The spontaneous expression of pride and shame:
Evidence for biologically innate nonverbal displays
Jessica L. Tracy*†and David Matsumoto‡
*Department of Psychology, University of British Columbia, 2136 West Mall, Vancouver, BC, Canada V6T 1Z4; and‡Department of Psychology,
San Francisco State University, 1600 Holloway Avenue, San Francisco, CA 94132
Edited by Frans B. M. de Waal, Emory University, Atlanta, GA, and approved June 16, 2008 (received for review March 18, 2008)
The present research examined whether the recognizable nonver-
bal expressions associated with pride and shame may be biologi-
cally innate behavioral responses to success and failure. Specifi-
cally, we tested whether sighted, blind, and congenitally blind
individuals across cultures spontaneously display pride and shame
behaviors in response to the same success and failure situations—
victory and defeat at the Olympic or Paralympic Games. Results
>30 nations displayed the behaviors associated with the proto-
typical pride expression in response to success. Sighted, blind, and
congenitally blind individuals from most cultures also displayed
behaviors associated with shame in response to failure. However,
culture moderated the shame response among sighted athletes: it
self-expression-valuing cultures, primarily in North America and
West Eurasia. Given that congenitally blind individuals across
consistent with the suggestion that the behavioral expressions
associated with both shame and pride are likely to be innate, but
the shame display may be intentionally inhibited by some sighted
individuals in accordance with cultural norms.
emotion ? innate behavioral response ? nonverbal expression ?
emotion ‘‘thrill.’’ Yet thrill may not be the most meaningful
emotion experienced in response to success. After winning an
athletic competition or succeeding at work or school, individuals
do not simply appear excited or happy. Rather, as social beings
focused on what such events mean for how we are perceived by
others and where we stand in the social hierarchy, we also feel
the emotion of pride. Similarly, the ‘‘agony’’ long associated with
defeat may in fact represent shame, the painful emotion expe-
rienced in response to failure. Pride and shame are typically not
included among the small set of emotions thought to be innate,
biologically based, pan-culturally experienced, shared with other
primates (possibly due to similar ancestral origins), and identi-
fiable via discrete, universal nonverbal expressions (1). Yet,
recent studies suggest that both emotions may meet several of
these criteria. Specifically, both are associated with distinct,
cross-culturally recognized nonverbal expressions, which resem-
ble the dominance and submission displays shown by nonhuman
The pride nonverbal expression is accurately identified by
children as young as 4-years old and adults from a range of
cultures including preliterate, highly isolated small-scale tradi-
tional societies, who are very unlikely to have learned the
expression through contact with other contemporary cultures
(2–4). The expression includes features such as expanded pos-
ture and head tilt back, behaviors similar to the ‘‘inflated
display’’ observed in dominant chimpanzees who have defeated
a rival (5), as well as the chest-beating intimidation displays seen
in mountain gorillas (6) and the ‘‘strutting. . . confident air’’ that
characterizes dominant Catarrhine monkeys (7). The shame
hanks to ABC’s ‘‘Wide World of Sports,’’ the word ‘‘victory’’
is, in the minds of many, inextricably associated with the
expression is also accurately identified across cultures, including
in the same isolated small-scale societies (4, 8, 9). Shame is
recognized from a simple head tilt downward, but based on
Darwin’s theory of antithesis (10) and the importance of ex-
panded posture in the pride expression, the full shame display
may include slumped shoulders and narrowed chest—behaviors
similar to the ‘‘cringing’’ and lowered posture associated with
submission in a range of animal species including chimpanzees,
macaques, baboons, rats, rabbits, crayfish, wolves, elephants,
seals, and salamanders (5, 11, 12). These findings raise the
possibility that pride and shame behavioral responses may be
human universals, evolved to serve unique adaptive functions.
Given that pride occurs in response to success, its nonverbal
expression may function to signal an individual’s success to
others, thereby boosting status. Emotion signals are thought to
have originated as purely functional (i.e., noncommunicative)
displays and over time became ‘‘ritualized’’ (i.e., simplified and
exaggerated) to the clearly communicative versions we see now
(13). Thus, the expanded posture and outstretched arms asso-
ciated with pride may have originated as a way of appearing
larger, allowing for the assertion of dominance and attracting
attention. The veracity of a behavioral signal may be established
on the basis of whether it is ‘‘handicapping’’—that is, perilous to
the sender (14). If individuals display signals despite inherent
risks (e.g., revealing oneself to a predator in the process of
alerting others to the danger), onlookers can trust the message’s
pride (and nonhuman primate dominance displays) may have
originated as a way of conveying the validity of the individual’s
belief in his/her dominance or success. Similarly, although
displaying behaviors associated with shame or submission re-
quires individuals to place themselves physically beneath adver-
saries and thus within their control, doing so may indicate the
veracity of their submission. This display likely originated as a
way of conveying acceptance of an aggressor’s power, thereby
the ancient submission display may have been ritualized into a
shame expression that also serves a secondary function: appeas-
ing onlookers who observed the failure (12, 15). By nonverbally
communicating an awareness of one’s transgression, the indi-
vidual can maintain his/her reputation as a trusted group
member who accepts social norms (16).
