Download Can Duchenne Smiles Be Feigned? New Evidence on Felt and

Emotion
2009, Vol. 9, No. 6, 807– 820
© 2009 American Psychological Association
1528-3542/09/$12.00 DOI: 10.1037/a0017844
Can Duchenne Smiles Be Feigned?
New Evidence on Felt and False Smiles
Eva G. Krumhuber
Antony S. R. Manstead
University of Geneva
Cardiff University
We investigated the value of the Duchenne (D) smile as a spontaneous sign of felt enjoyment.
Participants either smiled spontaneously in response to amusing material (spontaneous condition) or were
instructed to pose a smile (deliberate condition). Similar amounts of D and non-Duchenne (ND) smiles
were observed in these 2 conditions (Experiment 1). When subsets of these smiles were presented to other
participants, they generally rated spontaneous and deliberate D and ND smiles differently. Moreover,
they distinguished between D smiles of varying intensity within the spontaneous condition (Experiment
2). Such a differentiation was also made when seeing the upper or lower face only (Experiment 3), but
was impaired for static compared with dynamic displays (Experiment 4). The predictive value of the D
smile in these judgment studies was limited compared with other features such as asymmetry, apex
duration, and nonpositive facial actions, and was only significant for ratings of the upper face and static
displays. These findings raise doubts about the reliability and validity of the D smile and question the
usefulness of facial descriptions in identifying true feelings of enjoyment.
Keywords: emotion, facial expression, Duchenne smile, smile perception
smiles and as indicators of a positive emotion. In contrast, the
orbicularis oculi action (AU 6) should be absent in false or voluntary smiles (Ekman, 1993; Ekman & Friesen, 1982), which
involve movement only in the mouth region (AU 12, nonDuchenne [ND] smiles). These false smiles have also been labeled
social, polite, or masking smiles (Ekman, 1989). Although the
motives for posed smiles vary, their common function is to convince another person that enjoyment is occurring when it is not
(Ekman & Friesen, 1982; Frank, Ekman, & Friesen, 1993).
It is argued that this distinction between smile types has its
origins in different neural pathways (Frank & Ekman, 1993; Rinn,
1991). Most people are able to activate the zygomaticus major
muscle voluntarily; by contrast, the upper face muscles are less
involved in learned activities, with most people (80%) being
unable to contract the orbicularis oculi, pars lateralis muscle voluntarily (Ekman, Roper, & Hager, 1980; Levenson, Ekman, &
Friesen, 1990). The orbicularis oculi muscle (AU 6) involved in D
smiles has consequently been labeled a reliable muscle (Ekman,
1993, p. 390). Because this muscle is difficult to control, it is not
available for use in false expressions (Ekman, 1985, 1989); therefore, it provides a reliable and spontaneous sign of enjoyment. The
present research focused on this voluntary aspect and investigated
how reliable and valid the D smile actually is with respect to its
expression and perception.
According to discrete emotion theory, facial expressions show
emotions and are intrinsically linked to feelings (Ekman, 1982,
1992a). In this research tradition, the Duchenne (D) smile has been
proposed as a spontaneous and genuine expression of positive
emotions such as happiness, pleasure, or enjoyment (Ekman,
1992b; Ekman, Davidson, & Friesen, 1990; Frank & Ekman,
1993). This type of smile is named after the French neuroanatomist
Duchenne de Boulogne who described the smile and suggested
that it is “put into play by the sweet emotions of the soul . . .”
(Duchenne, 1862/1990, p. 72). In the Facial Action Coding System
(FACS; Ekman & Friesen, 1978; Ekman, Friesen, & Hager, 2002),
a system for coding visible facial movements, the appearance of
the D smile has been classified as a combination of two facial
muscle movements: the zygomaticus major muscle (Action Unit
[AU] 12 in FACS), which pulls the lip corners up, thereby producing a smiling mouth, and the orbicularis oculi, pars lateralis
muscle (AU 6 in FACS), which causes a lifting of the cheeks,
narrowing of the eye opening, and wrinkles around the eye socket
colloquially called crow’s feet.
It has been argued that only smiles that involve both facial
actions (i.e., D smiles) should be regarded as spontaneous, felt
Eva G. Krumhuber, Swiss Center for Affective Sciences, University of
Geneva; Antony S. R. Manstead, School of Psychology, Cardiff University.
We thank Magdalena Cholakova and Nicoletta Dimitrova for their help
with Experiment 1 and Gwenda Simons for FACS coding. We are grateful
to Paula Niedenthal, Geoff Thomas, and two anonymous reviewers for
valuable comments on previous versions of this article.
Correspondence concerning this article should be addressed to Eva
Krumhuber, Swiss Center for Affective Sciences, CISA—University of
Geneva, 7 Rue des Battoirs, 1205 Geneva, Switzerland. E-mail:
[email protected]
The Expression of Duchenne Smiles
There is evidence that D smiles occur when positive emotions
are elicited and when people enjoy themselves. Specifically, D
smiles are shown more often (than ND smiles) in response to
pleasant stimuli (Ekman et al., 1990; Soussignan & Schaal, 1996)
and when positive feelings such as amusement and enjoyment are
reported (Ekman, Friesen, & O’Sullivan, 1988; Frank et al., 1993).
D smiles are also associated with self-reports of love and good
807
808
KRUMHUBER AND MANSTEAD
mood (Gonzaga, Keltner, Londahl, & Smith, 2001; Scherer &
Ceschi, 2000), marital satisfaction (Harker & Keltner, 2001), and
health improvement in clinical populations (Ekman, Matsumoto,
& Friesen, 1997; Steiner, 1986). In children, more D smiles (than
ND smiles) occur in response to a familiar adult (Fox & Davidson,
1988), when children are gazing at their smiling mother (Messinger, Fogel, & Dickson, 1999, 2001), and during success in an
achievement game (Schneider & Unzer, 1992).
These studies show that D smiles tend to be shown under
circumstances of spontaneously experienced positive affect. However, there is also evidence that D smiles occur in posed and
negative contexts. Specifically, D smiles have been observed both
when participants smiled spontaneously and when they were instructed to do so (Schmidt, Ambadar, Cohn, & Reed, 2006). For
such posed smiles, the majority of the participants have been found
to show orbicularis oculi—AU 6 —activation (Schmidt, Bhattacharya, & Denlinger, 2009; Smith, Smith, & Ellgring, 1996); in
one study of 105 posed smiles, 67% were found to be D smiles
(Schmidt & Cohn, 2001). D smiles have also been found to occur
in response to negative film clips (Ekman et al., 1990), when
negative emotions are being concealed (Ekman et al., 1988), after
failure in a game (Schneider & Josephs, 1991; Zaalberg, Manstead,
& Fischer, 2002), during tickling in the absence of positive affect
(Harris & Alvarado, 2005), in expressions of embarrassment (Keltner, 1995), and when talking about negative events (Lee & Beattie,
1998). Bonnano and colleagues observed D laughter and smiles
when participants talked about a recently deceased spouse (Keltner
& Bonanno, 1997) and in response to a negative affect manipulation during a monologue task (Papa & Bonanno, 2008). It is
interesting that these D expressions and only these D expressions
were related to a reduction or dissociation of negative affect and to
better social relations at a later time. Ansfield (2007) suggested
that smiles serve a self-regulatory function of coping with negative
emotional experiences. In his study, the more participants smiled
while viewing distressing videos, the more positive affect they
reported afterward. The D expression, whether made spontaneously or deliberately, may therefore foster positive self-regulation
and lead to relevant physiological and affective changes through
facial feedback mechanisms.
Supportive evidence comes from Soussignan (2002), who
showed that through the unconscious facilitation of D smiles by
having participants hold a pen between their teeth, self-reports of
positive experience, skin conductance, and heart rate in response to
pleasant film clips all increased. These associations were not found
for ND smiles of similar intensity. However, the conscious voluntary production of D smiles by muscle-by-muscle instruction has
also been shown to generate relevant physiological changes (Ekman & Davidson, 1993). When contracting the zygomaticus major
and the orbicularis oculi, pars lateralis muscles on command,
patterns of regional brain activity were similar to those of spontaneous displays of D smiles. Moreover, 32 of 45 undergraduates
(71.1%) could produce such deliberate D smiles.
Such findings suggest that D smiles exert a positive feedback on
subjective and physiological variables that resemble those occurring during spontaneous positive affect. However, and in contrast
to what has been previously claimed, D smiles do not only occur
as a sign of genuine felt enjoyment. People can and do display D
smiles under deliberate circumstances and in the absence of positive feelings. The suggestion that 80% of the population would be
unable to voluntarily contract the orbicularis oculi, pars lateralis
muscle and thus be capable of producing a false D smile (Ekman
et al., 1990; Frank & Ekman, 1993) should therefore be treated
with caution. Indeed, the evidence generally cited in support of this
suggestion comes from observations and research with young
children in whom AU 6 was not measured at all (Ekman et al.,
1980) or from studies showing that happy D smiles were rated
actually as the easiest configuration to produce deliberately by
trained actors, college students, and local people (Levenson et al.,
1990).
The Perception of Duchenne Smiles
There are several studies showing that D and ND smiles are
perceived differently. Individuals displaying D smiles are evaluated as more emotional and interpersonally positive (Frank et al.,
1993; Harker & Keltner, 2001; Messinger, Cassel, Acosta, Ambadar, & Cohn, 2008), as being in better humor (Scherer & Ceschi,
2000), and as expressing more pleasantness (Soussignan & Schaal,
1996) than their ND smiling counterparts. D smiles also elicit more
positive emotional and affiliative responses in other people
(Gonzaga et al., 2001; Keltner & Bonanno, 1997; Surakka &
Hietanen, 1998), contribute to perceptions of greater spontaneity
and authenticity (Gosselin, Beaupré, & Boissonneault, 2002; Gosselin, Perron, Legault, & Campanella, 2002; Giudice & Colle,
2007; Hess & Kleck, 1994), and lead to more positive evaluations
of peripheral, unrelated attributes (Peace, Miles, & Johnston,
2006).
It is interesting that contraction of the orbicularis oculi muscle
occurs in negative emotions, not only in positive ones. Ekman and
Friesen (1982) noted that the outer part of the orbicularis oculi
muscle is also involved when distress, pain, or sadness are felt.
Messinger and colleagues (Bolzani Dinehart et al., 2005; Messinger, 2002) showed that eye constriction (AU 6) had significant
effects on raters’ perceptions of the intensity of positive emotions
as well as negative ones. In particular, when eye constriction
accompanied infants’ smiles, it made positive expressions appear
more positive; when it accompanied cry faces, it made these
expressions appear more negative. The same facial movement was
therefore seen as signaling increased intensity in two qualitatively
different emotions. A further complexity is that the D smile can
also occur in conjunction with facial features (i.e., frowning,
dimpling, lip tightening, lip corner depressing) indicative of negative emotions (Ekman, 1985). Such modulations of smiling have
been found in spontaneous as well as in deliberate or social smiles
(Schmidt, Cohn, & Tian, 2003; Schmidt et al., 2009) and have
been shown to influence observers’ ratings of the happiness and
amusement of the sender (Ambadar, Cohn, & Reed, 2009; Gosselin, Beaupré, & Boissonneault, 2002; Hess, Kappas, McHugo,
Kleck, & Lanzetta, 1989). Therefore, neither the D smile nor its
distinctive feature, AU 6, is uniquely associated with the perception of positive emotion; both can co-occur with nonpositive facial
actions.
