Vous êtes sur la page 1sur 23

Aspen Pub.

/IYC AS160-03 March 5, 2003 18:35 Char Count= 0

Infants and Young Children

Vol. 16, No. 2, pp. 120–142
c 2003 Lippincott Williams & Wilkins, Inc.

Emotional Expressions of
Young Infants and Children
A Practitioner’s Primer
Margaret Wolan Sullivan, PhD; Michael Lewis, PhD
Research on emotional development in infancy has benefited greatly from the use of video-
tape technology and coding systems that allow detailed coding of facial movements. Today we
know that a core set of human facial expressions, composed of specific movements in the brow,
eye/cheek, and mouth regions of the face, are probably innate. Theorists continue to debate the
meaning of these expressions as well as how they are organized and become regulated over the
first several years of life. Despite continuing debate and research on these issues, early facial ex-
pressions have practical, signal value for caregivers and practitioners alike. This article surveys
what is known about the appearance and early normative, developmental course of emotional
expressions, noting similarities and differences in special populations when available. Its goal is
to provide practitioners with basic information to help them and the parents they serve become
better able to recognize the expressive signals of the infants and young children in their care.
Key words: emotion, emotional development, facial expressions, infants, nonverbal commu-

S YSTEMS for deciphering the facial expres-

sions of infants and young children were
developed in the 1980s. These systems rep-
Lewis & Michalson, 1983). Despite this find-
ing, facial coding systems have only just be-
gun to move out of the laboratory into clinical
resent important and needed tools for pro- settings (Gilbert et al., 1999). This is unfor-
moting more accurate, empirical study of tunate, because information about facial ex-
early emotional development. With them, re- pressions potentially has practical import for
searchers are beginning to understand when those working with infants, older, nonverbal
infants first express particular emotion sig- children, and their caregivers.
nals facially, the organization of these signals, For practitioners, facial expressions are in-
and their relation to other aspects of moti- formative in 2 ways. First and foremost, they
vated behavior. While many important ques- are social signals to others. Crying, vocaliz-
tions about the meaning and developmental ing, and bodily movements combine with fa-
course of early expressions continue to be re- cial expression to provide cues to an infant’s
searched and debated, it is now recognized status. Savvy caregivers no doubt make use
that most, if not all, of the facial components of all of these cues in interpreting infant be-
of the human expression repertoire can be ob- havior. Although psychologists may be reluc-
served shortly after birth (Camras, Holland, tant to assign a specific emotion value to
& Patterson, 1993; Izard & Malatesta, 1987; infant facial expressions, parents have no
problem doing so. In fact, they routinely use
facial expressions to attribute personality and
From the Institute for the Study of Child intellectual characteristics to a young baby
Development, UMDNJ-Robert Wood Johnson Medical (Haviland, 1983). They also use facial expres-
School, New Brunswick, NJ.
sions to gauge their own responses, thus help-
Corresponding author: Margaret Wolan Sullivan, In- ing to regulate their infants’ arousal and teach-
stitute for the Study of Child Development, UMDNJ-
Robert Wood Johnson Medical School (e-mail: sullivan@ ing them display rules (Malatesta & Haviland,
umdnj.edu). 1982). Parents’ reading and interpretation of
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0

Emotional Expressions 121

the infant’s expressive cues in daily interac- neurological and cognitive status of the child.
tion are key to the child’s social development, Facial expressions are controlled through the
emotion regulation, and early language learn- facial cranial nerves but are also intimately
ing (Mundy & Willoughby, 1996; Walden & linked to cognitive development (Lewis &
Knieps, 1996). Consequently, whether par- Michalson, 1983). Consequently, young chil-
ents observe and are having difficulty inter- dren’s facial expressions have some clinical
preting the expressive signals of their infant significance. Although their initial appearance
should be an important screening question for is organized at the brain’s subcortical level,
the practitioner. changes in their form or developmental pat-
Although some adults seem to have “nat- tern over time reveals that the child’s higher
ural” skills at reading emotional expressions, cognitive and motivational systems are be-
others will need some training to recognize coming integrated as brain maturation pro-
the facial signals of young infants whose ex- ceeds. Appropriate developmental changes in
pressions are often fleeting, subtle, and per- facial signals imply that certain cognitive func-
haps less well organized than those of older tions are preserved in neurologically dam-
children. This problem can be compounded aged children, for example. This fact has
when there is neurological impairment or long been recognized for smiling (McCall,
developmental delay. Expressive behaviors 1972), but is likely to be true for other
of infants and children with various forms expressions as well.
of disability have been described as muted, Data on individual differences in facial ex-
hard-to-read, or excessively labile and incon- pressions, their developmental trajectory in
gruous, depending on the population un- normally developing populations, and spe-
der study (Mundy, Yirmiya, & Sigman, 1990; cific information about facial expressions in
Kasari & Sigman, 1996; Sigman, Kasari, Kwon, atypical populations is growing. This article
& Yirmiya, 1992). When parents are un- surveys the expressions of which infants and
able to recognize and interpret emotional sig- young children are capable as revealed by fa-
nals from their infants, they will be uncer- cial coding systems such as MAX (Maximally
tain about their child’s needs and less able Discriminative Facial Coding System), its com-
to share positive affect. Consequently, par- panion, whole-face scoring system, AFFEX
ents may become less expressive themselves (Affective Expressions Scoring System; Izard,
(Dawson, Hill, Spencer, Galpert, & Waton, 1982), and Baby FACS (Facial Action Coding
1990), further degrading the quality of interac- System; Oster, 1978) and will cover typical ex-
tion. Learning to recognize facial expressions pressions. The article will not treat related and
and how they evolve is a worthwhile effort important topics such as arousal, physiologi-
because these and other nonverbal cues of cal reactivity, or emotion regulation. Rather,
emotion offer another channel of communi- the focus is a pragmatic one of how to rec-
cation when children’s behavior or vocal ut- ognize expressions—whether they are muted
terances are absent, or unclear. This channel or intense, prototypic or idiosyncratic. Like-
signals the infant’s arousal level, something wise, we set aside the thorny theoretical ques-
about the quality of the infant’s positive or tion of whether the emotional experiences of
negative response to concurrent stimulation, very young children are similar to or different
and may help explain other aspects of motor, from those of older children and adults. Since
postural, and behavioral responses. In short, the discussion of expressions will make use
the ability to interpret infants’ expressive sig- of the movements described in MAX, AFFEX,
nals is extremely important to parents person- and FACS, we briefly consider the nature of
ally and to promoting mutually satisfying inter- these tools.
actions. Although some of the assumptions underly-
Another way expressions can be helpful to ing MAX and AFFEX, as opposed to the Baby
practitioners is that they offer clues to the FACS coding systems, remain controversial,
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0


research using these systems shows that most that infants’ expressions to any given context
of the facial movements comprising adult vary across individuals. Even within individ-
emotion expressions are present and map uals, multiple expressions seen in rapid suc-
onto positive and negative reactions to stim- cession are common. Expressions are also sel-
ulation during the first months of life. MAX, dom pure and may blend 2 emotion signals.
AFFEX, and Baby FACS also share many fea- Since blended expression may persist in some
tures. For example, all observe movements children (Yirmiya, Kasari, Sigman & Mundy,
of the facial musculature in the forehead, 1989), the developmental course of and
eye/cheek, and mouth regions of the face. individual differences in blends are of some
They also agree, for the most part, on those interest.
facial movements that constitute emotion
signals. The major differences between the THE INTEREST FAMILY
systems are their theoretical orientation and
comprehensiveness in scoring facial move- Some do not consider interest is an emo-
ments. tional expression, but since it is a sign of pos-
MAX and AFFEX focus on a theoret- itive approach and receptivity to people and
ically limited set of “prototypic” expres- objects, it is coded in MAX/AFFEX as a family
sions of adults. In contrast, Baby FACS of expressions distinct from a neutral, awake,
does not, and is comprehensive in scor- but nonexpressive face. Figure 1 illustrates
ing all possible movement combinations. Be- 3 varieties of interest occurring in young in-
cause MAX (Izard, 1983/1995) concentrates fants. Notice that in all of the photographs,
on those facial movements that are max- the infants appear alert and attentive. How-
imally discriminative of 9 specific human ever, there are subtle differences in their ex-
emotion expressions, it lends itself more pressions, signaling differing qualities of that
readily to applied situations. Besides the attention.
9 facially expressed emotions (ie, Interest, Figure 1a has been characterized as
Surprise, Enjoyment, Anger, Sadness, Fear, the “open” or relaxed interest expression
Disgust, Contempt, and Shame), several con- (Sullivan & Lewis, 1989; Sullivan, Lewis,
trol or regulatory movements are also in- & Alessandri, 1992). The brows are raised
cluded in MAX. Blended expressions, describ- slightly, eyes wide open. The mouth is re-
ing combinations of specific facial movements laxed and open, taking a bow-shape as shown
in the face, also are deciphered (for example, here. The mouth may be closed in some
Anger/Sad is a common expression blend in variants, but the key is that there is no sign
infants). In MAX, single component expres- of tension. This expression occurs in envi-
sions are thought to reflect a lower intensity ronments that offer the infant low intensity,
of expression than are signals occurring in all nonthreatening stimulation. Also described
3 facial regions (Abe & Izard, 1999). MAX, like as “curious” and “wide-eyed wonder” in the
Baby FACS, is designed for use with videotape nonscientific literature, open interest is by
or still photographs, but AFFEX can be used far the most common expression of young
as a live observational system. If you are inter- infants. Adults, on the other hand, typically
ested in what facial expression or expressions maintain a neutral/awake expression as our
can be “read” as emotional signals, MAX or modal expression. This is not the case for
AFFEX is the best application. We use it here infants, whose faces are rarely still and whose
to present the facial expressions of infancy, expressions change rapidly (Malatesta &
describing specific expressions, their devel- Haviland, 1982). In fact, a “neutral” or sober
opment, and what is currently known about face is rare in awake, attentive infants much
them in specific disabling conditions. The ex- before about 9 months of age (Lewis &
pressions are described as related sets or fam- Sullivan, 1988). The common occurrence of
ilies because of the well-established finding open interest reflects the “positivity offset”
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0