However, in both cases, these functionalist arguments are
premised on 2 central assumptions yet to be tested. First, are the
pride and shame behavioral expressions universally displayed
when individuals experience success and failure? It is possible
that individuals across cultures reliably recognize these expres-
Author contributions: D.M. designed research; J.L.T. performed research; J.L.T. analyzed
data; and J.L.T. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
†To whom correspondence should be addressed. E-mail: firstname.lastname@example.org.
This article contains supporting information online at www.pnas.org/cgi/content/full/
© 2008 by The National Academy of Sciences of the USA
August 19, 2008 ?
vol. 105 ?
no. 33 ?
sions not because they regularly see them, but rather because of
shared stereotypes (17). Furthermore, even if there is universal
agreement about behaviors that signify ‘‘pride’’ and ‘‘shame,’’
cultures could differ on whether those behaviors correspond to
success and failure. If pride and shame are not universally
to send messages relevant to these events.
The second question that needs to be addressed is whether the
pride and shame nonverbal expressions are likely to be innate
biological propensities rather than learned forms of social com-
munication. Even if individuals across cultures reliably display
these expressions in the predicted situations, we cannot know
whether they do so because they are modeling others or because
humans evolved to innately display these distinct behaviors,
perhaps as fixed action patterns, in these recurring, socially
important situations. To address this issue, we need to examine
spontaneous displays of pride- and shame-associated behaviors
in individuals who could not have learned to show them from
observing others (13). Thus, in the present research, we exam-
ined behavioral responses to success and failure in congenitally
blind individuals. These individuals have been unable to view
others’ expressions from birth or shortly thereafter and thus
cannot have learned to produce expressions through modeling.
If congenitally blind individuals display pride and shame expres-
sions in the same situations as sighted individuals, it would
provide compelling evidence for a biologically innate source of
these expressions, because it would be highly improbable for
blind individuals to have learned discrete behavioral configura-
tions that occur as automatic emotional reactions (13). This
conclusion is particularly likely if findings hold across congeni-
tally blind individuals from different countries and cultures.
Although no previous research has tested whether the recog-
nizable pride or shame expressions are cross-culturally displayed
in response to success and failure, several studies are consistent
with this possibility. Western children have been found to show
several components of both expressions in response to experi-
mentally manipulated achievements or failures (18–21) and
examination success (22). However, studies have not coded
behavioral responses to naturalistic successes and failures for all
components of the prototypical pride and shame expressions or
examined the issue cross-culturally. Furthermore, no previous
study has examined pride- and shame-associated behaviors in
blind individuals. Previous studies assessing spontaneous emo-
tional responses to naturalistic events have sought only those few
emotions that can be coded from the face alone (e.g., fear, anger,
happiness) (23–27); pride and shame expressions require head
and body movements outside the face (8, 28). Similarly, all
previous studies assessing blind individuals’ expressions exam-
ined only emotions shown in the face (e.g., 13, 29–31).
In the present research, we coded spontaneous behavioral
responses to winning or losing a judo match in the 2004 Olympic
and Paralympic Games. Sighted and blind athletes from 37
individualism (i.e., the extent to which emphasis is placed upon
the needs of the individual vs. the group) (32), traditional vs.
secular-rational values (i.e., the importance of religious and
traditional values vs. secular beliefs) (33), and survival vs.
self-expression values (i.e., the importance placed upon subsis-
tence and security vs. subjective well-being and self-expression)
(33). The latter 2 dimensions have been shown to account for the
concern. The former dimension has been shown to predict
differences in shame and pride subjective experiences, with
shame more strongly emphasized in collectivistic cultures and
35). Thus, if pride and shame responses to success and failure
differ across cultures, such differences would be most likely to
emerge in comparisons among cultures that vary on these 3
dimensions. Finally, we also tested whether larger geographic
region (i.e., regions with shared history and geography) influ-
enced these behavioral responses (36).
Results and Discussion
expression and several components of the shame expression
were spontaneously displayed in response to success and failure,
respectively. Specifically, pride-relevant behaviors of head-tilt
back, t (109) ? 4.13, d ? 0.84; smile, t (109) ? 6.85, d ? 1.45;
arms out from the body, t (107) ? 5.82, d ? 1.12; arms raised,
t (108) ? 5.37, d ? 1.03; hands in fists, t (106) ? 5.32, d ? 1.07;
chest expanded, t (102) ? 5.30, d ? 1.09; and torso pushed out,
t (107) ? 3.34, d ? 0.65; all ps ? 0.05; were greater in response
to winning than losing. In contrast, shame-relevant behaviors of
shoulders slumped forward, t (100) ? 4.10, d ? 0.82, and chest
narrowed, t (100) ? 3.12, d ? 0.62, both ps ? 0.05, were greater
in response to losing than winning (see Fig. 1). Losing did not
predict head-tilt down or face hiding, behavioral signatures of
the recognizable shame expression. In addition, winners were far
more likely than losers to show all pride components together
(i.e., the full pride expression), ?2(1) ? 24.75, P ? 0.05. Losers
were no more likely than winners to show the full shame
expression (head tilt down, face covered, and shoulders slumped
or chest narrowed), ?2(1) ? 0.52, ns, most likely because head
tilt down and face covering were not associated with failure.
