz-Snyder, 1998). The great hetero competencies, such as computer or mathe
geneity in results and lack of parallelism matical skills. Diffuse status characteristics,
between nonverbal gender differences and such as gender or race, not only are associ
power, dominance, or status have led many ated culturally with some specific skills but
scholars to conclude that research results on also carry general expectations for compe
the gender politics hypothesis are equivocal, tence that are diffuse and unbounded in
that there are more commonalities than range (Ridgeway & Walker, 1995). Many of
differences between the sexes in dominance these characteristics are signaled nonverbally
displays, that purported sex differences through demeanor and appearance, making
are often based on stereotypes, and that this theory especially relevant to nonverbal
observed differences may not be attributable dominance. Those who possess status-
to women holding more subservient roles valued external characteristics “are more
(Burgoon & Dillman, 1995; Dindia & likely (1) to have chances to perform, (2) to
Canary, 2006; Hall et al., 2005). initiate problem-solving performances, (3)
have their performances positively evaluated,
and (4) are less likely to be influenced when
EXPECTANCY THEORIES there are disagreements” (Berger, Ridgeway,
Fisek, & Norman, 1998, p. 381) than
Other models of power draw upon those lacking such characteristics or those
expectations. For example, expectation who possess negatively valued ones.
states theory, which focuses on influence Related to performance expectations
and task performance in groups, revolves are reward expectations: expectations
15-Manusov.qxd 6/30/2006 5:06 PM Page 286
286–––◆–––Functions
about whether the status characteristics are individual that are appropriate, desired, or
more or less likely to create benefits for preferred. Deviant or unexpected behaviors,
individual perceivers or the group. Three by virtue of their novelty, can heighten
classes of reward structures that have attention to the violation and the commu
been identified are categorical, ability, and nicator committing the violation. Reward
outcome (Ridgeway & Berger, 1986). valence refers to whether interactions with
Categorical structures are related to diffuse the communicator are viewed as desirable
social status characteristics such as age, or not. Someone who is physically attrac
gender, or physical strength. These expecta tive, has high status, controls valued
tions are like the physical attractiveness resources, or gives positive feedback, for
stereotypes discussed earlier in their ability instance, should be more positively regarded
to engender attraction and confer credibil than someone who is physically repulsive,
ity. Ability structures are associated with has lower status, controls nothing of
the specific task to be performed. Speaking value, or gives negative feedback. Tests of
with an authoritative voice or using dra EVT have demonstrated that for highly
matic gestures may imply greater confi rewarding communicators, violations of
dence and expertise (i.e., greater ability). expectancies can engender more positive
Outcome structures are associated with interpretations of ambiguous or polysemic
actual accomplishments during the group nonverbal behaviors, such that positive vio
task. Those with high expectation advan lations produce more favorable outcomes
tages not only are more likely to take than expectancy confirmations. The same
the initiative (talking first, establishing act committed by low-reward communica
seating arrangements, etc.) and to be more tors can backfire, eliciting uncharitable
participative but also are likely to be evaluations that make it a negative violation.
accorded more deferential treatment by For low-reward communicators, expectancy
others. In this manner, they will have more confirmation is therefore best (Burgoon,
of their recommendations acknowledged 1991; Burgoon, Newton, Walther, &
and accepted. Baesler, 1989; Le Poire & Burgoon, 1994).
Another expectancy-based theory is Nonverbal indicators of status, power, and
expectancy violations theory (EVT), dominance may be understood in many
which addresses the effects of noticeable contexts from an EVT frame to the extent
deviations from both societal and individ that they impinge on reward valence, con
ual expectations for nonverbal communi stitute expectancy violations, or are the
cation (Burgoon, 1978, 1995; Burgoon & result of violations.
Burgoon, 2001). EVT is relevant to nonver
bal dominance and power in several respects.
Nonverbal behaviors are the locus of the SUMMARY
expectations (or their violations). Inter
pretations of nonverbal behavior may include Scholars have identified many theoreti
dominance connotations, and effects may cal explanations for differences in status,
include perceived power and actual influ dominance, or power. Whereas some theo
ence. Additionally, effects may be moderated rists place the emphasis on personality or
by characteristics of the actor that include his individual characteristics, many scholars
or her status, dominance, and power. consider perceptions of power and the
In EVT, expectancies are enduring pat resultant dominance displays to be interper
terns of anticipated behavior for a particular sonal or relational in nature. They are often
15-Manusov.qxd 6/30/2006 5:06 PM Page 287
influenced by context in that interactants’ These displays map onto the most primitive
past experiential history or larger societal fight-flight and approach-avoidance response
norms can influence the dynamic inter patterns (Keltner, Gruenfeld, & Anderson,
play of dominance-submission. This view 2003). Threats may signal not only readi
allows us to examine the particular ways in ness to strike but also the ability to fight
which power-powerlessness or dominance- if provoked (Maynard Smith, 1982).
submission are typically manifest in inter Less indicative of impending physical
action. The next section introduces three aggression, but still intimidating, are threat
general principles for the nonverbal expres stares and penetrating gazes (Ellsworth,
sion of dominance and power. Carlsmith, & Henson, 1972; Exline,
Ellyson, & Long, 1975; Le Poire &
Burgoon, 1994). Although Hall et al.
♦ Nonverbal Strategies (2005) found that actual gaze was not asso
for Signaling Dominance- ciated with their dimension of verticality, it
Submission and Power- is possible that distinguishing among glares,
Powerlessness stares, and timing of breaks in eye contact
might produce differences. Hinde (1978,
1985) proposed that threat displays show
Based on an extensive survey of the empiri conflicting intentions to attack and to
cal nonverbal literature, Burgoon and col escape. Such ambivalence and the likely
leagues (Burgoon, Buller, & Woodall, 1996; development of finely graded rather than
Burgoon & Hoobler, 2002) have proposed crude signals of threat may account for
a number of principles (rules or abstrac mixed empirical findings.
tions) underlying the nonverbal expression Another symbolic form of threat comes
of dominance and power. We group these from silence, which can convert its target
principles into three general categories: (1) from personhood to nonperson (or object)
physical potency, (2) resource control, and status, with concomitant loss of belonging
(3) interaction control. ness and protection accorded members of
the same social unit. As explained by
Bruneau (1973), the “silent treatment” can
PHYSICAL POTENCY force
288–––◆–––Functions
crouching, drawing the head into the responses to visible physical attributes of
shoulders, a hesitant gait, a slow and ten potential harm or harmlessness that can
tative approach, retreating body orienta be deciphered from a safe distance. Mature
tions, and exposing vulnerable body faces are broader ones with square jaws,
regions such as the jugular vein or palm larger noses, more prominent eyebrows,
(Mehrabian, 1981). Targets of impending thinner lips, and smaller ratios of eye size to
attack may rely on these fright signals to face size, whereas baby-faced features
show submission symbolically. Those low included more rounded and softer features,
in power and in subservient roles may also smaller noses, less pronounced brows and
exhibit far less expressivity due to a host of eyebrows, larger lips, and larger eye sizes
inhibitory and avoidant tendencies associ relative to total face size (Keating, 1985,
ated with the position of vulnerability this volume; Rhodes & Zebrowitz, 2002;
(Keltner et al., 2003). Zebrowitz, Fellous, Mignault, & And
reoletti, 2003). Baby-faced and smiling
Size and Strength. All species appear to faces are seen as submissive, weak, and
respond to speed, agility, and energy expen helpless; mature and unsmiling or frowning
diture as indicators of potency. By contrast, faces with furrowed brows are seen as dom
lethargy and torpidity are associated typi inant, threatening or aggressive. Similarly,
cally with weakness and ineffectualness. high-pitched, thin voices are babyish;
Thus, any nonverbal action that entails a deep-pitched, louder, and more resonant
high degree of intensity and dynamic action ones are considered mature (Montepare &
is likely to connote power and to secure Zebrowitz-McArthur, 1987).
avoidance, flight, or submission from less
dominant others. In the human repertoire, Expressivity. In addition to threat and
height, weight, bulk, and muscularity may size, dominance has been associated with
signal sufficient physical strength and more energetic and animated behavior
endurance to render the displayer victorious such as variable facial expressions,
in a physical conflict. Other nonverbal inflected speech, high-pitched voice, head
behaviors that connote size or strength are shaking and nodding, frequent and broad
rapid gait, erect postures, firm stances, ani gestures, wide smiles, peppy, not sluggish
mated gesturing, loud and deep-pitched or lethargic movement, erect posture,
voices, rapid speaking tempo, clear articula quick movement, upright torso, vertical
tion, and clothing or hair styles that create a sitting posture, more expansive and
bulky appearance (Apple, Streeter & Krauss, emphatic postures and vocalizations, loud
1979; Burgoon, Birk, & Pfau, 1990; Hall voice, vigorous behaviors, heavy step, legs
et al., 2006; Mehrabian, 1981; Schwartz et wide apart, elbows akimbo, hands away
al., 1982). Size and strength can be signaled from body, emphatic and centrifugal ges
also by “strength in numbers,” such as a tures and movements, and more coordi
celebrity’s entourage, a gang member’s com nated nonverbal behavior (e.g., smooth
rades, or an army battalion’s compatriots. voice, fluid and graceful walk, rhythmic
Implicit signals of strength or weakness speech, flowing voice and speech)
are facial and vocal maturity or babyish (Gallaher, 1992). These behaviors con
ness. Nonverbal appearance features that note a high degree of actual or potential
connote maturity or immaturity, and energy expenditure.
actions that emphasize or deemphasize Gifford (1994) found several behaviors
these features, may capitalize on innate associated with perceptions of individuals
15-Manusov.qxd 6/30/2006 5:06 PM Page 289
290–––◆–––Functions
associated with the most and least powerful speaker through eye contact, gesturing, or
in a society. direct verbal address. They also may con
trol the conversational floor by initiating
Elevation. Like physical mass and size that and switching topics, picking up conversa
convey control of the horizontal sphere, tional turns more rapidly, interrupting
height may convey control of vertical space. others, and talking longer (Hall et al., 2005;
Elevated perches give predators an advan Wiemann, 1985), which results in them
tage over their prey; raised thrones, daises, talking more, influencing others more,
theater “box seats,” pedestals, penthouses, and being perceived as leaders (Leffler,
top-floor offices, and prison guard posts Gillespie, & Conaty, 1982).
give people greater surveillance and control
over others. The power bias toward Nonreciprocation. Another form of conver
height “is deeply embedded in the visual sational control is nonreciprocation of
grammar of western civilization. For a others’ nonverbal behavior patterns.
speaker, it has functional advantages. The Whereas people on an equal plane may sig
elevation . . . gives him or her a much larger nal their equality through matching and
field of vision. Elevation gathers and keeps mirroring of another’s kinesic, proxemic,
attention” (King, cited in Jaworski, 1993, and vocalic patterns, dominant individuals
p. 14). Behaviors that increase height differ may meet another’s smile with a blank
entials, such as standing over another expression or counter an expressive voice
person or “looking down at someone,” with a bored one. Dominant individuals
likewise function as dominant behaviors. It may also become the zeitgeber, the one who
follows, then, that tall people would be seen sets the interactional pace and to whom
as more powerful than short people (Frieze, others orient as they attempt to establish
Olson, & Good, 1990). Social rituals, such interactional synchrony (Burgoon, Stern, &
as bowing to the higher status individual in Dillman, 1995). All these patterns work to
Japan (Nixon & West, 1995), have evolved reflect the lack of power symmetry.
to accord elevation symbolically to individ
uals of higher rank. Task Performance Cues. A final principle of
conversational control is a culmination of
Initiation. A corollary to the principle of many of the foregoing principles applied
precedence discussed earlier is initiation. to a task context. Task performance cues
Deciding where people will sit or stand, are nonverbal indicators of status and task-
changing interactional distances, initiat related ability from which group members
ing touch, starting or stopping conversa infer one’s potential to contribute effec
tion, and setting interaction rhythms are all tively to a group’s task performance (Berger
interaction-based extensions of the “going et al., 1980; Ridgeway & Walker, 1995).
first” principle. For example, dominant High-status apparel, possession of artifacts
individuals may initiate handshakes, dictate that are culturally defined status symbols
conversational distancing, and set the (e.g., an expensive briefcase or pen), and
pattern for seated or standing interaction. actions indicative of centrality, privileged
Sheer proximity to, and surveillance of, access, precedence, and prerogative, con
more individuals enables a powerful person note likely expertise, experience, and lead
to dictate who talks to whom and when. By ership. These set up expectations for the
being accorded the privilege of speaking person to contribute favorably to a group’s
first, they then can nominate the next task and set into motion a self-fulfilling
15-Manusov.qxd 6/30/2006 5:06 PM Page 292
292–––◆–––Functions
prophecy whereby such individuals are should not expect simple and unidimen
accorded more and longer speaking turns, sional behavioral profiles of status, domi
are allowed to initiate or change topics, and nance-submission, or power-powerlessness.
thus exert more influence on the group, Moreover, because human actors are not
which then reinforces their prestige and bundles of instincts but rather active agents
position of power. In this manner, who choose among alternative strategies to
the “strong” may become stronger and the accomplish their goals, variability in forms
“weak,” weaker. of dominance and power expressions is to
be expected.
The nonexhaustive review we have
presented here illustrates the various ways
♦ Conclusions
in which nonverbal behaviors can be
enacted to accomplish these ends. These
The expression of dominance-submission include principles for displaying physical
and power-powerlessness is a form of com potency, signifying resource holding poten
munication that has a universal vocabulary. tial, and accomplishing interaction control.
The theories, lines of research, and princi We hope this review promotes insights into
ples for nonverbal display of dominance what behavioral patterns carry dominance-
and power that we have reviewed here submission and power-powerlessness
point to a strong grounding in ethological connotations as well as what theoretical
and related behavioral ecology perspectives. explanations can be advanced for under
Even those forms of expression that have standing their meanings and impact. It
clear social derivations and constitute sym provides a template for future research into
bolic rather than sign behavior often have the critical role played by nonverbal behav
more primitive biological analogs. Thus, ior in negotiating this elemental facet of
the study of nonverbal expressions of social life.
dominance, power, and status offer insights
more generally into societal hierarchies and
social functioning.
♦ References
Notwithstanding the evidence of univer
sality and possible innate foundations for
many power and dominance displays, Anderson, J. R. (1998). Cognitive psychology
we share with Hall et al. (2006) a convic and its implications. New York: Freeman.
tion that such displays are also responsive Andrew, R. J. (1972). The information poten
to culture, context, relationships, social tially available in mammal displays. In
motives, and the psychological makeup R. A. Hinde (Ed.), Non-verbal communica
of actors. This responsivity to a panoply of tion (pp. 179–204). Cambridge, UK:
antecedent factors helps explain what Cambridge University Press.
might otherwise seem to be a dismal record Apple, W., Streeter, L. A., & Krauss, R. M.
(1979). Effects of pitch and speech rate
of nonverbal expressions of status, power,
on personal attributions. Journal of
and dominance in predicting the vertical
Personality and Social Psychology, 37,
dimension of human relationships. The 715–727.
need for flexibility and adaptability for all Aries, E. J., Gold, C., & Weigel, R. H. (1983).
species argues for behavioral repertoires Dispositional and situational influences on
that are rich in alternative forms of expres dominance behavior in small groups.
sion and that can be adjusted or substituted Journal of Personality and Social Psycho
as circumstances demand. We, therefore, logy, 44, 779–786.
15-Manusov.qxd 6/30/2006 5:06 PM Page 293
Berger, C. R. (1994). Power, dominance, Burgoon, J. K., Buller, D. B., Hale, J. L., &
and social interaction. In M. L. Knapp & deTurck, M. (1984). Relational messages
G. R. Miller (Eds.), Handbook of inter associated with nonverbal behaviors.
personal communication (2nd ed., pp. 450– Human Communication Research, 10,
507). Thousand Oaks, CA: Sage. 351–378.
Berger, J., Conner, T. L., & Fisek, M. H. (1974). Burgoon, J. K., Buller, D. B., & Woodall, G. W.
Expectation states theory: A theoreti (1996). Nonverbal communication: The
cal research program. Cambridge, MA: unspoken dialogue. New York: McGraw-
Winthrop. Hill.
Berger, J., Ridgeway, C. L., Fisek, M. H., & Burgoon, J. K., & Burgoon, M. (2001).
Norman, R. Z. (1998). The legitimation Expectancy theories. In P. Robinson &
and delegitimation of power and prestige H. Giles (Eds.), Handbook of language and
orders. American Sociological Review, 63, social psychology (2nd ed., pp. 79–101).
379–405. Sussex, UK: Wiley.
Berger, J., Rosenholtz, S. J., & Zelditch, M., Jr. Burgoon, J. K., & Dillman, L. (1995). Gender,
(1980). States organizing processes. Annual immediacy and nonverbal communication.
Review of Sociology, 6, 479–508. In P. J. Kalbfleisch & M. J. Cody (Eds.),
Bernstein, I. S. (1980). Dominance: A theoreti Gender, power, and communication in
cal perspective for ethologists. In human relationships (pp. 63–81). Hillsdale,
D. R. Omark, F. F. Strayer, & D. G. NJ: Erlbaum.
Freedman (Eds.), Dominance relations Burgoon, J. K., & Dunbar, N. (2000). An interac
(pp. 71–84). New York: Garland. tionist perspective on dominance-submission:
Birdwhistell, R. (1970). Kinesics and context: Interpersonal dominance as a dynamically,
Essays on body motion communication. situationally contingent social skill. Com
Philadelphia: University of Pennsylvania munication Monographs, 67, 96–121.
Press. Burgoon, J. K., Dunbar, N. E., & Segrin, C.
Bruneau, T. J. (1973). Communicative silences: (2002). Nonverbal influence. In M. Pfau &
Forms and functions. Journal of Commun J. P. Dillard (Eds.), The persuasion hand
ication, 23, 17–46. book: Theory and practice (pp. 445–473).
Bugental, D. B., & Shennum, W. (2002). Thousand Oaks, CA: Sage.
Gender, power, and violence in the family. Burgoon, J. K., & Ebesu Hubbard, E. (2005).
Child Maltreatment, 7, 56–64. Expectancy violations theory and interac
Burgoon, J. K. (1978). A communication model tion adaptation theory. In W. Gudykunst
of personal space violations: Explication (Ed.), Theorizing about intercultural com
and an initial test. Human Communication munication (pp. 149–171). Thousand
Research, 4, 129–142. Oaks, CA: Sage.
Burgoon, J. K. (1991). Relational message inter Burgoon, J. K., & Hale, J. L. (1984). The fun
pretations of touch, conversational dis damental topoi of relational communi
tance, and posture. Journal of Nonverbal cation. Communication Monographs, 51,
Behavior, 15, 233–258. 193–214.
Burgoon, J. K. (1995). Cross-cultural and inter Burgoon, J. K., & Hale, J. L. (1988). Nonverbal
cultural applications of expectancy viola expectancy violations: Model elaboration
tions theory. In R. L. Wiseman (Ed.), and application to immediacy behaviors.
Intercultural communication theory (Inter Communication Monographs, 55, 58–79.
national and intercultural communication Burgoon, J. K., & Hoobler, G. (2002).
annual, Vol. 19, pp. 194–214). Thousand Nonverbal signals. In M. L. Knapp &
Oaks, CA: Sage. J. Daly (Ed.), Handbook of interpersonal
Burgoon, J. K., Birk, T., & Pfau, M. (1990). communication (pp. 240–299). Thousand
Nonverbal behaviors, persuasion, and Oaks, CA: Sage.
credibility. Human Communication Burgoon, J. K., Johnson, M. L., & Koch, P. T.
Research, 17, 140–169. (1998). The nature and measurement of
15-Manusov.qxd 6/30/2006 5:06 PM Page 294
294–––◆–––Functions
296–––◆–––Functions
Mehrabian, A. (1981). Silent messages: Implicit Ridgeway, C. L., Diekema, D., & Johnson, C.
communication of emotions and attitudes (1995). Legitimacy, compliance, and gender
(2nd ed.). Belmont, CA: Wadsworth. in peer groups. Social Psychology Quar
Montepare, J. M., & Zebrowitz-McArthur, L. terly, 58, 298–311.
(1987). Perceptions of adults and children Ridgeway, C. L., & Walker, H. A. (1995).
with childlike voices in two cultures. Status structures. In K. S. Cook, G. A. Fine,
Journal of Experimental Social Psychology, & J. S. House (Eds.), Sociological perspec
23, 331–349. tives on social psychology (pp. 281–310).
Moskowitz, D. S. (1988). Cross-situational Boston: Allyn & Bacon.
generality in the laboratory: Dominance Rogers-Millar, E. L., & Millar, F. E. (1979).
and friendliness. Journal of Personality and Domineeringness and dominance: A trans
Social Psychology, 54, 829–839. actional view. Human Communication
Moskowitz, D. S. (1990). Convergence of Research, 5, 238–246.
self-reports and independent observers: Rollins, B. C., & Bahr, S. J. (1976). A theory of
Dominance and friendliness. Journal of power relationships in marriage. Journal of
Personality and Social Psychology, 58, Marriage and the Family, 38, 619–627.
1096–1106. Rusbult, C. E., & Arriaga, X. B. (1997).
Neff, K. D., & Harter, S. (2002). The role of Interdependence theory. In S. Duck (Ed.),
power and authenticity in relationship styles Handbook of personal relationships (2nd
emphasizing autonomy, connectedness, or ed., pp. 221–250). Chichester, UK: Wiley.
mutuality among couples. Journal of Social Schwartz, B., Tesser, A., & Powell, E. (1982).
and Personal Relationships, 19, 835–857. Dominance cues in nonverbal behavior.
Nixon, J. C., & West, J. F. (1995). Intercultural Social Psychology Quarterly, 45, 114–120.
preparation for managers going to Japan. Sebeok, T. (1972). Perspectives in zoosemiotics.
Mid-American Journal of Business, 10, The Hague, The Netherlands: Mouton.
57–64. Siegel, S. M., Friedlander, M. L., &
Olson, D. H., & Cromwell, R. E. (1975). Power Heatherington, L. (1992). Nonverbal rela
in families. In R. E. Cromwell & D. H. tional control in family communica
Olson (Eds.), Power in families (pp. 3–11). tion. Journal of Nonverbal Behavior, 16,
Beverly Hills, CA: Sage. 117–139.
Omark, D. R. (1980). The group: A factor Smith, W. J. (1974). Displays and messages
or an epiphenomenon in evolution. in intraspecific communication. In S. Weitz
In D. R. Omark, F. F. Strayer, & (Ed.), Nonverbal communication: Readings
D. G. Freedman (Eds.), Dominance rela with commentary (pp. 331–340). New
tions (pp. 21–64). New York: Garland. York: Oxford University Press.
Remland, M. (1981). Developing leadership Snodgrass, S. E., Hecht, M. A., & Ploutz-
skills in nonverbal communication: A situa Snyder, R. (1998). Interpersonal sensitivity:
tional perspective. Journal of Business Expressivity or perceptivity? Journal of
Communication, 18, 18–31. Personality and Social Psychology, 74,
Rhodes, G., & Zebrowitz, L. A. (Eds.). (2002). 238–249.
Facial attractiveness: Evolutionary, cognitive, Sommer, R. (1971). Spatial parameters in
and social perspectives. Westport, CT: Ablex. naturalistic research. In A. H. Esser (Ed.),
Ridgeway, C. L. (1987). Nonverbal behavior, Behavior and environment: The use of
dominance, and the basis of status in task space in animals (pp. 281–290). New York:
groups. American Sociological Review, 52, Plenum.
683–694. Thibaut, J. W., & Kelley, H. H. (1959). The
Ridgeway, C. L., & Berger, J. (1986). social psychology of groups. New York:
Expectations, legitimation, and dominance Wiley.
behavior in task groups. American Socio Thorpe, W. H. (1972). The comparison of
logical Review, 62, 218–35. vocal communication in animals and man.
15-Manusov.qxd 6/30/2006 5:06 PM Page 297
In R. A. Hinde (Ed.), Non-verbal commu Street Jr., & J. N. Cappella (Eds.), Sequence
nication (pp. 27–48). Cambridge, UK: and pattern in communicative behaviour
Cambridge University Press. (pp. 85–102). London: Edward Arnold.
Wheeless, L. R., Barraclough, R., & Stewart, R. Zebrowitz, L. A., Fellous, J-M., Mignault, A., &
(1983). Compliance-gaining and power in Andreoletti, C. (2003). Trait impressions as
persuasion. Communication yearbook, 7 overgeneralized responses to adaptively sig
(pp. 105–145). Newbury Park, CA: Sage. nificant facial qualities: Evidence from con
Wiemann, J. M. (1985). Interpersonal control nectionist modeling. Personality and Social
and regulation in conversation. In R. L. Psychology Review, 7, 194–215.
15-Manusov.qxd 6/30/2006 5:06 PM Page 298
16-Manusov.qxd 6/30/2006 5:09 PM Page 299
16
THE FUNCTIONS
OF FACIAL EXPRESSIONS
What’s in a Face?
� Alan J. Fridlund
University of California, Santa Barbara
� James A. Russell
Boston College
◆ 299
16-Manusov.qxd 6/30/2006 5:09 PM Page 300
300–––◆–––Functions
facial paralanguage and suggest that the out” the Cartesian passions, much like the
traditional division between facial paralan face of the clock reads out the time. The
guage and “facial expressions of emotion” is French neurologist Guillaume Duchenne,
illusory and counterproductive. who invented the biopsy and discovered
Duchenne muscular dystrophy, pioneered
the scientific study of facial movement,
♦ Older Views of based on the assumption that the ability to
Facial Expression express emotions was God-given. But the
chief expositor of this theological view of
the facial muscles was the celebrated
History shows the variety of possible Scottish surgeon, anatomist, and artist
interpretations of faces. “Face reading,” or Charles Bell (discoverer of Bell’s palsy), who
physiognomy, goes back to antiquity (see published his celebrated Essays on the
Fridlund, 1994, for a more detailed review of Anatomy of Expression in Painting in 1806.
this history). It existed in ancient Egypt and Bell’s theological view was the implicit tar
Arabia and was a profession in China before get for Charles Darwin, whose Expression of
Confucius. Pythagoras probably originated the Emotions in Man and Animals (1872)
the scientific study of physiognomy, which aimed systematically to replace creationist
Hippocrates and Galen applied to their med accounts of human behavior by provid
ical diagnoses. Aristotle constructed a system ing an onslaught of natural-history ones. As
of physiognomy which likened people’s per Fridlund (1992a) discussed, Darwin (1872)
sonalities to the animals whose faces they countered Bell (1806) by arguing that
resembled. For example, people with promi human facial expressions were not there for
nent upper lips were said to be stupid like expression at all: They were either vestigial
apes and donkeys, and those with “hawk” reflexes or accidents of nervous system
noses were magnanimous. Two thousand wiring. If faces were vestiges or accidents,
years later, the Renaissance figure, then the argument of divine design was
Giambattista Della Porta, joined Aristotle’s undermined. Thus, for Darwin, we don’t
physiognomy with Hippocrates’ system of bare our teeth when we’re angry in order to
temperament, arguing that a person’s face show anger; we just do it because evolution
indicated whether he was sanguine, choleric, ary ancestors tended to bite when they were
phlegmatic, or melancholic. But physiog aggressive, and we simply inherited this out
nomy reached its zenith with the 18th-cen dated habit.
tury Swiss pastor and poet, Johann Caspar In the 20th century, accounts of facial
Lavater, who believed he could recognize the expression bifurcated (Russell & Fernandez-
God in man by divining his subjects’ traits Dols, 1997). Almost no one continued
from the shapes and lines in their faces. Darwin’s (1872) nonadaptationist account.
Much of physiognomy concerned static Those who focused on humans continued to
features of the face, but changes in the face see facial expressions as readouts of
were also thought to convey changes in the the passions, modernized with an evolution
person. Aristotle had read faces as indicat ary patina to say that facial expressions
ing the “passions,” which Descartes under evolved to express the passions. Those who
stood as turbulence in the bodily humors. studied nonhuman animals saw faces as sig
The 17th-century French painter Charles nal movements, which, like other animal
LeBrun, who ruled the arts in France for signals, evolved to convey likely actions and
two generations, argued that the face “reads thereby alter the course of ongoing social
16-Manusov.qxd 6/30/2006 5:09 PM Page 301
302–––◆–––Functions
Because evolutionary accounts of behavior blood through the dense, superficial vascula
amount to historical reconstructions, which ture of the face (and, often, the hands and
account applies is difficult to determine. the chest) exhausts heat. Prolonged flushing
The argument hinges on whether the sign in is accompanied by burning and sweating,
question is exaggerated beyond what would and this combination is not only painful but
be expected of an offshoot, often to the embarrassing (Gerlach, Wilhelm, Gruber, &
point that it becomes a physiological liabil Roth, 2001).
ity or handicap. For example, a male widow Vasodilation can also occur rapidly, as
bird drags along a tail of feathers four to in intense exercise. In blushing, the entire
five times the length of the female’s tail. His face (and usually, the ears and upper neck)
tail adds weight, is metabolically costly to reddens within 2 seconds of the event that
grow, and hinders flight. It does attract evoked it. Blushes last on average about 20
females, and that seems to be its only virtue seconds, although they may endure up to
(Andersson, 1982). Thus, it is reasonable to 15 min (Shields, Mallory, & Simon, 1990).
conclude that the male’s tail evolved to do People blush commonly in various kinds
so and that it is not simply an offshoot of of social interactions. The interactions are
the hormonal requirements of maleness. usually intense ones, such as when we (1)
No comparable clear-cut argument for lash out and yell (“reddening with anger”),
any human static facial feature has been (2) grieve, and (3) are made the subject
made. Perhaps the best candidate is the of undue social attention, whether being
eyebrows, which enhance the visibility of praised (“flushed with pride”), found guilty
our brow movements and—although some of a transgression, or caught off guard and
have conjectured that they are sweat catch made to go off role (e.g., embarrassment
ers (see Porter, 1993)—they have no other and other “self-presentational predica
convincing raison d’etre. Although static ments”; see Schlenker, 1980). As one might
features of the face provide information, predict, blushing is more likely among
much more research has focused on the people with low self-esteem or those who
dynamic features of the face, specifically its are especially sensitive about others’ judg
coloration and muscular movements. ments of them (Leary, Britt, Cutlip, &
Templeton, 1992).
Based on the fact that most blushing
DYNAMIC FEATURES OF THE FACE is social, Leary et al. (1992) suggested that
blushes are, at least in part, intrinsically
Facial Coloration Changes. Faces change communicative. In their account, our
color when the blood vessels constrict or blushes notify others that we care about
dilate, thereby regulating how much blood what they think of us, a strategy that may
flows through the facial muscles and skin. In appease those who would judge us harshly.
facial flushing, all or part of the face reddens, Social blushes can occur in solitude, as well,
often over a period of hours. Facial flushing when we imagine the appropriate social
may occur after changes in ambient temper circumstances, and then place ourselves
ature, eating spicy foods, or with fever. It can before an imaginary audience (see Fridlund,
also occur in a variety of disorders, such as Sabini, et al., 1990; Fridlund, Kenworthy,
rosacea, and after taking certain vasoactive & Jaffey, 1992; for the influence of imagi
substances, like alcohol or niacin. Flushing is nary audiences of facial expressions).
also common during menopause. It appears Whether blushes evolved as a signal
to be thermoregulatory, because running hinges on whether the coloration changes are
16-Manusov.qxd 6/30/2006 5:09 PM Page 303
Corrugator supercilii
Levator labii superioris aleque nasii Galea aponeurotica
Frontalis
Levator palpebrae superioris Procerus
Tarsus superioris Lacrimalis
Tarsus inferioris
Corrugator supercilii
Orbicularis oculi, Orbicularis oculi,
pars palpebralis pars orbitalis
Orbicularis oculi,
Levator labii
pars palpebralis
superioris
Orbicularis oculi,
Orbicularis oculi,
pars ciliaris
pars orbitalis
Septum/Lateral wall
Orbicularis oris
Buccinator
Risorius Risorius
Depressor septi/ Modiolus
Platysma
Orbicularis oris
Buccinator
Modiolus
Depressor anguli oris
Depressor labii Labial Salivary Orbicularis oris
inferioris Glands Mentalis
Figure 16.1 The Superficial (Mimetic) Muscles of the Face Seen From the Inside Out
SOURCE: Reprinted from Pernkopf (1938), with permission from Elsevier.
beyond what would be expected as inciden participant off to one side, the participant’s
tal offshoots of other vascular reactions. cheek temperature was higher on that side.
Specifically, the circumstances that might
make us blush socially are often those in Muscular Changes. Our facial coloration
which “the heat is on,” and when we would changes proceed over periods of seconds
blush as a natural part of thermoregulation. or minutes, but it is the muscular changes
That others might read blush as a sign of in our face that allow it to accomplish split-
that reaction would be incidental. The jury is second signaling and social engagement. Our
still out on the issue, but some evidence of exquisitely complex facial movements are
an intrinsic display component was found caused by fewer than 30 muscles (Figure 16.1)
by Drummond and Mirco (2004), who precisely controlled by more nerves than
measured the cheek temperatures of partici any other muscles save those of the fingers.
pants who were asked either to sing (an The face collects together the organs
embarrassing task) or read aloud (a less- for tasting, smelling, eating (including suck
embarrassing task). During the singing, ling), seeing, and speaking. All these involve
cheek temperatures were symmetrical if the orifices—the mouth, nose, and eyes—that
experimenter watched the participant sing require control and protection. These are
head-on. But if the experimenter watched the the primary function of our facial muscles.
16-Manusov.qxd 6/30/2006 5:09 PM Page 304
304–––◆–––Functions
At least some of the information a face reliability, and validity remains question
provides to an observer stems from these able. Terms such as “smile” and “frown”
simple functions: shouting or whispering, are vague, and we have no guarantee that
looking toward or away or down, sniffing they cover the range of facial behavior
or tasting, orienting or listening—information exhaustively or parse nature at its joints.
given off. The facial muscles also display The Maximally Discriminative Facial
messages—information given. But the same Movement Coding System developed by
question arises for the facial muscles as it Izard (1979) uses theoretically defined cate
did for static features and vascular changes: gories of facial features.
Were the muscles themselves shaped by nat A major advance in facial measurement
ural selection, at least in part, for these was an anatomically based system devised
display purposes? by the anatomist Hjortsjö (1970). His
Here, ironically, the case is harder to system was later revised by Ekman and
make. Any proof that muscles were shaped Friesen (1978) as their Facial Action
for signaling would require demonstrating Coding System (FACS). Specific facial
that they possess shapes or actions that do actions are given “action unit” codes; for
not accord with their physiological func example, brow knitting is an “Action Unit
tions, or perhaps even handicap their phys 4” (AU 4) and cheek-corner retraction is
iological functions (like the piloerection AU 12. A second major technique for facial
of feathers on the peacock’s tail). Muscles measurement is facial electromyography
used in producing speech, for example, (EMG), in which tiny electrodes are affixed
were likely just co-opted (technically, to the skin over major facial muscles. The
preadapted or exapted; Gould & Vrba, electrical signals produced by the contract
1982) because mammalian suckling had ing muscles are then digitized and recorded.
already required fine control and flexibility. The facial EMG technique can render pre
There is also little evidence on this issue. Of cise estimates of even weak muscle contrac
course, humans use facial muscles to create tions, but its limitations include lack of
displays, and the role of natural selection in selectivity (frequently, there is crosstalk
this use is a separate question. from neighboring muscles), reactivity
(participants are usually quite aware of the
Measuring Facial Muscular Changes. What electrodes), and imperfect relationship to
ever their use, studying the facial muscles visual appearance (co-contracting antago
requires measuring them precisely, and this nistic muscles can be electrically quite active
task is difficult. Most of the facial muscles but result in little change in facial appear
originate in bone and insert into skin. ance) (see Fridlund & Cacioppo, 1986).
Because they distend skin, we are seldom
able to measure the contractions of our
facial muscles exactly; we measure instead SUMMARY
the displacement of facial features as an
indication of their contractions. Perhaps the As we have seen, people make a wide
most common method of doing so is direct range of inferences about others from both
observation. In its simplest form, raters the appearance and the movements of faces.
watch people and tally the numbers of By far, most research on facial expression
smiles, frowns, grimaces, pouts, and so on, has focused on the face’s muscular move
in a given period. In this case, many raters ments, under the presupposition that they
are typically needed to achieve adequate “express emotion.” We now turn to the use
16-Manusov.qxd 6/30/2006 5:09 PM Page 305
of facial musculature in expression and the same specific basic emotions to those
evaluate this presupposition. universal facial patterns; and (3) those same
facial patterns are, indeed, manifestations of
those very same emotions in all humans.
♦ The Facial The principal method used in support of this
Expression Program premise was to show people of various cul
tures photographs of posed facial expres
sions (see Figure 16.2) and then ask each
Scientists since the end of the 18th century participant to match the photos with emo
have adopted the ancient and by then tionally laden scenarios or emotion-words.
common-sense assumption that what faces Peoples from many parts of the world were
express are emotions. Indeed, by the 1980s, tested, including, most famously, relatively
the psychology of facial expression was isolated, non-Western and illiterate peoples
dominated by a network of assumptions, from Papua New Guinea.1
theories, and methods that Russell and Russell (1994) reviewed the 34 cross-
Fernandez-Dols (1997) termed the Facial cultural studies that used the standard
Expression Program (FEP). Table 16.1 sum method just described. The overall findings
marizes the FEP. from the matching-to-word design are
What we offer is but a prototype, and so depicted in Figure 16.3. The leftmost set of
no one theorist would support all 12 of the bars comes from Western literate societies
following claims. The FEP guided research (mostly college students). The second set of
on facial expression, indeed on emotion, for bars comes from non-Western societies,
several decades. Some of the basic assump although the participants were also largely
tions have been severely questioned by the college students. The rightmost set is com
data; for evidence against the 12th claim, prised of the findings from the isolated, non-
for example, see Russell and Fehr (1987) Western, illiterate cultures. All groups
and Carroll and Russell (1996). More showed matching of still faces to words at
recently, questions have arisen about all of greater-than-chance rates (Elfenbein &
the assumptions (for reviews see Fridlund, Ambady, 2003; Russell, 1994), but as
1994; Russell, 1994; Russell & Fernandez- the participants’ level of westernization
Dols, 1997). Here we consider two central decreased, so did their matching rates. In the
issues: universality and the relationship of case of the studies involving non-Western,
facial expressions to emotions. isolated, illiterate participants, even the lower
rates are likely to be inflated by a concatena
ARE FACIAL EXPRESSIONS tion of technical shortcomings within the
OF EMOTION UNIVERSAL? studies themselves—namely, that they used a
within-subjects design (known to inflate
There are many arguments regarding matching relative to a between-subjects
the universality of facial expression. design), using a forced-choice response for
Demonstration of Strong Universality as mat (another known source of inflation),
presupposed in the FEP would require three with experimenters who were not blind to the
sets of findings: (1) the same patterns of stimuli or the hypotheses, and with question
facial movement occur in all normal able experimental control (Sorenson, 1975).
humans, with the specific patterns now The studies show something, to be sure,
offered illustrated in Figure 16.2; (2) but rejecting the null hypothesis of ran
observers in different societies attribute dom matching does not confirm Strong
16-Manusov.qxd 6/30/2006 5:09 PM Page 306
306–––◆–––Functions
10. The ability to recognize the emotion in a facial expression is innate rather than culturally
determined, and may be seen as early as just after birth.
11. The mental categories by means of which recognition occurs (in the self as facial
feedback or in others through facial signaling) are genetically rather than culturally
determined, and the emotion-words we use thus designate innate and universal
categories. Other languages may use other names, but the categories named are the
same.
12. The meaning (“signal value”) of a facial expression is fixed by nature and invariant across
changes in the context in which it occurs.
Figure 16.2 Four of the Facial Expressions of Emotion (“Happy,” “Sad,” “Angry,” and “Fear” Faces,
Respectively) That Are Claimed Within the Facial Expression Program to Be Innate,
Universally Produced and Understood, and Tied Intimately to Emotional State
SOURCE: Photographs by Alan Fridlund (model: Jason Fridlund)
(2) facial movements are coordinated with relationship between facial expressions
psychological states, which might include and emotion. At least in Western societies,
actions, preparations for actions, physical people interpret many stimuli—not just
states, emotional states, cognitive states, faces—in terms of emotion: angry storms,
and other psychological conditions (which weeping willows, joyous birdsong, and the
they are would require other evidence); melancholic baying of wolves. Conversely,
(3) most people everywhere can infer some humans interpret faces in many ways, not
thing of another’s psychological state from just as emotions. For example, people
facial movement, just as they can from interpret faces in terms of action tendencies
almost anything else that other person does; (e.g., Frijda & Tcherkassof, 1997) and
and (4) people in Western cultures have a social messages (e.g., Yik & Russell, 1999).
set of beliefs in which specific types of facial The assumption of a privileged relation is
actions are expressions of specific types of that faces and emotions are linked in some
emotions (e.g., the smile is a familiar sym more fundamental and specific way: Speci
bol for happiness). These four premises pre fic “basic” emotions cause specific facial
dict that, when asked, people everywhere expressions. Specific facial expressions were
can form nonrandom associations between shaped by evolution to express precisely
faces and emotions. People in Western cul those specific, basic emotions. What are
tures will be particularly adept at associat these “basic emotions” that are held to
ing faces with emotions. At this time, we cause specific facial expressions and to be
argue that anything beyond Minimal expressed by them? How would one deter
Universality has yet to be proved.2 mine them? Ortony and Turner (1990)
exposed a host of conceptual problems with
the supposition of “basic emotions,” but
IS THE RELATIONSHIP BETWEEN pragmatically, one could imagine sidestep
FACIAL EXPRESSIONS AND ping the complexities. Only three simple
EMOTION PRIVILEGED? steps would be required: (1) develop inclu
sion and exclusion criteria for stipulating
Untested in the famous cross-cultural whether people are having a particular emo
studies was the assumption of a privileged tion. In other words, what specific signs and
16-Manusov.qxd 6/30/2006 5:09 PM Page 308
308–––◆–––Functions
100
60
40
20
0
Western Literate Non-Western Literate Isolated Illiterate
(N = 20) (N = 11) (N = 3)
Culture Group
Expression
“Happy” “Surprise” “Anger”
“Sadness” “Fear” “Disgust”
symptoms must be present to say that the present at once. The absence of agreed upon
emotion is present, and what specific signs inclusion and exclusion criteria for “emotion”
and symptoms must be present to say that it means that the hypothesis of a privileged rela
is absent? This criterion cannot include the tion of emotion to facial expression remains
face, else one would simply be proving a fuzzy. It will remain untestable so long as
tautology; (2) measure what these people do “emotion” is operationally undefinable.
with their faces; and (3) investigate whether Setting aside the fuzzy edges of the con
there is a relationship between Steps 1 and 2. cept of emotion, one runs into difficulty
We argue that Step 3 has not been accom even with the prototypical cases created in
plished, because the prerequisite Step 1 has the laboratory. Empirical examination of
not been accomplished. For some, “emotion” specific emotions as highly coherent pack
means subjective emotional experience; for ages has produced surprisingly weak
others, it means emotional behavior or results. Research has repeatedly uncovered
changes in peripheral physiology; for still weak associations among components
others, it means certain kinds of cognition. of an emotion (Lang, 1968; Lazarus,
For many theorists, “emotion” is a complex Speisman, & Mordkoff, 1963; Mandler,
of all these components, and not all need be Mandler, Kremen, & Sholiton, 1961).
16-Manusov.qxd 6/30/2006 5:09 PM Page 309
In the most sophisticated set of laboratory Such correlations are compatible with many
studies on this topic, Reisenzein (2000) theories about facial expression, including
examined the intercorrelations among those that assume no privileged relation.
four components of surprise: & cognitive Altogether, two major planks of the
appraisal, self-reported experience of sur FEP—Strong Universality and a privile
prise, reaction time, and facial expression. ged link between facial expressions and
Correlations were modest, with the exception emotion—have been shown to be quite
of one relation that is close to a tautology: a controvertible. As a result, beginning in the
correlation between self-reported feeling of 1980s, theorists began to search for
surprise and self-reported appraisal of the accounts of our facial expressions that
stimulus as unexpected. better suited the evidence.
The assumption that emotions cause
facial expressions has fared poorly as well.
Little evidence is available on production of ♦ The Behavioral Ecology View
faces, and what exists largely goes against
the hypothesis that emotions are the imme
diate cause. Fernandez-Dols and Ruiz- Modern evolutionary theory has inspired
Belda (1997) reviewed available evidence new accounts of facial and vocal expres
on the question, “What is the actual behav sion (e.g., Owren & Bachorowski, 2001;
ior of a happy person, an angry person, and Owren & Rendall, 2001). Here we focus
so on?” (p. 256); their review led them to on Fridlund’s (1991a, 1992b, 1994, 1997,
conclude that the “conventional answer, 2003) BEV. BEV is a pragmatic, function
known to artists, actors, and everyone else alist view of facial displays centered on
throughout the ages, is wrong” (p. 256). In how we use our faces in everyday life. BEV
support of this contention, Camras (1992) assumes facial displays are simply signals
examined one child for over a year and that influence others’ actions. No assump
failed to find the predicted expressions in tion is made that the recipient’s response
the predicted emotional circumstances. is mediated by a conscious reading of
More notably, Fernandez-Dols and the sender’s display or a classification
Ruiz-Belda (1995) studied people in a clearly of that display into categories like
ecstatic state (having just won a gold medal “emotion.”
at the Olympic games) and noted that they Evolutionary theory provides strong
failed to smile except in specific social cir basic premises for the BEV of facial displays
cumstances. Reisenzein (in press) extended (see Floyd, this volume). In the first place,
his research program on surprise and found the generation of facial displays co-evolves
that the predicted “facial expression of sur with others’ vigilance for them. The display
prise” occurred rarely in surprised persons, must benefit the displayer, but the recipient
although, tellingly, those same persons is also indispensable: Signals that do not
reported (mistakenly) that their faces had end up influencing the behavior of recipi
shown the predicted pattern. In Parkinson’s ents, either because they are not received
(2005) thoughtful and thorough review of or because they are moot, would not be
the literature most relevant to the relation of selected. Furthermore, recipients of displays
emotion to faces, he argued that overall, the should attend only to cues that provide
research shows only moderate correlations predictions about the future behavior of
between facial expressions and emotion. the displayer. Therefore, facial displays
16-Manusov.qxd 6/30/2006 5:09 PM Page 310
310–––◆–––Functions
and vigilance for them must co-evolve like the signals of other animals, should be
and could do so only if displays provided no less dependent on motive and context.
mutually beneficial signals of contingent Because displays exert their influence in the
future action. These signals would allow particular context of their issuance, they
interactants to proceed with reciprocation may be interpretable only within that
or counteraction that promoted mutual context. The evidence supports this: For
survival (Alcock, 1984; Hinde, 1985; example, a face interpreted as “afraid” in
Smith, 1977, 1986). This does not imply one context may be interpreted as “angry”
that recipients can use only formalized cues; in another (Carroll & Russell, 1996).
everyday behavior in a given context also
allows prediction of the others’ next moves
(Argyle, 1972; Kendon, 1981). EMOTIONS OR MOTIVES?
Second, for the human face to broadcast
its owner’s ongoing internal state, just From these premises, the BEV offers
as the face of a clock reads out the time, an alternative to the FEP. In Table 16.2,
would be foolhardy in its altruism, because we provide contrasting interpretations of
it ignores the potential conflict of interest commonly studied facial displays. Just as
between signaler and recipient. Signals do Yik and Russell (1999) showed that people
not evolve to provide information detri in three different societies interpret faces in
mental to the signaler. Signalers must not terms of social motive with about as much
be honest automatically, but only when it is agreement as they do in terms of emotion,
beneficial to do so. Automatic readouts of in each case in the table, an emotion is
emotion would thus be extinguished early contrasted with a social motive or inten
in phylogeny in the service of deception, tion. The table must be interpreted with
economy, and privacy. two strong caveats, however. First, the
Third, individuals who survived conflict prototypical facial displays used in FEP
by signaling their intentions would include research are highly selected, posed, melo
not only those who produced more schema dramatic faces of unknown ecological
tized facial behavior (technically, “ritualized” validity (indeed, their ecological validity is
if the evolution is genetic and “conven dubious; see Russell & Carroll, 1999). The
tionalized” if it is cultural) but also those more appropriate faces to examine are
with a heightened sensitivity to faces (see those that actually occur in the ecology.
Nelson & de Haan, 1997). This “ecology” Second, as we discussed above, for BEV—
of signaling and vigilance, countersignaling but not for FEP—the meaning of a facial
and countervigilance, is analogous to the display depends on its context. As such, the
balance of resources and consumers, and interpretations offered for the BEV in
predator and prey, which characterize all Table 16.2 are but samples, and they would
natural ecosystems. change along with the context in which the
Fourth, the costs and benefits of signal faces occurred.
ing, and of emitting a particular kind of sig A common criticism of the BEV is that it
nal, would vary with the momentary social supplants one ineffable, internal psychologi
context and the animal’s intentions within cal construct, emotion, with another, social
it. This sociality of animal signals is well motive, or that “emotion” can be defined
documented in a number of animals and easily so as to subsume social motive. Doing
a variety of social settings (Fridlund, 1994; so would make “emotion” even slipperier
Marler & Evans, 1997). Human signals, than it already is. A better approach is to
16-Manusov.qxd 6/30/2006 5:09 PM Page 311
Table 16.2 Emotions and Behavioral Ecology Interpretations of Common Human Facial
Displays
“Felt” (“Duchenne”) smile (readout Readiness to affiliate or play (“Good to see you,”
of happiness) “Let’s play [keep playing],” or “Let’s be friends”)
“Sad” face (readout of sadness) Recruitment of succor (“Take care of me,” “Hold
me” or “Look what you [he/she] did to me”)
“Anger” face (readout of anger) Readiness to attack (“Back off or I’ll attack”)
“Leaked” anger (inhibited anger) Conflict about attacking (“I want to attack and
I don’t want to attack” or “I’m so close to
attacking you”)
“Fear” face (readout of fear) Readiness to submit or escape (“Don’t hurt me!” or
“Take what you want!”)
“Contempt” face (readout of Declaration of superiority (“I can’t even bother with
contempt you” or “You’re not worth the trouble”)
sharpen and narrow concepts through empir interaction trajectories, just like our words,
ical confrontation of competing hypotheses. our tone of voice, and our gestures. All these
Motives as used in BEV are not about feelings are like switches on a railway, diverting the
or, indeed, any necessarily conscious state. interactional train this way and that as it bar
Rather, motive is the projected plan of action rels down the track. Emotion is the accom
and its goal. BEV therefore offers a different panying plume of steam. We offer two
program of research, with different ques additional lines of research that add fuel to
tions, premises, and hypotheses than those this perspective.
comprising the FEP. Although Fridlund
(1994) has been critical of the concept of
emotion, within the BEV “emotion” is, at SOCIALITY OF FACES
best, secondary in understanding of faces. We
argue that the concept of emotion is not If facial displays are intrinsically social,
needed to understand how our facial expres as we contend, then it would seem that
sions evolved or operate in modern life. Our when alone we wouldn’t produce them.
expressions participate in and guide our As Ekman, Davidson, and Friesen (1990)
16-Manusov.qxd 6/30/2006 5:09 PM Page 312
312–––◆–––Functions
argued, “Facial expressions do occur when solitary faces has been documented in
people are alone . . . and contradict the the several experiments of affective imagery,
oretical proposals of those who view expres which showed that, controlling for self-
sions solely as social signals” (p. 351). We reported emotion, subjects showed more
know of no evidence that the kind of proto facial behavior in high- than in low-sociality
typical facial expressions of emotion seen in imagery. In other words, people display to
Figure 16.2 occur when the displayer is the people “in their heads” (Fridlund,
alone, and Reisenzein’s (in press) data that Sabini, et al., 1990; Fridlund, Kenworthy,
people believe they show faces when in fact & Jaffey, 1992).
they do not suggest that evidence ought to Fridlund (1991b) attempted to circum
replace assumption here. But being social vent the imprecision of imagery manipula
versus being alone does not conform to the tions by manipulating implicit audiences
all-or-none law. Even when an interactant directly. Subjects watched an amusing
has been removed physically from the room, videotape in one of four viewing con
he or she may still be present psychologi ditions: (1) alone; (2) alone, but with the
cally. Actually, the physical presence of belief that a friend nearby was performing
others is one of the least important ways of an irrelevant task; (3) alone, but with
assessing the sociality of facial displays. the belief that a friend nearby was viewing
There are several ways in which people the same videotape in another room; and
can be structurally alone, with their facial (4) when a friend was physically present.
behavior implicitly social (see Fridlund, Viewers’ smiles were measured using facial
1994, for a complete exposition). First, EMG over the zygomatic major muscles
when we are alone we often treat ourselves responsible for smiling. Smiling among
as interactants. We reward or punish our solitary viewers who believed a friend was
selves, hit, touch, and stroke ourselves, and viewing nearby equaled that shown in the
likely move our faces in the course of these actual presence of the friend, but it was
acts. Most important, we speak to ourselves greater than that shown by subjects who
(without undermining the social and simply viewed alone. Reported happiness
communicative nature of language), and did not differ among the viewing condi
the facial expressions involved may be a tions, and within conditions it correlated
part (see paralanguage section below). negligibly with smiling.
Second, we often act as if others are present Similar findings emerged from another
when they are not. In their absence, we study involving gradations in sociality.
curse them and utter words of love to them. Chovil (1991b) visually coded the types of
In many of these acts we deploy facial dis gestures made in different social contexts.
plays. These faces, too, are social and com Her subjects (all females) heard stories about
municative, although they are emitted when close calls in one of four conditions: (1) alone,
we are alone. from an audiotape recording; (2) alone, over
Third, we often imagine that others are the telephone; (3) from another subject across
present when they are not. In our imagina a partition; and (d) talking to another subject
tion we engage in interactions with others face to face. When these conditions were
who are not there; that is, we “simulate” ordered according to their “psychological
interaction with them. We imagine talking presence,” as determined by separate raters,
to them, arguing with them, making love Chovil’s subject’s exhibited facial displays—
with them, and we often deploy facial dis largely wincing and grimacing—that
plays. That “implicit audiences” mediate increased monotonically with sociality, a
16-Manusov.qxd 6/30/2006 5:09 PM Page 313
finding that mapped nearly identically onto smile when they had just made a spare or
that provided by Fridlund (1991b). strike, but were likely to do so when they
These kinds of audience effects have been turned around to meet the gaze of those in
found in many studies, although not all their bowling party. In analogous studies
found facial behavior to be independent of involving infants and their play toys, smil
reported emotion; the relationship of the ing was almost entirely dependent on visual
two has become a focus of debate (Kappas, contact with the caregiver (Jones, Collins,
1997, and the special 1997 issue of Journal & Hong, 1991; Jones & Raag, 1989).
of Nonverbal Behavior). Some studies Fifth, we often treat nonhumans, and ani
have found independence (Devereux & mate and inanimate objects, as interactants.
Ginsburg, 2001; Jakobs, Manstead, & That we often treat nonhuman animals as
Fischer, 1999a), whereas others have not humans is self-evident to any pet owner. We
(Hess, Banse, & Kappas, 1995; Jakobs, also treat inanimate objects as though they
Manstead, & Fischer, 1999b). In these social were social interactants. Devotees of indoor
ity studies, face-to-face interaction resulted in gardening talk, gesture, and make faces to
maximal facial behavior. This should not their houseplants. Children do the same to
always be the case, however, because in their stuffed animals, dolls, or toy soldiers
many contexts we do not issue communica (see exposition by Fridlund & Duchaine,
tions to others. Friends sharing a humorous 1996). Thus, even those “solitary” facial dis
experience face to face should exhibit plays that are offered as decisive evidence
greater facial behavior than if they are sepa against BEV actually support it, and they
rated by a partition; friends asked to play reveal the extent to which we are social, even
poker should exhibit less. Likewise, facial when we are alone.
behavior that is socially censored (e.g., cry
ing in front of strangers or casual friends)
may produce less facial behavior with FACIAL PARALANGUAGE
increasing sociality (see Jakobs, Manstead, AND GESTURE
& Fischer, 2001). Indeed, the social role we
play toward others has proven critical in For FEP theorists, facial expressions of
understanding the displays we make toward emotion are a thing apart from language.
them, whether they are present or not (see Indeed, Ekman and Friesen’s (1978) FACS
Wagner & Smith, 1991). specifies that certain Action Units cannot
Fourth, we often rehearse, forecast, or be scored if they occur during speech. This
prepare for interaction and deploy dis omission is not trivial, because most facial
plays appropriately, consciously or not, displays occur amid speech. Indeed, we
even though no interactant is immediately have seen various indications that facial dis
present. Some displays function to deter plays are tied to speech more closely than
interaction, like scowls, or the tongue- is commonly understood. Fridlund and
showing display seen in both gorillas and Gilbert (1985) argued that the face’s chief
humans (Dolgin & Sabini, 1982; Smith, display role was to accompany and supple
Chase, & Lieblich, 1974). On the other ment speech. Such a close association
hand, “readiness to interact” displays (cf., between facial movements and speech
Smith, 1977) are omnipresent. We deploy a might be expected, given that the entire
smile seconds before greeting a neighbor at orofacial apparatus (lips, cheeks, tongue,
the front door. Indeed, Kraut and Johnston pharynx) changes in the course of making
(1979) found that bowlers were unlikely to speech sounds.
16-Manusov.qxd 6/30/2006 5:09 PM Page 314
314–––◆–––Functions
our comprehension of the substance of the look away, stare, listen, taste, sniff, attend
speaker’s utterance, as if to say, “I know or ignore, speak or shout, and so on. Little
what you mean.” Fifth (25%) nonlinguistic we do fails to change the face, and
adaptors, occurring with no clear relation to observers make use of this information
ongoing speech, that include biting the lips “given off.” During conversation, the face
or wiping them with the tongue. supplements and complements the verbal
Chovil’s (1991a) findings are reminiscent message. The research we have reviewed in
of findings on facial expressions, in that her this chapter suggests that the boundary
facial paralanguage categories do not specify between facial paralanguage and facial
precise facial muscular patterns (except for expressions of emotion may be illusory: The
the brow movements of syntactic displays), face and the spoken word each modifies the
because she found no facial movements that other, and together, they form a package
were peculiar to any specific paralinguistic that modifies the trajectory of ongoing social
category. Rather, specific movements were interaction. We can indicate disbelief, for
categorizable only in the context of the example, by rolling our eyes, “scrunching
ongoing conversation. In addition, most up” one side of our face, or saying “Yeah,
facial movements (with the exception of the right” as innuendo, and each form of com
“personal reaction displays”) had nothing munication has a similar effect. In everyday
obvious to do with emotion. social commerce, face, voice, and gesture are
Chovil’s (1991a) data demonstrate all co-conspirators. They are all paralan
the overwhelming predominance of the guage, all signals that serve the same social
paralinguistic role of the face. Given this end, and none of them has a privileged rela
predominance, we argued that facial tionship to emotion (see Bavelas & Chovil,
expressions of emotion have received too 2000; this volume, for their Integrated
much attention relative to paralinguistic Message Model from this understanding).
functions of faces. Furthermore, the distinc New neurological findings support this
tion between faces that accompany speech suggestion by pointing to common neuro
and those that do not is superficial, and sev logical underpinnings and, possibly, evolu
eral explanations can be offered for it. One tionary origins of vocal, facial, and gestural
possibility is that “facial expressions” are a expressions (see Rizzolatti & Arbib, 1998;
type of interjection into speech. Like the Rizzolatti & Craighero, 2004). These find
verbal interjections we insert in our ongo ings begin with the discovery in monkeys
ing speech, the facial expressions of emo of the so-called mirror neurons in the
tion that are bracketed off by FEP theorists frontal lobe. These mirror neurons do
may actually be nonverbal interjections, not fire when the animal possessing them
once necessarily linked to vocalization but sees an object but rather when he sees a
now separable from it, making them appear certain action performed toward the
autonomous and wholly apart from the object—whether he is executing the action
stream of our facial paralanguage. or watching someone else do it. Mirror
neurons were found for actions that
included grasping between index finger
♦ Concluding Comment and thumb, placement on a plate, or
tearing.
Of course, the discovery of mirror
Faces differ in their static configura neurons is in one sense trivial. Because we
tions, and the face changes dynamically in imitate, we must have neurons for imita
coloration and movement as we look, tion. But the location of these neurons is
16-Manusov.qxd 6/30/2006 5:09 PM Page 316
316–––◆–––Functions
Dolgin, K. M., & Sabini, J. (1982). Experimental Fridlund, A. J. (1992b). The behavioral ecology
manipulation of a human non-verbal dis and sociality of human faces. In M. S. Clark
play: The tongue show affects an observer’s (Ed.), Review of personality and social psy
willingness to interact. Animal Behaviour, chology (Vol. 13, pp. 90–121). Newbury
30, 935–936. Park, CA: Sage.
Drummond, P. D., & Mirco, N. (2004). Staring Fridlund, A. J. (1994). Human facial expression:
at one side of the face increases blood flow An evolutionary view. San Diego, CA:
on that side of the face. Psychophysiology, Academic Press.
41, 281–287. Fridlund, A. J. (1997). The new ethology
Ekman, P., Davidson, R. J., & Friesen, W. V. of human facial expressions. In J. A. Russell
(1990). The Duchenne smile: Emotional & J.-M. Fernandez-Dols (Eds.), The
expression and brain physiology II. Journal psychology of facial expression (pp. 103–
of Personality and Social Psychology, 58, 129). New York: Cambridge University
342–353. Press.
Ekman, P., & Friesen, W. V. (1969). The reper Fridlund, A. J. (2003). The behavioral ecology
toire of nonverbal behavior: Categories, ori view of smiling and other facial expres
gins, usage, and coding. Semiotica, 1, 49–98. sions. In M. Abel (Ed.), An empirical reflec
Ekman, P., & Friesen, W. V. (1978). Facial tion on the smile (pp. 45–82). New York:
action coding system: A technique for the Mellen Press.
measurement of facial movement. Palo Fridlund, A. J., & Cacioppo, J. T. (1986).
Alto, CA: Consulting Psychologists Press. Publication guidelines for human elec
Elfenbein, H. A., & Ambady, N. (2003). When tromyographic research. Psychophysiology,
familiarity breeds accuracy: Cultural 23, 567–589.
exposure and facial emotion recognition. Fridlund, A. J., & Duchaine, B. (1996). “Facial
Journal of Personality and Social Psycho expressions of emotion” and the delusion
logy, 85, 276–290. of the hermetic self. In R. Harré & W. G.
Fernandez-Dols, J.-M., & Ruiz-Belda, M.-A. Parrott, The emotions (pp. 259–284).
(1995). Are smiles a sign of happiness? Cambridge, UK: Cambridge University
Gold medal winners at the Olympic Games. Press.
Journal of Personality and Social Psycho Fridlund, A. J., & Gilbert, A. N. (1985).
logy, 69, 1113–1119. Emotions and facial expressions (Letter to
Fernandez-Dols, J.-M., & Ruiz-Belda, M.-A. the Editor). Science, 230, 607–608.