Emotional Expressions 123

relaxed, open interest expression are ready to

explore—if only visually—the environment
around them.
Figure 1b shows a form of interest sug-
gesting greater intensity. It is “excited” or
“knit-brow” interest (Camras, 1992; Sullivan
& Lewis, 1989). The major difference from
Fig 1a is the upper face region. The brows
stand out more prominently because the fore-
head muscles have been contracted. They
are pulled together, sometimes slightly as
shown here, sometimes more strongly, mak-
ing central bulges. The eyes appear slightly
narrowed, resulting in crinkling or furrowing
of the outer eye muscles known as crow’s feet.
Infants showing this expression appear to be
intensely interested. They may vocalize while
showing this expression or momentarily be-
come still. They have a concentrated or even
a quizzical look, as if they are actively study-
ing what they are looking at. Knit brow in-
terest is observed during young infants’ so-
cial interactions, marking periods of gazing
at mother or father (Malatesta & Haviland,
1982; Oster, 1978). It persists as much as
10 seconds, thereafter either resolving into a
smile or, in some cases, fussing, if the inter-
action has been overwhelming (Oster, 1978).
This expression has a specific developmen-
tal trajectory (Lewis, Sullivan, & Alessandri,
1990; Malatesta & Haviland, 1982). Seen fre-
quently in young infants, it decreases between
2 and 8 months, appearing again regularly by
10–12 months (Malatesta & Haviland, 1982;
Malatesta, Culver, Tesman & Shipard, 1989;
Sullivan and Lewis, 1988). Knit-brow interest
at these ages typically occurs in situations that
challenge infants or requires problem-solving.
It has also been referred to as a wary face
in studies of response to novel objects. In
such contexts, this expression is coupled with
Fig 1. Three forms of interest: (a) open interest, (b) the inhibition of motor behavior (Bronson,
knit brow interest, and (c) interest with lip roll. 1972; Lewis & Michalson, 1983). Wariness
suggests a state of heightened vigilance or
of the central nervous system (Cacioppo uncertainty. Collectively, the findings sug-
& Gardner, 1999). That is, humans show gest that knit brow interest is a form of
a mild bias toward positive emotion and a highly focused, effortful attention associ-
motivation to approach novel objects, stimuli, ated with active information processing. The
or contexts. Very young infants showing the age change in its occurrence may mark a
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0


developmental shift in the type and quality and with an MA-matched group to observe
of stimulation that infants must “work” to differences in MAX-coded positive, negative,
assimilate. interest (all forms) versus neutral expressions
Figure 1c shows an interest expression that (e.g. Kasari et al, 1992; Yirmiya et al., 1992).
includes one of MAX’s regulatory movements, This work finds that autistic children display
a thinning and rolling inward of one or both primarily neutral and interest expressions
lips. This expression was seen widely some for somewhat longer periods than do other
years ago as the whimsical expression of groups. Increased interest in autistic children
Cabbage Patch@ baby dolls. This lip move- appears at the expense of less enjoyment
ment seems to be regulatory because it fre- expressions (Yirmiya et al., 1989). No pub-
quently suppresses smiling; however, its ex- lished study has examined developmental
act signal value in infants remains unclear. In trajectories or examined differences in
older infants and children however, it tends interest expressions in any atypical popula-
to be combined with or precede gaze aver- tion. Knowledge about the range, distribu-
sion. After about 12 months, coordination tion, and pattern of these expressions may
of this mouth expression with head and eye be of some clinical use in identifying those in
movements signals the child’s awareness of need of further assessment.
the other’s unwanted attention in social situa-
tions. Head lowering and gaze aversion func- SURPRISE
tion to withdraw from or reduce undesired
social interaction, and when combined with The Surprise expression is rarely seen in
this lip movement may signal either embar- young infants, especially in the full form
rassment or shyness toward a social part- shown in Fig 2. In surprise expressions, the
ner, depending on the context and presence brows are raised and prominently arched. The
of other bodily cues (Lewis & Brooks-Gunn, eyes are widened so that the white of the
1979; Lewis, Sullivan, Stanger, & Weiss, 1989). sclera is more evident than in relaxed inter-
After the onset of self-cognition at about 18 est. The mouth gapes with jaw slackened,
months of age, this lip expression may oc-
cur in evaluative settings. When coupled with
body collapse, mouth components suggest-
ing sadness, and/or withdrawal from a task,
it signals shame or evaluative embarrassment
(Lewis, Alessandri, & Sullivan, 1992; Lewis,
Infants with neurological impairment and
various disabling conditions are frequently
reported to have difficulties attending. These
difficulties are often noted during infancy or
are detected as learning disabilities as chil-
dren reach school age. Differences in the form
or patterning of interest expressions may
appear as well and may offer early markers.
Attention, unfortunately, has been studied
almost exclusively as motor behavior, ie,
visual, auditory, or even cardiac orienting to
stimuli, not as facial expressions. Exceptions
are studies of emotion during joint attention
primarily in children with autism and mental
retardation. In this work, children with these
disabilities are compared to each other Fig 2. Surprise in a 5-month-old.
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0

Emotional Expressions 125

assuming an “o” shape, and may be ac- various populations of infants and children
companied by abrupt, momentary stilling likely to be seen in clinical practice. Given its
of other ongoing behavior. The expression relation to arousal modulation in response to
in Fig 2 occurred when a 5-month-old ac- unexpected events and its tendency to habit-
cidentally turned on slides and taped mu- uate with familiarization, information about
sic by tugging on a ribbon attached to her surprise expressions in various groups is of
wrist (Sullivan & Lewis, in press). Surprise considerable interest in assessing children’s
is an appropriate expression in this context emotion regulation and emerging cognitive
of sudden, unexpected exposure to an au- skills.
diovisual event. Yet, investigators have been
stymied in their study of surprise because ENJOYMENT
it does not consistently occur in situations
that adults imagine would surprise babies. Because they are so readily recognized and
A jack-in-the-box does not produce surprise such widely acknowledged milestones of so-
expressions in a majority of babies, for ex- cial behavior, smiling and laughter, the 2
ample. Occasionally, surprise occurs in situ- major facial expressions of enjoyment, have
ations where the experimenter did not an- been the most intensely studied expressions
ticipate observing it (Bennett, Bendersky, & in infants. Their developmental trajectory
Lewis, 2002; Camras, 1992). When surprise and links to cognitive processing in infants
does occur, it appears briefly before resolv- have been extremely well documented in
ing into some other expression—either inter- full-term and preterm infants, as well as in
est, smiling, or a negative expression. Most infants with Down Syndrome, autism, and
researchers accept that infants show at least blindness (Calhoun & Kuczera, 1996; Carvajal
mild surprise expressions to novel events by 6 & Iglesias, 1997; Cicchetti & Sroufe, 1978;
months, and some may do so sooner (Bennett Kasari, Mundy, et al, 1990; Kasari, Sigman,
et al., 2002; Charlesworth & Kreutzer, 1973). et al., 1992; McCall, 1972; Sroufe & Waters,
But, there are wide individual differences in 1976; Sroufe & Wunsch, 1972; Vine, 1973).
whether babies display this expression even While smiling appears universally, cultural
within the standard laboratory situations, sug- and environmental differences also have been
gesting that surprise may occur only in the described (Camras et al., 1998; Gerwirtz,
most emotionally reactive infants. Besides the 1965; Kisilevsky, et al., 1998). Some have ar-
suddenness of the stimulus onset and its in- gued that additional enjoyment forms are de-
tensity, a key factor in surprise seems to be tectable in older infants (Fogel et al., 2000;
whether the stimulus event was expected. For Scanlon-Jones et al., 1990). In fact, the de-
instance, infants who learned that pulling a velopmental course of enjoyment provides a
string turned on a slide with music expressed model for comparison with the development
surprise at this contingent event only when of other expressions, although not all expres-
first learning this response. Once learning sions will follow enjoyment’s pattern.
had occurred, surprise expressions were no Smiling, like interest, is present from the
longer observed, suggesting that the infants opening days of life and is state-dependent
now expected something when they pulled during the neonatal period. Unlike the inter-
(Sullivan & Lewis, 1989). Surprise expressions est expressions, which presume an awake,
in very young infants also appear to grow alert infant, the first signals of enjoyment ap-
more intense across the first several repeated, pear during the newborn’s sleep. Even at this
sudden stimulus presentations. This is less early stage, newborn smiles include the 2
characteristic of surprise in adults, who if sur- most recognizable components of this expres-
prised by a stimulus more than once, rapidly sion: narrowed eyes, widened mouth, with
show an attenuated response. corners raised. These same features consis-
Like interest expressions, information tently appear in the smiles of older infants
about surprise expressions is lacking for (see Fig 3a–c), although later smiles are more
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0