These analyses are considerably more stringent than those
examining each component separately because spontaneously
be recognized from certain components alone (28, 37).
Neither gender nor any of the 3 cultural dimensions nor world
region moderated the effects of winning on pride behaviors.
Furthermore, in almost all cases pride-relevant behaviors were
shown to a greater extent in response to winning than losing
within each culture group [see supporting information (SI)]. The
full pride expression was also a more frequent response to
success than failure within each culture group, ?2(1) ? 7.45
(collectivistic), 9.71 (individualistic), 13.33 (traditional), 12.54
(secular), 13.18 (survival), and 9.89 (self-expression), all ps ?
0.05.§However, individualism/collectivism moderated the effect
of losing on the shame-relevant behavior of shoulders slumped,
B ? 0.30, P ? 0.05; the same interaction emerged with world
§These within-group analyses were not possible for world regions because in most cases
sample sizes were too small.
smilearms out arms
torso out chest
displayed in response to match wins and losses by sighted athletes, n ? 108.*,
P ? 0.05.
Mean levels of pride and shame nonverbal behaviors spontaneously
www.pnas.org?cgi?doi?10.1073?pnas.0802686105 Tracy and Matsumoto
region F(3, 83) ? 3.74, P ? 0.05.¶These interactions indicate a
weaker shame behavioral response among more individualistic,
these analyses at the country level (i.e., correlating mean be-
havioral responses to success and failure across all individuals
within a given country with country-level cultural dimension
scores). Based on these country-level analyses, none of the 3
dimensions were significantly correlated with any behavioral
responses except shoulders slumped and chest narrowed in
response to loss: these 2 behavioral responses to failure were
negatively correlated with individualism and self-expression
values, rs (19) ? ?0.53, ?0.53 (individualism) and rs (25) ?
?0.56, ?0.51 (self-expression), all ps ? 0.05; indicating that the
more individualistic and self-expression-valuing a given country
is, the less likely its athletes are to show the shame behavioral
response to failure.
One caveat to all these results is that behaviors may be due not
to the situation of winning vs. losing, but to personality. Thus, we
analyzed behaviors shown by 15 athletes (7 women) who both
winning again led to greater pride-relevant behaviors [i.e.,
smiling, t (14) ? 2.36 (Ms ? 1.51 vs. 0.35, d ? 0.93), arms
extended out, t (13) ? 4.98 (Ms ? 3.45 vs. 1.33, d ? 1.63), arms
raised, t (14) ? 2.52 (Ms ? 3.53 vs. 2.03, d ? 0.90), hands in fists,
t (13) ? 2.12 (Ms ? 2.50 vs. 0.79, d ? 0.94), and chest expanded,
t (12) ? 2.59 (Ms ? 2.28 vs. 1.31, d ? 0.77); all ps ? 0.05],
suggesting that the pride behavioral response to success can be
attributed to the situation of winning and not to the personality
of individuals who win. No differences emerged for shame-
We next tested whether pride- and shame-relevant behaviors
would remain significant predictors of win/loss outcomes when
controlling for other emotion-associated facial muscle move-
ments or ‘‘action units’’ (AUs) (38). In fact, AU 12 (lip corners
pulled up) and the pride behavior of arms extended out re-
mained significant when controlling for all other pride- and
happiness-relevant behaviors (B(exp)s ? 6.01, 3.67, respectively,
both ps ? 0.05). Both of these behaviors are part of the pride
expression; AU 12 is also part of happiness. When shared
variance between shame and sadness behaviors, shame and
anger behaviors, and shame and disgust behaviors was removed,
only shoulders slumped—a shame behavior—remained signifi-
cant in each equation, B(exp)s ? 0.30, 0.32, and 0.30, respec-
tively, all ps ? 0.05 (one-tailed). When shared variance between
shame and fear behaviors was removed, both shoulders slumped
and AU 1 (inner brow raiser—part of the fear expression)
remained significant, B(exp)s ? 0.23, 0.41, both ps ? 0.05. Thus,
unique components of both pride and shame expressions (arms
extended out and shoulders slumped) predicted win vs. loss
outcomes above and beyond what can be predicted from previ-
ously established emotion expressions, suggesting that shame
and pride expressions may be unique signals of success and
Turning to the blind athletes, all prototypical pride behaviors
were again shown to a greater extent in response to winning than
losing: head-tilt back, t (58) ? 1.86, d ? 1.11; smile, t (50) ? 3.13,
d ? 1.31; arms out, t (58) ? 3.66, d ? 1.05; arms raised, t (58) ?
4.48, d ? 1.26; hands in fists, t (57) ? 2.57, d ? 0.78; chest
expanded, t (58) ? 5.20, d ? 1.52; and torso pushed out, t (58) ?