Furthermore, factors other than the D marker have been proposed as differentiating between spontaneous and happy and posed
smiles (see Frank & Ekman, 1993; Frank et al., 1993). One such
factor is the greater symmetry of spontaneous smiles. Several
studies found that smiles made spontaneously in reaction to an
experimenter’s joke or a pleasant film were more symmetrical than
FELT AND FALSE SMILES
were deliberate smiles made on request (Ekman, Hager, & Friesen,
1981; Hager & Ekman, 1997; Skinner & Mullen, 1991). Schmidt
et al. (2006, 2009), however, did not find the proposed difference
in symmetry, and other studies have failed to show an effect of
symmetry on perceivers’ ratings of happiness and amusement
(Ambadar et al., 2009; Gosselin, Perron, et al., 2002).
It has also been argued that spontaneous smiles are smoother
than their deliberate counterparts. Spontaneous smiles have been
found to be smoother and to contain fewer irregularities (pauses
and stepped intensity changes) than deliberate smiles (Hess &
Kleck, 1990; Weiss, Blum, & Gleberman, 1987). Smoothness
versus irregularity also has been shown to influence observers’
judgments of the degree of control of the expression (Hess &
Kleck, 1994). Furthermore, temporal parameters such as the relative durations of onset, apex, and offset phases of a smile have
been proposed to distinguish between spontaneous and deliberate
smiles. Some studies have found longer onset and offset times for
spontaneous smiles than for deliberate ones (Hess & Kleck, 1990;
Schmidt et al., 2006, 2009). Effects of onset, apex, and offset
durations have been demonstrated for ratings of smile genuineness
(Krumhuber & Kappas, 2005; Krumhuber, Manstead, Cosker,
Marshall, & Rosin, 2009; Krumhuber, Manstead, & Kappas, 2007;
Krumhuber, Manstead, Cosker, et al., 2007), but not for accuracy
of expression discrimination or judgments of control when only
onset speed was considered (Hess & Kleck, 1994).
Of all the attributes of felt, genuine smiles, the D marker has
been proposed as the most reliable, robust, and diagnostic marker.
As Frank and Ekman (1993) stated, “All the other markers (i.e.,
symmetry, smoothness, duration) should be considered necessary,
but not necessarily sufficient indicators of an enjoyment smile” (p.
21). It is interesting that the signal value of the D smile rarely has
been studied alongside these other physical markers. In some
cases, only one feature (i.e., overall duration or symmetry or
nonpositive facial actions) was measured or manipulated in the
presence of the D smile, yielding evidence of the dominant role of
the D marker (Frank et al., 1993; Gosselin, Beaupré, & Boissonneault, 2002; Gosselin, Perron, et al., 2002). Hess and Kleck
(1994) explored a combination of cues, but their results pertained
more globally to spontaneous expressions, combining positive and
negative emotions. In a recent study by Ambadar et al. (2009), the
signal value of the D marker was tested in the context of other smile
characteristics such as asymmetry of amplitude, onset and offset
velocity, and nonpositive facial actions (called smile controls). Although AU 6 occurred more often during smiles perceived to reflect
amusement than during polite or embarrassed smiles, the presence of
the D marker on its own was not sufficient to signify amusement.
How much the D smile, as opposed to other physical markers, affects
perceptions of positive emotion therefore remains unclear.
Present Research
The objective of the present research was to investigate the
reliability and validity of the D smile. As we have seen, there is a
widespread assumption that D smiles are spontaneous expressions
that reflect felt positive emotion. We aimed to test this assumption
systematically by investigating both spontaneous and deliberate D
and ND smiles. In a first experiment, we focused on the expression
of these different types of smile. Here, we studied whether and
how often people activated the orbicularis oculi, pars lateralis
809
muscle (AU 6) when smiling voluntarily. We also examined
whether D smiles occurred solely or predominantly under spontaneous conditions, and whether ND smiles occurred solely or predominantly under deliberate conditions. If the presence of AU 6 in
D smiles is a reliable sign of spontaneous enjoyment, it should not
occur in false or voluntary smiles. Moreover, senders should report
subjective experience of positive emotion when exhibiting D
smiles, but not when exhibiting ND smiles.
In subsequent studies, we examined how D and ND smiles were
perceived, and whether people made different judgments of these
two smile types when they were made under spontaneous and
deliberate conditions. To date, no study has explored the perception of both spontaneous and deliberate D smiles. In a series of
experiments, we investigated perceivers’ ratings of these smiles
with respect to spontaneity, genuineness, and amusement. We also
examined whether judges differentiated between more and less
intense D smiles made in the spontaneous condition (Experiment
2), and whether ratings of smiles were affected by having access to
the upper versus the lower face (Experiment 3) or when seeing
static versus dynamic displays (Experiment 4). As well as this
focus on how smile types affected judgments, we tested the predictive value of the D marker alongside other behavioral features,
namely asymmetry; irregularity; onset, apex, and offset durations;
and nonpositive facial actions. If the D marker is the most diagnostic feature of an enjoyment smile, it should play a major role in
predicting how genuine and amused the smile is judged to be. We
therefore expected its utility in predicting the ratings of smiles in
Experiments 2– 4 to be consistent and at least as great as (if not
greater than) the predictive value of the other physical attributes.
Experiment 1
The purpose of this study was to test the notion that D smiles
signal spontaneous felt emotions, with AU 6 being a reliable
indicator of enjoyment (Ekman, 1993; Ekman & Friesen, 1982).
To this end, we compared the amount of D and ND smiles
exhibited under spontaneous and deliberate conditions and examined the levels of positive emotions that were reported by individuals showing these smile expressions.
Method
Participants
Thirty-two undergraduate students (16 men, 16 women), ages
18 to 35 years (M ⫽ 22.59 years, SD ⫽ 4.01) at Cardiff University
participated on a voluntary basis and were paid £3.
Materials and Procedure
In the spontaneous condition, participants saw six amusing
stimuli intended to induce positive emotional feelings. The stimuli
consisted of two jokes, one cartoon, and three film clips, each of
which was approximately 15 s in length. They were selected from
a larger pool of 34 stimuli that had been pretested on 19 participants (9 women, 10 men, ages 20 –26 years, M ⫽ 22.65 years,
SD ⫽ 2.06) for their capacity to evoke amusement (M ⫽ 5.37),
810
KRUMHUBER AND MANSTEAD
pleasure (M ⫽ 4.54), and happiness (M ⫽ 4.45) on a 7-point rating
scale (1 ⫽ not at all, 7 ⫽ extremely).1 Apart from one joke that had
been used in a previous study, all stimuli were chosen from the
Internet (e.g., http://www.allfunnypictures.com). The amusing
stimuli were separated by filler items, that is, pictures of inanimate
objects drawn from the International Affective Picture System (IAPS;
Lang, Bradley, & Cuthbert, 2001) that had low arousal and neutral
valence scores (M ⫽ 2.49 and M ⫽ 4.89, respectively, on a 9-point
scale). These filler stimuli were presented for 8 s each and were
intended to return participants to a neutral emotional state.
In the deliberate condition, participants saw a different set of six
IAPS slides of inanimate objects that had low arousal (M ⫽ 2.47)
and neutral valence scores (M ⫽ 5.01) and were intended to serve
as neutral stimuli. Each neutral stimulus was presented for 10 s.
Following Hess and Kleck (1990), participants were instructed to
pose in front of the video monitor an expression that made it look
as if they were watching something funny and in a way that would
be as convincing as possible. More specifically, they were asked
to smile as they would if they felt amused. They were not allowed
to talk and had to rely simply on positive facial expressions. To
motivate participants, an incentive of £10 was promised if their
smile expressions proved to be the most convincing of those made
by all participants (see Ekman et al., 1988; Schmidt et al., 2003, for
a similar procedure). The experimenter was not present during
either condition.
Participants always completed the spontaneous procedure first.
The rationale for this is the same as the one offered by Ekman et
al. (1988, 1990), namely that the reverse order would be likely to
result in interference effects between the two conditions. Within
each condition, the stimuli were presented in one of two different
sequences using MediaLab (Empirisoft) software. Participants
were asked to consent to being video recorded at the start of the
session, and their facial behavior was recorded during the whole
procedure. We also asked for participants’ permission to use the
video recordings of their behavior in future research projects.
Apart from three people who objected to the future use of their
recordings, all participants agreed to both requests.
For all samples in the spontaneous and deliberate epochs, we
scored both the frequency of AU 12 and AU 6 and the maximum
intensity (1 ⫽ trace to 5 ⫽ maximum) that AU 12 reached during
each episode. Two FACS-trained coders who were blind to the
different stimulus sequences within condition individually coded
all the facial activity shown by the 32 participants. In addition, a
FACS-certified coder scored 50% of the samples (i.e., 192 smiling
episodes). Intercoder reliability was assessed as mean percentage
agreement (%) and chance-corrected Cohen’s kappa (␬). Intensity
differences of 1 point were treated as agreements (for a similar
approach, see Messinger et al., 2008). Mean agreement was
98.44% (␬ ⫽ 0.94) for AU 12, 89.06% (␬ ⫽ 0.79) for AU 6, and
96.35% (␬ ⫽ 0.95) for AU 12 intensity, demonstrating high
reliability.
Subjective emotion. Immediately following each stimulus presentation, participants were asked to report the intensity with
which they had felt each of the following emotions while viewing
the stimulus just seen: amusement, pleasure, happiness, surprise,
and interest. Ratings were made on 7-point scales, with response
options ranging from 1 (not at all) to 7 (extremely). These questions were posed in a randomly selected order that varied with each
participant.
Dependent Variables
Facial Activity
Facial action units. For each participant, the first smile that
occurred in each of the six spontaneous and six deliberate epochs
was selected for analysis to ensure a balanced representation of
smiles (for a similar approach, see Cohn & Schmidt, 2004;
Schmidt et al., 2003, 2006). This resulted in a set of 384 smiling
episodes. Facial activity during these episodes was scored using
the FACS (Ekman & Friesen, 1978; Ekman et al., 2002). FACS
enables the measurement of all visible facial behavior and describes it in terms of 44 action units. For the purposes of the
present research, two of these action units, AU 12 and AU 6, were
treated as the primary dependent variables (for a similar procedure,
see Papa & Bonanno, 2008; Schmidt et al., 2006). AU 12 refers to
activity of the zygomaticus muscle and involves a lip corner pull
(smiling mouth). AU 6 refers to activity of the orbicularis oculi,
pars lateralis muscle that raises the cheeks, gathers the skin around
the eye, and produces crow’s feet wrinkles. Smile episodes that
included both AU 12 and AU 6 were defined as D smiles, whereas
episodes that included only AU 12 were defined as ND smiles (cf.
Frank et al., 1993).
ANOVA revealed a significant main effect of smile type on
frequency of smiling, F(1, 30) ⫽ 33.92, p ⬍ .001,
␩2p ⫽ .53. Overall, participants displayed more D (76.5%) than ND
smiles (23.3%). The condition main effect was not significant, F(1,
30) ⫽ 2.14, p ⬎ .05, ␩2p ⫽ .07, nor was the interaction between
condition and smile type, F(1, 30) ⫽ 3.81, p ⬎ .05, ␩2p ⫽ .11.