(1997). Spontaneous facial behavior during Fridlund, A. J., Kenworthy, K. G., & Jaffey, A.
intense emotional episodes: Artistic truth K. (1992). Audience effects in affective
and optical truth. In J. A. Russell & J.-M. imagery: Replication and extension to
Fernandez-Dols (Eds.), The psychology of affective imagery. Journal of Nonverbal
facial expression (pp. 255–274). New Behavior, 16, 191–212.
York: Cambridge University Press. Fridlund, A. J., Sabini, J. P., Hedlund, L. E.,
Fridlund, A. J. (1991a). Evolution and facial Schaut, J. A., Shenker, J. I., & Knauer,
action in reflex, social motive, and paralan M. J. (1990). Social determinants of
guage. Biological Psychology, 32, 3–100. facial expressions during affective imagery:
Fridlund, A. J. (1991b). Sociality of solitary Displaying to the people in your head.
smiling: Potentiation by an implicit audi Journal of Nonverbal Behavior, 14,
ence. Journal of Personality and Social Psy 113–137.
chology, 60, 229–240. Frijda, N. H., & Tcherkassof, A. (1997). Facial
Fridlund, A. J. (1992a). Darwin’s anti- expressions as modes of action readiness.
Darwinism and the Expression of the emo In J. A. Russell & J.-M. Fernández-Dols
tions in man and animals. In K. T. Strongman (Eds.), The psychology of facial expres
(Ed.), International review of emotion (Vol. sion (pp. 78–102). New York: Cambridge
2, pp. 117–137). New York: Wiley. University Press.
16-Manusov.qxd 6/30/2006 5:09 PM Page 318
318–––◆–––Functions
Gerlach, A. L., Wilhelm, F. H., Gruber, K., & 10-month-old infants. Psychological Science,
Roth, W. T. (2001). Blushing and physio 2, 45–49.
logical arousability in social phobia. Journal Jones, S. S., & Raag, T. (1989). Smile produc
of Abnormal Psychology, 110, 247–258. tion in older infants: The importance of a
Gilman, S. L. (1985). Difference and pathology: social recipient for the facial signal. Child
Stereotypes of sexuality, race and madness. Development, 60, 811–818.
Ithaca, NY: Cornell University Press. Kappas, A. (1997). The fascination with faces:
Gilman, S. L. (1996). Seeing the insane. Lincoln: Are they windows to our soul? Journal of
University of Nebraska Press. Nonverbal Behavior, 21, 157–161.
Goffman, E. (1959). The presentation of self in Kendon, A. (1981). Introduction: Current issues
everyday life. Garden City, NY: Doubleday. in the study of “nonverbal communi
Gould, S. J., & Vrba, E. S. (1982). Exaptation: cation.” In A. Kendon (Ed.), Nonverbal
A missing term in the science of form. communication, interaction, and gesture
Paleobiology, 8, 14–15. (pp. 1–53). Paris: Mouton.
Heatherton, T. F., Kleck, R. E., Hebl, M. R., & Kraut, R. E., & Johnston, R. E. (1979). Social and
Hull, J. G. (Eds.). (2000). The social psy emotional messages of smiling: An ethologi
chology of stigma. New York: Guilford. cal approach. Journal of Personality and
Hess, U., Banse, R., & Kappas, A. (1995). The Social Psychology, 37, 1539–1553.
intensity of facial expression is determined Lang, P. J. (1968). Fear reduction and fear
by underlying affective state and social behavior: Problems in treating a construct. In
situation. Journal of Personality and Social J. M. Shlien (Ed.), Research in Psycho
Psychology, 69, 280–288. therapy, 1, 90–102.
Hinde, R. A. (1985). Expression and negotia Lazarus, R. S., Speisman, J. C., & Mordkoff, A.
tion. In G. Zivin (Ed.), The development of M. (1963). The relationship between
expressive behavior (pp. 103–116). Orlando, autonomic indicators of psychological
FL: Academic Press. stress: Heart rate and skin conductance.
Hjortsjö, C. B. (1970). Man’s face and mimic Psychosomatic Medicine, 25, 19–30.
language. Lund: Student-litteratur. Leary, M. R., Britt, T. W., Cutlip, W. D., &
Izard, C. E. (1979). The Maximally Discriminative Templeton, J. L. (1992). Social blushing.
Facial Movement Coding System (MAX). Psychological Bulletin, 112, 446–460.
Unpublished manuscript. (Available from Mandler, G., Mandler, J. M., Kremen, I.,
Instructional Resource Center, University of & Sholiton, R. D. (1961). The response
Delaware, Newark, DE) to threat: Relations among verbal and phys
Jakobs, E., Manstead, A. S. R., & Fischer, A. H. iological indices. Psychological Mono
(1999a). Social motives and emotional graphs, 75(9, Whole No. 513), 22.
feelings as determinants of facial displays: Marler, P., & Evans, C. (1997). Animal sounds
The case of smiling. Personality and Social and human faces: Do they have anything in
Psychology Bulletin, 25, 424–435. common? In J. A. Russell & J.-M. Fernandez-
Jakobs, E., Manstead, A. S. R., & Fischer, A. H. Dols (Eds.), The psychology of facial expres
(1999b). Social motives, emotional feel sion (pp. 133–157). New York: Cambridge
ings, and smiling. Cognition & Emotion, University Press.
13, 321–345. Nelson, C. A., & de Haan, M. (1997). A neu
Jakobs, E., Manstead, A. S. R., & Fischer, A. H. robehavioral approach to the recognition
(2001). Social context effects on facial activ of facial expressions in infancy. In J. A.
ity in a negative emotional setting. Emotion, Russell & J.-M. Fernandez-Dols (Eds.),
1, 51–69. The psychology of facial expression
Johnson, V. S., & Franklin, M. (1993). Is beauty (pp. 176–204). New York: Cambridge
in the eye of the beholder? Ethology and University Press.
Sociobiology, 14, 183–199. Ortony, A., & Turner, T. J. (1990). What’s so
Jones, S. S., Collins, K., & Hong, H.-W. (1991). basic about basic emotions? Psychological
An audience effect on smile production in Review, 97, 315–331.
16-Manusov.qxd 6/30/2006 5:09 PM Page 319
Owren, M. J., & Bacharowski, J.-A. (2001). The Russell, J. A, & Carroll, J. M. (1999). The
evolution of emotional experience: A “self phoenix of bipolarity: Reply to Watson and
ish-gene” account of smiling and laughter Tellegen (1999). Psychological Bulletin,
in early hominids and humans. In T. J. 125, 611–617.
Mayne & G. A. Bonanno (Eds.), Emotions: Russell, J. A., & Fehr, B. (1987). Relativity in
Current issues and future directions. the perception of emotion in facial expres
Emotions and social behavior (pp. 152– sions. Journal of Experimental Psychology:
191). New York: Guilford Press. General, 116, 223–237.
Owren, M. J., & Rendall, D. (2001). Sound on Russell, J. A., & Fernandez-Dols (Eds.). (1997).
the rebound: Bringing form and function The psychology of facial expression
back to the forefront in understanding (pp. 295–320). New York: Cambridge
nonhuman primate vocal signaling. Evolu University Press.
tionary Anthropology, 10, 58–71. Schlenker, B. R. (1980). Impression manage
Parkinson, B. (2005). Do facial movements ment: The self-concept, social identity, and
express emotions or communicate motives? interpersonal relations. Monterey, CA:
Personality and Social Psychology Review, Brooks/Cole.
9, 278–311. Shields, S. A., Mallory, M. E., & Simon, A.
Penton-Voak, I. S., Perrett, D. I., Castles, D. L., (1990). The experience and symptoms of
Kobayashi, T., Burt, D. M., Murray, L. K., blushing as a function of age and reported
et al. (1999). Menstrual cycle alters face frequency of blushing. Journal of Nonver
preferences. Nature, 399, 741–742. bal Behavior. 14, 171–187.
Pernkopf, E. (1938). Atlas of topographical and Smith, W. J. (1977). The behavior of com
applied human anatomy. Philadelphia: municating. Cambridge, MA: Harvard
Saunders. University Press.
Perrett, D. I., May, K. A., & Yoshikawa, S. Smith, W. J. (1986). An “informational” per
(1994). Facial shape and judgments of spective on manipulation. In R. W. Mitchell
female attractiveness. Nature, 368, 239–242. & N. S. Thompson (Eds.), Perspectives on
Porter, A. M. W. (1993). Sweat and thermo human and nonhuman deceit (pp. 71–87).
regulation in hominids. Journal of Human Albany: State University of New York Press.
Evolution, 25, 417–423. Smith, W. J., Chase, J., & Lieblich, A. K. (1974).
Reisenzein, R. (2000). Exploring the strength Tongue showing: A facial display of
of association between the components of humans and other primate species.
emotion syndromes: The case of surprise. Semiotica, 11, 201–246.
Cognition & Emotion. 14, 1–38. Sorenson, E. R. (1975). Culture and the expres
Reisenzein, R. (in press). Evidence for strong sion of emotion. In. T. R. Williams (Ed.),
dissociation between emotion and facial Psychological anthropology (pp. 364–372).
displays: The case of surprise. Journal of The Hague: Mouton.
Personality and Social Psychology. Wagner, H. L., & Smith, J. (1991). Facial
Rizzolatti, G., & Arbib, M. A. (1998). Language expression in the presence of friends and
within our grasp. Trends in Neurosciences, strangers. Journal of Nonverbal Behavior,
21, 188–194. 15, 201–214.
Rizzolatti, G., & Craighero, L. (2004). The Yik, M. S. M., & Russell, J. A. (1999). Inter
mirror-neuron system. Annual Review of pretation of faces: A cross-cultural study of
Neuroscience, 27, 169–192. a prediction from Fridlund’s theory.
Russell, J. A. (1994). Is there universal recogni Cognition & Emotion, 13, 93–104.
tion of emotion from facial expression?
Psychological Bulletin, 115, 102–141.
Russell, J. A. (1995). Facial expressions of
emotion: What lies beyond minimal univer
sality? Psychological Bulletin, 118, 379–391.
16-Manusov.qxd 6/30/2006 5:09 PM Page 320
17-Manusov.qxd 6/30/2006 5:09 PM Page 321
17
WHY AND HOW THE
SILENT SELF SPEAKS VOLUMES
Functional Approaches to Nonverbal
Impression Management
� Caroline F. Keating
Colgate University
◆ 321
17-Manusov.qxd 6/30/2006 5:09 PM Page 322
322–––◆–––Functions
instance, whereas blatant, verbal self- behavior has with impression formation.
promotion can create a backlash against She noted that the impressions nonverbal
women who use it (Rudman, 1998), non cues generate are typically “off the record”
verbal tactics such as firm handshakes can in that they are resistant to precise identifi
be an effective way for women to self- cation and assessment (p. 206). Nonverbal
promote in a sexist environment (Chaplin, cues are like journalists’ anonymous
Phillips, Brown, Clanton, & Stein, 2000). sources: Their messages are crucial but hard
Functional explanations for impression to name. A literary example comes from
management combine how individuals columnist Maureen Dowd (2004), who
control the presentation of self-relevant attributed the Bush Administration’s deci
information with how others respond to sion to invade Iraq to “body language” that
it (Schlenker, 2003; Schlenker & Pontari, amplified verbal exchanges between the
2000). As DePaulo (1992) pointed out, President and his advisors. In this case,
nonverbal cues are more accessible to audi the President’s nonverbal actions made his
ences than to communicators, because words “speak louder” by rendering images
those signaling cannot see the messages of a man whose mind was already made up.
they send. Signalers are stuck relying largely Nonverbal impression management can
on audience reactions to gauge the impres also distract audiences from verbal mes
sions they leave. In the end, impression sages. Adaval and Wyers (2004) found that
management may depend as much on the when memory for impressions and nonver
audience as on the actor because it requires bal actions was sharp, recall for verbal mes
the dynamic interplay between the two sages was dull. In this instance, actions
(Goffman, 1959; Schlenker, 2003). Thus, appeared to speak louder than words,
the way to successful self-presentation is perhaps even drowning them out.
both to “Know Thy Self” and “Know Thy Muscling nonverbal cues into sculpted
Audience.” impressions is not always easy. It is particu
The importance of actor and audience larly difficult to manage impressions that
dynamics has encouraged some research require suppressing or neutralizing sponta
ers to analyze nonverbal interaction using neous nonverbal responses (Ekman, 1992).
dyads or groups as units of analysis (e.g., Individuals pressured to appear invulnerable
Bernieri, Gillis, & Davis, 1996). Approaches can often control verbal self-reports better
like these fuse the moment-by-moment than they can monitor nonverbal behavior.
interdependence of one person’s action with Gay child care workers, for example, may
another’s reaction. However, most analyses express little anxiety verbally when faced
of impression management processes have with stereotype threats, whereas kinesic cues
been extracted from the vantage point of the belie their verbal expression of confidence
individual actor or encoder (Burgoon et al., by conveying tension (Bosson, Haymovitz,
1996). To fortify these types of approaches, & Pinel, 2004). Moreover, the successful
Patterson (2001; see also Patterson, this neutralization of nonverbal expression is no
volume) hinged actor-audience dynamics guarantee that a desirable image will be pro
together by analyzing the actors’ dual tasks jected. For instance, overcontrolling nonver
as encoders of messages and decoders of bal output can make others suspicious of
audience reactions. highly motivated liars (DePaulo, Lindsay,
In her overview of nonverbal behavior Malone, & Muhlenbruck, 2003; see also
and self-presentation, DePaulo (1992) artic Vrij, this volume). Lack of expressivity
ulated the special relationship nonverbal tends to convey disinterest, aloofness, and
17-Manusov.qxd 6/30/2006 5:09 PM Page 324
324–––◆–––Functions
coldness (Burgoon et al., 1996; Mehrabian, assigned to them (Albright, Forest, &
1972). When it comes to the nonverbal Reiseter, 2001). How can perceivers be
engine that powers self-presentation, there so readily beguiled by others’ nonverbal
seems to be no “neutral” gear, only “for performances? At times, perceivers may
ward” and “reverse.” Cues such as mutual “want” to be fooled and accept impressions
gaze while smiling, nodding, and forward at face value. For example, socially anxious
lean generally draw approach, whereas people are especially poor lie detectors
mouth and brow frowns, gaze avoidance (DePaulo & Tang, 1994) and seem to avoid
or aggressive stares, tense body postures, and gazing at emotion-laden faces presumably
interpersonal space violations usually propel because they fear negative social appraisals
avoidance (Burgoon et al., 1996; Mehrabian, (Mansell, Clark, Ehlers, & Chen, 1999).
1972). Women often resist probing the veracity of
Managing images to regulate approach ingratiating comments (DePaulo, Stone, &
and avoidance could be considered a cross- Lassiter, 1985). Thus, accepting as well as
species phenomenon. As any horse rancher projecting contrived images may have a lot
knows, stallions never limp in the presence to do with the fear of looking too closely
of their mares: Revealing signs of lameness into the proverbial social mirror (Tice &
telegraphs vulnerability to mates and com Baumeister, 2001).
petitors and would be a stallion’s reproduc There are, nevertheless, nonverbal checks
tive undoing. Primate-care workers in and balances on human gullibility in
laboratories and zoos are often amazed to response to others’ managed impressions.
discover animals that, overnight, seem to One unlikely defense against skewed non
fall fatally ill and die. Sick or injured indi verbal presentations is the snap judgment.
viduals apparently protect themselves from Nonverbal behavior is often decoded early
becoming the target of conspecifics’ rejec and quickly during interactions (DePaulo,
tion and aggression by cloaking signs of 1992), and researchers have found, for
weakness up to the bitter end. The best- example, that deception is best detected
known human examples of this kind of rapidly (Vrij, Evans, Akehurst, & Mann,
impression management may come from 2004) before a person’s “acting” takes
the White House. The ability of American effect. Audiences also have defenses against
presidents to disguise physical and psycho bad acting: Self-presentations that do not
logical illness and project false images of seem genuine are not effective. So when
health and fortitude has kept many in office women read a script meant to project pow
despite the odds (Gilbert, 1998). From a erful leadership, they were not nearly as
functional perspective, honesty about one’s effective as when they performed the identi
physical condition may not always be the cal script in a mindful way, not straying from
best policy when the social goal is to main the content but adding personal, feminine
tain power. Indeed, the difficulty in detect touches to the tone (Kawakami, White, &
ing pain from gestures has stymied Langer, 2000). Overlearned, scripted behav
veterinarians and physicians alike for years ior can lead to stilted performances that are
(e.g., Hyden & Peolsson, 2002; Leary, not compelling, especially if the image one
Tchividjian, & Kraxberger, 1994). attempts to project does not come naturally
Individuals are proficient at “spinning” (Schlenker & Pontari, 2000).
images, even in experimental settings where Nevertheless, some individuals have
the image “spun” has been arbitrarily better self-presentational skills than others.
17-Manusov.qxd 6/30/2006 5:09 PM Page 325
Researchers have attributed these differ to do so (Hassin & Trope, 2000). At the
ences to variations in temperament, same time, perceivers are unaware that mor
appearance, or environment or to combi phological cues often guide their assessment
nations of traits, skills, and experiences and treatment of others (Keating, Randall,
(e.g., Anderson, John, Keltner & Kring, & Kendrick, 1999; Keating, Randall,
2001; DePaulo et al., 2003; Gangestad & Kendrick, & Gutshall, 2003; Todorov,
Snyder, 2000). Examples come from stud Mandisodza, Goren, & Hall, 2005).
ies of leadership. Undergraduates identi People direct remarkable amounts of
fied by their same-sex peers as socially time, effort, and resources toward modify
dominant seem to be especially good at ing outward appearances. Across cultures
disguising the truth (Keating & Heltman, and millennia, face and body parts have
1994). Presenting an image of intellectual been dressed, painted, pierced, shaved,
competence is just as good as the real thing plucked, injected, molded, stretched, cut,
(i.e., possessing intelligence) when leaders and sewn to manage images of self and iden
are judged for intelligence (Rubin, Bartels, tity (Guerrero & DeVito, 1999; Zebrowitz,
& Bommer, 2002). Furthermore, charis 1997). These (pre)occupations often reflect
matic nonverbal performances are conta cultural values. In parts of the West, fasci
gious and enhance leadership effectiveness nation with individual physical appearance
and liking (Cherulnick, Donely, Wiewel, has led to the popularity of television series
& Miller, 2001). Yet deciphering exactly such as ABC’s Extreme Makeover, which
what these powerfully appealing nonver on a typical night musters an audience of
bal impression management skills are and over 8 million U.S. viewers, who watch as
teaching them has proven to be difficult plastic surgeons, cosmetic artists, and
(Riggio, Riggio, Salinas, & Cole, 2003). physical trainers transform appearances
(C. Whipple, personal communication,
July 26, 2005; see Manusov & Jaworski,
NONVERBAL IMPRESSION this volume).
MANAGEMENT: Clothing is part of this nonverbal arsenal
TECHNIQUES OF THE TRADE of impression management techniques.
Women and men select clothing styles
Static Physical Appearance Cues. The sizes, strategically to make their bodies appear
shapes, qualities, and spatial relationships to fit cultural ideals (e.g., Frith & Gleeson,
of static morphological cues influence how 2004; see also Guerrero & DeVito, 1999).
individuals are perceived (see Rhodes & These physical ideals, and their accompany
Zebrowitz, 2002). Moreover, static signals ing fashions, change with the times in what
can alter the interpretation of dynamic non could arguably be an adaptive pattern
verbal cues: Thus, the same behavior may (Pettijohn & Jungeberg, 2004). Even the
get a different “read” when displayed by color of clothing can spark impressions in a
individuals with different facial structures, big way. For example, the aggressiveness of
body types, or genders (Keating, 2002). national athletic teams has been linked to
After all, morphological cues “arrive” first the color of their uniforms. Football and
and set expectations about traits and abili hockey teams wearing black uniforms
ties (Zebrowitz, 1997). So potent are these receive disproportionately high numbers of
cues that they are difficult for perceivers to penalties, in part because they are perceived
ignore even when given explicit instructions as more aggressive; cued by their own
17-Manusov.qxd 6/30/2006 5:09 PM Page 326
326–––◆–––Functions
328–––◆–––Functions
criteria were distinguishable by one of four and what messages are likely to be advan
overlapping, theoretical emphases: evolu tageous (or disadvantageous) to impression
tionary, ecological, emotional, or social- formation (for a more general discussion of
cognitive. Table 17.1 outlines distinguishing evolutionary perspectives on nonverbal
features and cites research related to each. communication, see Floyd, this volume).
The intellectual inspiration for evolutionary
approaches can be traced to Darwin
EVOLUTIONARY PERSPECTIVES (1872/1991), who applied notions of ran
ON IMPRESSION MANAGEMENT dom variation and selective retention of
genetically based traits to communication
Evolutionary perspectives are perhaps in animals and humans. Darwinian logic
best at projecting why individuals manage requires that genetic substrates (however
nonverbal cues to produce particular images direct or indirect) underlie appearances and
17-Manusov.qxd 6/30/2006 5:09 PM Page 329
330–––◆–––Functions
uncertainty, because deriving testable validity and cue utilization. It could be said
predictions and reasonable measures of fit that Brunswikian approaches are more gen
ness is difficult. In other words, the ques erous than evolutionary ones in the degree to
tion to be posed in this instance is not which they anticipate and tolerate error in
whether blonds have more fun, but whether the (social) perceptual system.
they have more offspring and kin who The application of Brunswik’s paradigm
themselves are reproductively successful. to nonverbal self-presentation reveals that
the cues encoded by a communicator may
or may not match those used (decoded) by
ECOLOGICAL PERSPECTIVES ON observers to derive impressions of the com
IMPRESSION MANAGEMENT municator. Gifford (1994) noted this slip
page when he adapted a Brunswikian lens
Although displays ultimately enhance model to nonverbal impression formation.
reproductive potential far downstream, He found that communicators who scored
proximate goals, say for social approval or high on the measurements of warmth and
financial reward, may be served in their agreeableness nodded their heads often dur
more immediate wake. From an ecological ing interactions. When observers judged the
perspective, the benefits of impression man communicators’ traits, however, they relied
agement involve adaptive social functioning on more than just the valid cue of nods;
in which the fit between signal and social observers used a host of postural cues
context or “ecology” is imperative. The the unassociated with the dispositions of com
oretical perspectives categorized as following municators. Similarly, object manipulation
an ecological tradition reflect the thinking predicted communicators’ scores on mea
of early functionalists like Brunswik (1955), surements of dominance and submissive
who applied his ideas to perception and ness, yet observers neglected to use this cue
social perception (e.g., to face cues, nations). in their assessments of them. One explana
Brunswik argued that perceptual systems tion for the discrepancy may relate to per
were adapted to their environments by ceiver’s motivation. When the dispositions
expectancies developed through experience judged are highly relevant to perceivers,
in a particular environment or ecology. they tend to increase their use of valid cues
Perception operated probabilistically, mean (Gangestad, Simpson, DiGeronimo, & Biek,
ing that cue perception was biased toward 1992).
interpretations that had worked in the past Ecological approaches relevant to
(e.g., Segall, Campbell, & Herskovitz, impression management include research
1966). Because probabilistic judgments were on social perception derived from static
mostly correct in specific ecologies, these physical appearances as well as behavior.
perceptual habits or attunements were The Brunswikian idea of affordances has
thought to be adaptive (McArthur & Baron, been championed by Zebrowitz and her
1983). Like a Vegas gambler at the blackjack colleagues (e.g., Montepare & Zebrowitz,
table, Brunswik (1955) reckoned that per 1998; Zebrowitz, 1997). For example,
ceivers need only beat the odds some per facial structures may convey affordances
centage of the time to come away with a defined as opportunities for certain types
winning perceptual formula. This implies of interactions. Sensitivity toward these
that a certain degree of error is acceptable signals is adaptive but can result in over-
in social perception. That is, there may be interpretation, a kind of perceptual error
some “slippage” in the matchup between cue we may be biologically prepared to
17-Manusov.qxd 6/30/2006 5:09 PM Page 331
332–––◆–––Functions
334–––◆–––Functions
graduate seminar classroom. Connie was within us sought to control our social world
stunned by this accusation; she did not feel (see Leary, 1995; Schlenker & Pontari,
sexually attracted to her professor. What 2003, for discussions). Contemporary evolu
kind of impression management was this? tionary and ecological perspectives give
Or was it? And whose was it? Maybe impression management, especially its non
Connie’s nonverbal behavior expressed an verbal forms, a place in the larger scheme
attraction toward her professor that she of adaptive social behavior. New knowledge
was unaware of and Connie was simply about connections between emotion and
self-deceptive. Alternatively, the professor cognition offers fresh insights into nonverbal
was perhaps motivated to misread Connie’s emotional self-presentation, self-regulation,
body cues and perceive messages that were and social context. Contemporary social-
not really there. But either way, could the cognitive approaches put conscious and
professor defend himself by claiming that nonconscious processes within the bound
he was victimized by Connie’s noncon aries of nonverbal impression management
scious goal of seduction played out on a frameworks and highlight the importance of
nonverbal stage? Is Connie responsible? Is nonverbal measurement techniques.
her professor responsible? At the same time, these new insights
This example highlights the fact that have a troubling side. Freed from the impo
nonconscious goal activation and its inter sition of conscious control, where is the
section with deceptive and self-deceptive “management” in nonverbal impression
cognitive processes have potentially impor management? How is successful nonverbal
tant practical consequences for the per impression management different from just
formers and audiences of nonverbal getting lucky in the context of impression
impression management. At the same time, formation? Is impression management
they present conceptual and measurement simply the mirror image of person percep
challenges for researchers studying impres tion, a contest for best performance of
sion management. These conceptual and expressive scripts, or a relatively honest
measurement issues are intertwined. Given form of deception? The usefulness of the
that nonverbal impression management can impression management concept may lie
be driven by nonconscious goals and stimu in a renewed emphasis on its most unique
lated by nonconscious processes (e.g., aspect: the interdependence of actor-
Cheng & Chartrand, 2003), nonverbal audience psychologies (Goffman, 1959;
behavior may be the best—or only—way to Patterson, 2001). Those who invest in
track their operation. These measurements future research on nonverbal impression
will have especially complex iterations in management must give due diligence by dis
“live” impression management situations tinguishing its functions and effects from
where nonconscious processes simultane those of other social influence processes,
ously energize actors, audiences, and their testing them in an orderly way, and reveal
relationships. Teasing nonverbal commu ing their superior predictive validity.
nicative processes apart as well as pasting
them together will be necessary to elucidate
how nonverbal impression management is ♦ References
orchestrated between actors and audiences.
Previous conceptualizations of impres Adaval, R., & Wyers, R. S. (2004).
sion management characterized it as a Communicating about a social interaction:
Machiavellian enterprise of strategy and Effects on member for protagonists’ state
manipulation in which the “mini-Mach” ment and nonverbal behaviors. Journal
17-Manusov.qxd 6/30/2006 5:09 PM Page 335
336–––◆–––Functions
338–––◆–––Functions
A perceiver’s perspective. Journal of Riggio, R., Riggio, H. R., Salinas, C., & Cole,
Nonverbal Behavior, 13, 15–24. E. J. (2003). The role of social and emotional
McArthur, L. Z., & Baron, R. M. (1983). communication skills in leaders emergence
Toward an ecological theory of social and effectiveness. Group Dynamics: Theory,
perception. Psychological Review, 90, Research, and Practice, 7, 83–103.
215–238. Rubin, R. S., Bartels, L. K., & Bommer. W. H.
Mehrabian, A. (1972). Nonverbal communica (2002). Are leaders smarter or do they just
tion. Chicago: Aldine-Atherton. seem that way? Exploring perceived intel
Montepare, J. M. (1995). The impact of varia lectual competence and leadership emer
tions in height on young children’s impres gence. Social Behavior and Personality,
sions of men and women. Journal of 30, 105–118.
Nonverbal Behavior 19, 31–47. Rudman, L. A. (1998). Self-promotion as a risk
Montepare, J. M., & Zebrowitz, L. A. (1998). factor for women: The costs and benefits
Person perception comes of age: The of counterstereotypical impression manage
salience and significance of age in social ment. Journal of Personality and Social
judgments. Advances in Experimental Psychology, 74, 629–645.
Social Psychology, 30, 93–161. Saarni, C. (1989). Children’s understanding of
Mueller, U., & Mazur, A. C. (1996). Facial strategic control of emotional expression in
dominance in Homo sapiens as honest sig social transactions. In C. Saarni & P. Harris
naling of male quality. Behavioral Ecology, (Eds.), Children’s understanding of emotion
8, 569–579. (pp. 181–208). New York: Cambridge
Pataki, S. P., & Clark, M. S. (2004). Self- University Press.
presentations of happiness: Sincere, polite, Saarni, C., & Weber, H. (1999). Emotional
or cautious? Personality and Social Psycho displays and dissemblance in childhood:
logy Bulletin, 30, 905–914. implications for self-presentation. In
Patterson, M. L. (1999). The evolution of a par P. Philippot, R. S. Feldman, & E. J. Coats
allel process model of nonverbal communi (Eds.), The social context of nonverbal
cation. In P. Philippot, R. S. Feldman, & behavior (pp. 317–347). Cambridge, UK:
E. J. Coats (Eds.), The social context of non Cambridge University Press.
verbal behavior (pp. 317–347). Cambridge, Schlenker, B. R. (2003). Self-presentation. In
UK: Cambridge University Press. M. R. Leary & J. P. Tangney (Eds.),
Patterson, M. L. (2001). Toward a comprehen Handbook of self and identity (pp. 462–
sive model of non-verbal communication. 518). New York: Guilford Press.
In W. P. Robinson & H. Giles (Eds.), The Schlenker, B. R., & Pontari, B. A. (2000). The
new handbook of language and social psy strategic control of information: Impression
chology (pp. 159–176). New York: Wiley. management and self-presentational in
Perrett, D. I., Lee, K. J., Penton-Voak, daily life. In A. Tesser, R. Felson, & J. Suils
I., Rowland, D., Yoshikawa, S., Burt, D. M., (Eds.), Perspective on self and identity
et al. (1998). Effects of sexual dimorphism (pp. 199–232). Washington, DC: American
on facial attractiveness. Nature, 394, Psychological Association.
884–887. Segall, M. H., Campbell, D. T., & Herskovitz,
Pettijohn, T. E., III, & Jungeberg, B. J. (2004). M. J. (1966). The influence of culture on
Playmate curves: Changes in facial and visual perception. Indianapolis, IN: Bobbs
body feature preferences across social Merrill.
and economic conditions. Personality and Singer, M. A., & Goldin-Meadows, S. (2005).
Social Psychology Bulletin, 30, 1186–1197. Children learn when their teacher’s gestures
Rhodes, G., & Zebrowitz, L. A. (2002). (Eds.). and speech differ. Psychological Science,
Facial attractiveness: Evolutionary, cogni 16, 85–89.
tive, and social perspectives. Westport, CT: Singh, D., & Bronstad, P. M. (1997). Sex
Ablex. differences in the anatomical locations of
17-Manusov.qxd 6/30/2006 5:09 PM Page 339
human body scarification and tattooing Personality and Social Psychology, 84,
as a function of pathogen prevalence. 558–568.
Evolution and Human Behavior, 18, Todorov, A., Mandisodza, A. N., Goren, A., &
403–416. Hall, C. C. (2005). Inferences of compe
Snodgrass, S. E. (1992). Further effects of role tence from faces predict election outcomes.
versus gender on interpersonal sensitivity. Science, 308, 1623–1626.
Journal of Personality and Social Psycho Vohs, K. D., Baumeister, R. F., & Ciarocco,
logy, 62, 154–158. N. J. (2005). Self-regulation and self-pre
Stolberg, S. G. (2004, January 25). Whoop, oops sentation: Regulatory resource depletion
and the state of the political slip. New York impairs impression management and
Times Week in Review, sec. 4, pp. 1, 3. effortful self-presentation depletes regula
Tedeschi, J. T., & Norman, N. (1985). tory resources. Journal of Personality and
Social power, self-presentation, and the Social Psychology, 88, 632–657.
self. In B. R. Schlenker (Ed.), The self in Vrij, A., Evans, H., Akehurst, L., & Mann, M.
social life (pp. 293–321). New York: (2004). Rapid judgements in assessing verbal
McGraw-Hill. and nonverbal cues: Their potential for
Tice, D. M., & Baumeister, R. F. (2001). deception researches and lie detection.
The primacy of the interpersonal self. Applied Cognitive Psychology, 18, 283–296.
In C. Sedikedes & M. Brewer (Eds.), Wroblewska, A. M. (1997). Androgen-anabolic
Individual self, relational self, collective self steroids and body dysmorphia in young
(pp. 71–88). Philadelphia: Psychology Press. men. Journal of Psychosomatic Research,
Tiedens, L. Z., & Fragale, A. R. (2003). Power 42, 225–234.
moves: Complementarity in dominant and Zebrowitz, L. A. (1997). Reading faces: Window
submissive nonverbal behavior. Journal of to the soul? Boulder, CO: Westview Press.
17-Manusov.qxd 6/30/2006 5:09 PM Page 340
18-Manusov.qxd 6/30/2006 5:10 PM Page 341
18
NONVERBAL COMMUNICATION
AND DECEPTION
� Aldert Vrij
University of Portsmouth, United Kingdom
◆ 341
18-Manusov.qxd 6/30/2006 5:10 PM Page 342
342–––◆–––Functions
different behaviors than do truth tellers. carefully to assess whether they are getting
According to Zuckerman, DePaulo, and away with their lie.
Rosenthal (1981), three factors could influ The extent to which lying is demanding
ence cues to deception: (1) emotional reac often depends on the type of lie. Telling an
tions, (2) content complexity, and (3) outright lie might be more cognitively chal
attempted behavioral control. Each of these lenging than concealing information, and
factors may influence a liar’s nonverbal telling an elaborate lie might well be more
behavior and emphasize a different aspect of demanding than providing short yes or no
deception. answers. Lying may be more demanding
Telling a lie is associated most com when the lie is not well prepared or
monly with three different emotions: fear, rehearsed. People engaged in cognitively
guilt, or delight (Ekman, 1985/2001). Liars complex tasks make more speech hesita
might feel guilty because they are lying, tions (e.g., stutters) and speech errors, speak
might be afraid of getting caught, or might slower, pause more, and wait longer before
be excited about having the opportunity to giving an answer (Goldman-Eisler, 1968).
fool someone. The strength of these emo Cognitive complexity also leads to fewer hand
tions depends on the personality of the liar and arm movements (Ekman & Friesen,
and on the circumstances under which 1972) and to more gaze aversion, because
the lie takes place (Ekman, 1985/2001; looking the conversation partner in the eye
Vrij, 2000a). Importantly, the experience of can be distracting (Doherty-Sneddon, Bruce,
guilt, fear, and excitement may influence a Bonner, Longbotham, & Doyle, 2002).
liar’s behavior. For example, guilt might Liars may realize that observers look at
result in gaze aversion if the liar does not their behavioral reactions to judge whether
dare to look the target straight in the eye they are lying and may, therefore, attempt
while telling a lie. Fear and excitement to control their behavior to appear credible.
might result in signs of stress, such as an To be successful, typically, liars must sup
increase in movements, an increase in press their nervousness while masking evi
speech hesitations (mm’s and errrr’s) and dence of having to think hard. They should
speech errors (stutters, repetition of words, also be able to show “honest-looking”
omission of words), or a higher pitched behaviors and avoid “dishonest-looking”
voice. The stronger the emotion, the more behaviors (Hocking & Leathers, 1980).
likely it is that some of these behaviors will These “requirements” mean that liars may
reveal deceit (Ekman, 1985/2001). need to act, but they must also avoid behav
Lying may be a cognitively demanding ior appearing planned, rehearsed, and
task. To avoid getting caught, liars need to lacking in spontaneity. According to this
provide plausible answers while avoiding theoretical perspective, liars’ motivation and
contradicting themselves. They must tell a efforts to control their behavior will increase
lie that is consistent with everything when the stakes (negative consequences of
the observer knows or might find out. getting caught or positive consequences of
Liars also need to remember what they succeeding) increase (Ekman, 1985/2001).
have said, so that they can say the same
things again when asked to repeat their
story. They may also feel an urge to control DEPAULO’S SELF
their demeanor so that they will appear PRESENTATIONAL PERSPECTIVE
honest (as emphasized in the attempted
control process below), and they may Zuckerman et al.’s (1981) perspective
observe the target person’s reactions predicts that the more liars experience one
18-Manusov.qxd 6/30/2006 5:10 PM Page 344
344–––◆–––Functions
or more of the three factors (emotion, con face-to-face encounters, liars must accom
tent complexity, behavioral control), the plish numerous communication tasks simul
more likely it is that cues to deception will taneously. They must produce a credible
occur. These factors are present only to a verbal message while projecting credible
limited extent in the majority of lies that nonverbal behavior simultaneously. They
people tell (DePaulo, Kashy, Kirkendol, must also manage their emotions, attend to
Wyer, & Epstein, 1996). In her self- their conversation partner while keeping the
presentational perspective, DePaulo (1992; dialogue running smoothly, send desired
DePaulo et al., 2003) argues that emotions, relational messages to their conversation
content complexity, and behavioral control partner and respond appropriately to what
might also influence truth tellers’ behavior. is said, and be discreet about any intentions
Thus, liars may be afraid of not being to deceive their partner. IDT embraces
believed in high-stakes situations, but so will Zuckerman et al.’s (1981) factors (emotion,
truth tellers, because they too could face content complexity, and attempted behavior
negative consequences if they fail to con control) as underlying reasons for cues to
vince others. Given the similarities between deceit (Burgoon, Buller, White, Afifi, &
liars and truth tellers, this perspective thus Buslig, 1999). In addition, it emphasizes
predicts that clear, diagnostic nonverbal that when deception occurs in interactive
cues to deception are unlikely to exist. contexts, it is not a unidirectional activity.
According to DePaulo et al. (2003), liars Rather, both liar and receiver mutually
and truth tellers will succeed in their social influence each other (Burgoon, Buller,
interaction goals only if they appear sincere. Floyd, & Grandpre, 1996).
The difference between lying and truth According to IDT, receivers’ behavior
telling is that the liar’s claim to honesty is may influence senders’ behavioral displays
illegitimate, and this lack of legitimacy has both directly, via synchrony, and indirectly,
two implications. First, deceptive self-presen because it may trigger behavioral adjust
tations might be less embraced as truthful ments (Burgoon et al., 1999). Regarding the
self-presentations less convincingly (e.g., direct effects, when people communicate
because liars have moral scruples, lack emo with each other, matching and synchrony
tional investment in their false claims, or lack may take place (Burgoon et al., 1999;
the knowledge and experience to back up Chartrand & Bargh, 1999; see Tickle-
their deceptive statements convincingly). Degnen, this volume). People may mirror
Second, liars typically experience a greater each other’s posture, or they may converge
sense of awareness and deliberateness in in how quickly and how loudly they speak.
their performances than truth tellers, because They may also reciprocate each other’s
they may take their credibility less for gazing, nodding, accents, and smiling behav
granted than truth tellers. Trying to appear ior (DePaulo & Friedman, 1998). This
convincing deliberately, however, might be “chameleon effect” (Chartrand & Bargh,
counterproductive. 1999) emerges even when strangers interact
with each other, and it happens typically
within a few minutes (Chartrand & Bargh,
BULLER AND BURGOON’S
INTERPERSONAL 1999). Furthermore, the indirect effects are
DECEPTION THEORY related to feedback from the receiver: When
liars are exposed to negative feedback from
A third perspective on deception, Buller the receiver, expressed through either verbal
and Burgoon’s (1996) interpersonal decep comments or through nonverbal behavior,
tion theory (IDT), postulates that during liars might realize that their performance is
18-Manusov.qxd 6/30/2006 5:10 PM Page 345
lacking credulity. Consequently, liars might review to date assessing the consistency
respond by making behavioral adjustments and strength of certain nonverbal cues
to diminish suspicions. indicating deception. Their meta-analysis
includes 116 studies, although not all these
SUMMARY projects focus on nonverbal cues to decep
tion, and involves 102 different nonverbal
The three perspectives discussed here cues. Most of the studies were experimental
make clear that the relationship between studies where university students lied or
lying and deceptive behavior is complex. told the truth for the sake of the experi
Zuckerman et al.’s (1981) assumptions that ment. Sometimes efforts were made to
liars might show signs of emotions and cog motivate the participants, for example, by
nitive load seem straightforward, yet liars promising them a financial reward if they
often do not experience emotions and were convincing.
high cognitive load (DePaulo et al., 1996). Significant findings emerged for 23 cues,
DePaulo et al.’s (2003) self-presentation per and these are listed in Table 18.1. Nine of
spective stresses that such experiences are not those cues, listed in the bottom half of
the exclusive domain of liars. Truth tellers Table 18.1, were investigated in only a few
may experience them as well, and, as a result, studies and will not be discussed further.
may also display nonverbal cues associated The cues are ranked in terms of their effect
with emotion or cognitive load. The sizes. Cohen (1977) suggested that effect
attempted behavioral control prediction is sizes of .20, .50, and .80 should be inter
not straightforward given that the behaviors preted as small, medium, and large effects,
shown by deceptive senders, as a result of this respectively. The effect sizes in the meta
deliberate control, will depend on both their analysis were typically small. No one cue
perceptions of what constitutes a credible was related significantly to deception across
nonverbal display and their acting skills in studies. The highest effect sizes were found
performing this display. Finally, IDT’s inter in the cues that were not often investigated
active approach implies that deceptive behav (bottom half of Table 18.1), but if the con
ior might be influenced directly by the centration is only on the cues that were
behavior of the receiver (a result of the investigated more often, then the largest
chameleon effect) or indirectly influenced by effect size was found for verbal and vocal
the suspicions raised by the receiver (Burgoon immediacy, d = –.55. The second highest
et al., 1999). The complex relationship cue, pupil dilation, obtained a d-score
between nonverbal communication and of .39. Most other cues obtained effect sizes
deception thus makes it unlikely that clear, of around .20 (see DePaulo et al., 2003, for
diagnostic, nonverbal cues to deception exist. all effect sizes, information about the indi
Deception research, summarized in the next vidual studies, definitions of the nonverbal
section, has supported this view. cues, and the impact of several moderating
factors on these cues).
346–––◆–––Functions
Cues d
348–––◆–––Functions
more metaphoric gestures, which are typi What happens in situations where the
cally made when people describe abstract stakes are really high, for example, when
ideas (McNeill, 1992). Illustrators as a guilty suspects lie in police interviews, when
whole entity (i.e., all the different types com smugglers go through customs at airports,
bined) were not linked to deception. Like or when adulterous husbands are chal
smiles, it was only when specific types of lenged by their wives? To examine how
distinctions were made among subclasses of liars respond in high-stakes situations, one
behavior that deception cues were detected. of the few options is to analyze such real-
life high-stakes situations. It is difficult,
however, to capture such lies on tape and to
INDIVIDUAL AND establish the ground truth in such situa
SITUATIONAL DIFFERENCES tions, that is, to know for certain that some
one was actually lying or telling the truth
Another reason for limited number of (Vrij & Mann, 2003), making such studies
nonverbal indicators of deceit may be that a difficult to undertake.
meta-analysis cannot capture signs of deceit Mann, Vrij, and Bull (2002) published
at an individual level. That is, different the most comprehensive study to date
individuals may give their lies away in dif about people’s behavior in real-life high-
ferent ways (DePaulo et al., 2003); such stakes situations. They examined the
idiosyncratic cues do not become apparent behavioral responses of 16 suspects while
when the focus is across studies. Similarly, they lied and told the truth during their
cues to deception could be dependent on the police interviews. The police interviews
situational context for the lie. A meta-analy were videotaped, and the tapes were
sis that accumulates findings across contexts made available for detailed scoring of the
would not apprehend those trends either. suspects’ behavioral responses. The sus
Furthermore, more cues to deception are pects were interviewed in connection with
likely to occur when the stakes are high serious crimes such as murder, rape, and
rather than low. In high-stakes situations, arson and were facing long custodial sen
liars might feel stronger emotions, might tences if found guilty. Regarding the
experience more cognitive demand, and ground truth, clips of video footage were
might be more motivated to manage their selected where other sources (reliable wit
behavior to appear credible. In their meta ness statements and forensic evidence) pro
analysis, DePaulo et al. (2003) com vided conclusive evidence that the suspect
pared higher-stakes studies (e.g., studies lied or told the truth.
where financial incentives were promised if Results revealed that compared with
the participant appeared credible) with when they told the truth, the suspects
lower-stakes studies. Some cues to decep exhibited more pauses, fewer eye blinks,
tion, such as an increase in blinking, a and fewer hand and arm movements (by
decrease in leg and foot movements, and an male suspects) when they lied. Indicators of
increased speech rate, appeared only in being tense (such as fidgeting and gaze aver
higher-stakes situations. The differences sion) did not emerge. These indicators are
between liars and truth tellers were still the behavioral patterns that police officers
small, however, perhaps because a high- expect typically in lying suspects. Mann
stakes situation will also affect truth tellers et al.’s results suggest that the suspects’ cues
or, alternatively, because the stakes were still to deception were more likely the result
not high enough in these laboratory-based, of increased cognitive demand than of
higher-stakes studies. nervousness. The strongest evidence for this
18-Manusov.qxd 6/30/2006 5:10 PM Page 349
was the reduction in eye blinks during the person was lying or telling the truth.
deception. Research has shown that ner There is no opportunity for the lie detectors
vousness results in an increase in eye to check the veracity of the statement via
blinking (Harrigan & O’Connell, 1996), physical evidence, third parties, and so on,
whereas increased cognitive load results in so the only source of information is the
a decrease in eye blinking (Wallbott & verbal and nonverbal behavior displayed in
Scherer, 1991). the videos. In such a study, simply guessing
The apparent predominance of cogni whether the person lied or spoke the truth
tive load processes compared with emo would result in an accuracy rate (percent
tional processes in those suspects is age of correct classifications) of 50%.
perhaps not surprising. Many of the sus Vrij (2000a) reviewed 37 of such stud
pects included in Mann et al.’s (2002) ies. In most studies, the accuracy rates
study have had regular contact with the varied between 45% and 60%, with an
police and were probably familiar with the average accuracy rate of 56.6%, only just
police interview situation, thereby decreas above the level of chance. When accuracy
ing their nervousness. Suspects in police at detecting lies was computed separately
interviews are typically of below average from accuracy at detecting truth, results
intelligence, however (Gudjonsson, 2003). showed that observers were reasonably
There is evidence that less intelligent good at detecting truths (correctly judging
people will have particular difficulty in that someone was telling the truth: 67%
inventing plausible and convincing stories accuracy rate) but particularly poor at
(Ekman & Frank, 1993). Alternatively, it detecting lies (correctly judging that some
might well be that the suspects were more one was lying: 44% accuracy rate). This
tense when they lied but that this was rate is below the level of chance, and
momentarily suppressed when they had to people would be more accurate at detecting
think hard. There is evidence that cognitive lies if they simply guessed.
demand results in an automatic and momen The superior accuracy rate for truthful
tary suppression of arousal (Jennings, messages is likely caused at least in part by
1992; Leal, 2005). the truth bias: Judges are more likely to
The absence of clear diagnostic nonver consider that messages are truthful than
bal cues to deceit makes detecting deceit a deceptive and, as a result, truthful messages
difficult task. People’s ability to detect are identified with more accuracy than are
deceit and reasons why they sometimes fail deceptive messages. There are at least four
to catch liars will be discussed in the next explanations for the truth bias (Gilbert,
two sections. 1991; Vrij & Mann, 2003). First, in daily
life people are more often confronted with
truthful than with deceptive statements,
♦ Accuracy in Detecting Deceit so they are more inclined to assume that
the behavior they observe is honest (what
LAYPERSONS’ LIE has been labeled the availability heuristic;
DETECTION IN STRANGERS O’Sullivan, Ekman, & Friesen, 1988).
Second, social conversation rules discour
In a typical lie detection study, observers age people from displaying suspicion. It is
(normally undergraduate students) are often necessary, however, to challenge what
shown video clips of strangers who are the person is saying and ask for more infor
either lying or telling the truth. They mation to detect deceit (Vrij, 2000a,
are asked to indicate after each clip whether 2004b). Third, people may be unsure as to
18-Manusov.qxd 6/30/2006 5:10 PM Page 350
350–––◆–––Functions
whether deception is actually occurring. The results of these studies did not sup
Given this uncertainty, the safest and port the idea that it is easier to detect lies in
most polite strategy may be to believe friends or lovers than in strangers. As rela
what is expressed overtly (DePaulo, Jordan, tionships become more intimate, partners
Irvine, & Laser, 1982). Fourth, based on a develop a strong tendency to judge the
Spinozan model of knowledge representa other, a tendency that has been called the
tion, Gilbert (1991) argues that everything relational truth-bias heuristic (Stiff, Kim, &
is taken to be true initially, and that disbe Ramesh, 1992). As the relationship between
lief requires extra effort. In other words, two people intensifies, they often become
the truth bias is the default setting when more confident that they can detect each
interacting with strangers. other’s lies. High levels of confidence tend
Lie detection is not only difficult when to result in the belief that the other person
adults are involved. When adults are asked to would probably not dare to lie, which sub
detect truths and lies in children with whom sequently reduces the need of trying to dis
they are not familiar, particularly in situa cover whether that person is lying (Levine
tions where they can only rely on the child’s & McCornack, 1992).
verbal and nonverbal behavior, they do not Anderson et al. (1999) provided addi
perform much better than when they are tional explanations as to why people may
asked to detect lies in adults. Parents, how fail to detect deceit in their close friends
ever, seem to be better than other adults at or romantic partners. They suggested that
detecting lies told by their own children (see when close relationship partners attempt
Vrij, 2002, for a review of child deception to detect deceit in each other, they bring
literature). Deception is also harder to spot to mind a great deal of information about
across cultures (e.g., when natives and immi each other. This information could be over
grants in a country interact) than within the whelming, and the lie detector might deal
same culture (see Bond & Rao, 2004, for a with this load by processing the informa
review of cross-cultural lie detection). tion heuristically instead of searching care
fully for genuine cues to deceit. Also, in
close relationship interactions, the lie detec
LAYPERSONS’ LIE DETECTION tor must engage simultaneously in social
IN FRIENDS, PARTNERS, behavior (e.g., the need to appear support
OR CHILDREN ive in those interactions) and social cogni
tion (e.g., decoding possible cues to
Boon and McLeod (2001) reported that deception) (Patterson, 1995). This might be
people believe that they are fairly successful too much for the lie detector and, as a
in detecting lies in their partners. They also result, valuable cues may remain unnoticed.
believe that they themselves are more suc Finally, it could be that as relationships
cessful in deceiving their partners than their develop, the partners become more
partners are at deceiving them. In several skilled at crafting communications designed
experiments, however, observers were uniquely to fool each other.
shown videotaped truthful or deceptive
statements of people they know, and were
asked to judge the veracity of each state PROFESSIONAL LIE CATCHERS
ment (Anderson, Ansfield, & DePaulo,
1999; Vrij, 2000a). As in most studies, Granhag and Vrij (2005) reviewed 10
there was no physical evidence or third- studies in which professional lie catchers,
party information to rely on. such as police officers, participated as
18-Manusov.qxd 6/30/2006 5:10 PM Page 351
judges. In these studies, the professionals cue uniquely related to deception. The
saw clips of people they did not know, and other reasons are a little more complicated
no physical or third-party evidence was and merit more discussion.
available to them. Granhag and Vrij found
an accuracy rate (lies and truths combined)
of 55%. This accuracy rate is similar to the THE USE OF HEURISTICS
accuracy rate found with students. Results
revealed, however, that professionals are Rather than scrutinizing another’s reac
equally good at detecting truths (55%) and tions actively for cues to deceit, observers
lies (55%), suggesting that the truth bias, may instead rely on rule of thumb decision
typically found in students, does not seem rules or cognitive heuristics (Levine, Park,
to occur with professional lie catchers. & McCornack, 1999). Such heuristics
In another study, Mann, Vrij, and Bull easily lead to systematic errors and biases.
(2004) showed 99 police officers videotapes Two heuristics that are thought to influence
consisting of a total of 54 truths and lies veracity judgments, the availability heuris
told by suspects during their police inter tic and the relational truth-bias heuristic,
views. These clips were derived from the were discussed earlier. Another heuristic
Mann et al. (2002) study described previ sis the probing heuristic (Levine &
ously. Again, no physical or third-party evi McCornack, 2001), and it refers to judges’
dence was available to the observers. In this tendency to believe a source more after the
study, however, the officers were capable source has been probed. Receivers often
of detecting, on average, 63% of the truths have a strong belief in the efficacy of prob
and 66% of the lies. These accuracy rates ing as a lie detection strategy. In cases
were higher than those found in most pre where probing does not result in clear signs
vious lie detection studies. One explanation of deceit, and it often will not (Levine &
for this greater accuracy is that the stakes McCornack), the source is more likely to
for the suspects in this study were consider be believed. The representativeness heuris
ably higher than the stakes for the liars and tic (Stiff et al., 1989) refers to the tendency
truth tellers in most lie detection studies, to evaluate a particular reaction (e.g., ner
and as noted, high-stake lies may be easier vous behavior) as an example of a broader
to detect than low-stake lies (see, Vrij, category (e.g., deception). The expectancy-
2000b, for a review). violation heuristic (Vrij, 2004b) relates to
observers’ tendency to judge reactions that
are odd or infrequent, such as keeping the
♦ Reasons Why Lies eyes closed or staring while speaking, as
deceptive (Bond et al., 1992). In addition,
Remain Undetected
the facial appearance heuristic (Vrij,
2004b) refers to observers’ tendency to
Despite some evidence that certain kinds of judge people with attractive faces or with
lies may be easier to detect, most lies are not a baby-faced appearance as more honest
detected easily. There are numerous reasons (Aune, Levine, Ching, & Yoshimoto, 1993;
why lies often remain undetected (Vrij, Bull & Rumsey, 1988).
2004a, 2004b). Those related to nonverbal Furthermore, O’Sullivan (2003) demon
communication are discussed here. As strated that the fundamental attribution
already argued, the most basic reason for error (FAE) can undermine lie detection.
the failure to detect lies is that there is no The FAE is the tendency, when forming
single verbal, nonverbal, or physiological impressions of others, to overestimate
18-Manusov.qxd 6/30/2006 5:10 PM Page 352
352–––◆–––Functions
dispositional causes of that person and beliefs are established, people tend to seek
to underestimate situational causes. When information that confirms rather than dis-
observers believe that someone is generally confirms these beliefs (confirmation bias)
a trustworthy person, they will have the and therefore have a biased evaluation of
tendency to judge that person as truthful new information. Consequently, they will
in any given situation. Similarly, when discount information that contradicts
observers believe that someone is an their beliefs (belief perseverance). These
untrustworthy person, they will be inclined processes result in people thinking that
to judge that person as dishonest in any their beliefs are more grounded in reality
given situation. than they are. In turn, this makes it unlikely
that they will change their beliefs (Granhag
& Vrij, 2005; Strömwall et al., 2004).
LOOKING AT THE WRONG CUES Fourth, it is often difficult for people to
test the accuracy of their beliefs because of
Studies investigating how people think a lack of reliable feedback (i.e., we seldom
liars behave have been carried out world learn whether our lie detection strategy is
wide, albeit overwhelmingly with Caucasian accurate). Interestingly, research suggests
participants. These studies suggest that that prisoners have the most accurate
people predominantly expect liars to react beliefs about deception (Hartwig, Granhag,
nervously, with “liars look away” being Strömwall, & Andersson, 2004; Vrij &
the most often mentioned belief (see Bond Semin, 1996). They may well live in a
& Rao, 2004; Strömwall, Granhag, & world that involves a lot of deception.
Hartwig, 2004; Vrij, Akehurst, & Knight, Success in such a culture may depend in
in press; Vrij, 2004b, for reviews). The part on being able to tell when one is the
belief that liars increase their movements is target of deception.
also widespread (Strömwall et al., 2004).
Gaze, however, does not appear to be reli
ably related to deception (DePaulo et al., PEOPLE DO NOT TAKE
2003; Vrij, 2000a). In addition, liars tend INDIVIDUAL AND SITUATIONAL
to decrease rather than increase their move DIFFERENCES INTO ACCOUNT
ments (DePaulo et al., 2003; Vrij, 2000a;
see Table 18.1). There are large individual differences in
There are several reasons why these people’s behavior. People may fail to catch
incorrect beliefs exist. First, because media liars because they do not take such individ
often promote the idea that liars look away ual differences into account when they
and fidget, many people may have been attempt to detect deceit. People whose
exposed to and absorbed this information. normative behavior looks suspicious are in
Even police manuals often suggest that liars a particularly disadvantageous position.
fidget and look away, which may explain Some individuals’ nonverbal behavior gives
why police officers also endorse these the impression that they are telling the truth
beliefs overwhelmingly (Vrij, 2004b). (honest demeanor bias), whereas others’
Second, people believe mistakenly that they natural behavior leaves the impression that
show nervous behaviors themselves when they are lying (dishonest demeanor bias)
they lie (Vrij, Edward, & Bull, 2001a), and, (Frank & Ekman, 2004; Riggio, Tucker, &
during lie detection, observers may look for Throckmorton, 1988).
cues they mistakenly believe they themselves Such demeanor biases are related to
show while lying. Third, once incorrect personality traits. Expressive people, for
18-Manusov.qxd 6/30/2006 5:10 PM Page 353
example, exude credibility, regardless of the differences); the same person also behaves
truth of their assertions, because their spon differently across situations (intrapersonal
taneity tends to disarm suspicion (Riggio, differences). For example, people show dif
1986). People with a strong sense of public ferent behaviors when with friends in a
self-consciousness tend to make a less credi restaurant than while interviewed for a
ble impression on others. When these indi job application. Establishing how people
viduals lie, they are concerned about being behave in the former setting can therefore
scrutinized by others, which changes their not be used to establish whether they are
behavior in such a way that it appears dis lying in the latter setting. Yet police manuals
honest. Introverts and socially anxious advise police detectives to establish a sus
people also tend to make a suspicious pect’s baseline behavior on the initial small
impression. The social clumsiness of many talk part of the interrogation, and to com
introverts and the impression of tension, pare this with the behavior shown during
nervousness, or fear of socially anxious indi the accusational part of the interrogation
viduals are often interpreted by observers as (Inbau, Reid, Buckley, & Jayne, 2001).
indicators of deceit. Interestingly, introverts’
demeanor seems not to reflect their actions
accurately. Specifically, research has found ♦ Discussion
that introverted people lie infrequently
(Kashy & DePaulo, 1996).
Errors in interpreting someone’s behavior This chapter makes clear that the relation
arise easily in cross-ethnic or cross-cultural ship between deception and nonverbal
interactions because of differences in behav communication is complex, and distin
ior that may be displayed by different groups. guishing accurately between lies and truths
For example, African Americans display is often challenging. Having difficulty with
more gaze aversion generally than do Euro- detecting deceit, however, does not harm
Americans (LaFrance & Mayo, 1976). Such people in many daily life situations. On the
differences are, in part, based on the fact that contrary, being ignorant about the truth
gaze patterns are influenced by culture, and often serves us well. What would we think
that looking into the eyes of a conversation of ourselves if we come to know the truth
partner is regarded as polite in some Western about every flattering comment we receive?
cultures but is considered to be rude in sev And how would we respond if other people
eral other cultures (Vrij & Winkel, 1991). In could detect every lie we tell, including our
support of this argument, several experimen white lies? Telling lies is very much part
tal studies revealed that behaviors displayed of everyday life (DePaulo et al., 1996;
typically by members of some ethnic groups Goffman, 1959; Vrij, in press) and some
living in the Netherlands make a suspicious thing that is expected in the smooth work
impression on Caucasian Dutch police offi ing of social relations. Sometimes, however,
cers, a phenomenon labeled cross-cultural it is important to be able to detect lies. For
nonverbal communication errors (see Vrij, example, it would benefit police investiga
2000a, for an overview of these studies). That tions if police detectives were able to tell
is, nonverbal behavioral patterns normative when suspects are lying and when they are
for certain ethnic groups may be interpreted telling the truth during police interviews.
by members of other ethnicities as revealing There are opportunities for improving lie
attempts to hide the truth. detection ability by paying attention to
Not only do different people behave dif people’s behavior. To conclude this chap
ferently in the same situation (interpersonal ter, I will discuss these ideas.
18-Manusov.qxd 6/30/2006 5:10 PM Page 354
354–––◆–––Functions
under investigation. Thus, if the response instantly the nonverbal behavior displayed
under investigation is a high-stakes situa by target persons (see Vrij, 2004b).
tion, then the baseline situation should also
be a high-stakes situation. Moreover,
there is evidence that some behaviors are INTERVIEWING TO
topic related. While analyzing the behavior DETECT DECEPTION
shown by Saddam Hussein in an interview
with Peter Arnett during the First Gulf War Mann et al.’s (2002) examination of
(the interview was broadcasted on CNN), real-life police interviews, discussed earlier,
Davis and Hadiks (1995) found that the suggested that when suspects lie, they
topic Hussein discussed was related to the experience high cognitive demand. Police
illustrators he made. Only when discussing interviewers could use this knowledge by
Israel and Zionism did Hussein make spe employing interview techniques that increase
cific movements with his left forearm. Likely, the cognitive demand in suspects. This
Israel has a special meaning to an Arab should have a greater effect on liars than on
leader, and issues such as personal involve truth tellers, thus facilitating discrimination
ment should be taken into account as well. between them. There are several ways in
Vrij and Mann (2001) employed the which cognitive demand could be increased.
comparable truth technique in an actual For example, suspects could be asked to
police interview. During a videotaped real- elaborate on or repeat what they have just
life police interview, a man suspected and said. Using this technique, lies could well
later convicted of murder was asked to fail if the liar did not plan the lie in suffi
describe his activities on a particular day. cient detail.
The murder suspect described his activities Moreover, liars tend to speak of events
during the morning (went to work), after in a more fixed chronological order (this
noon (visited a market), and evening (vis happened first, and then this, and then that,
ited a neighbor). Detailed analyses of the etc.) than truth tellers. It has been suggested
videotape revealed a sudden change in that it is difficult for liars to fabricate a
behavior as soon as he started to describe story in a nonchronological order (Vrij,
his activities during the afternoon and 2005). Lie detectors could exploit this diffi
evening. One reason for this change may culty by asking interviewees to tell their
have been that he was lying and evidence stories in a nonchronological order, for
supported this view. Police investigations example, in reverse order. Police officers
could confirm his story with regard to his could also use the evidence they have
morning activities but revealed that his against a suspect strategically. Inbau et al.
statements about the afternoon and even (2001), for example, advise the police to
ing were fabricated. In reality, he met the present such evidence at the beginning of
victim in the afternoon and killed her later the interview (e.g., “Our CCTV footage
on that day. In this case, my colleague and shows that you were in Commercial Road
I were able to make a good comparison. on Saturday evening at 8 p.m.”). The task
The comparable truth method would bene the lying suspect then faces is to fabricate
fit, however, from instant, accurate, and an alibi that is consistent with this factual
detailed coding of all sorts of behaviors evidence. Alternatively, the police do not
shown by a target person. Progress is being reveal the evidence initially but first let the
made in developing technical equipment suspect talk about his whereabouts. In an
that can observe nonintrusively and code experiment where the timing of presenting
18-Manusov.qxd 6/30/2006 5:10 PM Page 356
356–––◆–––Functions
the evidence was manipulated (it was expectancy violation. Journal of Personality
presented either before or after the intervie and Social Psychology, 63, 969–977.
wee was given the opportunity to discuss Bond, C. F., & Rao, S. R. (2004). Lies travel:
his or her activities) lies were more readily Mendacity in a mobile world. In P. A.