4.62, d ? 1.46; all ps ? 0.05 (one-tailed for head tilt back). In
addition, the 2 shame-relevant behaviors shown by sighted
athletes, chest narrowed and shoulders slumped, were shown by
blind athletes in response to failure, t (58) ? 2.14, d ? 0.57, P ?
0.05, and t (58) ? 1.89, d ? 0.50, P ? 0.05 one-tailed. None of
world region, or gender. Winners were again far more likely than
losers to show the full pride expression, ?2(1) ? 5.28, P ? 0.05;
losers were again no more likely than winners to show full the
shame expression, ?2(1) ? 3.64, ns.
The effects of winning on pride-relevant behaviors were not
moderated by blind status (i.e., congenital blindness vs. later
onset). However, blind status did moderate the effect of losing
on both shame-relevant behaviors, Fs (1, 31) ? 8.82, 6.42 for
shoulders slumped and chest narrowed, respectively, both ps ?
0.05, such that a larger behavioral response emerged in the
congenitally blind athletes; across the 2 behaviors, Ms ? 3.33
(failure) vs. 0.63 (success) for congenitally blind individuals, and
1.92 (failure) vs. 1.63 (success) for later-onset blind individuals.
Thus, the shame behavioral response to failure held within the
congenitally blind sample, t (10) ? 2.59, d ? 1.97, for shoulders
slumped; and t (10) ? 2.58, d ? 1.95, for narrowed chest, both
ps ? 0.05. In addition, the pride behavioral response to success
largely held within the congenitally blind sample: winners
t (7) ? 2.06, d ? 1.46, chest expanded, t (9) ? 3.15, d ? 1.88, and
torso pushed out; t (8) ? 3.25, d ? 2.04, all ps ? 0.05 (one-tailed
for arms raised and hands in fists; see Fig. 2). Effects for arms
extended and smiling were in the expected direction but did not
reach significance. However, we computed a scale based on the
mean of all pride-relevant behaviors (? ? 0.76) and found higher
scale scores for winners compared to losers within the congen-
itally blind sample, t (10) ? 2.05, d ? 1.74, P ? 0.05, one-tailed.
Thus, it appears that individuals who have never seen others
show pride and shame expressions in response to success and
failure spontaneously show precisely these expressions in these
situations. The large effect sizes that emerged within this sample
make it unlikely that the inclusion of additional participants—
even those who did not show the predicted behaviors—would
reduce effects to nonsignificance (39).
expressions on the basis of spontaneous, nonverbal behaviors
shown by sighted and blind individuals across cultures, in re-
sponse to the same naturalistic situation. The findings demon-
pride expression are displayed in response to success by indi-
viduals from collectivistic; individualistic; tradition-, secular-,
survival-, and self-expression-valuing cultures and by sighted,
¶World region also moderated the effect of losing on chest narrowed, F(3, 83) ? 3.51, P ?
0.05, but this effect did not hold in the smaller sample with independent data only.
P ? 0.05, but this effect also did not hold in the smaller sample.
displayed in response to match wins and losses by congenitally blind athletes,
n ? 12,*, P ? 0.05.
Mean levels of pride and shame nonverbal behaviors spontaneously
Tracy and MatsumotoPNAS ?
August 19, 2008 ?
vol. 105 ?
no. 33 ?
blind, and congenitally blind individuals across cultures. In all of
these analyses, success had a large effect on the display of
pride-relevant behaviors (39), which could not be attributed to
a third-factor personality variable or to shared variance with
facial expressions of happiness.
Second, several components of the shame expression
(slumped shoulders and narrowed chest) are displayed in re-
sponse to failure by sighted, blind, and congenitally blind
individuals. These findings could not be attributed to shared
variance with any other negative emotion expression; in fact,
shame-relevant behaviors were a better predictor of whether an
individual lost than were behaviors associated with any other
negative emotion except fear. However, the shame behavioral
response was weaker in sighted athletes from individualistic,
self-expression-valuing cultures within West Eurasian and North
American regions. In addition, the 2 behaviors previously asso-
ciated with the recognizable shame expression (head-tilt down,
averting/hiding the face) were not part of the spontaneous
behavioral response to failure. Findings from the congenitally
blind sample help clarify these ambiguities, as discussed below.
Implications. These findings imply, first, that the cross-culturally
recognized pride expression is not simply a widely held stereo-
type, but rather is a discrete behavioral configuration actually
produced in ecologically valid situations and may be an evolved
and innate behavioral response to success. The pride behaviors
across cultures; the only exception was the absence of hands on
not reliably displayed during a success experience. The finding
that congenitally blind individuals who could not have learned
to show the pride expression from watching others nonetheless
displayed these same behaviors in the same situation (see Fig. 3)
suggests that this behavioral response to success is unlikely to be
learned. Although parents may teach young children to engage
in some of these behaviors through direct physical contact (e.g.,
moving a child’s arms above his/her head), it is unlikely that
parents would or could teach the full configuration of behaviors
(e.g., expanded chest, hands in fists) in this manner?. Thus, the
most parsimonious interpretation of these findings is that con-
genitally blind individuals engage in these behaviors in response
to success because humans have an innate biological propensity
to do so (13).