Overall, 70% of smiles in the spontaneous condition were D
smiles, and the remaining 30% were ND smiles. By comparison,
83% of smiles in the deliberate condition were D smiles, and the
remaining 17% were ND smiles. Thus, D and ND smiles appeared
in approximately equal proportions in the spontaneous and deliberate conditions. This finding is consistent with evidence (Schmidt
& Cohn, 2001; Schmidt et al., 2006, 2009; Smith et al., 1996) that
Results and Discussion
Preliminary analyses showed that neither stimulus type (jokes,
cartoon, film clips, pictures) nor the sequence in which the stimuli
were presented had an effect on any of the dependent variables
( ps ⬎.05). These factors were dropped from further analyses. For
each participant, frequency or average scores (intensity, emotion)
for D and ND smiles were computed in the spontaneous and
deliberate conditions. These four scores were entered into mixed
analyses of variance (ANOVAs) with two within-subjects factors,
condition (spontaneous, deliberate) and smile type (D, ND), and
one between-subjects factor, sex of participant. Separate analyses
were carried out on the measures of facial activity (frequency,
intensity) and the measures of participants’ subjective emotion.
1
Arousal scores for the six amusing stimuli used in the spontaneous
condition were below the midpoint (M ⫽ 2.82, SD ⫽ 0.84, 7-point scale)
and similar to those of the stimuli used in the deliberate condition, as
indicated by an independent sample of 10 participants (5 men, 5 women),
ages 25 to 32 years (M ⫽ 28.9 years, SD ⫽ 2.42).
FELT AND FALSE SMILES
people are able to pose a D smile deliberately, and that they do so
quite frequently. Thus, there is no direct evidence that AU 6 in D
smiles reflected spontaneity, whereas its absence in ND smiles
reflected deliberateness (Parkinson, 2005). The sex of participant
main effect was not significant, F(1, 30) ⫽ 0.43, p ⬎ .05,
␩2p ⫽ .01.
With respect to smile intensity, D smiles had higher intensity
levels (M ⫽ 3.11) than ND smiles (M ⫽ 0.97), F(1, 30) ⫽ 106.67,
p ⬍ .001, ␩2p ⫽ .78, an effect observed by Ekman and Friesen
(1978) and also found in a recent study by Messinger et al. (2008).
The main effects of condition, F(1, 30) ⫽ 0.01, p ⬎ .05, ␩2p ⫽ .00,
and sex of participant, F(1, 30) ⫽ 1.37, p ⬎ .05, ␩2p ⫽ .04, were
not significant. However, there was a significant interaction between condition and smile type, F(1, 30) ⫽ 8.52, p ⬍ .01, ␩2p ⫽
.22. Analysis of simple effects (see Winer, Brown, & Michels,
1991) showed that when participants were instructed to pose
smiles, D smiles were more intense (M ⫽ 3.37) than were D smiles
observed in the spontaneous condition (M ⫽ 2.85), F(1, 30) ⫽
5.67, p ⬍ .05, ␩2p ⫽ .16. Thus, participants were able to simulate
the D smile by putting on a strong smile that also activated the
upper facial muscles. The corresponding difference for ND smiles
was not significant (spontaneous: M ⫽ 1.24, deliberate: M ⫽
0.71), F(1, 30) ⫽ 3.75, p ⬎ .05, ␩2p ⫽ .11.
Subjective Emotion
A factor analysis (principal components with varimax rotation)
of responses to the five emotion items showed that amusement,
pleasure, happiness, surprise, and interest all loaded on the same
factor (explaining 81.83% of the variance). They were therefore
combined to form a positive emotionality scale (Cronbach’s alpha ⫽ .94). ANOVA showed significant main effects of condition,
F(1, 30) ⫽ 53.78, p ⬍ .001, ␩2p ⫽ .64, and smile type, F(1, 30) ⫽
21.67, p ⬍ .001, ␩2p ⫽ .42. Smiles shown in the spontaneous
condition were accompanied by significantly more positive emotion (M ⫽ 3.03) than were those shown in the deliberate condition
(M ⫽ 1.25), confirming that the elicitation procedures gave rise to
substantially different levels of positive affect. Participants also
reported having experienced significantly more positive emotion
during epochs in which they displayed D smiles (M ⫽ 2.84) than
in epochs in which they displayed ND smiles (M ⫽ 1.43). Thus,
smile type was also related to positive emotion. Consistent with the
notion that the D smile reflects true felt enjoyment (Ekman et al.,
1988, 1990; Frank et al., 1993), spontaneous D smiles were accompanied by more positive emotion (M ⫽ 3.74) than were
deliberate D smiles (M ⫽ 1.94), F(1, 30) ⫽ 29.55, p ⬍ .001, ␩2p ⫽
.50. Similarly, spontaneous ND smiles were associated with more
positive emotion (M ⫽ 2.31) than were deliberate ND smiles (M ⫽
0.55), F(1, 30) ⫽ 22.89, p ⬍ .001, ␩2p ⫽ .43. The interaction
between condition and smile type was not significant, F(1, 30) ⫽
0.01, p ⬎ .05, ␩2p ⫽ .00, nor was the main effect of sex of
participant, F(1, 30) ⫽ 3.79, p ⬎ .05, ␩2p ⫽ .11. Additional
correlation analyses showed a significant positive relationship
between degree of positive emotion and smile intensity, r(32) ⫽
.54, p ⬍ .01.
Summarizing, neither D smiles nor ND smiles occurred solely
or predominantly under spontaneous or deliberate conditions.
Rather, whether or not a D smile was shown was associated with
the intensity of the expression and with the amount of positive
811
emotion experienced. In a series of judgment studies reported
below, we focused on the perception of smile expressions. Specifically, we investigated how D and ND smiles made under
spontaneous and deliberate conditions were perceived by naive
observers. Furthermore, we examined the extent to which the D
marker, along with other attributes (e.g., asymmetry; irregularity;
onset, apex, and offset durations; and nonpositive facial actions)
predicted the ratings made by observers.
Experiment 2
Our aim in this study was to examine whether observers would
distinguish in their judgments between D and ND smiles that were
made under spontaneous and deliberate conditions. For example,
would D smiles originating from a spontaneous condition be rated
differently from those arising from a deliberate condition? If there
is only type of enjoyment smile, spontaneous D smiles should be
seen as more genuine than deliberate D smiles or spontaneous ND
smiles. An additional goal was to study the effects of expression
intensity on judgments of D smiles. Much research has shown that
the intensity of an expression significantly influences ratings of
positive emotion (Bolzani Dinehart et al., 2005; Hess, Blairy, &
Kleck, 1997; Messinger, 2002; Messinger et al., 2008). We therefore wanted to test whether judges would distinguish between
spontaneous D smiles that differed in intensity. If the expression
intensity of the sender plays a role in the evaluation of D smiles,
more intense D smiles should be seen as reflecting greater subjective enjoyment than less intense D smiles.
Method
Participants
Fifty-two undergraduate students (22 men, 30 women), ages
18 –33 years (M ⫽ 20.17 years, SD ⫽ 2.25) at Cardiff University
took part in this study and were given either course credit or
payment of £2.
Stimulus Material and Design
It has been argued that the “true” spontaneous D smile should be
discernible mainly at a low or moderate intensity level, when the
smiling action of the mouth does not automatically lead to a
contraction of the upper face muscles, thereby giving the overall
appearance of a D smile (Ekman, 1985; Frank & Ekman, 1993).
The D and ND smiles selected from Experiment 1 were therefore
not of extreme intensity in FACS terms, but rather of slight or
moderate/pronounced intensity. We chose smiles that began and
ended with a neutral baseline expression and for which senders
reported having felt moderate or high emotions (i.e., pleasure,
amusement, and happiness ratings of 3 or higher on a 7-point scale
where 1 ⫽ not at all and 7 ⫽ extremely) in the spontaneous
condition and low or no emotions (i.e., pleasure, amusement and
happiness ratings of 2 and lower) in the deliberate condition (for a
similar procedure, see Hess & Kleck, 1994). From the pool of
smile expressions generated in Experiment 1, there were 30 smiles
of each type that met these criteria: (a) 6 spontaneous D smiles
(pronounced): Mintensity ⫽ 3.7, Mpositive emotion ⫽ 5.2; (b) 6 spontaneous D smiles (moderate): Mintensity ⫽ 3.0, Mpositive emotion ⫽
3.8; (c) 6 spontaneous ND smiles: Mintensity ⫽ 2.2, Mpositive emo-
812
KRUMHUBER AND MANSTEAD
tion ⫽ 3.9; (d) 6 deliberate D smiles: Mintensity ⫽ 3.5, Mpositive emotion ⫽ 1.6;
and (e) 6 deliberate ND smiles: Mintensity ⫽ 2.0, Mpositive emotion ⫽ 1.7.
Overall, the intensity of spontaneous D smiles was similar to that of
deliberate D smiles (M ⫽ 3.4 vs. M ⫽ 3.5). The same was the case for ND
smiles, which did not differ with respect to intensity between the spontaneous
and deliberate conditions (M ⫽ 2.2 vs. M ⫽ 2.0). In this way, we sought to
eliminate expression intensity as a possible explanation for differential responses to smiles in the spontaneous and deliberate conditions (for a similar
approach, see Miles & Johnston, 2007; Peace et al., 2006; Soussignan, 2002)
and rather study the effects of expression intensity in the perception of
spontaneous D smiles (pronounced vs. moderate).2
In addition to the D marker, all 30 smiles were scored by two
FACS-certified coders who were blind to condition (i.e., spontaneous vs. deliberate) for the following behavioral features: Asymmetry (mean agreement ⫽ 93.34%; ␬ ⫽ 0.86) was coded if the
smiling action was stronger on one side of the face (Ekman et al.,
1981). This was the case for deliberate ND smiles, which were
more asymmetric (83.3%) than both spontaneous D smiles (pronounced) and spontaneous ND smiles (0.0% and 16.7%, p ⬍ .01
and p ⬍ .05), F(4, 20) ⫽ 3.49, p ⬍ .05, ␩2p ⫽ .41.
Irregularity (mean agreement ⫽ 90%, ␬ ⫽ 0.75) was coded if
the smile expression faded and reintensified or was characterized
by discontinuities in the onset, apex, or offset phases (Hess &
Kleck, 1990). No significant differences in irregularity occurred
between the five smile types, F(4, 20) ⫽ 0.54, p ⬎ .05, ␩2p ⫽ .10.
Nonpositive facial actions (mean agreement ⫽ 93.34%, ␬ ⫽ 0.86)
were coded if frowning (AU 4), lip pressing (AU 24), dimpling (AU
14), lip corner depressing (AU 15), lid tightening (AU 7), or lower lip
depressing (AU 16) occurred during the smile. Such smile controls or
dampening movements (Ambadar et al., 2009; Schmidt et al., 2009)
previously have been associated with nonjoyful feelings (Ekman,
1985; Ekman et al., 2002). There were no significant differences in
the presence of nonpositive facial actions between the five smile
types, F(4, 20) ⫽ 0.73, p ⬎ .05, ␩2p ⫽ .13.
Onset time, apex time, and offset time were coded by a FACStrained coder and a FACS-certified coder within a 0.5-s tolerance
window (see Sayette, Cohn, Wertz, Perrott, & Parrott, 2001) and
yielded high correlations between the two coders (intraclass r ⫽ .79).
