Content uploaded by Claire D Vallotton
Author content
All content in this area was uploaded by Claire D Vallotton on Apr 12, 2018
Content may be subject to copyright.
SIGNS OF EMOTION: WHAT CAN PREVERBAL CHILDREN “SAY”
ABOUT INTERNAL STATES?
C.D. VALLOTTON
University of California, Davis
Abstract
Do infants explicitly recognize feelings and emotions in themselves and others? What would
preverbal children say about internal states if they had the words? Investigation of infants’ emotional
understanding is limited by the challenge of understanding infant mental states before the onset of
speech. I examined the use of symbolic gestures by normally hearing, preverbal children to discover
whether infants and toddlers represent emotion concepts such as sad and scared, and feeling words
such as sleepy. Participants were 22 children (5–28 months) in a childcare program where caregivers
modeled symbolic gestures. Gesture use by children and caregivers were videotaped and coded to
determine context, characteristics, and frequency. Twenty of 22 children used symbolic gestures; of
these, 6 used emotion gestures, and 5 used feeling gestures. These gestures were not imitations of
adult gestures, and qualitative data reveal their context and significance. Symbolic gestures reveal
the sophistication of infants’ internal worlds and their ability and desire to communicate thoughts
and feelings. Symbolic gestures are a promising methodology for investigating early explicit mental
processes. As a therapeutic communication tool, symbolic gestures may help children express
emotions, participate in conversations about emotion, and construct their own understanding of
internal states.
Katie, 27 months old, sat at the art table in the toddler classroom with her mother and caregiver
and played with the paint. Katie looked intently at the picture that she was drawing. She painted
a circle and said “Happy face,” and smiled. Then, looking at her mother, she said “Mama, you
make a happy face.” Katie’s mother painted a circle. Katie took her paint brush and put two
dots in her mother’s circle; she looked up and said “Eyes.” She smiled and, taking her brush
again, painted a line in the circle, saying “Smile.” She added a third dot, smiled, and said “I
made a nose!” Her classroom caregiver reflected, “Katie, you finished the happy face! You
put eyes and a nose and a mouth on the happy face!” Katie smiled and returned to her painting.
Meanwhile, 24-month-old Gerry stood by the front door of the toddler classroom crying, and
whimpered “Mommy.” He walked over to the book shelves, slapped a single hand at a few of
the books, and then paused as he cried. Gerry’s caregiver, Mandy, approached and sat down
next to him, saying “Gerry, I see that you are sad. You’re thinking about your mom.” He
continued to cry and crawled onto a soft chair. Mandy asked “Do you want to read a book?”
Gerry whimpered “Yes.” He climbed slowly out of the chair, walked to the books, and went
directly to the book about gorillas. He carried it to Mandy, sat between her legs, looked up at
her, still sad, and said “Mommy.” Mandy responded “This is the book your mom was reading
to you.” With his head down and hands busy flipping open the book pages, Gerry whispered
“Yes.” Mandy began to read.1
Direct correspondence to: C.D. Vallotton, Harvard Graduate School of Education, Larsen Hall, Appian Way, Cambridge, MA 02138;
vallotcl@gse.harvard.edu.
NIH Public Access
Author Manuscript
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
Published in final edited form as:
Infant Ment Health J. 2008 May 13; 29(3): 234–258. doi:10.1002/imhj.20175.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
In the first observation, Katie represented the concept of happy through both her words and
her drawing as she was painting with her mother. She associates the concept of happy with the
action of smiling and shows that she has a basic understanding of the emotion. Meanwhile, her
slightly younger peer Gerry uses just a few words to tell his caregiver why he is sad and what
will comfort him. Using these words, he is able to actively participate in and shape a routine
that will help him regulate his emotions. Most children begin to represent emotion concepts
vocally around 2 years of age, but was this the beginning of Katie’s understanding of the
concept of happy or Gerry’s ability to actively participate in regulatory interactions? In this
study, I show that these children’s abilities to represent and communicate about emotions and
emotional events are not as new as their emotion vocabularies. Both Katie and Gerry used
symbolic gestures to represent emotions, as well as other concepts, before they could speak,
and they used these symbols as a way to “converse” about emotions with their caregivers, to
express their own feelings, and to participate in regulatory interactions.
Though studies have shown that children understand spoken language before they vocalize
words (Fenson, et al., 1994), children are not generally believed to reflect on or have an explicit
understanding of emotions—either their own or those of others—until after they are verbal. In
this study, I examine infants’ and toddlers’ use of symbolic gestures to break the language
barrier that keeps scientists, practitioners, and parents from seeing and responding to the extent
of preverbal children’s understanding of feelings and emotions.
YOUNG CHILDREN’S UNDERSTANDING OF EMOTIONS
Young children learn about the social and emotional world in an amazingly short time.
Habituation studies have shown that infants as young as 2 to 3 months can discriminate between
positive (happy) and negative (angry or fearful) facial expressions, and by 7 months, infants
can even discriminate more subtle categories and intensities of emotions (as reviewed by
Nelson, 1987). Social referencing studies employing a visual cliff or novel object have
established that by 12 months, infants understand the contextual meaning of their mothers’
facial expressions and behave accordingly (i.e., Sorce, Emde, Campos, & Klinnert, 1985).
Thus, infants’ behaviors elicited by emotion stimuli reveal recognition of and responsiveness
to emotions. Their meaningful—contextually appropriate—response to emotion, paired with
the fact that preverbal-children’s verbal comprehension exceeds production, leads to the
question “Can preverbal children explicitly think about emotions; that is, can they represent
them mentally?”
The development of the capacity to represent emotions—to reflect on or name emotions—is
an integration of cognitive and social-emotional skills, an integration of domains which
continues to develop throughout childhood. Studies examining this integration as early as the
late second and early third year of life have focused on children’s use of spoken words to
describe the thoughts and feelings of others (Dunn, Bretherton, & Munn, 1987), and the
possible causes of emotion and emotional consequences of behavior (Lagattuta, 2001).
Typically developing children begin to express emotion concepts (e.g., sad, happy) and internal
feeling states (e.g., sleepy) between 18 and 20 months, shortly after the onset of verbal
language, with most children acquiring the use of basic internal state words around 2 years of
age (Bretherton & Beeghly, 1982; Dunn et al., 1987). Children’s articulation of emotion
understanding grows steadily, as seen in spontaneous and elicited conversations with parents;
as early as 3 years old, some children begin to understand the link between past experiences
and emotions (Lagattuta, 1999) and between thoughts or memories and emotions (Lagattuta,
Wellman, & Flavell, 1997). The complexity of children’s explanations of emotions and their
1Both anecdotes are drawn from systematic child observations recorded by teachers at the University of California, Davis Center for
Child and Family Studies. Names have been changed, but age and gender are accurate.