Overall then, the pride expression appears to meet one of the
central criteria for a functional universal (i.e., a psychological
entity that evolved to serve a particular adaptive function): it is
recognized and displayed across cultures in the same contexts
and situations (40, 41). These findings are thus consistent with
theoretical accounts of pride as an evolutionary adaptation for
securing status. By responding to success with behaviors that
expand the body and are reliably identified as pride, individuals
advertise their accomplishment, and thereby may ensure their
continued status and acceptance within their social group.
Similarly, the shame-relevant behaviors of shoulders slumped
and chest narrowed are not simply stereotypes associated with
shame but rather are behavioral responses actually produced in
ecologically valid shame-eliciting situations and thus may rep-
resent an evolved and innate behavioral response to failure.
Somewhat surprisingly, the expression previously found to be
recognized as shame (head tilt down, face covered) was not
shown in response to failure. However, this may be due to the
methodology used; the single photographer, who often had to
shoot from behind athletes, may not have captured all facial/
head movements. Regardless, it seems clear that the bodily
components of shame are spontaneously displayed in response
However, among individuals from individualistic, self-
expression-valuing, West Eurasian and North American cul-
tures, even these behaviors were not reliably associated with
failure. One explanation for this cultural difference is that these
athletes felt shame but suppressed its expression, in accordance
with cultural norms that stigmatize the display of shame and
(34, 42). In contrast, athletes from more collectivistic nations,
where shame is an appropriate response to social trespass and a
socially valued emotion, would not have needed to suppress their
shame in response to public failure (34, 43). The finding that
congenitally blind individuals from a range of cultures displayed
shame behaviors in response to failure, and did so to a greater
extent than individuals who acquired blindness later in life,
supports this interpretation.** Individuals who have never seen
of culture-specific norms of how emotions should be regulated,
and may be generally less sensitive to distinctions between
that these individuals showed the greatest evidence of a shame
behavioral response suggests that these behaviors are the
evolved, innate response, and the absence of a clear shame
expression among sighted athletes from certain cultures repre-
sents culture-specific emotion regulation. Although it is also
possible that these sighted athletes simply felt less shame after
losing, their lack of a shame response is unlikely to indicate that
the expression does not generalize to these cultures, given
previous evidence of shame recognition in American cultures
and behavioral displays of shame in response to failure among
American children and adolescents (8, 9, 18, 20).
At a broader level, these findings suggest that the expressions
associated with shame and pride can be assessed from sponta-
neous nonverbal behaviors. This finding highlights the impor-
tance of the body in emotion expression. Recent research has
demonstrated that bodies and faces are perceived through
similar cognitive and neural processes (44, 45); thus, it might be
fruitful to devote greater research attention to the role of the
body in emotion expression. Ethologically oriented researchers
?It also is unlikely that these behaviors were verbally or physically taught by judo coaches
or others involved in the sport; athletes are never instructed on specific nonverbal
behaviors to show after success or failure, nor are their limbs or body moved in any
particular manner in these situations.
**Within the congenitally blind sample, individualism/collectivism scores ranged from
20–89, M ? 54; 45% of these individuals were from survival-valuing nations, and 55%
from self-expression-valuing nations.
congenitally blind (right) athlete.
Pride expression in response to victory shown by a sighted (left) and
www.pnas.org?cgi?doi?10.1073?pnas.0802686105 Tracy and Matsumoto
interested in nonverbal emotion communication have long em-
phasized the role of posture (10, 13, 46), but the facial muscu-
lature has since received the lion’s share of research emphasis.
The development of a system for measuring basic emotions from
observable facial behaviors largely revolutionized the field of
emotion research (38), and the present findings, particularly the
pride and shame behavioral coding system that was used and
validated, provide a tool that may have similarly wide and varied
Limitations.Although it is likely that athletes felt pride and shame
in response to some of the most important successes and failures
of their lives, future studies should verify that these expressions
are associated with subjective feelings of pride and shame by
measuring nonverbal behaviors along with self-reports. None-
theless, regardless of these individuals’ subjective experiences,
the fact that they responded to success by showing behaviors
previously associated with pride and to failure with behaviors
previously associated with shame is informative about the
evolved signaling function of these behaviors and associated
emotions. A second limitation is that, because athletes per-
formed in front of an audience, we cannot rule out the possibility
that their expressions were intentional social communications.
However, it is highly unlikely that congenitally blind individuals
thought about the appearance of their expressive behaviors
enough to intentionally invoke (or inhibit) them.
The present findings add to our understanding of emotion
expression in several ways. By providing evidence that the
behaviors cross-culturally recognized as nonverbal expressions
of pride and shame are displayed in response to success and
failure by sighted and blind individuals across cultures, these
findings demonstrate that: (a) these expressions are not simply
stereotypes intuitively associated with pride and shame but
rather may be biologically innate behavioral responses to success
and failure; (b) the emotions of pride and shame may have
evolved, innate nonverbal expressions, challenging a longstand-
ing assumption in the emotion literature that only a small set of
emotions fit within the Darwinian framework; and (c) these
emotions may be assessed without reliance on self-report. In
sum, these findings support evolutionary accounts of pride and
shame as affective mechanisms of promoting and inhibiting
to the research goals) photographed athletes during and immediately after
each match, repeatedly for approximately 15 s, using a Nikon D2H profes-
sional digital camera (4.1 megapixels effective, 8 frames/s, 37 ms shutter-time
in JPEG formats. The ISO range was between 400 and 800, producing shutter
speeds of approximately 1/500th s, allowing for a series of moment-by-
moment images of each behavioral response. Although some photos showed
only the athlete’s back or profile, all were included to obtain the maximum
amount of information; photos that could be coded only for body, arm, or
head movements were coded only on those dimensions.