Onset duration was measured as the number of frames from the start
of the smile expression to its zenith. Apex duration was the number of
frames the smile expression was held at its peak, and offset duration
was measured as the number of frames from first evidence of decay
of the smile expression until it stopped decaying (Ekman et al., 2002;
Hess & Kleck, 1990). There were no significant differences between
the five smile types in onset duration, F(4, 20) ⫽ 1.83, p ⬎ .05, ␩2p ⫽
.27, or offset duration, F(4, 20) ⫽ 1.02, p ⬎ .05, ␩2p ⫽ .17. Onset
durations were on average 25.65 frames (1.03 s), whereas offset
durations were on average 49.80 frames (1.99 s). However, spontaneous ND smiles were shorter in their apex durations (M ⫽ 12 frames,
0.48 s) than the other smile types (M ⫽ 90.79 frames, 3.63 s), F(4,
20) ⫽ 12.04, p ⬍ .001, ␩2p ⫽ .71.
Correlation analyses of all smile features showed a significant
relationship only between temporal features of smiles. The longer the
onset, the longer the smile was held at its apex (r ⫽ .37, p ⬍ .05); the
longer the apex duration, the longer the offset (r ⫽ .42, p ⬍ .05).
Roughly the same number of smiles came from male and female
encoders (14:16). A given encoder could be shown more than once
with a different smile expression. On average, there were two smile
expressions per encoder that could be any of the types described
above. In this way, we tried to avoid presenting judges with a standard
baseline that they could rely on to rate the genuineness of an expression. The 30 smile expressions featuring five different types of smiles
were displayed as movie clips (750 ⫻ 576 pixels, 25 frames/s) in
MediaLab (Empirisoft) and shown in random order.
Procedure
After signing a consent form, participants were told that they
would see short video clips of several people. They were informed
that each person had been shown some amusing stimuli (e.g.,
funny pictures, movies, and texts) that were intended to elicit
positive emotional feelings and some neutral stimuli (i.e., pictures
of inanimate objects) that were intended to elicit no specific
emotion. They were further told that the amusing stimuli resulted
in smiles that were spontaneous and accompanied by genuine
amusement of the person, and that when presented with neutral
stimuli, the person viewing them had been asked to pose a smile as
if s/he was amused. We instructed participants to look carefully at
each smile and then to make judgments about it, including whether
it was spontaneous or posed. The video sequences were initiated
by using the mouse to click a Start button on the computer screen.
Dependent Variables
After each sequence, participants were asked to answer the
following questions on 7-point scales: (a) To what extent is the
smile you have just seen spontaneous or posed? (1 ⫽ deliberate,
7 ⫽ spontaneous), (b) to what extent is the smile you have just
seen genuine or fake? (1 ⫽ fake, 7 ⫽ genuine), and (c) to what
extent was the person you have just seen amused at the time of the
smile? (1 ⫽ not at all, 7 ⫽ very).
Results and Discussion
Analysis of Variance
A multivariate analysis of variance (MANOVA) with sex of
encoder and smile (five levels: pronounced spontaneous D, moderate spontaneous D, spontaneous ND, deliberate D, deliberate
ND) as within-subjects factors and sex of participant as a betweensubjects factor was conducted on the three dependent variables
(spontaneous, genuine, and amused).3 There were no significant
effects associated with sex of encoder, F(3, 48) ⫽ 0.28, p ⬎ .05,
␩2p ⫽ .02, and this factor was dropped in further analyses. For all
2
Examples of the different types of smiles can be obtained from Eva G.
Krumhuber on request.
3
Ratings of how spontaneous and genuine the smile appeared to be were
consistent and highly correlated with each other (rs ⬎ .86, ps ⬍ .001), and
analyses were virtually the same across all three experiments. Because
smile genuineness was theoretically the more important variable, we decided to focus on this variable and report results for smile genuineness
only, not considering perceived spontaneity. The same applied to the
regression analyses in which the results were identical for measures of
spontaneity and genuineness. Ratings of spontaneity and genuineness were
associated with those of amusement (rs ⬎ .46, p ⬍ .001). However,
correlation coefficients were lower than those between spontaneity and
genuineness, suggesting that these two measures tap related but different
aspects of smile perception.
FELT AND FALSE SMILES
univariate analyses (spontaneous, genuine, amused), we applied a
Greenhouse–Geisser adjustment to degrees of freedom. The sex of
participant main effect was not significant, F(3, 48) ⫽ 0.37, p ⬎ .05,
␩2p ⫽ .02. As expected, there was a main effect for smile, F(12, 39) ⫽
32.16, p ⬍ .001, ␩2p ⫽ .91. It was significant for all three variables:
spontaneous, F(2.87, 143.74) ⫽ 27.08, p ⬍ .001, ␩2p ⫽ .35; genuine,
F(3.09, 154.40) ⫽ 27.93, p ⬍ .001, ␩2p ⫽ .36; and amused, F(3.17,
158.46) ⫽ 135.35, p ⬍ .001, ␩2p ⫽ .73.
Post hoc comparisons showed that spontaneous D smiles were
generally rated as more genuine (Mdiff ⬎ 0.94, SEdiff ⬍ 0.18, ps ⬍
.001) and amused (Mdiff ⬎ 1.28, SEdiff ⬍ 0.12, ps ⬍ .001) than ND
smiles, in accordance with previous findings (Frank et al., 1993;
Gosselin, Beaupré, & Boissonneault, 2002; Gosselin, Perron, et al.,
2000; Hess & Kleck, 1994; Messinger et al., 2008). It is interesting
that participants also distinguished between more intense and less
intense D smiles within the spontaneous condition. More intense D
smiles (i.e., pronounced) were judged as more genuine (Mdiff ⫽
0.42, SEdiff ⫽ 0.12, p ⬍ .05) and more amused (Mdiff ⫽ 0.82,
SEdiff ⫽ 0.08, p ⬍ .001) than were D smiles of lower intensity (i.e.,
moderate; see Table 1, Experiment 2). Extending previous research (Bolzani Dinehart et al., 2005; Hess et al., 1997; Messinger
et al., 2008), these findings show that expression intensity leads to
differences in how spontaneous D smiles are perceived. The higher
the intensity of the expression, the more genuine and amused the
associated spontaneous D smile was perceived to be.
A different pattern of results was found for deliberate smiles.
Here, participants did not differ in their judgments of the genuineness of D and ND smiles (Mdiff ⫽ 0.19, SEdiff ⫽ 0.18, p ⬎ .05),
showing that they perceived these two types of smile as equally
insincere. They nevertheless made higher amusement ratings of D
than of ND smiles (Mdiff ⫽ 1.05, SEdiff ⫽ 0.12, p ⬍ .001), thereby
providing some support for the notion that D smiles convey greater
positive emotionality.
Post hoc comparisons showed that there was a significant difference in judgments of D smiles made under spontaneous and
deliberate conditions. Spontaneous D smiles attracted significantly
higher genuineness (Mdiff ⬎ 0.70, SEdiff ⬍ 0.16, ps ⬍ .001) and
amusement (Mdiff ⬎ 0.88, SEdiff ⬍ 0.09, p ⬍ .001) ratings than
813
deliberate D smiles. This is consistent with the assumption that
there is one type of genuine, felt smile: the spontaneous D smile.
No such difference emerged for ND smiles (genuine: Mdiff ⫽ 0.05,
SEdiff ⫽ 0.14, p ⬎ .05; amused: Mdiff ⫽ ⫺0.17, SEdiff ⫽ 0.09, p ⬎
.05), showing that elicitation condition had a significant impact
only on judgments of D smiles.
Regression Analyses
Multiple regression analyses were used to examine the degree to
which each of the facial features (D marker, asymmetry, irregularity, onset time, apex time, offset time, nonpositive facial actions) predicted ratings of spontaneity, genuineness, and amusement. Table 2 (Experiment 2) shows the beta coefficients
associated with these seven predictors. Apex duration ( ps ⱕ .05),
nonpositive facial actions ( ps ⬍ .05), and asymmetry ( ps ⬍ .05)
were significant predictors of how genuine and amused the smile
was perceived to be. The longer the apex duration and the fewer
nonpositive facial actions, the more genuine and amused a smile
was judged to be. Less asymmetry was also associated with higher
ratings of smile genuineness. Thus, symmetrical smiles were associated with higher ratings of genuineness of the expression. The
presence or absence of the D marker did not significantly predict
any of the ratings, nor did irregularity, onset time, or offset time
( ps ⬎ .05).
Experiment 3
Our aim in Experiment 3 was to investigate whether judgments
of smiles would be affected by having access to the upper or lower
face. Research on the relative importance of the upper and lower
face has shown either the mouth region (Bassili, 1979; Kerstenbaum, 1992) or both the mouth and the cheeks (Boucher & Ekman,
1975; Kohler et al., 2004; Nusseck, Cunningham, Wallraven, &
Bülthoff, 2008) to be crucial to the recognition of happiness.
However, these studies did not specifically examine D smiles.
Williams, Senior, David, Loughland, and Gordon (2001) showed
more frequent and longer eye fixations to the crow’s feet area in
Table 1
Means and Standard Errors for Ratings of Genuineness and Amusement as a Function of Condition and Smile Type in Experiments
2, 3, and 4
Spontaneous
Measure
Experiment 2 (n ⫽ 53)
Genuine
Amused
Experiment 3 (n ⫽ 50)
Genuine
Amused
Experiment 4 (n ⫽ 60)
Genuine
Amused
Deliberate
Duchenne (b)
Duchenne (a)
Non-Duchenne
Duchenne
Non-Duchenne
M
SE
M
SE
M
SE
M
SE
M
SE
4.42a
3.84a
0.12
0.10
4.83b
4.66b
0.12
0.09
3.47c
2.56c
0.14
0.08
3.71c
3.77a
0.12
0.09
3.52c
2.73c
0.11
0.08
—
—
—
—
4.82a
4.68a
0.11
0.10
3.41b
2.54b
0.13
0.09
4.08c
3.85c
0.10
0.09
3.36b
2.74d
0.12
0.10
—
—
—
—
4.61a
4.56a
0.11
0.09
3.37b
2.66b
0.12
0.09
3.91c
3.91c
0.09
0.08
3.57d
2.86d
0.10
0.09
Note. Duchenne (a) ⫽ pronounced, more intense smiles; Duchenne (b) ⫽ moderate, less intense smiles. All ratings were made on 1-to-7 Likert scales,
with higher scores indicating greater levels of that dimension. Row means with different subscripts differ at p ⱕ .05 or better.
KRUMHUBER AND MANSTEAD
814
Table 2
Multiple Regression Analyses of Ratings of Genuineness and Amusement of Smiles in Experiments 2, 3, and 4 on Seven Facial
Features (Values Are Standardized Regression Coefficients)
Experiment
Condition/Rating
2
Genuine
Amused
Upper face
Genuine
Amused
Lower face
Genuine
Amused
Dynamic
Genuine
Amused
Static
Genuine
Amused
3
4
ⴱ
p ⬍ .05.