VALLOTTON Page 2
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
ability to link cognitive and emotional states continues to grow as increasingly related skills
throughout early childhood (Harris, 1989). At around 5 years, children develop an explicit
Theory of Mind (ToM) (Astington, 1993; Flavell, 1999); that is, the understanding that others
have mental processes which can differ from their own, including beliefs, desires, and
emotions. Concordantly, typically developing 5-year-olds can articulate their understanding
of emotions and begin to succeed on ToM tasks, indicating an adultlike understanding of the
basic properties of thoughts and beliefs. Children continue to integrate this knowledge with
their understanding of the causes and consequences of emotion throughout the next 1 to 2 years
(deRosnay, Pons, & Harris, 2004). Finally, links have been found between use of mental state
language—including emotions—and children’s ToM performance. Parental mental-state
language has been found to predict children’s later ToM (Adrian, Clemente, Villanueva, &
Rieffe, 2005), and children’s own mental-state language is related to their concurrent ToM
(Symons, Peterson, Slaughter, Roche, & Doyle, 2005). Thus, young children’s representations
of emotion concepts may be seen as an early part of the developmental path to integrating
emotion and cognition.
More than simple maturational correlaries, representation skills may actually support the
development of social-emotional understanding. Showing a firm, and perhaps causal, link
between emotion understanding and language development, Cutting and Dunn (1999) found
that preschoolers’ language abilities, including both receptive and expressive language, were
correlated with their emotion understanding as indicated by their ability to label and predicts
others’ affects. Further, Pons, Lawson, Harris, and deRosney (2003) showed that age and
language ability together explain most (70%) individual differences in development of emotion
understanding in children followed between 4 and 11 years of age. These findings suggest that
children’s understanding of emotions may be partially dependent upon their language abilities.
Summarizing the literature on early development of emotion understanding, a gap is revealed
between the period of infancy (<1 year), when most studies focus on children’s abilities to
recognize basic differences between emotions, and the period of late toddlerhood or early
preschool (2–3 years), when studies focus on children’s abilities to verbalize explicit
representation and understanding of emotions. The question addressed in this study is whether
preverbal children in this age gap can explicitly represent an understanding of emotion.
PREVERBAL CHILDREN’S REPRESENTATIONAL REPERTOIRE: SYMBOLIC
PLAY AND GESTURE
Preverbal children are neither thought to intentionally cope with their emotions (Aldwin,
1994; Kopp, 1982; Murphy & Moriarty, 1976) nor have an explicit understanding of others’
emotional expressions—or a Theory of Mind for emotions. The failure of developmental
science to see these skills at younger ages may be due to (a) a reliance on children’s
verbalization of emotion concepts before emotion understanding can be measured and (b) the
assumption that children’s ability to use emotion concepts effectively in social situations
develops only after they can voice their feelings. What could be learned about infants’
knowledge of emotions if they could express what they know in a way adults could understand?
Young children begin to use spoken words as symbols to communicate in the first half of their
second year (Reich, 1986), but this is not the beginning of a child’s understanding and
production of symbols. Young children have other systems of symbols available with which
to represent concepts. Children’s representations of objects and concepts are expressed
physically (i.e., Piaget, 1952) before they are expressed vocally (Werner & Kaplan, 1963). To
be a true symbol, a representation must communicate the concept of the referent in its absence
(Werner & Kaplan, 1963). For example, a spoken or written word can communicate any
concept chosen in the absence of the object to which it is referring, as long as the meaning of
VALLOTTON Page 3
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
it is shared. Meeting this criteria, preverbal children use symbols in the absence of their
referents in play when they act out concepts for which they have not yet learned the words,
and use one object to represent another (Ungerer, Zelazo, & Kearsley, 1981). The use of objects
as a representational medium, however, limits representation and communication to situations
in which they have an object on hand that can be manipulated in a way similar to the referent
object. Another media of representation that is available prior to language and not so limited
is gesture.
Gestures are intentional motor actions usually intended by children as communicative cues
(Bakeman & Adamson, 1986; Bates, Camaioni, & Volterra, 1975). Typically, pointing emerges
as the first gesture around 9 months of age (Carpenter, Nagell, & Tomasello, 1998; Crais,
Douglas, & Campbell, 2004); it is a robust gesture, a version of which is even used by blind
infants (Iverson, Tencer, & Lany, 2000). The “showing” gesture also emerges around 9 months
(Bates, O’Connell, & Shore, 1987) and is used to engage an interaction partner. Typically
developing children use gestures both as a means of requesting (i.e., requesting to be given an
object out of reach) and directing another’s attention (Carpenter et al., 1998). Though these
gestures so commonly used by young children are not symbolic according to Werner and
Kaplan’s (1963) criteria, they are intentional and communicative, and thus indicate the
beginning of children’s attempts and abilities to create shared meaning.
In addition to the common nonsymbolic gestures, preverbal children are capable of learning
and even inventing truly symbolic gestures, representing absent referents (Acredolo &
Goodwyn, 1985, 1988). Acredolo and Goodwyn (1985) first documented the symbolic gestures
of one infant, Linda Acredolo’s daughter Kate, when Kate began inventing gestures for objects
in her environment at 12.5 months of age. Acredolo described her first observation of Kate’s
symbolic gestures with the following anecdote:
While about 3 feet away from [a rose bush], Kate pointed at it and sniffed. From then
on the sign occurred regularly in response to many flowers and pictures (always at a
distance) and in answer to various verbal prompts (e.g., “What’s that?”). (Acredolo
& Goodwyn, 1985, p. 42)
Kate used the “flower” gesture in both the presence and absence of flowers, and generalized
it to both real flowers and other representations (i.e., photographs) of flowers. Kate went on to
use a total of 29 different symbolic gestures—13 which she invented and 16 modeled by her
caregivers—before she began to talk prolifically at 18 months (Acredolo & Goodwyn, 1985).