Athletes. The sighted-athlete sample included 87 competitors (42 winners, 45
losers; 46% female) from 36 nations. Twenty-two of these individuals were
photographed in more than 1 match (e.g., semifinals and finals), producing a
20 nations. Seven of these individuals were photographed in more than 1
and losers, but only those that held in the smaller set (based on the last match
each athlete fought) are included to avoid issues associated with noninde-
pendent data.††Blind status was available for 68% of the full blind sample; of
these, 29% (n ? 12) were congenitally blind.
cultural dimensions: individualism/collectivism, secular-rational/traditional
values, and survival/self-expression values (32, 33). Individualism/collectivism
Belarus, Cuba, Georgia, Moldova, Mongolia, N. Korea, Slovenia, Tunisia,
Ukraine, ns ? 26 sighted athletes, 10 blind athletes). Secular-rational/
traditional values scores ranged from ?1.65 to 1.84; survival/self-expression
values scores ranged from ?1.86 to 2.05. For both dimensions scores were
based on Inglehart’s country-level findings of 2 dimensions of cross-cultural
variation (33), and were unavailable for 4 nations (Cuba, Mongolia, North
Korea, Tunisia, ns ? 9 sighted athletes, 2 blind athletes). Finally, each partic-
††Results for the smaller set are available from the first author.
Table 1. Nonverbal behaviors coded, with interrater reliability alphas, emotion predictions, and outcomes
actually associated with each behavior
Behavior Interrater alphas Predicted emotionActual outcome References for prediction
Moving hands to cover face
Hiding face by moving
One/both arms out from body
One/both arms raised
One/both hands in fists
Hands on hips
Torso pushed out/leaning back
Chest narrowed inward
2–4, 10, 18, 20, 21, 28, 47–50
4, 8, 9, 18, 47, 50–52
2–4, 18, 20, 28, 47, 53
8–10, 47, 51, 52
2–4, 19, 28, 47, 49, 54
2, 3, 28, 19, 47, 49, 54
2–4, 10, 18, 22, 28, 47, 53
2–4, 28, 18, 47, 20, 49
18, 20, 47, 48
10, 18, 20, 47, 48
Alpha reliabilities are first reported for the sighted sample, then, in parentheses, for the blind sample. References indicate previous
studies that demonstrated an association between the behavior and either knowledge of the relevant emotion or success/failure
Tracy and MatsumotoPNAS ?
August 19, 2008 ?
vol. 105 ?
no. 33 ?
ipant was scored as belonging to 1 of Murdock’s 6 world regions, a division of
the world’s nations based on shared history and geography (36; see SI).
Pride and Shame Behavioral Coding. All photos taken after match completion
(omitting those portraying physical interactions with opponents) were coded
for pride- and shame-relevant behaviors, based on previous research (see
study goals) rated the intensity of each movement on a scale from 0 (‘‘not at
all present’’) to 1 (‘‘visible but very mild intensity’’) to 5 (‘‘extreme intensity’’).
were represented by several photos, so the first coder to rate a match deter-
mined where each movement began and ended, then coded behaviors across
those photos. All photos were subsequently rated by 1 or 2 (nonblind sample)
or 3 (blind sample) other coders, who followed this delineation. If an athlete
was photographed making several movements, each was coded separately.
Total scores for an athlete’s behavioral responses to a match were computed
by taking the mean rating for each item (across coders) for each movement,
and then taking the highest mean rating across all movements. We used
highest mean ratings instead of overall means to ensure that athletes were
scored for their largest movement that was captured, without giving greater
weight to athletes who were photographed making more movements. Be-
havioral responses were thus operationalized as the intensity with which a
single (most intensely recorded) movement was displayed and not the fre-
quency with which a movement was displayed.
Facial Action Coding. For 69% of the sighted athletes (ns ? 62 for the full
Action Coding System (38). These expressions were coded by 2 certified FACS
coders; interrater reliability, calculated by doubling the number of codes on
5 (‘‘extreme intensity’’).
ACKNOWLEDGMENTS. We thank Bob Willingham for allowing us to use his
photographs for this research and the Social Sciences and Humanities Re-
search Council of Canada File 410-2006-1593.
1. Ekman P (2003) Emotions Revealed (Times Books, New York).
2. Tracy JL, Robins RW (2004) Show your pride: Evidence for a discrete emotion expres-
sion. Psychol Sci 15:194–197.
4. Tracy JL, Robins RW (2008) The nonverbal expression of pride: Evidence for cross-
cultural recognition. J Pers Soc Psychol 94:516–530.