Granhag & L. A. Strömwall (Eds.), Deception
detected by observers when the evidence
detection in forensic contexts (pp. 127–147).
was presented at a later stage (Hartwig,
Cambridge, UK: Cambridge University Press.
Granhag, Strömwall, & Vrij, 2005).
Boon, S. D., & McLeod, B. A. (2001). Deception
People have considerable difficulty in in romantic relationships: Subjective esti
accurately distinguishing between truths mates of success at deceiving and attitudes
and lies, among other factors, because they toward deception. Journal of Social and
often fail to take individual and situational Personal Relationships, 18, 463–476.
differences into account; attend to the Bull, R., & Rumsey, N. (1988). The social psy
wrong, nondiagnostic, cues, and use heuris chology of facial appearance. New York:
tics that lead to systematic errors and biases. Springer-Verlag.
It is argued in this chapter, however, that Buller, D. B., & Burgoon, J. K. (1996).
people could become better lie detectors if Interpersonal deception theory. Communi
cation Theory, 6, 203–242.
they used alternative lie detection strategies.
Buller, D. B., Stiff, J. B., & Burgoon, J. K.
(1996). Behavioral adaptation in deceptive
transactions. Human Communication
♦ References Research, 22, 589–603.
Burgoon, J. K., Buller, D. B., Floyd, K., &
Grandpre, J. (1996). Deceptive realities:
Anderson, D. E., Ansfield, M. E., & DePaulo, B. Sender, receiver, and observer perspectives
M. (1999). Love’s best habit: Deception in in deceptive conversations. Communication
the context of relationships. In P. Philippot, Research, 23, 724–748.
R. S. Feldman, & E. J. Coats (Eds.), Burgoon, J. K., Buller, D. B., White, C. H.,
The social context of nonverbal behavior Afifi, W., & Buslig, A. L. S. (1999). The
(pp. 372–409). Cambridge, UK: Cambridge role of conversation involvement in decep
University Press. tive interpersonal interactions. Personality
Anderson, D. E., DePaulo, B. M., Ansfield, M. and Social Psychology Bulletin, 25,
E., Tickle, J. J., & Green, E. (1999). Beliefs 669–685.
about cues to deception: Mindless stereo Caso, L., Maricchiolo, F., Bonaiuto, M., Vrij,
types or untapped wisdom? Journal of A., & Mann, S. (2006). The impact of
Nonverbal Behavior, 23, 67–89. deception and suspicion on different hand
Aune, R. K., Levine, T., Ching, P., & movements. Journal of Nonverbal Behavior,
Yoshimoto, J. (1993). The influence of per 30, 1–19.
ceived source reward value and attributions Chartrand, T. L., & Bargh, J. A. (1999). The
of deception. Communication Research chameleon effect: The perception-behavior
Reports, 10, 15–27. link and social interaction. Journal of
Bavelas, J. B., Chovil, N., Coates, L., & Roe, L. Personality and Social Psychology, 76,
(1995). Gestures specialized for dialogue. 893–910.
Personality and Social Psychology Bulletin, Cohen, J. (1977). Statistical power analysis
21, 394–405. for the behavioral sciences. New York:
Bond, C. F., & DePaulo, B. M. (2005). Accuracy Academic Press.
of deception judgments. Unpublished Davis, M., & Hadiks, D. (1995). Demeanor and
manuscript. credibility. Semiotica, 106, 5–54.
Bond, C. F., Omar, A., Pitre, U., Lashley, B. R., DePaulo, B. M. (1992). Nonverbal behavior and
Skaggs, L. M., & Kirk, C. T. (1992). Fishy- self-presentation. Psychological Bulletin,
looking liars: Deception judgment from 111, 203–243.
18-Manusov.qxd 6/30/2006 5:10 PM Page 357
DePaulo, B. M., & Friedman, H. S. (1998). and deception in everyday life (pp. 184–
Nonverbal communication. In D. T. Gilbert, 201). New York: Guildford Press.
S. T. Fiske, & G. Lindzey (Eds.), The hand Ekman, P., & Friesen, W. V. (1969). The reper
book of social psychology (pp. 3–40). toire of nonverbal behavior. Semiotica, 1,
Boston, MA: McGraw-Hill. 49–98.
DePaulo, B. M., Jordan, A., Irvine, A., & Laser, Ekman, P., & Friesen, W. V. (1972). Hand
P. S. (1982). Age changes in the detection movements. Journal of Communication,
of deception. Child Development, 53, 22, 353–374.
701–709. Ekman, P., Friesen, W. V., & O’Sullivan, M.
DePaulo, B. M., Kashy, D. A., Kirkendol, S. E., (1988). Smiles when lying. Journal of
Wyer, M. M., & Epstein, J. A. (1996). Lying Personality and Social Psychology, 54,
in everyday life. Journal of Personality and 414–420.
Social Psychology, 70, 979–995. Elaad, E. (2003). Effects of feedback on the
DePaulo, B. M., & Kirkendol, S. E. (1989). The overestimated capacity to detect lies and the
motivational impairment effect in the com underestimated ability to tell lies. Applied
munication of deception. In J. C. Yuille Cognitive Psychology, 17, 349–363.
(Ed.), Credibility assessment (pp. 51–70). Feeley, T. H., & Young, M. J. (2000). The
Dordrecht, The Netherlands: Kluwer. effects of cognitive capacity on beliefs about
DePaulo, B. M., Lindsay, J. L., Malone, B. E., deceptive communication. Communication
Muhlenbruck, L., Charlton, K., & Cooper, Quarterly, 48, 101–119.
H. (2003). Cues to deception. Psychological Frank, M. G., & Ekman, P. (1997). The ability to
Bulletin, 129, 74–118. detect deceit generalizes across different types
DePaulo, B. M., & Morris, W. L. (2004). of high-stake lies. Journal of Personality and
Discerning lies from truths: Behavioral cues Social Psychology, 72, 1429–1439.
to deception and the indirect pathway Frank, M. G., & Ekman, P. (2004). Appearing
of intuition. In P. A. Granhag & L. A. truthful generalizes across different decep
Strömwall (Eds.), Deception detection in tion situations. Journal of Personality and
forensic contexts (pp. 15–40). Cambridge, Social Psychology, 86, 486–495.
UK: Cambridge University Press. Frank, M. G., & Ekman, P., & Friesen, W. V.
DePaulo, B. M., Stone, J. L., & Lassiter, G. D. (1993). Behavioral markers and recogniz
(1985). Deceiving and detecting deceit. In ability of the smile of enjoyment. Journal of
B. R. Schlenker (Ed.), The self and social Personality and Social Psychology, 64,
life (pp. 323–370). New York: McGraw- 83–93.
Hill. Frank, M. G., & Feeley, T. H. (2003). To catch
Doherty-Sneddon, G., Bruce, V., Bonner, L., a liar: Challenges for research in lie detec
Longbotham, S., & Doyle, C. (2002). tion. Journal of Applied Communication
Development of gaze aversion as disengage Research, 31, 58–75.
ment of visual information. Developmental Friesen, W. V., Ekman, P., & Wallbott, H.
Psychology, 38, 438–445. (1979). Measuring hand movements.
Efron, D. (1941). Gesture and environment. Journal of Nonverbal Behavior, 4, 97–112.
New York: King’s Crown. Gilbert, D. T. (1991). How mental systems
Ekman, P. (2001). Telling lies. New York: W. believe. American Psychologist, 46,
W. Norton. (Original work published 107–119.
1985) Goffman, E. (1959). The presentation of self in
Ekman, P., Davidson, R. J., & Friesen, W. V. everyday life. New York: Doubleday.
(1990). The Duchenne smile: Emotional Goldman-Eisler, F. (1968). Psycholinguistics:
expression and brain physiology II. Journal Experiments in spontaneous speech. New
of Personality and Social Psychology, 58, York: Doubleday.
342–353. Granhag, P. A., & Vrij, A. (2005). Deception
Ekman, P., & Frank, M. G. (1993). Lies that detection. In N. Brewer & K. Williams
fail. In M. Lewis & C. Saarni (Eds.), Lying (Eds.), Psychology and law: An empirical
18-Manusov.qxd 6/30/2006 5:10 PM Page 358
358–––◆–––Functions
judgments of honesty and deceit. Journal Vrij, A., Edward, K., & Bull, R. (2001a).
of Personality and Social Psychology, 56, People’s insight into their own behaviour
555–564. and speech content while lying. British
Strömwall. L. A., Granhag, P. A., & Hartwig, M. Journal of Psychology, 92, 373–389.
(2004). Practitioners’ beliefs about decep Vrij, A., Edward, K., & Bull, R. (2001b). Police
tion. In P. A. Granhag & L. A. Strömwall officers’ ability to detect deceit: The benefit of
(Eds.), Deception detection in forensic con indirect deception detection measures. Legal
texts (pp. 229–250). Cambridge, UK: and Criminological Psychology, 6, 185–197.
Cambridge University Press. Vrij, A., & Mann, S. (2001). Telling and detect
Vrij, A. (2000a). Detecting lies and deceit: The ing lies in a high-stake situation: The case
psychology of lying and its implications for of a convicted murderer. Applied Cognitive
professional practice. Chichester: Wiley. Psychology, 15, 187–203.
Vrij, A. (2000b). Telling and detecting lies as a Vrij, A., & Mann, S. (2003). Telling and detect
function of raising the stakes. In C. M. ing true lies: Investigating and detecting the
Breur, M. M. Kommer, J. F. Nijboer, & lies of murderers and thieves during police
J. M. Reintjes (Eds.), New trends in criminal interviews. In M. Verhallen, G. Verkaeke,
investigation and evidence II (pp. 699–709). P. J. van Koppen, & J. Goethals (Eds.),
Antwerpen, Belgium: Intersentia. Much ado about crime: Chapters on psy
Vrij, A. (2001). Implicit lie detection. The chology and law (pp. 185–208). Brussels,
Psychologist, 14, 58–60. Belgium: Uitgeverij Politeia.
Vrij, A. (2002). Deception in children: A Vrij, A., & Mann, S. (2005). Police use of non
literature review and implications for verbal behavior as indicators of deception.
children’s testimony. In H. Westcott, G. In R. E. Riggio & R. S. Feldman (Eds.),
Davies, & R. Bull (Eds.), Children’s testi Applications of nonverbal communication
mony (pp. 175–194). Chichester: Wiley. (pp. 63–94). Mahwah, NJ: Erlbaum.
Vrij, A. (2004a). Guidelines to catch a liar. In Vrij, A., & Semin, G. R. (1996). Lie experts’
P. A. Granhag & L. A. Strömwall (Eds.), beliefs about nonverbal indicators of decep
Deception detection in forensic contexts tion. Journal of Nonverbal Behavior, 20,
(pp. 287–314). Cambridge, UK: Cambridge 65–80.
University Press. Vrij, A., & Winkel, F. W. (1991). Cultural pat
Vrij, A. (2004b). Invited article: Why profes terns in Dutch and Surinam nonverbal
sionals fail to catch liars and how they behaviour: An analysis of simulated police/
can improve. Legal and Criminological citizen encounters. Journal of Nonverbal
Psychology, 9, 159–181. Behavior, 15, 169–184.
Vrij, A. (2005). Criteria-based content analysis: Wallbott, H. G., & Scherer, K. R. (1991). Stress
A qualitative review of the first 37 studies. specifics: Differential effects of coping style,
Psychology, Public Policy, and Law, 11, gender, and type of stressor on automatic
3–41. arousal, facial expression, and subjective
Vrij, A. (in press). Deception: A social lubricant feeling. Journal of Personality and Social
and a selfish act. In K. Fiedler (Ed.), Psychology, 61, 147–156.
Frontiers of social psychology: Social Zuckerman, M., DePaulo, B. M., & Rosenthal,
communication. New York: Psychology R. (1981). Verbal and nonverbal communi
Press. cation of deception. In L. Berkowitz (Ed.),
Vrij, A., Akehurst, L., & Knight, S. (in press). Advances in experimental social psychology
Police officers’, social workers’, teachers’ (Vol. 14, pp. 1–57). New York: Academic
and general public’s beliefs about deception Press.
in children, adolescents and adults. Legal
and Criminological Psychology.
18-Manusov.qxd 6/30/2006 5:10 PM Page 360
19-Manusov.qxd 6/30/2006 5:12 PM Page 361
19
THE INTERACTION MANAGEMENT
FUNCTION OF NONVERBAL CUES
Theory and Research About Mutual
Behavioral Influence in Face-to-Face Settings
� Joseph N. Cappella
University of Pennsylvania
� Darren M. Schreiber
University of California, San Diego
◆ 361
19-Manusov.qxd 6/30/2006 5:12 PM Page 362
362–––◆–––Functions
are often “multifunctional.” These behav M. levator labii, the frontal M. lateralis, and
iors often have several interpretations the M. corrugator supercilii muscle regions
depending on the context. The physiolo were obtained as were emotional experi
gical responses are less susceptible— ences. The results revealed that people both
but certainly not immune—to interpretive mimicked and experienced an emotion sim
variation. The less behavioral, more macro ilar to that expressed by the stimulus person.
scopic behaviors, move the research on The focus on coordination in emotions
coordination into a different research has extended to vocal emotion. Neuman
domain (e.g., the role of emotional and and Strack (2000) had people listen
other types of contagion in groups, organi to philosophical essays read in a slightly
zations, and other social units). The impli happy or slightly sad tone of voice.
cation of these two developments is to Listeners reported a more positive mood
direct that theories be capable of explaining with the happier reading and a more
both the automatic responses characteristic negative mood with the sadder reading. In
of physiological coordination and the a second study, they also repeated the essay
more subjective, deliberate processes of in a tone that independent raters found
emotional yoking between partners, per to be happier or sadder depending on
haps with expressive imitation mediating hearing the slightly happy or sad initial
the subjective yoking of emotional experi rendition—a kind of vocal coordination
ence. The following details examples of of emotion. Additionally, Hietanen et al.
research with these characteristics. (1998) obtained facial EMG responses to
The first of these concerns microcoordi vocal affect expressions as participants lis
nation. Electromyography (EMG) is the tened to single words uttered by two actors
study of facial muscle activity in response stimulating different emotions. Three cate
to various stimuli. Visual stimuli can elicit gories of expressions were selected: emo
facial muscle activity depending on the tional neutrality, anger, and contentment.
affective valence of the stimuli to the The EMG activity over two facial muscle
subject. For example, studies suggest that regions was measured: corrugator supercilii
stimuli related to positive affect increase (the muscle that knits the brows together)
activity in the cheek muscle region— and orbicularis oculi (the muscle that pro
smiling—and stimuli related to negative duces bagging below the eyes and wrinkles
affect increase activity in the brow muscle in the corners of the eyes). Hearing the
region—frowning (Hietanen, Surakka, & expression of anger increased EMG activity
Linnankoski, 1998). Some studies have in the participants’ brow region more than
investigated facial electromyographic respon did hearing the expression of contentment.
ses during actual interaction. Lundqvist In contrast, the expression of contentment
(1995), for example, explored whether activated the periocular muscle region more
people exposed to facial expressions than did anger. The results support the view
responded with specific facial muscle that negative and positive affects are “con
reaction patterns that correspond to spe tagious” from hearing human vocal affect
cific emotional experiences. Participants expressions.
were shown pictures of faces expressing The linkage between emotional expres
sadness, anger, fear, surprise, disgust, sion and felt emotion within the person is
happiness, as well as neutral facial expres an important set of facts that theories must
sions. At the same time, facial electromyo explain. In interpersonal contexts, if person
graphs from the M. zygomaticus major, the A mimics person B and if both A and B
19-Manusov.qxd 6/30/2006 5:12 PM Page 364
364–––◆–––Functions
experience emotion consistent with their depression between spouses where one was
facial displays, then we might reasonably primarily the caretaker and one the care
expect coordination between A and B in giver. Depression scores for the care receiver
their subjective experiences of emotional tended to determine those of the caretaker
or mood. This coordination in subjective at a later point in time, suggesting coordina
emotion and mood is examined next. tion but with the caregiver dominant. The
Specifically, Totterdell, Kellett, Teuchmann, opposite influences were not found with
and Briner (1998) explored “mood link regard to depression or a variety of other
age” in the context of work groups. They emotional states. Specifically, Goodman and
investigated whether people’s moods are Shippy (2002) studied elderly spouses where
influenced by the collective mood of their one was experiencing serious vision prob
work teammates over time. lems. After controlling for other factors,
In their first study, a time-series analysis depression by one spouse predicted the
showed a significant association between partner’s depression.
the nurses’ moods and the collective mood Individual differences in sensitivity to
of their teammates independent of shared emotions from others and the ability to trans
hassles. In their second study, a team of mit emotions to others may affect the exis
accountants rated their own moods and the tence and strength of observed contagion.
moods of their teammates three times a day Verbeke (1997) explored whether these
for 4 weeks using pocket computers. The individual differences are assets or liabilities
accountants’ moods and their judgments of over the long term for salespersons. Doherty
their teammates’ moods were significantly (1997) explored the individual differences in
associated with the collective mood of susceptibility to emotional coordination by
their teammates. The findings suggest that crafting an Emotional Contagion Scale, a
people’s moods can become linked to the 15-item measure of individual differences in
mood of their coworkers, offering a com susceptibility to catching others’ emotions.
pelling line of evidence for affective or, Participants were videotaped while watching
in this case, mood synchrony during adult videotapes of emotionally expressive stimulus
interaction. persons relating their happiest and saddest
Similar effects have been reported for memories. Doherty found that susceptibility
teacher burnout (i.e., emotional exhaus to emotional contagion was positively related
tion and depersonalization; Bakker & to reactivity, emotionality, sensitivity to
Schaufeli, 2001), in sales contexts (Verbeke, others, social functioning, and self-esteem.
1997), and in clinical environments (Hsee, Doherty, Orimoto, Singelis, Hatfield, and
Hatfield, Carlson, & Chemtob, 1992). Hebb (1995) showed further that women
Thompson, Nadler, and Kim (1999), for in a variety of occupations illustrated
example, argued that the ability of negotia higher total emotional contagion scores than
tors at the bargaining table is enhanced to did men.
the extent that they are successful in per Findings on mood transfer observed
ceiving emotions of participants, reacting in more applied settings have also been
appropriately to them, and being “in tune” obtained in more controlled environments
with those emotions. Pugh (1998) found (Gump & Kulik, 1997). In one study
that in a service context sales people were (Neuman & Strack, 2000), participants
more effective when they were emotionally were tested on their listening comprehen
congruent with their customers. Furthermore, sion in response to a neutral text that was
Ingram (1997) studied the coordination of read to them in a happy or sad tone of
19-Manusov.qxd 6/30/2006 5:12 PM Page 365
voice. Listeners reported being in an (but see Ullstadius, 1998, who offered
emotional state that was like that of the contrary evidence for the imitation of
read materials and when required to repeat tongue protrusions and mouth openings in
portions of the text read to them, employed 18 newborn infants). Other studies support
a vocal tone similar to the one they had the claim of mutual coordination between
heard. Similar findings by Hess and Blairy infants and adults. Stack and Arnold
(2001) suggest that aspects of facial mim (1998), for example, focused on maternal
icry of emotion may account for the subjec touch and its ability to influence infants’
tive emotional reactions that viewers of the gaze and affect during interactions. The
videotaped faces reported. Direct evidence results from this study indicated that (1)
of mediation from facial stimuli to facial infants were sensitive to subtle changes
mimicry by viewers to reports of emotional in maternal tactile-gestural behavior, (2)
state by viewers was not obtained. These maternal touch and hand gestures can
reactions to facial displays are likely to attract infants’ attention to their faces even
occur quickly (presentations less than half when still and expressionless, and (3) there
a second) and exhibit a dose-response rela were associations between infant expres
tionship with more intense displays eliciting siveness and gaze at mothers’ faces and
more intense emotional reactions (Wild, hands during these periods.
Erb, & Bartels, 2001). Symons and Moran (1994) extended the
idea of maternal influence to mutual influ
Summary. Overall, four conclusions can be ence—that is, observing infants’ smiling
drawn from the recent literature: (1) rela behavior as being both responsive to and
tively automatic responses to emotional dependent on maternal smiling behavior.
stimuli are manifest in facial and vocal reac Twenty-five mothers were observed engaged
tions; (2) these automatic reactions are in face-to-face interactions with their 8-,
sometimes accompanied by subjective feel 12-, 16-, and 20-week-old infants, the ages
ings of emotion; (3) mood and emotional at which face-to-face interaction is most
contagion in subjective experience—emo common. Maternal dependency and infant
tional yoking—is common in applied and responsiveness were not found to have
more controlled settings; and (4) the ques occurred at significant levels, but maternal
tion of which mechanisms might account responsiveness and infant dependency
for emotional yoking is an open question, were, and at all ages. Although mothers
although the possibility of expressive mim were responsive to their infants, mothers
icry is suggested. smiled a lot independently of their infants’
smiling behavior; hence, infant behavior
is sufficient but not necessary to elicit
MUTUAL COORDINATION IN smiles from the mother. Infants responded
INFANT-MOTHER INTERACTION to their mothers’ smiles with smiles of
their own. The proportion of mother
Studies reviewed by Cappella (1997, smiles followed by infant smiles did not
1998) and Field (1987) show that infants exceed the levels expected by chance.
weeks and, in a few cases, even hours old Mother smiling behavior seems necessary,
adapt to their adult partners in vocal, gaze, but not sufficient, to elicit smiles from the
facial, and movement behaviors. Such evi infants.
dence underscores the centrality of mutual In an important, and related, line of
coordination in human social interaction study, empathic responsiveness and affective
19-Manusov.qxd 6/30/2006 5:12 PM Page 366
366–––◆–––Functions
reactivity to infant stimuli were studied in interaction video exhibited more behavioral
mothers at high and low risk for physical co-occurrences between infant and mother
child abuse (Milner, Halsey, & Fultz, involving vocal, touch, and gaze behaviors.
1995). Compared with baseline, high-risk A meta-analysis tracked the value of inter
mothers reported no change in empathy ventions in improving parental sensitivity
across infant conditions such as baseline, in interacting with their infants across
smiling, quiet, and crying. Low-risk moth available studies (Bakermans-Kranenburg,
ers did report an increase in empathy. van Ijzendoorn, & Juffer, 2003). Inter
Following the presentation of a crying ventions that altered parental sensitivity
infant, high-risk mothers reported more also had favorable consequences on subse
distress and hostility. These data agree with quent attachment, adding evidence to the
other studies showing that child abusers are claim that responsive interactions are impor
less empathic and more hostile in response tant for the development of attachment.
to a crying child. The absence of responsive interaction in
Although these conclusions are com the early weeks and months of an infant’s
pelling, one of the problems with many of life can be deleterious to the infant’s devel
the studies of infant-adult interaction is that opment in the absence of other buffer
the samples are usually small and unrepre ing social and psychological conditions.
sentative of the population at large. The Despite this understanding, many parents
National Institute of Child Health and may not be responsive to their child. There
Human Development (NICHD), however, are many reasons for lower responsive
has undertaken a large-scale study of more ness by parents, including depression. Field
than 1,100 infants and their mothers at 10 (1998) argues that early maternal depres
different sites around the country. One of sion is associated with two different inter
the first reports from this study (NICHD actional styles: withdrawn or intrusive.
Early Child Care Research Team, 1997) Both can affect the infant’s physiology and
focused on the relationship between child biochemistry through inadequate stimula
care and quality of later attachment between tion or its opposite, overstimulation. Field
the mother and her infant. Although the argues that others in the infant’s life who
quality, amount, and type of child care out are not themselves depressed may buffer the
side the home (6–15 months) were unre negative consequences of interaction with a
lated to attachment quality at 15 months, depressed mother.
maternal responsiveness to the infant did In addition to the more one-sided
interact with child care arrangements. responsiveness, synchrony in the expressed
When mothers were low in responsiveness behaviors between infants and their care
and their children also experienced poor givers has become a staple of researchers
quality child care outside the home, the and is slowly achieving the stature of a
infants tended to be less securely attached diagnostic tool in assessing developmental
at 15 months. progression and responsive parenting.
Parents can also be trained to be more Some recent studies have used synchrony
responsive to their infants. Wendland- between infant and mother (and sometimes
Carro, Piccinini, and Millar (1999) exposed father) to assess the risk status of triplets
parents of newborns (2–3 days old) to a in comparison to twins and singletons
video on the importance of parental inter (Feldman & Eidelman, 2004), the develop
action or a video on basic caregiving. ment of symbolic competence at 2 years
One month later, those exposed to the from synchrony at 3 and 9 months
19-Manusov.qxd 6/30/2006 5:12 PM Page 367
(Feldman & Greenbaum, 1997), the devel evaluations have been associated, for
opment of self-control at 2 years (Feldman, example, with coordination in speech latency
Greenbaum, & Yirmiya, 1999), and the (Welkowitz & Kuc, 1973), speech rate and
development of emotional differentiation duration (Street, 1982), and pronunciation
with mothers and with fathers (Feldman, (Giles & Smith, 1979). Generalized respon
2003). Not only does infant-adult coordi siveness is associated with attraction (Davis
nation occur early in the infant’s develop & Martin, 1978) and the provision of plea
ment, but the presence of these behaviors is surable stimulation (Davis & Perkowitz,
also predictive of subsequent attachment 1979). Movement synchrony and mimicry
and, very possibly, other cognitive and are associated with rapport (Bernieri, 1988;
behavioral advances. The fact that parents Hess et al., 1999; see Tickle-Degnen, this
can be primed to be responsive and sensi volume). Our own research has produced
tive to their infants’ behaviors is evidence modest positive correlations between mea
that intervention can benefit infants and sures of dynamic coordination and inter
their parents by enhancing the bonding personal attraction (see Cappella, 1996,
process as well as other desirable social and 1998; Cappella & Flagg, 1992; Cappella,
cognitive competencies. Palmer, & Donzella, 1991).
In the context of marital relationships,
Summary. Despite the occasional contrary Gottman’s (1979) widely cited findings are
study, research on coordination between still the exemplar. Although all his couples
infants and adults (1) has been consistent tended to show reciprocity in hostile affect
with earlier research; (2) has been consis in discussions about common problems in
tent across studies in meta-analytical sum their marriages, the less well-adjusted cou
maries; (3) has been manifest in more ples showed greater hostile affect than did
representative populations; and (4) has the better-adjusted couples. Pike and Sillars
begun to be treated as an indicator of nor (1985) also found greater reciprocity in
mal behavioral, emotional, and cognitive negative vocal affect for dissatisfied as
development. These conclusions imply that opposed to satisfied married couples. Using
coordination in adult social interaction has face-directed gaze rather than negative
its roots in infant-adult social interaction. affect, Noller’s (1984) satisfied couples
exhibited greater correlation between part
ners than did the dissatisfied couples (see
MUTUAL COORDINATION also Manusov, 1995). Overall, partners in
AND RELATIONAL AND satisfying, established relationships appear
INDIVIDUAL CONDITIONS to differ in the type of mutual influence that
their interaction shows relative to those in
Coordination between partners in less satisfied relationships.
expressed emotion (and other behaviors) Some of the recent studies in emotional
and in subjectively experienced emotion does contagion and relational outcomes have
not imply necessarily desirable or undesir focused more on the similarity of reported
able outcomes for the persons or their rela emotion than on expressed emotion (e.g.,
tionship. Earlier research on this issue Totterdell, 2000; Totterdell, Wall, Holman,
(e.g., Cappella, 1991, 1998) has shown Diamond, & Epitropaki, 2004). Anderson,
mostly, but not exclusively, beneficial out Keltner, and John (2003) completed a con
comes for infants and mostly favorable out trolled version of the contagion hypothesis.
comes in adult relationships. Positive social Their study evaluated the development of
19-Manusov.qxd 6/30/2006 5:12 PM Page 368
368–––◆–––Functions
emotional similarity over time. They made from the available empirical results
defined emotional similarity as the coordi without inducing some skepticism. The
nation of thoughts and behaviors leading to “causality problem” is both theoretical and
greater understanding and cohesion among empirical. On empirical grounds, even if
partners. Studies 1 and 2 in Anderson et al. there is both covariation and temporal
were longitudinal, investigating dating part order, as is the case in Gottman’s research
ners and college roommates at two points and in many of the infant-adult studies,
in time separated by 6 months. The third these criteria do not eliminate the possibil
was experimental. Study 1 showed an ity of spurious correlation or mutual attrac
increase in reported emotional similarity— tion prior to the interaction affecting the
both positive and negative—over time, initial levels of coordination. Whereas the
whereas personality similarity remained rel data for infants are more convincing, those
atively stable in the same time period. Over for adults are less convincing. On theoreti
the same time period, positive emotional cal grounds, the mechanisms through
convergence was associated with relational which rapport might grow from behavioral
satisfaction, but negative emotional similar coordination between partners are not well
ity was not. Relational breakup from Time described by prevailing explanations. In the
1 to Time 2 was also predicted by emo remainder of this section, we will focus on
tional similarity at Time 1: Couples with empirical considerations.
greater emotional similarity at Time 1 Although this issue was not their pri
stayed together, whereas those without mary motivation, Chartrand and Bargh
emotional similarity tended to part. (1999) have taken on the causal question
Although emotional contagion is com directly in a series of studies. In their first
mon among coworkers, dating couples, study, the authors established a mimicry
college roommates, and spouses, and its effect between confederate and respondent
suppression can disrupt communication using two uncommon behaviors: face rub
and relationship formation and elevate and foot shake. When confederates used
blood pressure (Butler et al., 2003), it is less one of the unusual behaviors, then so did
clear whether emotional contagion and syn the respondent. In effect, there was imita
chrony in behavior are as consequential to tion of the behavior over and above base
the success of a relationship. Gottman and line. These imitations were independent of
Levenson (1999) compared four classes of whether the confederate was smiling or not.
predictors in accounting for deterioration In their second study, confederates imitated
in marital interaction over a 4-year period. the “posture, movement, and mannerisms”
Two classes of predictors were physiologi (p. 902) of the respondent while maintain
cal, one cognitive, and one interactional. ing a neutral facial expression and avoiding
The ratio of positive to negative expressions gaze. Results indicated that when the
was the best predictor of deterioration of confederates imitated the behavior of
interaction, which, in turn, was an excellent the participants, the participants rated the
predictor of future marital dissolution and interactions as smoother and reported lik
dissatisfaction. ing the confederate more. Careful checking
Despite the association between some of the confederates’ other nonverbal behav
type of coordination and some form of rap iors indicated no differences in rated eye
port or attachment for adult and infant- contact, smiling, friendliness, or liking of
adult dyads, a strong claim of a causal the participant by the confederates across
relationship between the two cannot be the mimicry/no mimicry conditions. The
19-Manusov.qxd 6/30/2006 5:12 PM Page 369
authors employed a careful debriefing mimicry. Their results suggest that being
procedure to determine whether the partic mimicked enhanced a prosocial attitude in
ipants were aware of imitation by the con general. Several personality and situation
federate. They were not. factors enhance or retard the likelihood
The results of this study establish a clear of mimicry, including self-monitoring
causal relationship between imitation (Cheng & Chartrand, 2003), context depen
and positive relational consequences. What dence (Van Baaren, Horgan, Chartrand, &
differentiates this study from several others Dijkmans, 2004), self-construal orientation
that seem to test the same hypothesis is the (Van Baaren, Maddux, Chartrand, de
careful control over confederate behavior Bouter, & van Knippenberg, 2003), affilia
and attention to initial levels of liking or tion goals (Lakin & Chartrand, 2003), and
cues to liking and attraction (such as smil attachment patterns (Sonnby-Borgstrom &
ing or other nonverbal cues). Other studies Jonsson, 2004).
have used confederates to enact behaviors
that are themselves clear indicators of posi Summary. The research on the association
tive feeling, such as eye gaze and smiling. between coordination and relational out
These behaviors can create positive regard comes finds that at both the micro- and
right away. Instead of studying the impact macrolevels, coordination affects relational
of coordination on subsequent positive feel outcomes. Studies at the microlevel show
ing, a design allowing initially positive non clearly that mimicry of unobtrusive behav
verbal behavior confounds initial positive iors is causally linked to rapport, and those
regard with coordination. In the Chartrand at the macrolevel show that emotional yok
and Bargh (1999) study, however, the imi ing between partners is a necessary condi
tated behaviors are hardly noticed by tion for bonding.
participants, and initial behaviors by the
confederate are not inherently positive. Yet
they create positive social perceptions auto ♦ Implications for Theory
matically. The authors argue strongly for
an automatic perception-behavior linkage
based on these and other data. The review of previous and newer empirical
This recent work in social psychology on findings produces four broad conclusions
imitation of behaviors has produced a surge that should guide our tour into the theoret
of other studies operating under the label ical realms. The first of these is automatic
mimicry. These studies have reinvigorated ity. Many behaviors produce coordination
the study of contagion and coordination between partners automatically, operating
processes but have done so, at least initially, well below conscious awareness (see Lakin,
in ignorance of a long history of research on this volume). The consistent observation of
similar, if not exactly the same, processes. automatic coordination of behaviors sug
New findings supporting and extending gests strongly that theory must be based on
Chartrand and Bargh’s initial work have mechanisms that allow for automatic, non-
cascaded into the literature. Van Baaren, conscious behavioral and emotional coordi
Holland, Kawakami, and van Knippenberg nation. Second, the evidence establishing a
(2004), for example, found in three sepa causal linkage between behavioral coordi
rate studies that people who were mimicked nation and some form of positive relational
by others were more helpful and generous outcome, particularly rapport, is difficult
toward third parties not involved in the to treat with skepticism any longer. The
19-Manusov.qxd 6/30/2006 5:12 PM Page 370
370–––◆–––Functions
absence of a good explanation for this limitations will not allow a comprehensive
relationship leaves the empirical causal review, but we will discuss extant theories
findings without a strong explanatory in terms of their ability to account for
basis, however, and therefore undermines the recent findings reported above (see
their believability. Patterson, this volume). Specifically, what
Third, research in various domains has can theories tell us about automatic
begun to produce findings showing that responding, the linking between synchrony
persons in work groups, in relationships, and outcome, emotional yoking, and how
in social groups, living together, and so coordination came to be so central a
on develop (and report) similar emotional process in human social interaction?
responses over the course of time. This sug
gests an emotional yoking in social groups.
The impact for successful relationships and EVALUATING THEORIES
performance of emotional similarity and OF COORDINATION
dissimilarity is only beginning to be under
stood. Theories must begin to explain the A number of competing accounts have
mechanisms through which yoked emotion been put forward to meet the basic require
develops (other than through spurious ments of explaining coordination. These
external events that must be controlled), include drive explanations (Argyle & Dean,
especially given the failure of several studies 1965; Firestone, 1977), arousal-mediated
to show that expressed emotion mediates explanations (Andersen, 1985; Burgoon,
the relationship between partners’ reports of 1978; Burgoon & Jones, 1976; Burgoon
yoked emotion. Theory must account for & Hale, 1988; Cappella & Greene, 1982;
the link between synchrony in expressed Patterson, 1976, 1982), cognitive explana
emotion and emotional contagion. Finally, tions (Giles & Powesland, 1975; Giles,
the strong and consistent findings of syn Mulac, Bradac, & Johnson, 1987; Street
chrony in infant-adult interactions and the & Giles, 1982), and various combinations
possibility that synchrony (or its absence) of these (Andersen, Guerrero, Buller, &
might be a diagnostic tool for normal devel Jorgensen, 1998; Burgoon et al., 1995).
opmental progression of infants argue that With the upsurge of research emphasizing
coordination is a deep-seated and abiding the automatic nature of certain aspects of
process in human social and behavioral life. coordination in the adult (Chartrand &
Explaining how and why these processes Bargh, 1999) and in the infant-adult arenas
came to be is an important goal for theory, (De Wolff & van Ijzendoorn, 1997), how
particularly evolutionary theories, which ever, and the clear causal evidence for a
aim to explain how processes came to be in coordination-rapport link, theories must be
the first place (Cappella, 1991; Buck & capable of accounting for these develop
Renfrow Powers, this volume). ments, not as an afterthought but as a
central feature.
Despite their elegance, careful attention
♦ Theories About Coordination to the research literature, and attempts to
be comprehensive, none of these theories
has risen to the top of the empirical heap in
In this section, we take up theoretical contrastive tests. In three such tests, the
approaches to explaining the existing data findings are mostly mixed, with one or the
on coordination in social interaction. Space other theory taking precedence in some
19-Manusov.qxd 6/30/2006 5:12 PM Page 371
results but no one theory clearly accounting explanation. In particular, two components
for all findings. O’Connor and Gifford of their findings need explaining: mimicry
(1988) tested their social cognitive approach and rapport. What explains people’s
against arousal labeling and discrepancy mimicry of each other’s (inconsequential)
arousal theories, reporting that the social behaviors? Furthermore, what explains
cognitive model fared best in accounting for why mimicry should be associated with
behavioral responses but the self-report positive social regard for the partner? Other
results were less clearly supported. Other issues arise as we interrogate this process.
contrastive tests have produced a mixture Do people always imitate? The answer of
of findings favoring no one explanation course is “no,” but what are the conditions
indicative of the complexity of realistic promoting mimicry and its absence? Can
social interactions or the difficulty of pro mimicry create hostility rather than rap
ducing true critical tests (Andersen et al., port?
1998; Le Poire & Burgoon, 1996). Bargh and his colleagues have offered
One possible explanation for the inabil some answers (Bargh, 2003; Bargh &
ity of one of the several extant theories to Chartrand, 1999; Bargh & Ferguson, 2000;
account successfully for the results is not Ferguson & Bargh, 2004), even carrying
only the complexity of the testing environ their views into the evolutionary domain
ments and the requirements of careful, con (Lakin, Jefferis, Cheng, & Chartrand,
trolled manipulation by confederates but 2003). Their essential claim is that a much
also the breadth of the theories themselves. larger proportion of human activity is dri
Each of the theories makes a concerted ven by automatic processes than people and
effort to encompass the full range of behav psychology have been willing to admit.
iors, explain the conditions promoting Although there has been a very rich
compensation and reciprocity, and take tradition of environmental primes stimulat
into account mitigating and aggravating con ing cognitions and evaluations auto
ditions. For example, discrepancy-arousal matically (Bargh & Ferguson, 2000), it is
theory (Cappella & Greene, 1982) particu only recently that automatic primes for
larly tries to offer an account of infant- behavior have been investigated and found
adult as well as adult-adult patterns of operational.
coordination. This very strength, however, Bargh and his colleagues posit a percep
might produce generalities in the theories tion-behavior linkage that shows itself
that make them less able to predict particu in media effects on behavior (Berkowitz,
lar outcomes in particular social contexts. 1984, 1997), in behavior consistent with
A different strategy is to craft theories of the activation of stereotypes (Dijksterhuis
much narrower scope that seek to provide & van Knippenberg, 1998), and, of course,
very specific predictions of specific empiri in social interaction (Chartrand & Bargh,
cal phenomena. We turn now to such a 1999). The mechanisms through which
case, which provides explanations for perception leads to behavior and the
results about behavioral mimicry. limiting conditions are as yet not com
pletely explored or understood. “Mirror
neurons” are a viable candidate for per
EXPLAINING MIMICRY ception-behavior activation and inter
personal facial feedback (IFF) a potential
The intriguing findings reported by mechanism for establishing rapport
Chartrand and Bargh (1999) beg for an through imitation.
19-Manusov.qxd 6/30/2006 5:12 PM Page 372
372–––◆–––Functions
explained. The IFFH assumes that behav action of another. Additional research soon
ioral adaptation is temporally prior to identified clusters of mirror neurons in a
one’s own affective response. By invoking number of different locations in the brain
the IFFH, physiological and, as we will see (Rizzolatti & Craighero, 2004). Support
below, neurological pathways are impli for the theory that these neurons were pro
cated as the causal linkages from behavioral viding mental representations of the action
activation in the muscles of the face and increased when it was discovered that neu
voice to the subjective affective response rons representing the final part of a motor
(Zajonc, Murphy, & Englehart, 1989). In sequence continued to fire even when the
effect, a clear, if controversial, causal mech final portion of the sequence was hidden
anism is posited to account for the associa from the monkey’s view (Umilta et al.,
tion between behavioral coordination and 2001).
interpersonal affect. Mirror neurons have been posited as the
foundation on which imitation (Buccino
et al., 2004), empathy (Carr, Iacoboni,
NEUROLOGICAL BASES FOR Dubeau, Mazziotta, & Lenzi, 2003), and
INTERPERSONAL FACIAL even our capacity to understand another’s
FEEDBACK, INTERPERSONAL state of mind rests (Gallese & Goldman,
VOCAL FEEDBACK, AND MIMICRY 1998; Schulkin, 2000; Williams, Whiten,
Suddendorf, & Perrett, 2001). Evidence
One line of research that supports from a variety of sources including neuro
Bargh’s (2003) explanation of mimicry as logical impairment, direct neuron record
well as the IFFH is found in recent work in ings, evolutionary biology, and neuroimaging
neuroscience, specifically the isolation of have been marshaled in support of these
a mirror neuron reflex. In the mid-1990s, arguments, at least in part because mirror
neuroscientist Vitorio Gallese was observing neurons enable a plausible story to be told
neural activity in the cortex of a macaque about how we have come to the capacity to
monkey during object manipulation. After communicate emotionally.
returning to his laboratory with an ice- There is also mounting evidence suggest
cream cone, Gallese noticed that each time ing that mirror neurons allow humans to
he licked the ice-cream cone, the neurons use the same neural mechanisms both to
in the monkey’s premotor cortex fired. This express emotions and perceive the expres
was intriguing, given that the monkey was sion of emotion in others. Furthermore,
not making any motor movements. these neural substrates appear connected to
Further study revealed a set of 92 neu the emotion (changes in body and brain
rons in the premotor cortex that were active states triggered by the content of percep
both when the monkey performed an tions) as well as feelings (those changes
action and when the experimenter per in brain state that reach sufficient intensity
formed the same action (Gallese, Fadiga, to be perceived by conscious awareness)
Fogassi, & Rizzolatti, 1996; Rizzolatti, (Damasio, 1999, 2003). The insula, for
Fadiga, Gallese, & Fogassi, 1996). Although instance, appears to become activated not
canonical neurons in the premotor cortex only when we experience a disgusting smell
would activate only when the monkey per but also when we perceive someone else’s
formed a motor hand action, these neurons experience of disgust or imitate a disgusted
were described as mirror neurons because expression. The region appears to facilitate
of their apparent ability to represent the our recognition of our physical, emotional,
19-Manusov.qxd 6/30/2006 5:12 PM Page 374
374–––◆–––Functions
and mental states and the physical, emo others in the social environment. Importantly,
tional, and mental states of others. And it the evidence is strong with causal direction
may well be that the insula is one part of clear and possible confounders minimized.
the network involved in the experience Third, coordination in the form of syn
of disgust after adopting a disgusted facial chrony between infants and their parents
expression. has become sufficiently well established as
Mirror neurons offer a plausible, causal to be a sign of normal (and, in its absence,
mechanism for explaining the “perception abnormal) social development. These find
behavior” linkage that is central to the new ings are evidence of the maturation of a
research on automatic mimicry and older field of study.
research on automatic responding in infants At the theoretical level, the active formu
and adults as well. Coupled with the IFF lation of general explanations of coordina
hypothesis, the two mechanisms provide a tion that has marked the past 40 years of its
plausible, if speculative, account of auto study has not produced conceptual or theo
matic behavioral responding, imitation (both retical consensus on which of several simi
gestural and facial), the occurrence of rap lar competitors is the most effective. We
port, and emotional yoking. Empirical test speculate that the move toward broad the
ing for these regulative processes awaits. ories that encompass the range of behaviors
and circumstances characteristic of coordi
nation may be part of the problem along
with the difficulty of providing unequivocal
♦ Conclusions
contrastive tests. One alterative is explana
tions that are more limited in scope. We
In this chapter, we have tried to bring explored “automatic perception-behavior”
previous reviews of the literature on non account of mimicry, arguing that it offers
verbal coordination up to date by high clear predictions and links well with other
lighting key studies and trends in the theories of automatic responding. Whereas
literature. In addition, we have tried to the perception-behavior link works well, it
offer some speculations for theory develop is also clear that the behavior-rapport link is
ment that would help to resolve some puz less obvious a consequence of mimicry. As a
zles and paradoxes in the existing literature. suggestive resolution, we presented IFFH to
In closing, however, three empirical conclu account for the behavior-rapport link and
sions should be brought out, lest they be some evidence from the neurosciences on
lost in the details that the review entails. mirror neurons to strengthen the association
First, yoked emotional response among between perception and behaviors enacted
social actors characterizes work groups, through imitation (or coordination).
dating and longer term relationships, room
mates, and even professional sports teams.
The mechanism of the production of this ♦ References
yoking of emotion is less well established
and its consequences—for good or ill—
Adelman, P. K., & Zajonc, R. B. (1989).
have not yet been fully explored. Second, Facial efference and the experience of
coordination shows itself through mimicry emotion. Annual Review of Psychology,
of unnoticed behaviors. Mimicry, in turn, 40, 249–280.
produces a sense of rapport with the person Andersen, P. A. (1985). Nonverbal immedi
mimicking that may even generalize to acy in interpersonal communication. In
19-Manusov.qxd 6/30/2006 5:12 PM Page 375
A. W. Siegman & S. Feldstein (Eds.), Bernieri, F., & Rosenthal, R. (1991). Inter
Multichannel integrations of nonverbal personal coordination: Behavior matching
behavior (pp. 1–36). Hillsdale, NJ: Erlbaum. and interactional synchrony. In R. S. Feldman
Andersen, P. A., Guerrero, L. K., Buller, D. B., & B. Rime (Eds.), The fundamentals of non
& Jorgensen, P. F. (1998). An empirical verbal behavior (pp. 401–432). New York:
comparison of three theories of nonverbal Cambridge University Press.
immediacy exchange. Human Communica Buccino, G., Vogt, S., Ritzl, A., Fink, G. R.,
tion Research, 24, 501–535. Zilles, K., Freund, H. J., et al. (2004).
Anderson, C., Keltner, D., & John, O. P. (2003). Neural circuits underlying imitation learn
Emotional convergence between people ing of hand actions: An event-related fMRI
over time. Journal of Personality and Social study. Neuron, 42, 323–334.
Psychology, 84, 1054–1068. Burgoon, J. K. (1978). A communication model
Argyle, M., & Dean, J. (1965). Eye contact, of interpersonal space violations: Explica
distance, and affiliation. Sociometry, 28, tion and initial test. Human Communication
289–304. Research, 4, 129–142.
Bakermans-Kranenburg, M. J., van Ijzendoorn, Burgoon, J. K. (1993). Interpersonal expecta
M. H., & Juffer, F. (2003). Less is more: tions, expectancy violations, and emotional
Meta-analyses of sensitivity and attachment communication. Journal of Language and
interventions in early childhood. Psycholo Social Psychology, 12, 30–48.
gical Bulletin, 129, 195–215. Burgoon J. K. (1994). Nonverbal signals. In
Bakker, A. B., & Schaufeli, W. B. (2001). Burnout M. L. Knapp & G. R. Miller (Eds.), Hand
contagion processes among teachers. Journal of book of interpersonal communication (2nd
Applied Social Psychology, 30, 2289–2308. ed., pp. 229–271). Thousand Oaks, CA: Sage.
Bargh, J. A. (2003). Why we thought we could Burgoon, J. K., & Hale, J. L. (1988). Nonverbal
prime social behavior. Psychological Inquiry, expectancies violations: Model elaboration
14, 216–218. and application to immediacy behaviors.
Bargh, J. A., & Chartrand, T. L. (1999). The Communication Monographs, 55, 58–79.
unbearable automaticity of being. Burgoon, J. K., & Jones, S. B. (1976). Toward a
American Psychologist, 54, 462–479. theory of personal space expectations and
Bargh, J. A., & Ferguson, M. J. (2000). Beyond their violations. Human Communication
behaviorism: On the automaticity of higher Research, 2, 131–146.
mental processes. Psychological Bulletin, Burgoon, J. K., Stern, L. A., & Dillman, L.
126, 925–945. (1995). Interpersonal adaptation: Dyadic
Berkowitz, L. (1984). Some effects of thoughts interaction patterns. New York: Cambridge
on anti- and prosocial influences of media University Press.
events: A cognitive neo-association analy Butler, E. A., Egloff, B., Wilhelm, F. H., Smith,
sis. Psychological Bulletin, 95, 410–427. N. C., Erickson, E. A., & Gross, J. J.
Berkowitz, L. (1997). Some thoughts extending (2003). The social consequences of expres
Bargh’s argument. In R. S. Wyer (Ed.), sive suppression. Emotion, 3, 48–67.
Advances in social cognition (Vol. 10, Cappella, J. N. (1981). Mutual influence in
pp. 83–94). Mahwah, NJ: Erlbaum. expressive behavior: Adult and infant-adult
Berger, C. R., & Roloff, M. E. (1980). Social cog dyadic interaction. Psychological Bulletin,
nition, self awareness, and interpersonal 89, 101–132.
communication. In B. Dervin & M. J. Voight Cappella, J. N. (1985). The management of con
(Eds.), Progress in communication sciences versations. In M. L. Knapp & G. R. Miller
(Vol. 2, pp. 1–50). Norwood, NJ: Ablex. (Eds.), The handbook of interpersonal commu
Bernieri, F. J. (1988). Coordinated movement nication (pp. 393–438). Beverly Hills, CA: Sage.
and rapport in teacher-student interac Cappella, J. N. (1991). The biological origins of
tions. Journal of Nonverbal Behavior, 12, automated patterns of human interaction.
120–138. Communication Theory, 1, 4–35.
19-Manusov.qxd 6/30/2006 5:12 PM Page 376
376–––◆–––Functions
Cappella, J. N. (1993). The facial feedback expressive groups. Paper presented at the
hypothesis in human interaction: Review International Communication Association
and speculations. Journal of Language and conference, Chicago.
Social Psychology, 12, 13–29. Carr, L., Iacoboni, M., Dubeau, M. C.,
Cappella, J. N. (1994). The management of Mazziotta, J. C., & Lenzi, G. L. (2003).
conversational interaction in adults and Neural mechanisms of empathy in humans:
infants. In M. L. Knapp & G. R. Miller A relay from neural systems for imitation to
(Eds.), Handbook of interpersonal commu limbic areas. Proceedings of the National
nication (2nd ed., pp. 380–419). Thousand Academy of Sciences, 100, 5497–5502.
Oaks, CA: Sage. Chartrand, T. L., & Bargh, J. A. (1999). The
Cappella, J. N. (1996). Dynamic coordination chameleon effect: The perception-behavior
of vocal and kinesic behavior. In J. Watt & link and social interaction. Journal of
C. A. VanLear (Eds.), Dynamic patterns Personality and Social Psychology, 76,
in communication processes (pp. 353–386). 893–910.
Thousand Oaks, CA: Sage. Cheng, C. M., & Chartrand, T. L. (2003). Self-
Cappella, J. N. (1997). The development of monitoring without awareness: Using
theory about automated patterns of face mimicry as a non-conscious affiliation strat
to-face human interaction. In G. Philipsen egy. Journal of Personality and Social
& T. Albrecht (Eds.), Theories in human Psychology, 85, 1170–1179.
communication (pp. 57–84). Albany, NY: Damasio, A. R. (1999). The feeling of what hap
SUNY Press. pens: Body and emotion in the making of
Cappella, J. N. (1998). The dynamics of nonver consciousness. New York: Harcourt Brace.
bal coordination and attachment: Problems Damasio, A. R. (2003). Looking for Spinoza:
of causal direction and causal mechanism. Joy, sorrow, and the feeling brain. New
In M. T. Palmer (Ed.), Progress in commu York: Harcourt.
nication sciences: Vol. 14. Mutual influence Davis, D., & Martin, H. J. (1978). When plea
in interpersonal communication: Cognitive, sure begets pleasure: Recipient responsive
behavioral, and affective components ness as a determinant of physical pleasuring
(pp. 19–37). Greenwich, CT: Ablex. between heterosexual dating couples and
Cappella, J. N., & Flagg, M. E. (1992, July). strangers. Journal of Personality and Social
Interactional adaptation, expressiveness Psychology, 36, 767–777.
and attraction: Kinesic and vocal respon Davis, D., & Perkowitz, W. T. (1979).
siveness patterns in initial liking. Paper pre Consequences of responsiveness in dyadic
sented at the 6th International Conference interaction: Effects of probability of
on Personal Relationships, University of response and proportion of content-related
Maine, Orono. responses on interpersonal attraction.
Cappella, J. N., & Greene, J. O. (1982). A Journal of Personality and Social Psycho
discrepancy-arousal explanation of mutual logy, 37, 534–550.
influence in expressive behaviors for De Wolff, M. S., & van Ijzendoorn, M. H.
adult-adult and infant-adult interactions. (1997). Sensitivity and attachment: A meta
Communication Monographs, 49, 89–114. analysis on parental antecedents of infant
Cappella, J. N., & Palmer, M. T. (1990). Attitude attachment. Child Development, 68,
similarity, relational history, and attraction: 571–591.
The mediating effects of kinesic and vocal Dijksterhuis, A., & van Knippenberg, A. (1998).
behaviors. Communication Monographs, The relation between perception and
57, 161–183. behavior, or how to win a game of Trivial
Cappella, J. N., Palmer, M. T., & Donzella, B. Pursuit. Journal of Personality and Social
(1991, May). Individual consistency in tem Psychology, 74, 865–877.
poral adaptations in nonverbal behavior Doherty, R. W. (1997). The Emotional
in dyadic conversations: High and low Contagion Scale: A measure of individual
19-Manusov.qxd 6/30/2006 5:12 PM Page 377
differences. Journal of Nonverbal Behavior, Giles, H., Mulac, A., Bradac, J. J., & Johnson, P.
21, 131–153. (1987). Speech accommodation theory:
Doherty, R. W., Orimoto, L., Singelis, T. M., The first decade and beyond. In M. L.
Hatfield, E., & Hebb, J. (1995). Emotional McLaughlin (Ed.), Communication year
contagion: Gender and occupational differ book 10 (pp. 13–48). Newbury Park, CA:
ences. Psychology of Women Quarterly, Sage.
19, 355–371. Giles, H., & Smith, P. M. (1979).
Feldman, R. (2003). Infant-mother and infant- Accommodation theory: Optimal levels
father synchrony: The coregulation of posi of convergence. In H. Giles & R. St. Clair
tive arousal. Infant Mental Health Journal, (Eds.), Language and social psychology
24, 1–23. (pp. 45–65). Baltimore, MD: University
Feldman, R., & Eidelman, A. I. (2004). Parent- Park Press.
infant synchrony and the social-emotional Goodman, C. R., & Shippy, R. A. (2002). Is it
development of triplets. Developmental contagious? Affect similarity among
Psychology, 40, 1133–1147. spouses. Aging & Mental Health, 6,
Feldman, R., & Greenbaum, C. W. (1997). 266–274.
Affect regulation and synchrony in mother- Gottman, J. M. (1979). Marital interaction.
infant play as precursors to the develop New York: Academic Press.
ment of symbolic competence. Infant Gottman, J. M., & Levenson, R. W. (1999).
Mental Health Journal, 18, 4–23. What predicts change in marital interaction
Feldman, R., Greenbaum, C. W., & Yirmiya, N. over time? A study of alternative models.
(1999). Mother-infant affect synchrony Family Process, 38, 143–158.
as an antecedent of the emergence of Gump, B. B., & Kulik, J. A. (1997). Stress, affil
self-control. Developmental Psychology, iation, and emotional contagion. Journal
35, 223–231. of Personality and Social Psychology, 72,
Ferguson, M. J., & Bargh, J.A. (2004). How 305–319.
social perception can automatically influ Hatfield, E., Cacioppo, J., & Rapson, R. L.
ence behavior. Trends in Cognitive Sciences, (1994). Emotional contagion. Paris:
8, 33–39. Cambridge University Press.
Field, T. (1987). Affective and interactive distur Hess, U., & Blairy, S. (2001). Facial mimicry
bances in infants. In J. D. Osofsky and emotional contagion to dynamic emo
(Ed.), Handbook of infant development tional facial expressions and their influence
(2nd ed., pp. 972–1007). New York: Wiley. on decoding accuracy. International Journal
Field, T. (1998). Maternal depression effects on of Psychophysiology, 40, 129–141.
infants and early interventions. Preventive Hess, U., Philippot, P., & Blairy, S. (1999).
Medicine: An International Journal Devoted Mimicry: Facts and fiction. In P. Philippot,
to Practice & Theory, 27, 200–203. R. S. Feldman, & E. J. Coats (Eds.), The
Firestone, I. (1977). Reconciling verbal and non social context of nonverbal behavior (pp.
verbal models of dyadic communication. 213–241). New York: Cambridge University
Environmental Psychology and Nonverbal Press.
Behavior, 2, 30–44. Hietanen, J. K., Surakka, V., & Linnankoski, I.
Gallese, V., Fadiga, L., Fogassi, L., & Rizzolatti, (1998). Facial electromyographic responses
G. (1996). Action recognition in the premo to vocal affect expressions. Psychophysio
tor cortex. Brain, 119, 593–609. logy, 35, 530–536.
Gallese, V., & Goldman, A. (1998). Mirror neu Hsee, C. K., Hatfield, E., Carlson, J. G., &
rons and the simulation theory of mind read Chemtob, C. (1992). Assessments of the
ing. Trends in Cognitive Science, 2, 493–501. emotional state of others—Conscious judg
Giles, H., & Powesland, P. F. (1975). Speech style ments versus emotional contagion. Journal
and social evaluation. London: Academic of Social and Clinical Psychology, 11,
Press. 119–128.
19-Manusov.qxd 6/30/2006 5:12 PM Page 378
378–––◆–––Functions
Street, R. L. (1982). Evaluation of noncontent Van Baaren, R. B., Holland, R. W., Kawakami,
speech accommodation. Language and K., & Van Knippenberg, A. (2004). Mimicry
Communication, 2, 13–31. and prosocial behavior. Psychological
Street, R. L., & Giles, H. (1982). Speech accom Science, 15, 71–74.
modation theory: A social cognitive approach Van Baaren, R. B., Horgan, T. G., Chartrand,
to language and speech behavior. In T. L., & Dijkmans, M. (2004). The forest,
M. Roloff & C. R. Berger (Eds.), Social the trees, and the chameleon: Context depen
cognition and communication (pp. 193– dence and mimicry. Journal of Personality &
226). Beverly Hills, CA: Sage. Social Psychology, 86, 453–459.
Symons, D., & Moran, G. (1994). Responsiveness Van Baaren, R. B., Maddux, W. W., Chartrand,
and dependency are different aspects of T. L., de Bouter, C., & van Knippenberg, A.
social contingencies: An example from (2003). It takes two to mimic: Behavioral
mother and infant smiles. Infant Behavior consequences of self-construals. Journal of
and Development, 17, 209–214. Personality and Social Psychology, 84,
Thompson, L. L., Nadler, J., & Kim, P. H. 1093–1102.
(1999). Some like it hot: The case for the Verbeke, W. (1997). Individual differences in
emotional negotiator. In L. L. Thompson emotional contagion of salespersons: Its
& J. M. Levine (Eds.), Shared cognition in effect on performance and burnout.
organizations: The management of knowl Psychology & Marketing, 14, 617–636.
edge (pp. 139–161). Mahwah, NJ: Erlbaum. Welkowitz, J., & Kuc, M. (1973). Inter
Totterdell, P. (2000). Catching moods and relationships among warmth, genuineness,
hitting runs: Mood linkage and subjective empathy, and temporal speech patterns in
performance in professional sport teams. interpersonal interaction. Journal of
Journal of Applied Psychology, 85, Consulting and Clinical Psychology, 41,
848–859. 472–473.
Totterdell, P., Kellett, S., Teuchmann, K., & Wendland-Carro, J., Piccinini, C. A., & Millar,
Briner, R. B. (1998). Evidence of mood W. S. (1999). The role of an early interven
linkage in work groups. Journal of Per tion on enhancing the quality of mother-
sonality and Social Psychology, 74, infant interaction. Child Development,
1504–1515. 70, 713–721.
Totterdell, P., Wall, T., Holman, D., Diamond, Wild, B., Erb, M., & Bartels, M. (2001). Are emo
H., & Epitropaki, O. (2004). Affect net tions contagious? Evoked emotions while
works: A structural analysis of the relation viewing emotionally expressive faces: Quality,
ship between work ties and job-related quantity, time course and gender differences.
affect. Journal of Applied Psychology, 89, Psychiatry Research, 102, 109–124.
854–867. Williams, J. H., Whiten, A., Suddendorf, T., &
Ullstadius, E. (1998). Neonatal imitation in a Perrett, D. I. (2001). Imitation, mirror neu
mother-infant setting. Early Development rons and autism. Neuroscience Biobehavior
and Parenting, 7, 1–8. Review, 25, 287–295.
Umilta, M. A., Kohler, E., Gallese, V., Fogassi, Zajonc, R. B., Murphy, S. T., & Englehart, M.
L., Fadiga, L., Keysers, C., et al. (2001). I (1989). Feeling and facial efference:
know what you are doing: A neurophysio Implications of the vascular theory of emo
logical study. Neuron, 31, 155–165. tion. Psychological Review, 96, 395–416.
19-Manusov.qxd 6/30/2006 5:12 PM Page 380
20-Manusov.qxd 6/30/2006 5:12 PM Page 381
20
NONVERBAL BEHAVIOR
AND ITS FUNCTIONS IN
THE ECOSYSTEM OF RAPPORT
� Linda Tickle-Degnen
Boston University
Author’s Note: This chapter was supported in part by Grant No. NS048059
from the National Institute of Neurological Disorders and Stroke of the
National Institutes of Health (NIH). Its contents are solely the responsibility
of the author and do not necessarily represent the official views of NIH.