Fig 3. Four examples of enjoyment from 2- to 9-month-olds in various contexts. (a) Response to tactile
teasing at 8 weeks, (b) response to audiovisual contingency at 4.5 months, (c) response to play in an older
infant, and (d) response to audiovisual contingency in an older infant with Down syndrome.
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0

Emotional Expressions 127

intense and may include other components. month-old smile (Ambrose, 1963; Vine, 1973).
Originally thought to be related to digestion Figure 3a shows that gentle tactile stimula-
(older literature may refer to them as “vegeta- tion also successfully elicits enjoyment in an
tive” or “gas” smiles), sleepy smiles are proba- 8-week-old. Such demonstrations show that
bly related to the discharge of pleasant stimu- expressions of enjoyment in early life are
lation of some kind by the infant’s immature closely tied to the physical quality of stimula-
central nervous system (CNS) during rapid tion, which can be auditory, tactile, or visual,
eye movement sleep. The source of this stim- but is usually multisensory in quality. Very
ulation need not be internal, but can be linked soon, however, enjoyment begins to occur to
to the external environment through what- social events.
ever attention and perceptual processing sys- Expressions of enjoyment to social stimula-
tems are functional at birth. For example, one tion increase dramatically by 16 weeks. Social
of the author’s newborn daughter slept in a smiling peaks between 12–14 weeks of age
room with a musical, chiming clock on her in home-reared infants in Western culture. In-
first day at home from the hospital. Two days fants at this age clearly seem to enjoy people
later, when she was again sleeping nearby, the and will smile readily at most adults who in-
clock chimes elicited smiling. This incident teract pleasantly with them. After 16 weeks,
in a healthy newborn shows that the CNS is however, many infants become more discrim-
primed to recognize familiar external auditory inating about whom they will smile for. The
information and to signal this recognition to familiarity and the behavioral style of the in-
others. Repetition alone may be enough to teractive partner become important factors.
promote positive emotional responses to non- In this way, an emotional signal provides an
threatening stimulation through simple asso- index of the infant’s growing social-cognitive
ciative learning (Zajonc, 2001). development.
Between 6 and 8 weeks, smiles of en- In 12- to 17- month-old infants, variants
joyment can be observed during waking to of social smiling have been described (Fogel
both visual and auditory stimulation. Visual et al., 2000; Scanlon-Jones et al., 1990). These
stimuli alone elicit smiling in young infants appear to signal somewhat different quali-
but must have a face-like quality. For ex- ties of enjoyment within social situations but
ample, a gently bobbing oval with 2 small more work is needed to determine if they
black circles, suggesting the eyes of a hu- are actually different expressions or just dif-
man face, are sufficient to make a 2- to 3- fering intensities of enjoyment (see Table 1).

Table 1. Variants of enjoyment

Facial actions codes Context

Closed-mouth, simple smile 0-0-52 CL Social and Toy play (Scalon-Jones, Raag,
(ie, grin) & Collins, 1990)
Enjoyment Readiness to engage, peekaboo game (Fogel,
Nelson-Jones, & Hsu, 2002)
“Duchenne” smile (with 0-33-52 Social interactiona (Fogel et al., 2000;
cheek raise/eye crinkle) Scalon-Jones et al., 1990)
Enjoyment Play with mother, tickle game (Fogel et al., 2002)
Bared-teeth play or 0-0-52/50∗ Social and Toy play (Scalon-Jones et al., 1990)
open-mouth Enjoyment Social Play (Scalon-Jones et al., 1990)
Duplay (play + cheek 0-33-52/50 Game resolution, peek-a-boo and
raise/eye crinkle) Enjoyment tickle (Fogel et al., 2000)

∗The combination of codes 52/50 can be used to designate a smile with jaw drop.
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0


None are treated as unique emotions in ing this expression to one of the 9-month-old
MAX/AFFEX, but are simply coded as enjoy- at play with something novel (Fig 3c), one can
ment. All of these expressions were more fre- see that the crinkled eyes, widened mouth,
quent during interaction with mother than and raised, dimpled cheeks are very similar,
with toys. The play face, however, has special despite the fact that motor and cognitive skills
meaning. have grown considerably in the intervening
5 months. Thus, it seems that the form of
The laughing or play face the enjoyment expression does not change,
Wide-open mouthed, gaping enjoyment ex- only the contexts and qualities of stimula-
pressions occur with characteristic vocaliza- tion required to elicit it. Enjoyment of mas-
tions known as laughter. In intense forms of tery of an object or task becomes especially
this expression, the lips may be rolled in- common after 12 months and has been stud-
ward as the jaw gapes. Either form signals ied as one measure of “mastery motivation”
high excitement and positive arousal. Such ex- (Busch-Rossnagel, 1997). Once the infant is
pressions haven been called “play face” be- capable of self-referential and self-evaluative
cause they appear to be the human equiv- behavior, the integration of enjoyment of
alent of a primate expression of the same mastery and postural and gestural signals
quality (Blurton-Jones, 1972; Preuschoft & van will lead to more elaborated expressions of
Hoof, 1997). The play face makes its appear- pride.
ance by 5 months of age in normally devel- Because of the connection of enjoyment
oping infants (Sroufe & Waters, 1976). It oc- with management of arousal and emerging
curs at first to vigorous, auditory, and tactile cognitive competencies, smiling and laughter
stimulation, ie, tickling as did smiling a few have been the most studied expressions of in-
weeks previously. After 7 months, visual stim- fants and children with disabilities. Much is
ulation becomes more effective in eliciting known, for example, about the emergence of
this expression. By about 12 months, incon- enjoyment in infants with Down syndrome
gruity and novelty, especially if it involves the (DS) (DS; Cicchetti & Sroufe, 1978; Sroufe &
infant’s own participation, will elicit laugh- Wunsch, 1972). As can be seen in Fig 3d, the
ter and the play face, immediately or in an- characteristic expressive components of mas-
ticipation of game-like interaction with a so- tery enjoyment are present in this 8-month-
cial partner, such as peek-a-boo (Fogel et al., old child. Research shows that the general
2000). course of both smiling and the play face is
the same in infants with DS, although delayed
Enjoyment of mastery compared to infants without this syndrome.
During the second half of the first year, in- For example, social smiling peaks at the
fants continue to smile and laugh at nonsocial same mental age (ie, 4–5 months) and smiles
events but infrequently compared to social are directed appropriately to social partners
situations. Smiling occurs in nonsocial con- (Carvajal & Iglesias, 1997). The intensity of
texts especially when infants learn that inter- enjoyment is less, however, and infants with
esting, familiar, and therefore, nonthreaten- DS may not sustain social enjoyment in spon-
ing stimulation is under their own control. taneous interaction to the same degree as do
Figure 3b shows a 4-month-old expressing en- infants without the syndrome and may be less
joyment of a slide and music that she has likely to initiate smiling (Carvajal & Iglesias,
learned to expect when she tugs a ribbon 2000; Kasari & Sigman, 1996). Such differ-
attached to her wrist (Lewis et al., 1990; ences can have important effects on social
Sullivan & Lewis, 1989). This is the beginning interaction. Smiling to auditory stimulation
of playful enjoyment of mastery of the physi- also may be more limited in some children
cal environment that will continue to develop with DS, but can be increased with interven-
as infants learn to play and explore. Compar- tion (Calhoun & Kuczera, 1996). In fact, by
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0