5. DeWaal F (1989) Chimpanzee Politics (Johns Hopkins, Baltimore, MD).
6. Schaller GB (1963) The Mountain Gorilla: Ecology and Behavior (Univ of Chicago Press,
7. Maslow AH (1936) The role of dominance in the social and sexual behavior of infra-
human primates: I. Observations at Vilas Park Zoo. J Gerontol 48:261–277.
8. Keltner D (1995) Signs of appeasement: Evidence for the distinct displays of embar-
rassment, amusement, and shame. J Pers Soc Psychol 68:441–454.
9. Izard CE (1971) The Face of Emotion (Appleton-Century-Crofts, East Norwalk, CT).
10. Darwin C (1872) The Expression of the Emotions in Man and Animals (Oxford Univ
Press, New York).
11. Maslow AH (1936) The role of dominance in the social and sexual behavior of infra-
12. Keltner D, Buswell BN (1997) Embarrassment: Its distinct form and appeasement
functions. Psychol Bull 122:250–270.
13. Eibl-Eibesfeldt I (1989) HumanEethology (DeGruyter, New York).
14. Zahavi A, Zahavi A (1997) The Handicap Principle: A Missing Piece of Darwin’s Puzzle
(Oxford Univ Press, New York).
15. Keltner D, Buswell BN, Young R (1997) Appeasement in human emotion, personality,
and social practice. Aggress Behav 23:359–374.
16. Fessler DMT (2007) in The Self-Conscious Emotions: Theory and Research, eds Tracy JL,
Robins RW, Tangney JP (Guilford, New York), pp 174–193.
17. Russell JA (1994) Is there universal recognition of emotion from facial expressions? A
review of the cross-cultural studies. Psychol Bull 115:102–141.
18. Belsky J, Domitrovich C, Crnic K (1997) Temperament and parenting antecedents of
individual difference in three-year-old boys’ pride and shame reactions. Child Dev
19. Heckhausen H (1984) Emergent achievement behavior: Some early developments.
Advan Motiv Achiev 3:1–32.
20. Lewis M, Alessandri SM, Sullivan MW (1992) Differences in shame and pride as a
function of children’s gender and task difficulty. Child Dev 63:630–638.
21. Stipek DJ, Recchia S, McClintic S (1992) Self-evaluation in young children. Monogr Soc
Res Child Dev 57:Serial No. 226.
22. Weisfeld GE, Beresford JM (1982) Erectness of posture as an indicator of dominance or
success in humans. Motiv Emot 6:113–131.
23. Kraut RE, Johnson RE (1979) Social and emotional messages of smiling: An ethological
approach. J Pers Soc Psychol 37:1539–1553.
24. Fernandez-Dols JM, Ruiz-Belda MA (1995) Are smiles signs of happiness: Gold medal
winners at the Olympic Games. J Pers Soc Psychol 69:1113–1119.
25. Medvec VH, Madey SF, Gilovich T (1995) When less is more: Counterfactual thinking
and satisfaction among Olympic medalists. J Pers Soc Psychol 69:603–610.
26. Matsumoto D, Willingham B (2006) The thrill of victory and the agony of defeat:
Spontaneous expressions of medal winners of the 2004 Athens Olympic Games. J Pers
Soc Psychol 91:568–581.
27. Ruiz-Belda MA, Fernandez-Dols JM, Carrera P, Barchard K (2003) Spontaneous facial
expressions of happy bowlers and soccer fans. Cognit Emot 17:315–326.
28. Tracy JL, Robins RW (2007) The prototypical pride expression: Development of a
nonverbal behavioral coding system. Emotion 7:789–801.
29. Cole PM, Jenkins PA, Shott CT (1989) Spontaneous expressive control in blind and
sighted children. Child Dev 60:683–688.
blind and sighted children aged 8–11. J Vis Impair Blindness:418–428.
31. Peleg G, Katzir G, Peleg O, Kamara M, Brodsky L, et al. (2006) Hereditary family
signature of facial expression. Proc Natl Acad Sci USA 103:15921–15926.
32. Hoftstede G (2001) Culture’s Consequence: Comparing Values, Behaviors, Institutions,
and Organizations Across Nations (Sage, Thousand Oaks, CA).
33. Inglehart R, Norris P (2003) Rising Tide: Gender Equality and Cultural Change Around
the World (Cambridge Univ Press, Cambridge, UK).
34. Markus H, Kitayama S (1991) Culture and the self: Implications for cognition, emotion,
and motivation. Psychol Rev 98:224–253.
intranational differences. J Pers Soc Psychol 81:869–885.
36. Murdock GP (1949) Social Structure (MacMillan, New York).
37. Carroll JM, Russell JA (1997) Facial expressions in Hollywood’s portrayal of emotion. J
Pers Soc Psychol 72:164–176.
38. Ekman P, Friesen WV (1978). Facial Action Coding System: Investigator’s Guide.
(Consulting Psychologists, Palo Alto, CA).