◆ 381
20-Manusov.qxd 6/30/2006 5:12 PM Page 382
382–––◆–––Functions
Rapport creates a powerful medium for perceived qualities (Dijksterhuis & Bargh,
social influence (Freud, 1914/1924; LaFrance, 2001), or a strategic process (Brunswik,
1990) and for the accomplishment of tasks 1955) is beyond the scope of this chapter
that are challenging and require mutual com (but see Lakin, this volume), although
mitment to accomplish (Tickle-Degnen & arguably both automaticity and cognitive
Gavett, 2003). Individuals in rapport are mediation are involved (Patterson, 1999,
likely to cooperate with one another to 2003; Strack & Deutsch, 2004). Once in
accomplish tasks and objectives that could rapport, however, individuals form a phys
not be accomplished alone as effectively, ical and social ecosystem—an interde
as efficiently, or at all. Cooperation facili pendence of two entities’ perceptions and
tates many of the tasks of everyday living, actions—that operates in relation to the
including those most basic to species larger physical and social environment in
survival—eating, protection against harm, which it is embedded. People develop, main
or procreation—as well as those involved in tain, and indicate their rapport through a
the actualization of human potential. The stream of interlinked signals and responses
child in rapport with the parent is fed and that are shaped by their personal physical
nourished, workers in rapport achieve the and psychological properties, the parame
project deadline, and the student and ters of the task in which they are engaged,
teacher in rapport pass knowledge on to the and the physical and social environment of
next generation. their actions.
Importantly for this chapter, rapport is Research studies on rapport, with a
created, sustained, and marked by action. few exceptions (Bernieri, Gillis, Davis, &
More specifically, it involves the coupling Grahe, 1996; Puccinelli, Tickle-Degnen,
of internal experience and purposeful, often & Rosenthal, 2003, 2004), have not
nonverbal, action into a perception-action addressed nonverbal behavior as emergent
system that has adaptive value for the indi from context explicitly and as part of an
vidual, the dyad, and the human species ecosystem of rapport, that is, from an eco
(Dijksterhuis & Bargh, 2001; Fridlund, logical perspective. As a result, accumulat
1997; Patterson, 1999). The perception- ing evidence has little theoretical coherence
action system of rapport is not separate and appears to be complex and conflicting.
from its context but is rather emergent The purpose of this chapter is to explore
within it (Gibson, 1986; McArthur & the role of nonverbal behavior within the
Baron, 1983; Zebrowitz & Collins, 1997), ecosystem of rapport and to demonstrate
and the immediate contextual elements for that a systematic investigation of context
one individual are the features of the other resolves apparently conflicting results and
individual, including static nonverbal fea promotes the theoretical coherence needed
tures such as facial structure and signs of to move this area of research forward. To
gender as well as dynamic features from do so, the chapter describes the functions
verbal and nonverbal behavior. of nonverbal behavior and presents a three-
Perceivers respond rapidly to informa component model of rapport that is consis
tive physical features in the face, body, and tent with an ecological perspective.
voice of potential cooperative partners as The chapter also reviews recent research
they negotiate the natural social environ findings on nonverbal behavior and rapport
ment. Whether this response is due to an and describes how a model of optimal expe
attunement of attention that is learned rience developed by Csikszentmihalyi and
(Gibson, 1986), an automatic expression of his colleague (Csikszentmihalyi, 1990;
20-Manusov.qxd 6/30/2006 5:12 PM Page 383
Csikszentmihalyi & Csikszentmihalyi, 1988) dyad (Ambady, Bernieri, & Richeson, 2000;
can be applied flexibly to the systematic Patterson, 1999). This ecosystem as a whole
investigation of rapport and nonverbal serves the function of regulating dyadic
behavior in context. Specifically, this model connection and rapport to accomplish inter
places optimal experiences such as rapport personal tasks and meet personal, dyadic,
within the broader theoretical perspective and group needs.
of how dyads regulate themselves to achieve In contrast to a view that specific
smooth, harmonious, and enjoyable interac patterns of nonverbal behavior map one-
tion as constrained by the tasks that they are to-one with discrete emotions (Ekman &
doing. The model provides a framework for Keltner, 1997; Izard, 1994) or that nonver
showing how nonverbal behavior varies in bal behavior expresses an actor’s social
relation to the degree that task parameters intentions, not emotions (Fridlund, 1997),
support or undermine the achievement of the position taken here is that a nonver
rapport. The chapter concludes with implica bal action or event is best understood as
tions for future research. expressive of many psychological states,
traits, and action tendencies at once. The
stream of perception and behavior is such
♦ The Functions of that participants’ intentions, attitudes, and
Nonverbal Behavior in the feelings are not parsed out into entities that
are expressed in mutually exclusive behav
Ecosystem of Rapport
ioral markers. Rather internal states, per
ceptions, and behaviors are convergent
As part of the dyadic ecosystem, nonverbal aspects of a meaningful event that is under
behavior is functional (Patterson, 1994). stood within context (Heft, 2003; James,
Nonverbal behavior provides information 1890). That is, from an ecological perspec
about each interacting participant’s initial tive, nonverbal behavior is constrained, or
intentions, attitudes, and feelings and, as shaped, in relation to (a) the capacities and
such, serves an expressive function in the tendencies of the interacting individuals to
development and maintenance of rapport. experience, perceive, and express rapport-
Furthermore, each participant’s nonverbal relevant intentions, attitudes, and feelings;
behavior functions to create affordances (b) parameters of the interactive task; and
(Gibson, 1986) or opportunities for action (c) features of the physical and societal
by the other participant. To the perceiver, environment.
an actor’s behavior affords differing degrees More specifically, individuals vary in
of interpersonal contact, enjoyment, infor their sensitivity to experiencing rapport,
mation exchange, mutual task engagement, their expressiveness of internal states, and
and so on. The nonverbal reactions of the their ability to pick up information rele
perceiver, now actor, function to signal the vant to others’ states. For one individual, a
degree of “pick up” of the other partici vertical wrinkle between the brows may be
pant’s intentions, attitudes, and feelings as an expression of anger, an emotion rele
well as the acceptance or agreement with vant to rapport, and, in another, it may be
them. These functions operate within a con an indication of far-sightedness, a condi
tinuous signal-perception-action-signal loop tion less relevant to rapport. Also, interac
of interpersonal interaction, and they often tive tasks vary in difficulty, required
occur spontaneously, within seconds and in actions and object manipulations, bodily
parallel across both participants within a positioning to accomplish the task, and
20-Manusov.qxd 6/30/2006 5:12 PM Page 384
384–––◆–––Functions
2000) and in the explanation of how own needs and inner states. Expressivity
individuals form their judgments of rapport tends to elicit liking, rapport, and other
(Grahe & Bernieri, 2002). It is the approach positive social outcomes (Riggio & Riggio,
taken in the three-component model of 2002). In an experimental manipulation
rapport (Tickle-Degnen & Rosenthal, of the suppression of emotional behavior
1987, 1990). (Butler et al., 2003), female dyads watched
an upsetting film and, then, after receiving
the experimental manipulation, discussed it.
♦ Research on Nonverbal Women who had been instructed to sup
Behavior and Rapport press their emotional behavior were less
expressive, both positively and negatively,
less responsive, and more distracted than
The three-component model is partially were those women who had not received
represented in the recent research on non any specific instructions. The uninstructed
verbal behavior and rapport. The bulk of partners of the “suppressing” women report
recent and relevant research is on interper edly felt less rapport and willingness to form
sonal coordination, followed by nonverbal friendships with those women. Dyads in
expressivity, a construct that encompasses which both members were uninstructed,
but is not limited to positivity. Little and, therefore, more expressive, had better
research has systematically examined rapport outcomes.
the component of mutual attentiveness. The
review that follows describes first the Expressivity and Positivity. That expression
findings for expressivity, and second, of both positive and negative feelings is
the findings for coordination, because non conducive to rapport is counter to the
verbal expressivity appears to be the raw proposal that positivity alone, not general
behavioral material for the development expressivity, is a basic component of rap
of interpersonal coordination. Both sets of port (Tickle-Degnen & Rosenthal, 1987,
findings demonstrate the need for a more 1990). Some research studies, however,
systematic investigation of context and of have implied the importance of positivity
more complex patterns among the different by finding that negative expressivity is a
behavioral components of rapport. negative correlate of rapport. In a study of
women who interviewed one another about
their daily lives and aspirations, dyads
NONVERBAL EXPRESSIVITY experienced lower rapport when the more
expressive partner of the dyad was also the
From their review of the literature on partner with the more negatively emotional
emotional expressivity, Boone and Buck personality (Tickle-Degnen & Puccinelli,
(2003) conclude that to be openly self- 1999). Rapport was higher when the less
expressive is a claim to one’s own trust negatively emotional partner was the more
worthiness and one’s ability to trust. By expressive partner of the dyad. Other find
displaying feelings openly, vividly, and ings suggest that dyads containing partners
dynamically, people invite others to know who are sensitive to detecting negative
their feelings, intentions, and attitudes. Such emotionality in others tend to have com
openness allows the perceiver to determine promised experiences of rapport (Elfenbein
the potential for an interaction with the & Ambady, 2002; Puccinelli & Tickle-
actor that is responsive to the perceiver’s Degnen, 2004).
20-Manusov.qxd 6/30/2006 5:12 PM Page 386
386–––◆–––Functions
Context factors may explain the inconsis (2003) note that expressivity of an unusu
tency in research results about negative expres ally intense or chaotic nature among newly
sivity. In Tickle-Degnen and Puccinelli’s acquainted individuals would be perceived
(1999) study, the task of discussing one’s daily as untrustworthy and not conducive to
life and aspirations with a relatively new cooperation. From this perspective, a mod
acquaintance implicitly required a pleasant erate level of positivity and negativity of
self-presentation (DePaulo, 1992). In this con expression would be more favorable to rap
text, rapport was higher when expressivity port than either very low levels (unexpres
was more positive than negative. Expressed sive) or very high levels (unusually intense).
negativity by one partner may have been off- Related to this nonlinear perspective is
putting given the context demands. In con Gottman’s (1994) finding that when hap
trast, in Butler et al.’s (2003) study, the task pily married couples openly express nega
of watching an upsetting film with another tive emotion in a manner that is modulated
person would have implicitly required a by positive interpersonal strategies, such
social acknowledgment of the shared nega as humor, they return to a nonnegative
tive experience. Suppression of negative state. Unhappy couples, on the other hand,
expressivity may have been off-putting in this descend into unalleviated and intense nega
context. Perhaps, this suppression was a mes tivity. Positive and negative expressivity
sage of dissimilarity or nonacceptance of not events may cycle relative to one another,
only the suppressing woman’s inner state and the dynamic pattern of these cycles may
but also the partner’s inner state. Further drive the rapport experience.
more, suppression would afford the partner Tickle-Degnen and Gavett (2003)
little responsive action, such as commis observed nonlinear and dynamic patterning
eration. In the context of this study, unsup of nonverbal behavior in six speech thera
pressed negative expressivity would be pists and their new child clients over eight
“positive” because it would imply mutual weekly clinic sessions. There appeared to be
acceptance and understanding of one a higher degree of fluctuating, up-down
another’s experience. changes in levels of nonverbal positivity
If such an interpretation is correct, the and attentiveness within dyads from session
implication is that expressivity can have to session during the first four sessions
meaning that is positive, neutral, or nega compared with the last four sessions, which
tive, depending on the context. Specifically, showed a stable leveling of behavior. The
it is expressivity that has positive meaning therapists’ self-reported experience of rap
in-context that is the correlate or facilitator port, competence, and success, as well as
of rapport. Lending support to this contex their supervisors’ analysis of the sessions,
tual explanation, Bernieri et al. (1996) suggested that the greater degree of behav
found that nonverbal expressivity had a ioral fluctuation in the earlier set of sessions
lower correlation with experienced rapport was due to conditions that challenged effec
in an adversarial context, in which expres tive communication and collaboration. The
sivity may have had a less positive, more greater degree of behavioral stability in the
negative quality, than in a cooperative one, later set appeared to be due to less chal
in which expressivity may have had a more lenging conditions. More concretely, the
positive, less negative quality. therapists and children were learning about
each other and how to communicate and
Nonlinear and Dynamic Patterns of work together effectively in their earlier ses
Nonverbal Expressivity. Boone and Buck sions, and they were more informed and
20-Manusov.qxd 6/30/2006 5:12 PM Page 387
skilled in their interactive patterns in later patterns that are not mutually exclusive of
sessions. one another. Three frequently measured
Furthermore, across the entire set of ses coordinative patterns are matching, interac
sions, therapists who felt the highest level of tional synchrony, and mimicry (Burgoon,
rapport demonstrated intermediate levels of Stern, & Dillman, 1995). Matching refers to
attentiveness and positivity. Therapists who behavioral similarity, often measured as a
felt the lowest level of rapport demonstrated couple’s shared positioning of arms and legs.
low or high—not intermediate—attentive In my work, I have operationalized matching
ness and positivity. These findings support as the similarity in the degree of observed
a view of nonlinearity in the association attentiveness and positivity of partners
between nonverbal behavior and rapport (Tickle-Degnen & Gavett, 2003). Inter
that is consistent with Boone and Buck’s actional synchrony refers to similarity in the
(2003) view that a moderate level of expres rhythm of behavior and to a measurable
sivity is more conducive to feelings of trust enmeshment or interdependence of the tim
worthiness and motives to cooperate than ing of movements. Mimicry is a matching of
extremely low or high levels of expressivity. the behavior of one participant, such as foot
shaking, by another participant close in time
INTERPERSONAL COORDINATION to the original behavior.
Of the three hypothesized components
The research on nonverbal expressivity of rapport (Tickle-Degnen & Rosenthal,
points toward rapport as emerging from 1987, 1990), interpersonal coordination has
multicomponent, nonlinear, and dynamic the most evidence supporting its validity.
behavior in context. The research on inter Parent-infant studies suggest that there are
personal coordination leads to a similar innate pressures for individuals to accom
conclusion. The concept of interpersonal modate to one another behaviorally (Crown,
coordination is particularly useful for Feldstein, Jasnow, Beebe, & Jaffe, 2001;
understanding rapport because the concept Papoušek & Papoušek, 1997). When an
of interpersonal coordination incorporates interaction is going well or supported by
the dynamic and bidirectional, person-to the context, coordinated patterning tends
person interdependence of behavior. Each to increase over the course of an encounter
partner is part of the context for the other (Bernieri, Reznick, & Rosenthal, 1988;
partner, and each partner’s expressivity is Cappella, 1996, this volume; Warner,
the “raw action” material that affords the 1992). Such patterning, however, does not
responsive action necessary to link part always occur. Kritzer and Valenti (1990)
ners into a coordinative unit. Coordination showed that synchrony did not increase
emerges from the context of interpersonal among untrained therapists and their
action. For example, individuals who move clients, but it did so with trained therapists.
more are more likely to show a behavioral Bernieri et al. (1988) found synchrony to
responsiveness to one another beyond that increase in mothers with their own infants
expected by chance alone (Grammer, Kruck, but not among mothers with infants
& Magnusson, 1998). It is this above-chance with whom they were interacting for the
responsiveness that defines interpersonal first time.
coordination. It is possible that individuals who coordi
The terms coordination, responsiveness, nate behaviorally recognize consciously that
and accommodation are used interchange their actions are easy, predictable, efficient,
ably to refer to a whole class of coordinative and coordinated and, consequently, infer
20-Manusov.qxd 6/30/2006 5:12 PM Page 388
388–––◆–––Functions
that these qualities are due to an experiential degrees of synchrony than those who were
understanding of one another (Bem, 1972). less involved. Renfro and Rauh (2005)
But recent research demonstrates that con found that manipulated disruption of
sciousness is not required. Studies in which interpersonal coordination during video-
mimicry is manipulated experimentally sug mediated discussions had a more negative
gest that another’s mimicry of one’s own effect on communication satisfaction and
behavior induces liking of and rapport with emotion when individuals were more rather
the mimicker unconsciously (Chartrand than less interested and involved in the
& Bargh, 1999; Lakin, Jefferis, Cheng, & topic of their discussion.
Chartrand, 2003; see Lakin, this volume)
and promotes prosocial behavior toward Interpersonal Coordination and Positivity.
mimickers as well as those not involved Another example that coordination is not a
in mimicking (van Baaren, Holland, sufficient component of rapport is demon
Kawakami, & van Knippenberg, 2004). strated in the mixed findings of research
on postural matching, one type of coordi
Interpersonal Coordination and Attentive native structure. In naturalistic interactions
ness. Despite the persuasive evidence that of people in established relationships,
interpersonal coordination is a behavioral matching is a positive correlate of rapport
correlate and antecedent of rapport across experience (LaFrance, 1979; LaFrance &
various contexts, it is possible that coordi Broadbent, 1976). Alternatively, in interac
nation is an automatic and spontaneous tions of strangers involved in experimental
correlate of any type of interpersonal expe tasks that had unusual or ambiguous
rience that involves engagement and mutual purposes, matching has been found to be
attention (Bernieri et al., 1996; Burgoon indicative of communication anxiety or
et al, 1995; Dijksterhuis & Bargh, 2001; nervousness rather than rapport (Bernieri,
Hatfield, Cacioppo, & Rapson, 1994). An 1988; LaFrance & Ickes, 1981). The match
example that coordination is not a suffi ing in these latter contexts was relatively sta
cient component of rapport is demonstrated tic and unchanging, appearing rigid and
in the findings of a study of people interact strained (Bernieri, 1988). LaFrance and
ing via typing in an Internet-style chat Ickes (1981) have suggested such matching
room (Niederhoffer & Pennebaker, 2002). in anxiety-provoking contexts may reflect
Accommodation to one another’s linguistic an intention or effort to connect rather than
styles occurred regardless of rapport levels an actual expression of rapport. Even
(Giles & Coupland, 1991). Engagement in though coordinative behavior in these con
the form of attentive interest appeared to be texts reflects feelings that are negative, such
a better indicator of rapport than the coordi patterns also reflect a prosocial, and there
native process of accommodation. fore, positive intention; that is, the intention
Other studies suggest that attentiveness or desire to be in rapport. Such an inten
and synchrony are additive or interactive tion, in the long run, actually may promote
with respect to rapport. Kendon (1970) rapport but, perhaps, not in short-lived,
found that observers who were present and experimental encounters.
attentive to, but not involved in, a conver
sation between two people tended to syn Nonlinear and Dynamic Patterns of
chronize their movement to the speaker. Interpersonal Coordination. Some research
The two people who were engaged with has demonstrated that interpersonal coordi
each other showed more pronounced nation varies along a looseness-to-tightness
20-Manusov.qxd 6/30/2006 5:12 PM Page 389
dimension, and that this dimension is with the therapist accommodating more to
important for understanding its role in rap the client than vice versa (Lichtenberg et al.,
port (Bernieri & Rosenthal, 1991; Burgoon 1998).
et al., 1995; Cappella, 1996; Gottman, These findings suggest that the varia
1994; Warner, Malloy, Schneider, Knoth, tion in degree of coordination reflects the
& Wilder, 1987; Watzlawick, Beavin, & dyad’s effort to optimize dyadic function
Jackson, 1967). In general, a moderate ing. Dyadic functioning involves maintain
degree of coordination between individuals ing connection to others in order to
indicates effective communication, bond accomplish a socially mediated task. Just as
ing, and positive interpersonal outcomes the attainment and maintenance of connec
(Bernieri et al., 1988; Jaffe, Beebe, Feldstein, tion to others drives much social behavior
Crown, & Jasnow, 2001; Kritzer & (e.g., Burgoon et al., 1995; Cappella, 1996;
Valenti, 1990; Warner, 1992). Based on Goffman, 1967), it also appears, more
this research, Tickle-Degnen and Gavett specifically, to drive nonverbal behavior
(2003) divided speech therapy dyads into associated with rapport. Research on the
three categories—those having interper automatic functions of simple forms of
sonal coordination that was loose, tight, behavioral mimicry (e.g., face touching
or intermediate—based on the absolute or foot shaking) suggests that individuals
difference score between the therapist’s and mimic and respond to mimicry in order to
child’s rated levels of attentiveness or posi affiliate with one another (Lakin et al.,
tivity, across three slices in time within a 2003). Many factors (including subliminal
session. The higher the difference score, priming and recent failure to affiliate satis
the “looser” the pattern of coordination. factorily with another person) that stimu
Consistent with previous research findings, late affiliation and affiliation goals induce
therapists experienced higher rapport when mimicry and associated feelings of liking
there was an intermediate degree of (Lakin & Chartrand, 2003).
matched attentiveness or matched positiv
ity, and experienced lower rapport when
the match was loose or tight. CONTEXT AND DIMENSIONALITY
The results from some studies suggest IN THE NONVERBAL BEHAVIOR
that very tightly coordinated action, as OF THE RAPPORT ECOSYSTEM
reflected in a relatively high degree of
symmetry, regularity, periodicity, or pre This overview of the recent research on
dictability in the interaction pattern, rapport demonstrates that the behavioral
appears to be indicative of a challenging, components of prosocial or positive expres
uncomfortable social situation or an effort sivity, interpersonal coordination, and an
to overcome a problematic interaction. attentive engagement combine in function to
Cappella (1996), for example, found that initiate, produce, indicate, and sustain rap
strangers were more likely to accommodate port. These behavioral components of rap
to one another in their speech turn-taking port emerge in relation (a) to challenges in
than were well-acquainted individuals, sug the context of interaction, including chal
gesting that “hyperpoliteness” operates in lenges to effective communication, connec
initial encounters. In another study, thera tion, and relationship and (b) to achieving
pist-client dyads with ultimately unsuccess other beneficial dyadic outcomes. There is
ful, compared with successful, outcomes moderation, or a slight degree of “loose
were characterized by tighter coordination, ness,” in the intensity and structure of
20-Manusov.qxd 6/30/2006 5:12 PM Page 390
390–––◆–––Functions
behavior when individuals experience rap different interpersonal tasks. Bernieri et al.
port. Relatively low or high levels of inten (1996) and Puccinelli et al. (2003, 2004)
sity in the behavioral components of rapport had each of their participating dyads per
may indicate that individuals are either not form two tasks: one in which the purpose
connecting or attempting to do so in condi was to communicate important personal
tions that are challenging for preserving the information via face-to-face conversation,
interpersonal connection required to achieve and another one, in which the purpose was
beneficial dyadic outcomes. to solve a problem using props and physical
Outcome goals and process subgoals objects creatively in an object-focused task.
operate simultaneously and interactively Bernieri et al. (1996) compared an adver
during interpersonal interaction (Austin sarial debate with a cooperative trip-
& Vancouver, 1996; Dijksterhuis & Bargh, planning task. They found more behavioral
2001) and shape the physical, spatial, elements to be associated with the dyad
and temporal structure of interaction. With members’ experienced rapport during the
respect to rapport, relevant outcome goals debate than during the trip-planning task.
include affiliation, enjoyment, and the In addition, the behavioral elements indica
accomplishment of an interpersonal task tive of rapport during the debate were less
(e.g., getting a child to eat, completing a intercorrelated—that is, more multidimen
project on deadline, or facilitating student sional—compared with those elements
learning). Relevant process subgoals include indicative of rapport during the trip-plan
communicating internal states, enabling ning task, which were more intercorre
partner responses, and showing agreement lated—that is, more unidimensional. In the
and cooperation. debate task, behaviors associated with rap
There may be, at times, a simpler goal port included those involved in expressing
structure, that is, fewer goals and more and connecting, such as synchrony, proxim
integrated or congruent goals, in contexts ity, and to a small degree expressivity, as
that support rapport than in contexts that well as those involved in conversing effec
do not. Under a simple goal structure, tively face to face, such as the back-channel
behavioral positivity, coordination, and responses of head nodding and “uh hmms,”
attentiveness may covary with one another eye contact, and talkativeness. In the trip-
in a relatively stable fashion across the planning task, on the other hand, individu
interaction, all being correlated positively als were active and expressive or were
with rapport. A complex goal structure relatively quiet and nonexpressive, and if
may occur in contexts that do not support they were active and expressive they had
rapport. Individuals may express negativity, higher rapport. Furthermore, observers
poor responsiveness, and inattention, and were able to detect the rapport experienced
mix or alternate these signals with others by the participants more accurately from the
that indicate their intention to attempt to more simple configuration of behavior
establish or restore rapport. In these situa encoded during the trip-planning task
tions, there may be more than one dimen than from the more complex configuration
sion in the correlational structure of encoded during the debate.
behavior and the experience of rapport. In a similar vein, Puccinelli et al. (2003,
The most direct support for this idea 2004) compared conversations about daily
of uni- versus multidimensionality to the life activities and aspirations with behav
correlational structure of behavior and iors during a puzzle construction task. In
experienced rapport comes from studies of contrast with Bernieri et al. (1996), who
20-Manusov.qxd 6/30/2006 5:12 PM Page 391
392–––◆–––Functions
challenges of the situation. At the same time, skill. A complete research agenda requires
self-consciousness declined. Similarly, in that additional features and pressures, such
studies of therapist-child interactions as those related to individual and group
(Dunkerley et al., 1997; Tickle-Degnen & differences, be investigated systematically
Coster, 1995), as behavioral attentiveness, as well. For example, the individual in the
positivity, and coordination increased so dyad who has the highest level of trait
did the match between the challenge of expressivity drives rapport (Sullins, 1991;
therapeutic tasks and the child’s demon Tickle-Degnen & Puccinelli, 1999).
strated ability level increasingly. On the Relative to individuals who are less accu
other hand, these behavioral correlates of rate in interpreting others’ inner states from
rapport showed very little relationship to nonverbal behavior, individuals who are
the task challenge as measured indepen more accurate have rapport experiences
dently of the child’s ability, which indicates that are more reactive or sensitive to the
that task challenge alone did not moderate other’s emotional traits (Puccinelli &
this behavior. Rather, nonverbal behavior Tickle-Degnen, 2004). Similarly, individu
was associated with the child’s own capaci als who have strong social motives, such as
ties in interaction with the context of action. a need to belong or a tendency toward high
Table 20.1 shows the distinctions between self-monitoring, are more attentive to non
optimal and suboptimal rapport experi verbal behavior and respond with greater
ences based on Csikszentmihalyi’s (1990) empathic accuracy (Pickett, Gardner, &
model, which I have modified for the pur Knowles, 2004) and behavioral mimicry
poses of understanding the rapport ecosys (Cheng & Chartrand, 2003).
tem. When conditions provide the just-right Emotional expressivity, nonverbal sensi
challenge (i.e., not over- and not under- tivity, and social motives vary according to
challenging) to a dyad’s skill level, the gender, health conditions, and age as well.
dyad’s nonverbal behavior would clearly Females are more emotionally expressive
and simply express the rapport experience. generally, their own emotions are more
When conditions are overchallenging to responsive to the behavior of others
dyad skill, the behavior would show signs (Fischer, 2000), and their interactions are
of effort and nervousness and, perhaps, driven more by mutual participation goals
more behavioral elements or more intense (Strough & Berg, 2000) compared with
behavior indicative of attempts to establish males. Some health conditions, such as the
or recover rapport. When underchalleng movement disorder of Parkinson’s disease,
ing, there would be very little purposeful prevent individuals from expressing their
interpersonal behavior, unless there are inner states with clarity and responding
other implicit contextual pressures to estab to others’ states in a coordinated manner.
lish rapport. Observers often have negative impressions
and presume social apathy or negative social
motives, such as deceptiveness, inaccurately
♦ Individual and Group in these individuals, leading to stigmatiza
Differences in the tion processes that prevent or erode rapport
(Lyons & Tickle-Degnen, 2003; Tickle-
Rapport Ecosystem
Degnen & Lyons, 2004). Interestingly, the
movement disorder of Parkinson’s disease
The nonverbal correlates of rapport are mimics the expressive behavior of the very
shaped by many contextual features and old (Singer, 1974) and can occur in younger
pressures besides task challenge and dyad and older adults who, otherwise, are
20-Manusov.qxd 6/30/2006 5:12 PM Page 394
394–––◆–––Functions
Task conditions
Too little of a challenge Task challenge matches Overly challenging for dyad
for dyad members dyad members’ skills members
Unstructured Structured Complicated structure
Purposeless Clear goals Ambiguous goals
Feedback irrelevant Ongoing feedback about Little feedback about success
success in achieving goals in achieving goals
Nonverbal behavior
Unfocused Absorbed attention Vigilant attention
Inactive, perhaps Expressive, prosocial Disregulated affect, perhaps
prosocial intentions affect and intentions prosocial intentions
No effort Engaged but relaxed Effortful, intense
Lethargic, off-task Calm, socially clear Nervous or socially
behavior behavior ambiguous behavior
Hypo-responsivity to Moderate responsivity to Excessive responsivity to
partner’s behavior partner’s behavior partner’s behavior
Rapport ecosystem
Unengaged Resilient, secure bond Insecure bond
Failure to develop or Ease in sustaining Difficulty effectively
maintain bond. There interaction toward interacting toward goal
may be attempts to goal achievement achievement. Failure to
create or restore bond. develop or maintain bond.
There may be attempts to
create or restore bond.
NOTE: Table influenced by Csikszentmihalyi (1990).
emotion and expression are seen as central to to the rapport experience is far from
achieving harmonious relationships with understood and requires more research.
others, an achievement highly valued within Despite the call in this chapter to integrate
the Confucian tradition and collectivistic cul an understanding of rapport with optimal
ture (Chang & Holt, 1994). This emphasis experience, rapport has been studied rarely in
on moderation contrasts to the emphasis contexts in which conditions are favorable
on personal expression in individualistic for such experience. Yet social enjoyment
cultures. For example, Asians compared is one of the primary types of satisfaction
with North Americans show more control of that individuals experience and seek out
facial expressivity in response to emotionally (Csikszentmihalyi & LeFevre, 1989). To fur
evocative stimuli (Okazaki, Liu, Longworth, ther understand the implications of this con
& Minn, 2002). tention, more research is needed on describing
people’s tasks when they have optimal social
experience and what their behavior looks like
♦ Implications for
under these optimal conditions. Specifically,
Future Research the review of the literature on rapport and
nonverbal behavior suggests the following
People experience rapport with their hypothesis: When contextual factors support
family members, friends, physicians, rapport, nonverbal behavior will show a clear
teachers, and coworkers as they engage in and simple relationship to experienced rap
tasks that vary in how challenging, struc port. On the others hand, when contextual
tured, comfortable, or intimate they are. In factors threaten or undermine rapport, non
contrast to the variety of task conditions verbal behavior will show a complex or
found in natural social life, the laboratory, ambiguous relationship to experienced rap
in which most research on rapport is con port. To test this hypothesis, we need to
ducted, manipulates or represents only a clarify what is meant by “simple” or “com
limited array of conditions. It presents spe plex” relationship, and the following con
cialized and unique sets of task constraints structs require more careful development and
that influence patterns of nonverbal operationalization: linearity, loose versus tight
behavior and their functional associations coordination, and dimensionality in the asso
with outcomes (Grahe & Sherman, 2005; ciation of nonverbal behavior with rapport.
Tamir et al., 2004). Experimental interac The experience of rapport also requires
tion tasks involve unacquainted or mildly more careful elaboration from a temporal
acquainted participants typically, and or dynamic perspective (see Cappella &
implicit norms require pleasantness and Schreiber, this volume). Individuals may
appropriate modulation of emotion. fluctuate in the strength of their rapport
Research on rapport and nonverbal com over time, yet they may have strong motives
munication must take into account these to retain rapport. At times, they may not
constraints operating on behavior. We feel in rapport with one another, but their
should also consider how laboratory task behavior demonstrates features that are
constraints related to interpersonal expres somewhat congruent with what would be
sivity and coordination are similar to or demonstrated if their rapport experience
deviate from the everyday situations of were stronger. Tracking the development,
participants outside the laboratory. waning, and restoration of rapport
Finally, how individual and group differ across time in enduring relationships will
ences interact with task constraints as shed light on these differences in rapport
nonverbal behavior mediates and responds experiences. Finally, there is a need for
20-Manusov.qxd 6/30/2006 5:12 PM Page 396
396–––◆–––Functions
measures that assess not only being in rap of Personality and Social Psychology, 54,
port but also desire to be in rapport. 243–253.
Together, this work will enable a theo Bernieri, F. J., & Rosenthal, R. (1991). Inter
retically more coherent approach to study personal coordination: Behavior matching
and interactional synchrony. In R. Feldman
ing the nonverbal correlates of rapport and
& R. Rime (Eds.), Fundamentals of nonver
interpreting the results. Models such as
bal behavior (pp. 401–432). Cambridge,
that of Csikszentmihalyi (Csikszentmihalyi,
UK: Cambridge University Press.
1990; Csikszentmihalyi & Csikszentmihalyi, Boone, T., & Buck, R. (2003). Emotional
1988) are helpful in guiding this work for expressivity and trustworthiness: The role
ward. They illustrate how optimal experi of nonverbal behavior in the evolution
ences, such as rapport, emerge in the of cooperation. Journal of Nonverbal
context of accomplishing interpersonal Behavior, 27, 163–182.
tasks to meet personal, dyadic, and group Brunswik, E. (1955). Representative design and
needs. These models can be used to organize probabilistic theory. Psychological Review,
the systematic investigation of how nonver 62, 193–217.
bal behavior functions in context to initiate, Burgoon, J. K, Stern, L. A., & Dillman, L.
(1995). Interpersonal adaptation: Dyadic
produce, indicate, and sustain rapport.
interaction patterns. Cambridge, UK:
Cambridge University Press.
Butler, E. A., Egloff, B., Wilhelm, F. H., Smith,
♦ References N. C., Erickson, A., & Gross, J. J. (2003).
The social consequence of expression sup
pression. Emotion, 3, 48–67.
Ambady, N., Bernieri, F. J., & Richeson, J. A. Cappella, J. N. (1996). Dynamic coordination of
(2000). Toward a histology of social vocal and kinesic behavior in dyadic inter
behavior: Judgmental accuracy from thin action: Methods, problems, and interper
slices of the behavioral stream. Advances in sonal outcomes. In J. H. Watt & C. A.
Experimental Social Psychology, 32, VanLear (Eds.), Dynamic patterns in
201–271. communication processes (pp. 353–386).
Austin, J. T., & Vancouver, J. B. (1996). Goal Thousand Oaks, CA: Sage.
constructs in psychology: Structure, process, Chang, H-C, & Holt, G. R. (1994). A Chinese
and content. Psychological Bulletin, 120, perspective on face as inter-relational con
338–375. cern. In S. Ting-Toomey (Ed.), The chal
Bem, D. J. (1972). Self-perception theory. In lenge of facework: Cross-cultural and
L. Berkowitz (Ed.), Advances in experimen interpersonal issues (pp. 95–132). Albany:
tal social psychology (Vol. 6, pp. 1–62). State University of New York Press.
New York: Academic Press. Chartrand, T. L., & Bargh, J. A. (1999). The
Bernieri, F. (1988). Coordinated movement and chameleon effect: The perception-behavior
rapport in teacher-student interactions. link and social interaction. Journal of
Journal of Nonverbal Behavior, 12, Personality and Social Psychology, 76,
120–138. 893–910.
Bernieri, F. J., Gillis, J. S., Davis, J. M., & Grahe, Cheng, C. M., & Chartrand, T. L. (2003). Self-
J. E. (1996). Dyad rapport and the accuracy monitoring without awareness: Using mim
of its judgments across situations: A lens icry as a nonconscious affiliation strategy.
model analysis. Journal of Personality and Journal of Personality and Social Psy
Social Psychology, 71, 110–129. chology, 85, 1170–1179.
Bernieri, F. J., Reznick, J. S., & Rosenthal, R. Compact Oxford English dictionary. (2005).
(1988). Synchrony, pseudosynchrony, and Oxford, UK: Oxford University Press.
dissynchrony: Measuring the entrainment Retrieved May 5, 2005, from www
process in mother-infant interactions. Journal .askoxford.com
20-Manusov.qxd 6/30/2006 5:12 PM Page 397
Crown, C. L., Feldstein, S., Jasnow, M. D., J. M. Fernandez-Dols (Eds.), The psychol
Beebe, B., & Jaffe, J. (2001). The cross- ogy of facial expression (pp. 103–129).
modal coordination of interpersonal tim Paris: Cambridge University Press.
ing: Six-week-old infants’ gaze with adults’ Gibson, J. J. (1986). The ecological approach to
vocal behavior. Journal of Psycholinguistic visual perception. Hillsdale, NJ: Erlbaum.
Research, 30, 1–23. (Original work published 1979).
Csikszentmihalyi, M. (1990). Flow: The psy Giles, H., & Coupland, N. (1991). Language:
chology of optimal experience. New York: Contexts and consequences. Pacific Grove,
HarperCollins. CA: Brooks/Cole.
Csikszentmihalyi, M., & Csikszentmihalyi, I. S. Goffman, E. (1967). Interaction ritual: Essays on
(1988). Optimal experience: Psychologi face-to-face behavior. New York: Pantheon.
cal studies of flow in consciousness. Gottman, J. M. (1994). What predicts divorce?
Cambridge,UK: Cambridge University Press. The relationship between marital proc
Csikszentmihalyi, M., & LeFevre, J. (1989). esses and marital outcomes. Hillsdale, NJ:
Optimal experience in work and leisure. Erlbaum.
Journal of Personality and Social Psycho Grahe, J. E., & Bernieri, F. J. (2002). Self-
logy, 56, 815–822. awareness of judgment policies of rapport.
DePaulo, B. M. (1992). Nonverbal behavior and Personality and Social Psychology Bulletin,
self-presentation. Psychological Bulletin, 28, 1407–1418.
111, 203–243. Grahe, J. E., & Sherman, R. A. (2005). An eco
Dijksterhuis, A., & Bargh, J. A. (2001). The per logical examination of rapport using
ception-behavior expressway: Automatic a dyadic puzzle task. Unpublished manu
effects of social perception on social behav script.
ior. In M. Zanna (Ed.), Advances in exper Grammer, K. Kruck, K. B., & Magnusson, M. S.
imental social psychology (Vol. 33, (1998). The courtship dance: Patterns of
pp. 1–40). San Diego, CA: Academic Press. nonverbal synchronization in opposite-sex
Dunkerley, E., Tickle-Degnen, L., & Coster, W. encounters. Journal of Nonverbal Behavior,
(1997). Therapist-child interaction in the 22, 3–29.
middle minutes of sensory integration treat Hatfield, E., Cacioppo, J. T., & Rapson, R. L.
ment. American Journal of Occupational (1994). Emotional contagion. Paris:
Therapy, 51, 799–805. Cambridge University Press.
Ekman, P., & Keltner, D. (1997). Universal Heft, H. (2003). Affordances, dynamic
facial expressions of emotion: An old con experience, and the challenge of reifica
troversy and new findings. In U. Segerstråle tion. Ecological Psychology, 15, 149–180.
& P. Molnár (Eds.), Nonverbal communi Izard, C. (1994). Innate and universal facial
cation: Where nature meets culture (pp. expressions: Evidence from developmen
27–46). Mahwah, NJ: Erlbaum. tal cross-cultural research. Psychological
Elfenbein, H. A., & Ambady, N. (2002). Bulletin, 115, 288–299.
Predicting workplace outcomes from the Jaffe, J., Beebe, B., Feldstein, S., Crown, C. L., &
ability to eavesdrop on feelings. Journal of Jasnow, M. D. (2001). Rhythms of dialogue
Applied Psychology, 87, 963–971. in infancy: Coordinated timing in develop
Fischer, A. H. (Ed.). (2000). Gender and emo ment. Monographs of the Society for
tion: Social psychological perspectives. Research in Child Development, 66(2),
Paris: Cambridge University Press. vi-131.
Freud, S. (1924). On the history of the psycho James, W. (1890). Principles of psychology.
analytic movement. In J. Riviere (Trans.) New York: Holt.
& E. Jones (Ed.), Collected papers (Vol. I, Kendon, A. (1970). Movement coordination in
pp. 287–359). London: Hogarth. (Original social interaction: Some examples described.
work published 1914) Acta Psychologica, 32, 100–125.
Fridlund, A. J. (1997). The new ethology of Kritzer, R., & Valenti, S. S. (1990, March).
human facial expression. In J. A. Russell & Rapport in therapist-client interactions:
20-Manusov.qxd 6/30/2006 5:12 PM Page 398
398–––◆–––Functions
PART IV
CONTEXTS AND
CONSEQUENCES
21-Manusov.qxd 6/30/2006 6:11 PM Page 402
21-Manusov.qxd 6/30/2006 6:11 PM Page 403
21
NONVERBAL COMMUNICATION
IN CLOSE RELATIONSHIPS
� Patricia Noller
University of Queensland
◆ 403
21-Manusov.qxd 6/30/2006 6:11 PM Page 404
emotion and also feel that emotion with the less close friends were more accurate than
person. The authors review studies con close friends, particularly if the senders had
cerned with empathy-related responding (or been asked to conceal the negative emotion
responding in a way that lets the other (making the message more ambiguous).
person know that one empathizes with their These researchers suggest that close friends
situation), including those studies involving may have more to lose than do less close
the assessment of facial and gestural cues of friends if they perceive concealed negativity
young children (e.g., Miller, Eisenberg, accurately. These friends may believe that
Fabes, & Shell, 1996; Zahn-Waxler, if the negative message is ignored, the issue
Radke-Yarrow, Wagner, & Chapman, 1992; will go away.
Zahn-Waxler, Robinson, & Emde, 1992). Another relational feature is overall sat
Using actual facial expressions of emo isfaction, and it too may be related to sensi
tion in research studies to assess people’s tivity. In support of this suggestion, Noller
empathy can help deal with the problem (1984, 2001), using the standard-content
of social desirability affecting self-report paradigm (a research procedure that
responses and the tendency, particularly involves the use of ambiguous words to
of older children and adults, to react in ensure that decoders are focusing on the
socially expected ways in more constrained nonverbal cues; see Noller, 2005a), found
research paradigms. Losoya and Eisenberg a relation between decoding accuracy and
(2001) show that affective empathy can be relationship satisfaction: Those who were
distinguished from sympathy (which does happy in their marriages, and particularly
not involve feeling the person’s distress), husbands, were more accurate at decoding
however, and that affective empathy is their spouses than those who were unhappy
correlated with the overall quality of social in their marriages. The study also showed
functioning. that satisfied spouses were more accurate
Besides individual skill in affective empa at decoding their spouses than they were
thy, there is also some evidence for a “close at decoding strangers, whereas those who
ness of relationship effect” on decoding were unhappy with their marriages were
sensitivity, with close relationship partners’ more accurate at decoding strangers than
greater knowledge about each other’s lives they were at decoding their spouses. In addi
leading to greater accuracy at decoding tion, Noller and Ruzzene (1991) showed
nonverbal behavior. For example, romantic that happy couples were more accurate
partners (Noller & Ruzzene, 1991) and than unhappy couples at identifying the
friends (Fleming, Darley, Hilton, & type of affect being experienced by their
Kojetin, 1990; Stinson & Ickes, 1992) are partners and were also better able to iden
generally better at decoding one another’s tify their partners’ goals and intentions in
nonverbal cues than are strangers. More an interaction.
specifically, Sternglanz and DePaulo (2004) In a later study involving newlyweds,
compared the accuracy of friends and and also using the standard-content para
strangers at decoding nonverbal cues to digm, Noller and Feeney (1998) reported
emotion and found that, overall, friends that nonverbal accuracy increased over
were more accurate than strangers at time for all three message types: positive,
decoding cues to emotion, and close friends neutral, and negative. This finding suggests
were more accurate at decoding clearly that couples may become better at under
expressed negative emotions. For some standing each other’s nonverbal behavior
negative emotions in their study, however, over time, although practice effects, while
21-Manusov.qxd 6/30/2006 6:11 PM Page 406
unlikely, cannot be ruled out (see Noller & affected negatively in relationships char
Feeney, 1994, for more discussion of this acterized by high levels of interdependence,
issue). We also found that wife-to-husband high levels of insecurity, and consequent
communications were decoded more accu concerns about the vulnerability of the rela
rately than were husband-to-wife commu tionship. On the other hand, Simpson,
nications, particularly if the messages were Ickes, and Grich (1999) found that anx
positive. This finding is similar to that in iously attached women became hypervigi
the earlier study where the effect was a lant and hence more accurate when the
function of the more accurate encoding relationship was under threat. This latter
of positive messages by wives. Husbands’ finding suggests that a complex model
accuracy at decoding positive messages is needed to account for the role of cogni
from their wives was associated with their tive biases in explaining decoding in close
current marital satisfaction, and husbands’ relationships.
earlier satisfaction predicted their later Simpson, Orina, and Ickes (2003) tested
accuracy at decoding their wives’ commu their empathic accuracy model further in a
nications across all three types of messages. sample of married couples and found sup
It seems likely that these husbands’ decod port for their predictions. Feelings of close
ing was more accurate because their ability ness decreased for those who were more
was not affected by negative perceptions of empathically accurate when their partners
the relationship with the partner. reported relationship-threatening thoughts
and feelings, and feelings of closeness
increased for those accurate decoders whose
MOTIVATED INACCURACY partners reported nonthreatening thoughts
and feelings. Similar results were found
Ickes and Simpson (1997, 2001) have when outsiders rated the level of threat con
proposed an empathic accuracy model that tained in the partners’ thoughts and feel
may help to explain some of the findings ings. There was also evidence in these data
just discussed. They suggested that there are that an individual’s own perception of rela
times when those in close relationships are tionship vulnerability, and not just his or
actually motivated to be inaccurate in terms her perception of the partner’s sense of rela
of their decoding of one another’s nonver tionship vulnerability, affected the sense of
bal behavior. The authors argue that this closeness with the partner. As before, these
motivated inaccuracy can be explained in findings fit with those of Sternglanz and
terms of the threatening versus nonthreat DePaulo (2004), who found some evidence
ening nature of the discussion in which for motivated inaccuracy in the decoding of
the partners are engaged. They proposed a close friends.
positive relation between accuracy and
relationship satisfaction when the material
being discussed is relatively nonthreatening BIAS IN DECODING
and a negative relation when the discussion
is contentious, conflicted, and threatening The decoding of nonverbal cues may be
to the relationship. biased in various ways, particularly by
Simpson, Ickes, and their colleagues cognitive processes that make individuals
have carried out several studies to test this prone to distortions of perception in a nega
model. For example, Simpson, Ickes, and tive direction. For example, although Noller
Blackstone (1995) showed, in a sample of (1984) did not find an expected bias related
dating couples, that empathic accuracy was to marital satisfaction, the study did reveal a
21-Manusov.qxd 6/30/2006 6:11 PM Page 407
bias related to gender. Wives tended to maintained irrespective of the valence of the
make errors in a positive direction, decod stimuli. In a more complex empathic accu
ing negative messages as neutral and neutral racy task, these researchers were also able
messages as positive, whereas husbands to show an effect of need to belong on indi
tended to make them in a negative direction, viduals’ sensitivity to the thoughts and feel
decoding positive messages as neutral and ings of another. Thus, those with a strong
neutral messages as negative. Gaelick, need to belong are likely to be highly sensi
Bodenhausen, and Wyer (1985) found a tive to close others’ nonverbal communi
similar bias: Spouses reciprocated the emo cation. In a study assessing the extent to
tion they thought their partner was express which need to belong and rejection sensitiv
ing, but because they were less accurate at ity could predict accuracy on the social per
decoding partners’ expressions of love, only ception task, known as The Diagnostic
hostility tended to be reciprocated. Analysis of Nonverbal Accuracy 2 (Nowicki
Downey and Feldman (1996) have & Duke, 1994), however, Pickett et al.
explored the construct of rejection sensitiv (2004) found that accuracy was predicted
ity as leading to bias in the decoding of by the need to belong, but it was not pre
nonverbal behavior. They argued that rejec dicted by rejection sensitivity. They explain
tion-sensitive individuals are prone, because this difference in terms of the nature of
of their heightened anxiety and sensitiv the task, arguing that rejection sensitivity
ity, to perceive ambiguous nonverbal cues is likely to be more influential in a task
as indications of dislike and rejection. involving ambiguous stimuli that could be
Rejection sensitivity tends to be maladap interpreted as rejecting.
tive for relationships (Downey, Freitas,
Michaelis, & Khouri, 1998), because rejec
tion-sensitive individuals tend to focus on SUMMARY
any potential negativity in a message and
ignore or fail to perceive positivity. Likewise, Overall, decoding accuracy tends to be
Pickett, Gardner, and Knowles (2004) argue affected by the closeness of the relationship,
that individuals who are chronically high in although accuracy may be lower in very
the need to belong are likely to be hypervig close relationships when communications
ilant in scanning the social environment for are perceived as threatening the stability of
indications of acceptance or rejection, espe the relationship. The quality of the relation
cially if they are lonely or have experienced ship also influences decoding accuracy, at
rejection continually (Gardner, Jefferis, least in some studies. There is also evidence
Knowles, & Pickett, 2003; Gardner, Pickett that decoding accuracy can be affected by
& Brewer, 2000). gender and by such cognitive processes as
Need to belong and rejection sensitivity rejection sensitivity and the need to belong.
are likely to be related in terms of their As shall be discussed in a later section of
effects on nonverbal decoding. Pickett et al. this chapter, decoding accuracy may also be
(2004) explored the association between affected by attachment insecurity.
the need to belong and accuracy in the iden
tification of facial expression and vocal
tone. They found that accuracy in identify ♦ Nonverbal Expressiveness
ing both facial expressions and vocal tone
was positively correlated with scores on
a measure of need to belong. The authors Accurate expression of emotion (or inten
also reported that this relationship was tion, according to Fridlund, 1994; Fridlund
21-Manusov.qxd 6/30/2006 6:11 PM Page 408
& Russell, this volume) is important in negative behavior) when wives were most
close relationships. Much of the time, each stressed. Neff and Karney argue, on the
partner needs to know how the other is feel basis of these data, that husbands and wives
ing so that those feelings can be taken into are equally capable of providing support
account in their attempts to solve problems to their partners, but that wives seem able
and make decisions. It is also important for to provide that support when it is most
partners to let one another know when they needed. It may be that husbands do not rec
are feeling sad or disappointed so that the ognize the cues indicating their wives’
appropriate level of social support can be greater levels of stress or that they believe
provided. Nevertheless, individuals differ support will be more effective if offered
in their willingness to express their feelings, in “calmer” moments. Alternatively, hus
and this willingness can be affected by gen bands more than wives may resent the
der, by their ambivalence about the value expectation that they will support their
of communicating feelings, by the type of wives at times of distress and become nega
relationship in which they are involved, tive in response. A further possibility is that
and by the quality of their relationship. husbands, because of their greater emphasis
Specifically, nonverbal behaviors are on autonomy and independence, are less
central to the expression of love in close sympathetic to those who need to seek
relationships. A study by Gonzaga, Keltner, support (MacGeorge, 2003).
Londahl, and Smith (2001) indicated that Willingness to express feelings depends,
behaviors, such as head nods, forward at least to some extent, on the individuals’
leans, and Duchenne (felt as opposed to “ambivalence about emotional expression”
contrived) smiles were central to the expres (King & Emmons, 1990, p. 864), which has
sion of love. In addition, nonverbal behav been defined as experiencing a conflict
iors that create closeness, such as gaze, between “the competing goals of wanting
close distance, body orientation, touch, and to show how one is feeling yet fearing
smiling, are important in the nonverbal the consequences of such self-expression”
expression of love (Andersen, 1985). These (Mongrain & Vettese, 2003, p. 545).
behaviors (often termed immediacy behav Mongrain and Vettese found that women’s
iors) are seen as signaling warmth, commu self-reported ambivalence over the expres
nicating availability, decreasing distance, sion of emotion was related to their reports
and promoting involvement in interactions of suppressed anger toward their boyfriends.
(see Andersen, Guerrero, & Jones, this In other words, “ambivalent” women
volume). reported that they were not prepared to
Because social support is important for express the anger that they felt about the
relationship satisfaction (Coyne & Smith, behavior of their boyfriends. In addition,
1994; Pasch & Bradbury, 1998), it often when these highly ambivalent (about emo
needs to be expressed clearly, as does the tional expression) women were observed in
supportive response. As Neff and Karney a conflict resolution task, they were less
(2005) have pointed out, the timing of such congruent in their communication (i.e., there
a response may be critical. These research was a mismatch between their verbal and
ers found that husbands and wives pro nonverbal channels). Furthermore, when
vided equal proportions of supportive mood and personality constructs were con
behaviors, but wives tended to provide the trolled for, they displayed more overt sub
most support when the husbands’ levels of missiveness toward their boyfriends than did
stress were highest, whereas husbands pro women who were low in ambivalence. These
vided the least support (in fact the most findings suggest that women are likely to
21-Manusov.qxd 6/30/2006 6:11 PM Page 409
tolerate bad behavior from their boyfriends of negative danger signs, such as invalida
and even be submissive toward them in such tion, escalation, and withdrawal, these
circumstances. The chances of these couples researchers found that those who cohabited
having healthy, egalitarian, and respectful before engagement reported more danger
relationships seem bleak. signs both before and after marriage than
Guerrero (1997) explored whether type of did those who cohabited only after engage
relationship affects encoding. She explored ment or marriage.
nonverbal behavior using a partial round- Cohan and Kleinbaum (2002) proposed
robin design, so that an individual’s behav that cohabiters treat each other less posi
ior could be compared across relationships: tively than other couples and that this ten
with a same-sex friend, an opposite-sex dency to less positive behavior carries over
friend, and a romantic partner. She found into the marriage and may contribute to
that romantic relationships could be distin the higher divorce rate among those who
guished from other types of relationships by cohabit before marriage (see, also, Heaton,
the closer proximity and higher levels of 2002; Kieran, 2002). In the study by Kline
touch and gaze in those relationships. In et al. (2004), those who had cohabited
addition, there were more silences, longer before their engagement were coded as more
response latencies, and less fluency in negative than those who cohabited only
romantic relationships. Nodding and vocal after engagement or after marriage. Those
interest were more common in friendships who cohabited before marriage were also
than in romantic relationships, and there coded as less positive than those who cohab
was more evidence of postural congruence ited only after marriage, and those who
in same-sex than in opposite-sex dyads. cohabited after engagement were also coded
Particular patterns of nonverbal expres as less positive than those who cohabited
sion are also linked to relationship quality after marriage. These differences in the pos
or satisfaction. Unhappy couples display itivity of communication were found before
more negative nonverbal behaviors and less marriage and were also in evidence after the
positive emotion than more satisfied cou marriage. These findings also fit with those
ples (Burgoon, Buller, & Woodall, 1996). of Cohan and Kleinbaum (2002): that mar
Specifically, Gottman (1994) has argued ried couples who had cohabited before
that happy couples tend to display five pos marriage tended to treat each other less pos
itive behaviors for every negative behavior, itively than those who had not.
whereas unhappy couples have a much Other research has explored the associa
lower ratio of positive to negative behaviors tions between emotional expression and
and may even display more negative behav marital satisfaction in couples. Waldinger,
iors than positive. Thus, if we think of this Schulz, Hauser, Allen, and Crowell (2004),
issue in terms of a bank account model, for example, used naive coders’ observa
these couples would not have any positive tional data based on conflict interactions to
resources to draw on in terms of expressing assess four factors of emotional expression:
their feelings about their relationships. hostility, empathy, affection, and distress.
In recent years, there has been an The researchers found that men who were
increased interest in comparing married part of satisfied couples expressed more
couples with different histories of cohabi empathy and personal distress, such as sad
tation. In one study, Kline et al. (2004) ness and anxiety, less hostility, and margin
assessed the association between the timing ally more affection than did less satisfied
of cohabitation and couple interaction men. Women in satisfied couples expressed
before and after marriage. Using self-reports greater empathy in their marital interactions,
21-Manusov.qxd 6/30/2006 6:11 PM Page 410
and marginally more affection, than did less Holtzworth-Munroe, Smutzler, and Stuart
satisfied women. These results were similar (1998) have also suggested that withdrawal
whether the marriage qualities were assessed can lead to ongoing power struggles and
using spouse self-report or an interview- “a pattern of coercive efforts can gradually
based assessment. The researchers also develop, creating a rigid pattern of negative
explored the association between emotional polarized interaction” (p. 72) (see also
expression and relationship dissolution Gottman et al., 2002).
(within 5 years). Using the four emotion Researchers have pointed to a gendered
scales scores in a logistic regression, the pattern in the use of withdrawal in mar
researchers were able to identify correctly riage, particularly with regard to the
85% of those whose relationships had demand-withdraw pattern of conflict inter
remained intact, and overall relationship action. Females are more likely to be in the
status at follow-up was predicted with role of demander and males in the role of
83% accuracy using the women’s emotional withdrawer, although whose issue is being
expression and 81% accuracy using the discussed is also important (Christensen
men’s emotional expression. & Heavey, 1990). For example, Noller,
The encoding of nonverbal behavior also Feeney, Roberts, and Christensen (2005)
involves the expression of both power and reported an observational study of the
intimacy. Burgoon and Dillman (1995), nonverbal concomitants of withdrawal in
for example, have argued that knowing married couples engaged in conflict interac
who wields power and how that power is tions. They were interested in whether the
expressed through nonverbal communica nonverbal behavior related to withdrawal
tion is central to understanding the dynam would be affected by gender and whose
ics of any relationship (see also Burgoon issue was being discussed. They found that
& Dunbar, this volume). Nonverbal behav the clearest pattern of nonverbal behavior
iors, such as touch, gaze, body movements, related to withdrawal was for husbands
and spacing, can be used to express power during discussion of their wives’ issues.
and control (Edinger & Patterson, 1983; Husbands’ withdrawal was indicated by a
Guerrero, Andersen, & Afifi, 2001). A lack of gaze and open gestures and by head
study comparing violent couples with dis down and head turn. These behaviors
tressed nonviolent and nondistressed nonvi may reflect avoidance and a lack of imme
olent couples found that distressed couples diacy, with husbands seeming to avoid
engaged in lower levels of mutual gaze any eye contact with their wives. Wives’
than did other couples, although the vio withdrawal was indicated by a lack of
lent couples seemed to be maintaining gaze open or neutral gestures and head down.
with different affect (e.g., the angry stare; Interestingly, and as would be expected,
Gottman, Driver, Yashimoto, & Rushe, highly satisfied spouses tended to be more
2002). Violent couples also rated their open and less withdrawn in their conflict
interaction less positively than did either of interactions with partners than did spouses
the nonviolent groups. low in satisfaction.
One behavior generally seen as problem
atic in couple relationships is withdrawal
from conflict. Because the likely conse SUMMARY
quences of withdrawal from conflict are
that issues are not resolved and resentment Overall, nonverbal communication is
may increase, withdrawal is generally seen as important for the expression of love, for the
having a negative impact on relationships. expression of support, and to indicate the
21-Manusov.qxd 6/30/2006 6:11 PM Page 411
need for support. Individuals differ in their involvement desired are particularly likely
willingness to express their feelings, with to be problematic, especially if the female is
women often not prepared to express their signaling interest in “getting to know you,”
anger in romantic relationships. Gender and the male decodes that message as inter
also affects the extent to which individuals est in sexual activity.
are able to express support to their partners Studies of nonverbal courtship behavior,
and express that support when it is needed however, have generally failed to find a
most. In addition, researchers have found specific list of behaviors that could be con
a pattern of men’s nonverbal withdrawal sidered as signs of interest in pursuing a rela
when issues raised by their wives are being tionship. Moore (1985), for example,
discussed. There is also evidence of effects catalogued 52 behaviors that occurred in a
related to the type and quality of relation singles bar; however, because of the large
ships including the following: (1) negativity amount of variability between participants,
is particularly marked in low-quality rela he was unable to identify specific behaviors
tionships, (2) empathy and lack of hostility that could be classified as courtship behav
are especially noticeable in high-quality iors. Grammer, Honda, Juette, and Schmitt
relationships, and (3) nonverbal behaviors (1999), in their study of initial interactions
indicating withdrawal are less evident in the between opposite-sex strangers in two cul
conflict interactions of satisfied couples. tures (Germany and Japan), were unable to
identify behaviors that could be consistently
classified as courtship behaviors. They did
♦ Nonverbal Behavior find, however, that when female interest was
high, mutual gaze was initiated and partners
and Courtship
made a similar number of comments. They
also found that women were more nonver
The research on decoding and encoding just bally involved and open in their interactions
reviewed has centered largely on ongoing than were men during initial conversations.
relationships, such as marriage and friend Argyle (1988) also identified a number of
ship. In addition to nonverbal cues in estab courtship behaviors including high levels
lished relationships, courtship, or the of mutual gaze, touch, movement, smiling,
potential establishment of romantic rela erect and open posture, and proximity.
tionships, is also an area where nonverbal Differences between males and females
behavior is critical. Expressions of interest have been found, at least within hetero
in specific members of the opposite sex sexual courtship displays. Analysis of
are often nonverbal (Renninger, Wade, & courtship interactions by Grammer et al.