Acredolo and Goodwyn (1988) followed their case study by investigating whether and when
other children use symbolic gestures and whether use of symbolic gestures could slow language
development. An interview study with parents of 17-month-olds established that the
phenomenon of spontaneous use of symbolic gestures was not unique to the case-study infant;
and a longitudinal parent diary study detailed the concurrent development of gesture and
language in infants from 11 to 20 months (Acredolo & Goodwyn, 1988), showing that use of
gestures to communicate faded as children began to effectively use vocal words. This study
also documented that infants’ first gestures were typically requests such as up or outside,
followed by gestures naming objects such as dog or car and events. Finally, Goodwyn and
Acredolo (1993) used an experimental design to test the effects of symbolic gesturing on
language development; an experimental group of families used symbolic gestures with their
children while a control group of parents focused on their children’s oral language
development. Results revealed that children could learn a symbolic gesture for a given concept
before the corresponding vocal word (Goodwyn & Acredolo, 1993), indicating that conceptual
and representational capacities precede language development. Further, those who were taught
to use symbolic gestures used vocal language earlier and scored higher on measures of
expressive language (Goodwyn, Acredolo, & Brown, 1989), assuaging fears that use of
VALLOTTON Page 4
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
gesturing would prevent children from learning vocal language, and suggesting that gesturing
may help children build symbolic capacity.
Approaching these findings on preverbal-children’s representational capacities with an interest
in the beginning of children’s understanding of internal states leads to the following questions:
If preverbal children can use symbolic gestures to communicate desires and observations, can
they also explicitly represent emotion concepts? And if so, when will these emotion gestures
appear in relation to gestures for requests and objects?
THE CURRENT STUDY
The current study examines the use of symbolic gestures by normally hearing infants and
toddlers to determine whether young children can represent emotions, feelings, and time
concepts in dialogue with others. Infants and toddlers were those enrolled in an early childhood
education program in which symbolic gestures were modeled by caregivers.
Use of Symbolic Gesturing in Infant and Toddler Care
The children who participated in this study were in a unique, university-based childcare and
education program in which caregivers (university students) used symbolic gestures with the
children during typical routines and spontaneous interactions. Caregivers were also taught, in
accordance with the program’s philosophy, ways to support children’s emotional and cognitive
development from an attachment theory perspective.2 Caregivers were taught to use gestures
through modeling by their classroom supervisors; they were given a list of common gestures
and their descriptions, and small posters describing specific gestures were placed around the
classrooms as reminders.
Children were never directed to use gestures, but learned and used the gestures modeled by
caregivers in conjunction with speech during typical daily interactions. Children occasionally
invented new gestures themselves which, if understood, would then be used by the caregivers
in interactions with the children. Caregivers were encouraged to respond to children’s gestures
regardless of how precisely children performed them, and to pay particular attention to gestures
children invented to discover their meaning and respond appropriately.
Caregivers modeled a total of 71 symbolic gestures including signs for the feeling, emotion,
and time concepts of interest in the current study. Feeling concepts include hurt, cold, loud,
gentle, and sleepy; emotion concepts include sad, angry, scared, and happy; and time concepts
include wait, later, and Popsicle Time. Descriptions of a sample of symbolic gestures observed
in the Infant and Toddler Program during the study period can be found in the Appendix.
Defining Feelings and Emotions
Though both feelings and emotions can be internal states, there are reasons for separating them
into two different categories. In accordance with the work of Bretherton and Beeghly (1982),
feelings are primarily physiological sensations that can be a result of internal (sleepy, hurt) or
external stimuli (cold, loud) whereas emotion is used to label the more complex psychosocial
internal states such as sad or happy.
The Inclusion of Time Concepts—It may not be readily apparent why time concepts are
included in a study of feelings and emotions; however, time concepts are integral to the ability
to regulate emotions and behavior. Self-regulation of emotions and behavior has much to do
2This program philosophy and training is reflected in the caregivers’ responsiveness to children’s emotional cues as seen in the qualitative
transcripts.
VALLOTTON Page 5
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
with anticipation of future events. For example, one of the definitions of self-regulation
includes the ability “to postpone acting upon a desired object or goal” (Kopp, 1982, p. 199).
Waiting is one of the primary challenges to any infant or toddler, and learning to maintain one’s
emotional state while waiting until later or until a specific time may be a critical task in utilizing
cognitive skills in the service of self-regulation once a sense of the future becomes conscious.
Waiting is also critical in the ability to get along with others—such as peers who may be holding
the only toy a child wants and adults who insist on keeping schedules despite a child’s
immediate desires. Thus, young children’s abilities to understand or symbolize time-related
concepts are also of interest in this study, particularly the concepts wait, later, and Popsicle
Time. Popsicle Time is a concept specific to the study population, referring to the program’s
daily routine in which parents return for their children. Children, caregivers, and parents eat
popsicles (ice made from water with a little juice) together each day at the end of the program,
providing the children with a specific reference for the daily event of reunion with their parents.
Questions
The spontaneous use of gestures by infants and toddlers in a childcare setting was examined
to investigate whether and to what extent preverbal children—who had been systematically
introduced to the use of symbolic gestures by adults—used emotion concepts in their early
communications. The questions addressed in this study are:
•Do infants and toddlers use symbols to communicate emotion-state concepts such as
sad, happy, and angry?
•Do they use feeling concepts, indicating a perception of internal sensations?
•Do they use time concepts, indicating an anticipation of events?
•In what contexts do young children express emotion concepts through gesture? Do
they use gestures to communicate their own feelings, or comment on others’ feelings?
•When do they begin to use these types of gestures in relation to other types of gestures,
such as those for requests or objects?
METHODS
Participants
Participants were 10 infants and 12 toddlers in an Infant and Toddler Program at the University
of California, Davis Center for Child and Family Studies (CCFS), where the use of symbolic
gesturing was modeled by caregivers. Infants were 7 females and 3 males who were 4.5 to 11.5
months old when the study began and 12.5 to 19.5 months when the study ended. Toddlers
were 5 females and 7 males who were 17.3 to 24.8 months old when the study began and
between 20.8 and 28.3 months when the study ended. Though enrollment in the CCFS Infant
and Toddler Program was open to the entire community, priority was given to the children of
University students, staff, and faculty; thus, many of the study participants were children of
University community members.
Videotaping Procedures
Children were videotaped during normal program routines at the CCFS. Videotapes of infants
were collected for 8 months whereas videotapes of toddlers were collected for 3.5 months.3
3Initial data on infants and toddlers were gathered separately for two different studies on symbolic gesture use; thus, the methods of
initial data collection are not equivalent, though coding of symbolic gestures from the videotapes was consistent. The datasets are presented
together in this article because they provide complimentary information; however, analyses of the two groups were done separately, and
they were not compared statistically.