ⴱⴱ
Duchenne
marker
Asymmetry
Irregularity
.17
.29
⫺.39ⴱ
⫺.21
.53ⴱ
.56ⴱ
.19
.37
⫺.14
.34
.31
.61ⴱ
Onset
time
Apex
time
Offset
time
Nonpositive
facial action
⫺.08
⫺.12
.10
.13
.42ⴱ
.44ⴱⴱ
⫺.02
⫺.08
⫺.32ⴱ
⫺.42ⴱⴱ
⫺.21
⫺.20
⫺.18
⫺.11
⫺.07
.02
.19
.22
⫺.12
⫺.06
⫺.23
⫺.24
⫺.40ⴱ
⫺.25
⫺.00
⫺.06
.02
.03
.56ⴱ
.46ⴱ
.10
.08
⫺.11
⫺.23
⫺.55ⴱⴱ
⫺.19
⫺.02
⫺.08
.16
.10
.62ⴱ
.41ⴱ
.03
⫺.04
⫺.29
⫺.37ⴱ
.02
.03
⫺.15
⫺.09
.01
.00
.34
.21
⫺.01
⫺.04
⫺.46ⴱ
⫺.36ⴱ
p ⬍ .01.
the upper face of D smiles, but their findings pertain to comparisons with nonexpressive faces rather than with ND smiles. Tremblay et al. (1993) used image manipulation techniques to vary the
presence or absence of the D marker in still photographs of smiling
or neutral faces. Although crow’s feet (a distinctive feature of D
smiles) influenced perceptions of smile genuineness, this was only
in the case for medium intensity smiles. In general, the intensity of
mouth region activity was most indicative of a genuine smile
expression.
We examined the relative importance of the upper and lower
part of the face in dynamic D and ND smiles made under spontaneous and deliberate conditions. Participants were presented with
partial faces that showed either the upper or the lower face. Given
that the D marker is a feature of the upper face, it should play a
significant role in the distinction between D and ND smiles only in
the upper face. Differences in ratings of genuineness and amusement of D and ND smiles should therefore be absent or at least
attenuated when judges have access solely to the lower face.
Method
Participants
Fifty-one undergraduate students (22 men, 29 women), ages 18
to 34 years (M ⫽ 20.57 years, SD ⫽ 2.84) at Cardiff University
took part individually either for course credit or for a payment
of £3.
Stimulus Material and Design
The smile stimuli were the same as those used in Experiment 2,
except that only spontaneous D smiles of higher intensity (pronounced) were included. This resulted in a design with four smile
types: (a) 6 spontaneous D smiles, (b) 6 spontaneous ND smiles,
(c) 6 deliberate D smiles, and (d) 6 deliberate ND smiles. Smile
type (D, ND) and condition (spontaneous, deliberate) were treated
as two independent factors. Two partial face versions were created
for each of the 24 expressions, showing either the lower or upper
face with the other half covered by a black mask. Lower face
versions showed the mouth area only, including the nasolabial
furrows. Upper face versions showed the eye area only, including
the cheek bulges. The resulting 48 facial stimuli were displayed as
movie clips in MediaLab (Empirisoft) and shown in random order.
Procedure and Dependent Variables
These were the same as in Experiment 2, except that participants
were told that for each stimulus person only the lower or upper part
of the face would be visible. The dependent variables were identical to those used in Experiment 2.
Results and Discussion
Analysis of Variance
The three dependent measures were entered into a MANOVA,
with face (lower, upper), condition (spontaneous, deliberate), and
smile type (D, ND) as within-subjects factors, and sex of participant (male, female) as a between-subjects factor. The main effect
of sex of participant was not significant, F(3, 47) ⫽ 2.60, p ⬎ .05,
␩2p ⫽ .14. A significant main effect was found for face, F(3, 47) ⫽
10.65, p ⬍ .001, ␩2p ⫽ .40. This was significant only for ratings of
amusement, F(1, 49) ⫽ 8.31, p ⬍ .01, ␩2p ⫽ .14. Smiles shown in
the lower face attracted significantly higher amusement ratings
(M ⫽ 3.53) than those shown in the upper face (M ⫽ 3.38). This
is consistent with the notion that the lower face is more influential
in shaping attributions of happiness and amusement (Bassili, 1979;
Kerstenbaum, 1992; Tremblay et al., 1993). An upward turning
mouth therefore appears to be the strongest determinant of how
much amusement smiles are seen as conveying.
Significant main effects also emerged for condition, F(3, 47) ⫽
11.49, p ⬍ .001, ␩2p ⫽ .42, and smile type, F(3, 47) ⫽ 105.66, p ⬍
.001, ␩2p ⫽ .87. These two main effects were qualified by a
significant interaction between condition and smile type, F(3,
47) ⫽ 31.64, p ⬍ .001, ␩2p ⫽ .67. The interaction was significant
for all three dependent variables: spontaneous, F(1, 49) ⫽ 23.35,
FELT AND FALSE SMILES
p ⬍ .001, ␩2p ⫽ .32; genuine, F(1, 49) ⫽ 27.40, p ⬍ .001, ␩2p ⫽
.36; and amused, F(1, 49) ⫽ 89.88, p ⬍ .001, ␩2p ⫽ .65. Analyses
of simple effects showed that D smiles made under both spontaneous and deliberate conditions were rated as more genuine—
spontaneous, F(1, 49) ⫽ 52.46, p ⬍ .001, ␩2p ⫽ .52; deliberate,
F(1, 49) ⫽ 20.24, p ⬍ .001, ␩ p2 ⫽ .29 —and amused—
spontaneous, F(1, 49) ⫽ 200.38, p ⬍ .001, ␩2p ⫽ .80; deliberate,
F(1, 49) ⫽ 95.78, p ⬍ .001, ␩2p ⫽ .66 —than were ND smiles.
Thus, participants were sensitive to the presence or absence of the D
marker within each condition, although such differences in the ratings
of D and ND smiles were greater in the spontaneous condition than in
the deliberate condition (see Table 1, Experiment 3).
As in Experiment 2, judgments of D smiles varied significantly
between the spontaneous and deliberate conditions. In particular,
spontaneous D smiles attracted higher ratings of genuineness, F(1,
49) ⫽ 43.19, p ⬍ .001, ␩2p ⫽ .47, and amusement, F(1, 49) ⫽
85.56, p ⬍ .001, ␩2p ⫽ .64, than did deliberate D smiles. Replicating the findings of Experiment 2, this supports the notion that
there is just one type of genuine felt smile, namely the spontaneous
D smile.
Elicitation condition did not influence the perceived genuineness of ND smiles, F(1, 49) ⫽ 0.35, p ⬎ .05, ␩2p ⫽ .01. However,
participants judged deliberate ND smiles to be more amused than
spontaneous ones, F(1, 49) ⫽ 10.67, p ⬍ .01, ␩2p ⫽ .18. Whereas
D smiles are perceived as reflecting greater amusement when they
are shown spontaneously rather than deliberately, the opposite is
true for ND smiles.
Unexpectedly, these effects were not qualified by the face
variable. The interaction between face and smile type was not
significant, F(3, 47) ⫽ 1.83, p ⬎ .05, ␩2p ⫽ .10. Thus, the
differences between ratings of D and ND smiles did not vary as a
function of whether the upper or the lower face was shown. This
seems odd given that the principal physical difference between D
and ND smiles is visible only in the upper face. As we will see in
the regression analyses, however, other features contributed to
participants‘ ratings, and these may have enabled them to distinguish between smile types simply on the basis of the lower face.
815
Experiment 4
Our aim in this study was to examine whether judgments of
smiles would be affected by seeing static images versus dynamic
displays. Several studies have shown that dynamic information
increases the recognition of emotional expressions (Ambadar,
Schooler, & Cohn, 2005; Bassili, 1979; Bould & Morris, 2008;
Bruce & Valentine, 1988; Wehrle, Kaiser, Schmidt, & Scherer,
2000) and influences perceptions of the naturalness and spontaneity of expressions (Krumhuber & Kappas, 2005; Sato & Yoshikawa, 2004). These benefits of dynamic displays were found to
be due to the motion signal itself rather than additional static
information contained in moving sequences. Ambadar et al. (2005)
and Bould and Morris (2008) showed that “multistatic sequences,”
containing the same number of frames as dynamic images but with
a mask between each frame to prevent apparent motion, were
recognized with lower accuracy than dynamic displays.
We used a similar paradigm in which participants were presented with either multistatic images or dynamic displays of spontaneous and deliberate D and ND smiles. Bearing in mind that the
D smile is defined in terms of its morphology, we expected
perceivers to be able to distinguish between D and ND smiles
regardless of whether the smiles were static or dynamic. However,
the ability to distinguish between spontaneous and deliberate
smiles should be impaired when seeing static displays because
dynamic information concerning the spontaneity and genuineness
of a smile is absent. Participants should therefore distinguish
between spontaneous and posed smiles only when exposed to
dynamic faces, with spontaneous smiles being rated as more genuine and amused.
Method
Participants
Sixty undergraduate students (30 men, 30 women), ages 18 to
32 years (M ⫽ 22.35 years, SD ⫽ 3.46) at Cardiff University took
part individually in this study either for course credit or for a
payment of £2. Half were randomly allocated to the static mode,
the other half to the dynamic mode.
Regression Analyses
Multiple regression analyses showed that the pattern of results
varied depending on whether the lower or upper face was shown
(see Table 2, Experiment 3). As expected, in the upper face
condition only the D marker was a significant and positive predictor of how genuine and amused ( ps ⬍ .05) the smile was
perceived to be. However, other features were significant predictors of ratings made in the lower face condition. Here, apex
duration ( ps ⬍ .05) and asymmetry ( ps ⬍ .05) were associated
with ratings. As in Experiment 2, the longer the apex duration, the
more genuine and amused the smile was perceived to be. In
addition, the less asymmetrical the smile was, the more genuine it
was seen to be. Given that apex duration and asymmetry differed
between the smile types (see above), these features may have
contributed not only to overall ratings of genuineness and amusement in the lower face, but also to the distinction between D and
ND smiles. Thus, the role played by the D marker in judgments of
the upper face seems to have been taken over by asymmetry and
apex duration when only the lower face was visible.
Stimulus Material and Design
The same four smile types as in Experiment 3 were employed:
(a) 6 spontaneous D smiles, (b) 6 spontaneous ND smiles, (c) 6
deliberate D smiles, and (d) 6 deliberate ND smiles. In the dynamic mode condition, these 24 smiles were shown to the participants as video clips. In the static mode condition, each smile was
represented by a sequence of five still images extracted from the
video sequences. Images reflected the time course of the expression and showed a neutral face (first frames), followed by the smile
halfway through its onset (frame at midpoint from the start of the
smile to its peak), at its apex (frame at peak intensity), halfway
through its offset (frame at midpoint from the decrease of the smile
to its neutral position), and a neutral face (last frames) at the end.
These five images were displayed for the same total duration as the
corresponding clips in the dynamic mode. To prevent any perception of motion in the static sequences, we separated successive
pairs of static images by a blank screen that lasted 5 frames. The
resulting 24 sequences of static images were shown to all partic-
KRUMHUBER AND MANSTEAD
816
ipants in the static mode condition. In both conditions (static and
dynamic), stimuli were presented in random order via MediaLab
(Empirisoft).
Procedure and Dependent Variables
The procedure in the dynamic condition was identical to that in
Experiment 2. In the static condition, participants were instructed
as follows: “What you are going to see is a series of still images,
each of which depicts a person expressing a spontaneous or posed
smile. Each picture in the set shows one segment of the smile
episode. The pictures are separated by a blank screen. In total, you
will only see 5 still pictures of each episode. The duration for
which the pictures are displayed will vary.” Further instructions
regarding the nature of spontaneous and deliberate smiles, as well
as the experimental task and procedures for responding, were the
same as in the previous studies. The dependent variables were the
same as those used in Experiments 2 and 3.