39. Cohen J (1962) The statistical power of abnormal-social psychological research: A
review. J Abnorm Soc Psychol 65:145–153.
eds Berry JW, Poortinga YH, Pandey J (Allyn and Bacon, Boston), pp 43–83.
know? Psychol Bull 131:763–784.
42. Scheff TJ (1988) Shame and conformity: The deference-emotion system. Amer Sociol
43. Wong L, Tsai J (2007) in The Self-Conscious Emotions: Theory and Research, eds Tracy
JL, Robins RW, Tangney JP (Guilford, New York), pp 209–223.
of facial expression and emotional body language. Proc Natl Acad Sci USA 102:16518–
45. Slaughter V, Stone VE, Reed C (2004) Perception of faces and bodies: Similar or
different? Curr Dir Psychol Sci 13:219–223.
46. Hinde RA (1972) Non-Verbal Communication (Cambridge Univ Press, Oxford, UK).
47. Geppert U (1986) A Coding System for Analyzing Behavioral Expressions of Self-
Evaluative Emotions (Max Planck Institute for Psychological Research, Munich, Ger-
48. Wallbott HG (1998) Bodily Expression of Emotion. Eur J Soc Psychol 28:879–896.
49. Coulson M (2004) Attributing emotion to static body postures: Recognition accuracy,
confusions, and viewpoint dependence. J Nonverbal Behav 28:117–139.
50. Zivin G (1977) Facial gestures predict preschoolers’ encounter outcomes. Soc Sci Info
expressions and a gradient of recognition. Cognit Emot 13:225–266.
52. Keltner D, Buswell BN (1996) Evidence for the distinctness of embarrassment, shame,
and guilt: A study of recalled antecedents and facial expressions of emotion. Cognit
53. Kasari C, Sigman MD, Baumgartner P, Stipek DJ (1993) Pride and mastery in children
with autism. J Child Psychiatry 34:353–362.
54. Reissland N, Harris P (1991) Children’s use of display rules in pride-eliciting situations.
Brit J Dev Psychol 9:431–435.
www.pnas.org?cgi?doi?10.1073?pnas.0802686105Tracy and Matsumoto
INAUGURAL ARTICLE, APPLIED PHYSICAL SCIENCES. For the article
‘‘Inaugural Article: Dynamic interfaces in an organic thin film,’’
by Chenggang Tao, Qiang Liu, Blake S. Riddick, William G.
Cullen, Janice Reutt-Robey, John D. Weeks, and Ellen D.
Williams, which appeared in issue 43, October 28, 2008, of Proc
Natl Acad Sci USA (105:16418–16425; first published September
2, 2008; 10.1073?pnas.0805811105), the authors note that the
author name Blake S. Riddick should have appeared as Blake C.
Riddick. The author line has been corrected online. The cor-
rected author line and related author contributions footnote
Chenggang Tao, Qiang Liu, Blake C. Riddick, William G.
Cullen, Janice Reutt-Robey, John D. Weeks, and Ellen D.
Author contributions: C.T., J.D.W., and E.D.W. designed research; C.T., Q.L., and W.G.C.
performed research; W.G.C. and J.R.-R. contributed new reagents/analytic tools; C.T., Q.L.,
B.C.R., and E.D.W. analyzed data; and C.T., W.G.C., J.R.-R., J.D.W., and E.D.W. wrote the
PSYCHOLOGY. For the article ‘‘The spontaneous expression of
pride and shame: Evidence for biologically innate nonverbal
displays,’’ by Jessica L. Tracy and David Matsumoto, which
appeared in issue 33, August 19, 2008, of Proc Natl Acad Sci USA
(105:11655–11660; first published August 11, 2008; 10.1073?
pnas.0802686105), the authors note that Fig. 3 is copyrighted by
its corrected legend appear below.
IMMUNOLOGY. For the article ‘‘Reciprocal patterns of methylation
of H3K36 and H3K27 on proximal vs. distal IgVH genes are
modulated by IL-7 and Pax5,’’ by Cheng-Ran Xu, Lana Schaffer,
Steven R. Head, and Ann J. Feeney, which appeared in issue 25,
June 24, 2008, of Proc Natl Acad Sci USA (105:8685–8690; first
published June 17, 2008; 10.1073?pnas.0711758105), the authors
note that due to a printer’s error, Fig. 2 appeared incorrectly and
was a duplicate of Supporting Information Fig. S2. The correct
figure and its legend appear below.
congenitally blind (Right) athlete. [Reproduced with permission (copyright
2004, Bob Willingham).]
Pride expression in response to victory shown by a sighted (Left) and
using antibodies reactive with H3K27me3 or H3K36me2 on pro-/pre-B cells
liver, and pro-B cells (B220?CD19?) from 3- to 4-week-old ?MT bone marrow.
(B) ChIP assays were performed by using Ezh2 antibody on pro-B cells from
?MT fetal liver and 3- to 4-week-old ?MT bone marrow. Data are presented
as relative to the positive control of the Neuregulin gene (Neuregulin ? 1).
The patterns of histones H3K27me3 and H3K36me2 and Ezh2 in fetal
December 16, 2008 ?
vol. 105 ?
no. 16 www.pnas.org