Emotional Expressions 129

preschool age, children with DS are usually dren also fail to share enjoyment with social
reported to be highly positive in their social partners by coordinating their enjoyment
interactions. One study finds however that expressions with eye contact (Dawson
they are uniformly high in expression enjoy- et al., 1990; Sigman et al., 1992). During toy
ment across situations whereas MA-matched play, autistic children are as likely to express
peers displayed more enjoyment only in those enjoyment, a finding that underscores the
situations involving joint attention and affect nonsocial nature of their expressions (Sigman
sharing (Kasari, Mundy, et al., 1990; Kasari, et al, 1992). Finally, autistic children
Sigman, et al., 1992). Thus the child with are likely to display incongruous blends
DS seems to display enjoyment more indis- expressions that combine positive and nega-
criminately, suggesting a need to focus on tive facial elements (Yirmiya et al., 1989). It
emotion regulation in some contexts. Finally, is unclear if this occurs because of problems
enjoyment of mastery may be affected in in regulating expressions themselves, or be-
some children with DS because motor lim- cause of ambivalent or competing emotional
itations restrict early forms of object play, arousal in these children. However, incongru-
but this also can be improved by interven- ous expressions provide unclear, conflicting
tion (Brinker & Lewis, 1982; Sullivan & Lewis, signals to caregivers who rely much on
1990). context or trial and error in responding to
In severe cerebral palsy (CP), enjoyment ex- them.
pressions, especially in the mouth region, may
be less well modulated or appear asymmetri- PHYSICAL PAIN
cal. If motor impairment is severe, the play
face may be the only form of smiling available Some argue that pain is not an emotion.
to these children because subcortical centers Yet, painful stimulation clearly causes a strong
modulating enjoyment are activated. negative emotional response and promotes
The course of smiling in blind children also other negative expressions. The developmen-
has been studied to understand the degree tal course of pain expression has been stud-
of visual input that is necessary to produce ied in some detail because of its theoretical
recognizable expressions of happiness. Visual interest and the very practical need for as-
input does not appear to be necessary for sessment and management of pain in pedi-
blind children to produce recognizable, spon- atric procedures (Grunau, Oberlander, Holsti,
taneous expressions of enjoyment. Instead, it & Whitfield, 1998; Oberlander, 2001). Pain
is their negative expressions that are more expressions can be also observed in situa-
difficult to recognize compared to those of tions of distress that are not physically painful
sighted children (Galati, Sini, Tinit, & Miceli, (Oster, Hegley, & Nagel, 1992). Thus, the pain
2001). However, blind children may be less expression and those that follow it provide
skillful at management of social expressions, clues to emotional and regulatory responses
contributing to the social inhibition that is fre- to all forms of aversive stimulation.
quently reported in these children (Castanho Acute pain in response to tissue damage
& Otta, 1999). Lower rates and poorer qual- during standard pediatric procedures (eg, cir-
ity of enjoyment expressions may prompt the cumcision, heel lance, or inoculation) pro-
perception of inhibition in children with dis- vides a naturalistic and ethical way to observe
abilities generally. how facially and behaviorally expressed pain
Enjoyment expressions in autistic children responses change with the developmental
have been studied around the issue of shared and neurological status of the infant. Pain in
affect between child and adult partners. Less response to such procedures is signaled by
enjoyment is expressed by these children distinctive and intense facial actions includ-
when compared with MA-matched controls ing the drawing together and lowering of the
in semistructured interactions. Autistic chil- brows to create a midbrow bulge, a deepened
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0


Fig 4. Pain in response to DPT inoculation in 4-month-olds showing (a) cupped tongue and (b) lateral
retraction of the mouth.

nasolabial furrow, and tight squeezing of commonly at these ages (see Fig 4b). Since
the eye orbit muscles, resulting in a strong few longitudinal investigations of pain have
squint. From the newborn period through 18 followed infants’ pain expressions from the
months, few changes occur in the pain ex- newborn period, the meaning of these vari-
pression’s appearance or components (Craig, ations and age changes in the mouth compo-
1992; Craig & Grunau, 1993; Izard, Hem- nents of pain are unknown. We do not know if
bree, Dougherty, & Spizziri, 1983; Johnston, they reflect individual differences in pain sen-
Stevens, Craig, & Granan, 1993; Lilley, Craig, sitivity or in regulatory responses to pain.
& Grunau,1997). Throughout this time, the fa- Surprisingly, increased crying in preterms
cial response to acute pain reliably includes and newborns is not a reliable marker for
all of the upper face movements listed. Mouth pain in response to a heel lance (Grunau
movements are more variable but include lat- & Craig, 1987). Young babies vary in their
eral stretching of lips, especially in older in- irritability and many will cry in response
fants and children. In young infants, one of to handling prior to the actual procedure
two common mouth variations can be ob- (Grunau, Johnston, & Craig, 1990; Owens
served. Prominent especially in newborns is & Todt, 1985). Very low birth weight pre-
a dropped jaw with taut or ”cupped” tongue mature infants between 26–31 weeks gesta-
within an angular, wide mouth (see Fig 4a; tional age, show the upper face pain actions
Lilley et al., 1997). From 1 to 5 years, these when their heels are lanced to obtain blood.
facial movements cohere to form pain expres- The upper facial response is specific to the
sions in pediatric patients. A greater number piercing of the skin, rather than other po-
of components shown is related to higher tentially stressful aspects of handling that oc-
pain ratings by clinicians (Gilbert et al., 1999). cur as part of the medical procedure, and is
The laterally stretched mouth also occurs accompanied by the maximum increase in
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0

Emotional Expressions 131

heart rate. Noxious stimulation and the high mission of pain become functional at this
negative arousal they produce appear to si- age. Following this important transition point,
multaneously activate many different nega- 6-month-olds show a shorter duration of pain
tively toned neurological systems in the young response and less of a rise in cortisol in re-
baby. Pain expressions are associated with a sponse to immunizations, suggesting better
rise in cortisol levels in newborns, also sug- internal physiological regulation in response
gesting that heel lancing is a highly stressful to pain (Lewis & Thomas, 1990; Ramsay &
procedure for the young baby (Oberlander, Lewis, 1994).
Gilbert, Chambers, O’Donnell, & Craig, 1999; Although the appearance of the pain ex-
Owens & Todt, 1985; Ramsay & Lewis, 1994). pression may change little with age, older in-
The pain expression and its accompanying fants have more complex responses to pain.
physiological response are related to the de- Typically, they display facial pain for a smaller
velopmental age of the preterms, appearing proportion of time prior to quieting, dis-
more consistent and robust in older babies playing anger and blended expressions in-
(Johnston, Stevens, Yang, & Horton, 1995). stead (Izard et al., 1983; Izard, Hembree,
It is unclear if this age change reflects bet- & Heubner, 1987). By 18 months, pain-
ter neurological regulation of the pain re- specific expressions comprised only 10% of
sponse, or the gradual recovery from illness the postinoculation distress (Izard et al.,
and trauma experienced by many of these sick 1987). Thus, anger and other negative ex-
babies. pressions become rapid after-reactions to the
Although the expression of pain appears initial pain response. This pain after-reaction
relatively invariant over the first 2 years of is most likely to influence the appropri-
life, a number of important changes occur ate soothing strategy and might also be sta-
that possibly reflect a combination of neuro- ble across individuals. For example, these
physiological maturation, life experience, and postpain facial signals likely reflect some
a growing ability to remember prior painful combination of differences in pain sensitiv-
experiences. In contrast to newborns, in ity and social experience among individuals.
2-month-olds, the expressive components of For, example, Japanese infants seem to have
pain occur at low frequency during a preinoc- less pain sensitivity and qualitatively different
ulation or baseline period. This observation emotional responses to inoculation than do
suggests that pain expressions, occurring as American infants; pain expressions are less
nonspecific generalized distress reactions to intense and are not typically followed by
handling, decrease with age. However, be- anger and crying, but by surprise (Lewis,
cause a significant and dramatic increase in Ramsay, & Kawakami, 1993). Differences be-
all pain components is observed in response tween Chinese infants and other groups are
to inoculation, infants’ pain response shows even more pronounced for a variety of expres-
some specificity to skin trauma at every age sions (Camras et al., 1998).
studied (Lilley et al., 1997). Healthy term in- Greater behavioral coordination and mat-
fants between 2 and 4 months of age have uration of motor skills also lead to elabora-
the most robust response to pain (Lewis & tion of responses to pain in older infants.
Thomas, 1990; Maikler, 1991). By 4 months, Infants over 12 months of age are more
pain expressions are highly specific to in- likely to show goal-focused behaviors fol-
oculation, with very few pain signals occur- lowing immunizations (withdrawing, avoid-
ring during the baseline period. Likewise, ing), compared to younger infants who dis-
4 month-olds have quicker recovery from play more unorganized distress responses
pain (Lilley et al., 1997; Lewis & Thomas, (Craig & Grunau, 1993). Between 12 and 18
1990; Ramsay & Lewis, 1994), suggesting months, pain responses become anticipatory.
that CNS mechanisms inhibiting the trans- Lilley et al (1997) reported that baseline
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0