Grammer, 2004). In addition, although (2000), for example, showed that unless
males tend to be seen as the initiators of males received nonverbal indicators of
courtship, they may seek nonverbal signs interest, such as eye contact, they were less
that females are interested before making likely to approach females. But females may
any approaches (Grammer, Kruck, Juette, rely on males’ signals in deciding whether
& Fink, 2000). Because nonverbal signals to encourage the approach of a male. In
of interest are likely to be somewhat support of this, Renninger et al. (2004)
ambiguous, there is a strong possibility that observed the nonverbal behaviors of males
misunderstandings can occur, both in terms in a bar and found that those who made
of the level of interest expressed and in contact with females successfully tended to
terms of the desired level of involvement. engage in more glancing behaviors, more
Misunderstandings about the levels of space-maximization movements (generally
21-Manusov.qxd 6/30/2006 6:11 PM Page 412
Mikulnicer (2005) note, anxious individuals difficult for them to believe that their
are likely to have difficulty decoding nega partner cares. A similar filter can make it
tive messages that arouse unwanted attach difficult for individuals to express their
ment needs and positive messages that feelings. Specifically, “avoidant” individuals
promote intimacy and closeness. were less expressive in interactions with their
In studying the association between partners (Fraley & Shaver, 2000; Guerrero,
attachment security and the decoding of 1996; Le Poire, Sheppard, & Duggan, 1999;
nonverbal messages by young married cou Tucker & Anders, 1998), perhaps because
ples (Noller & Feeney, 1994), my colleague of their lack of trust in others. Alternatively,
and I found that decoding accuracy was expressivity may be perceived as involving
affected by both gender and attachment the type of intimacy likely to threaten their
security. Husbands’ attachment anxiety autonomy and independence.
was associated with low levels of accuracy Tucker and Anders also assessed the
in decoding wives’ messages, irrespective encoding of nonverbal behavior in dating
of whether the messages were positive, neu couples and related expressions of nonver
tral, or negative, suggesting a general bal closeness to attachment style. They
decoding deficit related to anxiety. For found that individuals with a secure attach
wives, however, anxiety did not affect ment style were rated as less tense, as more
decoding of husbands’ messages, but avoid nonverbally expressive, and as seeming to
ance was associated with low levels of enjoy the conversation more than other
accuracy in decoding neutral and negative individuals. Their interactions with their
messages only. Thus, highly anxious partners tended to involve more laughing,
husbands were less accurate decoders touching, gazing, and smiling than was true
overall, and wives who were high in com for insecure individuals. In contrast, the
fort with closeness tended to be more accu interactions of “preoccupied” individuals
rate for neutral and negative messages. were rated as less nonverbally expres
Also, attachment security after a year of sive, and these individuals were rated as
marriage predicted decoding accuracy after experiencing less enjoyment of the conversa
almost 2 years, following the same pattern tion. Both avoidants and preoccupieds
as reported above for husbands and wives. tended to smile and gaze less than secures. It
Attachment security or insecurity is also is interesting to note, however, that preoccu
likely to influence people’s willingness to pieds were not seen as engaging in more
express their feelings to a partner. The clingy behaviors, as may have been expected.
behaviors most likely to be affected are Noller et al. (2005) report a study of
those related to expressing discomfort and attachment and emotional reactions in a
distress and to seeking support and comfort. partner-distant interaction (where a partner
Moreover, because of problems with decod or a confederate was asked to act distant
ing, people may assume incorrectly that with his or her partner). For females, com
their partner does not care about their dis fort with closeness was related to low levels
tress. Thus, “insecures” (i.e., people catego of negative affect (both hostile and worried
rized as having an insecure form of affects), and anxiety over abandonment
attachment) may struggle in their close rela was related to high levels of avoidance
tionships, not because their partner does not (turning the head or body away from the
care about their distress, but because the partner or resisting contact) and low levels
filter of their attachment insecurity distorts of touch (hugging, hand holding, etc.). For
their perception in such a way that it is males, comfort with closeness predicted
21-Manusov.qxd 6/30/2006 6:11 PM Page 414
high levels of touch and low levels of Manusov (1990) laid out a set of princi
avoidance, and anxiety over abandon ples detailing the ways in which attribution
ment was unrelated to nonverbal behavior. theories applied to the meanings people give
Highly anxious males did, however, engage to nonverbal cues. Following work by other
in negative verbal behavior in response to relational scholars who used attribution the
their partner’s distancing. ories to predict couples’ sense-making for
Overall, attachment security seems to act other behaviors such as conflict (e.g.,
as a filter and affect both the encoding and Fincham & Bradbury, 1992), she argued
the decoding of nonverbal behavior in close that couples make attributions to explain
relationships. Highly anxious individuals, why the partner acted as he or she did (i.e.,
especially males, seem to have a general sat closer, sighed, avoided eye contact).
decoding deficit, to be less nonverbally Furthermore, these attributions are linked to
expressive than other partners, and to marital satisfaction, such that more satisfied
engage in more avoidance. Thus, highly anx couples interpret the same behaviors differ
ious individuals are likely to have difficulty ently than less satisfied couples.
in maintaining close relationships. But it is In a follow-up study, Manusov et al.
important to keep in mind that attachment (1997) again looked to see if those in satis
theory is not the only theory relevant to fying relationships make more “spouse
nonverbal behavior in close relationships. enhancing attributions” and those in
Attributions for nonverbal behavior are also unhappy relationships make more “dis
important, particularly for understanding tress-maintaining attributions,” as tends to
where communication differences may arise. occur for other marital events (Fincham,
Beach, & Baucom, 1987; Fincham &
Bradbury, 1992). Spouse-enhancing attri
♦ Attributions for butions tend to involve internal and stable
attributions for positive behaviors and
Nonverbal Behavior
external and unstable attributions for nega
tive behavior. Distress-maintaining attribu
Attributions are the explanations we con tions, on the other hand, tend to involve
struct to explain our own and others’ behav external and unstable attributions for
ior. Because of the ambiguity of nonverbal positive behavior and internal and stable
cues noted earlier, it is particularly important attributions for negative behavior.
in the context of close relationships to under As with previous research, Manusov
stand the attributions that those in intimate et al. (1997) found that, in general, couples
relationships make for one another’s non were more likely to notice (and hence
verbal behavior (Abbey, 1982; Manusov, report) negative behavior than positive
Floyd, & Kerssen-Griep, 1997). In support behavior, although relationship satisfaction
of this proposition, Solomon (2001) notes tended to increase the likelihood of noticing
that the interplay among affect, attributions, positive behavior, particularly for males. In
and communication is particularly impor addition, those in satisfying relationships
tant in the context of close relationships. In were more likely to adopt a spouse-enhanc
this section, I will focus on Manusov’s work, ing attribution pattern by attributing simi
which has explored (1) the kinds of attri lar causes to their own and their spouse’s
butions partners in romantic relationships behavior. In contrast, those who were more
make for partners’ nonverbal behaviors and dissatisfied were more likely to attribute
(2) the effect of the attribution on the attrib negative causes to their spouse’s behavior
utors’ own nonverbal response. than to their own behavior. In spite of this
21-Manusov.qxd 6/30/2006 6:11 PM Page 415
Burgoon, J. K., & Dillman, L. (1995). Gender, Fraley, R. C., & Shaver, P. R. (2000). Adult
immediacy, and nonverbal communication. romantic attachment: Theoretical develop
In P. J. Kalbfleisch & M. J. Cody (Eds.), ments, emerging controversies, and unan
Gender, power and communication in swered questions. Review of General
human relationships (pp. 63–81). Hillsdale, Psychology, 4, 132–154.
NJ: Erlbaum. Fridlund, A. J. (1994). Human facial expression:
Christensen, A., & Heavey, C. L. (1990). An evolutionary view. San Diego, CA:
Gender and social structure in the demand/ Academic Press.
withdraw pattern of marital conflict. Friedman, H. S. (2001). Paradoxes of nonverbal
Journal of Personality and Social Psy detection, expression and responding:
chology, 59, 73–81. Points to ponder. In J. A. Hall & F. J.
Cohan, C., & Kleinbaum, S. (2002). Toward a Bernieri (Eds.), Interpersonal sensitivity:
greater understanding of the cohabitation Theory and measurement (pp. 351–362).
effect: Premarital cohabitation and subse Mahwah, NJ: Erlbaum.
quent marital stability. Journal of Marriage Gaelick, L., Bodenhausen, G. V., & Wyer, R.
and the Family, 64, 180–192. (1985). Emotional communication in close
Coyne, J. C., & Smith, D. A. F. (1994). Couples relationships. Journal of Personality and
coping with a myocardial infarction: Contex Social Psychology, 49, 1246–1265.
tual perspective on patient self-efficacy. Gardner, W. L., Jefferis, V., Knowles, M. L., &
Journal of Family Psychology, 8, 43–54. Pickett, C. L. (2003). Loneliness and selec
Downey, G., & Feldman, S. (1996). Implica tive memory: A general bias for remember
tions of rejection sensitivity for intimate ing social events. Unpublished manuscript.
relationships. Journal of Personality and Gardner, W. L., Pickett, C. L., & Brewer, M. B.
Social Psychology, 70, 1327–1343. (2000). Social exclusion and selective mem
Downey, G., Freitas, A. L., Michaelis, B., & ory: How the need to belong influences
Khouri, H. (1998). The self-fulfilling memory for social events. Personality and
prophecy in close relationships: Do rejec Social Psychology Bulletin, 26, 486–496.
tion-sensitive women get rejected by Gonzaga, G. C., Keltner, D., Londahl, E. A., &
romantic partners? Journal of Personality Smith, M. D. (2001). Love and the commit
and Social Psychology, 75, 545–560. ment problems in romantic relations and
Edinger, J. A., & Patterson, M. L. (1983). friendship. Journal of Personality and
Nonverbal involvement and social control. Social Psychology, 81, 247–262.
Psychological Bulletin, 93, 30–56. Gottman, J. M. (1994). What predicts divorce:
Fincham, F. D., Beach, S. R. H., & Baucom, The relationship between marital processes
D. H. (1987). Attribution processing in and marital outcomes. Hillsdale, NJ:
distressed and nondistressed couples: 4. Erlbaum.
Self-partner attribution differences. Journal Gottman, J. M., Driver, J., Yoshimoto, D., &
of Personality and Social Psychology, 52, Rushe, R. (2002). Approaches to the study
739–748. of power in violent and nonviolent mar
Fincham, F. D., & Bradbury, T. N. (1992). riages, and in gay male and lesbian cohabit
Assessing attributions in marriage: The ing relationships. In P. Noller & J. A.
relationship attribution measure. Journal Feeney (Eds.), Understanding marriage:
of Personality and Social Psychology, 62, Developments in the study of couple
457–468. interaction (pp. 323–347). New York:
Fleming, J. H., Darley, J. M., Hilton, J. L., & Cambridge University Press.
Kojetin, B. A. (1990). Multiple audience Gottman, J. M., Markman, H. J., & Notarius,
problem: A strategic communication per C. I. (1977). The topography of marital
ception on social perception. Journal of conflict: A sequential analysis of verbal and
Personality and Social Psychology, 58, nonverbal behavior. Journal of Marriage
593–609. and the Family, 39, 461–477.
21-Manusov.qxd 6/30/2006 6:11 PM Page 418
Grammer, K., Honda, M., Juette, A., & Schmitt, In A. Booth & A. Crouter (Eds.), Just living
A. (1999). Fuzziness of nonverbal court together: Implications of cohabitation
ship communication unblurred by motion for families, children and social policy
energy detection. Journal of Personality and (pp. 3–32). Mahwah, NJ: Erlbaum.
Social Psychology, 77, 487–508. King, L. A., & Emmons, R. A. (1990). Conflict
Grammer, K., Kruck, K., Juette, A., & Fink, B. over emotional expression: Psychologi
(2000). Nonverbal behavior as courtship cal and physical correlates. Journal of
signals: The role of control and choice in Personality and Social Psychology, 58,
selecting partners. Evolution and Human 864–877.
Behavior, 21, 371–390. Kline, G. H., Stanley, S. M., Markman, H. J.,
Greer, A. E., & Buss, D. M. (1994). Tactics Olmos-Gallo, P. A., St Peters, M., Whitton,
for promoting sexual encounters. Journal of S. W., et al. (2004). Timing is everything:
Sex Research, 31, 185–201. Pre-engagement cohabitation and increased
Guerrero, L. K. (1996). Attachment-style differ risk for poor marital outcomes. Journal of
ences in intimacy and involvement: A test of Family Psychology, 18, 311–318.
the four-category model. Communication Le Poire, B. A., Sheppard, C., & Duggan, A.
Monographs, 63, 269–292. (1999). Nonverbal involvement, expressive
Guerrero, L. K. (1997). Nonverbal involvement ness, and pleasantness as predicted by
across interactions with same-sex friends, parental and partner attachment style.
opposite-sex friends and romantic partners: Communication Monographs, 66, 293–311.
Consistency or change? Journal of Social Losoya, S. H., & Eisenberg, N. (2001). Affective
and Personal Relationships, 14, 31–58. empathy. In J. A. Hall & F. J. Bernieri
Guerrero, L. K., Andersen, P. A., & Afifi, W. A. (Eds.), Interpersonal sensitivity: Theory and
(2001). Close encounters: Communicat measurement (pp. 21–43). Mahwah, NJ:
ing in relationships. Mountain View, CA: Erlbaum.
Mayfield. MacGeorge, E. L. (2003). Gender differences
Heaton, T. B. (2002). Factors contributing in attributions and emotions in helping
to increasing marital stability in the U. S. contexts. Sex Roles, 48, 175–182.
Journal of Family Issues, 23, 392–409. Manusov, V. (1990). An application of attribu
Henningsen, D. D. (2004). Flirting with mean tion principles to nonverbal behavior in
ing: An examination of miscommunication romantic dyads. Communication Mono
in flirting interactions. Sex Roles, 50, graphs, 57, 104–118.
481–489. Manusov, V. (2002). Thought and action:
Holtzworth-Munroe, A., Smutzler, N., & Stuart, Connecting attributions to behaviors in
G. L. (1998). Demand and withdraw com married couples’ interactions. In P. Noller
munication among couples experiencing & J. A. Feeney (Eds.), Understanding mar
husband violence. Journal of Consulting riage: Developments in the study of marital
and Clinical Psychology, 66, 731–743. interaction (pp. 14–31). New York:
Ickes, W., & Simpson, J. A. (1997). Managing Cambridge University Press.
empathic accuracy in close relationships. Manusov, V., Floyd, K., & Kerssen-Griep, J.
In W. Ickes (Ed.), Empathic accuracy (1997). Yours, mine, and ours: Mutual attri
(pp. 218–250). New York: Guilford Press. butions for nonverbal behavior in couples’
Ickes, W., & Simpson, J. A. (2001). Moti interactions. Communication Research, 24,
vational aspects of empathic accuracy. In 234–260.
G. J. O. Fletcher & M. Clark (Eds.), The Mikulincer, M., & Shaver, P. R. (2003). The
Blackwell handbook of social psychology: attachment behavioral system in adulthood:
Interpersonal processes (pp. 229–249). Activation, psychodynamics, and interper
Oxford, UK: Blackwell. sonal processes. In M. P. Zanna (Ed.),
Kieran, K. (2002). Cohabitation in Western Advances in experimental social psychology
Europe: Trends, issues and implications. (Vol. 35). New York: Academic Press.
21-Manusov.qxd 6/30/2006 6:11 PM Page 419
Miller, P. A., Eisenberg, N., Fabes, R. A., & tion (pp. 11–43). New York: Cambridge
Shell, R. (1996). Relations of moral rea University Press.
soning and vicarious emotion to young Noller, P., Feeney, J. A., Roberts, N., &
children’s prosocial behavior towards peers Christensen, A. (2005). Withdrawal in cou
and adults. Developmental Psychology, ple interactions: Exploring the causes and
32, 210–219. consequences. In R. E. Riggio & R. S.
Mongrain, M., & Vettese, L. C. (2003). Conflict Feldman (Eds.), Applications of nonverbal
over emotional expression: Implications communication (pp. 195–213). Mahwah,
for interpersonal communication. Person NJ: Erlbaum.
ality and Social Psychology Bulletin, 29, Noller, P., & Ruzzene, M. (1991). Com
545–555. munication in marriage: The influence of
Moore, M. (1985). Nonverbal courtship pat affect and cognition. In G. J. O. Fletcher &
terns in women: Context and consequences. F. D. Fincham (Eds.), Cognition in close
Ethology and Sociobiology, 6, 237–247. relationships (pp. 203–233). Hillsdale, NJ:
Neff, L. A., & Karney, B. R. (2005). Gender dif Erlbaum.
ferences in social support: A question of skill Nowicki, S., & Duke, M. P. (1994). Individual
or responsiveness. Journal of Personality differences in the nonverbal communica
and Social Psychology, 88, 79–90. tion of affect: The Diagnostic Analysis of
Noller, P. (1984). Nonverbal communication Nonverbal Accuracy scale. Journal of
and marital interaction. Oxford, UK: Nonverbal Behavior, 18, 9–35.
Pergamon. Nowicki, S., & Duke, M. P. (2001). Nonverbal
Noller, P. (1995). Parent-adolescent relation receptivity: The Diagnostic Analysis of
ships. In M. A. Fitzpatrick & A. L. Nonverbal Accuracy (DANVA). In J. A.
Vangelisti (Eds.), Explaining family interac Hall & F. J. Bernieri (Eds.), Interpersonal
tions (pp. 77–111). Thousand Oaks, CA: sensitivity: Theory and measurement
Sage. (pp. 183–198). Mahwah, NJ: Erlbaum.
Noller, P. (2001). Using standard content Pasch, L. A, & Bradbury, T. N. (1998). Social
methodology to assess nonverbal sensitivity support, conflict, and the development of
in dyads. In J. A. Hall & F. J. Bernieri marital dysfunction. Journal of Consulting
(Eds.), Interpersonal sensitivity: Theory and and Clinical Psychology, 66, 219–230.
measurement (pp. 243–264). Mahwah, NJ: Pickett, C. L., Gardner, W. L., & Knowles,
Erlbaum. M. (2004). Getting a cue: The need to
Noller, P. (2005a). Standard content methodol belong and enhanced sensitivity to social
ogy: Controlling the verbal channel. In cues. Personality and Social Psychology
V. Manusov (Ed.), The sourcebook of non Bulletin, 30, 1095–1107.
verbal measures: Going beyond words Renninger, L. A., Wade, J. T., & Grammer, K.
(pp. 417–430). Mahwah, NJ: Erlbaum. (2004). Getting that female glance: Patterns
Noller, P. (2005b). Attachment insecurity as a fil and consequences of male nonverbal behav
ter in the encoding and decoding of nonver ior in courtship contexts. Evolution and
bal behavior in close relationships. Journal Human Behavior, 25, 416–431.
of Nonverbal Behavior, 29, 171–176. Schachner, D. A., Shaver, P. R., & Mikulincer,
Noller, P., & Feeney, J. A. (1994). Relationship M. (2005). Patterns of nonverbal behavior
satisfaction, attachment, and nonverbal and sensitivity in the context of attach
accuracy in early marriage. Journal of ment relationships. Journal of Nonverbal
Nonverbal Behavior, 18, 199–221. Behavior, 29, 141–169.
Noller, P., & Feeney, J. A. (1998). Communica Simpson, J. A., Gangestad, S. W., & Biek, M.
tion in early marriage: Responses to con (1993). Personality and nonverbal social
flict, nonverbal accuracy and conversational behavior: An ethological perspective of
patterns. In T. N. Bradbury (Ed.), The relationship initiation. Journal of Experi
developmental course of marital dysfunc mental Social Psychology, 29, 434–461.
21-Manusov.qxd 6/30/2006 6:11 PM Page 420
Simpson, J. A., Ickes, W., & Blackstone, T. Trees, A. R. (2000). Nonverbal communication
(1995). When the head protects the heart: and the support process: Interactional
Empathic accuracy in dating relation sensitivity in interactions between mothers
ships. Journal of Personality and Social and young adult children. Communication
Psychology, 69, 629–641. Monographs, 67, 239–261.
Simpson, J. A., Ickes, W., & Grich, J. (1999). Tucker, J. S., & Anders, S. L. (1998). Adult
When accuracy hurts: Reactions of anxious, attachment style and nonverbal closeness
uncertain individuals to a relationship- in dating couples. Journal of Nonverbal
threatening situation. Journal of Personality Behavior, 22, 109–124.
and Social Psychology, 76, 754–769. Waldinger, R. J., Schulz, M. S., Hauser, S. T.,
Simpson, J. A., Orina, M. M., & Ickes, W. (2003). Allen, J. P., & Crowell, J. A. (2004).
When accuracy hurts, and when it helps: Reading others’ emotions: The role of intu
A test of the empathic accuracy model in itive judgments in predicting marital satis
marital interactions. Journal of Personality faction, quality and stability. Journal of
and Social Psychology, 85, 881–893. Family Psychology, 18, 58–71.
Solomon, D. H. (2001). Affect, attribution, Wood, J. T. (1995). Relational communication:
and communication: Uniting interaction Continuity and change in personal relation
episodes and global relationship judgments. ships. Belmont, CA: Wadsworth.
In V. Manusov & J. H. Harvey (Eds.), Zahn-Waxler, C., Radke-Yarrow, M., Wagner,
Attribution, communication behavior, and E., & Chapman, M. (1992). Development
close relationships (pp. 79–90). Cambridge, of concern for others. Developmental
UK: Cambridge University Press. Psychology, 28, 126–136.
Sternglanz, R. W., & DePaulo, B. M. (2004). Zahn-Waxler, C., Robinson, J., & Emde, R. N.
Reading nonverbal cues to emotions: The (1992). The development of empathy in
advantages and liabilities of relationship twins. Developmental Psychology, 28,
closeness. Journal of Nonverbal Behavior, 1038–1047.
28, 245–266. Zebrowitz, L. A. (2001). Groping for the ele
Stinson, L., & Ickes, W. (1992). Empathic accu phant of nonverbal sensitivity. In J. A. Hall
racy in the interactions of male friends ver & F. J. Bernieri (Eds.), Interpersonal
sus male strangers. Journal of Personality sensitivity: Theory and measurement
and Social Psychology, 62, 787–797. (pp. 333–350). Mahwah, NJ: Erlbaum.
22-Manusov.qxd 6/30/2006 5:16 PM Page 421
22
NONVERBAL COMMUNICATION
IN INSTRUCTIONAL CONTEXTS
� James C. McCroskey and
Virginia P. Richmond
University of West Virginia
� Linda L. McCroskey
California State University, Long Beach
◆ 421
22-Manusov.qxd 6/30/2006 5:16 PM Page 422
well thought out in advance; others are advanced what he called the principle of
recognized as “choices” only long after they immediacy. He suggested that “people are
were actually made. Instructional/nonver drawn toward persons and things they like,
bal researchers have made four important evaluate highly, and prefer; they avoid or
choices: (1) to focus on but reframe the focus move away from things they dislike, evalu
of the “immediacy principle,” (2) to focus on ate negatively, or do not prefer” (p. 1). This
affective outcomes, (3) to focus on short- principle advances immediacy as a marker
term outcomes, and (4) to focus on the forest for how people feel about other people,
rather than the trees. Within this discussion which can be considered a psychological
we assess the implications of these choices perspective. Even though many credit
for past, current, and future research. Mehrabian’s writings as the foundation for
their work, communication scholars have
moved away from this psychological orien
REFRAMING THE tation and assumed that the causal path was
IMMEDIACY PRINCIPLE the reverse of what Mehrabian assumed.
He argued that orientation toward another
Immediacy is defined as the degree of per produces one’s immediacy choices, whereas
ceived physical or psychological closeness communication scholars suggest that
between two (or more) people (Mehrabian, one’s immediacy choices produce others’
1971; see Andersen, Guerrero, & Jones, this orientations.
volume). Immediacy behaviors (verbal and Richmond and McCroskey (2000b)
nonverbal) are those that result in people formalized this “production” approach
perceiving others to be closer or more distant recently in their principle of immediate
interactionally. Whereas Mehrabian (1971) communication. As they put it, “The more
suggested that there were both verbal and communicators employ immediate behav
nonverbal elements related to immediacy, iors, the more others will like, evaluate
communication research has dealt primar highly, and prefer such communicators;
ily with nonverbal immediacy. This may be and the less communicators employ imme
in part because development of measures of diate behaviors, the more others will dis
verbal immediacy has proven to be prob like, evaluate negatively, and reject such
lematic. Although there is no extant valid communicators” (p. 212). Immediate or
measure of verbal immediacy (for a critique nonimmediate communication behaviors
of existing measures, see Robinson & are seen as causes of communication out
Richmond, 1995), research reported by comes, reflecting what can be called a com
Mottet and Richmond (1998) suggests munication orientation. This view is
there are 10 verbal strategies (approach consistent with a larger perspective of com
messages) that are likely to increase percep munication researchers, particularly those
tions of immediacy, and there are 8 verbal interested in social influence: that commu
strategies (avoidance messages) that are nicative messages produce outcomes (e.g.,
likely to decrease perceptions of imme McCroskey, 1988).
diacy. All of these strategies are likely Whereas these two conceptualizations of
to be employed simultaneously with com immediacy may appear to be the opposite of
patible nonverbal behaviors to produce one another, it is not asserted (at least in
increased or decreased perceptions of published papers) that one is correct and the
immediacy. other is incorrect. Research in Psychology,
To help make sense of the influence that for example, has provided strong support for
immediacy may have, Mehrabian (1971) the principle of immediacy (e.g., Mehrabian,
22-Manusov.qxd 6/30/2006 5:16 PM Page 424
1971), and research in several areas of Whereas considerable research has exam
Communication has provided strong ined both cognitive and affective learning
support for the principle of immediate outcomes, that which has centered on non
communication (e.g., Richmond, Lane, & verbal communication generally has focused
McCroskey, 2006). Indeed, organizational most attention on affective learning. This
research results suggest that, rather than focus has been chosen for two reasons: (1)
being at odds with one another, psycho the belief that verbal communication has
logical and communicative approaches to its largest impact on cognitive learning and
immediacy may well coexist and together nonverbal communication has its largest
produce a cyclical process leading to both impact on affective learning, and (2)
higher or lower immediacy and higher or measuring cognitive learning has been con
lower liking between individuals over time fronted with the numerous challenges that
(Richmond & McCroskey, 2000a). Such have not yet been overcome to everyone’s
research suggests that the immediacy behav satisfaction. These two areas are reviewed
iors of either a supervisor or a subordinate here.
tend to produce reciprocal immediacy on Although verbal and nonverbal messages
the part of the other person and are associ are both present in the vast majority of
ated with more subordinate job satis instructional environments, it has been
faction. This result is likely to transfer to argued that they perform different func
instructional contexts. tions in those environments. More specifi
cally, researchers assert that the verbal
messages stimulate primarily cognitive
FOCUS ON AFFECTIVE meanings in receivers (what the students
LEARNING OUTCOMES learn about the subject matter), whereas
the nonverbal messages stimulate affective
In addition to a focus on immediacy, meanings in receivers (i.e., the feelings and
instructional communication researchers attitudes toward the content as well as feel
have been concerned with learning out ings and attitudes toward the teacher)
comes associated with communication (McCroskey, Richmond, & McCroskey,
behaviors, including immediacy. The tradi 2006; Richmond & McCroskey, 2000b).
tional view of learning advanced in From this perspective, subject competence
Educational Psychology posits three distinct of the instructor is said to be the critical ele
types of learning: cognitive, psychomotor, ment in cognitive learning, whereas behav
and affective (Bloom, 1956; Krathwohl, ior, particularly nonverbal communication,
Bloom, & Masia, 1964). From the earliest produces a relational impact on the student
reported instructional research involving attitudes and feelings. Although this view
nonverbal communication (Andersen, 1978), cannot be said to be scientifically proven,
instructional communication researchers virtually all the available research results
have embraced this conceptualization. are consistent with it (see Mottet & Beebe,
Most of this work has sought to determine 2006).
relationships between various communica Besides a theoretical push toward affec
tion variables and cognitive and/or affective tive learning, scholars may focus less on cog
learning (e.g., Richmond et al., 2006). Little nitive learning for methodological reasons.
work has been directed toward psychomo Specifically, the development of measures,
tor learning, but some studies have exam or even approaches to the measurement, of
ined affect toward behaviors learned cognitive learning has frustrated instructional
instead (e.g., Andersen, 1978). communication researchers consistently. No
22-Manusov.qxd 6/30/2006 5:16 PM Page 425
primary focus for scholars interested in the communication, and in textbooks con
role of nonverbal communication in learn cerned with nonverbal communication,
ing environments. As noted, this decision is communication scholars have chosen to
reflected in much of the current research take two different approaches, which we
being published in this area. refer to as the “trees” or the “forest.” The
A theoretical view advanced by some “tree” approach (also referred to as the
instructional communication researchers “categorical” approach; Knapp, 1972;
(including the present authors) is that long- Malandro & Barker, 1983) separates the
term learning is the most important form of elements into categories and centers on how
learning (McCroskey, 1998). This view each of these elements affects communica
holds that schools are created primarily to tion outcomes across contexts. The “for
acculturate young people and prepare them est” approach, also referred to as the
for success in the surrounding culture. “functional” approach (Burgoon & Saine,
Whereas schools are seen as teaching the 1978), centers on how multiple elements of
content of many disciplines, as well as the nonverbal behavior influence communica
values and attitudes of the people in that tion outcomes collectively in specific con
culture, their primary function is, arguably, texts. Most researchers have chosen one or
to teach the students the differences between the other of these to organize or frame their
right and wrong (as seen by their culture), work on nonverbal communication. Some
and prepare them with lifelong learning textbook authors also have followed this
skills. Much of this socialization deals with pattern, whereas others have chosen to
matters of affect, such as developing posi include both approaches in their works
tive affect for learning and the norms and (e.g., Hickson & Stacks, 1985; Richmond
values of the culture. Given that nonverbal & McCroskey, 2004).
communication has been found to have a Instructional communication researchers,
major impact on the development of affect for the most part, have taken the “forest”
(Richmond et al., 2006), it may well be the approach, and their emphasis on nonverbal
critical element producing the desired (as immediacy is representative of this choice.
well that which is not desired) outcomes of They recognize that single nonverbal
educational institutions. The long-term behaviors rarely manifest alone (e.g.,
impact of nonverbal communication, there Hickson & Stacks, 1985) and that, typi
fore, should be a major focus of future cally, there are many nonverbal messages
instructional and nonverbal communica transmitted simultaneously in instructional
tion researchers. communication by both teachers/trainers
and students. Whereas single nonverbal
behaviors may have a direct impact in
STUDYING THE TREES this context, it is most likely that outcomes
OR THE FOREST are also influenced by multiple nonver
bal behaviors interacting with one another.
In addition to choices made regarding This impact of both single nonverbal
what is to be the focus of research and at behaviors and the interaction of nonver
what point the focus is captured, scholars bal behaviors has been demonstrated
also make choices as to the perspective in experimental research (e.g., Kelley &
from which to approach the topics. Gorham, 1988).
From the beginning of formal communi Yet, because most of the nonverbal cate
cation research relating to nonverbal gories were conceptualized by nonverbal
22-Manusov.qxd 6/30/2006 5:16 PM Page 427
interest in this nonverbal behavior problem, Summary. As we have noted above, there
however. are very many different kinds of nonverbal
Many factors of the environment either behaviors that may individually affect
send constant messages to people or control instruction. Each has the potential to
the communication that exists within the enhance or impede communication in this
environment directly. These effects occur context. This concludes our discussion of
in educational environments. Considerable the specific categories of nonverbal behav
research on instructional environments has iors. Now we turn our attention to collec
been conducted, but virtually none by tions of these variables.
instructional communication researchers.
Such research has been reported in fields such
Collective Nonverbal Constructs:
as Architecture, Education, Psychology, and
The “Forest” Approach
Sociology. What little research that has been
reported by instructional communication As noted, it is rare that a single non
researchers, however, indicates that even the verbal variable will exist alone in an
arrangement of seating in a classroom can instructional environment. Whereas many
have an important impact on student com nonverbal cues function within and affect
munication behavior and student learning most instructional contexts, their collective
(e.g., McCroskey & McVetta, 1978). impact may be of much greater theoretical
The use of time, also known as chrone and practical importance. Hence, instruc
mics, can be highly communicative in the tional researchers have sought to develop or
classroom. How people react to time will adopt constructs that can explain the gen
vary greatly based on the culture in which eral effects of the combined forces of multi
they were born or in which they have been ple nonverbal behaviors (i.e., the nonverbal
enculturated. Not only do large national cul “forest”). The collective nonverbal con
tures have a system of time, but individual structs we focus on here are immediacy,
organizations have a culture as well that clarity, and sociocommunicative style.
addresses the time rules for its members. As noted, immediacy is the degree of per
Schools, at least within the general North ceived physical or psychological closeness
American culture, are examples of the most between people (Mehrabian, 1971), and it
rigid ways of dealing with time. Everything may be reflected in or a result of immediacy
within a given day usually is timed down to behaviors. Nonverbal immediacy research
the minute. One of the major concerns in has determined that it has a powerful impact
most elementary schools is teaching the on instructional outcomes. These outcomes
system of time required by its rules to young include (1) increased affinity for the teacher,
children who have not already been taught (2) increased affective learning (affect for the
the system enforced in their culture. Time content of the subject matter), (3) increased
imposes rules for communicating (or not student perceptions of their own cognitive
communicating). It determines when students learning, and (4) the granting of increased
can talk, how long they can talk, where they referent power to teachers by students. Each
can talk, and what they can talk about. of these will be discussed in turn.
Penalties that can be very severe are imposed The affinity construct was first explored
on students who do not follow the commu by interpersonal communication research
nication time rules. This is an area that ers (Bell & Daly, 1984; McCroskey &
instructional communication researchers Wheeless, 1976) who argued that develop
have generally ignored up until this time. ment of affinity is the “first function of
22-Manusov.qxd 6/30/2006 5:16 PM Page 430
communication” (McCroskey & Wheeless, associated with both actual content learned
1976, pp. 21–22). Having affinity for (Kelley & Gorham, 1988) and students’
another person essentially means liking perceptions of their own learning (e.g.,
that other person. Nonverbal immediacy is Christophel, 1990; McCroskey et al., 2004;
considered one of the main means one McCroskey, Sallinen, Fayer, Richmond,
person can use to get another person to like & Barraclough, 1996; Richmond, 1990;
him or her. This “liking” is believed to Richmond et al., 1987). Actual learning
include a number of positive affects—for (measured via a content test) and student
example, having increased interpersonal perceived learning have been found to be
attraction (social and/or task) for the other correlated substantially (r = .50; Chesebro
person, having increased respect for the & McCroskey, 2000). Collectively, the
other person, seeing the other person as research evidence suggests a strong positive
more credible (competence, caring, and/or association between teacher nonverbal
trustworthiness), being more responsive to immediacy and cognitive learning of their
the other person, and so on (McCroskey & students.
Wheeless, 1976). Instructional communica The final factor involved with immediacy
tion research has observed that teachers to be discussed here that can have a strong
employ nonverbal immediacy consciously impact on students is the teacher’s referent
to increase students’ affinity for them power. French and Raven (1959) conceptu
(McCroskey & McCroskey, 1986). It also alized five types of power. Referent power
has produced consistent results showing is one of those, the one that has been found
that teachers’ increased nonverbal immedi to produce the most positive influence on
acy is associated with similar increases in students when used by teachers (Richmond
students’ evaluations of the teacher, both in & McCroskey, 1984). Opinion leaders,
the United States and in other cultures (e.g., mentors, and others who are seen as positive
McCroskey, Richmond, Sallinen, Fayer, & models and/or people whom others wish to
Barraclough, 1995; McCroskey, Valencic, emulate are said to have referent power.
& Richmond, 2004). These individuals are seen to be both credi
As noted earlier in this chapter, the con ble and task attractive. Teachers who are
struct of affective learning was advanced in nonverbally immediate have been found to
educational psychology (Krathwohl et al., be perceived as both more credible and task
1964). It refers to students’ development of attractive than are teachers who are less
positive (or negative) affect toward the nonverbally immediate (McCroskey et al.,
subject matter taught. Many studies have 2004; Thweatt & McCroskey, 1998).
produced results indicating a strong, positive Overall, then, nonverbally immediate
relationship (in the United States and other teachers tend to produce substantially more
cultures) between the nonverbal immediacy positive instructional outcomes. Students
of teachers and the affective learning of have more affinity for immediate teachers,
their students (e.g., Christophel, 1990; they have more affective learning, they have
McCroskey, Fayer, Richmond, Sallinen, & more cognitive learning, and they grant the
Barraclough, 1996; McCroskey et al., 2004; teacher more referent power in the instruc
Richmond, 1990). Cognitive learning was tional environment.
also advanced in educational psychology Another collective construct implicating
(Bloom, 1956). Whereas the measurement nonverbal cues is clarity. Clarity has long
of cognitive learning has been controversial, been believed to be an important factor in
nonverbal immediacy has been found to be effective instruction. Considerable research
22-Manusov.qxd 6/30/2006 5:16 PM Page 431
viewed as a positive characteristic for both more positive instructional outcomes: more
females and males, however, and necessary positive student affect, more affective learn
for attaining communication competence ing, more cognitive learning, more moti
for members of both sexes. vated students, more student trust for the
Likewise, responsiveness is seen as other- teacher, and so on (e.g., Kearney &
oriented communication behavior that McCroskey, 1980; McCroskey et al., 2004;
reflects an individual’s ability to recognize Thomas, 1994; Wanzer & McCroskey,
the needs of others without being submis 1998; Wooten & McCroskey, 1996).
sive to the others. Communicators who are Those who are less assertive and/or respon
responsive are sensitive to others, consider sive are less effective. As was the case with
others’ feelings, and listen carefully to their teacher clarity, teacher immediacy is corre
communication. Bem (1974) has argued lated with teacher assertiveness and res
that this type of communication behavior is ponsiveness. Although teacher assertiveness
stereotypically characteristic of females in and teacher responsiveness have been found
the general North American culture (and not to be statistically or meaningfully corre
many other cultures). As is the case with lated with each other, they have been found
assertive communication, however, respon to be correlated consistently with nonver
sive communication is viewed as a positive bal immediacy (McCroskey et al., 2004;
characteristic of both males and females, Thomas, Richmond, & McCroskey, 1994).
one that is often necessary for attaining
communication competence for members Summary. Based on the research reported
of both sexes. up to this time, it is clear that nonverbal
Whereas both assertiveness and respon communication has a major impact on
siveness are seen as characteristic traits of teachers’ effectiveness in the classroom. In
an individual’s behavior, versatility is seen particular, nonverbal immediacy, clarity,
as an individual’s ability to be flexible and and sociocommunicative style all are impor
adaptable with regard to his or her use of tant factors in that success or failure. Recent
either or both assertiveness and responsive research indicates that immediacy and two
ness. Competent communicators, then, are components of sociocommunication style,
seen as being assertive when it is needed responsiveness and assertiveness, have a
and responsive when it is needed as well as strong genetic base (McCroskey, Heisel, &
reducing their assertiveness and/or respon Richmond, 2001); however, no research of
siveness when that would be appropriate. this type related to clarity has been reported.
In the instructional environment, for
example, a teacher who is teaching a mass
♦ Implications
lecture class might need to be more
assertive, whereas that teacher, when teach
ing an interactive class such as a seminar, The research choices reviewed in this chap
might need to be more responsive. Teachers ter, and the subsequent findings of that
who are highly assertive or highly respon research, help us understand much of what
sive, and not versatile, might be more effec nonverbal behaviors do in instructional
tive in one type of class and less effective in contexts. This research is important to
another type of class. those who study instructional communica
In support of this, research has indicated tion and nonverbal communication. It is
that, in general, teachers who are more also important to those who may use non
assertive and more responsive produce verbal cues in the instructional context. We
22-Manusov.qxd 6/30/2006 5:16 PM Page 433
close this chapter with a discussion of some constructs as predictors of the differences
recent research that focuses specifically on between student perceptions of domestic
the importance of teacher effectiveness in and foreign teachers as well as their pre
multi-cultural classrooms. dictability in predicting the outcomes
Two studies reported by Linda McCroskey within the two teacher groups separately.
(2002, 2003) have provided considerable The specific constructs she employed were
assistance in determining what nonverbal the ones discussed in the previous section:
constructs are particularly important to immediacy, clarity, and sociocommunica
the effectiveness of teachers in the instruc tive style. Student ethnocentrism was also
tional environment. In these two studies, included as a predictor of the outcome vari
McCroskey sought to determine whether, ables (affect toward instructor, instructor
and, if so, why, foreign instructors are eval evaluation, affective learning, and perceived
uated less positively than are domestic cognitive learning). This study employed a
instructors. The first study (McCroskey, more reliable and valid measure of ethno
2002) examined student ethnocentrism and centrism than the one used in the first
several student communication traits (e.g., study. The results of the research replicated
intercultural communication apprehension, the results of the first study with regard to
willingness to communicate). The student ethnocentrism. The correlations of ethno
participants were asked to respond to two centrism with all the outcome variables
of their current teachers: one a domestic were trivial (r = .04 to .11) and nonsignifi
teacher and the other a foreign teacher. cant. Ethnocentrism could not predict
Her results confirmed the expected students’ varying responses to foreign and
difference between students’ reactions to domestic instructors.
domestic and foreign teachers. The mean The students’ responses to each of the
differences between the two groups outcomes were substantially different when
were significant and large. The results also they were responding to a domestic or a
indicated that student ethnocentrism had foreign instructor. All the mean differences
a low but significant relationship with strongly favored the domestic teachers. It is
the various outcome variables (perceived important to note, however, that approxi
effectiveness of the teacher, willingness to mately 30% of the students responded
initiate communication with the teacher, more positively to their foreign teacher than
student motivation to work with the their domestic teacher. Where the teacher
teacher, affective learning, perceived cogni came from was not a consistent indicator of
tive learning). The student trait predictors student preferences. In contrast, each of the
produced correlations that were trivial communication variables served individu
(average r = .11) and nonsignificant. The ally as a significant predictor of all the out
“bottom line” in this study was that come variables. The multiple correlations
although the students perceived very large of just assertiveness and responsiveness
differences between the foreign and domes with each of the outcome variables were
tic teachers, neither ethnocentrism nor the extremely strong, ranging from .54 to .72.
student communication traits could predict The multiple correlations involving imme
a large amount of this variation between diacy and clarity predicting the outcome
domestic and foreign teachers. variables were equally strong, ranging from
The second study (McCroskey, 2003) r = .52 to .81. The results of this research
was designed to examine the predictability suggest that effective instruction is based on
of teacher nonverbal communication such teacher behaviors as are collectively
22-Manusov.qxd 6/30/2006 5:16 PM Page 434
included in the constructs of nonverbal Burgoon, J. K., & Saine, T. J. (1978). The
immediacy, clarity, and sociocommunica unspoken dialogue: An introduction to
tive style, all of which include at their base nonverbal communication. Boston, MA:
nonverbal communication. Houghton Mifflin.
Chesebro, J. L., & McCroskey, J. C. (1998). The
relationship of teacher clarity and teacher
immediacy with students’ experiences of
♦ Conclusion state receiver apprehension. Commu
nication Quarterly, 46, 446–456.
Chesebro, J. L., & McCroskey, J. C. (2000). The
Effective teaching is dependent upon
relationship between students’ reports of
“appropriate” nonverbal communication
learning and their actual recall of lecture
of teachers. In our opinion, certification of
material: A validity test. Communication
teachers without substantial instruction in Education, 49, 297–301.
nonverbal communication concepts and Chesebro, J. L., & McCroskey, J. C. (2001). The
skills would be pure folly. The success of relationship of teacher clarity and immedi
teachers at all levels depends on how they acy with student state receiver apprehen
communicate nonverbally. It may be that sion, affect, and cognitive learning.
some teachers are genetically programmed Communication Education, 50, 59–68.
to be more effective, whereas others are Christophel, D. M. (1990). The relationships
equally programmed to fail. Individuals among teacher immediacy behaviors,
considering teaching as a profession need to student motivation, and learning. Com
munication Education, 39, 323–340.
be aware that these nonverbal behaviors are
French, J. R. P., & Raven, B. (1959). The bases
expected of them, at least in the U.S. cul
of social power. In D. Cartwright (Ed.),
ture, and measure that in their decision.
Studies in social power (pp. 150–167). Ann
Similarly, administrators of teacher educa Arbor, MI: Institute for Social Research.
tion and teacher/trainer selection should Hickson, M. L., III, & Stacks, D. W. (1985).
consider these communication abilities (or Nonverbal communication: Studies and
lack of) in their decisions to admit or hire applications. Dubuque, IA: Wm. C. Brown.
people for their programs. Hines, C., Cruickshank, D., & Kennedy, J.
(1985). Teacher clarity and its relationship
to student achievement and satisfaction.
♦ References American Educational Research Journal,
22, 87–99.
Kearney, P., & McCroskey, J. C. (1980).
Andersen, J. A. (1978). The relationship Relationships among teacher communica
between teacher immediacy and teach tion style, trait and state communication
ing effectiveness. Unpublished doctoral apprehension, and teacher effectiveness.
dissertation, West Virginia University, Communication Yearbook, 4, 533–551.
Morgantown. Kelley, D. H., & Gorham, J. (1988). Effects
Bell, R., & Daly, J. A. (1984). The affinity-seek of immediacy on recall of information.
ing function of communication. Commu Communication Education, 37, 198–207.
nication Monographs, 51, 91–115. Knapp, M. L. (1972). Nonverbal communica
Bem, S. L. (1974). The measurement of psycho tion in human interaction. New York: Holt,
logical androgyny. Journal of Consulting Rinehart, and Winston.
and Clinical Psychology, 47, 155–162. Krathwohl, D. R., Bloom, B. S., & Masia, B. B.
Bloom, B. S. (1956). Taxonomy of educational (1964). Taxonomy of educational objec
objectives (Handbook I: Cognitive domain). tives (Handbook II: Affective domain).
New York: McKay. New York: McKay.
22-Manusov.qxd 6/30/2006 5:16 PM Page 435
Malandro, L. A., & Barker, L. (1983). McCroskey, J. C., Valencic, K. M., &
Nonverbal communication. Reading, MA: Richmond, V. P. (2004). Toward a general
Addison-Wesley. model of instructional communication.
McCroskey, J. C. (1998). An introduction to Communication Quarterly, 53, 197–210.
communication in the classroom (2nd ed.). McCroskey, J. C., & Wheeless, L. R. (1976).
Acton, MA: Tapestry Press. Introduction to human communication.
McCroskey, J. C., Fayer, J. M., Richmond, Boston, MA: Allyn & Bacon.
V. P., Sallinen, A., & Barraclough, R. A. McCroskey, L. L. (2002). Domestic and interna
(1996). A multi-cultural examination of the tional college instructors: An examination
relationship between nonverbal immediacy of perceived differences and their correlates.
and affective learning. Communication Journal of Intercultural Communication
Quarterly, 44, 297–307. Research, 31, 63–83.
McCroskey, J. C., Heisel, A. D., & Richmond, McCroskey, L. L. (2003). Relationships of instruc
V. P. (2001). Eysenck’s BIG THREE and tional communication styles of domestic
communication traits: Three correlational and foreign instructors with instructional
studies. Communication Monographs, 68, outcomes. Journal of Intercultural Commu
360–366. nication Research, 32, 75–96.
McCroskey, J. C., & McCroskey, L. L. (1986). Mehrabian, A. (1971). Silent messages. Belmont,
The affinity-seeking of classroom teachers. CA: Wadsworth.
Communication Research Reports, 3, Merrill, D. W., & Reid, R. H. (1981). Personal
158–167. styles and effective performance: Make
McCroskey, J. C., & McVetta, R. W. (1978). your style work for you. Radnor, PA:
Classroom seating arrangements: Instruc Chilton Book.
tional communication theory versus student Mottet, T. P., & Beebe, S. A. (2006).
preferences. Communication Education, Foundations of instructional communica
27, 99–111. tion. In T. P. Mottet, V. P. Richmond, &
McCroskey, J. C., & Richmond, V. P. (1996). J. C. McCroskey (Eds.), The handbook
Fundamentals of human communication. of instructional communication: Rhetorical
Prospect Heights, IL: Waveland. and relational perspectives (pp. 3–32).
McCroskey, J. C., Richmond, V. P., & Boston, MA: Allyn & Bacon.
McCroskey, L. L. (2006). An introduction Mottet, T., & Richmond, V. P. (1998). Verbal
to communication in the classroom: The approach and avoidance items. Communica
role of communication in teaching and tion Quarterly, 46, 25–40.
training. Boston, MA: Allyn & Bacon. Norton, R. (1983). Communicator style: Theory,
McCroskey, J. C., Richmond, V. P., Plax, T. G., applications, and measures. Beverly Hills,
& Kearney, P. (1985). Power in the class CA: Sage.
room V: Behavior alteration techniques, Plax, T. G., Kearney, P., McCroskey, J. C.,
communication training, and learning. & Richmond, V. P. (1986). Power in
Communication Education, 34, 214–226. the classroom VI: Verbal control strategies,
McCroskey, J. C., Richmond, V. P., Sallinen, A., nonverbal immediacy, and affective learn
Fayer, J. M., & Barraclough, R. A. (1995). ing. Communication Education, 35, 43–55.
A cross-cultural and multi-behavioral Richmond, V. P. (1990). Communication in
analysis of the relationship between non the classroom: Power and motivation.
verbal immediacy and teacher evaluation. Communication Education, 39, 181–195.
Communication Education, 44, 281–290. Richmond, V. P. (1997). Nonverbal communi
McCroskey, J. C. Sallinen, A., Fayer, J. M., cation in the classroom. Acton, MA:
Richmond, V. P., & Barraclough, R. A. Tapestry Press.
(1996). Nonverbal immediacy and cogni Richmond, V. P. (2002). Socio-communicative
tive learning: A cross-cultural investigation. style and orientation in instruction. In
Communication Education, 45, 200–211. J. L. Chesebro & J. C. McCroskey (Eds.),
22-Manusov.qxd 6/30/2006 5:16 PM Page 436
Communication for teachers (pp. 104–115). Richmond, V.P., & McCroskey, J.C. (2004).
Boston, MA: Allyn & Bacon. Nonverbal behavior in interpersonal rela
Richmond, V. P., Gorham, J. S., & McCroskey, tions (5th ed.). Boston, MA: Allyn & Bacon.
J. C. (1987). The relationship between Robinson, R. Y., & Richmond, V. P. (1995).
selected immediacy behaviors and cognitive Validity of the verbal immediacy scale.
learning. Communication Yearbook, 10, Communication Research Reports, 12,
574–590. 80–84.
Richmond, V. P., Lane, D. R., & McCroskey, Sidelinger, R. J., & McCroskey, J. C. (1997).
J. C. (2006). Teacher immediacy and the Communication correlates of teacher clarity
teacher–student relationship. In T. P. Mottet, in the college classroom. Communication
V. P. Richmond, & J. C. McCroskey (Eds.), Research Reports, 14, 1–10.
Handbook of instructional communication: Thomas, C. E. (1994). An analysis of teacher
Rhetorical and relational perspectives. socio-communicative style as a predictor
Boston, MA: Allyn & Bacon. of classroom communication behaviors,
Richmond, V. P., & Martin, M. M. (1998). student liking, motivation, and learning.
Socio-communicative style and socio-com Unpublished doctoral dissertation, West
municative orientation. In J. C. McCroskey, Virginia University, Morgantown.
J. A. Daly, M. M. Martin, & M. J. Beatty Thomas, C. E., Richmond, V. P., & McCroskey,
(Eds.), Communication and personality: J. C. (1994). Is immediacy anything more
Trait perspectives (pp. 133–148). Cresskill, than just being nice? The association
NJ: Hampton Press. between immediacy and socio-communica
Richmond, V. P., & McCroskey, J. C. (1984). tive style. Communication Research
Power in the classroom II: Power and Reports, 11, 107–115.
learning. Communication Education, 33, Thweatt, K. S., & McCroskey, J. C. (1998). The
125–136. impact of teacher immediacy and misbehav
Richmond, V. P., & McCroskey, J. C. (1990). iors on teacher credibility. Communication
Reliability and separation of factors on Education, 47, 348–358.
the assertiveness–responsiveness measure. Wanzer, M. B., & McCroskey, J. C. (1998).
Psychological Reports, 67, 449–450. Teacher socio-communicative style as a
Richmond, V. P., & McCroskey, J. C. (2000a). correlated of student affect toward teacher
The impact of supervisor and subordinate and course material. Communication Edu
immediacy on relational and organizational cation, 47, 43–52.
outcomes. Communication Monographs, Wooten, A. G., & McCroskey, J. C. (1996).
66, 85–95. Student trust of teacher as a function
Richmond, V. P., & McCroskey, J. C. (2000b). of socio-communicative style of teacher
Nonverbal behavior in interpersonal and socio-communicative orientation of
relations (5th ed.). Boston, MA: Allyn & student. Communication Research Reports,
Bacon. 13, 94–100.
23-Manusov.qxd 6/30/2006 7:34 PM Page 437
23
NONVERBAL COMMUNICATION
AND PHYSICIAN-PATIENT
INTERACTION
Review and New Directions
� Jeffrey D. Robinson
Rutgers University
Author’s Note: The author thanks James Dillard, Jenny Mandelbaum, Valerie
Manusov, and Richard Street for comments on earlier drafts.
◆ 437
23-Manusov.qxd 6/30/2006 7:34 PM Page 438
that physicians’ time spent reading patients’ delivering bad medical news). The analytic
medical records, which was also physicians’ “payoff” of looking at the larger interac
time spent gazing away from patients, was tion, rather than discrete variables, will be
negatively associated with patients’ post- discussed later in this chapter.
visit satisfaction and understanding of
medical information. Furthermore, Bensing
(1991) showed that physicians’ gaze toward PHYSICIANS’ HEAD NODDING
patients was positively associated with
external physician-raters’ evaluations of the When people gaze at speakers, especially
quality of participant-physicians’ psychoso when speakers are producing multi-unit
cial care. turns (e.g., when patients produce illness
One explanation of this possible con narratives or when physicians explain treat
tradiction lies in an analysis of where physi ments), gazers nod their head frequently,
cians are gazing—and what physicians are which, at a minimum, communicates atten
doing—while patients are talking. For tion (Schegloff, 1982). Hall, Irish, Roter,
example, Harrigan et al. (1985) also found Ehrlich, and Miller (1994) found that, com
that, compared with low-rapport physicians, pared with male physicians, female physi
high-rapport physicians gazed at (i.e., read) cians nod more overall and they nod more to
patients’ medical records more often when female patients. Harrigan and Rosenthal
not gazing at patients, but they were more (1983) discovered external raters’ evalua
likely to continue to gaze at patients when tions of physicians’ nodding to be positively
patients were talking. In support of this, Giron, associated with raters’ perceptions of physi
Manjon-Arce, Puerto-Barber, Sanchez-Garcia, cians’ rapport. In a later study, however,
and Gomez-Beneyto (1998) revealed that Harrigan et al. (1985) found no association
physicians’ eye contact while patients spoke between nodding and rapport. Nodding is,
was positively associated with physicians’ however, more commonly studied in associ
psychodiagnostic abilities. ation with other variables than as an isolated
In sum, physicians’ gaze toward (rather cue, and the findings when nodding is viewed
than away from) patients appears to be as part of a larger communicative function are
positively associated with patients’ giving more robust. For example, Weinberger,
of psychosocial information, which may Greene, and Mamlin (1981) found physicians’
explain the concomitant positive associa nonverbal encouragement—operationalized
tions with physicians’ psychodiagnostic in terms of nodding and gesture—was posi
abilities and patients’ positive evaluations tively associated with patients’ post-visit sat
of at least the affective-relational dimension isfaction. Duggan and Parrott (2001)
of physicians’ communication. These find showed likewise that physicians’ facial rein
ings are confounded by a lack of control forcement—operationalized in terms of nod
for what physicians and patients are doing ding and facial animation—was positively
while gazing, however, as well as where associated with patients’ self-disclosure (e.g.,
patients are gazing (e.g., toward or away about life beyond symptoms).
from physicians). A particularly salient
issue seems to be whether physicians and
patients are talking generally and, specifi PHYSICIANS’ BODY ORIENTATION
cally, what social actions are getting accom
plished through such talk (e.g., instructing Although head movement and gaze
patients to sit on the examination table vs. orientation communicate persons’ current
23-Manusov.qxd 6/30/2006 7:34 PM Page 441
engagement, the front of a person’s body body orientation and patients’ satisfaction
communicates a frame of dominant orienta with physicians. As with gaze, however,
tion: a frame of space wherein long-term and these findings may be confounded by a lack
dominant social actions are most likely of control for what physicians and patients
to be focused (Goodwin, 1981; Kendon, are doing.
1990; Schegloff, 1998). The orientation of
persons’ bodies communicates their availabil
ity or nonavailability for collaborative action. SUMMARY
When two persons bring each other into (or
remove the other from) their frame of domi A variety of individual nonverbal behav
nant orientation, they establish (or dismantle) iors have been associated with communica
a participation framework (Goodwin, 1981). tive, social, and psychological outcomes. The
This orientation appears to have a majority of these outcomes relate to the affec
number of outcomes. Larsen and Smith tive-relational (vs. medical-technical) dimen
(1981) found that the amount of time sion of communication, such as ratings of
physicians spend with their bodies oriented physicians’ dominance, rapport, and likeabil
toward patients was positively associated ity, and to psychosocial (vs. somatic) aspects
with patients’ post-visit satisfaction and of care, such as patients’ self-disclosure of
understanding. Street and Buller (1987) lifeworld events and physicians’ psychodiag
showed that physicians’ indirect body ori nostic abilities. Together, they suggest the
entation (i.e., away from patients) was pos important role that nonverbal cues may play
itively associated with patients’ perceptions in physician-patient interactions.
of physicians’ dominance. Harrigan et al. Despite the overall strength of their con
(1985) reported that physicians’ body ori clusions, the findings just reviewed can be
entation away from patients was negatively extended in a number of ways that highlight
associated with external raters’ evaluations new directions for research on nonverbal
of physicians’ rapport. In a different vein, communication. Put generally, to under
Giron et al. (1998) found physicians’ open stand the process more completely—that is,
face-to-face posture while patients spoke how these behaviors come to have the con
to be positively associated with physicians’ sequences that they do—research needs to
psychodiagnostic abilities. A range of stud be situated within a larger framework
ies has also found positive relationships for understanding communication per se.
between physicians’ proximity to and lean Whereas communication simultaneously
toward patients and outcomes such as involves multiple, mutually influential
patients’ post-visit understanding (Larsen & modalities of meaning (nonverbal, verbal,
Smith, 1981; Smith, Polis, & Hadac, 1981) artifactual) and is interactive inherently,
and external raters’ evaluations of physi much of the research done in the medical
cians’ rapport (Harrigan & Rosenthal, 1983; context focuses in isolation on one modal
Harrigan et al., 1985). ity of behavior (i.e., nonverbal) produced
Overall, this work shows that physicians’ by one participant (i.e., the physician). To
body orientation toward (and physical make its largest contribution—to help
proximity to) patients appears to be posi understand how communication in the
tively and negatively associated with ratings medical context comes to work as it does—
of physicians’ rapport and dominance, research on physician-patient communica
respectively, and this may partially explain tion needs to be situated within (i.e., needs
the positive association between physicians’ to control for) aspects of interactional
23-Manusov.qxd 6/30/2006 7:34 PM Page 442
context that have been shown to shape how conceptualizing the relationship between
nonverbal behavior is produced and under verbal and nonverbal behavior that focus
stood. The following section suggests how on their co-occurrence in social meaning.
this may be done. The first conceptualization is that verbal
behavior and nonverbal behavior consti
tute two distinct channels of communica
♦ New Directions for tion that are attended to and processed
Research Dealing With separately by receivers (e.g., Ekman &
Interactional Context Friesen, 1969). Researchers adopt this posi
tion tacitly whenever they examine phe
nomena whose functions entail both verbal
Street (2003) proposed an ecological model
and nonverbal communication (e.g., domi
of communication that recognizes that “visit
nance) yet analyze such phenomena exclu
communication” and its outcomes are orga
sively in terms of one channel, or modality,
nized by reference to organizational, politi
of meaning.
cal, media, cultural, and interpersonal
The second, alternative, conceptualiza
contexts. This section of the chapter extends
tion (which this chapter adopts) is that the
Street’s model generally, and his interper
meaning of communicative events is shaped
sonal context specifically (which includes
by, and thus depends on, the “context” in
verbal and nonverbal communication), by
which it is situated and that verbal and non
recognizing the organizing effects of interac
verbal behavior are each forms of context
tional context. That is, in addition to tradi
(Goodwin, 1995; Sanders, 1987). From this
tional conceptions of context, interaction has
perspective, the relationship between verbal
its own, independent orders of social organi
and nonverbal behavior is neither additive
zation (Goffman, 1983) that can affect both
nor multiplicative, in the sense that each
the production and the understanding of
constitutes a separate yet combinable factor
nonverbal communication. Over the past 30
of meaning. Rather, their relationship is
years, three of the most robust “interaction
holistic and metamorphic, with a multitude
orders” involve turn taking, social action,
of modalities (e.g., verbal, nonverbal, arti
and sequences of talk and action. Before
factual) working together to convey a single
addressing these issues, however, this section
meaning (for more on this, see McNeill,
begins with a discussion of the inseparability
Cassell, & McCullough, 1994).
of nonverbal and verbal behavior.
This second conceptualization shifts
analytic attention away from the function
THE INSEPARABILITY of individual nonverbal behaviors to how
OF NONVERBAL AND they achieve their social meanings in and
VERBAL BEHAVIOR through interaction—that is, to the multi-
modal array of communication practices
Almost 15 years ago, Streeck and Knapp that participants rely on to accomplish cer
(1992) asserted that “the classification of tain meanings (Sanders, 1987). This con
communicative behavior as either ‘verbal’ ceptualization is in line with Burgoon’s
or ‘nonverbal’ is misleading and obsolete” (1994) message perspective and Stamp and
(p. 3). Although this position is not new, Knapp’s (1990) interaction perspective on
and has continued to be a mantra of the nature of nonverbal communication,
research reform (see Bavelas & Chovil, this and Robinson and Stivers (2001) supported
volume), its implications often go ignored. the validity of a multimodal perspective in
There are at least two different ways of physician-patient interaction specifically.
23-Manusov.qxd 6/30/2006 7:34 PM Page 443
From this multimodal perspective, the first operationalizing talk is unitization. In stud
conceptualization (stated above) is statisti ies of physician-patient communication,
cally reified (rather than supported) by nonverbal behavior has been unitized his
physician-patient studies showing that, torically in terms of its duration and fre
when controlling for verbal variables, non quency (e.g., in seconds) across (sometimes
verbal variables retain independent signifi randomly selected) segments of (or entire)
cance (e.g., Bensing, 1991; Griffith, Wilson, visits. Contrary to this, physicians and
Langer, & Haist, 2003). patients organize much of their nonverbal
behavior relative to talk, and organize their
talk according to turn-taking rules for
THREE ASPECTS OF
ordinary conversation (Sacks, Schegloff, &
INTERACTIONAL CONTEXT
Jefferson, 1974; for a review, see Robinson,
2001a).3
Taking into account the multimodal con
One example occurs with gaze orienta
ceptualization, and because in face-to-face
tion. Because gaze can communicate that
interaction, nonverbal behavior is produced
one is “listening,” interactants orient to the
and understood largely by reference to talk,
general rule that recipients (e.g., physicians)
researchers studying nonverbal communi
should gaze at speakers (e.g., patients)
cation need to account for the organizing
when being spoken to (Goodwin, 1981).
effects of at least three aspects of interac
Turns of talk have consequential “posi
tional context: (1) turn taking, (2) social
tions,” such as beginnings, middles, and
action, and (3) sequences of talk and action.
endings (Schegloff, 1996), and Goodwin
Furthermore, researchers need to account
(1981) showed that recipients’ gaze toward
for the fact that (4) individual nonverbal
speakers is particularly salient at turn
behaviors are (almost always) produced and
beginnings. For instance, in Extract 1, when
understood by reference to each other.
the physician begins to ask his question
These four issues are discussed in order.
(Line 1), he is gazing at the computer screen
(see Figure 23.1, which corresponds posi
Turn Taking
tionally to the “1” in the transcript).
Buller and Street (1992) noted that tra Precisely at the completion of his ques
ditional measures of nonverbal behavior tion—that is, just as he is about to become
“do not account for how communicators a recipient of the patient’s talk—the physi
qualitatively interpret the behaviors being cian shifts his gaze to the patient (Figure
quantified” (p. 135, emphasis in original). 23.2; commas “,” symbolize movement of
The underlying issue in their statement is the physician’s head, the “X” symbolizes
whether operationalizations of nonverbal the point at which the physician’s gaze
behavior are ecologically valid (i.e., relevant reaches the patient, and brackets “[ ]” sym
to participants). An integral component of bolize simultaneous behavior).