VALLOTTON Page 6
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Videographers were in the classrooms and playgrounds with the children and could move to
follow the children as necessary. Children in each classroom were filmed approximately the
same number of times, but the order of filming varied randomly each time the videographers
cycled through the list of participating children.
Infants—Each infant was videotaped for a 5-min period an average of 40 times over the course
of 8 months. All videotapes were taken when infants were interacting with a caregiver during
typical program routines: snack time and free play. Snack time and free play were chosen for
data-collection times because these routines do not elicit specific emotions as would separation
or diaper-changing routines. Thus, the timing of infant observations was designed to gather a
sample of all spontaneous gesturing behavior in this classroom context rather than specifically
feeling, emotion, and time concepts. On average, each infant was filmed a total of 200 min
(2.5 hr), an average of approximately 1% (0.93) of their 360 hours in the classroom over the
8 months of data collection.
Toddlers—Each toddler was videotaped an average of 15 times over the course of 3.5 months.
Toddlers were videotaped during regularly occurring distressing routines (e.g., separation from
parents, diaper changes, and conflicts between children) in which they were typically
interacting with a caregiver; however, presence of a caregiver was not required in the case of
filming spontaneously occurring peer conflict. Videographers shadowed each child and began
filming before the anticipated distressing event (except in the case of spontaneously occurring
conflict), and continued recording until the child had resumed normal play—a sign that the
distress was mostly resolved for the child. Videotape lengths ranged from 0.5 to 19 min (M =
5.5 min). Thus, the timing of toddler observations was designed to gather a sample of
emotionally relevant gesture behavior.
Coding
All videotaped episodes were coded unless there were technical problems rendering the
behavior of children or caregivers unrecognizable. Observers coded videotaped episodes in
real time; that is, they marked time, in minutes and seconds, as each behavior occurred. Next,
I describe the coding in terms of gesture content, conversational context, and social-emotional
context. I then describe the process of coder training and reliability.
Content of Coding
Gesture content: Gestures were defined as intentional and communicative motor behaviors
performed in the context of interaction. Because gestures were performed largely the same way
across participants, coders recorded the concept represented by each gesture rather than
describing the form of the gesture. For example, if a child or caregiver traced a finger down
her cheek from her eye, coders recorded “sad;” if she tapped her fingers against her mouth,
coders recorded “eat/snack” (for examples of gestures and descriptions of how they are
performed, see the Appendix). To be sure that children’s gestures were indeed meaningful
rather than random motor behavior, behaviors were recorded as gestures only if they were in
the context of an interaction with a caregiver as indicated by body positioning and eye contact.
Each gesture performed by children and caregivers in each episode was recorded. Because the
timing of each gesture was recorded, the data also include the sequence and frequency of
gestures.
Categories of gestures: The gestures used by caregivers and children can be placed in 11
different categories: Action/Event, Adverb/Adjective, Emotion, Feeling, Nonsymbolic,
Object, Parent, Question, Request, Time, and Yes/No. These categories reflect both the varying
definitions of the gestures (i.e., those for actions such as play and those for nouns such as
bird) and how the children typically use the symbols pragmatically. Numerous gestures could
VALLOTTON Page 7
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
be placed in more than one category. For example, outside could be used by a child or caregiver
as an adjective (e.g., “See the bird outside”) or as a request (e.g., “Want to go outside”). For
simplicity, I categorized each gesture in 1 of the 11 mutually exclusive categories listed earlier
based on the gesture’s most common use among children. For example, though the symbol for
eat/snack may be seen as an event or a noun, it was most often used by the infants as a request;
thus, it was grouped as a request in the current study. Because the gestures were preassigned
to categories, coders did not assess the categories of the gestures but simply recorded which
gestures they observed (for examples of gestures in each category, refer to the Appendix).
Conversational context: To establish that infants were expressing their own thoughts and
feelings through their gestures rather than simply imitating gestures performed by caregivers,
it was important to code whether the gesture was (a) initiated by the infant, (b) in reply to a
caregiver, or (c) an imitation of a caregiver’s gesture. Thus, for the infant data, observers also
coded the conversational context of each gesture. That is, gestures were coded in one of four
conversational categories; (a) Initiation: Gesture was not preceded by any other gesture in the
last 5 s;4(b) Continuation: Gesture was preceded by a different gesture by the same individual
within 5 s (e.g., child gestures “mom,” then he gestures “later”); (c) Response: Gesture was
preceded by the same gesture by a different person within 5 s (e.g., caregiver gestures “snack,”
then child gestures “snack”); and (d) Reply: Gesture was preceded by a different gesture by a
different person within 5 s (e.g., caregiver gestures “bird,” then child gestures “where?”).
Importantly, the Response category includes gesturing that may be an imitation of another’s
gesture.
Social-emotional context: To examine the social-emotional relevance of children’s gesturing,
other child and caregiver behavior and significant events occurring in the classroom were
described in real time to clarify the meaning of symbolic gestures. These descriptions include
child’s affect and vocalizations, caregiver’s gestures and vocalizations, and events such as a
mother’s arrival or departure, or another child’s crying.
Process of Coding
Coder training: Coders of the infant videotapes were University students who had not
interacted with the children in the study or with any child using symbolic gestures. Coders of
the toddler data were University students who had never directly interacted with the children
in the study but who had observed children using gestures while collecting the videotape data.
Coders were trained to recognize gestures–symbolic and nonsymbolic—by learning behavioral
descriptions of the gestures and seeing examples of them on coding-training videos. If there
was more than one way to perform a sign, the coders were given both descriptions, and both
versions of the sign were coded the same.
Intercoder reliability: Observers obtained a κ of .75 or above—on the timing, content, and
conversational context of each gesture—before beginning to code independently; agreement
among coders was reassessed on 15% of the episodes using Cohen’s κ to ensure consistency
in interpretation of the codes. Agreement on all codes (timing, content, and conversational
context of symbolic gestures) was considered critical for accurately interpreting their meaning.
For the infant data, coders’ agreement on both type (i.e., which gestures were being performed)
and conversational context of gestures was assessed concurrently; that is, coders had to agree
on both which gesture was performed and whether the infant had initiated, continued,
responded, or replied to an adult for codes to be considered in agreement. There was high
4Five seconds was used as a conservative, yet somewhat arbitrary, marker of conversational timing. Through visual and auditory review
of the gesturing episodes, it was determined that if a child or caregiver were to respond to another’s gesture, it would happen within 5 s,
and that those occurring more than 5 s later were not responses as indicated by changes in attention and gesture content.