pain expressions were low between 4 and expressed appear to index increasing concen-
12 months, but rose again between 12 and trations of the solutions presented (Granchow
18 months to a level similar to that of 2-month- et al., 1983; Rosenstein & Oster, 1988; Steiner,
olds! These differences in the baseline or pre- 1979). The response to a bitter taste typically
inoculation period suggest that the older in- involves turning away from the source of the
fants anticipated the painful procedures in stimulus and gaping of the mouth and lower
store for them. If so, pediatricians may be- lip, as shown in Fig 5a. Sometimes, the gag
gin to encounter anticipatory emotional re- reflex is observed (Steiner, 1979). Nose wrin-
sponses to pain, cued by the environment kling and upper lip raising are components of
and salient memories of prior procedures by the full disgust reaction, but occur alone as
12 months, perhaps even earlier. Collectively, observed in about half of the newborns as a
these changes suggest that the pain expres- less intense form of the reaction (Rosenstein
sion is stable over the first 2 years of life and & Oster, 1988). A flattened tongue and drool-
there is a close link between facially signaled ing may also be visible, depending on the state
pain and neurophysiological development. of the child and the concentration of the fluid
Significant neurological impairment ap- (Steiner, 1979). These responses signal strong
pears to dampen the pain response, but the distaste on the part of the infant. They are a
basis for dampened or atypical pain responses defensive reflex helping the infant to rid itself
is unclear (Oberlander et al., 1999). Signif- of the unappealing taste. Similar responses are
icant neurological, motor, and/or cognitive observed to odors that most adults would find
impairment can produce a decreased pain objectionable, such as fishy and “rotten egg”
response for any number of reasons, lead- smells (Steiner, 1979).
ing to underrecognition and poor manage- Responses to a sour taste, such as citric
ment of pain in children with disabilities or acid solution, are more variable, seem to be
neurological impairment. The development milder, and evolve over several seconds in
of expression-based pain scales for pediatric contrast to disgust. Lip pursing is the com-
practice is a relatively recent application di- monly observed initial response, accompa-
rected toward addressing this problem and nied by nose wrinkling, narrowed eyes, and
discovering methods to assess pain based on blinking (Rosenstein & Oster, 1988; Steiner,
facial responses (Oberlander, 2001). Hope- 1979). Figure 5b shows this response in a 4-
fully, better quantification of pain will lead to month-old. A closed mouth, either with cor-
better recognition and effective management. ners down or retracted lips, is a feature as-
sociated with milder aversive reactions. This
DISGUST AND ITS VARIANTS “sour grimace,” signaling dislike, occurs as the
response develops (Granchow et al., 1983;
The disgust expression is another that has Steiner, 1979).
been studied in some detail in newborns. The The variability of sour expressions in re-
newborn’s response to bitter and sour sub- sponse to sour tastes is even more pro-
stances is distinct from responses to water nounced in 4-month-olds than in newborns.
and sweet solutions (Granchow, Steiner, & The most common response to a natural
Daher, 1983; Rosenstein & Oster, 1988; sour taste (lemon juice) at this age is in-
Steiner, 1979). Saltiness does not trigger dis- terest, following the initial puckering and
gust expressions, with less than half of in- rapid lip movements associated with tasting
fants showing any negative response to it or mouthing of the flavor. In some infants,
(Rosenstein & Oster, 1988). However, quinine however, negative expressions indicating dis-
and other bitter tastes are potent and rapid like (sad frown or lip retraction movements)
elicitors of disgust responses regardless of in- appear, as observed in newborns (see Fig 5c).
fant state. The intensity of the disgust expres- However, a few infants actually smile, the next
sions as well as the pattern of components most common response to this taste (Bennett,
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0

Emotional Expressions 133

Fig 5. Disgust (a) and sour (b, c) responses of 4-month-old infants elicited by a lemon juice swab.

et al, 2000). This finding shows that early in- anencephaly, the congenitally blind, children
dividual differences in taste preferences and with craniofacial malformations, and mentally
rapid resolution of initial mild disgust reac- retarded adults. Disgust and sour grimace ex-
tions into either enjoyment or dislike emerge pressions were recognizable in all popula-
early in life. tions studied, despite the considerable vari-
The developmental course of the disgust ation in the cognitive and motor control
family of expressions has not been traced (Steiner, 1979). He argued that the preser-
further, but it appears that the disgust re- vation of disgust responses across this wide
sponse is so well-developed in infants and range of cognitive and motor functioning
specific to both the quality of olfactory and means that disgust expressions are controlled
gustatory stimulation, that its form changes by the brainstem and so undergoes little mor-
little. Steiner (1979) undertook extensive phological change. However, the emergence
study of disgust and enjoyment taste re- of the positive or negative reactions to sour
actions in various populations of infants also points to 4 months as an important
and children with various forms of disabil- developmental milestone in organization of
ity, including newborns with hydrocephaly, expressions.
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0


Perhaps the most interesting aspect of dis- ANGER OR ‘‘CRY FACE” EXPRESSIONS
gust expressions, and the one about which
little is known, is how the subtle variations The prototypic anger or cry face appears
in responsiveness observed even in newborns in Fig 6a. It is characterized by brows drawn
are related to individual differences in ner- together and lowered, sometimes strongly as
vous system functioning or other factors, such seen in this view. Deep nasolabial folds frame
as temperament. A second unanswered ques- a wide-opened square mouth. Anger expres-
tion is when these expressions begin to occur sions of this intensity are almost always ac-
in response to non-taste stimuli, and espe- companied by a rolling cry in infants, but
cially social stimulation. Like smiling, dis- milder versions are observed without any vo-
gust expressions are elicited initially to phys- calization. This expression is the infant’s most
ical rather than social-psychological elicitors. common negative expression.
Non-taste stimuli that might be sufficient to A major quandary for theorists is that the
produce disgust in infants have not yet been anger/cry face rarely occurs in young infants
reported, but can be imagined. Disgust sig- without the co-occurrence of other negative
nals stimulus rejection. Therefore, too rapid expressions or blends. Blended anger and sad-
or sudden occurrence of a stimulus might ness, as seen in Fig 5b, is commonly seen. In
elicit disgust expressions if the stimulus over- this expression, knit, lowered brows and nar-
whelms infants’ ability to process it. We have rowed eyes of the upper face region are ac-
observed this on some occasions in the infant companied by a sad ”pout” in the lower face.
learning lab: sometimes an infant pulls suffi- This particular blend (upper face anger with
ciently rapidly that the slide and music ap- lower face sadness) is often observed during
pears within a second of its previous appear- social interaction with the mother and it tends
ance and before the infant’s reaction to the to decrease with age in this setting (Izard
first appearance has subsided. When this oc- et al., 1995). This might not be so in all situa-
curs, the infant may display nose wrinkling, or tions that recruit anger expressions, but more
asymmetrical mouth and lip movements in re- information is needed. The co-occurrence of
sponse, suggesting a kind of recoil reaction to multiple negative and blended expressions
the overwhelming, too rapid reoccurrence of with the anger/cry face has led some to argue
the stimulus. While clearly not yet an instance that these negative expressions collectively re-
of “psychological disgust,” such observations flect general distress or unhappiness instead
suggest that prior to 6 months of age, compo- of a specific facial signal of anger or other
nents of disgust may occur in response to vi- negative emotions (Camras, 1992). However,
sual and auditory stimulation, setting the stage the co-occurrence of multiple negative ex-
for their later function in social situations. pressions in the early months can also be
But, the forms of disgust observed in this case explained by the well-known quality of the
do not involve the intense, gaping reactions CNS to respond more intensely to negative
observed in response to bitter tastes. than to positive stimuli (Peters & Czapinski,
Toilet training and introduction of solid 1990). Although the CNS idles in a mildly pos-
foods when the child is able to eat inde- itive mode (positivity offset), when a threaten-
pendently are thought to be important con- ing stimulus is encountered, a vigorous nega-
texts for parent-child socialization of disgust tive response is observed. This phenomenon,
(Rozen, Hadt, & McCauley, 2000). It is also known as the negativity bias, may be espe-
possible that disgust, signaling stimulus rejec- cially evident in young infants, in whom in-
tion, is more common in neurologically dam- hibitory control is still limited and arousal is
aged infants and children in whom inhibitory poorly regulated. The negativity bias could
control is slow to develop or absent. This re- result in an initially unmodulated negative
sponse would be an important clue to care- response that simultaneously activates several
givers to reduce the intensity or pacing of competing systems when an aversive event
stimulation. is initially perceived. It is not that there is
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0

Emotional Expressions 135

Fig 6. Anger (a) and upper face anger blended with sad (b) expressions in 5-month-olds.

one “undifferentiated” distress state, but that that action abruptly fails to produce the event.
all negative emotion systems are primed. The Anger expressions increase, but other nega-
strongest response tendency is an energiz- tive expressions do not. The increase is there-
ing response to action in order to resist or fore specific to infants’ having learned a re-
overcome the aversive stimulus. It is signaled lation between action and outcome (Lewis
by increased motor activity and crying as et al., 1990; Sullivan & Lewis, in press). Situa-
well as by the prototypical anger or cry face tions that produce anger expressions before
expressions. 6 months of age are those in which ac-
Although anger expressions may not ap- cess to desired objects or goals is blocked
pear exclusively when infants encounter cer- or thwarted in some way (Bennett et al.,
tain types of negative events, by 3–4 months 2002; Lewis et al., 1990; Stenberg et al., 1983;
of age, a number of situations seem to re- Sullivan et al, 1992) or, those causing pain af-
sult in anger expressions predominantly. Re- ter about 4 months (Izard et al., 1983). There
straining an infant’s arms, for example, pro- is stability in individuals’ expressions of anger
duces a variety of facial expressions, not in these contexts as well (Izard et al., 1995;
all of them negative. Anger, however, is the Sullivan et al, 1992).
predominant negative expression (Bennett Crying and fussing are widely recognized
et al., 2002; Braungart-Reiker & Stifter, 1996; vocal signals of negative emotion in infants,
Stenberg, Campos, & Emde, 1983; but see but no any one quality of crying appears
Camras et al., 1998, regarding ethnic differ- to be specific to the “cry face” or to any
ences). An even more powerful example is other negative expression. Negative facial
the response of infants to loss of a contingent expressions can precede negative vocal be-
event (Lewis et al, 1990; Sullivan, Lewis, & havior, suggesting that greater or increas-
Alessandri, 1992). Infants who learn to con- ing arousal is needed for vocal crying. Cries
trol a pleasant event by pulling a ribbon show can be differentiated by frequency and tonal
anger expressions and increase pulling when quality but it is difficult for observers to
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0