Extract 1 [MC:12:03]
1 2
02 DOC: [,,,,,X]
Figure 23.1
Figure 23.2
23-Manusov.qxd 6/30/2006 7:34 PM Page 445
Figure 23.3
Due to the rules of gaze orientation in her answer with “Uh:m.” In this case, the
ordinary conversation (Goodwin, 1981), if patient does not succeed in securing the
speakers (e.g., patients) do not secure recip physician’s gaze, and she continues to pro
ients’ (e.g., physicians’) gaze at the begin duce a self-diagnosis: “I believe I have a
ning of their turns, speakers use vocal sinus infection.” (Line 2).4 Research shows,
hitches and perturbations—such as pauses, however, that although recipients’ gaze is
cutoffs, and other marked prosodic pat particularly salient at the beginning of
terns—to secure recipients’ gaze prior to speakers’ turns, it is relevant throughout
beginning or completing their turns. This is speakers’ turns, and speakers continue fre
equally true in physician-patient interaction quently to work to secure recipients’ gaze
(Heath, 1986; Ruusuvuori, 2001). In (e.g., Goodwin, 1981; Heath, 1986). When
Extract 1, the patient does secure the the patient in Extract 2 continues to produce
physician’s gaze prior to beginning her more talk (Line 3), she again attempts to
turn, and she produces a fluent response secure the physician’s gaze by cutting herself
(Line 3). This, however, is not the case in off (symbolized by the hyphen, “a-”) and
Extract 2. When the physician completes pausing for three tenths of a second (each
his question “What’s up.” (Line 1), he is tenth is symbolized by a dash, “(—)”). This
gazing down at the computer keyboard time, she succeeds—that is, immediately
(Figure 23.3). after her cutoff and two tenths of a second
At the outset of her response (Line 2), the of silence, the physician shifts his gaze
patient attempts to secure the physician’s toward the patient (Figure 23.4). The
gaze by pausing briefly (i.e., breathing in, patient begins to speak again precisely as the
symbolized by “.h”) and further delaying physician’s gaze arrives.
23-Manusov.qxd 6/30/2006 7:34 PM Page 446
Extract 2 [MC:20:02]
4 5
Figure 23.4
Figure 23.5
23-Manusov.qxd 6/30/2006 7:34 PM Page 448
Figure 23.6
Figure 23.7
23-Manusov.qxd 6/30/2006 7:34 PM Page 449
physicians’ psychodiagnostic abilities only across Extract 3, the physician maintains his
when patients were speaking. gaze and body orientation toward the patient
Other nonverbal behaviors, such as head (Figure 23.7). In the transcript, “/” and “\”
nodding, also appear to be organized by ref symbolize the upward and downward nod
erence to turn construction. For example, ding of the physician’s head, respectively.
Extract 3 [MC:11:02]
05 DOC: [ /\/ ]
06 PAT: la:st week w’s- thuh sore throat [ca:me an’- (.)] no:w
07 DOC: [ /\ ]
09 DOC: [ \/ ]
10 DOC: °Okay.°
touching, gesturing, and smiling (Beach & sensitive or embarrassing events (Beach &
LeBaron, 2002; Haakana, 2002). LeBaron, 2002; Heath, 1986, 1988), or
when they receive bad medical news (e.g.,
Social Action cancer diagnoses; Maynard, 2003), they
sometimes look away from their interlocu
Once nonverbal and verbal communica tors or cover their faces. One systematic
tion are reunited, researchers must recognize finding, then, is that while physicians
that (1) persons produce and understand engage in physical examination, patients
all communication primarily in terms of the tend to adopt a middle distance gaze orien
action(s) it performs (e.g., explaining, advis tation “away from the doctor yet at no
ing, informing; Schegloff, 1995); (2) actions particular object within the local environ
are organized by social rules that transcend ment” (Heath, 1988, p. 149).
individual actors (Heritage, 1984); and (3) The possibility that the social organiza
different social rules—for example, the rules tion of different verbal actions structures
for (or practices of) providing good (vs. bad) nonverbal behavior differently has direct
diagnostic news (e.g., Maynard, 2003)— implications for the types of outcomes dis
shape at least verbal behavior differentially cussed earlier. For example, it has already
(Heritage, 1984). been documented that physicians’ nonverbal
Although the claim that the social behavior affects raters’ evaluations of the
organization of action structures nonverbal affective-relational dimension of communi
behavior is in need of further support, it is cation. Such evaluations are, however, also
buttressed by a variety of findings. Harrigan affected by physicians’ verbal behaviors,
(1985) found that patients self-touched more which shape outcomes. For example, when
when answering questions than when being physicians are more verbally empathetic
asked questions and concluded that “the (e.g., provide more reassurance), patients are
semantic context of an utterance may be more satisfied and adherent to medical rec
expected to exert the strongest influence on ommendations and less willing to sue for
the expression of self touching” (p. 1164), malpractice (for a review, see Frankel,
which “is more complex than a simple cue 1995). When providers are less verbally
of anxiety” (p. 1167). Certain actions, such domineering or controlling (e.g., less direc
as apologizing for causing emotional pain tive), patients are more assertive, expressive,
(Beach & LeBaron, 2002), may make physi and disclose more information (Street,
cians’ touching of patients more normative 1992b); are more satisfied (Street, 1992a);
relative to other actions and thus affect and experience better physical-health out
patients’ evaluations of physicians’ touch. comes (e.g., metabolic control; Street et al.,
In line with Goffman’s (1963) notion of 1993). Researchers need to determine if
civil inattention, while patients engage in a these types of verbal behaviors structure
variety of private actions such as undress nonverbal behavior in particular ways, and
ing (Heath, 1986), it may be more norma they need to analyze nonverbal behaviors in
tive for physicians to avert their gaze, even conjunction with them.
if engaged in conversation. The action of If researchers are going to link nonverbal
“remembering” publicly is associated com behavior to talk in interaction, in this case
monly with gazing away from interlocutors physical-patient interaction, and if different
(Beach & LeBaron, 2002; Goodwin, social organizations of verbal action struc
1987). When patients discuss personally ture nonverbal behavior differentially, then,
23-Manusov.qxd 6/30/2006 7:34 PM Page 451
at the very least, researchers need to refine First, first-pair parts affect second-pair
extant coding schemata to better control parts by normatively obligating, and
for social action. The bulk of most coding constraining what counts as, relevant
schemata categorize talk not in terms of responses. For example, “Yes”/“No”-for
social action but rather in terms of a com matted questions constrain initial responses
bination of grammatical form and topical to versions of “Yes” or “No” (Raymond,
content. These limitations have been 2003), and such constraints can have impli
addressed partially by narrowing analyses cations for patient participation (Heritage
to, and developing new coding schemata & Robinson, 2006; Robinson, 2001a).
around, particular (classes of) social actions, Second, first-pair parts establish frame
such as patients’ requests for medical works for understanding second-pair parts,
services (Kravitz, Bell, & Franz, 1999).5 and thus the meaning of communicative
behavior is heavily influenced by its sequen
tial positioning (Schegloff & Sacks, 1973).
Sequencing
For example, in Extract 3, the physicians’
One of the most fundamental interac first two head nods (Lines 5 and 7) are pro
tional contexts organizing social action is the duced after the patient’s response and are
adjacency-pair sequence (Schegloff & Sacks understood as acknowledging the patient’s
1973). In its basic form, the adjacency-pair talk and as encouragement to continue
sequence is composed of two turns of talk: a (Schegloff, 1982). Evidence for this claim is
first-pair part, produced by one speaker, that in each case, the patient continues to
which initiates a course of action, and a sec produce a new turn-constructional unit. In
ond-pair part, produced by a different contrast, in Extract 4, the physician’s head
speaker, which responds to the initiated nod (Line 4) is produced as a response to the
action. Space limitations prohibit full expli patient’s request for confirmation (Line 1);
cation of the adjacency-pair sequence, and because of its sequential positioning, it gets
only two points are made here. understood differently as a confirmation.
Extract 4 [MC:14:02]
02 (1.2)
04 DOC: [ /\[/
Research attempting to discover vari- not accounting for “the two-way and con
ables that affect physicians’ and patients’ tingent (i.e., sequential) nature of physician-
communication has been criticized for (1) patient interaction” (Hall, Harrigan, &
23-Manusov.qxd 6/30/2006 7:34 PM Page 452
Rosenthal, 1995, p. 25) and (2) implying physician’s gaze shift to a female patient’s
claims about causation while using the cor chest can lead directly to the patient gazing
relation statistic (for other critiques, see at her own chest. Similarly, Heath (1986)
Buller & Street, 1992; Street & Buller, showed that physicians’ gaze shifts to
1987). Because correlations do not reveal objects of discussion, such as X rays, can
the direction of causality, because much of lead directly to patients’ gazing at the same
social action is organized by the adjacency- objects. In the same way that patients can
pair sequence, and because first-pair parts use vocal hitches and perturbations to solicit
affect second-pair parts, if sequencing is not physicians’ gaze (see above), patients can
accounted for, then significant correlations use nonverbal behaviors—such as gaze
are constantly in danger of being sequen shifts toward physicians, hand gestures,
tially spurious. torso shifts, and leg movements—to solicit
It also appears that nonverbal behavior both physicians’ gaze and talk (Heath,
can be sequenced independent of verbal 1986). For example, patients sometimes
adjacency-pair sequences (although nonver seek physicians’ gaze nonvocally for pur
bal sequences are still often organized by poses of exhibiting embodied characteristics
reference to talk). That is, there is prelimi of their suffering (Beach & LeBaron, 2002;
nary evidence that physicians’ and patients’ Heath, 1986, 2002). In Figure 23.6, in addi
nonverbal behaviors are nonrandom, pat tion to using vocal hitches and perturbations
terned, and synchronized. For instance, to solicit the physician’s gaze, the patient
Street and Buller (1987) provided evidence additionally gestures with her right hand to
that physicians and patients “matched” locate the position of her symptoms (i.e.,
gaze orientation, body orientation, and her “e:ye e:ar”). Although more research is
illustrative gestures, and Street and Buller necessary, Heath (1986) noted that some
(1988) demonstrated physician-patient rec of these nonverbal sequences operate in a
iprocation or convergence regarding body fashion similar to verbal summons-answer
orientation. Koss and Rosenthal (1997) sequences (Schegloff, 1968).
found external raters’ evaluations of physi
cian-patient nonverbal synchrony to be
The Inseparability of Nonverbal
positively associated with raters’ evalua
Behaviors From Themselves
tions of physician-patient rapport. Several
reviews (Kiesler & Auerbach, 2003; Lepper, In addition to turn taking, social action,
Martin, & DiMatteo, 1995) have suggested and sequences, nonverbal behavior embod
that the presence or absence of physicians’ ies its own interactional context. That is,
and patients’ nonverbal synchrony (or the social meaning of individual nonverbal
exchange) plays a role in a variety of par behaviors can be altered when they are
ticipants’ affective-relational attributions, employed simultaneously. For example,
such as affiliation and dominance, respec Harrigan and Rosenthal (1983) found
tively (for more on synchrony, see Tickle- external raters’ evaluations of physicians’
Degnen, this volume). rapport to be associated with interactions
The above correlational research is sup between physicians’ torso position (i.e.,
ported by focused studies of interaction. For forward or backward lean), head nodding,
example, in line with the observation that and leg position (i.e., crossed or uncrossed).
gaze communicates persons’ current focus Based on these types of findings, Harrigan
of attention, Heath (1988) showed that a and Rosenthal (1986) later asserted,
23-Manusov.qxd 6/30/2006 7:34 PM Page 453
opening questions (e.g., “What can I do for empathy, rapport, and generally a “positive”
you today?”) and showed that when physi affective-relational communication style—
cians removed their gaze from patients has wide-ranging effects (direct and indirect)
prior to patients having completed their on patients’ communicative, social, psycho
responses, patients tended to produce dis logical, and physiological health outcomes.
fluencies to (re)solicit physicians’ gaze. Despite these empirical strides, research on
Ruusuvuori found that patients produced nonverbal communication can be improved.
fewer disfluencies when physicians’ bodies The bulk of prior research has focused
were oriented toward (vs. away) from exclusively on individual nonverbal behav
patients (e.g., Figure 23.7 compared with iors (e.g., body orientation) or coherent
Figure 23.6). Ruusuvuori’s findings suggest aggregates of nonverbal behaviors (e.g.,
that in terms of physicians’ levels of reinforcement). This chapter argues that the
engagement with patients (i.e., attention social meaning of nonverbal behavior—and
to patient’s responses), patients understand thus its production, understanding, and
the absence or removal of physicians’ gaze effects—is fundamentally and irremediably
differently depending on the orientation of shaped by, and thus must be studied rela
physicians’ bodies. tive to, its situation within a variety of
The interrelationship between gaze and aspects of interactional context. This is not
body orientation can be complicated by to say that the effects of nonverbal behavior
other nonverbal behaviors. For example, in are relative. Rather, nonverbal behavior is
a nonmedical context, Goodwin (1981) sug organized systematically and finely by ref
gested that recipients’ lack of attention com erence to both talk and other nonverbal
municated by shifting their gaze away from behavior.
speakers can be partially offset by recipients This chapter demonstrates that physi
nodding during and after their gaze cians’ production of individual nonverbal
removal. To further complicate matters, cues (e.g., gaze, body orientation, head
because gaze and body orientation are used nodding), and patients’ understandings of
primarily to communicate (dis)engagement, such cues, is organized by rules associated
their understanding cannot be separated with turn taking, the construction of partic
from that of the objects being (dis)engaged. ular social actions, the sequencing of
Patients can differentiate between physi actions, and the organization of other non
cians’ gaze at patients’ eyes versus other verbal behaviors (e.g., the interorganization
body parts (e.g., legs, breasts, backs), versus of gaze and body orientation). Not taking
medical records, with different interactional these aspects of interactional context into
and attributional implications (Heath, account obscures our understanding of the
1986, 1988; Robinson, 1998). cause-effect relationships between nonver
bal (as well as verbal) communication vari
ables and their outcomes.
♦ Conclusion Relative to research on nonverbal com
munication between 1965 and 1995,
research over the last 10 years has
This chapter reviews findings related to languished. New directions in research
nonverbal behavior and physician-patient point toward developing new ways of
interaction and shows that physicians’ classifying—that is, conceptualizing and
nonverbal behavior—which is integral to measuring—nonverbal communication. On
the construction and management of the one hand, “without classification, there
23-Manusov.qxd 6/30/2006 7:34 PM Page 455
♦ References
1. Because of space limitations, this
chapter excludes research on patient-nurse
and patient-psychotherapist interaction (for Aruguete, M. S., & Roberts, C. A. (2002).
a review, see Caris-Verhallen, Kerkstra, & Participants’ ratings of male physicians
Bensing, 1999, and Hall, Harrigan, & who vary in race and communication style.
Rosenthal, 1995, respectively). Psychological Reports, 91, 793–806.
2. Compared with self-report measures of Bailey, K. D. (1994). Typologies and tax
emotional expressiveness, those that are based onomies: An introduction to classification
on external raters’ observations of concrete techniques. Thousand Oaks, CA: Sage.
behaviors are more conservative and thus more Beach, W. A., & LeBaron, C. D. (2002). Body
appropriate for isolating nonverbal “predictors” disclosures: Attending to personal problems
of behavior (Riggio & Riggio, 2002). and reported sexual abuse during a medical
3. Overall, researchers seldom measure encounter. Journal of Communication, 52,
physicians’ and patients’ turns. For an excep 617–639.
tion, see Street and Buller (1988), who found Bensing, J. M. (1991). Doctor-patient communi
physicians’ turns to be longer than those of cation and the quality of care. Social
patients and older patients’ turns to be longer Science & Medicine, 32, 1301–1310.
than those of younger patients. Bensing, J. M., & Dronkers, J. (1992).
4. The fact that the patient’s “.h Uh:m” does Instrumental and affective aspects of physi
not secure the physician’s gaze in this particular cian behavior. Medical Care, 30, 283–298.
instance does not invalidate it as being a member Bensing, J. M., Kerssens, J. J., & van der Pasch,
of a class of practices designed to do so M. (1995). Patient-directed gaze as a tool
(Goodwin, 1981; Heath, 1986; Ruusuvuori, for discovering and handling psychosocial
23-Manusov.qxd 6/30/2006 7:34 PM Page 456
problems in general practice. Journal of Frankel, R. M. (1995). Emotion and the physi
Nonverbal Behavior, 19, 223–242. cian-patient relationship. Motivation &
Ben-Sira, Z. (1982). Stress potential and eso Emotion, 19, 163–173.
tericity of health problems: The signifi Giron, M., Manjon-Arce, P., Puerto-Barber, J.,
cance of the physician’s affective behavior. Sanchez-Garcia, E., & Gomez-Beneyto
Medical Care, 20, 414–424. (1998). Clinical interview skills and identi
Buller, D. B., & Street, R. L., Jr. (1992). fication of emotional disorders in primary
Physician-patient relationships. In R. S. care. American Journal of Psychiatry, 155,
Feldman (Ed.), Applications of nonverbal 530–535.
behavior theories and research (pp. 119– Glassman, M., & Glassman, N. (1981). A mar
141). Hillsdale, NJ: Erlbaum. keting analysis of physician selection and
Burgoon, J. K. (1994). Nonverbal signals. In patient satisfaction. Journal of Health Care
M. L. Knapp & G. R. Miller (Eds.), Marketing, 1, 25–31.
Handbook of interpersonal communication Goffman, E. (1963). Behavior in public places.
(2nd ed., pp. 229–285). Thousand Oaks, New York: Free Press.
CA: Sage. Goffman, E. (1983). The interaction order.
Caris-Verhallen, W. M., Kerkstra, A., & American Sociological Review, 48, 1–17.
Bensing, J. M. (1999). Non-verbal behav Goodwin, C. (1981). Conversational organiza
iour in nurse-elderly patient communica tion: Interaction between speakers and
tion. Journal of Advanced Nursing, 29, hearers. New York: Academic Press.
808–818. Goodwin, C. (1987). Forgetfulness as an interac
Cegala, D. J., McNeilis, K. S., McGee, D. S., tive resource. Social Psychology Quarterly,
& Jonas, A. P. (1995). A study of doctors’ 50, 115–131.
and patients’ perceptions of information Goodwin, C. (1995). Seeing in depth. Social
processing and communication competence Studies of Science, 25, 237–274.
during the medical interview. Health Griffith, C. H., Wilson, J. F., Langer, S., &
Communication, 7, 179–203. Haist, S. A. (2003). House staff nonverbal
Duffy, F. D., Gordon, G. H., Whelan, G., Cole- communication skills and standardized
Kelly, K., & Frankel, R. (2004). Assessing patient satisfaction. Journal of General
competence in communication and inter Internal Medicine, 18, 170–174.
personal skills: The Kalamazoo II Report. Haakana, M. (2002). Laughter in medical
Academic Medicine, 79, 495–507. interaction: From quantification to analy
Duggan, A. P., & Parrott, R. L. (2001). sis, and back. Journal of Sociolinguistics,
Physicians’ nonverbal rapport building and 6, 207–235.
patients’ talk about the subjective compo Hall, J. A., Harrigan, J. A., & Rosenthal, R.
nent of illness. Human Communication (1995). Nonverbal behavior in clinician-
Research, 27, 299–311. patient interaction. Applied and Preventive
Egbert, L. D., Battit, G. E., Welch, C. E., & Psychology, 4, 21–37.
Bartlett, M. K. (1964). Reduction of post Hall, J. A., Irish, J. T., Roter, D. L., Ehrlich,
operative pain by encouragement and C. M., & Miller, L. H. (1994). Gender in
instruction of patients: A study of doctor- medical encounters: An analysis of physician
patient rapport. New England Journal of and patient communication in a primary care
Medicine, 270, 825. setting. Health Psychology, 13, 384–392.
Ekman, P., & Friesen, W. V. (1969). The reper Harrigan, J. A. (1985). Self-touching as an indi
toire of nonverbal behavior: Categories, ori cator of underlying affect and language
gins, usage, and coding. Semiotica, 1, 49–98. processes. Social Science & Medicine, 20,
Farace, R. V., Monge, P. R., & Russell, 1161–1168.
M. (1977). Communicating in organiza Harrigan, J. A., Oxman, T. E., & Rosenthal, R.
tions. Reading, MA: Addison-Wesley. (1985). Rapport expressed through nonverbal
23-Manusov.qxd 6/30/2006 7:34 PM Page 457
Street, R. L., Jr., & Buller, D. B. (1987). Nonverbal Van Dulmen, A. M., Verhaak, P. F. M., & Bilo,
response patterns in physician-patient interac H. J. G. (1997). Shifts in doctor-patient
tions: A functional analysis. Journal of communication during a series of outpa
Nonverbal Behavior, 11, 234–253. tient consultations in non-insulin-dependent
Street, R. L., Jr., & Buller, D. B. (1988). Patients’ diabetes mellitus. Patient Education &
characteristics affecting physician-patient Counselling, 30, 227–237.
nonverbal communication. Human Com Verhaak, P. F. M. (1988). Detection of psycho
munication Research, 15, 60–90. logical complaints by general practitioners.
Street, R. L., Jr., Piziak, V. K., Carpentier, W. S., Medical Care, 26, 1009–1020.
Herzog, J., Hejl, J., Skinnner, G., et al. Weinberger, M., Greene, J. Y., & Mamlin, J. J.
(1993). Provider-patient communication (1981). The impact of clinical encounter
and metabolic control. Diabetes Care, 16, events on patient and physician satisfaction.
714–721. Social Science & Medicine, 15E, 239–244.
23-Manusov.qxd 6/30/2006 7:34 PM Page 460
24-Manusov.qxd 6/30/2006 7:35 PM Page 461
24
NONVERBAL DYNAMICS IN
COMPUTER-MEDIATED
COMMUNICATION,
OR :( AND THE NET :( ’S WITH
YOU, :) AND YOU :) ALONE
� Joseph B. Walther
Michigan State University
◆ 461
24-Manusov.qxd 6/30/2006 7:35 PM Page 462
known and easily recognized although their theoretical and technological development
utility is questionable. requires more exacting research on nonver
The limited research on extant nonverbal bal communication in an area once thought
cues or their substitutes, as well as emerging to be devoid of such features.
research on specific reintroductions of non
verbal features through avatars, videocon
ferencing, and virtual reality systems is
♦ Hypothesizing About the
leading to a more functionally oriented per
spective on mediated human communica
“Lack” of Nonverbal Cues
tion. Newer research is focusing on what in Early CMC Theories
people communicate, and the variety of
means by which to do so, some of which SOCIAL PRESENCE THEORY
means were considered previously the exclu
sive domain of nonverbal cues. As a result, a The earliest predictions applied to CMC
major consequence of contemporary CMC stressed the depersonalizing, predominantly
research is to help us learn more about com negative effects of communication without
munication symbol systems and their func nonverbal cues. The first of these theories
tions in general, by observing both their was social presence theory (Short, Williams,
absence and their systematic replacement. & Christie, 1976), the original treatment of
This chapter reviews the major theories which is noteworthy for its comprehensive
and their research traditions on CMC and treatment of the role of nonverbal cues
the similarities and differences among them in communication. Originally focused on
with respect to how the relative absence of video- and audioconferencing, its theoreti
nonverbal cues may affect communication cal specifications have also been applied to
and social perceptions. As will be argued, text-only communication (Hiltz, Johnson,
most of these approaches have relegated & Agle, 1978; Rice, 1984; Rice & Case,
nonverbal communication to a “black box,” 1983). The theory deals with decrements in
in a kind of all-or-nothing fashion, assum interpersonal affect as communication
ing that all nonverbal cues lead to a variety systems incrementally reduce the cue
of functions, and that the cues and func systems that users may employ. Thus, as
tions are isomorphic (i.e., that nonverbal communicators shift from FtF to videocon
cues are tied directly and exclusively to ferencing, many proxemic, as well as
communicative social functions, such that haptic, cues are unavailable. Moving
the absence of such cues precludes func to audioconferencing, kinesics and any
tional effects from occurring). remaining proxemic cues are also removed.
The chapter then discusses the potency Short et al. (1976) equate the uses or
of chronemics in CMC (e.g., alternative absence of these cue systems with the
temporal scales, time pressure, and the degree of “social presence” that communi
implicit and explicit effects of timing cues cators may experience, positing that social
on interpersonal judgments online). It turns presence declines as the number of cue
next to a variety of ways in which users or systems declines. Social presence, in turn,
technology designers attempt to reintro is conceptualized as the communicator’s
duce nonverbal cues into CMC or other involvement with the target of the conver
electronic communication systems. Finally, sation, and it is associated with warmth
considering exemplary approaches to online and friendliness. Many studies have sup
deception and some research employing ported the premises of social presence
virtual reality systems, we see that future theory (for review, see Walther & Parks,
24-Manusov.qxd 6/30/2006 7:35 PM Page 463
2002), although it has also received much to use natural and varied language, and
criticism insofar as its application to CMC the extent to which message exchanges
is concerned (e.g., Lea, 1991; Walther, offer immediate feedback (i.e., sender and
1992). receiver exchanges are bidirectional, or
they are asynchronous and responses are
delayed). Together, these dimensions define
THE LACK OF SOCIAL “media richness.”
CONTEXT CUES HYPOTHESIS The second important difference
between this theory and the others is the
A similar perspective to social presence specification regarding the predicted effec
theory is the lack of social context cues tiveness and efficiency of richer versus
hypothesis (Kiesler, 1987; Kiesler, Siegel, & leaner media when considering the degree
McGuire, 1984; Siegel, Dubrovsky, Kiesler, of equivocality and uncertainty involved in
& McGuire, 1986; Sproull & Kiesler, 1986). the communication task. Thus, for highly
This position argues that nonverbal cues in equivocal tasks, richer media are posited to
FtF settings establish the social context of be more efficient, whereas for simpler tasks,
interaction, and with the awareness of social although a rich or lean medium might be
context, participants infer and perform equally as effective, a leaner medium may
normative behavior. Without social context be more efficient (Daft & Lengel, 1984).
cues, participants are deindividuated and For instance, to ask a colleague what time a
thus behave aberrantly, including being meeting is scheduled to take place, one may
self- rather than other-focused, task- go down the hallway FtF, but a phone call
oriented, and disinhibited. These states lead or e-mail would work as well, possibly
not only to colder and more task-oriented more quickly, and with less effort.
communication, it is argued, but also to The third difference is that, whereas this
engage in “flaming” (name-calling, swearing, theory, like others, places a premium on
or other uninhibited expressions) online and nonverbal and other aspects of communica
more attitude polarization. This position, like tive flexibility, it stresses the role of multiple
social presence theory, suggests that the cues as sources to facilitate the comprehen
absence of nonverbal cues is the causal factor sion of information rather than as a source
distinguishing FtF and online interaction. of individuation, social presence, or social
context. In media richness theory, the avail
MEDIA RICHNESS ability of nonverbal cues (without differen
tiation) and other communication system
A third theory also regards the differ attributes are expected to help make media,
ences among media and their effects due to and messages, richer, leading to the reduc
the range of nonverbal cue systems media tion of equivocality in shorter periods of
carry, although media richness theory (Daft time. Although interpersonal effects have
& Lengel, 1984, 1986) differs from the pre been imputed as derivatives of this theory
vious positions in three important respects. (Markus, 1994), the original formulation of
First, although the number of cue systems the theory makes no such claim.
supported is a primary difference among
communication media in this theory, cue
systems are joined by three other elements OVERVIEW AND SUMMARY
in differentiating media capacity: the ability
to personalize messages (i.e., to tailor mes As a group, all three of these theories sug
sages for a specific recipient), the capacity gest a “black box” approach to the role of
24-Manusov.qxd 6/30/2006 7:35 PM Page 464
expressed through a variety of modalities. and CMC transcripts for verbal indications
Like media richness theory (Daft & Lengel, of affinity. Regression analyses of the cues
1984), SIP acknowledges that written cues in both conditions were used to identify the
alone may be less efficient within a given variations in cues and channels that most
time interval compared with a simultane strongly predicted the variations in partner
ously multimodal (i.e., kinesic, vocalic, ratings of affect. As expected, that FtF part
and verbal) exchange, but given sufficient ners expressed affinity through nonverbal
time and exchange, the two systems may cues primarily, with vocalic cues (pleasant
be functionally equivalent, and CMC users ness, vocal sharpness, vocal condescension,
make these adaptations fluidly. and timber) predominating over kinesics;
The SIP theory has been supported in verbal cues were not significantly associated
several empirical studies (for review, see with FtF liking in comparison with these
Walther & Parks, 2002). For instance, Liu, nonverbal variations. In CMC conversa
Ginther, and Zelhart (2002) found that tions, however, an equivalent proportion
impression development in CMC was sensi of the variance in liking was accomplished
tive to both the length of e-mail messages through verbal behaviors (explicit verbal
and the frequency of e-mail messages from a statements of affection, changing the
partner over time, and Walther and Burgoon subject, and various forms of disagree
(1992) found that relational communica ment), demonstrating comparability and
tion levels changed more or less in parallel substitutability of verbal cues in CMC for
between CMC and FtF groups in response to vocalic and kinesic cues in FtF interaction.
time accrual rather than to the differences
between communication conditions (see
also Chidambaram & Bostrom, 1993). SUMMARY
These studies lend credence to the model’s
causal factors and predicted effects; they did The major theories of CMC each portray
not examine the microprocesses implicated significant effects of the reduction of non
by the theory—that is, the substitution of verbal cues online. Positions range from the
verbal cues in the service of functions for austere, early formulations, where nonver
which nonverbal cues are employed offline. bal cues were isomorphic with certain com
A recent study addressed this gap by municative functions, to the more adaptive
assessing the specific behaviors in alternative models of hyperpersonal CMC and SIP, in
channels that express affinity. Walther et al.’s which users exploit or work through the rel
(2005) experiment employed decision- ative lack of nonverbal cues. Cutting across
making dyads meeting FtF or via synchro these models, other research has focused on
nous computer chat. One member was specific cues—natural or stylized—and the
prompted to enact greater or lesser levels of degree to which CMC users adapt affective
liking toward his or her partner after an ini meaning to their usage.
tial interaction period, by whatever way he
or she chose to display the affect. The other
dyad partner rated the ad hoc confederates’ ♦ The Cues That
performance on perceived immediacy and Remain: Chronemics
affection. Coders rated kinesic cues from
videotapes of the FtF confederates and
independently rated the vocalic perfor Whereas physical behavior, voice, space,
mances as heard through a content-filtering and appearance cues are indeed absent in
device. Additional coders analyzed both FtF text-based CMC, the chronemic cue system
24-Manusov.qxd 6/30/2006 7:35 PM Page 468
remains, although it is frequently over (Reid, Ball, Morley, & Evans, 1997; Reid,
looked in descriptions of CMC’s nonver Malinek, Stott, & Evans, 1996) determined
bal capacity. Hesse, Werner, and Altman that the relational tone of CMC is more
(1988) were among the first to recognize sensitive to time scarcity than is FtF inter
the potential for temporal dynamics and action: In CMC groups, more than in
cues to play a significant role in CMC, FtF groups, discourse became less rational,
noting potential departures from traditional and less affective content appeared, as pres
interaction patterns in terms of temporal sure increased with shorter time limits.
scale (the temporal scope and duration of Conversely, it appears that a long duration
events and relationships), the sequencing of of time spent in CMC with a partner is
actions, the pace, and the salience of past, inferred as a token of relational intimacy
present, or future issues in ongoing CMC (Henderson & Gilding, 2004).
interactions. Although several studies can Response latencies are another familiar
now be said to address some of these issues, chronemic characteristic, and their effects
few of them have noted Hesse et al.’s origi have been studied in several CMC contexts.
nal thinking. Members make biased attributions for
Among those studies examining tem response delays, assuming personal rather
poral factors in CMC, chronemic dynamics than situational causes for lags by distant
are potent forces in the experience of team members (Cramton, 2001). Failure
CMC users. As Kalman and Rafaeli to get responses may erode initial levels of
(2005) observed, for example, “One of the trust in virtual groups (Jarvenpaa, Knoll, &
unknowns of emailing is the time it will Leidner, 1998), and frequent messaging is
take the receiver to form and post a reply. noted consistently as a critical factor in
Response times vary considerably, and virtual group trust, affective relations, and
the chronemics of email are an important effectiveness (Walther & Bunz, 2005), espe
non-verbal cue which can convey meaning” cially with regard to partners’ replies to
(p. 1). According to Rice (1990), e-mail an individual’s conversational initiations or
users attend to the time stamps that are requests (Iacono & Weisband, 1997).
placed on messages automatically, inferring Latencies also have mixed effects in dyadic,
from them when a message was sent and synchronous CMC. In organizational
how much latency occurred before one of settings where members use Instant Messen
their own messages received a reply. ger, a query that goes without a response is
Temporal dynamics affect virtual groups frequently attributed to one’s partner being
in a variety of ways, although specific busy (Nardi, Whittaker, & Bradner, 2000).
chronemic cues may or may not play a role In social chatting, however, individuals
in these effects. Orlikowski and Yates who find themselves waiting for replies
(2002) found that virtual groups’ activity grow increasingly frustrated if not hostile
cycles became oriented more toward critical (Rintel & Pittam, 1997; see also Feenberg,
events, such as the occasional exchange of 1989).
collaborative documents, than toward the One study tested the interpersonal impres
influence of predetermined deadlines. In sions affected by variations in e-mail response
a field study of organizational CMC, latency, as well as whether messages were
Steinfield (1986) found that CMC becomes sent at night or during the day (Walther &
more task oriented and less socially ori Tidwell, 1995). Researchers created several
ented as collaborators get closer to project pairs of e-mail message facsimiles featuring an
deadlines. In a closer inspection of time initial message and a reply that appeared to be
pressure and CMC, Reid and colleagues initiated by a vice president and replied to by
24-Manusov.qxd 6/30/2006 7:35 PM Page 469
a manager who were separated geographi and emerging technologies selectively rein
cally within a corporation. The time stamps troduce additional cues into communicative
on these e-mail facsimiles made one pair exchanges among people who do not meet
appear to have been sent shortly after 10 a.m. FtF. Whether executed by users or technol
and another pair after 10 p.m. This factor was ogy designers, these developments and their
crossed by the apparent response lag. In some impacts inform nonverbal communication
pairs, the reply seemed to occur several min principles.
utes after the initial message; alternatively, 24
hours and several minutes appeared to have
elapsed. The time stamps were further crossed CUE SURROGATES: EMOTICONS
over two kinds of message exchanges, a task-
oriented request versus social banter. A considerable amount of attention has
Ratings of these various stimuli con been devoted to the use of “emoticons” in
firmed chronemics-based hypotheses. When CMC. Emoticons are the presentation of
task messages were sent at night the sender keyboard symbols used in such manner as
was rated highest on dominance compared to resemble facial expressions. They are
with the same message sent during the day. assumed widely to express emotion and are
The pattern was opposite for social mes frequently described as emotional surro
sages, which signaled more dominance by gates in CMC for facial expressions and
day than at night. The amount of affection other nonverbal cues to emotion. “Because
ascribed to a sender’s message was affected the use of e-mail eliminates visual cues such
by an interaction between day and night, the as head nodding, facial expressions, pos
promptness of the reply, and the thematic ture, and eye contact found in FtF commu
content. The most affection accorded to nication, CMC users often incorporate
task exchanges occurred when there was a emoticons as visual cues to augment the
quick reply to a daytime request, and the meaning of textual electronic messages”
least affection was associated with a prompt (Rezabek & Cochenour, 1998, pp. 201–
response to a nighttime message. As for 202). The use of emoticons in CMC dates
social messages, more affection was per back at least as far the early 1980s, and for
ceived in a slower reply to a daytime mes many years “smiley dictionaries” circulated
sage than a fast reply, but a fast reply at in the Internet, containing hundreds of vari
night showed more affection than a slow ations and the verbal labels of their alleged
one. Consistent with Hall’s (1959) observa emotional equivalents (e.g., Godin, 1993;
tions about FtF speech lags, it appears that Sanderson, 1993). The best known of these
expectations of quick e-mail replies are symbols are “a smile, wink, and frown,
relaxed within established social relation respectively: :-) ;-) :-(” (Danet, Ruedenberg-
ships, although reactions to response latency Wright, & Rosenbaum-Tamari, 1997,
are quite different within impersonal rela n.p.). These symbols are well recognized
tions, both online and offline. within the CMC-using community. Among
one college student sample, basic emoticons
were interpreted more reliably than were
♦ Reintroducing Cues photos of human facial expressions of emo
tion: Whereas Ekman and Friesen (1975)
report percentages of agreement about the
Whereas chronemic cues have always been association of facial photos depicting basic
available, thus countering the ideas that human emotions from 97% for happiness
nonverbal cues are lacking from CMC, new to 67% for anger, Walther and D’Addario
24-Manusov.qxd 6/30/2006 7:35 PM Page 470
(2001) found that the :) and :( emoticons sender’s affective state and attitude about
achieved 98% consensus for happiness and the course. There was very little effect of
sadness, respectively, and associations of emoticons on attitude and interpretation;
other emoticons with anger, disgust, and what impact they did exhibit was not in
fear ranged from 88% to 85%. accord with the hypotheses from facial
Although the literature on emoticons expression research. Specifically, smiley
asserts frequently that they function as emoticons had no effect on message inter
nonverbal (facial) expressions, very little pretation whatsoever. Frown emoticons
research has examined the functional reduced the positivity of a positive verbal
impact of these symbols. Most of the message, but frowns did not affect interpre
research on emoticons has analyzed pat tations of negative verbal messages—that
terns of their use based on demographic is, did not make them even more negative.
factors: Females use them more frequently Overall, there appeared to be a negativity
than do males (Witmer & Katzman, 1997) effect: When any negative message element
and their usage even depends on which part appeared, whether it was an emoticon or
of the United States e-mail users reside a verbal statement, the interpretation was
(Rezabek & Cochenour, 1998). Walther negative. Additionally, the combination of
and D’Addario (2001) explored their func verbal statements and their opposite emoti
tional dynamics. Reviewing the facial affect con were not significantly different in
literature, they derived hypotheses predict sarcasm from other combinations; only a
ing relationships between emoticons and positive verbal message with a ;) emoticon
accompanying verbal messages on affective was rated higher in sarcasm than other
message interpretation. These relationships combinations, suggesting that the wink
included a variety of additive effects, by symbol has some iconic value in CMC but
which the emotional valence of the emoti that a negative verbal statement may over
con would be added to the emotional ride the emoticon effect. Given that only
valence of a verbal message, leading to sup the frown emoticon affected meaning, it
plementation (for a positive emoticon plus appears to be the case that :( and the net
a positive verbal message, a negative emoti :( ’s with you, but :) and you :) alone.
con plus a negative verbal message) or mod
ification (a positive element plus a negative
“canceling out” or neutralizing overall AVATARS AND VIDEO
affect). Alternatively, visual primacy was
posited: An emoticon’s valence might over In addition to stylized affective cues such
ride that of the verbal statement. The com as emoticons, developers and users explore
bination of positive and negative messages, the utility of reintroducing certain visual
among emoticon and verbal statements, cues into distributed interaction. This has
might also result in an interpretation of sar been done primarily through the use of
casm, as might the iconic ;) or “winkie.” avatars, icons, and videoconferencing.
In a 4 by 2 experimental procedure, :) ;)
:( or no emoticon were inserted alternately Avatars. Avatars are two-dimensional
in simulated e-mail message mock-ups that representations on a computer screen that
contained either a positive or a negative chat users can select and move around the
verbal statement about a college course. screen during online interaction. Avatars
Participants viewed one of these mock-ups are used frequently in various multiplayer
and then rated the supposed message computer games such as Everquest and the
24-Manusov.qxd 6/30/2006 7:35 PM Page 471
SimsOnline, although they have a somewhat (see also Bengtsson, Burgoon, Cederberg,
longer history in multiuser chat spaces such Bonito, & Lundberg, 1999).
as the Palace (www.thepalace.com/; see Krikorian, Lee, and Chock (2000)
Suler, 1999). In most environments, one examined proxemic responses using
selects an avatar initially from among a avatars. Within a Palace chat space, partic
stock of available figures. Avatars are often ipants engaged in a get-to-know-you con
cartoonish and range from very generic versation online, exchanging text and
with few distinguishing features to rather manipulating the positions of their respec
elaborate in design. It is also possible to tive avatars. Researchers captured the
create an individual avatar using graphics video images and developed an automated
software or to craft an avatar from a pho system for measuring the dynamic distance
tograph and upload it to the interaction between avatars based on the pixels in the
space. During interaction, dialogue often center of the avatars and the relative dis
appears as text, as though emanating from tances between them. Results showed rela
an avatar like the conversational bubbles tively even proportions among the pairs of
that appear in comic strips. Advocates of participants who moved their avatars
these systems argue that they help orient closer, farther, or not at all over the course
players and that they reduce the impersonal of the conversation. Among those who
nature of text-based systems. Much of the moved farther apart, there was also an
research employing avatars focuses on the increase in avatar expressive movement—
psychoanalytic dimensions of avatar selec as if being too close inhibited other
tion and usage, such as how an avatar both kinesics—which was accompanied by self-
reflects aspects of the user’s personalities as reports of greater conversational appropri
it also shapes the online persona through ateness and conversational involvement
social interaction (Suler, 1999). In some among participants. There was also a
online multiplayer games, avatars are used curvilinear trend on other ratings, how
to duel or fight, although users socialize ever. In general, the correspondence of par
through text to a large extent alongside the ticipants’ social attraction ratings and
avatar battles (Peña & Hancock, 2006). avatar distances mapped onto the predic
Two avatar studies bear immediate tions of nonverbal expectancy violations
relevance to nonverbal communication theory (Burgoon & Hale, 1988), in that
research. Nowak and Biocca (2003) exam attraction was greater when avatars inter
ined avatars varying in anthropomorphic acted either at relatively close or far dis
appearance, representing conversational tances rather than at median ranges.
partners. The more anthropomorphic
representations depicted 3-D drawings of Anthropomorphic Icons. A less manipula
heads and faces, whereas less anthropo ble form of avatars is anthropomorphic
morphic versions featured disembodied, graphics, or icons, accompanying CMC
cartoonish pairs of eyes and lips. Contrary messages or appearing fixed on a screen
to hypotheses, the less anthropomorphic during chat. Icons have the capacity to
the avatar, the greater the participants’ influence receiver’s interpretations of mes
responses on various measures of presence. sages, even if receivers are aware that the
The authors concluded that the more realis icon does not necessarily represent the char
tic but imperfect human resemblances frus acteristics of the actual message sender.
trated users’ expectations, whereas the Isotalus (2003) found that receivers’
more abstract images drew greater interest responses to news stories delivered to
24-Manusov.qxd 6/30/2006 7:35 PM Page 472
handheld computers differed based on the conditions obtained few differences on task
apparent gender of an icon accompanying performance quality or interpersonal per
the story. Participants paid more attention ceptions due to main effects of the inter
when the icon appeared to be female. faces (Nowak, Watt, & Walther, 2005).
Furthermore, males found the news more There was greater perceived involvement
credible when accompanied by a male icon, with others group members in those condi
whereas females’ credibility assessments tions with fewer visual cues, which, in turn,
were higher for female icons; evaluations of led to increases in social attraction and
the stories’ entertainment followed an credibility ratings of partners.
opposite pattern. Lee (2005) also used gen Interestingly, by focusing video on the
dered icons to accompany spontaneous, objects that collaborators discuss rather
dyadic CMC chat messages, but the partic than on facial displays (but including par
ipants were aware that the gendered icon ticipants’ voices) seems to be superior to
had been randomly assigned to users; that face-oriented videoconferencing in many
is, there was an even chance that the gender cases (Brittan, 1992). This may be due to
of the icon and the user were mismatched. the communication efficiency with which
Despite this awareness, participants (espe humans process multimodal messages,
cially female participants) attributed the when one level of content traverses the
gender of the chat partner on the basis of vocal-to-auditory channel, leaving vision
the icon’s gender. This over-interpretation free to focus on a common object. When
of gender based on a simple physical appear videoconferencing depicts the communica
ance representation suggests, as previous tors rather than the objects they are dis
perspectives have argued, that CMC plus a cussing, the objects and the image of
little nonverbal representation leads to partners compete for visual attention,
potentially exaggerated perceptions. leading to decrements in efficiency and
performance.
Videoconferencing. Research exploring Fussell and colleagues (Fussell, Kraut, &
videoconferencing to enhance social pres Siegel, 2000; Gergle, Kraut, & Fussell, 2004;
ence and improve remote collaborations Kraut, Fussell, & Siegel, 2003) have
predates the Internet and digital technol employed audio/videoconferencing with the
ogy considerably (see for review Chapanis, visual field aimed at an object that one part
Ochsman, Parrish, & Weeks, 1972). Most ner manipulates but both can see. In one
videoconferencing arrangements and stud study (Kraut et al., 2003) a head-mounted
ies involve real-time visual conveyance of camera focused on a bicycle that one partner
participants’ faces to remote partners, repaired while an expert helper elsewhere
accompanied by their voices. The results of viewed the bicycle (and the repairer’s manip
this research have been generally disap ulations of it) via video, and instructed the
pointing. Whereas users report greater sub repairer via audio while both looked at the
jective presence when video is available, bike. In another study (Gergle et al., 2004),
their communication effectiveness and task both partners viewed puzzle pieces on an
output tends to be no better, and sometimes electronic video display, while one partner
worse, than non-visual interfaces provide guided the other via voice toward the puz
(Gale, 1991; Storck & Sproull, 1995). zle’s completion. Compared with other video
Similarly, a recent study comparing asyn foci, or no video, participants performed
chronous videoconferencing with synchro more accurately and quickly at the tasks
nous video, text-only systems, and FtF that were employed in these studies. In the
24-Manusov.qxd 6/30/2006 7:35 PM Page 473
intergroup relations and group processes. Krikorian, D. H., Lee, J., & Chock, T. M.
London: Routledge. (2000). Isn’t that spatial? Distance and
Hollingshead, A. B. (2000). Truth and lying in communication in a 2D virtual environ
computer-mediated groups. In M. A. Neale, ment. Journal of Computer-Mediated
E. A. Mannix, & T. Griffith (Eds.), Communication, 5. Retrieved March 7,
Research in managing groups and teams 2006, from http://www.jcmc.indiana.edu/
(Vol. 3: Technology and teams, pp. 157– vol5/issue4/krikorian.html
173). Greenwich, CT: JAI Press. Lea, M. (1991). Rationalist assumptions in
Iacono, C. S., & Weisband, S. (1997). cross-media comparisons of computer-
Developing trust in virtual teams. mediated communication. Behaviour &
Proceedings of the 30th Annual Hawaii Information Technology, 10, 153–172.
International Conference on System Lea, M., & Spears, R. (1992). Paralanguage and
Sciences. Retrieved July 30, 2005, from social perception in computer-mediated
http://ieeexplore.ieee.org/search/wrapper communication. Journal of Organizational
.jsp?arnumber=665615 Computing, 2, 321–341.
Isotalus, P. (2003, November). Gender and Lea, M., Spears, R., & de Groot, D. (2001).
interface agents in the on-line news. Paper Knowing me, knowing you: Anonymity
presented at the annual meeting of the effects on social identity processes within
National Communication Association, groups. Personality and Social Psychology
Miami Beach, FL. Bulletin, 27, 526–537.
Jarvenpaa, S. L., Knoll, K., & Leidner, D. E. Lee, E. (2005). Effects of the influence agent’s
(1998). Is anybody out there? The implica sex and self-confidence on informational
tions of trust in global virtual teams. Journal social influence in computer-mediated com
of Management Information Systems, 14, munication: Quantitative versus qualitative
29–64. presentation. Communication Research,
Jones, S. E., & LeBaron, C. D. (2002). Guest 32, 29–58.
editors’ introduction. Journal of Communi Liu, Y. L., Ginther, D., & Zelhart, P. (2002).
cation, 52, 499–521. An exploratory study of the effects of
Kalman, Y. M., & Rafaeli, S. (2005). Email frequency and duration of messaging on
chronemics: Unobtrusive profiling of impression development in computer-
response times. Proceedings of the 38th mediated communication. Social Science
Annual Hawaii International Conference Computer Review, 20, 73–80.
on System Sciences. Retrieved July 30, Markus, M. L. (1994). Finding a happy
2005, from http://ieeexplore.ieee.org/search/ medium: Explaining the negative effects of
wrapper. jsp?arnumber =1385456 electronic communication on social life at
Kiesler, S. (1987). Social aspects of computer work. ACM Transactions on Information
environments. Social Science, 72, 23–28. Systems, 12, 119–149.
Kiesler, S., Siegel, J., & McGuire, T. W. (1984). Nardi, B., Whittaker, S., & Bradner, E. (2000).
Social psychological aspects of computer- Interaction and outeraction: Instant mes
mediated communication. American Psy saging in action. In Proceedings of the 2000
chologist, 39, 1123–1134. ACM Conference on Computer Supported
Kim, J., Kim, H., Tay, B. K., Muniyandi, M., Cooperative Work (pp. 79–88). New York:
Srinivasan, M. A., Jordan, J., et al. (2004). ACM Press. Retrieved March 1, 2006, from
Transatlantic touch: A study of haptic http://www.dis.shef.ac.uk/stevewhittaker/
collaboration over long distance. Presence, outeraction_cscw2000.pdf
13, 328–337. Nowak, K. L., & Biocca, F. (2003). The effect
Kraut, R. E., Fussell, S. R., & Siegel, J. (2003). of the agency and anthropomorphism on
Visual information as a conversational users’ sense of telepresence, copresence, and
resource in collaborative physical tasks. social presence in virtual environments.
Human-Computer Interaction, 18, 13–49. Presence, 12, 481–494.
24-Manusov.qxd 6/30/2006 7:35 PM Page 478
Nowak, K., Watt, J. H., & Walther, J. (2005). The new media: Communication, research, and
influence of synchrony and sensory modality technology (pp. 129–156). Beverly Hills,
on the person perception process in computer CA: Sage.
mediated groups. Journal of Computer- Rice, R. E. (1990). Computer-mediated commu
Mediated Communication, 10(3), article 3. nication system network data: Theoretical
Retrieved July 1, 2005, from http://jcmc. concerns and empirical examples. Inter
indiana.edu/vol10/issue3/nowak.html national Journal of Man-Machine Studies,
Orlikowski, W. J., & Yates, J. (2002). It’s about 32, 627–647.
time: Temporal structuring in organiza Rice, R. E., & Case, D. (1983). Electronic
tions. Organization Science, 13, 684–700. message systems in the university: A
O’Sullivan, P. B. (2000). What you don’t know description of use and utility. Journal of
won’t hurt me: Impression management Communication, 33, 131–154.
functions of communication channels in Rintel, E. S., & Pittam, J. (1997). Strangers
relationships. Human Communication in a strange land: Interaction management
Research, 26, 403–431. on Internet Relay Chat. Human Com
Pen, J., & Hancock, J. T. (2006). An analysis munication Research, 23, 507–534.
of socioemotional and task communication Rogers, P., & Lea, M. (2004). Cohesion in
in online multiplayer videogames. Com online groups. In K. Morgan, C. A.
munication Research, 33, 92–109. Sanchez, C. A. Brebbia, & A. Vioskounsky
Postmes, T., & Spears, R. (2002). Contextual (Eds.), Human perspectives on the Internet
moderators of gender differences and society: Culture, psychology and gender
stereotyping in computer-mediated group (pp. 115–124). Southampton, England:
discussions. Personality and Social WIT Press.
Psychology Bulletin, 28, 1073–1083. Sanderson, D. (1993). Smileys. Sebastopol, CA:
Postmes, T., Spears, R., & Lea, M. (1998). O’Reilly & Associates.
Breaching or building social boundaries? Short, J., Williams, E., & Christie, B. (1976).
SIDE-effects of computer-mediated com The social psychology of telecommunica
munication. Communication Research, 25, tions. London: Wiley.
689–715. Siegel, J., Dubrovsky, V., Kiesler, S., & McGuire,
Postmes, T., Spears, R., Lea, M., & Reicher, S. T. W. (1986). Group processes in computer-
D. (2000). SIDE issues centre stage: Recent mediated communication. Organizational
developments in studies of deindividuation Behavior and Human Decision Processes,
in groups. Amsterdam: Royal Netherlands 37, 157–187.
Academy of Arts and Sciences. Sproull, L., & Kiesler, S. (1986). Reducing social
Reid, F. J. M., Ball, L. J., Morley, A. M., & context cues: Electronic mail in organiza
Evans, J. S. B. T. (1997). Styles of group tional communication. Management
discussion in computer-mediated decision Science, 32, 1492–1512.
making. British Journal of Social Psy Steinfield, C. W. (1986). Computer-mediated
chology, 36, 241–262. communication in an organizational
Reid, F. J. M., Malinek, V., Stott, C., & Evans, setting: Explaining task-related and
J. S. B. T. (1996). The messaging threshold socioemotional uses. In M. L. McLaughlin
in computer-mediated communication. (Ed.), Communication yearbook 9 (pp. 777–
Ergonomics, 39, 1017–1037. 804). Beverly Hills, CA: Sage.
Rezabek, L. L., & Cochenour, J. J. (1998). Storck, J., & Sproull, L. (1995). Through a
Visual cues in computer-mediated commu glass darkly: What do people learn in video-
nication: Supplementing text with emoti conferences? Human Communication
cons. Journal of Visual Literacy, 18, Research, 22, 197–219.
210–215. Suler, J. (1999). The psychology of avatars and
Rice, R. E. (1984). Mediated group communica graphical space in multimedia chat commu
tion. In R. E. Rice & Associates (Eds.), The nities, or, how I learned to stop worrying
24-Manusov.qxd 6/30/2006 7:35 PM Page 479
and love my Palace props. Retrieved Walther, J. B., Loh, T., & Granka, L. (2005).
March 6, 2005, from http://www.rider Let me count the ways: The interchange of
.edu/~suler/psycyber/psyav.html verbal and nonverbal cues in computer-
Tidwell, L. C., & Walther, J. B. (2002). mediated and face-to-face affinity. Journal
Computer-mediated communication effects of Language and Social Psychology, 24,
on disclosure, impressions, and interper 36–65.
sonal evaluations: Getting to know one Walther, J. B., & Parks, M. R. (2002). Cues
another a bit at a time. Human Com filtered out, cues filtered in: Computer-
munication Research, 28, 317–348. mediated communication and relationships.
Walther, J. B. (1992). Interpersonal effects in In M. L. Knapp & J. A. Daly (Eds.),
computer-mediated interaction: A relational Handbook of interpersonal communication
perspective. Communication Research, 19, (3rd ed., pp. 529–563). Thousand Oaks,
52–90. CA: Sage.
Walther, J. B. (1996). Computer-mediated com Walther, J. B., Slovacek, C., & Tidwell, L. C.
munication: Impersonal, interpersonal, and (2001). Is a picture worth a thousand
hyperpersonal interaction. Communication words? Photographic images in long term
Research, 23, 3–43. and short term virtual teams. Communica
Walther, J. B. (1997). Group and interpersonal tion Research, 28, 105–134.
effects in international computer-mediated Walther, J. B., & Tidwell, L. C. (1995).
collaboration. Human Communication Nonverbal cues in computer-mediated
Research, 23, 342–369. communication, and the effect of chrone
Walther, J. B. (in press). Selective self-presenta mics on relational communication. Journal
tion in computer-mediated communication: of Organizational Computing, 5, 355–378.
Hyperpersonal dimensions of technology, Witmer, D., & Katzman, S. (1997). On-line
language, and cognition. Computers in smiles: Does gender make a difference in
Human Behavior. the use of graphic accents? Journal of
Walther, J. B., & Bunz, U. (2005). The rules of Computer-Mediated Communication, 2(4).
virtual groups: Trust, liking, and perfor Retrieved March 16, 2006, from http://
mance in computer-mediated commu www.jcmc.indiana.edu/vol2/issue4/wit
nication. Journal of Communication, 55, mer1.html
828–846. Woodworth, M. T., Hancock, J. T., & Goorha,
Walther, J. B., & Burgoon, J. K. (1992). Relational S. (2005). The motivational enhancement
communication in computer-mediated inter effect: Implications for our chosen modes
action. Human Communication Research, of communication in the 21st century.
19, 50–88. Proceedings of the 38th Annual Hawaii
Walther, J. B., & D’Addario, K. P. (2001). The International Conference on System
impacts of emoticons on message interpre Sciences. Retrieved July 30, 2005, from
tation in computer-mediated communica http://ieeexplore.ieee.org/search/wrapper
tion. Social Science Computer Review, 19, .jsp?arnumber=1385273
323–345.
24-Manusov.qxd 6/30/2006 7:35 PM Page 480
25-Manusov.qxd 6/30/2006 7:37 PM Page 481
25
NONVERBAL COMMUNICATION,
RACE, AND INTERGROUP
INTERACTION
� John F. Dovidio
University of Connecticut
� Michelle Hebl
Rice University
� Jennifer A. Richeson
Northwestern University
� J. Nicole Shelton
Princeton University
◆ 481
25-Manusov.qxd 6/30/2006 7:37 PM Page 482
representative of that group but also as the hypothesized dynamics of intergroup iden
embodiment of that group’s values, beliefs, tity and nonverbal behavior.
and interests (Turner, Hogg, Oakes, Reicher,
& Wetherell, 1987). Thus, interpersonal
interactions under these conditions become, ♦ Group Differences in
in essence, intergroup encounters. Because
Nonverbal Behavior:
of the importance of social identity in
everyday activity (Tajfel & Turner, 1979),
Status and Race
and the need to manage the complexity and
demands of social functioning (Fiske, Lin, In this section of the chapter, we provide
& Neuberg, 1999), such intergroup inter an overview of theoretical perspectives that
actions are common. suggest why and how members of majority
This chapter examines nonverbal behav and minority groups in general may differ
ior in an intergroup context. We begin by in the types of nonverbal behaviors they
considering how members of different display regularly. The focus, thus, is on sys
groups, defined by majority and minority tematic group differences in nonverbal
status, may differ systematically in their behavior during intragroup interaction. We
intragroup nonverbal behavior. Next, we summarize the empirical literature building
focus on the dynamics of interactions between on Halberstadt’s (1985) comprehensive
members of different groups, identifying review of racial differences in nonverbal
important elements that can influence non behavior by drawing on more recent work.
verbal behavior. We present a general model Because of personal and group motiva
of “mixed social interaction” and explore tions for esteem and social dominance, status
the role of nonverbal behavior in these inter is a fundamental dimension of intergroup
actions, considering the causes and conse encounters (see Burgoon & Dunbar, this
quences of these nonverbal behaviors. We volume). Power and status are relational
conclude with a discussion of the impor concepts that are often determined contex
tance of understanding the reciprocal tually. For instance, a college senior may be
relationship between interpersonal and high status in an interaction with a college
intergroup interaction and the fundamental freshman but low status in a meeting with a
importance of nonverbal behavior in these professor. In general, high-status individu
interactions. als may have more freedom of movement
Although we examine a range of inter and thus may be more open in their pos
group contexts in this chapter, we illustrate tures and approach others more closely
the relationship between intergroup rela than do low-status individuals (Ellyson
tions and nonverbal behavior mainly within & Dovidio, 1985). Low-status individuals
the context of Black-White relations and may be more inhibited in their actions and
primarily within the cultural context of the monitor their partners more closely than do
United States. Even though there has been high-status individuals. Keltner, Gruenfeld,
substantial recognition of the importance of and Anderson (2003) hypothesize that high
culture (see Matsumoto, this volume) and power and status are associated with a gen
gender (see Hall, this volume) for nonverbal eral approach orientation, whereas lower
behavior, the literature on nonverbal power and status are related to inhibition.
behavior and intergroup relations is rela Consistent with this view, individual sta
tively sparse. Focusing on Black-White tus has been found to exert a strong influ
relations thus helps provide a coherent ence on nonverbal behavior between people
test, within a defined context, of the (see Hall, Coats, & Smith LeBeau, 2005).
25-Manusov.qxd 6/30/2006 7:37 PM Page 483
results as consistent with the oppression same-race interactions, Blacks displayed sig
hypothesis. In general, though, Whites and nificantly lower levels of visual contact than
Blacks are more accurate in decoding the did Whites. In addition, other studies of
nonverbal behavior of members of their adults (Smith, 1983) and children (Reid
own race than they are of other races et al., 1989), beyond the literature Halberstadt
(Bailey, Nowicki, & Cole, 1998; Weathers (1985) reviewed, have also shown that
et al., 2004). Greater intragroup than Blacks display lower levels of visual contact
intergroup accuracy is a function of greater than do Whites.
familiarity and more experience with in- In summary, the results of studies that
group members than with out-group mem examined racial differences in nonverbal
bers (Elfenbein & Ambady, 2002). behavior have revealed systematic differ
Racial differences in nonverbal behav ences in spatial behavior, touch, and visual
iors have been studied most commonly in contact. The pattern of findings does not
terms of proxemics (interpersonal distance provide clear support for either the oppres
and body orientation), touch, and visual sion hypothesis or Hall’s (1966) cultural
contact (gaze). Halberstadt’s (1985) review hypothesis, however. In particular, the
of studies on interpersonal distances main findings for spatial behavior (i.e., greater
tained between Blacks and between Whites interpersonal distance among Black than
showed no overall racial differences. Age White adults) are more consistent with
was an important moderating factor, oppression than with the cultural hypothe
however. Halberstadt’s review revealed that sis, the results for touch (more touching by
Black children maintained closer interper Blacks than Whites) are more consistent
sonal distances than did White children, but with the cultural hypothesis, and the find
Black adults maintained greater interper ings for visual behaviors (less eye contact by
sonal distance than did White adults. Reid, Blacks than by Whites) are inconsistent
Tate, and Berman (1989), however, found with both positions. Although racial differ
that Black children do not always maintain ences in nonverbal behavior do not con
closer distances than White children. They form uniformly to predictions derived from
found that Black children (ages 4–7) stood the oppression hypothesis or the cultural
farther away from an infant of the same race hypothesis, it is important to recognize that
than did White children. nonverbal behavior is highly sensitive to
Across the eight studies of body orienta social context. Racial differences may there
tion reviewed by Halberstadt (1985), Blacks fore be more pronounced, apparent, and
exhibited a less direct body orientation than consistent in situations in which social
did Whites when interacting with others, and identity is salient, such as in dyadic interac
this difference tended to increase with age. tions between members of different social
With respect to touch, Halberstadt’s review identity groups.
of eight studies revealed that Blacks touched
one another during their interactions more
often than Whites did. Finally, with respect ♦ Social Identity and
to visual contact, both the oppression Nonverbal Behavior in
hypothesis and E. T. Hall’s (1966) cultural Intergroup Interaction
hypothesis predict that Blacks will exhibit
higher levels of eye gaze than will Whites.