VALLOTTON Page 8
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
agreement on both of these gesture qualities. Coders of infant data achieved interrater reliability
scores of κ = ≥.75 on five tapes in a row before coding independently, and upon reassessment
of 15% of all tapes, had scores of κ = ≥.83. Coders of the toddlers’ types of gestures (not coded
for conversational context) also achieved interrater reliability scores of κ =≥.75 on five episodes
in a row before coding independently, and reassessment had maintained κ scores above .75.
Transcripts
The coded gestures, conversational context, and descriptions of social-emotional context were
used to create transcripts of the gesturing episodes that included children using emotion of
feeling gestures. These transcripts are presented as qualitative data in the results of this study;
they corroborate the meaningfulness of the infants’ use of emotion, feeling, and time gestures.
RESULTS
Gestures Used in the Classrooms
All of the caregivers used symbolic gestures in their interactions with children. Caregivers
performed a total of 71 different symbolic gestures between them, and 20 of the 22 children
displayed a subset of 51 different gestures between them (A sample list of gestures observed
in the classroom during the study period is presented in the Appendix.) Most gestures by
children and caregivers were observed more than once per individual; however, some were
captured on videotape only once. Some gestures were more likely to be used in the infant
classroom whereas others were seen only in the toddler classroom, highlighting the fact that
though caregivers were all trained the same way, there was no attempt to control their use of
gestures, resulting in individual differences. This also means that there was no control of
differences in how many times a child was exposed to a given gesture nor whether they were
exposed to it at home.
Children Represent Both Emotion States and Time Concepts Through Gesture
Of the 20 children observed to use symbolic gestures, 9 of them used at least one emotion or
feeling gesture, or both. Six children used emotion gestures, the earliest recorded on video at
10.9 months of age, and 5 children used feelings gestures, the earliest recorded at 14.7 months
of age. Additionally, 11 children used time-related gestures, the earliest at 12.8 months of age.
Table 1 displays the number of infants and toddlers who were observed to use each of the
emotion, feeling, and time gestures, including the youngest age at which the gesture was
observed to be used. The second column in Table 1 names the gesture, the third column
describes the total number (of 22 observed) of infants and toddlers who were observed on
videotape to perform the gesture at least once, and the fourth column describes the earliest age
at which any child was observed on videotape to perform the gesture.
Preverbal Children Use Emotion and Feelings Gestures to Talk About Their Own and Others’
Feelings in Socially Meaningful Contexts
Each episode in which a child used a gesture communicatively was transcribed in terms of its
social and emotional context. There were 37 transcribed gesturing episodes (across 9 children)
that included a child using emotion or feeling gestures, excluding those which included
emotionally relevant content (i.e., mom leaving or returning) but did not include an emotion
or feeling gesture. Based on the transcription of the child’s behavior, caregivers’ behavior, and
salient events in these episodes, it was possible to discern whether a child was representing his
or her own internal state or the internal state of another for 26 of the 37 episodes. The numbers
of episodes about self and other, as well as those episodes which could not be determined due
to insufficient information, are presented in Table 2. Column 1 of Table 2 identifies the 9
children who were observed to use emotion or feeling gestures during a videotaped interaction;
VALLOTTON Page 9
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
columns 2, 3, and 4 indicate the number of episodes for each individual child in which the
emotion gestures were judged to be about the child’s own emotions (column 2), the emotions
of another (column 3), or whether there was insufficient information to make such a
determination (column 4). Columns 5, 6, and 7 provide the same information about episodes
that included feeling gestures.
To conservatively test whether children’s use of emotion and feeling gestures could be random
rather than meaningful, I tested whether the number of interpretable episodes (n =26) was
greater than chance. Testing conservatively, I assumed that all 11 episodes in which the use of
emotion and feeling gestures could not be determined as self- or other-referent were random,
or not meaningful behavior. In a binomial distribution, the 70% (26/37) observed to be
contextually appropriate is statistically above chance (p < .01).
Table 3 presents six samples of the transcribed observations of children’s use of feeling and
emotion gestures. Five of the observations in Table 3 were selected from the 26 socially
meaningful episodes transcribed from the videotapes; however, Observation 4 was recounted
by the University of California, Davis Academic Child Development Specialist. These
qualitative observations demonstrate infants’ ability to articulate their own internal states as
well as their ability to reflect on the internal states of others.
Further qualitative interpretation of the transcripts showed that 3 children demonstrated the
beginnings of explicit empathy by commenting on the emotional expressions of another child
(for examples see Observations 5 and 6 in Table 3); the earliest age at which this appeared was
15.4 months. Further qualitative analysis of the other transcribed observations revealed that 2
children demonstrated the ability to distinguish between negative emotions by correcting their
caregivers as to which emotion they were feeling (i.e., mad, not sad), as described in
Observation 4 in Table 3. Two children demonstrated an understanding of the causes of sadness
by describing what another child needed to feel better (see Observation 6 in Table 3); the earliest
age at which this was observed was 15.5 months.
Emotion Gestures Used Earlier Than Time and Feeling Gestures
Only data from the infant participants were used to describe the order in which each gesture
category appears because only the infants were observed from their first exposure to symbolic
gestures whereas many toddlers had been exposed to symbolic gestures during the previous
year if they had been enrolled in the program as infants. To describe the developmental
sequence of the appearance of gesture categories, the age at which each child was first observed
to use a gesture in each category was determined. I then averaged the ages of first use across
children within each category. Figure 1 displays the sequence of gesture category development
according to average age at first observation. The black bars show the median age for each
category. The top and bottom of the gray boxes show the 25th and 75th percentiles,
respectively, of ages. The horizontal lines above and below the boxes show the oldest and
youngest ages, respectively, at which children were first observed to use each category of
gesture. The black circle below the Yes/No bar represents an outlier. Below each bar are the
numbers of infants who were observed to use at least one gesture in that category.
By ordering the gesture categories according to average earliest age of appearance across
infants, the following order of appearance is found: (a) nonsymbolic gestures at 9.83 months,
(b) request gestures at 10.32 months, (c) gestures naming objects at 11.37 months, (d) gestures
naming actions or events at 13.52 months, (e) gestures representing questions at 13.60 months,
then (f) gestures naming emotions at 13.83 months, (g) yes and no gestures at 14.16 months,
followed by (h) time concepts at 14.23 months, (i) Mom and Dad at 14.82 months, and (j)
feelings/sensations at 14.70 months. These results are consistent with and suggest an
elaboration of the sequence of appearance reported by Acredolo and Goodwyn (1988), who
VALLOTTON Page 10
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
described the development of naturally occurring (rather than intentionally modeled) symbolic
gestures as reported in parent interviews. According to the parents’ reports, objects and requests
were the most common categories of symbolic gestures observed, and infants’ use of requests
preceded gestures for objects and events by an average of .59 months.