Results and Discussion
Analysis of Variance
A MANOVA with condition (spontaneous, deliberate) and
smile type (D, ND) as within-subjects factors and display mode
(static, dynamic) and sex of participant (male, female) as betweensubjects factors was conducted on the three dependent variables.
The main effect of sex of participant was not significant, F(3,
54) ⫽ 2.29, p ⬎ .05, ␩2p ⫽ .11. There was a significant main effect
of display mode, F(3, 54) ⫽ 3.53, p ⬍ .05, ␩2p ⫽ .16. This was
significant only for amusement, F(1, 56) ⫽ 5.87, p ⬍ .05, ␩2p ⫽
.09. Overall, static displays attracted significantly higher amusement ratings (M ⫽ 3.64) than did dynamic displays (M ⫽ 3.36).
Consistent with expectations, there was a significant interaction
between condition and display mode, F(3, 54) ⫽ 7.43, p ⬍ .001,
␩2p ⫽ .29. It was significant for all three variables: spontaneous, F(1,
56) ⫽ 21.41, p ⬍ .001, ␩2p ⫽ .28; genuine, F(1, 56) ⫽ 15.06, p ⬍
.001, ␩p2 ⫽ .21; and amused, F(1, 56) ⫽ 14.34, p ⬍ .001,
␩2p ⫽ .20. Paired samples t tests showed that participants discriminated between spontaneous and deliberate smiles in the dynamic,
but not in the static, mode. As can be seen in Figure 1, spontaneous
smiles were rated as more genuine, t(29) ⫽ 3.94, p ⬍ .001, and
amused, t(29) ⫽ 5.38, p ⬍ .001, than deliberate smiles. No such
difference emerged for static displays— genuine, t(29) ⫽ ⫺1.03,
p ⬎ .05; amused, t(29) ⫽ 0.22, p ⬎ .05— indicating that significant information about the spontaneity of the expression was not
transmitted via static cues. The finding is consistent with previous
research and suggests that dynamic displays convey information relevant to expression perception over and above that contained in static
representations (see Ambadar et al., 2005; Wehrle et al., 2000).
There was also evidence that participants were misled by the
static representations. Independent samples t tests showed that
significantly higher ratings of genuineness, t(58) ⫽ 2.78, p ⬍ .01,
and amusement, t(58) ⫽ 3.69, p ⬍ .001, were given to deliberate
smiles, and that spontaneous smiles were perceived as less genuine
when shown in the static compared with the dynamic mode,
t(58) ⫽ ⫺1.98, p ⱕ .05. This shows that static displays resulted in
less accurate judgments of both spontaneous and deliberate smiles
and also to an overattribution of emotion in the case of deliberate
smiles.
The influence of the D marker should not depend on whether a
face is static or dynamic. Indeed, the effect of the presence or
absence of this marker was not moderated by display mode. There
were significant main effects for condition, F(3, 54) ⫽ 6.35, p ⬍
.01, ␩2p ⫽ .26, and smile type, F(3, 54) ⫽ 115.78, p ⬍ .001, ␩2p ⫽
.86. These were qualified by a significant interaction between
condition and smile type, F(3, 54) ⫽ 17.71, p ⬍ .001, ␩2p ⫽ .50.
The interaction was significant for all three dependent measures:
spontaneous, F(1, 56) ⫽ 24.44, p ⬍ .001, ␩2p ⫽ .30; genuine, F(1,
56) ⫽ 34.25, p ⬍ .001, ␩2p ⫽ .38; and amused, F(1, 56) ⫽ 53.81,
p ⬍ .001, ␩2p ⫽ .49. Analyses of simple effects revealed that both
spontaneous and deliberate D smiles were rated as more genuine—
spontaneous, F(1, 56) ⫽ 57.94, p ⬍ .001, ␩2p ⫽ .51; deliberate,
F(1, 56) ⫽ 5.44, p ⬍ .05, ␩2p ⫽ .09 —and amused—spontaneous,
F(1, 56) ⫽ 228.24, p ⬍ .001, ␩2p ⫽ .80; deliberate, F(1, 56) ⫽
109.42, p ⬍ .001, ␩2p ⫽ .66 —than ND smiles (see Table 1,
Experiment 4), thereby replicating the findings of Experiment 3.
Comparing smile types across elicitation condition, spontaneous
D smiles were judged to be significantly more genuine, F(1, 56) ⫽
29.71, p ⬍ .001, ␩2p ⫽ .35, and amused, F(1, 56) ⫽ 62.15, p ⬍
.001, ␩2p ⫽ .53, than deliberate D smiles were. This is in line with
the two previous studies and shows the critical role played by
condition in shaping the perceived quality of D smiles.
It is interesting that participants’ ratings of ND smiles also
varied as a function of elicitation condition, such that less genu-
Figure 1. Genuineness and amusement ratings of spontaneous and deliberate smiles in the static and dynamic
conditions of Experiment 4. Error bars show standard errors.
FELT AND FALSE SMILES
ineness, F(1, 56) ⫽ 3.89, p ⱕ .05, ␩2p ⫽ .06, and amusement, F(1,
56) ⫽ 7.01, p ⬍ .05, ␩2p ⫽ .11, were attributed to spontaneous than
to deliberate ND smiles. The finding provides further evidence of
opposing effects on judgments of D and ND smiles. Among ND
smiles, those made deliberately appear more genuine and as reflecting greater amusement than those made spontaneously.
Regression Analyses
Multiple regression analyses showed that the results varied as a
function of static and dynamic displays (see Table 2, Experiment
4). For dynamic displays, the same three facial features as in
Experiment 2 emerged as significant predictors: apex duration
( ps ⬍ .05), asymmetry ( ps ⬍ .01), and nonpositive facial actions
( ps ⬍ .05). The longer the apex, the more genuine and amused the
smile was judged to be. Similarly, less asymmetry and fewer
nonpositive facial actions were associated with higher ratings of
genuineness and amusement. For static displays, only nonpositive
facial actions ( ps ⬍ .05) and the D marker ( ps ⬍ .05) significantly
predicted judgments. Higher ratings of amusement were given to
smiles characterized by fewer nonpositive facial actions and by the
D marker. Similarly, fewer nonpositive facial actions were associated with increased ratings of smile genuineness.
General Discussion
The aim of this research was to test the value of the D smile as
a spontaneous and genuine sign of enjoyment. It has been argued
that D smiles signal spontaneous, felt emotions that differ from
posed or false smiles with respect to the presence of AU 6. In the
present research, we explored the reliability and validity of the D
smile with respect to both expression and perception. In Experiment 1, we focused on expression and investigated whether people
were able to activate the orbicularis oculi, pars lateralis muscle
(AU 6) to produce a voluntary D smile and how frequently they
did so. We examined whether D smiles occur solely or predominantly under spontaneous conditions, and whether ND smiles
occur solely or predominantly under deliberate conditions. Participants either spontaneously smiled in reaction to genuinely amusing stimuli (spontaneous condition) or were instructed to pose a
smile expression (deliberate condition). It is interesting that there
were similar proportions of smiles with and without the D marker,
AU 6, in these two conditions. Thus, senders could produce D
smiles deliberately, and they did so quite often. Furthermore, not
all the observed ND smiles were posed; some occurred in the
spontaneous condition. These findings are in line with increasing
evidence that D smiles and ND smiles can occur under both
spontaneous and deliberate conditions (Schmidt & Cohn, 2001;
Schmidt et al., 2006, 2009; Smith et al., 1996) and cast doubt on
claims that D smiles are spontaneous signs of felt enjoyment that
cannot be feigned (Ekman, 1985, 1989, 1993).
Whether a D or ND smile was shown was less a function of the
spontaneity of the expression than of the intensity of the smile. D
smiles were more intense than ND smiles, and when they were
made deliberately, they were even more intense than D smiles
made in the spontaneous condition. This is consistent with observations that D smiles tend to involve stronger smiling actions than
ND smiles and that they covary with the degree of smile strength
(Ekman & Friesen, 1978; Frank et al., 1993). In a recent study by
817
Messinger et al. (2008), changes in smile intensity were also found
to be linked to changes in eye constriction (AU 6). That is, the
more intense the smile expression, the higher was the probability
of a D smile. Messinger et al. (2008) consequently suggested
treating AU 6 in D smiles not as a dichotomous signal that
distinguishes D from ND smiles (see Abe, Beetham, & Izard,
2002; Messinger, 2002; Messinger et al., 1999, for similar claims),
but rather as an ordinally graded index of smile intensity. An
important source of such association may be muscular synergies
(Messinger et al., 2008). That is, as the zygomaticus major muscle
pulls the lip corners upward, it raises the cheeks toward the eyes.
These synergies provide a basis for explaining the effects of
expression intensity by linking the occurrence of AU 6 directly to
the intensity of the smiling action (AU 12).4
Self-reports of positive experience varied as a function of both
elicitation condition and smile type. Higher ratings accompanied D
smiles occurring in the spontaneous condition than in the deliberate condition. This supports the notion that only the spontaneous D
smile reflects genuine enjoyment (Ekman et al., 1988, 1990; Frank
et al., 1993). By contrast, in the deliberate condition senders were
able to produce D smiles that did not reflect felt positive affect.
Whether made spontaneously or deliberately, however, D smiles
were associated with more positive emotion than ND smiles,
suggesting that this relation holds even when a D smile is being
posed. A possible avenue for future research would be to study the
relation between D smiles and emotional experience under varying
conditions, including negative ones. Whereas Bonanno and colleagues (Keltner & Bonanno, 1997; Papa & Bonanno, 2008)
showed that D smiles led to less negative feelings on the part of
individuals talking about their recently deceased spouse or in
response to a negative affect manipulation during a monologue
task, such mitigation of negative affect by D smiles was not found
by Soussignan (2002). Moreover, the content of participants’ talk
in the monologue task (Papa & Bonanno, 2008) and the bereavement interview (Keltner & Bonanno, 1997) during which facial
behaviors were measured was not taken into account. An important
step, therefore, would be to conduct a time-based analysis of facial
behavior in order to clarify the extent to which D smiles serve a
self-regulatory function. Preliminary work of this nature has already been done by Wagner and Lee (1999; see also Lee & Beattie,
1998), who tracked the occurrence of D and ND smiles in relation
to specific events in time. These researchers found considerable D
smiling accompanying negative statements. Further research is
clearly needed to shed more light on the intrapersonal consequences of D smiles.
In Experiments 2– 4, we focused on the perception of smiles.
Specifically, we examined how genuine and amused D and ND
smiles were judged to be, and whether people distinguished be4
It is worth noting that the neurological claims for different neural
pathways involved in spontaneous (D: AU 6 ⫹ 12) and deliberate (ND: AU
12) smiles are not unequivocal. The neurological literature shows that it is
difficult to provide a precise definition of the neural systems involved in
these two processes (Pizzamiglio, Caltagirone, & Zoccolotti, 1989). Indeed, Ekman and colleagues concluded that the voluntary–involuntary
dichotomy is too simple (Ekman et al., 1981). Instead, these authors
suggested that there are several types of voluntary and involuntary expressions, each likely to vary with respect to the underlying neural pathways
(Ekman, 1984).