The results of Halberstadt’s (1985) analysis One of the most influential theories of
of eight visual contact studies, however, group influences on intergroup interaction
are inconsistent with both positions. In and nonverbal behavior is expectation
25-Manusov.qxd 6/30/2006 7:37 PM Page 485
states theory (Berger, Wagner, & Zelditch, latter conditions, collective needs, goals,
1985; Correll & Ridgeway, 2003). Accord and standards are primary.
ing to this sociological framework, interac Social categorization and social identity
tions between members of different groups can influence the nature of communication,
are accompanied by differential expecta and thus nonverbal behavior, through a
tions about the status of the interactants in broad range of cognitive, affective, and
cases in which group membership is associ motivational mechanisms. Social catego
ated systematically with prestige and status rization activates, often without awareness
in everyday life. Berger et al. (1985) or control, stereotypic associations that can
referred to characteristics of individuals influence expectations of the encounter and
that give rise to differential status expecta online attributions of the behavior of the
tions as “diffuse status characteristics.” other person in the interaction (see Lakin,
These expectations, in turn, can generalize this volume). In addition, the general evalu
to a broad range of situations and, through ative biases that accompany recognition of
a process of behavioral confirmation of different group memberships can produce
expectancies similar to the self-fulfilling general approach or avoidance tendencies
prophecy, affect power-related behavior that are systematically reflected in nonver
and perceptions across a variety of social bal behaviors. These prejudices may be bla
contexts. Expectation states theory has tant or subtle (Gaertner & Dovidio, 1986),
received substantial empirical support with and people who have explicitly nonpreju
respect to interaction between women and diced attitudes may still harbor implicit
men, and it has received some support in intergroup biases (Dovidio & Gaertner,
the realm of interracial behavior (see Berger 2004).
et al., 1985; Correll & Ridgeway, 2003). Recognition of different group member
Whereas expectation states theory views ships in the interaction also typically arouses
the effect of group membership on interper intergroup anxiety (Stephan & Stephan,
sonal interaction as a relatively rational, 1985, 2000). Part of this anxiety may be due
albeit not necessarily conscious, process of to uncertainty about how to behave in this
evaluating the relative contributions of intergroup context, and part may be a func
interactants in creating social organization, tion of feelings of real or symbolic threat. In
psychological perspectives have posited a addition to anxiety, intergroup interactions
much more pervasive and fundamental can arouse a number of different motiva
influence of group membership and iden tions. For members of majority groups, for
tity. Social identity theory (Tajfel & Turner, instance, intergroup interactions can arouse
1979) and self-categorization theory (Turner dominance orientations (Sidanius & Pratto,
et al., 1987) view the distinction between 1999) or the desire to appear nonprejudiced
personal identity and social identity as (Gaertner & Dovidio, 1986). For members of
critical. According to these theories, when minority groups, the desire to detect and
personal identity is salient, a person’s potentially compensate for anticipated dis
individual needs, standards, beliefs, and crimination by majority-group interaction
motives determine behavior. In contrast, partners may be activated (Miller & Myers,
when social identity is salient, “people 1998; Shelton, Richeson, & Salvatore, 2005).
come to perceive themselves as more inter In general, these positions suggest that
changeable exemplars of a social category majority-group members would be likely to
than as unique personalities defined by display nonverbal behaviors associated with
their individual differences from others” lower levels of liking or attraction and with
(Turner et al., 1987, p. 50). Under these higher levels of social dominance or power
25-Manusov.qxd 6/30/2006 7:37 PM Page 486
6/30/2006
S8
Type of Motivations
Stigma and Goals
S10
Personal Nonverbal
7:37 PM
S7 Behaviors
S3 Behavioral
Predisposition S6 Assessment of
S4 Meeting Goals
Experiential S2
S1 Affect and Verbal
Page 487
Arousal Assessment
S5 Behaviors of Interactant
Relational-
Situational Stereotypes
and Cognitions Coping Costs/Benefits
Strategies
S9
STIGMATIZER
Termination
TARGET T9
Relational- Coping Costs/Benefits
Situational Stereotypes Strategies
and Cognitions
Assessment
T1 Affect and T5 Verbal of Interactant
Experiential Arousal Behaviors
T2 T4 Assessment of
T10 Behavioral
T3 T6 Meeting Goals
Personal Predisposition T7 Nonverbal
Behaviors
Type of Motivations
Stigma and Goals
T8
Figure 25.1 Hebl and Dovidio’s (2005) Model of Mixed Social Interactions Based on Patterson’s (1982) Sequential Functional Model of Nonverbal
◆
Exchange
487
25-Manusov.qxd 6/30/2006 7:37 PM Page 488
applicants than with the nonstigmatized majority-group members, the less frequent
applicants. research conducted on minority-group
Systematic differences in communication members suggests that they, too, behave in
are also evident in other important types ways that influence intergroup interaction.
of intergroup encounters. Johnson, Roter, In particular, minority-group members
Powe, and Cooper (2004), for example, attempt to act in strategic ways to manage
found that physicians displayed greater ver intergroup interactions. These behaviors
bal dominance, less positive affect, and less may take a number of different forms, but
patient-centered communication with Black they are characterized generally as ways to
patients than with White patients. It is cope with actual or anticipated discrimina
perhaps because of such differences in tion. Majority and minority members engage
communication orientation that same-race in coping strategies in intergroup interac
interactions between physicians and patients tions. They are adjusting to different types of
(Cooper et al., 2003), and same-race inter threats, however, and thus the methods and
actions between teachers and students consequences of coping are quite different.
(Feldman & Donohoe, 1978) are experi As suggested by Gaertner and Dovidio
enced more positively than cross-race (1986) (see also Dovidio & Gaertner, 2004),
interactions. Whites commonly find interracial interac
Importantly, subtle differences in these tion to be anxiety arousing and, thus, they
types of interaction behaviors can exert sig may avoid interracial interactions when pos
nificant impact on the nature and outcomes sible. When they cannot avoid these inter
of intergroup interactions. Specifically, non actions, they may become focused on not
verbal behavior is an important mechanism acting inappropriately, particularly in ways
in self-fulfilling prophecies. Word et al. that can be attributed to racism.
(1974) demonstrated that Whites exhibited Preoccupation with behaving in a non-
less immediate, more negative nonverbal prejudiced manner can further contribute to
behaviors when interviewing Black relative inconsistencies in Whites’ verbal and nonver
to White confederates. In addition, they bal behaviors in interracial interaction. Hebl
showed that participants with interviewers and Dovidio (2005) found in their review
who showed low-immediacy behaviors that across a range of different types of
(behaviors mirroring Whites’ interactions interactions, stigmatizers’ (i.e., members of
with Blacks), compared with participants majority or socially dominant groups) dis
with interviewers who displayed high- play of negative nonverbal behaviors was
immediacy behaviors, responded in a less frequently at odds with their verbal behav
favorable and responsive way during the iors (see paths S4 and T4 in Figure 25.1). In
interaction and were judged by independent general, stigmatizers often report feeling pos
raters as less suitable for the position. Thus, itively toward targets, whereas their nonver
the nonverbal behaviors of the interviewers, bal and paraverbal behaviors indicate more
which reflected the differences displayed in negative reactions. This divergence between
interactions with Blacks and Whites, elicited self-reported favorable orientations and
complementary verbal and nonverbal respon negative nonverbal behaviors is typically
ses from the interviewees. This study illus observed for Whites in interracial interac
trates the powerful role of nonverbal tions (see Crosby, Bromley, & Saxe, 1980).
behavior in perpetuating racial disparities. Because Whites may be concerned about
Whereas researchers have amassed a great acting in a prejudiced or otherwise inappro
many insights regarding the perspective of priate way in interracial interactions
25-Manusov.qxd 6/30/2006 7:37 PM Page 493
(Gaertner & Dovidio, 1986), they may focus Repeated disengagement may lead to a
the majority of their attention on managing more long-term strategy of disidentifica
their verbal behaviors, which can be easier to tion, in which Black individuals disengage
monitor and control than nonverbal behav permanently from the domain of evalua
iors. Moreover, to the extent that monitor tion. Disengagement may be one reason
ing and controlling of verbal responses why Blacks tend to be less emotionally
involve high cognitive demand, these activi expressive in interracial interaction than are
ties may actually facilitate the expression Whites (Ickes, 1984).
of more spontaneous responses (see also Alternatively, minority-group members,
Patterson, 1995). As a consequence, Whites such as U.S. Blacks, often cope with antici
(and other types of stigmatizers) may be less pated discrimination by either being
adept at managing affect-driven behaviors particularly vigilant to cues of bias or com
that occur in interactions spontaneously and pensating for potential bias. With respect to
without time for deliberation. ethnic minority/White relations, daily
Vorauer and Turpie (2004) found simi encounters with potential discrimination
lar effects in interactions between Canadian may lead ethnic minorities to interpreta
majority- (White) and minority- (First tions that confirm and reconfirm that prej
Nations) group members. Whereas higher udice exists and to label ambiguous
evaluative concerns reduced bias in intimacy- behaviors as discriminatory (Sellers &
building behaviors (e.g., eye contact, self- Shelton, 2003). Instead of monitoring for
disclosure) among high-prejudiced Whites, bias that is occurring, minorities can also
higher evaluative concerns interfered with compensate for potential bias before it has
intimacy-building behaviors among low- the opportunity to affect the interracial
prejudiced Whites. Vorauer and Turpie interactions in which they engage (Miller &
interpreted these results as low-prejudiced Myers, 1998). Specifically, when they are
Whites “choking” under the pressure of concerned about the potential bias of their
high evaluative concerns. White interaction partners, ethnic minority
Blacks’ coping strategies in interracial participants often engage in compensatory
interaction are directed generally at coping strategies, such as smiling and talking more,
with actual, perceived, or anticipated bias. to ward off potentially negative outcomes
Like Whites, Blacks may also avoid interra (Shelton et al., 2005). Thus, coping styles
cial interactions when possible (Patchen, can influence the assessment processes iden
1983), but when avoidance is not an tified in the secondary process phase sys
option, they can make use of several coping tematically (see Figure 25.1).
strategies, including disengagement, vigi
lance, and compensation. Disengagement
involves limiting the extent to which feel SECONDARY INTERACTION PHASE
ings of self-worth are dependent on feed
back within the interaction. If Blacks After individuals have exchanged verbal
disengage, they do not allow the biases of and nonverbal behaviors, they are likely to
Whites to influence them in substantially assess the interaction in an attempt to con
negative, or at least direct, ways. As a con tinue or terminate it. In deciding which
sequence, Blacks may be less responsive course to pursue, both individuals often
to feedback, either positive or negative, engage in an assessment of their goals, the
from Whites than are Whites in interac other interactant, and their outcomes. In
tions (Major, Quinton, & McCoy, 2002). terms of assessing one’s goals, interactants,
25-Manusov.qxd 6/30/2006 7:37 PM Page 494
for example, might examine their social behaved in the interaction were essentially
goals (e.g., did they make a good impres uncorrelated. Other researchers have
sion on their interaction partner?) or their shown similarly that Whites’ implicit inter
task-oriented goals (e.g., did they get the group attitudes and stereotypes, of which
job or other outcome they sought?). In they have limited awareness, predict the
terms of assessing the interactant, interac impressions that ethnic minorities form of
tants both evaluate and judge each other them during interactions (Fazio, Jackson,
actively during the interaction (e.g., what Dunton, & Williams, 1995; McConnell
does the other person think of me and my & Leibold, 2001; Sekaquaptewa, Espinoza,
contributions?). Thompson, Vargas, & von Hippel, 2003).
The dissociation of majority-group Because of heightened awareness and
members’ explicit and implicit attitudes and anticipated rejection in interracial interac
their consequent effect on verbal and non tions, Whites tend to overestimate the
verbal behavior can produce a significant extent to which racial minorities will per
divergence in their self-assessments and how ceive their behavior as friendly (Vorauer
they are evaluated by their interaction part & Sakamoto, 2005). Furthermore, both
ners. Dovidio et al. (2002) showed that Whites and Blacks often misinterpret anxi
Whites’ explicit attitudes and verbal behav ety-related behaviors, such a shorter gaze
iors were related, whereas their implicit atti durations and more frequent self-touching,
tudes and nonverbal behaviors were related. as signals of unfriendliness more frequently
Specifically, Whites’ explicit racial attitudes in interracial than in intraracial encounters
predicted the positivity of their verbal com (Devine & Vasquez, 1998; Dovidio &
munications with Black interaction part Johnson, 2005). This vigilance may pro
ners, but their implicit racial attitudes duce more accurate sensitivity to racial bias
predicted the positivity of their nonverbal among Blacks, however (Rollman, 1978).
communication. In addition, Whites based Richeson and Shelton (2005) found, for
their impressions of how friendly they instance, that Black judges (college student
behaved during their interracial interactions participants) were, on average, better able
on the attitudes that were accessible (i.e., to detect both the explicit and the implicit
their explicit attitudes) and the behaviors racial bias levels of White individuals from
that they could readily monitor (i.e., their 20 seconds of their nonverbal behavior dur
verbal behaviors). Because most of the ing interracial interactions than were White
Whites in the study perceived themselves as judges. Specifically, Black judges’ ratings
nonprejudiced, they generally believed that of how positively a sample of White targets
they behaved in a friendly and unbiased behaved during an interracial interaction
manner toward their Black partners. were more highly correlated (albeit nega
Blacks, in contrast, relied on their White tively) with those targets’ automatic racial
interaction partners’ nonverbal behavior in bias scores than were the ratings made by
making their assessments of how friendly White judges. Furthermore, Black judges’
their partners behaved. Because Whites’ ratings of the White targets’ prejudice levels
nonverbal behavior was correlated with were more highly correlated with those
their implicit attitudes that were, on aver targets’ explicit prejudice scores than were
age, negative, Blacks often left the interac the same ratings made by White judges.
tion with a negative perception of their In addition to assessing their goals and
White partners. Thus, Whites’ and Blacks’ their interaction partner, individuals
assessments of how the White person engage in ongoing assessment of the costs
25-Manusov.qxd 6/30/2006 7:37 PM Page 495
Blacks in forming their impressions can lead involved in intergroup as well as intragroup
to vastly divergent views not only about their interactions. Building on Patterson’s (1982)
interpersonal relations but also ultimately to Sequential Functional Model of Nonverbal
race relations in general. These dynamics pro Exchange, we presented a model of mixed
vide insight into why Blacks and Whites view social interaction (Hebl & Dovidio, 2005)
race relations so differently. For instance, in that outlines key elements in the dynamics
the United States, whereas most (69%) of of communication in an intergroup context.
Whites perceive that Blacks are treated “the Thus, although nonverbal behavior can be
same as Whites”, the majority of Blacks studied in terms of separate encoding and
(59%) report that Blacks are treated worse decoding processes, the dynamic nature
than Whites (Gallup Organization, 2002). of nonverbal communication can best be
Understanding the role of nonverbal behav studied during interactions. Unfortunately,
ior in interaction can thus provide funda studies of actual intergroup interaction,
mental insights for understanding and including measures of verbal and nonverbal
improving intergroup relations. behavior, remain all too rare. Yet we believe
that this type of research, along with appro
priate theoretical development, is crucial to
understanding the important role of nonver
♦ Conclusion
bal behavior in intergroup communication
and ultimately intergroup relations.
Despite its obvious practical importance
and theoretical value, nonverbal behavior
in intergroup contexts is a curiously under ♦ References
studied topic. In 1985, Halberstadt observed
specifically that “the first research on race
and socioeconomic differences in nonverbal Ashburn-Nardo, L., Knowles, M. L., &
Monteith, M. J. (2003). Black Americans’
communication was conducted in the
implicit racial associations and their impli
1930s . . . , but interest in these issues was
cations for intergroup judgment. Social
not sustained until the early 1970s” Cognition, 21, 61–87.
(p. 228). Our review of the literature Bailey, W., Nowicki, S., Cole, S. P. (1998).
reveals that the interest in race and nonver The ability to decode nonverbal informa
bal communication peaked in the 1970s; tion in African American, African, and
research activity on this topic has waned Afro-Caribbean and European American
since then. The bulk of this seminal work adults. Journal of Black Psychology, 24,
on nonverbal behavior, race, and inter 418–431.
group interaction focuses on the contention Berger, J., Wagner, D. G., & Zelditch, M.
that majority- and minority-group members, (1985). Introduction: Expectation states
perhaps because of differences in status, theory. In J. Berger & M. Zelditch
Jr. (Eds.), Status, rewards, and influence
social power, and stigmatization, are likely
(pp. 1–71). San Francisco: Jossey-Bass.
to develop different styles of communica
Blascovich, J., Mendes, W. B., Hunter, S. B., Lickel,
tion, reflected in nonverbal and verbal B., & Kowai-Bell, N. (2001). Perceiver threat
behaviors. The data, however, are sugges in social interactions with stigmatized individ
tive but not yet conclusive. uals. Journal of Personality and Social
We propose that a comprehensive under Psychology, 80, 253–267.
standing of nonverbal behavior requires a Caporael, L. R. (1981). The paralanguage of care-
consideration of the complex processes giving: Baby talk to the institutionalized
25-Manusov.qxd 6/30/2006 7:37 PM Page 497
Fazio, R. H., & Olson, M. A. (2003). Implicit Hall, J. A., Coats, E. J., & Smith LeBeau, L.
measures in social cognition research: Their (2005). Nonverbal behavior and the verti
meaning and uses. Annual Review of cal dimension of social relations: A meta
Psychology, 54, 297–327. analysis. Psychological Bulletin, 131,
Feldman, R. S. (1985). Nonverbal behavior, 898–924.
race, and the classroom teacher. Theory Hebl, M. R., & Dovidio, J. F. (2005). Promoting
Into Practice, 24, 45–49. the “social” in the examination of social
Feldman, R. S., & Donohoe, L. F. (1978). stigmas. Personality and Social Psychology
Nonverbal communication of affect in Review, 9, 156–182.
interracial dyads. Journal of Educational Hebl, M. R., Foster, J., Mannix, L. M., &
Psychology, 70, 979–987. Dovidio, J. F. (2002). Formal and interper
Fiske, S. T., Lin, M., & Neuberg, S. L. (1999). sonal discrimination: A field study under
The continuum model: Ten years later. In standing of applicant bias. Personality and
S. Chaiken & Y. Trope (Eds.), Dual process Social Psychological Bulletin, 28, 815–825.
theories in social psychology (pp. 231– Hebl, M., Tickle, J., & Heatherton, T. F. (2000).
254). New York: Guilford. Awkward moments in interactions between
Frable, D. E. S., Blackstone, T., & Scherbaum, nonstigmatized and stigmatized individuals.
C. (1990). Marginal and mindful deviants In T. F. Heatherton, R. E. Kleck, M. R.
in social interactions. Journal of Personality Hebl, & J. G. Hull (Eds.), Stigma: Social
and Social Psychology, 59, 140–149. psychological perspectives (pp. 273–306).
Fugita, S. S., Wexley, K. N., & Hillery, J. M. New York: Guilford Press.
(1974). Black–white differences in nonver Henley, N. M. (1977). Body politics: Power,
bal behavior in an interview setting. sex, and nonverbal communication. Engle
Journal of Applied Social Psychology, wood Cliffs, NJ: Prentice Hall.
4, 343–350. Hyers, L. L., & Swim, J. K. (1998). A compari
Gaertner, S. L., & Dovidio, J. F. (1986). The son of experiences of dominant and minor
aversive form of racism. In J. F. Dovidio & ity group members during an intergroup
S. L. Gaertner (Eds.), Prejudice, discrimina encounter. Group Processes and Intergroup
tion, and racism (pp. 61–89). Orlando, FL: Relations, 12, 143–163.
Academic Press. Ickes, W. (1984). Compositions in black and
Galinsky, A. D., Gruenfeld, D. H., & Magee, J. C. white: Determinants of interaction in inter
(2003). From power to action. Journal of racial dyads. Journal of Personality and
Personality and Social Psychology, 85, Social Psychology, 47, 330–341.
453–466. Johnson, R. L., Roter, D., Powe, N. R., &
Gallup Organization. (2002). Poll topics & Cooper, L. A. (2004). Patient race/ethnicity
trends: Race relations. Washington, DC: and quality of patient–physician communi
Author. Retrieved October 3, 2004, from cation during medical visits. American
http://www.gallup.com/poll/topics/race.asp Journal of Public Health, 94, 2084–2090.
Guinote, A., Judd, C. M., & Brauer, M. (2002). Jones, J. M. (1986). Racism: A cultural analysis
Effects of power on perceived and objective of the problem. In J. F. Dovidio & S. L.
group variability: Evidence that more Gaertner (Eds.), Prejudice, discrimination,
powerful groups are more variable. Journal and racism (pp. 279–314). Orlando, FL:
of Personality and Social Psychology, 82, Academic Press.
708–721. Keltner, D., Gruenfield, D. H., & Anderson, C.
Halberstadt, A. G. (1985). Race, socioeconomic (2003). Power, approach, and inhibition.
status, and nonverbal behavior. In A. W. Psychological Review, 110, 265–284.
Siegman & S. Feldstein (Eds.), Multichannel Kemper, S., Vandeputte, D., Rice, K., Cheung,
integrations of nonverbal behavior (pp. 227– H., & Gubarchuk, J. (1995). Spontaneous
266). Hillsdale, NJ: Erlbaum. adoption of elderspeak during referential
Hall, E. T. (1966). The hidden dimension. New communication tasks. Journal of Language
York: Doubleday. and Social Psychology, 14, 40–59.
25-Manusov.qxd 6/30/2006 7:37 PM Page 499
LaFrance, M., & Henley, N. M. (1994). On situations with infants: Effects of race
oppressing hypotheses: Or differences in and sex. Child Development, 60,
nonverbal sensitivity revisited. In H. L. 710–714.
Radtke & H. J. Stam (Eds.), Power/gender: Richeson, J. A., & Ambady, N. (2003). Effects
Social relations theory and practice of situational power on automatic racial
(pp. 287–311). London: Sage. prejudice. Journal of Experimental Social
Major, B., Quinton, W., & McCoy, S. (2002). Psychology, 39, 177–183.
Antecedents and consequences of attribu Richeson, J. A., & Shelton, J. N. (2003). When
tions to discrimination: Theoretical and prejudice does not pay: Effects of interracial
empirical advances. In M. P. Zanna contact on executive function. Psychological
(Ed.), Advances of Experimental Social Science, 14, 287–290.
Psychology (Vol. 34, pp. 251–330). New Richeson, J. A., & Shelton, J. N. (2005). Thin
York: Academic Press. slices of racial bias. Journal of Nonverbal
McConnell, A. R., & Leibold, J. M. (2001). Behavior, 29, 75–86.
Relations among the Implicit Association Richeson, J. A., & Trawalter, S. (2005). Why
Test, discriminatory behavior, and explicit do interracial interactions impair execu
measures of racial attitudes. Journal of tive function? A resource depletion
Experimental Social Psychology, 37, account. Journal of Personality and Social
435–442. Psychology, 88, 934–947.
Miller, C. T., & Myers, A. M. (1998). Rollman, S. A. (1978). The sensitivity of Black
Compensating for prejudice: How heavy and White Americans to nonverbal cues
weight people (and others) control out of prejudice. Journal of Social Psychology,
comes despite prejudice. In J. K. Swim & 105, 73–77.
C. Stangor (Eds.), Prejudice: The target’s Sekaquaptewa, D., Espinoza, P., Thompson, M.,
perspective (pp. 191–218). San Diego, Vargas, P., & von Hippel, W. (2003).
CA: Academic Press. Stereotypic explanatory bias: Implicit
Patchen, M. (1983). Students’ own racial atti stereotyping as a predictor of discrimina
tudes and those of peers of both races, as tion. Journal of Experimental Social
related to interracial behaviors. Sociology Psychology, 39, 75–82.
and Social Research, 68, 59–77. Sellers, R. M., & Shelton, J. N. (2003). The role
Patterson, M. L. (1982). A sequential functional of racial identity in perceived racial discrim
model of nonverbal exchange. Psychological ination. Journal of Personality and Social
Review, 89, 231–249. Psychology, 84, 1079–1092.
Patterson, M. L. (1995). Invited article: A parallel Shelton, J. N. (2003). Interpersonal concerns
process model of nonverbal communication. in social encounters between majority
Journal of Nonverbal Behavior, 19, 3–29. and minority group members. Group
Pettigrew, T. F., & Tropp, L. R. (2000). Does Processes & Intergroup Relations, 6,
intergroup contact reduce prejudice? Recent 171–185.
meta-analytic findings. In S. Oskamp (Ed.), Shelton, J. N., Richeson, J. A., & Salvatore, J.
Reducing prejudice and discrimination (2005). Expecting to be the target of preju
(pp. 93–114). Hillsdale, NJ: Erlbaum. dice. Implications for interethnic interac
Pinel, E. C. (1999). Stigma consciousness: tions. Personality and Social Psychology
The psychological legacy of social stereo Bulletin, 39, 1189–1202.
types. Journal of Personality and Social Sidanius, J., & Pratto, F. (1999). Social
Psychology, 76, 114–128. dominance: An intergroup theory of social
Plant, E. A., & Devine, P. G. (1998). Internal hierarchy and oppression. New York:
and external motivation to respond without Cambridge University Press.
prejudice. Journal of Personality and Social Smith, A. (1983). Nonverbal communication
Psychology, 75, 811–832. among Black female dyads: An assessment
Reid, P. T., Tate, C. S., & Berman, P. W. (1989). of intimacy and race. Journal of Social
Preschool children’s self-presentations in Issues, 39, 55–67.
25-Manusov.qxd 6/30/2006 7:37 PM Page 500
Stangor, C., Sullivan, L. A, & Ford, T. E. Rediscovering the social group: A self-
(1991). Affective and cognitive determi categorization theory. Oxford, UK: Basil
nants of prejudice. Social Cognition, 9, Blackwell.
359–380. Vorauer, J., Main, K. J., & O’Connell, G. B.
Stephan, W., & Stephan, C. W. (1985). (1998). How do individuals expect to be
Intergroup anxiety. Journal of Social Issues, viewed by members of lower status groups?
41, 157–175. Content and implications of meta
Stephan, W. G., & Stephan, C. W. (2000). An stereotypes. Journal of Personality and
integrated threat theory of prejudice. In Social Psychology, 75, 917–937.
S. Oskamp (Ed.), Reducing prejudice and Vorauer, J. D., & Sakamoto, Y. (2005).
discrimination: The Claremont Symposium I thought we could be friends, but
on Applied Psychology (pp. 23–45). . . . . Systematic miscommunication and
Mahwah, NJ: Erlbaum. defensive distancing as obstacles to cross-
Tajfel, H., & Turner, J. C. (1979). An integra group friendship formation. Unpublished
tive theory of intergroup conflict. In W. G. manuscript.
Austin & S. Worchel (Eds.), The social psy Vorauer, J. D., & Turpie, C. (2004). Disruptive
chology of intergroup relations (pp. 33–48). effects of vigilance on dominant group
Monterey, CA: Brooks/Cole. members’ treatment of outgroup members:
Tiedens, L. Z. (2000). Powerful emotions: The Choking versus shining under pressure.
vicious cycle of social status positions and Journal of Personality and Social Psy
emotions. In N. Ashkanasy, W. Zerbe, & C. chology, 27, 706–709.
Hartel (Eds.), Emotions in the workplace: Weathers, M. D., Kitsantas, P., Lever, T.,
Research, theory, and practice (pp. 71–81). O’Brien, P., Campbell, S., & Rastatter, M.
Westport, CT: Quorum Books. (2004). Recognition accuracy and reaction
Tiedens, L. Z., & Fragale, A. R. (2003). Power time of vocal expressions of emotion by
moves: Complementarity in dominant and African-American and Euro-American col
submissive nonverbal behavior. Journal of lege women. Perceptual and Motor Skills,
Personality and Social Psychology, 3, 99, 662–668.
558–568. Weitz, S. (1972). Attitude, voice, and behavior:
Tropp, L. R. (2003). The psychological impact A repressed affect model of interracial
of prejudice: Implications for intergroup interaction. Journal of Personality and
contact. Group Processes and Intergroup Social Psychology, 24, 14–21.
Relations, 6, 131–149. Word, C. O., Zanna, M. P., & Cooper, J.
Turkstra, L., Ciccia, A., & Seaton, C. (2003). (1974). The nonverbal mediation of self-
Interactive behaviors in adolescent conver fulfilling prophecies in interracial interac
sation dyads. Language, Speech, and tion. Journal of Experimental Social
Hearing Services in Schools, 34, 117–127. Psychology, 10, 109–120.
Turner, J. C., Hogg, M. A., Oakes, P. J.,
Reicher, S. D., & Wetherell, M. S. (1987).
26-Manusov.qxd 6/30/2006 5:21 PM Page 501
26
USES AND CONSEQUENCES OF
NONVERBAL COMMUNICATION
IN THE CONTEXT OF
ORGANIZATIONAL LIFE
� Martin S. Remland
West Chester University
◆ 501
26-Manusov.qxd 6/30/2006 5:21 PM Page 502
group (e.g., unique styles of dress) and indi (e.g., “I’m telling you to do this”), the
cating their affiliation with the organization greater ambiguity of nonverbal displays
(e.g., dressing like everyone else). People also carries the potential to produce unintended
confront and cope with discriminatory prac consequences that interfere with the goals
tices based on a person’s race, sex, age, and of an organization. Moreover, because we
so forth. are generally less aware of these “below
In this chapter, I discuss those challenges the radar” signals, the consequences may
about which there is sufficient empirical be especially difficult to avoid. The section
research of use to scholars and practition below addresses three possible conse
ers. The first major set of challenges, which quences of status cues: dysfunctional lead
stems most directly from the identification ership, sexual harassment, and workplace
and relationship functions of nonverbal discrimination.
communication, is how to manage the dis
plays of status that are such an integral part
of organizational life. The second major set STATUS DISPLAYS AND
of challenges deals with the emotion func DYSFUNCTIONAL LEADERSHIP
tion and focuses on the impact of emotional
Social exchange theory provides a useful
exchanges on task performance. The third
framework for appreciating how status
major set of challenges addresses the
differentials between leaders and followers
delivery function, which involves the
can result in a leader’s loss of power (Blau,
coordination and integration of verbal and
1964; Jacobs, 1970; see also Burgoon &
nonverbal channels of communication.
Dunbar, this volume). Based on social
Table 26.1 offers one view of how these
exchange theory principles, a leader can lose
communication challenges emerge from the
influence over subordinates when subordi
interplay of context and function. The table
nates begin to assess their relationship with
also serves to identify and organize much of
a leader as more costly than rewarding—
the research that has been done on the uses
that is, in a state of disequilibrium. A social
and consequences of nonverbal communi
exchange view of leadership emphasizes the
cation in organizations.
interactive or transactional nature of the
leadership process. In each superior-subor
♦ Managing Nonverbal dinate interaction, an exchange of resources
is negotiated in a way that is perceived as
Displays of Status
equitable by both parties. Effective com
munication allows the leader to keep the
Research suggests that the nonverbal com exchange in a state of equilibrium. Jacobs
munication of high-status persons differs in (1970) suggests, for example, that supervi
fundamental ways from that of lower-sta sors avoid interactions “that make evident
tus persons (Andersen & Bowman, 1990; power or status differentials” (p. 237).
Burgoon & Dunbar, this volume; Edinger The interactive nature of social exchange
& Patterson, 1983; Hall, 2005). These non theory suggests that the nonverbal commu
verbal displays of status serve a useful func nication of both superior and subordinate
tion: They clarify and reinforce the role should shape perceptions of leadership, as
relationships that exist in an organization, the status displays of either person can steer
helping to sustain the organization’s hierar the relationship into a state of disequilib
chy. But unlike verbal reminders of status rium (Remland, 1981). Whereas there is no
26-Manusov.qxd 6/30/2006 5:21 PM Page 503
lack of research on the effects of a leader’s however, Remland (1984) produced four
nonverbal cues (e.g., Heintzman, Leathers, videotapes of the same two male actors,
Parrott, & Cairns, 1993; Remland, 1984; role-playing a scene in which a superior
Richmond & McCroskey, 2000), there is reprimands his subordinate. Although the
little regarding the impact of a subordi script was the same in each role play, the
nate’s nonverbal behavior. In one study, actors altered their nonverbal cues so that
26-Manusov.qxd 6/30/2006 5:21 PM Page 504
each had a high-status and a low-status person spoke. As the researchers point out,
performance. In the high-status (HS) perfor “this seemingly paradoxical pattern is
mance, they used a relaxed posture, indirect understandable if the [high-status] person is
body orientation, loud voice, inattentive motivated to downplay his or her own sta
behavior, and an act of spatial invasion. In tus in the service of comfortable social
the low-status (LS) performance, they used a interaction by (as one example) encourag
tense posture, direct body orientation, soft ing the partner to speak more” (Hall &
and hesitating speech, and attentive gaze. Friedman, 1999, p. 1088). But, despite the
Not surprisingly, judges rated the supe benefit of minimizing status differentials,
rior as more considerate when he used LS some research suggests that leaders often
behaviors than when he used HS behaviors. prefer asymmetrical relations. For example,
But they also rated him as more considerate Yukl (2002) reports studies showing a pos
when the subordinate he interacted with itive correlation between ingratiation by
used HS behaviors rather than LS behav subordinates and leaders’ liking for those
iors. This finding suggests that observers subordinates. Moreover, a recent study
may perceive a male leader as more consid found that persons in low-status positions
erate when status differentials are reduced, often choose low-status behavior, which
either from a reduction in the leader’s dis may be more “comfortable” for them,
plays of status, or an increase in the subor when interacting with higher-status persons
dinate’s. Perhaps the leader receives some in task-oriented encounters (Tiedens &
“credit” for allowing or encouraging subor Fragale, 2003).
dinates to behave in a high-status manner, The maintenance of status differentials is
empowering them in the process. Unfortu not just about exchange, however. It is also
nately, however, the results of this experi a cultural artifact. Hofstede (1982) main
ment do not permit any inferences about tains that cultures classified as high in
how subordinates judge such interactions “power distance” tend to embrace authori
with their superiors, a central component in tarian values and encourage actions that
the social exchange hypothesis. perpetuate status distinctions. For example,
Some research indicates that the nonver Kowner and Wiseman (2003) asked
bal communication of leaders frequently Japanese and American participants to imag
includes a mix of high- and low-status cues ine various interactions between high-status
that reduces the differential that exists and lower-status individuals. Although
between them and their subordinates. In there was considerable agreement on the
one study, Hall and Friedman (1999) found specific behaviors differentiating high- from
that higher-status persons spoke more, used low-status persons, the magnitude of the dif
more hand gestures, and leaned forward ferences varied, with Japanese, a more hier
less than did lower-status persons. But the archical, collectivistic, and high-context
higher-status persons also nodded more people, reporting greater differences than did
frequently. Unlike the first set of behaviors, Americans, representing a more egalitarian,
all of which signal higher status, head nod individualistic, and lower context society.
ding, which implies attentiveness, agree Thus, what seems “excessive” in one culture
ment, or the desire for approval, tends to may seem quite ordinary in another.
signal lower status. The challenge of managing displays of
One especially interesting finding in the status is complicated further by the effects
study was that the greater the disparity was of sex-role stereotyping and the correspond
between the high-status person and the ing claim that sex constitutes a diffuse status
low-status person, the less the high-status characteristic, with women viewed as lower
26-Manusov.qxd 6/30/2006 5:21 PM Page 505
in status (Lockheed & Hall, 1976). A sizeable judged as less effective by males than were
body of research shows how status takes root male speakers who used the same task style.
in the communication behavior of men and Moreover, when female speakers injected
women. Specifically, and although contested some warmth and friendliness into their
in the larger literature, research shows that presentations (a social style) they were
women’s nonverbal communication differs more persuasive with male judges than
to some degree from that of men’s along the when they used the “cooler” task style; this
dimension of status and power (Hall, 1984, was not true for the male speakers. The
this volume; Henley, 1995). But researchers male judges also rated female speakers
also find that many of these differences dis using the task style as less likeable and more
appear when women assume positions of threatening than the male speakers who
leadership or possess levels of power equal to used the same style. Although the results of
those of men. That is, the influence of author this study show that women were, in this
ity and power on nonverbal communication case, better off using a task-oriented style
may be greater than that of gender (Johnson, than one that highlights feminine (submis
1994; Dovidio, Ellyson, Keating, Heltman, sive) or masculine (dominant) traits, it still
& Brown, 1988). In addition, women’s reveals the presence of a double standard:
nonverbal cues become more “powerful” For the same performance, women appar
than men’s when men and women work ently get less credit from men than do their
together on “feminine” tasks (Dovidio, male counterparts.
Brown, Heltman, Ellyson, & Keating, 1988). Almost certainly there is some link
Research also shows that women use more between nonverbal displays of status and
assertive (high status) nonverbal communica leadership success. Social exchange theory
tion, such as a more confident tone of voice, raises one possibility for this link: Dysfuncti
when they interact with superiors and subor onal leadership results from asymmetrical
dinates compared with their interactions with patterns of nonverbal communication in
peers (Steckler & Rosenthal, 1985). superior-subordinate interactions that favor
But women in leadership positions may the superior or the subordinate to excess,
still face the challenge of overcoming creating a state of disequilibrium. Yet despite
sex-role expectations. Carli, LaFleur, and the intuitive nature of this proposition, it has
Loeber (1995) compared the effectiveness of not yet been adequately tested. Research on
a task style of nonverbal communication gender differences in nonverbal communica
(competent) with that of a dominant style, a tion suggests another possibility: that female
submissive style, and a social style (friendly leaders may be judged more harshly than
and competent). They prepared videotapes their male counterparts for using the same
of a male or female speaker using one of the displays of status. But researchers are still a
four styles to deliver the same persuasive long way from identifying systematically the
message to a seated listener. The judges who conditions under which organizational
watched the tapes were persuaded most by members are most likely to impose such
male and female speakers when those speak double standards.
ers used the task and social styles.
Contrary to expectations, female speak
ers were not penalized more than male STATUS DISPLAYS AND
speakers were for using a dominant (i.e., SEXUAL HARASSMENT
masculine) style: Male and female speakers
using this style were equally ineffective. But Another potential workplace problem
female speakers who used a task style were involving nonverbal displays of status is
26-Manusov.qxd 6/30/2006 5:21 PM Page 506
sexual harassment. Although many cases there is some degree of ambiguity in the
of harassment involve the deliberate abuse actions of either party (Jones & Remland,
of authority and power, other cases likely 1997; for an application to courtship con
occur because one person misses or mis texts, see Noller, this volume).
reads the signals of another. For example, The danger of being misunderstood is
a recent experiment by Woodzicka and particularly acute in asymmetrical relation
LaFrance (2005) demonstrates how a smile ships, where a nonverbal display of status
in response to sexually provocative ques can take on sexual connotations. A superior’s
tions, which do at times occur in the work use of immediacy behaviors—touching, star
place, has the potential for prompting ing, and getting close, for example—has
inappropriate sexual conduct. Based on the long been the prerogative of higher-status
idea that women use social smiles for a individuals. But because these actions are
variety of reasons, they discovered that subject to multiple interpretations (e.g.,
female job applicants were more likely to friendliness, intimidation, sexual interest),
use “masking” smiles (concealing negative there is always the chance of misreading the
feelings) in response to questions such as signals (Le Poire, Burgoon, & Parrot,
“Do you have a boyfriend?” than in 1992). In addition, a subordinate’s use of
response to questions such as, “Do you submissive or low-status behaviors, such
have a best friend?” Interestingly, these as smiling, head nodding, silence, eye con
smiling responses (coded by the researchers tact, and direct body orientation, can
as unfelt or “non-Duchenne” smiles) were make it equally difficult to tell whether
correlated with perceptions of the inter the subordinate welcomes the superior’s
viewer as sexist and sexually harassing. advances or is behaving like a subordinate
What is more, men were less able to read is expected to behave. Moreover, the ambi
these uncomfortable smiles correctly than guity of nonverbal signals makes it possible
women were, and men who scored higher for harassers to deny the charges against
on an instrument that measures likelihood them (“I didn’t mean anything by it”).
to sexually harass were most likely to inter Nonverbal displays of status may also
pret the smiles as flirtatious. reveal whether someone is prone to engage
Explanations of sexual harassment usu in sexual harassment. Studies on the atti
ally refer to the actions of both the perpe tudes, beliefs, and perceptions of persons
trator and the victim: inappropriate sexual likely to sexually harass show that such
behavior by the perpetrator alongside some persons tend to describe themselves in
form of resistance, or at least disapproval, ways that emphasize social and sexual
by the victim. Studies show that judgments dominance (Pryor, 1987). Some research
of whether an individual is guilty of sexual indicates that nonverbal displays of status
harassment depend on the actions of both may be symptomatic of persons likely to
parties (e.g., Jones & Remland, 1997). The sexually harass. In one study, participants
more people view a behavior as inappropri viewed silent clips of videotaped interviews
ate, and the more unwelcome the behavior of men being interviewed by an attractive
is, the more likely those people are to define female subordinate (who could not be seen
it as a case of sexual harassment. But stud by the viewers). Only observing the men’s
ies also show that men and women often nonverbal behavior, the participants were
don’t agree on what actions constitute sex able to predict which men scored high on
ual harassment. In general, men are less a test that measured likelihood to sex
likely than women to see the same actions ually harass and which men scored low
as sexually harassing, particularly when (Driscoll, Kelly, & Henderson, 1998).
26-Manusov.qxd 6/30/2006 5:21 PM Page 507
(2000) discovered that negative affectivity with this surge of interest in emotions gen
correlated with the reporting of indirect erally has come a focused examination of
rather than direct forms of victimization nonverbal communication in three particu
(e.g., sabotaging work). They suggest that lar areas: emotional contagion, emotional
displays of negative affect are more likely to labor, and emotion recognition.
elicit acts of contempt than of anger or
aggression, making persons who exhibit
such behavior “appear as vulnerable targets EMOTIONAL CONTAGION
for exploitation, gossip, and other less
obtrusive forms of mistreatment” (Aquino Emotional contagion refers to a phenom
& Bradfield, 2000, p. 533). enon in which emotions spread from person
The research reviewed in this section to person. Primitive emotional contagion
supports a connection between nonverbal theory maintains that we “catch” others’
communication and various organizational emotions by means of automatic mimicry of
outcomes originating from the identification emotional expressions and the subsequent
and relationship functions of nonverbal feedback that results from our emotional
communication. In the context of organi displays (Hatfield, Cacioppo, & Rapson,
zational life, nonverbal displays of status 1994). Laboratory experiments show that
include an array of behaviors that symbolize mere exposure to a facial expression is suffi
one person’s power over another. These dis cient to produce muscle contractions in
plays not only reinforce the role relation observers that mirror the expression they
ships that exist in an organization, but they see (Dimberg & Ohman, 1996). In addition,
also can create conditions that can lead to studies confirm the facial feedback hypothe
dysfunctional leadership, sexual harass sis: that an individual’s facial expression of
ment, and workplace discrimination. an emotion can influence the person’s expe
rience of that emotion directly and immedi
ately (McIntosh, 1996).
♦ Managing Nonverbal The implications of facial feedback for
organizations has not escaped the attention
Displays of Emotion
of scholars, such as Goleman et al. (2002),
who consider it a ubiquitous process that
The preceding discussion addressed the uses leaders should harness for the good of the
and consequences of nonverbal communica organization. Among the studies they cite
tion in the service of an organization’s hier in their review is one where researchers
archy: maintaining the social order. In this observed 70 work teams across diverse
section, I direct attention to how nonverbal industries and found that members who sat
displays of emotion may affect the routine in meetings together ended up sharing
performances of organizational members. moods in a relatively short period of time
The subject of emotions in organizations is (Bartel & Saavedra, 2000). Goleman et al.
worthy of serious investigation (Fineman, (2002) claim that the more cohesive a work
1993), and a great deal of interest has been group is, the more contagious the emo
shown specifically in the development of tional displays will be. Furthermore, they
“emotional intelligence” in the workplace, argue that leaders are most likely to control
as advocated fervently by Goleman (1998), the contagion that takes place because
Goleman, Boyatzis, and McKee (2002), group members generally see the leader’s
and Dulewicz and Higgs (2003). Along emotional reaction as the most valid
26-Manusov.qxd 6/30/2006 5:21 PM Page 509
response, and therefore, members tend to and suppressing a felt emotion are acts of
model their own reactions on the leader’s, emotional labor in the workplace, which
particularly in emotionally ambiguous situ Hochschild (1983) defined as “the manage
ations. They also suggest that a leader’s ment of feeling to create a publicly observ
ability to spread emotions depends on his able facial and bodily display [that] is sold
or her capacity to convey those emotions. for a wage [and] therefore has exchange
That is, a leader with a highly expressive value” (p. 7). The management of emo
face, voice, and body is more likely to acti tions, according to Hochschild, requires
vate the emotional contagion process than a worker to engage in either surface acting
is a leader who is much less expressive. or deep acting. Whereas surface acting only
Emotional contagion may occur wher requires the actor to display an emotion
ever individuals work together in face-to with no attendant feelings, deep acting
face groups, or meet directly with the requires the actor to elicit the correspond
public. But does it affect task performance? ing emotion in some way, as a method
Some research suggests that the spread of actor might do to prepare for an emotion
positive emotions can boost the perfor ally charged scene. Curiously, whereas the
mance of work groups (Barsade, 2002), short-term effort involved in deep acting
predict job satisfaction among employees may surpass that needed for surface acting,
(Fisher, 2000), increase cooperation and the long-term effort required for the latter
minimize conflict (Barsade, 2002), improve appears to take a heavier toll (Grandey,
sales performance, and increase customer 2003; Totterdell & Holman, 2003).
satisfaction (Homburg & Stock, 2004; In her early research, Hochschild (1983)
Verbeke, 1997). Researchers have also dis estimated that “roughly one-third of
covered, however, that the spread of nega American workers have jobs that subject
tive emotions is a contributing factor to them to substantial demands for emotional
stress and burnout among physicians labor” (p. 11). Mann (1999) surveyed 12
(Bakker, Schaufeli, Sixma, & Bosveld, U.K. companies and found moderate levels
2001), nurses (Omdahl & O’Donnel, of emotional labor in almost two thirds of
1999), teachers (Bakker & Schaufeli, the communications reported by respon
2000), and sales personnel (Verbeke, dents and high levels in about one third of
1997). Thus, studies show that emotional the reported communications. More than
contagion may have positive and negative half of the participants reported that they
effects on the health and well-being of an laughed or frowned, not because they
organization. wanted to but because they were expected
to. Sixty percent of the reported communi
cations involved suppressing an emotion,
EMOTIONAL LABOR mostly anger. In addition, those higher up
in the organization reported less emotional
The contagion process depends on the labor than did those lower in the chain of
genuine (i.e., spontaneous) expression of command, supporting Van Maanen and
emotions. But the workplace often also Kunda’s (1989) astute observation that
demands that individuals engage in various “only the dominant and the dormant have
kinds of emotional dissimulation, pretend relative freedom from emotional con
ing to be cheerful when really annoyed or straints in organizational life” (p. 55).
frustrated, for instance. Expressing an Early qualitative studies of flight atten
unfelt emotion, exaggerating a felt emotion, dants, nurses, cashiers, and others led to a
26-Manusov.qxd 6/30/2006 5:21 PM Page 510
conception of emotional labor as a multidi training and experience rather than their
mensional construct consisting of (1) the fre looks or demeanor.
quency, duration, and intensity of emotional
displays; (2) the variety of emotions dis
played; (3) attentiveness to display rules; EMOTION RECOGNITION
and (4) the discrepancy between the felt and
the displayed emotion, referred to as emo Whereas the research on emotional con
tional dissonance (Mann, 1999; Morris & tagion and emotional labor generally
Feldman, 1996). Subsequent surveys have focuses on the expression and regulation
identified emotional dissonance consistently of emotion, other studies have examined
as a strong predictor of job dissatisfaction, the recognition of emotion. The ability to
emotional exhaustion, depersonalization, recognize emotions in others is a mainstay
and other factors contributing to job in the research on interpersonal sensitivity,
burnout (Diefendorff & Richard, 2003; Lee which Riggio (2001, this volume) contends
& Ashforth, 1996; Pugliesi, 1999). There is is necessary for leadership success, person
also evidence that emotional dissonance, nel functions of hiring and performance
particularly the suppression of negative appraisal, the development and functioning
emotions, can produce health consequences of work teams, and successful customer
related to prolonged stress (Maslach, 1982). service. Emotion recognition is also the
But the research also points to factors that most reliably valid component of emotional
moderate the impact of these negative intelligence (Elfenbein & Ambady, 2002).
consequences, such as job autonomy and Early research on nonverbal decoding
social support (Morris & Feldman, 1996; ability using the profile of nonverbal sensi
Wharton, 1996). tivity (PONS) reported positive correlations
Although most of the research on emo between PONS scores and measures of job
tional labor highlights negative effects, effectiveness of foreign service officers,
some work identifies benefits that arise leadership skills of school principals, and
under certain conditions. For example, the job ratings of human service workers
use of “deep acting” and the regular display (Rosenthal, Hall, DiMatteo, Rogers, &
of positive emotions can result in decreased Archer, 1979). More recently, it has been
dissonance, improved performance, and linked to effective leadership in organiza
increased satisfaction (Diefendorff & tions (Goleman et al., 2002).
Richard, 2003; Grandey, 2003; Totterdell But recent studies have also begun to
& Holman, 2003). Other researchers con raise questions about the benefits of emo
tend that any requirement to display posi tion recognition, finding support for the
tive emotions leads ultimately to improved counterintuitive claim that “people reading”
performance (e.g., increased sales) and a has a downside. Using the diagnostic analy
heightened sense of accomplishment (Rafaeli sis of nonverbal accuracy (DANVA),
& Sutton, 1987). In a qualitative study of Elfenbein and Ambady (2002) found that
sales workers, for example, Abiala (1999) the ability to read negative emotions con
found that emotional labor was most likely veyed through the voice rather than the face
to produce positive effects when interacting damaged workplace evaluations received
with customers was a small part of the from peers and supervisors. The ability to
workers’day, there were few rules to fol pick up emotions from less controllable
low, the intent to sell was not concealed, nonverbal communication channels, which
and the workers were hired for their the researchers call “eavesdropping,” may
26-Manusov.qxd 6/30/2006 5:21 PM Page 511
focus more on getting authority figures to persons were more likely to use positive
raise their expectations than on the more nonverbal cues, regardless of verbal con
difficult task of having them monitor and tent. Apparently, low status prompts indi
control their nonverbal cues. Some research viduals to follow a distinct set of rules that
shows, however, that training managers to emphasize clarity when delivering news,
hold higher expectations for their subordi and politeness when making requests.
nates, as well as to convey those expectations In their study of gender and mixed mes
to the subordinates, is difficult at best (Eden sages, LaPlante and Ambady (2002) varied
et al., 2000; White & Locke, 2000). the performance feedback of a supervisor in
terms of content and delivery (tone of voice).
They found that male supervisors were most
MIXED MESSAGES successful when the content was negative
and their tone of voice was positive. In con
One of the properties of communication trast, female supervisors were most success
is the potential for sending mixed messages, ful with feedback that included positive
where a message conveyed across one content paired with a negative tone of voice.
channel (e.g., facial) may not be consistent Overall, however, a positive tone of voice
with messages conveyed across one or more was only more effective in the male dyads.
other channels (e.g., vocal, verbal). Early The results for verbal content are consistent
experiments revealed that message recei with sex-role expectations: People are more
vers may use a “weighted sum” method of likely to expect compliments from women
inferring the attitudes of a speaker, placing and criticism from men. The difference in
more weight on facial and vocal expression tone of voice seems to suggest that a negative
than on words (Mehrabian & Wiener, tone adds seriousness to the female supervi
1967). In the workplace, Newcombe and sor’s feedback, reinforcing her legitimate
Ashkanasy (2000) reported that percep authority, whereas the positive tone softens
tions of a leader delivering feedback to sub the impact of criticism given to male subor
ordinates were affected more by the leader’s dinates, making it less threatening.
facial expressions than by the verbal con
tent of the message. Negative facial expres
sions elicited the most unfavorable judgments NONVERBAL
of the leader. Other studies show that INVOLVEMENT BEHAVIOR
receivers weigh verbal and nonverbal mes
sages when making inferences about polite Most research findings on the uses and
ness (LaPlante & Ambady, 2003; Trees & consequences of nonverbal communication
Manusov, 1998) and sexual harassment in formal workplace interactions, such as
(Remland & Jones, 1985). customer service transactions, interviews,
The uses and consequences of mixed and oral presentations, converge on the
messages may be related to status and gen construct of nonverbal involvement behav
der. LaPlante (2001) found status differ ior. Gaze, body orientation, facial expres
ences in the delivery of mixed messages. siveness, gesticulation, head nods, vocal
When delivering news, low-status persons animation, and more indicate the degree
were more likely than higher-status persons to which a person is overtly involved in
to use nonverbal communication that an interaction (Coker & Burgoon, 1987;
was consistent with their verbal content; Edinger & Patterson, 1983; see also
but when making requests, lower-status Andersen, Guerrero, & Jones, this volume).
26-Manusov.qxd 6/30/2006 5:21 PM Page 513
Generally, research shows that nonver presentations. One study found that profes
bal involvement behavior leads to positive sional buyers rated a salesperson as more
outcomes. Richmond and McCroskey believable when he used a steady gaze,
(2000) found that perceptions of a supervi and more interesting and persuasive when
sor’s nonverbal immediacy behavior were he avoided speech hesitations (Leigh &
associated with favorable evaluations of the Summers, 2002). Awamleh and Gardner
supervisor, subordinate motivation, and (1999) manipulated the speech delivery of
job satisfaction. In customer service trans a bogus CEO and found, not surprisingly,
actions, studies generally show that the use that the CEO’s presentation was more effec
of touch and eye contact generates more tive with eye contact, fluency, smiles, and
positive reactions from patrons (Crusco dynamic gestures. They also found that
& Wetzel, 1984; Kaufman & Mahoney, delivery was a more important predictor of
1999). Information gathering interviews performance than either the leader’s vision
also tend to benefit from involvement cues. or the organizational performance of the
Certain forms of touching and making eye leader’s company. Howell and Frost (1989)
contact with a person increase the likeli reported higher levels of task performance
hood that the person will comply with a and satisfaction when a leader’s delivery
request to participate in a survey (Hornik included vocal variety, eye contact, relaxed
& Ellis, 1988). Even apprehensive respon gestures, and animated facial expressions.
dents will talk more and like their inter Holladay and Coombs (1993, 1994) found
viewer more when the interviewer uses high that a leader’s nonverbal cues were more
levels of nonverbal involvement—direct predictive of charisma than was the “vision
body orientation, forward lean, head nods, ary content” of the leader’s message.
backchannels, and gazing while listening— Furthermore, Kirkpatrick and Locke (1996)
as opposed to much lower levels (Remland documented that delivery boosted percep
& Jones, 1989). tions of a leader’s charisma, but it did not
In the employment interview, research help the performance of subordinates.
confirms the positive impact of an applicant’s This section reviewed the research on the
nonverbal involvement behavior (Gifford, delivery function of nonverbal communi
Cheuk, & Wilkinson, 1985; McGovern cation, which refers to the coordination and
& Tinsley, 1978; Young & Beier, 1977). integration of verbal and nonverbal mes
Instances of such behavior may even predict sages. In the context of organizational life,
subsequent job performance evaluations researchers have been interested chiefly in
(DeGroot & Motowidlo, 1999). The impor outcomes associated with expectancy effects,
tance of nonverbal communication notwith mixed messages, and various nonverbal
standing, some research suggests that what involvement behaviors. In this regard, there
applicants say, the verbal content, influ is strong support for the claim that a
ences hiring decisions more than how they speaker’s nonverbal cues are often more
say it (Riggio & Throckmorton, 1988). In important than the speaker’s words (see also
addition, because social skills are more Giles & Le Poire, this volume). Specifically,
important for some types of jobs than through nonverbal channels, a speaker can
others, the impact of nonverbal communi signal positive or negative expectations,
cation may depend on the skills needed for modify the meaning of a verbal message,
a particular job (DePaulo, 1992). and influence the reactions and judgments
Research also recommends the use of of listeners. Studies show that in each of
nonverbal involvement behaviors in oral these ways, a speaker’s delivery can have a
26-Manusov.qxd 6/30/2006 5:21 PM Page 514
Power displays between women and men in Gladwell, M. (2005). Blink: The power of thinking
discussions of gender-linked tasks: A multi without thinking. New York: Little, Brown.
channel study. Journal of Personality and Goleman, D. (1998). Working with emotional
Social Psychology, 55, 580–587. intelligence. New York: Bantam.
Dovidio, J. F., Ellyson, S. L., Keating, C. F., Goleman, D., Boyatzis, R., & McKee, A. (2002).
Heltman, K., & Brown, C. E. (1988). The Primal leadership. Boston, MA: Harvard
relationship between social power and Business School Press.
visual displays of dominance between men Grandey, A. A. (2003). When “the show must
and women. Journal of Personality and go on”: Surface acting and deep acting
Social Psychology, 54, 233–242. as determinants of emotional exhaustion
Driscoll, D. M., Kelly, J. R., & Henderson, W. and peer-rated service delivery. Academy of
M. (1998). Can perceivers identify likeli Management Journal, 46, 86–96.
hood to sexually harass? Sex Roles, 38, Hall, J. A. (1984). Nonverbal sex differences:
557–588. Communication accuracy and expressive
Dulewicz, V., & Higgs, M. (2003). Leadership style. Baltimore, MD: Johns Hopkins
at the top: The need for emotional intelli University Press.
gence in organizations. The International Hall, J. A. (2005). Meta-analysis of nonverbal
Journal of Organizational Analysis, 11, behavior. In V. Manusov (Ed.), The source-
193–210. book of nonverbal measures (pp. 483–492).
Eden, D. (2003). Self-fulfilling prophecies in orga Mahwah, NJ: Erlbaum.
nizations. In J. Greenberg (Ed.), Organiza Hall, J. A., & Friedman, G. (1999). Status, gender,
tional behavior: The state of the science and nonverbal behavior: A study of struc
(pp. 91–122). Mahwah, NJ: Erlbaum. tured interactions between employers of a
Eden, D., Geller, D., Gerwirtz, A., Gordon- company. Personality and Social Psychology
Terner, R., Inbar, I., Liberman, M., et al. Bulletin, 25, 1082–1091.
(2000). Implanting Pygmalion leadership Hammermesh, D., & Biddle, J.E. (1994). Beauty
style through workshop training: Seven and the labor market. American Economic
field experiments. Leadership Quarterly, Review, 84, 1174–1194.
11, 171–210. Hatfield, E., Cacioppo, J. T., & Rapson, R.
Edinger, J. A., & Patterson, M. L. (1983). (1994). Emotional contagion. Cambridge,
Nonverbal involvement and social control. UK: Cambridge University Press.
Psychological Bulletin, 93, 30–56. Heintzman, M., Leathers, D. G., Parrott, R. L., &
Elfenbein, H. A., & Ambady, N. (2002). Cairns, A. B. (1993). Nonverbal rapport-
Predicting workplace outcomes from the building behaviors’ effects on perceptions of
ability to eavesdrop on feelings. Journal of a supervisor. Management Communication
Applied Psychology, 87, 963–971. Quarterly, 7, 181–209.
Fineman, S. (1993). Emotion in organizations. Henley, N. M. (1995). Body politics revisited:
Newbury Park, CA: Sage. What do we know today? In P. J.
Fisher, C. D. (2000). Mood and emotions while Kalbfleisch & M. J. Cody (Eds.), Gender,
working: Missing pieces of job satisfaction? power, and communication in human
Journal of Organizational Behavior, 21, relationships (pp. 27–62). Hillsdale, NJ:
185–202. Erlbaum.
French, M. T. (2002). Physical appearance and Hochschild, A. (1983). The managed heart:
earnings: Further evidence. Applied Eco Commercialization of human feeling.
nomics, 34, 569–572. Berkeley: University of California Press.
Gifford, R., Cheuk, F. N., Wilkinson, M. Hofstede, G. (1982). Culture’s consequences
(1985). Nonverbal cues in the employment (abridged ed.). Beverly Hills, CA: Sage.
interview: Links between applicant qualities Holladay, S. J., & Coombs, W. T. (1993).
and interviewer judgments. Journal of Communicating visions: An exploration of
Applied Psychology, 4, 729–736. the role of delivery in the creation of leader
26-Manusov.qxd 6/30/2006 5:21 PM Page 517
McGovern, T. V., & Tinsley, H. E. A. (1978). Rafaeli, A., & Sutton, R. I. (1987). Expression of
Interviewer evaluations of interviewee emotion as part of the work role. Academy
nonverbal behavior. Journal of Vocational of Management Review, 12, 23–37.
Behavior, 13, 163–171. Remland, M. S. (1981). Developing leadership
McIntosh, D. N. (1996). Facial feedback skills in nonverbal communication: A situa
hypotheses: Evidence, implications, and tional perspective. Journal of Business
directions. Motivation and Emotion, 20, Communication, 18, 17–29.
121–147. Remland, M. S. (1984). Leadership impressions
McNatt, D. B. (2000). Ancient Pygmalion and nonverbal communication in a superior-
joins contemporary management: A meta subordinate interaction. Communication
analysis of the result. Journal of Applied Quarterly, 32, 41–48.
Psychology, 85, 314–322. Remland, M. S. (2004). Nonverbal communica
Mehrabian, A., & Wiener, M. (1967). Decoding tion in everyday life (2nd ed.). Boston:
of inconsistent communications. Journal Houghton Mifflin.
of Personality and Social Psychology, 6, Remland, M. S., & Jones, T. S. (1985). Sex dif
108–114. ferences, communication consistency and
Morris, J. A., & Feldman, D. C. (1996). The judgments of sexual harassment in a perfor
dimensions, antecedents, and consequences mance appraisal interview. Southern Speech
of emotional labor. Academy of Manage Communication Journal, 50, 156–176.
ment Review, 21, 986–1010. Remland, M. S., & Jones, T. S. (1989). The
Murphy, J. D., Driscoll, D. M., & Kelly, J. R. effect of nonverbal involvement and com
(1999). Differences in the nonverbal behav munication apprehension on state anxi
ior of men who vary in the likelihood to ety, interpersonal attraction, and speech
sexually harass. Journal of Social Behavior duration. Communication Quarterly, 37,
and Personality, 14, 113–128. 170–183.
Newcombe, M. J., & Ashkanasy, N. M. (2002). Richmond, V. P., & McCroskey, J. C. (2000).
The role of affect and affective congruence The impact of supervisor and subordinate
in perceptions of leaders: An experimental immediacy on relational and organizational
study. Leadership Quarterly, 13, 601–615. outcomes. Communication Monographs,
Omdahl, B. L., & O’Donnel, C. (1999). 67, 85–95.
Emotional contagion, empathic concern, and Riggio, R. E. (2001). Interpersonal sensitivity
communicative responsiveness as variables and organizational psychology: Theoretical
affecting nurses’ stress and occupational and methodological applications. In J. A.
commitment. Journal of Advanced Nursing, Hall & F. J. Bernieri (Eds.), Interpersonal
29, 1351–1360. sensitivity: Theory and measurement
Patterson, M. L. (1983). Nonverbal behavior: (pp. 305–318). Mahwah, NJ: Erlbaum.
A functional perspective. New York: Riggio, R. E. (2005). Business applications of
Springer-Verlag. nonverbal communication. In R. E. Riggio
Pryor, J. B. (1987). Sexual harassment proclivi & R. S. Feldman (Eds.), Applications of
ties in men. Sex Roles, 17, 269–290. nonverbal communication (pp. 121–138).
Puccinelli, N. M., & Tickle-Degnen, L. (2004). Mahwah, NJ: Erlbaum.
Knowing too much about others: Modera Riggio, R. E., & Throckmorton, B. (1988). The
tors of the relationship between eavesdrop relative effects of verbal and nonverbal
ping and rapport in social interaction. behavior, appearance, and social skills in
Journal of Nonverbal Behavior, 28, evaluations made in hiring interviews.