Not Just Imitation: Infants Initiated More Than Imitated Symbolic Gestures
To be certain that infants’ expression of emotion, feeling, and time concepts through gestures
were not simply a result of imitation of adult gestures, the conversational context was assessed
for each of these categories of gestures. A binomial distribution was used to test whether the
infants’ emotion, feeling, and time gestures were as likely to be imitations as nonimitations,
as judged by coders. Results are presented in Table 4. As seen in the table, infants’ use of
emotion, feeling, and time gestures were rarely imitations of an adult gesture; that is, infants’
use of emotion, feeling, and time gestures were not preceded by an adult gesture in the prior 5
s.
DISCUSSION
This article began with two stories of toddlers’ use of words to communicate about emotions.
Katie used her drawing and words to initiate a conversation with her mother about the concept
of happy, and Gerry used a few simple words to draw comfort from his caregiver when he was
sad that has mother had left. The question I posed was whether their concepts of and abilities
to communicate about emotions were predicated upon the children’s use of words or whether,
given the right tools for representation, these children might have communicated their
understanding of emotions before they could talk. The results of this study indicate that
preverbal children exposed to the systematic use of symbolic gestures can use these gestures
to represent and communicate about their own and others’ emotions, initiating and maintaining
meaningful, emotion-related, explicit interactions with their caregivers.
This study provides a first report on preverbal-children’s use of emotion gestures. In
interpreting the results, note that the data were collected as an observation of children’s
spontaneous behavior and thus describe what preverbal children in this unique classroom
context did rather than testing what they were capable of doing when their gesturing behaviors
were elicited and supported. Next, I address the limitations of the study, then describe both the
theoretical implications and potential clinical applications of the results.
Limitations
Can We Generalize to Other Children and Other Contexts?—The most important
limitations of the current study are to generalizability. Even more critical than the small sample
size is the unique culture of the CCFS. The findings that at least some preverbal infants and
toddlers can use gestures to represent feelings, emotions, and time concepts must be
contextualized by the highly gesture-rich environment in which gestures were consistently
modeled and responded to. There is no reason to believe that this aspect of the context could
not be replicated in other childcare centers or the home environment, but the current results
must be interpreted with the specific context in mind. With the growing popularity of the Baby
Signs Program (Acredolo & Goodwyn, 1992) among both parents and childcare centers in the
United States and other nations, there are growing opportunities to further the study of
preverbal-children’s use of symbolic gestures in interaction contexts.
Additionally, the generalizability of the findings on the order of symbolic gesture development
is limited by the somewhat cross-sectional nature of the current study design. Though infants
were followed intensively for 8 months and toddlers for 3.5 months, to understand the
development of different symbolic gestures and emotion concepts over time, a larger sample
VALLOTTON Page 11
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
of children should be followed for approximately 2 years—from before the beginning of
nonsymbolic gesture use to after the beginning of word use.
Why were There so Few Observed Uses of Emotion and Feeling Gestures?—
One limitation of this study is that the data most likely underestimated the number of children
in this sample who could and did use emotion, feeling, and time gestures, and the number of
times each child used the gestures. Rather than constructing a circumstance in which children’s
emotion gestures were specifically elicited, testing whether all infants were capable of using
them, I gathered a sample of spontaneous gesturing behavior as it was used in everyday
caregiver–child interactions in this real, but unique, classroom environment. I observed only
1% of the children’s time in the classroom, and likely only 1% of infants’ gesturing behavior,
but these observations were typical of everyday use of gesturing behavior within the daily
routines around which data collection was structured. Though this design was limited in its
ability to verify the full abilities of each child, it describes meaningful behavior by children
and caregivers in the important context of childcare. Future studies should be designed with
complementary controlled tests of elicited gestures as well as observation of spontaneous
gesturing in an interaction context.
How Do We Know Infant Gestures are Really Meaningful?—I did not construct a
situation in which infants could name the wrong emotion; thus, I did not test children’s
discriminatory abilities and confirm that they can always name the correct emotion when faced
with a choice. A further challenge of this type of observation is that even if an emotion gesture
was observed that did not seem to fit the social circumstance of the child, it would be very hard
to say that the child was incorrectly using the gesture because the child might be responding
to internal (i.e., invisible) stimuli—such as a memory or their own current internal state. With
no way to verify what the child is responding to, we have to take their word for it, so to speak.
Thus, there is no way to verify in this type of observation that 100% of children’s emotion
gestures were appropriate to their social-emotional context. However, the qualitative
transcripts (see the examples in Table 3) provide valuable information on children’s abilities
to represent emotions in a meaningful, context-appropriate way. By describing instances in
which children’s symbolic gestures do match the social-emotional context—including the
child’s own internal state or the emotional state of a peer—I have illustrated the ability of
symbolic gestures to make visible children’s social-emotional understanding. Confirming the
meaningfulness of children’s symbolic gesture use, in 26 of 37 (70%) episodes, children’s use
of feeling or emotion gestures could be determined by the social context to be self- or other-
referent.
Finally, if infant gestures were indeed random, or lacked meaning to the infants, then one would
expect that children would more frequently perform one of these meaningless gestures in
imitation of another’s gesture rather than performing symbolic gestures coded as “initiations”
or “replies.” The test of infants’ imitation of adult gestures showed that infants were far less
likely to imitate symbolic gestures than to initiate them, including emotion, feeling, and time
gestures. Thus, I conclude that in most instances, children’s use of emotion, feeling, and time
gestures is meaningful to the infant and appropriate to the social and emotional context.
Limitations of Observation of Spontaneous Behavior in the Childcare
Classroom Setting—There are two limitations of the observational design I used in the
classroom setting. First, because of the nature of the childcare setting and the procedures for
filming, almost all episodes of children’s use of symbolic gestures were in the context of
interaction with a caregiver. Thus, I did not observe how infants may use gestures with other
children or by themselves. Second, although the children were all in the same program with
caregivers who were trained the same way, there was no way to control differences in how
many times a child was exposed to a given sign or how much they were exposed to the use of
VALLOTTON Page 12
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
gestures at home. This limits confidence about which gestures would come first, particularly
within the emotion and feeling categories. However, the current study provides a first look at
a context in which children demonstrate their social-emotional understanding through
symbolic gestures. This information can serve as a guide in developing meaningful
experimental procedures to elicit more of these behaviors to further our study of the integration
between cognition and emotion in preverbal children.