discriminate these features reliably and expe- abling conditions significantly affect adult per-
rience is an important factor when differen- ceptions of young children. For example, in-
tiating vocal crying in the mild to moderate fants and children with DS are reportedly less
range (Papousek, 1989). Facial expressions irritable than the norm and those with CP and
probably provide more graded information auditory and visual handicaps even more so
about the onset, quality, and intensity of neg- (Field, 1996). Autistic children are also some-
ative reactivity, making such discriminations times found to be more negative (Yirmiya
possible. et al., 1989). Greater attention to the quality of
The form, development, and regulation of expressive behavior of children with disabil-
anger expressions and their relation to vo- ities might provide more clinically useful in-
cal crying in atypical development has not formation and suggest possible interventions.
been addressed in any significant detail. Like Given their physical limitations, it is reason-
smiling, negative vocalizations undergo an im- able to expect greater frustration on the part
portant developmental transition in the first of these children. Whether it is expressed as
4 months (Hopkins, 2001). Presumably this anger, other negative expressions, or undiffer-
change parallels changes in the patterning entiated distress is not known
of anger and other negative expressions, al-
though this has not been studied. Much avail- SAD OR ‘‘POUT FACE” EXPRESSIONS
able work has focused on the clinical utility
of vocal cry behavior and unexplained infant Figure 7a shows the sad frown or “pout
irritability (Barr, Hopkins, & Green, 2000). Be- face” expression. The brows are raised and an-
cause adults rely on vocal and facial informa- gular in appearance over narrowed eyes. The
tion before correctly identifying that infants nasolabial folds appear prominent, as they
are indeed upset (Greene & Gustafson, 2001), do in all negative expressions. The mouth
the lack of information about the coordination corners are down-turned in a “horseshoe”
of expression and voice in the development of shape (Oster, 1978). The chin is raised, some-
typical and atypical children is frustrating. Dis- times prominently as in Fig 7b. The latter

Fig 7. Sad expressions without (a) and with (b) lower lip protrusion in infants under 6 months of age.
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0

Emotional Expressions 137

movement, especially when combined with approach of a stranger, highly novel masks,
forward projection of the lower lip, as in and startling mechanical toys) do not elicit
Fig 7b, gives it the pouting quality. In both fear expressions (Bennett et al., 2002; Camras,
examples of this expression, the mouth is 1992; Izard et al., 1995; Lewis & Michalson,
closed, although open-mouthed sad pouts are 1983). Therefore, some suggest that infants
also possible. Sad pouts are equally likely to may not be able to display fear until some-
appear as an upper face signal or as a full ex- time after 7–12 months because greater expe-
pression in young infants (Izard et al., 1995; rience and cognitive abilities are needed for
Sullivan & Lewis, in press). When blended appraisal of strange and dangerous environ-
with another expression, sad brows are espe- mental events (Izard & Malatesta, 1987). Both
cially likely to combine with anger or, less of- independent locomotion and ability to inhibit
ten, with interest in the lower face. In either behavior may also be prerequisites of fear. For
case, such blends are interpreted as a signal of example, wariness of strangers, a mild form of
dislike or unhappiness. fear, is not observed in a majority of children
Sad pouts have never been observed to until after 7 months of age (Bronson, 1972).
be specific to any stimulus or context. For By 11 months, the brow components of fear
this reason, little is known about them. They do occur to a variety of aversive stimuli, but
are not a dominant negative expression in are not specific to fear-inducing toys, such as
any context that has been studied, except for a growling mechanical gorilla (Camras et al.,
pouts associated with the after-sour response 1998). Fear systems in the brain do have a
(Bennett et al., 2002). Sad pouts occur at rela- high degree of plasticity (LeDoux & Phelps,
tively low but stable levels during both social 2000), supporting the view that learning and
interaction and learning/frustration episodes experience are very important to this emo-
(Izard et al., 1995; Sullivan et al., 1992). tion. The question remains how little experi-
Some have suggested that this expression ence is needed for fear expressions to occur.
occurs as an infant or young child either Figure 8a shows the response of a 6-month-
arouses to an anger expression or as a reg- old infant to the return of a nurse for the
ulatory movement inhibiting anger (Camras, second of 2 inoculations administered at this
1992; Oster, 1978). It is possible that sad well-child visit. This observation suggests that
expressions are never completely indepen- a single salient experience may be suffi-
dent of anger expressions even in adults cient. By 2 years of age, fear can be rapidly
(Barr-Zisowitz, 2000). Caregiver intervention learned, resulting in phobic reactions that
in response to sad expressions may forestall may spread readily to previously nonfeared
or more effectively help dampen more in- objects through learning (Watson & Rayner,
tense negative response, but this has not been 1920).
demonstrated. Even in young infants, fear blends are ob-
served on occasion in situations that are aver-
FEAR EXPRESSIONS sive. For example, fear blends occurred in
8% of infants in response to arm restraint
Fear expressions are notoriously difficult (Camras, Oster, Campos, Miyake, & Bradshaw,
to observe in infants. In adults and children, 1997). This raises the possibility that most
this expression involves raised and straight- stimulus situations studied thus far are not
ened brows, widened eyes with tense lower adequate to elicit fear expressions and that
eyelids, and horizontally retracted lips. Less blends are observed because milder stimula-
intense versions feature brow movements tion results in less intense reactions. As with
alone, possibly blended with interest or other disgust, very specific and intense stimulation
facial movements in the lower face, such as may be necessary to elicit fear expressions.
anger or sadness. Situations that could con- Looming objects, very loud sounds, and the
ceivably frighten infants (eg, a visual cliff, the loss of support are good candidates. However,
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0


Fig 8. A fear expression in response to approach of a nurse following inoculation (a) and a fear blend in
response to the sudden appearance of a stimulus (b).

some of these are impossible to study exper- fuss and are unable to complete the proce-
imentally because of ethical considerations. dure. These infants are reported by mothers
Alternatively, aversive situations may produce to have more fearful temperaments (Sullivan,
fear only in the most temperamentally fearful Ramsay, & Lewis, 1992). It seems reasonable
children. In our studies of infant condition- to infer that infants with fearful temperaments
ing, fear blends appear as negatively toned sur- may be expressing fear.
prise expressions (see Fig 7b). These surprise- Because so few negative situations elicit
fear blends appear early in the conditioning fear and the percentage of children for whom
session, before infants realize that their own fear expressions have been observed is so
movements cause the sudden onset of slides low (Bennett et al., 2002), the developmen-
and music. While most infants are surprised tal course of fear expressions has not been
and interested in these first appearances of studied extensively. The dependence of fear
slides and music, others show surprise reac- on cognitive skills and appraisal of danger
tions that grow more extreme across several in the environment may mean that most
presentations, ultimately developing a nega- infants must learn to fear. Hence, fear ex-
tive quality including fear components (brow pressions will vary greatly across individ-
straightening, tensed lower eyelid, and/or lat- uals. Some infants may express excessive
eral movement of the mouth corners). In most amounts of fear, either because neurologi-
cases, fear-surprise responses subside quickly cal damage has interfered with the normal
as infants learn the relation between their buffering of the fear expression, or because
own movements and the slide’s appearance prolonged experience with invasive medical
(Lewis, Sullivan, & Michalson, 1984; Sullivan procedures have resulted in conditioning of
& Lewis, 1989). But, some infants begin to fearfulness.
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0

Emotional Expressions 139

SUMMARY it seems clear that facial expressions provide

important information to caregivers and prac-
Despite differing approaches to measure- titioners alike. Expressions, especially when
ment, considerable evidence shows that combined with vocal and postural behaviors,
context-appropriate emotion expressions oc- provide important clues to the motivational
cur in infants and young children from birth state of infants who cannot otherwise re-
or shortly thereafter. No matter what sys- port what they feel. Ability to read the facial
tem is used, interest, enjoyment, anger, pain, behavior of children with disabilities gives
and disgust have been shown to signal be- parents an important tool for promoting posi-
havioral dispositions congruent with these tive emotion, better management of their chil-
emotions in young infants. This is also true dren’s negative behavior, and preventing frus-
for fear and sad expressions in older infants. tration or helpless withdrawal of the parents.
Pain, enjoyment, interest, and possibly anger It is also clear that social environments have
and disgust show little change over the first substantial influence on the emotional be-
2 years. Instead, the contexts that elicit haviors of children with disabilities as well
these emotions shift at several major transi- (Field, 1996), suggesting that well-designed
tions in cognitive development and cerebral interventions may help parents who have dif-
maturation, ie, 4 months, 7–9 months, and ficulty in understanding their children’s ex-
18–24 months. At these developmental ages, pression become better attuned to subtle, in-
expressions can be particularly informative congruous, or ambiguous signals. Awareness
about the developmental status of children. of these expressions and their developmen-
At all ages, individual differences in expres- tal trajectories by the practitioner can as-
sive behavior can provide clues to children’s sist in assigning meaning to infant behavior,
preferences and style of response. Although skills that parents are especially concerned
a number of interesting and important ques- about in day to day commerce with their
tions remain about expressive development, children.