818
KRUMHUBER AND MANSTEAD
tween these two smile types when they were made under spontaneous and deliberate conditions. In each study, we also investigated the predictive value of the D marker alongside other smile
attributes such as asymmetry; irregularity; onset, apex, and offset
durations; and nonpositive facial actions. If the D marker is indeed
the most diagnostic symptom of a felt smile, we expected its
capacity to predict perceivers’ ratings of the smiles to be consistent
and at least as great as the predictive value of the other smile
attributes.
Participants generally rated spontaneous and deliberate smiles
differently. In all three studies, the spontaneous D smile was rated
as most genuine and amused, consistent with the notion that this
type of smile is perceived as expressing felt positive emotion
(Frank et al., 1993; Gosselin, Beaupré, & Boissonneault, 2002;
Gosselin, Perron, et al., 2002). In Experiment 2, perceivers also
distinguished between spontaneous D smiles that varied in intensity. D smiles of higher intensity were judged to reflect greater
subjective enjoyment than D smiles of lower intensity. This shows
that the intensity of the expression made by a sender significantly
affected how spontaneous his or her D smiles were perceived to be
and extends previous research on the effects of expression intensity (e.g., Bolzani Dinehart et al., 2005; Hess et al., 1997; Messinger et al., 2008).
In Experiment 3, perceivers were shown either the upper or the
lower face. Contrary to expectations, similar ratings were made
when seeing the upper or lower face, showing that judges were
able to use cues other than the D marker itself to distinguish
between D and ND smiles. In Experiment 4, perceivers saw static
or dynamic versions of the smiles, and this had a significant impact
on judges’ ratings. When smiles were presented in their dynamic
form, spontaneous smiles were judged as more genuine and
amused than deliberate smiles. However, this difference did not
emerge for static displays. This is consistent with past research on
dynamic displays (Ambadar et al., 2005; Wehrle et al., 2000) and
shows that dynamic information contains important cues concerning the spontaneity of the expression.
In comparing the utility of several physical attributes of smiles
in predicting how smiles were judged, we found that asymmetry of
expression and apex duration were consistent predictors across the
three studies. The less asymmetric the smile was and the longer its
apex duration, the more genuine and amused it was judged to be.
The presence of nonpositive facial actions also significantly predicted judgments, with smiles being rated as more genuine and
amused when they had fewer nonpositive facial actions, such as
frowning and lip pressing. These findings extend previous research
(Ambadar et al., 2009; Gosselin, Beaupré, & Boissonneault, 2002;
Gosselin, Perron, et al., 2002; Krumhuber & Kappas, 2005) and
suggest that all three variables contribute to how smile expressions
are evaluated. It is interesting that the predictive utility of the D
marker was rather limited. It only significantly predicted ratings
when judges saw the upper face or when they viewed static rather
than dynamic displays. Furthermore, in the case of static displays,
the D marker predicted only how amused (but not how genuine)
the smile was judged to be. These results call into question the
received wisdom that the D marker is the most reliable index of
felt positive emotion (Frank & Ekman, 1993). Our findings point
to a different conclusion, namely that the D marker is one among
several attributes that lead judges to regard smiles as genuine and
as reflecting amusement. In the present research, the D marker was
not the most diagnostic attribute, nor was it a consistent cue for
rating smiles as expressing spontaneous amusement.
Much of our current understanding of facial expressions of
emotion derives from research using static facial images (Russell,
Bachorowski, & Fernández-Dols, 2003). A distinction is often
drawn between “spontaneous” and “deliberate” smiles on the basis
of the presence or absence of the D marker. Clearly, this morphological marker is discernible in static displays and, as the present
research shows, it has reliable effects on how smiles are judged on
the basis of static images. In everyday life, however, judgments of
the genuineness of smiles are made on the basis of dynamic
expressions, and the evidence from the present research is that the
D marker is a less important cue when dynamic information is
available. Future research should focus on dynamic displays and
consider a broader range of behavioral cues than the D marker.
A feature of the present work that might be regarded as a
limitation stems from the way in which we equated smile intensity
across conditions in the three judgment studies, with D and ND
smiles being of comparable intensity in both the spontaneous and
deliberate conditions. There is no doubt that expression intensity
influences the perceived intensity of the associated emotion (see
Bolzani Dinehart et al., 2005; Hess et al., 1997; Messinger et al.,
2008). However, in the present research, we wanted to rule out
differences in smile intensity as a possible explanation for differential responses to smiles in the spontaneous and deliberate conditions. We therefore kept this variable constant across conditions
and considered the effects of expression intensity separately in
Experiment 2, where we examined the perception of D smiles that
differed in intensity within the spontaneous condition. Future
research could include a broader range of variation in smile intensity and examine whether the influence of other physical attributes
such as the ones studied in the current research is moderated by
smile intensity.
In conclusion, we have shown that D smiles occurred as both
felt and posed expressions, and that the presence or absence of the
D marker was not the only basis on which those who judged smiles
made ratings of genuineness and amusement. The role of the D
marker, AU 6, therefore needs to be reconsidered. The D smile is
operationalized as the combination of AUs 6 and 12. It is clear that
D smiles can and do occur when the individual experiences genuine amusement. However, it is not the case that all smiles involving the D marker reflect genuine amusement. D smiles also occur
when the individual poses amusement and are even shown under
negative emotional circumstances. The usefulness of facial muscle
units such as AUs 6 and 12 in identifying whether a smile expresses enjoyment is therefore thrown into doubt. The evidence
from the present research is that the D marker by itself is not a
reliable basis for inferring whether a smile is spontaneous, genuine, or reflective of subjective amusement.
References
Abe, J. A. A., Beetham, M., & Izard, C. E. (2002). What do smiles mean?
An analysis in terms of differential emotions theory. In M. H. Abel (Ed.),
An empirical reflection on the smile (pp. 83–109). New York: Edwin
Mellen Press.
Ambadar, Z., Cohn, J. F., & Reed, L. I. (2009). All smiles are not created
equal: Morphology and timing of smiles perceived as amused, polite,
and embarrassed/nervous. Journal of Nonverbal Behavior, 33, 17–34.
Ambadar, Z., Schooler, J., & Cohn, J. (2005). Deciphering the enigmatic
FELT AND FALSE SMILES
face: The importance of facial dynamics in interpreting subtle facial
expressions. Psychological Science, 16, 403– 410.
Ansfield, M. R. (2007). Smiling when distressed: When a smile is a frown
turned upside down. Personality and Social Psychology Bulletin, 33,
763–775.
Bassili, J. N. (1979). Emotion recognition: The role of facial movement
and the relative importance of upper and lower areas of the face. Journal
of Personality and Social Psychology, 37, 2049 –2058.
Bolzani Dinehart, L. H., Messinger, D. S., Acosta, S. I., Cassel, T.,
Ambadar, Z., & Cohn, J. (2005). Adult perceptions of positive and
negative infant emotional expressions. Infancy, 8, 279 –303.
Boucher, J. D., & Ekman, P. (1975). Facial areas and emotional information. Journal of Communication, 25, 21–29.
Bould, E., & Morris, N. (2008). Role of motion signals in recognizing
subtle facial expressions of emotion. British Journal of Psychology, 99,
167–189.
Bruce, V., & Valentine, T. (1988). When a nod’s as good as a wink: The
role of dynamic information in facial recognition. In M. M. Gruneberg,
P. E. Morris, & R. N. Sykes (Eds.), Practical aspects of memory:
Current research and issues (Vol. 1, pp. 169 –174). New York: Wiley.
Cohn, J. F., & Schmidt, K. L. (2004). The timing of facial motion in posed
and spontaneous smiles. International Journal of Wavelets, Multiresolution, and Information Processing, 2, 1–12.
Duchenne, G. B. A. (1990). The mechanism of human facial expression.
R. A. Cuthbertson (Ed. and Trans.). Cambridge, UK: Cambridge University Press. (Original work published 1862)
Ekman, P. (1982). Emotion in the human face (2nd ed.). New York:
Cambridge University Press.
Ekman, P. (1984). Expression and the nature of emotion. In K. R. Scherer
& P. Ekman (Eds.), Approaches to emotion (pp. 319 –344). Hillsdale,
NJ: Erlbaum.
Ekman, P. (1985). Telling lies. New York: Norton.
Ekman, P. (1989). The argument and evidence about universals in facial
expressions of emotion. In H. Wagner & A. Manstead (Eds.), Handbook
of psychophysiology: The biological psychology of the emotions and
social processes (pp. 143–164). London: Wiley.
Ekman, P. (1992a). Are there basic emotions? Psychological Review, 99,
550 –553.
Ekman, P. (1992b). An argument for basic emotions. Cognition and
Emotion, 6, 169 –200.
Ekman, P. (1993). Facial expression and emotion. American Psychologist,
48, 384 –392.
Ekman, P., & Davidson, R. J. (1993). Voluntary smiling changes regional
brain activity. Psychological Science, 4, 342–345.
Ekman, P., Davidson, R., & Friesen, W. V. (1990). The Duchenne smile:
Emotional expression and brain physiology II. Journal of Personality
and Social Psychology, 58, 342–353.
Ekman, P., & Friesen, W. V. (1978). The Facial Action Coding System.
Palo Alto, CA: Consulting Psychologists Press.
Ekman, P., & Friesen, W. V. (1982). Felt, false and miserable smiles.
Journal of Nonverbal Behavior, 6, 238 –252.
Ekman, P., Friesen, W. V., & Hager, J. C. (2002). The Facial Action
Coding System (2nd ed.). Salt Lake City, UT: Research Nexus eBook.
Ekman, P., Friesen, W. V., & O’Sullivan, M. (1988). Smiles when lying.
Journal of Personality and Social Psychology, 54, 414 – 420.
Ekman, P., Hager, J. C., & Friesen, W. V. (1981). The symmetry of
emotional and deliberate facial actions. Psychophysiology, 18, 101–106.
Ekman, P., Matsumoto, D. R., & Friesen, W. V. (1997). Facial expression
in affective disorders. In P. Ekman & E. L. Rosenberg (Eds.), What the
face reveals: Basic and applied studies of spontaneous expression using
the Facial Action Coding System (pp. 331–342). New York: Oxford
University Press.
Ekman, P., Roper, G., & Hager, J. C. (1980). Deliberate facial movement.
Child Development, 51, 886 – 891.
819
Fox, N. A., & Davidson, R. J. (1988). Patterns of brain electrical activity
during facial signs of emotion in 10-month-old infants. Developmental
Psychology, 24, 230 –236.
Frank, M. G., & Ekman, P. (1993). Not all smiles are created equal: The
differences between enjoyment and nonenjoyment smiles. Humor: International Journal of Humor Research, 6, 9 –26.
Frank, M. G., Ekman, P., & Friesen, W. V. (1993). Behavioral markers and
recognizability of the smile of enjoyment. Journal of Personality and
Social Psychology, 64, 83–93.
Giudice, M. D., & Colle, L. (2007). Differences between children and
adults in the recognition of enjoyment smiles. Developmental Psychology, 43, 796 – 803.
Gonzaga, G. C., Keltner, D., Londahl, E. A., & Smith, M. D. (2001). Love
and the commitment problem in romantic relations and friendship.
Journal of Personality and Social Psychology, 81, 247–262.
Gosselin, P., Beaupré, M., & Boissonneault, A. (2002). Perception of
genuine and masking smiles in children and adults: Sensitivity to traces
of anger. The Journal of Genetic Psychology, 163, 58 –71.