Theoretical and Research Implications
Are Children Representing Internal States or External Expressions?—These
results bring up a number of questions about the details and sophistication of children’s explicit
understanding of emotion. Could infants simply be describing actions that are the results of
emotions rather than truly representing internal states? For example, when children use the
sad gesture, are they instead saying “cry?” When they use the tired gesture, are they instead
saying “sleep?” This may seem logical, particularly since Werner and Kaplan (1963) argued
that our symbols are built upon action. However, this cannot be the case for all of the emotion
gestures. For example, scared and angry have no obvious or consistent specific actions beyond
facial configurations that consistently accompany them. Thus, the symbols used in this setting
by children and caregivers to say “scared” or “angry” do not directly mimic an emotion-related
action but are somewhat more abstract. Further, as seen in Observation 2 (Table 2), when
children refer to themselves as sleepy, they are most likely referring to an internal state rather
than to an action because to refer to the self performing the action of sleep (which is not
happening at the moment the child is gesturing) would indicate hypothetical projection of self
into the past or future, thus demonstrating explicit memory or imaginative play. Future studies
should test whether gestures with specific feeling/emotion-related actions (e.g., sleep and cry)
develop before the more abstract ones (e.g., fear and anger).
Early Emotion-Cognition Integration—Children in this study used emotion gestures to
express internal states including sadness, anger, hurt, sleepiness, and fear. Additionally, the
qualitative observation data showed that infants demonstrated emotion understanding by using
emotion gestures in context-specific ways; for example, 3 children showed what could be seen
as the beginnings of empathy by gesturing to their caregivers about the emotions of another
child. Two children used gestures to communicate both their own state and another’s. These
findings challenge the view that explicit awareness of emotions as well as complex and
differentiated emotional experiences develop later in childhood.
Infants’ use of symbolic gestures representing emotions and feelings catalyzes another line of
questions regarding the potential of symbolic gestures to help children integrate cognitive tools
(i.e., symbols) with emotional experiences. These findings show an early link between
cognition and emotion, using representation for emotional experience. Given the literature
reviewed earlier showing that language skills support emotion understanding, future research
should test whether the early opportunity to apply symbol skills (i.e., gestures) to the social-
emotional domain in infancy leads to earlier or more complex emotion understanding in
preschool children.
Overcoming Methodological Limitations—Granting the foregoing interpretation of the
data, and if children’s use of emotion gestures can be further verified in other populations,
symbolic gestures may give us the opportunity to overcome two barriers to the study of
emotional understanding and social development in preverbal children. First, investigators of
child development have had trouble breaking the language barrier to examine children’s
conscious or explicit knowledge of emotions prior to the third year of life, and how that
knowledge might affect behavior. It would be particularly useful to see if and how preverbal
children’s explicit knowledge of emotions affects their coping strategies with their own
VALLOTTON Page 13
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
emotions and their reactions to the emotional signals of others. If we can verify the position
that preverbal children can use symbols to represent internal states, we can look earlier in
development to examine the role of symbol skills in social development. If Vygotsky
(1934/1986) is correct in the proposal that symbols act as the “psychological tools” of our
higher mental processes, tools with which we think, reflect on, and modify our behavior, we
may expect that children who have a usable symbol system at an earlier age will be able to use
those symbols to understand their experiences and modify their behavior in a way they did not
before.
The second methodological barrier which could be overcome by symbolic gestures is that of
experimenting with children’s language development. Although many studies have shown a
correlation between the domains of children’s language and social development (i.e., Baldwin
& Moses, 2001; Cassidy, Werner, Rourke, Zubernis, & Balarman, 2003; Zeidner, Matthews,
Roberts, & MacCann, 2003), none have experimentally examined the nature of this relation
because all typical children develop language. To understand how these two domains of
development—symbolic competence and social skills—are related, we must separate them
somehow. Symbolic gestures may give us the opportunity to experiment with the relation
between language and social development by providing one group of children symbols prior
to their ability to produce verbal language and comparing their reflective functioning and
social-emotional skills to a comparable group of children who are not using symbolic gestures.
Thus, we can explore the putative causal relations between language development and social
skills.
Further Research on Early Symbol Use and Emotion—First, children’s ability to use
symbolic gestures to express emotion concepts should be verified, both in other populations
of children and in situations designed to elicit specific emotion gestures. For example, children
could be provided stimuli related to specific emotions and be tested on their spontaneous
responses. Additionally, children’s ability to distinguish matching emotions and correct
mismatched labels could be tested by experimentally correctly labeling and mislabeling
emotion stimuli and testing whether children will more often correct a mistaken label.
After such experimental verification of infants’ abilities to accurately use and understand
emotion-related gestures, I suggest that an experimental design with longitudinal data
collection could answer some questions useful to application. By providing one group of
children with consistent modeling in symbolic gestures and comparing their cognitive and
social-emotional behavior to a group of children who are not provided with this gesture-rich
environment, we may be able answer such questions as: (a) Do children who had earlier
developing symbol systems including emotion gestures show more competent reflective
functioning (i.e., explicit understanding of internal states such as emotions, desires, thoughts,
etc.) later in development? (b) Do children who had earlier developing symbol systems
including emotion gestures show more competent social functioning (i.e., prosocial behavior,
self-regulation, empathetic responses to peers, etc.) later in development?
Clinical Implications of the Use of Symbolic Gestures With Preverbal Children
There are several possible clinical implications for the use of symbolic gestures as a
communication tool between parents and children. Because further study first should verify
use of emotion gestures in additional populations of children, the clinical applications reliant
on children’s use of emotion gestures should be taken as speculative. I propose four clinically
applicable benefits of using symbolic gestures with preverbal children; each of which could
potentially be achieved within the context of either a parent–child or a nonparental caregiver–
child relationship.
VALLOTTON Page 14
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Using Symbolic Gestures in General: Enhanced Communication and Reduced
Frustration—Incidental to their seminal study of young children’s symbolic-gesture use,
Acredolo and Goodwyn (1988) heard from many parents about the relationship benefits of
using the Baby Signs Program—involving systematic use of symbolic gestures similar to the
use at the CCFS. Qualitative interviews with parents verified that the use of symbolic gestures
with their preverbal children eased their relationships (Acredolo & Goodwyn, 1992) during a
time of typically frequent frustration, when children’s receptive language is developing rapidly
but expressive language is not. Parents reported that the use of the Baby Signs Program allowed
their toddlers to communicate their needs and desires instead of becoming frustrated by less
effective pointing and nonspecific grunts. Though they did not include gestures for emotions
in their study, the results of Goodwyn and Acredolo’s (1988) parent interviews have promising
implications for enhancing parent–child communication about emotions. This finding should
be investigated experimentally to more clearly understand how the early parent–child
relationship could be affected by the use of symbolic gestures.