Abe, J. A., & Izard, C. (1999). A longitudinal study of emo- Brinker, R. P., & Lewis, M. (1982). Discovering the com-
tion expression and personality relations in early de- petent handicapped infant: A process approach to as-
velopment. Journal of Personality and Social Psy- sessment and intervention. Topics in Early Childhood
chology, 77, 566–577. Special Education, 2(2), 1–16.
Ambrose, J. A. (1963). The development of the smiling Bronson, G. W. (1972). Infants’ reactions to unfamiliar
response in infancy. In B. Foss (Ed.), Determinants persons and novel objects. Monographs of the Soci-
of infant behavior (Vol. 1, pp. 179–196). New York: ety of Research in Child Development, 37(Serial No.
Wiley. 148).
Barr, R. G., Hopkins, B., & Green, J. A. (Eds.). (2000). Cry- Busch-Rossnagel, N. (1997). Mastery motivation in tod-
ing as a sign, a symptom, & a signal. London: MacK- dlers. Infants and Young Children, 9(4), 1–11.
eith Press. Cacioppo, J. T., & Gardner, W. L. (1999). Emotion. Annual
Barr-Zisowitz, C. (2000). “Sadness”—Is there such a Review of Psychology, 50, 191–214.
thing? In M. Lewis & J. M. Haviland-Jones (Eds.), The Calhoun, M., & Kuczera, M. (1996). Increasing social
handbook of emotions (2nd ed., pp. 607–622). New smiles of young children with disabilities. Perceptual
York: Guilford. and Motor Skills, 82, 1265–1266.
Bennett, D., Bendersky, M., & Lewis, M. (2002). Facial ex- Camras, L. (1992). Expressive development and basic
pressivity at 4 months: A context by expression analy- emotions. Cognition and Emotion, 6, 269–283.
sis. Infancy, 3, 97–114. Camras, L., Holland, E. A., & Patterson, M. J. (1993). Fa-
Blurton-Jones, J. (1972). Ethological studies of child be- cial expressions. In M. Lewis & J. M. Haviland (Eds.),
havior. London: Cambridge University Press. Handbook of emotions (pp. 199–208). New York:
Braungart-Reiker, J. M., & Stifter, C. E. (1996). Infants’ re- Guildford.
sponses to frustrating situations: Change and continu- Camras, L., Oster, H., Campos, J., Campos, R., Ujiie, T.,
ity in reactivity and regulation. Child Development, Miyake, K., et al. (1998). Production of emotional fa-
67, 1767–1779. cial expression in European American, Japanese, and
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0


Chinese infants. Developmental Psychology, 34, 616– Gilbert, C., Lilley, C., Craig, K., McGrath, P., Court, C.,
628. Bennett, S., et al. (1999). Postoperative pain expres-
Camras, L., Oster, H., Campos, J, Miyake, K., & Bradshaw, sion in preschool children: Validation of the child fa-
D. (1997). Japanese and American infants’ responses cial coding system. The Clinical Journal of Pain, 15,
to arm restraint. In P. Ekman & L. Erika (Eds.), What 192–200.
the face reveals: Basic and applied studies of sponta- Greene, J. A., & Gustafson, G. (2001). Crying: Multiple
neous expression using the Facial Action Coding Sys- determinants of perceived meaning. In R. Barr, I. St.
tem. Series in Affective Science (pp. 289–301). New James-Roberts, & M. Keefe (Eds.), New evidence on
York: Oxford University Press. unexplained early infant crying: Its origins, nature
Carvajal, F., & Iglesias, J. (1997). Mother and infant smil- and management (pp. 105–120). Skillman, NJ: John-
ing exchanges during face-to-face interaction in infants son & Johnson Pediatric Institute.
with and without Down Syndrome. Developmental Grunau, R., & Craig, K. (1987). Pain expressions in
Psychobiology, 31, 277–286. neonates: Facial action and cry. Pain, 28, 395–
Carvajal, F., & Iglesias, J. (2000). Looking behavior and 410.
smiling in Down Syndrome infants. Journal of Non- Grunau, R., Johnston, C. C., & Craig, K. (1990). Neonatal
verbal Behavior, 24, 225–236. facial and cry responses to invasive and noninvasive
Castanho, A. P., & Otta, E. (1999). Decoding spontaneous procedures. Pain, 76, 277–286.
and posed smiles of children who are visually im- Grunau, R. E., Oberlander, T., Holsti, L., & Whitfield,
paired and sighted. Journal of Visual Impairment M. F. (1998). Bedside application of the Neonatal Fa-
and Blindness, 93, 659–665. cial Coding System in pain assessment of premature
Charlesworth, W. R., & Kreutzer, M. (1973). Facial expres- neonates. Pain, 76, 277–286.
sions of infants and children. In P. Ekman (Ed.), Dar- Haviland, J. M. (1983). Looking smart. In M. Lewis (Ed.),
win and facial expression: A century of research in Origins of Intelligence (2nd ed., pp. 423–450). New
review. New York: Academic Press. York: Plenum.
Cicchetti, D., & Sroufe, L. (1978). The relationship be- Hopkins, B. (2001). Development of crying: The origin
tween affective and cognitive development in Down and change problem. In R. Barr, I. St. James-Roberts, &
Syndrome infants. Child Development, 47, 920–929. M. Keefe (Eds.), New evidence on unexplained early
Craig, K. D. (1992). The facial display of pain: Better than infant crying: Its origins, nature and management
a thousand words. American Pain Journal, 1, 153– (pp. 71–86). Skillman, NJ: Johnson & Johnson Pedi-
162. atric Institute.
Craig, K. D., & Grunau, R. V. (1993). Neonatal pain per- Izard. C. (1982). A system for identifying affect expres-
ception and pain measurement. In K. Anand & P. Mc- sions by holistic judgements (AFFEX). Newark, DE:
Grath (Eds.), Pain in neonates (pp. 67–105). Amster- Instructional Resources Center.
dam: Elsevier. Izard, C. (1983/1995). The maximally discriminative fa-
Dawson, G., Hill, D., Spencer, A. Galpert, L., & Waton, cial movement coding system (MAX-revised edition).
L. (1990). Affective exchanges between young autis- Newark, DE: Instructional Resources Center.
tic children and their mothers. Journal of Abnormal Izard, C., Fantauzzo, C. A., Castle, O. , Haynes, M., Rayias,
Child Psychology, 10, 335–345. M. F., & Putnam, P. (1995). The ontogeny and sig-
Field, T. (1996). Expressivity in physically and emotionally nificance of infants’ facial expressions in the first 9
handicapped children. In M. Lewis & M. W. Sullivan months of life. Developmental Psychology, 31, 997–
(Eds.), Emotional development in atypical children 1013.
(pp. 1–28). Mahwah, NJ: Erlbaum. Izard, C., Hembree, E. A., Dougherty, L. M., & Spizziri, C.
Fogel, A., Nelson-Goens, C., & Hsu, H. (2000). Do differ- L. (1983). Changes in facial expressions of 2- to 19-
ent infant smiles reflect different positive emotions? month-old infants following acute pain. Developmen-
Social Development, 9, 497–520. tal Psychology, 19, 418–426.
Galati, D., Sini, B., Tinit, C., & Miceli, R. (2001). Facial Izard, C., Hembree, E. A., & Heubner, R. B. (1987). In-
expressions of emotion in congenitally blind children fants’ emotion expressions to acute pain: Develop-
from 8–11 years of age: Observing and encoding the mental change and stability of individual differences.
expression of emotions (Italian). Research in Psychol- Developmental Psychology, 23, 105–113.
ogy, 23, 103–126. Izard, C., & Malatesta, C. (1987). Perspectives on emo-
Ganchrow, J. R., Steiner, J. E., & Daher, M. (1983). Neona- tional development. In J. D. Osofsky (Ed.), Handbook
tal facial expression in response to different qualities of infant development (2nd ed., pp. 494–554). New
and intensities of gustatory stimuli. Infant Behavior York: Wiley.
and Development, 6, 473–484. Johnston, C. C., Stevens, B., Craig, K., & Granau, R. K.
Gerwirtz, J. L. (1965). The course of infant smiling in four (1993). Developmental changes in pain expressions in
child-rearing environments in Israel. In B. Foss (Ed.), premature, full-term, 2- and 4- month-old infants. Pain,
Determinants of infant behavior (Vol. 3, pp. 205– 52, 201–208.
249). London: Methuen. Johnston, C. C., Stevens, B., Yang, F., & Horton, L.
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0