223–243. Journal of Applied Social Psychology, 18,
Pugliesi, K. (1999). The consequences of 331–348.
emotional labor: Effects on work stress, Rosenthal, R. (1966). Experimenter effects in
job satisfaction, and well-being. Motivation behavioral research. New York: Appleton
and Emotion, 23, 125–154. Century-Crofts.
26-Manusov.qxd 6/30/2006 5:21 PM Page 519
Rosenthal, R. (1994). Interpersonal expectancy behavior (Vol. 11, pp. 43–103). Greenwich,
effects: A 30-year perspective. Current CT: JAI.
Directions in Psychological Science, 3, Verbeke, W. (1997). Individual differences in
176–179. emotional contagion of salespersons: Its
Rosenthal, R., Hall, J. A., DiMatteo, M. R., effect on performance and burnout.
Rogers, P. L, & Archer, D. (1979). Psychology and Marketing, 14, 617–637.
Sensitivity in nonverbal communication: Wharton, A. S. (1996). Service with a smile:
The PONS test. Baltimore: Johns Hopkins Understanding the consequences of emo
University Press. tional labor. In C. L. McDonald &
Rosenthal, R., & Jacobson, L. (1968). C. Sirianni (Eds.), Working in the service
Pygmalion in the classroom. New York: society (pp. 91–112). Philadelphia: Temple
Holt, Rinehart & Winston. University Press.
Steckler, N. A., & Rosenthal, R. (1985). Sex White, S. S., & Locke, E. A. (2000). Problems with
differences in nonverbal and verbal com the Pygmalian effect and some proposed solu
munication with bosses, peers, and subordi tions. Leadership Quarterly, 11, 389–415.
nates. Journal of Applied Psychology, 70, Woodzicka, J. A., & LaFrance, M. (2005).
157–163. Working on a smile: Responding to sexual
Tiedens, L. Z., & Fragale, A. R. (2003). Power provocation in the workplace. In R. E.
moves: Complementarity in dominant and Riggio & R. S. Feldman (Eds.), Applications
submissive nonverbal behavior. Journal of nonverbal communication (pp. 141–155).
of Personality and Social Psychology, 84, Mahwah, NJ: Erlbaum.
558–568. Word, C. O., Zanna, M. P., & Cooper, J.
Totterdell, P., & Holman, D. (2003). Emotion (1974). The nonverbal mediation of self-
regulation in customer service roles: Testing fulfilling prophecies in interracial inter
a model of emotional labor. Journal of action. Journal of Experimental Social
Occupational Health Psychology, 8, 55–73. Psychology, 10, 109–120.
Trees, A. R., & Manusov, V. (1998). Managing Young, D. M., & Beier, E. G. (1977). The role
face concerns in criticism: Integrating non of applicant nonverbal communication
verbal behaviors as a dimension of polite in the employment interview. Journal of
ness in female friendship dyads. Human Employment Counseling, 14, 154–165.
Communication Research, 24, 564–582. Yukl, G. (2002). Leadership in organizations.
Van Maanen, J., & Kunda, G. (1989). Real Upper Saddle River, NJ: Prentice Hall.
feelings: Emotional expression and organi Zebrowitz, L. A. (1997). Reading faces: Window
zational culture. In L. L. Cummings & to the soul? Boulder, CO: Westview Press.
B. M. Staw (Eds.), Research in organizational
26-Manusov.qxd 6/30/2006 5:21 PM Page 520
Part-V-Manusov.qxd 6/30/2006 5:25 PM Page 521
PART V
FINAL THOUGHTS
Part-V-Manusov.qxd 6/30/2006 5:25 PM Page 522
NONVERBAL COMMUNICATION
Basic Issues and Future Prospects
� Miles L. Patterson
University of Missouri, St. Louis
� Valerie Manusov
University of Washington
522 ◆
Part-V-Manusov.qxd 6/30/2006 5:25 PM Page 523
Final Thoughts–––◆–––523
emphasized the multivariate nature of non by Remland, this volume), expectations for
verbal communication in a coordinated set normative behavior (see Robinson, this vol
of behaviors. Conceptual recognition of the ume), and an awareness of what a particu
multidimensional nature of nonverbal com lar speech community values or stigmatizes
munication was not, however, the same as (see Manusov & Jaworksi, this volume).
actually measuring it. Even with the Ethnographic work on nonverbal behav
advances in videotaping and sophisticated ior is one way to reveal more qualitatively
techniques of computerized recording and how some of these issues work themselves
analysis, multivariate studies of nonverbal out in interaction. One exemplar by Donal
communication are still complicated and Carbaugh (1999) provided a discussion
time-consuming. Nevertheless, conceptualiz of the functions of silence among members
ing nonverbal communication as multivari of the Blackfeet, a confederacy of three
ate in nature set the stage for two important independent Indian tribes presently living in
developments in the field: (1) the formula Montana and Alberta, Canada. In his
tion of theories of interactive behavior (see analysis, Carbaugh described the process
Patterson, this volume) and (2) emphasis on of “listening” (i.e., being silent, usually in
the various functions of nonverbal commu sacred spaces) as both a mode of learning
nication (Argyle, 1972; Patterson, 1983; see and a reflection of cultural values. This
chapters in Part III of this volume). practice does not occur—at least not in the
Consistent with these two developments, same form—in other cultural groups, and
the focus of research also changed as the it can only be understood by reference to
field matured. In the 1970s and 1980s, the values of that group. In providing this
research moved away from the study of spe description, Carbaugh was able to show
cific channels as isolated aspects of subtle effectively how certain rules for and inter
communication to a greater focus on the pretation of nonverbal behaviors can only
processes mediating nonverbal communica be understood when the larger communica
tion and the consequences and applications tive context is also understood.
of nonverbal communication (see the chap Although these issues—mediating pro
ters in Part IV of this volume). Furthermore, cesses, applications, and moderating
there was increased attention to the way in variables—continue to affect the contempo
which the moderating variables of culture, rary study of nonverbal communication,
sex or gender, age, and personality affected another emerging development from the
nonverbal communication (see the chapters mid-1980s through the present day changed
in Part II of this volume). the way that researchers viewed nonverbal
The focus on various determinants facil communication. Specifically, and in some
itated an appreciation of nonverbal behav ways inconsistent with the more relativistic
ior as part of a larger set of communicative perspectives of studies that look at differ
practices shaped and constrained by the ences across groups, there was a growing
cultural context or code in which they recognition of the importance of evolution
occur. Thus, for many researchers, under in shaping nonverbal communication (see
standing the communicative value of non the Floyd and Buck & Renfro Powers chap
verbal communication depends on its ters, this volume). Basic elements of every
integration with language (see Bavelas & day life, particularly those related to
Chovil, this volume). For others, the non survival, reproduction, and care of off
verbal behaviors can be understood only spring were linked to natural selection.
when seen within a framework of gender Thus, there is an increased appreciation of
rules and power (see chapters by Hall and hardwired reactions predisposing people to
Part-V-Manusov.qxd 6/30/2006 5:25 PM Page 525
Final Thoughts–––◆–––525
behave and to make judgments in a fashion active agents, initiating behavior and testing
promoting survival of the species. As preliminary judgments in social settings.
noted, other factors, such as culture (see Fourth, although people possess some adap
Matsumoto, this volume), individual dif tive flexibility in relating to others, and
ferences (see Gifford, this volume), and some more than others (e.g., high self-
social norms (see Dovidio, Hebl, Richeson, monitors), we are constrained substantially
& Shelton, this volume) can moderate in our social judgments and behavior. That
this hardwiring. There is also the pre is, the residual effects of biology, culture,
dictable disagreement over the relative gender, and personality constitute the “bag
impact of these hardwired predispositions, gage” that limits the range of our judgments
but it is clear that the biological bases and behavior (for more on this, see Part II of
of nonverbal communication cannot be this volume), while also providing a rich set
ignored. of interpretive practices that reflect those
At the risk of oversimplifying the field in forces.
its contemporary form, what do these So from this basic foundation, where are
developments suggest about the nature of we headed and what merits special atten
nonverbal communication? First, nonver tion in the years to come? In the remainder
bal communication is best represented as a of this chapter, we highlight a few issues
system comprising interdependent compo that may shape the direction of research
nents (see chapters by Bavelas & Chovil and theory in the future. Specifically, we
and Patterson, this volume) and not a hap discuss the automaticity of much nonverbal
hazard collection of unrelated cues and communication, advances in methods for
behaviors. We cannot attend to all the assessing nonverbal communication, and
pieces in any single study, but we can rec some ways in which new technologies may
ognize that any specific focus is part of a work to extend how we conceptualize and
larger system of communication, embedded what we attend to in the study of nonverbal
within a larger set of communicative prac communication.
tices. Second, social judgments and behav
ior are pragmatic, serving a variety of
♦ A Glimpse at Future Issues
different functions and having important
social, personal, and relational conse
quences (see Part IV of this volume). These AUTOMATIC PROCESSES
consequences can be valuable (as in the
case of developing close relationships) or A growing body of research and theory
harmful (as in the case of unwarranted points to the dominance of automatic
affection). The reflection of functions in processes in the sending and receiving
goals can even be activated outside of of nonverbal communication (see Bargh,
awareness (Bargh, 1997). That is, particular 1997; Lakin, this volume). Although we
situations, such as an interview or a social may prefer to see ourselves as rational and
gathering, may be sufficient to prime a spe deliberate in our social contacts, automatic
cific goal and activate adaptive social judg processes seem to be the “default setting” in
ment and behavioral processes, all outside relating to others. In many cases, neither
of awareness (see chapters by Lakin and judgments nor behaviors can wait for a log
Patterson, this volume). ical assessment of alternatives before some
Third, individuals are not simply reactive decision or action is required. Even when
in their social environments, as the early individuals have the luxury of adequate
interactive theories suggested. They are also time, they are biased toward the quick and
Part-V-Manusov.qxd 6/30/2006 5:25 PM Page 526
easy solutions. That is, people are inclined more accurate than chance still leaves a lot
to be “cognitive misers,” avoiding unneces of room for knowing when we can be con
sary (and sometimes, necessary) cognitive fident in the accuracy of our judgments.
effort (Fiske & Taylor, 1995). This is especially the case when confidence
It is clear that receivers’ automatic judg in judgments does not predict their accu
ments are typically neither arbitrary nor racy (DePaulo, Charlton, Cooper, Lindsay,
maladaptive (Bargh, 1997). Perceivers make & Muhlenbruck, 1997; Patterson, Foster,
relatively accurate judgments of others & Bellmer, 2001). Future research might
quickly on the basis of limited appearance well consider the circumstances related to
and behavioral information (Ambady & accuracy in judgments, especially automatic
Rosenthal, 1992). Greater accuracy might judgments. There is research documenting
be expected in practical, affordance judg individual differences in judgment accu
ments dealing with how we relate to others racy, but different measures of decoding
than in more abstract trait judgments accuracy are not highly correlated (see
(McArthur & Baron, 1983; Swann, 1984; Riggio, this volume).
Zebrowitz & Collins, 1997). Of course, So is there a simple, general dimension
automatic judgments are sometimes wrong underlying decoding accuracy? What are
(e.g., the fundamental attribution error). the particular circumstances that contribute
Correcting such errors requires not only the to increased accuracy in judgments? Are
availability of cognitive resources, but it also some specific pieces of appearance and
necessitates the motivation to apply those behavioral information especially diagnos
resources in resolving inconsistencies (Gilbert tic or is the overall gestalt, even in very brief
& Malone, 1995). Thus, without the neces glimpses of others, more important? In
sary cognitive resources and motivation, ini everyday experience, are we more attuned
tial, automatic judgments will dominate. On to reading the emotions of others or to
the other hand, because the automatic judg reading their behavioral intentions (see
ments are relatively accurate, more is not Fridlund & Russell, this volume)? To what
necessarily better when it comes to applying extent can training improve accuracy in
cognitive resources to our judgments. judgments, and if training makes us more
Thinking too much about initially auto aware, what does that do to the automatic
matic judgments can actually decrease accu ity of judgments? These are just a few of the
racy (Patterson & Stockbridge, 1998; questions that may merit attention in future
Wilson & Schooler, 1991). research.
So there is something of a dilemma here. Recent research on encoding nonverbal
Automatic judgments, particularly affor communication has also emphasized the
dance judgments, are generally accurate. primacy of automatic processes for a wide
But when they are not, cognitive effort range of behaviors. Like social judgment, it
must be applied to correct the judgments. is possible to initiate controlled, deliberate
Without independent evidence, however, patterns of behavior. But individuals must
we do not usually know when our judg be motivated enough to expend the effort
ments are inaccurate. Occasionally, such and have the necessary cognitive resources
mistakes catch up with us, as problems to monitor and manage the behavior.
arise in relationships or bank accounts are Nevertheless, thinking more about manag
drained, but by then the damage is done. ing one’s behavior is not necessarily desir
Documenting that automatic judgments are able, especially with behavioral routines
Part-V-Manusov.qxd 6/30/2006 5:25 PM Page 527
Final Thoughts–––◆–––527
that are otherwise automatic. Vallacher and support, automatic mimicry provided an
Wegner (1987) make this point in their efficient and adaptive way to increase the
action identification theory. Specifically, social bonds.
when actions are well learned or automatic, Although it is convenient to discuss the
people tend to conceptualize them in gen social judgment and behavioral “sides” of
eral terms, focused primarily on the goals interaction as isolated processes, they are
or consequences of actions. For example, interdependent elements in a larger system of
deciding to go to the supermarket to get communication (see Patterson, this volume).
milk and bread is sufficient to get into the In social settings, specific goals activate adap
car and drive 3 miles to the store. A person tive judgments and behavior simultaneously.
does not have to think about the specific Much of this process is initiated automati
elements of driving a car, including the cally and often proceeds to conclusion with
coordinated movements involved with a out any reflection. But unexpected or unusual
stick shift. If attention is directed to the par behavior from others can prompt attention
ticular components’ actions (e.g., the timing and effort in forming judgments, just as our
of applying the clutch, shifting the gear, and own awkward behavior can prompt attempts
accelerating), the efficiency of the sequence at behavior management. It seems likely that
may suffer. This is similar to Bargh’s (1997) once awareness and effort are applied in one
goal-dependent automaticity. For Bargh, process, they are likely to spread to the com
individuals are aware of particular goals, plementary process.
but once they decide to pursue them, the For example, if a friend behaves in an
instrumental behaviors frequently run on unusual fashion, controlled processes are
automatic. activated to determine just what this means.
In addition, when goals are commonly At the same time, our own behavior may be
pursued in a particular setting, the context more deliberately managed to facilitate an
itself may be sufficient to activate goals and accurate judgment. In a similar fashion, the
goal-directed behavior. A common form of realization of one’s own behavioral faux
automaticity in social settings is behavioral pas directs attention to correcting the mis
mimicry. Individuals mimic speech pat take and, simultaneously, focuses attention
terns, facial expressions, posture, and on the reactions of those around us.
movement in interactions routinely and Following these kinds of events, it may take
automatically (Lakin, Jefferis, Cheng, & some time before automaticity becomes the
Chartrand, 2003; see also chapters by norm again. Although automatic processes
Cappella & Schreiber and Tickle-Degnen, generally constitute the “default” setting
this volume). Behavioral mimicry is clearly for social judgments and behavior, as
adaptive, not only facilitating the immedi circumstances change, more controlled
ate interaction but also increasing liking processes can be activated. Of course, indi
and interdependence between partners. viduals have to be sufficiently motivated
One suggestion for how mimicry happens and have adequate cognitive resources to
so quickly and smoothly is that there is apply in evaluating their judgments or man
a direct perception-behavior link selected aging their behavior. Nevertheless, more
over the course of evolution because it was is not necessarily better when it comes to
critical for our survival (Dijksterhuis & thinking about our judgments and manag
Bargh, 2001). That is, because social ani ing our behavior. Automaticity is there for
mals depend on others for affiliation and a reason.
Part-V-Manusov.qxd 6/30/2006 5:25 PM Page 528
Final Thoughts–––◆–––529
nonverbal communication and its integra have not received much attention in research
tion with language (Bavelas & Chovil, this (for a notable exception, see Levine, 1997).
volume). They also reflect the degree to Specifically, Walther argues that chrone
which people can use an array of behaviors mics make up an important set of cues in
to communicate the same messages or per CMC and are available without any
form the same functions: When one cue is technological advances. For example,
not available, others take its place. how quickly people respond to one
Second, an exploration of online commu another’s messages can send important
nicating reveals the importance of nonverbal communicative messages of relationships,
cues to an array of vital social issues (see importance, and communicative skill (see
chapters in Part IV of this volume). Riggio, the volume). Yet chronemic cues
Specifically, research on CMC has revealed have been neglected as relevant in the
the extent to which we rely on nonverbal cues CMC context.
to display—and constrain—identity. When Closer attention to the new technologies
our physical attractiveness, cues to gender, may provide a new perspective on how
race, and age, and physical abilities are absent nonverbal communication works and what
from the interaction, as occurs in many but is most central in our research endeavors.
not all online interactions, we get a chance to Many of the assumptions that we hold
see the ways in which the nonverbal cues of (including the automaticity of our cue use)
identification sometimes limit, and certainly can be reinvestigated when nonverbal com
configure, our face-to-face interactions. So, munication takes on a more strategic form.
for example, research reveals the extent to Finally, although the new technologies
which we require nonverbal cues to personal advance the range and efficiency of interper
ize another; without them, people may sonal communication, there may also be
categorize others vaguely as “out-group” unwanted side effects. As people spend more
members and be less likely to treat them fairly time online, talking on cell phones, and text
(Postmes, Spears, Lea, & Reicher, 2000; see messaging, what happens to the important
also, Dovidio et al., this volume). social benefits possible only in face-to-face
At the same time, the absence of certain interactions? Some scholars suggest that the
nonverbal cues that work typically to impersonal nature of our new technologies
categorize a person into a stigmatized reduces our sense of belonging and affects
group (e.g., skin color, body size) may the social fabric adversely (see Bugeja, 2005).
allow for enhancement of certain peoples’ The effect of technologically mediated com
identities when they can minimize the influ munication on the frequency and quality
ence of the cue. Furthermore, as Walther of face-to-face interactions is an important
(this volume) notes, CMC allows for issue for future research.
selective identity presentation in a way that
face-to-face communication does not. For
example, people can take more time to ♦ Conclusion
think through what they want to say, they
may choose to send a photograph of them
selves that is particularly flattering (or sev The study of nonverbal communication has
eral years old), or they may state that their changed dramatically over the last 50 years.
age or sex is other than what it actually is. Conceptual and methodological advances
Third, research on CMC reveals the have facilitated a growing appreciation of
importance of certain nonverbal cues that the influence of nonverbal communication
Part-V-Manusov.qxd 6/30/2006 5:25 PM Page 531
Final Thoughts–––◆–––531
in our social world. We now have a better Blascovich, J., Loomis, J., Beall, A., Swinth, K.,
understanding of how biology, culture, gen Hoyt, C., & Bailenson, J. N. (2002).
der, and personality help shape the course Immersive virtual environment technology
of nonverbal communication. In face-to as a methodological tool for social psychol
ogy. Psychological Inquiry, 13, 103–124.
face interactions, there is a growing appreci
Bugeja, M. (2005). Interpersonal divide.
ation for the interdependence of parallel
Oxford, UK: Oxford University Press.
processes of sending and receiving infor
Carbaugh, D. (1999). “Just listen”: “Listening”
mation in the service of specific goals. and landscape among the Blackfeet.
Furthermore, this system of communication Western Journal of Communication, 63,
operates primarily on automatic, but it can 250–270.
shift to more controlled processes as correc DePaulo, B. M., Charlton, C., Cooper, H.,
tions and adjustments are required in inter Lindsay, J. J., & Muhlenbruck, L. (1997).
actions. New methods, such as brain imaging The accuracy-confidence correlation in
and virtual reality techniques, provide us the detection of deception. Personality and
with interesting and powerful alternatives for Social Psychology Review, 1, 346–357.
studying nonverbal communication. There is, Dijksterhuis, A., & Bargh, J. A. (2001). The per
ception-behavior expressway: Automatic
however, still much to learn from careful
effects of social perception on social behav
research in mundane public settings where
ior. In M. P. Zanna (Ed.), Advances in
nonverbal communication is the means of
experimental social psychology (Vol. 33,
negotiating unfocused interactions. New pp. 1–46). San Diego, CA: Academic Press.
media and computer technologies also enlist Ekman, P. (1965). Differential communication
nonverbal elements, but not in the fashion of affect by head and body cues. Journal
we have seen in the past. Thus, as some of Personality and Social Psychology, 2,
issues seem closer to resolution, new ques 726–735.
tions will keep us busy in the future. Exline, R. V. (1963). Explorations in the process
of person perception: Visual interaction in
relation to competition, sex, and need for
affiliation. Journal of Personality, 31, 1–20.
♦ References Fiske, S. T., & Taylor, S. E. (1995). Social cog
nition (2nd ed.). New York: McGraw-Hill.
Gilbert, D. T., & Malone, P. S. (1995). The cor
Ambady, N., & Rosenthal, R. (1992). Thin respondence bias. Psychological Bulletin,
slices of behavior as predictors of inter 117, 21–38.
personal consequences: A meta-analysis. Goffman, E. (1963). Behavior in public places.
Psychological Bulletin, 111, 256–274. New York: Free Press.
Argyle, M. (1972). The psychology of interper Kolb, B., & Whishaw, I. Q. (2003).
sonal behavior (2nd ed.). London: Penguin. Fundamentals of human neuropsychology
Argyle, M., & Dean, J. (1965). Eye-contact, (5th ed.). New York: Worth.
distance, and affiliation. Sociometry, Lakin, J. L., Jefferis, V. E., Cheng, C. M., &
28, 289–304. Chartrand, T. L. (2003). The chameleon
Bailenson, J. N., & Yee, N. (2005). Digital effect as social glue: Evidence for the
chameleons: Automatic assimilation of non evolutionary significance of nonconscious
verbal gestures in immersive virtual environ mimicry. Journal of Nonverbal Behavior,
ments. Psychological Science, 16, 814–819. 27, 145–162.
Bargh, J. A. (1997). The automaticity of every Levine, R. (1997). A geography of time. New
day life. In R. S. Wyer Jr. (Ed.), Advances York: Basic Books.
in social cognition (Vol. 10, pp. 1–61). Mahl, G. F. (1956). Disturbances and silences
Mahwah, NJ: Erlbaum. in the patient’s speech in psychotherapy.
Part-V-Manusov.qxd 6/30/2006 5:25 PM Page 532
Journal of Abnormal and Social Psychology, and visibility on recognition and avoidance
53, 1–15. in passing encounters. Western Journal of
Manusov, V. (2005). (Ed.). The sourcebook of Communication, 69, 219–231.
nonverbal measures: Going beyond words. Patterson, M. L., Webb, A., & Schwartz, W.
Mahwah, NJ: Erlbaum. (2002). Passing encounters: Patterns of recog
McArthur, L. Z., & Baron, R. (1983). Toward nition and avoidance in pedestrians. Basic
an ecological theory of social perception. and Applied Social Psychology, 24, 57–66.
Psychological Review, 90, 215–238. Pelphrey, K. A., Viola, R. J., & McCarthy, G.
Mehrabian, A. (1969). Significance of posture (2004). When strangers pass: Processing of
and position in the communication of atti mutual and averted gaze in the superior
tude and status relationships. Psychological temporal sulcus. Psychological Science,
Bulletin, 71, 359–372. 15, 598–603.
Patterson, M. L. (1983). Nonverbal behavior: Postmes, T., Spears, R., Lea, M., & Reicher, S. D.
A functional perspective. New York: (2000). SIDE issues centre stage: Recent
Springer-Verlag. developments in studies of deindividuation
Patterson, M. L. (2005). The passing encounters in groups. Amsterdam: Royal Netherlands
paradigm: Monitoring microinteractions Academy of Arts and Sciences.
between pedestrians. In V. Manusov (Ed.), Rizzolatti, G., & Craighero, L. (2004). The
The sourcebook of nonverbal measures: mirror-neuron system. Annual Review of
Going beyond words (pp. 431–440). Neuroscience, 27, 169–192.
Mahwah NJ: Erlbaum. Sommer, R. (1959). Studies in personal space.
Patterson, M. L., Foster, J. L., & Bellmer, C. D. Sociometry, 22, 247–260.
(2001). Another look at accuracy and con Swann, W. B., Jr. (1984). Quest for accuracy in
fidence in social judgments. Journal of person perception: A matter of pragmatics.
Nonverbal Behavior, 25, 207–219. Psychological Review, 91, 457–477.
Patterson, M. L., Iizuka, Y., Tubbs, M., Ansel, Vallacher, R. R., & Wegner, D. M. (1987).
J., & Anson, J. (2006, January). Passing What do people think they’re doing? Action
encounters East and West: A comparison of identification and human behavior. Psy
Japanese and American pedestrian interac chological Review, 94, 3–15.
tions. Poster presented at the annual con Wilson, T. D., & Schooler, J. W. (1991).
vention of the Society of Personality and Thinking too much: Introspection can
Social Psychology, Palm Springs, CA. reduce the quality of preferences and deci
Patterson, M. L., & Stockbridge, E. (1998). sions. Journal of Personality and Social
Effects of cognitive demand and judgment Psychology, 60, 181–192.
strategy on person perception accuracy. Zebrowitz, L. A., & Collins, M. A. (1997).
Journal of Nonverbal Behavior, 22, Accurate social perception at zero acquain
253–263. tance: The affordances of a Gibsonian
Patterson, M. L., & Tubbs, M. E. (2005). approach. Personality and Social Psychology
Through a glass darkly: Effects of smiling Review, 1, 204–223.
AU-Index.qxd 6/30/2006 5:25 PM Page 533
AUTHOR INDEX
◆ 533
AU-Index.qxd 6/30/2006 5:25 PM Page 534
Author Index–––◆–––535
Author Index–––◆–––537
Carter, J. D., 87, 207, 209, 212, 214, Cheng, C. M., 69, 333, 334, 369, 371, 388,
285, 288, 292 389, 393, 526
Carton, E. E. R., 165 Cherulnick, P. D., 325
Carton, J. S., xix, 165 Chesebro, J. L., 425, 430, 431
Caruso, D. R., 80, 87, 90 Cheuk, F. N., 513
Carver, C. S., 62 Cheung, H., 486
Carver, V. H., 332 Chevalier-Skolnikoff, S., 224
Case, D., 462 Chiao, J. Y., 53
Casella, D. F., 86 Chidambaram, L., 467
Casey, F., 108 Ching, P., 351
Casey, S., 108 Chiu, P., 53
Caso, L., 347 Chock, T. M., 471
Casper, R. C., 247 Choi, J. W., 226, 228, 229
Cassell, J., 442 Choi, Y. S., 47, 64
Castel, J. M., 189 Chovil, N., 13, 98, 100, 101, 102, 103,
Castles, D. L., 301 106, 107, 108, 109, 110, 312, 314,
Catt, K. J., 123 315, 347, 537
Cattell, R. B., 280 Christensen, A., xx, xxi, 271, 284, 410, 413
Caughlin, J. P., xx Christie, B., 462
Caul, W. F., 81, 129 Christophel, D. M., 430
Cawley, J., 507 Chu, L., 326
Cederberg, C., 471 Churchill, M. E., 67, 248, 266
Cegala, D. J., 438 Cialdini, R. B., 326, 474
Ceschi, G., 189, 190 Ciarocco, N. J., 322, 333
Chaiken, S., 62, 250 Ciarrochi, J., 90
Chambers, W., 62 Ciccia, A., 491
Champion, C., 87 Cicero, M. T., 4
Chan, A., 90, 227 Clancy, S. M., 212
Chang, H. C., 395 Clanton, N. R., 323
Chapanis, A., 472 Clark, D. M., 324
Chaplin, W. F., 323 Clark, H. H., 99, 102, 103, 105, 109,
Chapman, A. J., 23 110, 473
Chapple, E., 8 Clark, M. S., 332
Charlesworth, W. R., 147, 148, 224 Clarke, D. D., 261
Charlton, C., 525 Clarke, H. H., 98
Charlton, K., 85, 344, 345, 346, 347, 348, Clausen, A., 188
352, 354 Clayton, F., 50
Chartrand, T. L., 32, 49, 52, 60, 61, Cleeton, G. U., 165
63, 65, 69, 70, 71, 333, 334, Clore, G. L., 384
344, 368, 369, 370, 371, 388, Cloven, D. H., 284
389, 393, 526 Coates, L., 98, 99, 109, 347
Chase, J., 313 Coats, E., 13, 190, 212, 238, 239, 240, 280,
Chasiotis, A., xviii 284, 287, 291, 295, 482
Chassin, L., xviii Cochenour, J. J., 469, 470
Chaudhuri, A., 247 Coe, N. P. W., 438
Chaumeton, N., 47, 53 Cohan, C., 409
Chazan, E., 183 Cohen, A. A., 11, 100, 108
Cheek, J. M., 88, 164 Cohen, D., 182
Chemtob, C., 364 Cohen, J., 209, 210, 345
Chen, M., 61, 69, 490 Cohn, D. F., 124
Chen, Y., 108, 324 Cohn, J. F., 189
Cheney, D. L., 122 Coifman, K., 89
AU-Index.qxd 6/30/2006 5:25 PM Page 538
Author Index–––◆–––539
Author Index–––◆–––541
Gifford, R., 163, 165, 168, 169, 170, 171, Gotlib, I. H., 53
172, 173, 282, 288, 326, 330, 371, 513 Gottman, J. M., xv, xix, xx, xxi,
Gilbert, A. N., 104, 313 52, 190, 271, 368, 386, 389,
Gilbert, D. T., 30, 34, 47, 68, 349, 350, 525 403, 409, 410
Gilbert, R. E., 324 Gould, S. J., 304
Gilding, M., 468 Govender, R., 62
Giles, H., xxi, xxii, 51, 367, 370, 388, 538 Graham, J. A., 108
Gillespie, D. L., 291 Grahe, J. E., 382, 385, 386, 388, 390,
Gillis, J. S., 83, 323, 382, 386, 388, 390, 391 391, 395
Gilman, S. L., 301 Grammer, K., 150, 326, 327, 387, 411
Gilmour, R., xvi Granhag, P. A., 342, 350, 352, 356
Gilovich, T., 62, 124, 134, 135, 326 Granka, L., 464, 467
Gingrich, B., 127 Graves, T. D., 221
Ginsburg, B. E., 121, 128, 135 Gray, A. R., 47
Ginsburg, G., xix, 313 Gray, H. M., 47, 64, 538–539
Ginsburg, S. L., 242 Green, E., 354
Giron, M., 440, 441, 446 Green, J. Y., 440
Givskov, M., 125 Greenbaum, C. W., 367
Gladwell, M., 514 Greenbaum, P., 108
Glaser, R., xvii, xviii Greenberg, B. S., 245
Glassman, M., 438 Greenberg, R., 182
Glassman, N., 438 Greene, J. O., 24, 25, 87, 143, 262,
Gleeson, K., 325 269, 370, 371
Godin, S., 469 Greenfield, S., 439
Goffman, E., 10, 52, 104, 240, 243, 301, 322, Greenwald, A. G., 43, 53
323, 334, 353, 389, 442, 450, 528 Greer, A. E., 412
Gold, C., 164, 281, 282 Gregory, S. W., 51
Gold, D. J., 267 Gregory, S. W., Jr., 248
Goldberg, L. R., 168 Grice, H. P., 103, 106
Goldin-Meadow, S., 326 Grich, J., 406
Goldman, A., 373 Griffith, C. H., 438, 443
Goldman-Eisler, F., 10, 343 Gronbeck, B. E., 6
Goldsmith, D. J., 270 Gross, J. J., 87, 88, 202, 368, 385, 386
Goldsmith, H. H., 68 Grossman, A., 242
Goldstein, S., 188 Grossman, E., 165
Goleman, D., 80, 508, 510 Gruber, K., 302
Gollwitzer, P. M., 63 Gruenfeld, D. H., 287, 288, 482, 488
Gomez-Beneyto, M., 440, 441, 446 Gruzen, J., 150
Gonzaga, G. C., 408 Guadagno, R. E., 474
Good, D. C., 291 Gubarchuk, J., 486
Goodman, C. R., 364 Gudjonsson, G. H., 349
Goodwin, C., 12, 98, 101, 439, 441, 442, Gudykunst, W. B., 220, 225
443, 445, 446, 450, 454, 455 Guerrero, L. K., 25, 49, 144, 164, 259, 261,
Goodwin, M. H., 98 264, 265, 266, 267, 268, 270, 271, 325,
Goodwyn, S. W., 192 370, 371, 409, 410, 412, 413, 539
Goorha, S., 474 Guinote, A., 488
Gordon, G. H., 438 Gump, B. B., 364
Gordon, V. V., xvii Gunnar, M., 183
Gordon-Terner, R., 512 Gunnoe, J., 206
Goren, A., 325 Guthrie, I. K., xxi
Gorham, J. S., 426, 427, 428, 430 Guthrie, R. D., 329
Goss, A. E., 190 Gwiazda, J., 183
AU-Index.qxd 6/30/2006 5:25 PM Page 543
Author Index–––◆–––543
Haakana, M., 450 Hatfield, E., 65, 84, 262, 268, 270, 331,
Habbema, J. D. F., 151 362, 364, 388, 508
Hadac, R. R., 439, 441 Hauser, M. D., 120, 123, 224
Hadiks, D., 355 Hauser, S. T., 409
Haefner, P., xx Hay, J., 322
Hager, J., 12, 220 Hayes, C. W., 204
Haidt, J., 45 Haymovitz, E. L., 323
Hairfield, J. G., 86 Haynes, M. T., 143, 151
Haist, S. A., 438, 443 Haynes, O. M., 189
Halberstadt, A., 237 Hays, R. B., 270
Halberstadt, A. G., 204, 212, 482, Hearn, V., 208, 227
483, 484, 496 Heatch, C., 452
Hale, J. L., xxi, 23, 24, 143, 264, 265, Heath, C., 445, 446, 450, 454, 455
267, 281, 289, 370, 471 Heath, D. C., 108
Hall, C. C., 325 Heatherington, L., 282
Hall, E. T., 9, 11, 164, 221, 265, 280, Heatherton, T. F., 301, 489
469, 483, 484 Heaton, T. B., 409
Hall, J. A., xvii, 13, 14, 45, 46, 53, 64, 82, Heavey, C. L., xxi, 271, 284
83, 85, 87, 202, 203, 204, 205, 206, 207, Hebb, J., 364
208, 209, 210, 212, 214, 267, 268, 280, Heberlein, A. S., 46, 53
284, 285, 287, 288, 291, 292, 295, 440, Hebl, M. R., 27, 301, 486, 488, 489, 491,
451, 455, 482, 502, 504, 505, 510, 539 492, 495, 496, 539
Hall, K. P., 505 Hecht, M. A., 204, 213, 214, 285
Hallahan, M., 45, 46, 82, 165, 166, 208 Hecht, M. L., 264, 266, 268
Hallam, M., 150 Hedlund, L. E., 302, 312
Halsey, L. B., 366 Heft, H., 383
Halverson, H. M., 8 Heider, F., 45, 187
Hamilton, W. D., 141 Heider, K., 208, 224, 227
Hammermesh, D., 507 Heinrichs, M., 128
Hancock, J. T., 471, 474 Heintzman, M., 503
Hansen, J. S., 164 Heisel, A. D., 432
Harbaugh, C. R., 128 Hejl, J., 450
Hargreaves, D., 240, 247 Held, R., 183
Harizuka, S., 44, 229 Helmreich, R., 202
Harlow, H. F., 10, 128, 129, 135 Heltman, K. R., 205, 325, 332, 489, 505
Haroff, P., xxi Henderson, M., 225
Harpham, B., 23 Henderson, S., 468
Harrigan, J. A., 66, 183, 187, 349, 439, Henderson, W. M., 506, 507
440, 441, 446, 449, 450, 451, 452, 455 Henderson-King, D., 247
Harris, K., 248 Henderson-King, E., 247
Harris, M., xix, 42, 80 Hendricks, A., 250
Harrison, D., 184 Henggeler, S. W., xvii
Harrison, K., 247 Henley, N. M., 11, 89, 206, 211, 212, 214,
Harrison, R. P., 11, 100, 108 284, 483, 505
Harszlak, J. J., 80 Henningsen, D. D., 412
Hart, A. J., 165 Henry, C., 45
Harter, S., 281 Henson, A., 287
Hartman, B., 192 Hentzer, M., 125
Hartwig, M., 352, 356 Heritage, J., 450, 455
Harwood, J., xxii Hermsen, S., 71
Haslett, B. B., 192 Herring, C., 507
Hassin, R., 59, 325 Herskovitz, M. J., 330
AU-Index.qxd 6/30/2006 5:25 PM Page 544
Author Index–––◆–––545
Jaworski, A., 238, 242, 243, 244, 245, 246, Kaitz, M., 165
247, 291, 539 Kalakanis, L., 150
Jaworski, J. S., 325 Kalman, Y. M., 468
Jayne, B. C., 353, 355 Kanki, B. G., 163
Jefferis, A. E., 388, 389 Kaplan, S. H., 439
Jefferis, V. E., 69, 371, 526 Kappas, A., 148, 313
Jefferson, G., 439, 443, 446 Karau, S. J., 489
Jenkins, L., 190 Kardes, F. R., 62
Jennings, J. R., 349 Karney, B. R., 271, 408
Johansson, G., 46, 186 Kash, K. L., 53
John, O. P., 88, 169, 202, 325, 367 Kashy, D. A., 326, 344, 345, 353
Johnson, B., 66, 67, 488, 491 Kasri, F., 227
Johnson, C., 66, 67, 281, 488, 491, 505 Kassin, S. M., 187
Johnson, E., 13, 90 Kastner, M. P., 193
Johnson, J., 202, 495 Kaswan, J. W., 106
Johnson, K., 462 Katzman, S., 470
Johnson, M. H., 182 Kaufman, D., 513
Johnson, M. L., 280 Kaufman, W., 322
Johnson, P., 370 Kawakami, C., 324
Johnson, R. L., 492 Kawakami, K., 62, 66, 67, 71, 369, 388,
Johnson, S. L., 53 488, 489, 491, 494
Johnson, T., 98, 99, 101, 106, 108 Kaye, S. A., 151
Johnson, V. S., 301 Kean, K. J., 187
Johnston, A., 45 Kearney, P., 425, 432
Johnston, L., 62 Keating, C. F., 205, 288, 325, 329, 332,
Johnston, R. E., 103, 108 489, 505
Johnstone, T., 45 Kees, W., 11
Jonas, A. P., 438 Keith, V. M., 507
Jones, A. J., 205 Keller, H., xviii
Jones, B. C., 44 Kellett, S., 364
Jones, C. R., 61 Kelley, D. H., 426, 427, 430
Jones, D., 150 Kelley, H. H., 284
Jones, E., xvi Kelly, A. B., 270
Jones, E. E., 220, 322 Kelly, J. R., 506, 507
Jones, J. M., 483 Keltner, D., 45, 49, 283, 287, 288, 322,
Jones, P., 221 325, 367, 383, 408, 482, 488
Jones, S. B., 23, 370 Kemper, S., 486
Jones, S. E., 98, 466 Kendon, A., 6, 7, 9, 10, 11, 12, 14, 98,
Jones, S. M., 270, 539–540 102, 108, 310, 388, 439, 441, 453
Jones, S. S., 108, 313 Kendrick, T., 325
Jones, T. S., 506, 512, 513 Kennedy, G., 4
Jonsson, P., 369 Kennedy, J., 431
Joorman, J., 53 Kenny, D. A., 48, 121, 162, 163
Jordan, A., 25, 350 Kenny, R. D., 326
Jordan, J., 474 Kenrick, D. T., 142
Jorgensen, P. F., 25, 370, 371 Kenwood, C., 98, 101, 108
Josephs, I., 108 Kenworthy, K. G., 302, 312
Judd, C. M., 209, 488 Kerkstra, A., xxi, 203, 455
Juette, A., 326, 327, 411 Kersens, J. J., 439
Juffer, F., 366 Kerssen-Griep, J., 271, 414
Jungeberg, B. J., 325 Kestenbaum, R., 183
Jussim, L., 32 Khouri, H., 407
AU-Index.qxd 6/30/2006 5:25 PM Page 546
Author Index–––◆–––547
Author Index–––◆–––549
Author Index–––◆–––551
Randall, D. W., 325 Riggio, R. E., 14, 53, 80, 84, 85, 86, 87, 88,
Rao, S. R., 350, 352 89, 164, 325, 352, 353, 385, 455, 510,
Rapson, R. L., 65, 84, 268, 270, 331, 362, 513, 514, 542–543
388, 508 Rilling, J. K., 128
Rastatter, M., 484 Rime, B., 108
Raste, Y., 208 Rinck, C. M., 206
Rather, P., 126 Rintel, E. S., 468
Rauh, C., 124, 388 Risau Peters, J., xviii
Raven, B., 280, 430 Risenhoover, N., 183
Ray, G. B., 264, 266, 268 Ritts, V., 32
Raymond, G., 451 Ritzl, A., 373
Reeder, H., 266 Rizzolatti, G., 315, 316, 373, 527
Reeder, M. E., 268 Roach, M. A., xix
Regan, D. T., 124, 134, 135 Robers, T., 264
Reichenbach, H., 169 Roberts, A. R., 149, 301
Reicher, S. D., 464, 482, 485, 529 Roberts, C. A., xvii, 438
Reid, F. J. M., 468 Roberts, K. P., 51
Reid, J. E., 353, 355 Roberts, L. J., 49, 260, 261, 270
Reid, P. T., 484 Roberts, N., 410, 413
Reid, R. H., 431 Roberts, R., 90
Reis, H. T., 150, 259, 261 Robertson, R. R., 68
Reisenzein, R., 309 Robins, R. W., 43
Reiser, M., xxi Robinson, E., 405
Reiseter, K., 324, 327 Robinson, H., xxii
Remland, M. S., 289, 501, 502, 503, 506, Robinson, J., 405
512, 513, 542 Robinson, J. D., 442, 451, 454,
Ren, X., 127, 128 455, 543
Rendall, D., 309 Robinson, M. D., 384
Renfro, S., 388 Robinson, R. Y., 423
Renninger, L. A., 411 Robles, T. F., xvii, xviii
Restemeier, R., xviii Rodriguez, J. S., 326
Rezabek, L. L., 469, 470 Roe, L., 109, 347
Reznick, J. S., 69, 387, 389 Rogan, R., 351
Rhodes, G., 13, 42, 325 Rogers, P. L., 45, 46, 82, 208, 465, 510
Rholes, W. S., 61 Rogers, R. D., 128
Rice, K., 486 Rogers-Millar, E. L., 281, 282
Rice, R. E., 462, 468 Roggman, L. A., 149
Richard, E. M., 510 Rohner, R. P., 220
Richard, F. D., 210 Rollins, B. C., 281, 284
Richards, M. H., 247 Rollman, S. A., 486, 494
Richeson, J. A., 42, 46, 68, 327, 383, Roloff, M. E., 284, 361
384, 485, 486, 488, 489, 490, 493, 494, Romney, A. K., 224
495, 542 Roodenburg, H., 3
Richmond, V. P., 423, 424, 425, Rose, S. N., xviii
426, 427, 428, 430, 432, 503, Rosenbaum, R., 9
513, 542 Rosenbaum-Tamari, Y., 469
Ridgeway, C. L., 280, 281, 285, 286, Rosenberg, E., 13, 226
291, 485 Rosenberg, S. W., 248
Riedel, K., 126 Rosenfeld, H. M., 108, 182
Riesler, S., 463 Rosenholtz, S. J., 285, 291
Riggio, H. R., 53, 84, 86, 87, 88, 164, Rosenkrantz, P., 202
325, 455 Rosenstein, D., 148
AU-Index.qxd 6/30/2006 5:25 PM Page 553
Author Index–––◆–––553
Rosenthal, R., xvii, xix, 10, 12, 32, 34, 42, Saltz, J. L., 66
44, 45, 46, 47, 53, 64, 66, 68, 69, 80, Salvatore, J., 485, 489, 490, 493
82, 83, 90, 124, 134, 165, 166, 182, 185, Samter, W., 192, 270
187, 207, 208, 209, 214, 267, 268, 343, Sanbonmatsu, D. M., 62
344, 345, 362, 382, 384, 385, 387, 389, Sanchez de Rota, G., 202
390, 391, 439, 440, 441, 446, 449, 450, Sanchez-Garcia, E., 440, 441, 446
452, 455, 505, 511, 514, 525 Sanders, R. E., 98, 106, 442
Rosip, J. C., 53, 208, 214, 285, 288, 292 Sanderson, D., 469
Rosnow, R. L., 44 Sanfey, A. G., 128
Ross, E. D., 130 Sargent, J., 53
Roter, D., 440, 492 Satyaprasad, C., 186
Roth, W. T., 302 Saville-Troike, M., 238
Rothman, A. D., 46 Savin, V. J., 81
Rowland, D. A., 150 Sawyer, J., 265
Rowland-Morin, P. A., 438 Saxe, L., 66, 492
Roy, A., 241 Schaap, C., xxi
Rubenstein, A. J., 149, 150 Schachner, D. A., 412
Rubin, D. B., xix Schachter, S., 23
Rubin, M., 124 Schaufeli, W. B., 364, 509
Rubin, R. S., 325 Schaut, J. A., 302, 312
Rubinow, S., 11, 98 Scheflen, A., 9, 98
Rudman, L. A., 323 Schegloff, E. A., 439, 440, 441, 443, 446,
Ruedenberg-Wright, L., 469 450, 451, 452
Ruesch, J., 10, 11 Scherer, K. R., 12, 13, 44, 45, 46, 51, 98,
Ruiz-Belda, M. A., 108, 147, 309 168, 189, 190, 208, 224, 349
Rumsey, N., 351 Schieffelin, B., 238
Rusch, C. D., 224 Schiffrin, R. M., 60
Rushbult, C. E., 284 Schimmack, U., 227
Rushe, R., 410 Schlenker, B. R., 68, 302, 322, 323, 324,
Rushton, J. P., 163 332, 334
Russell, J. A., 11, 32, 43, 44, 146, 300, 305, Schmid Mast, M., 207, 208, 212
306, 307, 310, 331, 543 Schmidt, K. L., 124
Russell, M., 438 Schmitt, A., 326, 327, 411
Russin, A., 71 Schneider, F. W., 25
Rutherford, D., 262, 267, 268 Schneider, K., 108, 389
Ruusuvuori, J., 445, 453, 455 Schneider, S. L., 507
Ruzzene, M., 405 Schneider, W., 60
Ryan, E. B., xxi Schober, M. F., 108
Schooler, J. W., 34, 525
Saarni, C., 183, 189, 239, 327, 331 Schreiber, D. M., 543
Saavedra, R., 508 Schulkin, J., 373
Sabatelli, R. M., 124 Schwartz, A. M., 240
Sabini, J. P., 104, 302, 312, 313 Schwartz, B., 165, 174, 282, 288
Sachs, V. B., 190 Schwartz, G., 148
Sacks, E. A., 451 Schwartz, J., 224, 264
Sacks, H., 439, 443, 446 Schwartz, W., 528
Saine, T. J., 426 Scollon, R., 251
Sakamoto, Y., 494 Scott, M. B., 290
Salinas, C., 88, 325 Seaton, C., 491
Sallinen, A., 430 Sebeok, T., 281
Salovey, P., 80, 87, 90 Segall, M. H., 220, 330
Salter, V., xvi Segerstrale, U., 14
AU-Index.qxd 6/30/2006 5:25 PM Page 554
Author Index–––◆–––555
St. Peters, M., 409 Street, R. L., Jr., xvi, 367, 438, 441, 443,
Stack, D. M., 365 450, 452, 455
Stacks, D. W., 426 Streeter, L. A., 288
Stafford, B. M., 244 Stromwall, L. A., 352, 356
Stairs, D., 249 Strong, R. K., 47
Stallings, J., xviii Strongman, K., 229
Stamp, G. H., 442 Strough, J., 393
Stangor, C., 489 Stuart, G. L., 410
Stanley, S. M., 409 Sturgill, G., 126
Stark, R. E., xviii Suddendorf, T., 373
Steckler, N. A., 505 Suler, J., 471
Steele, C. M., 62 Sullivan, D. G., 249
Steenaert, B., 70 Sullivan, L. A., 489
Stein, J. L., 323 Sullivan, M., 147, 148
Steinberg, M., 12 Sullivan, M. W., 190
Steiner, M., xviii Summers, J. O., 513
Steinfield, C. W., 468 Surakka, V., 363
Stelzner, M. A., 268 Sutton, C., 106, 108, 110
Stenberg, C., 148 Sutton, M., 107, 108
Stennett, J., 45 Sutton, R. I., 510
Stenson, L. L., 25 Suzuki, R., 227, 228
Stephan, C. W., 485, 489 Swaim, G. W., 86
Stephan, W., 485, 489 Swan, K. P., 464
Stepper, S., 213 Swann, W. B., Jr., 32, 47, 525
Sterling, R. S., 187 Swim, J. K., 490
Stern, D. N., xix Swinkels, A., 165
Stern, L. A., 29, 35, 48, 268, 291, 362, 370, Swinth, K., 527
387, 388, 389 Symmes, D., xviii
Sternberg, R. J., 261 Symons, D., 365
Sternglanz, R. W., 405, 406 Synnott, A., 248
Stevenson, M. B., xix
Stewart, R., 280 Tabert, M., 207
Stice, E., xviii Tajfel, H., 226, 482, 485
Stier, D. S., 205, 206 Takeuchi, S., 226
Stiff, J. B., 346, 350, 351 Tamir, M., 384
Stillwell, A. M., 67 Tanford, S., 25, 26
Stinson, L., 405 Tang, J., 324
Stivers, T., 442 Taranta, A., 82
Stock, R. M., 509 Tascovich, J., 527
Stockbridge, E., 34, 64, 68, 525 Tassinary, L. G., 51
Stokes-Zoota, J. J., 210 Tate, C. S., 484
Stolberg, S. G., 331 Tatsuoka, M. M., 280
Stone, J. I., 324 Tay, B. K., 474
Stone, J. L., 354 Taylor, D. A., 262
Stoner, G. M., 474 Taylor, M. J., 331
Storck, J., 472 Taylor, S. E., 30, 34, 525
Stott, C., 468 Taylor, S. J., 80
Stoufe, L. A., 147 Tcherkassof, A., 307
Strack, F., 34, 35, 45, 66, 213, 363, Tchividjian, L. R., 324
364, 382 Tedeschi, J. T., 332
Strauss, E., 184 Templeton, J. L., 65, 302
Streeck, J., 98, 102, 442, 449 Terry, J., 243, 248
AU-Index.qxd 6/30/2006 5:25 PM Page 556
Author Index–––◆–––557
Wilkinson, M., 168, 169, 173, 513 Yamada, H., 227, 228
Williams, C. J., 494 Yamaguchi, S., 226, 228
Williams, E., 462 Yarczower, M., 228, 229
Williams, J. H., 373 Yashimoto, D., 410
Williams, J. R., 128 Yates, S., 250
Williams, K. D., 70 Yates, Y., 468
Williams, M., xvi Yee, N., 70, 527
Willis, F. N., 186, 206 Yeo, R. A., 329
Wilmot, W. W., 270 Yik, M. S. M., 307, 310
Wilson, B. J., 239, 240 Yirmiya, N., 367
Wilson, D. B., 209 Yoo, S. E., 87
Wilson, E. O., 123, 127 Yoo, S. H., 226
Wilson, J. F., 438, 443 Yoon, H. J., 220
Wilson, K., 66 Yoshikawa, S., 301
Wilson, T., 49 Yoshimoto, J., 351
Wilson, T. D., 34, 525 Yoshioka, M., 220
Winkel, F. W., 221, 353 Young, D. M., 513
Winograd, C. H., xvii, 185 Young, L., 127, 128, 160
Winter, L., 61 Young, M. J., 354
Wiseman, R., 504 Young, R. C., 283
Witkowski, S., 163 Young, R. K., 151
Witmer, D., 470 Young, S. L., xx
Witton, S. W., 409 Young-Browne, G., 182
Woike, B. A., 187, 194 Yukl, G., 504
Wolfer, J. E., 439
Wolff, W., 187 Zaadstra, B. M., 151
Wood, J. T., 403 Zach, U., xviii
Wood, R. M., 128 Zahn-Waxler, C., 405
Wood, W., 212, 481 Zaidel, S. F., 86
Woodall, W. G., 262, 267, 287, 322, 323, Zaitchik, D., 185
324, 326, 333, 409 Zajonc, R. B., 62, 64, 372, 373
Woods, J., 102, 110 Zak, P. J., 128
Woodson, R., 182 Zammuner, V., 225
Woodworth, M., 474 Zanna, M. P., 68, 69, 486, 491, 492, 511
Woodzicka, J. A., 506 Zebrowitz, L. A., 13, 31, 32, 42, 43,
Woolpy, J. H., 128, 135 67, 288, 325, 326, 330, 331, 382, 384,
Wooten, A. G., 432 404, 507
Word, C. O., 68, 69, 486, 491, 492, 511 Zebrowitz-McArthur, L., 188, 194, 288
Worthman, C., xviii Zelditch, M., 285, 291, 485
Wroblewska, A. M., 326 Zeman, J., 190
Wu, C. H., 149, 301 Zhou, Q., xxi
Wu, S., 224 Zilles, K., 373
Wyer, M. M., 344, 345, 353 Zimmerman, J. A., 88
Wyers, R. S., 61, 62, 271, 323, 407 Zinober, B., 192
Zuckerman, M., 80, 83, 84, 86, 202, 207,
Xin, K. R., xxiii 208, 343, 344, 345
Sub-Index.qxd 6/30/2006 5:25 PM Page 559
SUBJECT INDEX
◆ 559
Sub-Index.qxd 6/30/2006 5:25 PM Page 560
Subject Index–––◆–––561
Subject Index–––◆–––563
Subject Index–––◆–––565
Subject Index–––◆–––567
Subject Index–––◆–––569
Subject Index–––◆–––571
Subject Index–––◆–––573
◆ 575
ABE-Manusov.qxd 6/30/2006 5:21 PM Page 576
◆ 577
ATC-Manusov.qxd 7/6/2006 5:28 PM Page 578
methods. His research has been supported received SPSSI’s Kurt Lewin Award in 2004
by grants from the National Institutes (with S. L. Gaertner) for his career contribu
of Mental Health, the National Institute tions to the study of prejudice and discrimi
on Drug Abuse, the National Science nation. He received the Gordon Allport
Foundation, the National Cancer Institute, Intergroup Relations Prize in 1985, 1998,
the Twentieth Century Fund, and the and 2001 for his research on intergroup rela
Markle, Ford, Carnegie, Pew, and Robert tions. He has an MA and PhD from the
Wood Johnson foundations. He has served University of Delaware.
on the editorial boards of 15 different jour
nals. He is a Fellow of the International Norah E. Dunbar is Associate Professor of
Communication Association, a Distinguished Communication Studies at California State
Scholar of the National Communication University Long Beach. Her research inter
Association, a Past President of ICA, and a ests are in relational conflict, deception,
recipient of the B. Aubrey Fisher Mentorship power and dominance, and nonverbal
Award. He received his PhD in 1974 from communication. Currently, she is working
Michigan State University. on several projects, including a study on
the nonverbal expressions of dominance
Nicole Chovil is an author of Equivocal in close relationships. Her work can be
Communication (with Bavelas, Black, & found in journals such as the Journal
Mullett), plus 15 articles and chapters on of Social and Personal Relationships and
equivocal communication, hand and facial the Journal of Family Communication. She
displays in dialogue, and motor mimicry. She teaches undergraduate and graduate classes
conducted the first systematic experimental in interpersonal communication, per
studies of nonemotional functions of facial suasion, nonverbal communication,
displays in face-to-face dialogue. She was also research methods, and communication
co-investigator and collaborator on a project theory. She received her PhD in 2000 from
using discourse analysis to study the language the University of Arizona.
characterizing sexualized assault in legal judg
ments. She has a BA, with honors (psychol Robert S. Feldman is Associate Dean for
ogy), from the University of Victoria; an MA Faculty and Student Development and
(psychology) from Simon Fraser University; Professor of Psychology at the University of
and a PhD (psychology) from the University Massachusetts at Amherst. A winner of
of Victoria. In 2006, she left her position as the College Distinguished Teacher award,
Director of Education for the British he is a fellow of the American Psychological
Columbia Schizophrenia Society to become Association and the American Psycho
an independent research and education con logical Society. He is a winner of a
sultant specializing in mental illness. Fulbright Senior Research Scholar and
Lecturer award, and he has written more
John F. Dovidio is Professor of Psychology at than 100 books, book chapters, and scien
the University of Connecticut. He is cur tific articles. His research has been sup
rently Editor of the Journal of Personality ported by grants from the National
and Social Psychology—Interpersonal Institute of Mental Health and the National
Relations and Group Processes and has pre Institute on Disabilities and Rehabilitation
viously been Editor of Personality and Social Research. His research interests include
Psychology Bulletin. His research interests the development of nonverbal behavior in
are in intergroup relation, nonverbal impression management and honesty and
communication, and prosocial behavior. He deception.
ATC-Manusov.qxd 6/30/2006 7:41 PM Page 580
fellowship and has worked as a research race and gender influence the way people
assistant on projects relating to the com think, feel, and behave. More specifically,
munication of trustworthiness (funded by her research investigates multiple dynamics
the Russell Sage Foundation), the commu of prejudice and stereotyping from the per
nication of emotion in Huntington’s spectives of members of both traditionally
Disease patients and caregivers (funded by stigmatized and dominant social groups.
the University of Connecticut Health She is currently working on three primary
Center Huntington’s Disease Program), lines of research: the dynamics and conse
and brain mechanisms of empathic ability quences of interracial contact, detecting
(funded by the Olin Neuropsychiatry and controlling racial bias, and racial cate
Center of the Hartford Hospital Institute gorization and identity. She earned her BS
of Living). She is certified in the Facial in psychology from Brown University in
Action Coding System (FACS) and has 1994 and her PhD in social psychology
received training in functional brain imag from Harvard University in 2000.
ing (fMRI) study design and data analysis.
Her main research interests are in facial Virginia P. Richmond is Professor of
expressivity, empathic ability, interper Communication Studies at West Virginia
sonal coordination, and sex differences in University. She has authored or coau
emotion communication. thored over 25 books and 150 journal
articles and book chapters. Her book on
Martin S. Remland is Associate Professor of
nonverbal communication, Nonverbal
Communication Studies at West Chester
Behavior in Interpersonal Relations (coau
University of Pennsylvania. His research
thored with J. C. McCroskey), is now in
interests include nonverbal displays of sta
its fifth edition (2004). Her research,
tus and power and cross-cultural differ
writing, and teaching have focused on
ences in nonverbal involvement behaviors.
nonverbal communication, instructional
He is the author of Nonverbal Com
communication, interpersonal communi
munication in Everyday Life (Houghton
cation, organizational communication,
Mifflin, 2004) and coauthor of Inter
communication traits, social influence,
personal Communication Through the
and training and development. She has
Lifespan (Houghton Mifflin, in press). His
received numerous awards for her teach
work has appeared in numerous journals in
ing and research from her university and a
the fields of communication and psychol
wide variety of national professional asso
ogy. He received his BA from Western
ciations in communication, teacher educa
Illinois University, his MA from Central
tion, and pharmacy education. She has
Michigan University, and his PhD from
been recognized as one of the top five pro
Southern Illinois University.
lific published scholars in the history of
Jennifer A. Richeson is Associate Professor the field of communication.
in the Department of Psychology and
Faculty Fellow at the Institute for Policy Ronald E. Riggio is the Henry R. Kravis
Research, both at Northwestern Uni Professor of Leadership and Organizational
versity. Her research is in the areas of Psychology and Director of the Kravis
prejudice, stereotyping, and intergroup Leadership Institute at Claremont McKenna
relations. Her work considers the ways in College. He is the author of over 100 books,
which social group memberships such as book chapters, and research articles in the
ATC-Manusov.qxd 6/30/2006 7:41 PM Page 585
areas of leadership, assessment centers, orga together in a model called the psychological
nizational psychology, and social psychology. construction of emotion. He is a fellow of
His research work has included studies on the the APA and the APS. He received his PhD
role of social skills and emotions in leadership from the University of California at Los
potential and success, empathy, social intelli Angeles in 1974.
gence, emotional skill, and charisma. He is
Associate Editor of The Leadership Darren M. Schreiber is Assistant Professor
Quarterly and is on the editorial boards of of Political Science at the University of
Leadership and the Journal of Nonverbal California at San Diego. His research centers
Behavior. His recent books are Multiple on emergence and complexity in political
Intelligences and Leadership and The Future systems. He studied politics, philosophy,
of Leadership Development (coedited with and economics as an undergraduate at
Susan Murphy, 2002, 2003), Improving Claremont McKenna College, later attend
Leadership in Nonprofit Organizations ing the U.C. Davis School of Law, where he
(coedited with Sarah Smith Orr, 2004), focused on civil rights litigation. He then spe
Applications of Nonverbal Behavior cialized in federal litigation at the law firm of
(coedited with Robert S. Feldman, 2005), Neumiller and Beardslee. His dissertation
and Transformational Leadership (2nd ed., research used functional brain imaging
coauthored with Bernard M. Bass, 2006). (fMRI) to study the neural substrates of
political cognition and affect. He has shown
Jeffrey D. Robinson is Associate Professor that ideological sophisticates differ from
in the Department of Communication at political novices in their heightened use of
Rutgers University. He is broadly interested the posterior cingulate, a brain region associ
in interpersonal communication, health ated with automatic social evaluation. His
communication, and language and social goal is to integrate agent-based models of
interaction. He specializes in conversation macropolitical dynamics with his computa
analysis and physician-patient interaction. tional model of political cognition in individ
He received his BA (communication) from uals in order to illuminate the emergence of
the University of California, Santa Barbara, political ideology in the mass public. He
his MA (communication), from the served as Research Director at the Center
University of Southern California, and his of Excellence in Cancer Communication
PhD (sociology) from the University of Research at the Annenberg School of
California at Los Angeles. Communication, University of Pennsylvania,
during 2004 to 2005. He earned his PhD in
James A. Russell is Professor and Chair in political science at the University of
the Psychology Department at Boston California at Los Angeles in 2005.
College. He spent most of his academic
career at the University of British J. Nicole Shelton is Associate Professor of
Columbia. An initial interest in the Psychology at Princeton University. She
emotional impact of large-scale physical was a postdoctoral fellow at the University
environments led to studies on the of Michigan from 1998 to 2000. Her pri
language of emotion, taxonomies of emo mary research, which has been funded by
tion, facial expressions of emotion, cultural the National Institute of Mental Health
differences in emotion, the developmental and the National Science Foundation,
course of emotion knowledge, and theories focuses on how whites and ethnic minori
of emotion. This research was brought ties navigate issues of prejudice in
ATC-Manusov.qxd 7/6/2006 5:04 PM Page 586