Gosselin, P., Perron, M., Legault, M., & Campanella, P. (2002). Children’s
and adults’ knowledge of the distinction between enjoyment and nonenjoyment smiles. Journal of Nonverbal Behavior, 26, 83–108.
Hager, J. C., & Ekman, P. (1997). The asymmetry of facial actions is
inconsistent with models of hemispheric specialization. In P. Ekman &
E. Rosenberg (Eds.), What the face reveals: Basic and applied studies of
spontaneous expression using the Facial Action Coding System (pp.
40 – 62). New York: Oxford University Press.
Harker, L. A., & Keltner, D. (2001). Expressions of positive emotion in
women’s college yearbook pictures and their relationship to personality
and life outcomes across adulthood. Journal of Personality and Social
Psychology, 80, 112–124.
Harris, C. R., & Alvarado, N. (2005). Facial expressions, smile types, and
self-report during humour, tickle, and pain. Cognition & Emotion, 19,
655– 669.
Hess, U., Blairy, S., & Kleck, R. E. (1997). The intensity of emotional
facial expressions and decoding accuracy. Journal of Nonverbal Behavior, 21, 241–257.
Hess, U., Kappas, A., McHugo, G. J., Kleck, R. E., & Lanzetta, J. T.
(1989). An analysis of the encoding and decoding of spontaneous and
posed smiles: The use of facial electromyography. Journal of Nonverbal
Behavior, 13, 121–136.
Hess, U., & Kleck, R. E. (1990). Differentiating emotion elicited and
deliberate emotional facial expressions. European Journal of Social
Psychology, 20, 369 –385.
Hess, U., & Kleck, R. E. (1994). The cues decoders use in attempting to
differentiate emotion-elicited and posed facial expressions. European
Journal of Social Psychology, 24, 367–381.
Keltner, D. (1995). Signs of appeasement: Evidence for the distinct displays of embarrassment, amusement, and shame. Journal of Personality
and Social Psychology, 68, 441– 454.
Keltner, D., & Bonanno, G. A. (1997). A study of laughter and dissociation: Distinct correlates of laughter and smiling during bereavement.
Journal of Personality and Social Psychology, 73, 687–702.
Kerstenbaum, R. (1992). Feeling happy versus feeling good: The processing of discrete and global categories of emotional expressions by children and adults. Developmental Psychology, 28, 1132–1142.
Kohler, C. G., Turner, T., Stolar, N. M., Bilker, W. B., Brensinger, C. M.,
Gur, R. E., & Gur, R. C. ( 2004). Differences in facial expressions of
four universal emotions. Psychiatry Research, 128, 235–244.
Krumhuber, E., & Kappas, A. (2005). Moving smiles: The role of dynamic
components for the perception of the genuineness of smiles. Journal of
Nonverbal Behavior, 29, 3–24.
Krumhuber, E., Manstead, A. S. R., Cosker, D., Marshall, D., & Rosin,
P. L. (2009). Effects of dynamic attributes of smiles in human and
820
KRUMHUBER AND MANSTEAD
synthetic faces: A simulated job interview setting. Journal of Nonverbal
Behavior, 33, 1–15.
Krumhuber, E., Manstead, A. S. R., Cosker, D., Marshall, D., Rosin, P. L.,
& Kappas, A. (2007). Facial dynamics as indicators of trustworthiness
and cooperative behavior. Emotion, 7, 730 –735.
Krumhuber, E., Manstead, A. S. R., & Kappas, A. (2007). Temporal
aspects of facial displays in person and expression perception: The
effects of smile-dynamics, head-tilt and gender. Journal of Nonverbal
Behavior, 31, 39 –56.
Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (2001). International
Affective Picture System (IAPS): Instruction manual and affective ratings (Tech. Rep. No. A-5). Gainesville: The Center for Research in
Psychophysiology, University of Florida.
Lee, V., & Beattie, G. (1998). The rhetorical organization of verbal and
nonverbal behaviour in emotion talk. Semiotica, 120, 39 –92.
Levenson, R. W., Ekman, P., & Friesen, W. V. (1990). Voluntary facial
action generates emotion-specific autonomic nervous system activity.
Psychophysiology, 27, 363–384.
Messinger, D. S. (2002). Positive and negative: Infant facial expressions
and emotions. Current Directions in Psychological Science, 11, 1– 6.
Messinger, D. S., Cassel, T. D., Acosta, S. I., Ambadar, Z., & Cohn, J. F.
(2008). Infant smiling dynamics and perceived positive emotion. Journal of Nonverbal Behavior, 32, 133–155.
Messinger, D. S., Fogel, A., & Dickson, K. L. (1999). What’s in a smile?
Developmental Psychology, 35, 701–708.
Messinger, D. S., Fogel, A., & Dickson, K. L. (2001). All smiles are
positive, but some smiles are more positive than others. Developmental
Psychology, 37, 642– 653.
Miles, L., & Johnston, L. (2007). Detecting happiness: Perceiver sensitivity
to enjoyment and non-enjoyment smiles. Journal of Nonverbal Behavior, 31, 259 –275.
Nusseck, M., Cunningham, D. W., Wallraven, C., & Bülthoff, H. H.
(2008). The contribution of different facial regions to the recognition of
conversational expressions. Journal of Vision, 8, 1–23.
Papa, A., & Bonanno, G. A. (2008). Smiling in the face of adversity: The
interpersonal and intrapersonal functions of smiling. Emotion, 8, 1–12.
Parkinson, B. (2005). Do facial movements express emotions or communicate motives? Personality and Social Psychology Review, 9, 278 –311.
Peace, V., Miles, L., & Johnston, L. (2006). It doesn’t matter what you
wear: The impact of posed and genuine expressions of happiness on
product evaluation. Social Cognition, 24, 137–168.
Pizzamiglio, L., Caltagirone, C., & Zoccolotti, P. (1989). Facial expression
of emotion. In F. Boller & J. Grafman (Eds.), Handbook of neuropsychology (pp. 383– 401). New York: Elsevier Science.
Rinn, W. E. (1991). Neuropsychology of facial expression. In R. S.
Feldmann & B. Rimé (Eds.), Fundamentals of nonverbal behavior (pp.
3–30). Cambridge, UK/Paris: Cambridge University Press/Editions de la
Maison des Science de 1’Homme.
Russell, J. A., Bachorowski, J. A., & Fernández-Dols, J. M. (2003). Facial
and vocal expressions of emotion. Annual Review of Psychology, 54,
329 –349.
Sato, W., & Yoshikawa, S. (2004). The dynamic aspects of emotional
facial expressions. Cognition and Emotion, 18, 701–710.
Sayette, M. A., Cohn, J. F., Wertz, J. M., Perrott, M. A., & Parrott, D. J.
(2001). A psychometric evaluation of the Facial Action Coding System
for assessing spontaneous expression. Journal of Nonverbal Behavior,
25, 167–185.
Scherer, K. R., & Ceschi, G. (2000). Criteria for emotion recognition from
verbal and nonverbal expression: Studying baggage loss in the airport.
Personality and Social Psychology Bulletin, 26, 327–339.
Schmidt, K. L., Ambadar, Z., Cohn, J. F., & Reed, L. I. (2006). Movement
differences between deliberate and spontaneous facial expressions: Zygomaticus major action in smiling. Journal of Nonverbal Behavior, 30, 37–52.
Schmidt, K. L., Bhattacharya, S., & Denlinger, R. (2009). Comparison of
deliberate and spontaneous facial movement in smiles and eyebrow
raises. Journal of Nonverbal Behavior, 33, 35– 45.
Schmidt, K. L., & Cohn, J. F. (2001). Human facial expressions as
adaptations: Evolutionary questions in facial expression research. Yearbook of Physical Anthropology, 44, 3–24.
Schmidt, K. L., Cohn, J. F., & Tian, Y. (2003). Signal characteristics of
spontaneous facial expressions: Automatic movement in solitary and
social smiles. Biological Psychology, 65, 49 – 66.
Schneider, K., & Josephs, I. (1991). The expressive and communicative
functions of preschool children’s smiles in an achievement situation.
Journal of Nonverbal Behavior, 15, 185–198.
Schneider, K., & Unzer, L. (1992). Preschoolers’ attention and emotion in
an achievement and an effect game: A longitudinal study. Cognition &
Emotion, 6, 37– 63.
Skinner, M., & Mullen, B. (1991). Facial asymmetry in emotional expression: A meta-analysis of research. British Journal of Social Psychology,
30, 113–124.
Smith, M. C., Smith, M. K., & Ellgring, H. (1996). Spontaneous and posed
facial expression in Parkinson’s disease. Journal of the International
Neuropsychological Society, 2, 383–391.
Soussignan, R. (2002). Duchenne smile, emotional experience, and autonomic
reactivity: A test of the facial feedback hypothesis. Emotion, 2, 52–74.
Soussignan, R., & Schaal, B. (1996). Forms and social signal value of
smiles associated with pleasant and unpleasant sensory experience.
Ethology, 102, 1020 –1041.
Steiner, F. (1986). Differentiating smiles. In E. Bänninger-Huber & F.
Steiner (Eds.), FACS in psychotherapy research (pp. 139 –148). Zürich:
Department of Clinical Psychology, Universität Zürich.
Surakka, V., & Hietanen, J. K. (1998). Facial and emotional reactions to
Duchenne and non-Duchenne smiles. International Journal of Psychophysiology, 29, 23–33.
Tremblay, A., Deschenes, J., Poulin, B., Roy, J., Kirouac, G., & Kappas, A.
(1993). Identification de régions faciales critiques dans le jugement
d’authenticité d’un sourire d’expressions de joie véritablement ressentie
[Identification of critical facial regions in the judgment of the genuineness of smiles expressing experienced joy]. Presented at the XVIéme
Congrés annuel de la Société Québécoise pour la Recherche en Psychologie, Québec, Canada (November).
Wagner, H., & Lee, V. (1999). Facial behavior alone and in the presence
of others. In P. Philippot, R. S. Feldman, & E. J. Coats (Eds.), The social
context of nonverbal behavior (pp. 262–286). Cambridge, UK: Cambridge University Press.
Wehrle, T., Kaiser, S., Schmidt, S., & Scherer, K. R. (2000). Studying the
dynamics of emotional expression using synthesized facial muscle
movements. Journal of Personality and Social Psychology, 78, 105–119.
Weiss, F., Blum, G. S., & Gleberman, L. (1987). Anatomically based
measurement of facial expression in simulated versus hypnotically induced affect. Motivation and Emotion, 11, 67– 81.
Williams, L. M., Senior, C., David, A. S., Loughland, C. M., & Gordon, E.
(2001). In search of the “Duchenne smile”: Evidence from eye movements. Journal of Psychophysiology, 15, 122–127.
Winer, B. J., Brown, D. R., & Michels, K. M. (1991). Statistical principles
in experimental design. New York: McGraw-Hill.
Zaalberg, R., Manstead, A. S. R., & Fischer, A. H. (2002, June). Facial
behavior in game-playing situations: The mediating role of display rules
and social motives. Paper presented as part of symposium “Emotion in
Social Context” (convenors: Antony Manstead and Agneta Fischer) at
General Meeting of the European Association of Experimental Social
Psychology, San Sebastian, Basque Country.
Received December 12, 2007
Revision received October 1, 2009
Accepted October 1, 2009 䡲