Emotional Expressions 141

(1995). Differential response to pain by very prema- in response to inoculation. Child Development, 61,
ture neonates. Pain, 61, 471–479. 50–59.
Kasari, C., Mundy, P., Yirmiya, N., & Sigman, M. (1990). Af- Lilley, C. M., Craig, K., & Grunau, R. E. (1997). The ex-
fect and attention in children with Down Syndrome. pression of pain in infants and toddlers: Developmen-
American Journal of Mental Retardation, 95, 55– tal changes in facial action. Pain, 72, 161–170.
67. Maikler, V. E. (1991). Effects of a skin refriger-
Kasari, C., & Sigman, M. (1996). Expression and under- ant/anesthetic and age on the pain responses of
standing of emotion in atypical development: Autism infants receiving immunizations. Research in Nurs-
and Down Syndrome. In M. Lewis & M. W. Sullivan ing and Health, 14, 397–403.
(Eds.), Emotional development in atypical children Malatesta, C. Z., Culver, C., Tesman, J. R., & Shepard, B.
(pp. 109–130). Mahwah, NJ: Erlbaum. (1989). The development of emotion expression dur-
Kasari, D., Sigman, M., Mundy, P., & Yirmiya, N. (1992). ing the first two years of life. Monographs of the So-
Affective sharing in the context of joint attention inter- ciety for Research in Child Development, 54 (Serial
actions of normal, autistic, and mentally-retarded chil- Nos. 1–2).
dren. Malatesta, C. Z., & Haviland, J. M. (1982). Learning display
Kisilevsky, B., Hains, S., Lee, K., Muir, D., Xu, F., Fu, G., rules: The socialization of affect expression in infancy.
et al. (1998). The still-face effect in Chinese and Cana- Child Development, 53, 991–1003.
dian 3- to- 6-month-old infants. Developmental Psy- McCall, R. B. (1972). Smiling and vocalization in infants
chology, 34, 629–639. as indices of perceptual-cognitive progress. Merrill-
LeDoux, J. E., & Phelps, E. A. (2000). Emotion networks Palmer, 18, 341–347.
in the brain. In M. Lewis & J. M. Haviland-Jones (Eds.), Mundy, P., & Willoughby, J. (1996). Nonverbal commu-
Handbook of Emotions (2nd ed., pp. 157–172). New nication, joint attention and early socioemotional de-
York: Guilford. velopment. In M. Lewis & M. W. Sullivan (Eds.), Emo-
Lewis, M. (2000). The self-conscious emotions: Embar- tional development in atypical children (pp. 65–88).
rassment, shame, pride and guilt. In M. Lewis & J. Mahwah, NJ: Erlbaum.
M. Haviland-Jones (Eds.), The handbook of emotions Oberlander, T. F. (2001). Pain assessment and manage-
(2nd ed., pp. 623–636). New York: Guilford. ment in infants and young children. Infants and
Lewis, M., Alessandri, S., & Sullivan, M.W. (1992). Differ- Young Children, 14, 33–47.
ences in shame and pride as a function of children’s Oberlander, T. F., Gilbert, G., Chambers, C. T., O’Donnell,
gender and task difficulty. Child Development, 63, M. E., & Craig, K. (1999). Biobehavioral responses to
630–638. acute pain in adolescents with a significant neurolog-
Lewis, M., & Brooks-Gunn, J. (1979). Social cognition ical impairment. The Clinical Journal of Pain, 15,
and the acquisition of self. New York: Plenum. 201–209.
Lewis, M., & Michalson, L. (1983). Children’s emotions Oster, H. (1978). Facial expression and affect devel-
and moods: Developmental theory and measure- opment. In M. Lewis & L. Rosenblum (Eds.), The
ment. New York: Plenum. development of affect (pp. 43–75). New York:
Lewis, M., Ramsay, D. S., & Kawakami, K. (1993). Dif- Plenum.
ferences between Japanese infants and Caucasian Oster, H., Hegley, D., & Nagel, L. (1992). Adult judge-
American infants in behavioral and cortisol re- ments and fine-grained analysis of infant facial ex-
sponse to inoculation. Child Development, 64, 1722– pressions: Testing the validity of a priori coding
1731. formulas. Developmental Psychology, 28, 1115–
Lewis, M., Sullivan, M. W., & Alessandri, S. M. (1990). Vi- 1131.
olation of expectancy and frustration in early infancy: Owens, M., & Todt, E. H. (1985). Pain in infancy: Neonatal
The effect of loss of control. Developmental Psychol- reaction to a heel lance. Pain, 20, 77–86.
ogy, 26(5), 744–753. Papousek, M. (1989). Determinants of responsiveness to
Lewis, M., Sullivan, M., & Brooks-Gunn, J. (1985). Emo- infant vocal expression of emotional state. Infant be-
tional behaviour during the learning of a contingency havior and development, 12, 507–524.
in early infancy. British Journal of Developmental Peters, G., & Czapinski, J. (1990). Positive-negative asym-
Psychology, 3, 307–316. metry in evaluations: The distinction between affec-
Lewis, M., Sullivan, M. W., & Michalson, L. (1984). The tive and informational negativity effects. European
cognitive emotional fugue. In C. Izard, J. Kagen, & Review of Psychology, 1, 33–60.
R. Zajonc (Eds.), Emotions, cognition, and behav- Preuschoft, S., & van Hoof, Jan A. R. A. (1997). The social
ior (pp. 264–288). New York: Cambridge University function of “smile” and “laughter”: Variations across
Press. primate species and societies. In U. Segerstrale & P.
Lewis, M., Sullivan, M. W., Stanger, C. & Weiss, M. (1989). Monar (Eds.), Nonverbal communication: Where na-
Self-development and self-conscious emotions. Child ture meets culture (pp. 171–190). Mahwah, NJ: Erl-
Development, 60, 146–156. baum.
Lewis, M., & Thomas, D. (1990). Cortisol release in infants Ramsay, D. S., & Lewis, M. (1994). Developmental change
Aspen Pub./IYC AS160-03 March 5, 2003 18:35 Char Count= 0


in infant cortisol and behavioral response to inocula- learning. International Journal of Behavioral Devel-
tion. Child Development, 65, 1491–1502. opment, 12(2), 221–237.
Rosenstein, D., & Oster, H. (1988). Differential response Sullivan, M.W., & Lewis, M. (1990). Contingency Interven-
to four basic tastes in newborns. Child Development, tion: A program portrait. Journal of Early Interven-
59, 1555–1568. tion, 14(4), 367–375.
Rozen, P., Hadt, J., & McCauley, C. (2000). Disgust. In Sullivan, M. W., & Lewis, M. (in press). Contextual deter-
M. Lewis & J. M. Haviland-Jones (Eds.), The hand- minants of infant anger. Developmental Psychology.
book of emotions (2nd ed., pp. 637–653). New York: Sullivan, M. W., Lewis, M., & Alessandri, M. (1992). Cross-
Guilford. age stability in emotional expressions during learning
Scanlon-Jones, S., Raag T., & Collins, K. (1990). Smiling and extinction. Developmental Psychology, 28, 58–
in older infants: Form and maternal response. Infant 63.
Behavior and Development, 13, 147–165. Sullivan, M. W., Ramsay, D., & Lewis, M. (1992). Learning
Sigman, M., Kasari, D., Kwon, J., & Yirmiya, N. (1992). Re- and temperament in young infants: A longitudinal
sponses to the negative emotions of others by autistic, study of 2- to 8-month-olds. A paper presented at the
mentally-retarded, and normal children. Child Devel- Biennial Meeting of the International Society for Infant
opment, 63, 796–807. Studies, Miami, FL.
Sroufe, L., & Waters, E. (1976). The ontogenesis of smil- Vine, I. (1973). Facial visual signalling and social devel-
ing and laughter: A perspective on the organization opment. In M. von Cranash & I. Vine (Eds.), Social
of development in infancy. Psychological Review, 83, communication (pp. 195–289). New York: Academic
173–189. Press.
Sroufe, L., & Wunsch, J. P. (1972). The development of Walden, T., & Knieps, L. (1996). Reading and respond-
laughter in the first year of life. Child Development, ing to social signals. In M. Lewis & M. W. Sullivan
43, 1326–1344. (Eds.), Emotional development in atypical children
Steiner, J. E. (1979). Human facial expressions in response (pp. 29–42). Mahwah, NJ: Erlbaum.
to taste and smell stimulation. Advances in Child De- Watson, J. B., & Rayner, R. (1920). Conditioned emotional
velopment and Behavior, 13, 257–296. reactions. Journal of Experimental Psychology, 3, 1–
Stenberg, C. R., Campos, J., & Emde, R. (1983). The facial 14.
expression of anger in seven-month-old infants. Child Yirmiya, N., Kasari, C., Sigman, M., & Mundy, P. (1989).
Development, 54, 178–184. Facial expressions of affect in autistic, mentally re-
Sullivan, M. W., & Lewis, M. (1988). Facial expressions tarded, and normal children. Journal of Child Psychol-
during learning in 1-year-old infants. Infant Behavior ogy and Psychiatry, 30, 725–735.
and Development, 11, 369–373. Zajonc, R. B. (2001). Mere exposure: A gateway to the
Sullivan, M., & Lewis, M. (1989). Emotion and cogni- subliminal. Current Directions in Psychological Sci-
tion in infancy: Facial expressions during contingency ence, 10, 224–228.

Back to lists of previous articles- http://depts.washington.edu/isei/iyc/iyc_previous.html

Lists of current articles- http://depts.washington.edu/isei/iyc/iyc_comments.html