Using Symbolic Gestures for Emotions: Expressing, Conversing, and
Constructing Understanding of Emotion—In addition to the benefits of clearer adult–
child communication at an age during which frustration is the norm, the use of emotion gestures
specifically may have additional clinical applications; (a) Expressing emotions: Emotion
gestures could be used to encourage children to express their own emotions, in both positive
and challenging circumstances. Programs such as Early Head Start and NAEYC-accredited
childcare classrooms include expression of emotions as part of their curriculum standards;
however, these curricula and practices are typically developed only for verbal children. (b)
Conversing about emotions: Use of emotion gestures could create opportunities for children
to initiate and hold two-way conversations with their caregivers about their own and others’
emotions, potentially overcoming caregivers’ tendencies to avoid talking about negative
emotions with their very young children. (c) Constructing an understanding of emotions and
internal states: Vygotsky (1934/1986) proposed that words—or symbols—are mental tools
for actively constructing understanding of experiences and knowledge of the world. Thus, the
use of emotion gestures may give the preverbal child more than a set of communication tools
for expression of emotion, it may also provide a set of mental tools for constructing an
understanding of emotions. If so, then early use of emotion gestures may help advance the
development of a Theory of Mind for emotion.
SUMMARY
The current study provides a first report on preverbal children’s use of symbolic gestures to
represent emotion concepts. In addition to opening a new set of questions about preverbal
children’s explicit social-emotional concepts, these results further reveal the sophistication of
the infant’s social-emotional experience and understanding as well as demonstrate the potential
of symbolic gesture as a methodology for understanding the intersection of emotional and
cognitive development prior to the onset of vocal language.
Acknowledgments
I thank the children, parents, and staff of the Center for Child and Family Studies at the University of California-Davis
for their time and patience in the studies involved in this report, as well as the dedicated research assistants who so
competently collected, coded, and transcribed data, and without whom this study would have been impossible. I also
thank friend and colleague Livna Grinbaum for the use of her valuable data on toddlers’ symbolic gesture use. Finally,
I thank the mentors who guided me through both the research and writing processes to produce this article, including
Larry Harper, Linda Acredolo, Kathleen Grey, and Elizabeth Fivaz. Preparation of this article was supported in part
by a grant from the Eichhorn Family Trust and by National Institute of Child Health and Human Development Grant
1 F32 HD050040–01.
VALLOTTON Page 15
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
References
Acredolo L, Goodwyn S. Symbolic gesturing in language development: A case study. Human
Development 1985;28:40–49.
Acredolo L, Goodwyn S. Symbolic gesturing in normal infants. Child Development 1988;59:450–466.
[PubMed: 2452052]
Acredolo, L.; Goodwyn, S. Baby signs: How to talk with your baby before your baby can talk. Chicago:
Contemporary Books; 1992.
Adrian JE, Clemente RA, Villanueva L, Rieffe C. Parent–child picture-book reading, mothers’ mental
state language and children’s theory of mind. Journal of Child Language 2005;32:673–686. [PubMed:
16220639]
Aldwin, CM. Stress, coping, and development: An integrative perspective. New York: Guilford Press;
1994.
Astington, JW. The child’s discovery of the mind. Cambridge, MA: Harvard University Press; 1993.
Bakeman R, Adamson LB. Infants’ conventionalized acts: Gestures and words with mothers and peers.
Infant Behavior & Development 1986;9:215–230.
Baldwin D, Moses LJ. Links between social understanding and early word learning: Challenges to current
accounts. Social Development 2001;10:309–329.
Bates E, Camaioni L, Volterra V. The acquisition of performatives prior to speech. Merrill-Palmer
Quarterly 1975;21:205–226.
Bates, E.; O’Connell, B.; Shore, C. Language and communication in infancy. In: Osofsky, J., editor.
Handbook of infant development. Vol. 2. Oxford, England: Wiley; 1987. p. 149-203.
Bretherton I, Beeghly M. Talking about internal states: The acquisition of an explicit theory of mind.
Developmental Psychology 1982;18:906–921.
Carpenter M, Nagell K, Tomasello M. Social cognition, joint attention, and communicative competence
from 9 to 15 months of age. Monographs of the Society for Research in Child Development 1998;63
(4)
Cassidy KW, Werner RS, Rourke M, Zubernis LS, Balarman G. The relationship between psychosocial
understanding and positive social behaviors. Social Development 2003;12:198–221.
Crais E, Douglas DD, Campbell CC. The intersection of the development of gestures and intentionality.
Journal of Speech, Language, and Hearing Research 2004;47:678–694.
Cutting AL, Dunn J. Theory of mind, emotion understanding, language, and family background:
Individual differences and interrelations. Child Development 1999;70:853–865. [PubMed:
10446724]
de Rosnay M, Pons F, Harris PL. A lag between understanding false belief and emotion attribution in
young children: Relationships with linguistic ability and mothers’ mental-state language. British
Journal of Developmental Psychology 2004;22:197–218.
Dunn J, Bretherton I, Munn P. Conversations about feeling states between mothers and their young
children. Developmental Psychology 1987;23:132–139.
Fenson L, Dale P, Reznick JS, Bates E, Thal D, Pethick S. Variability in early communicative
development. Monographs of the Society for Research in Child Development 1994;59(5)
Flavell JH. Cognitive development: Children’s knowledge about the mind. Annual Review of Psychology
1999;50:21–45.
Goodwyn S, Acredolo L. Symbolic gesture versus word: Is there a modality advantage for onset of symbol
use? Child Development 1993;64:688–701. [PubMed: 8339689]
Goodwyn SW, Acredolo LP, Brown CA. Impact of symbolic gesturing on early language development.
Journal of Nonverbal Behavior 1989;24:81–103.
Harris, PL. Children and emotion: The development of psychological understanding. Cambridge, MA:
Basil Blackwell; 1989.
Iverson JM, Tencer HL, Lany J. The relation between gesture and speech in congenitally blind and sighted
language-learners. Journal of Nonverbal Behavior 2000;24:105–130.
Kopp C. Antecedents of self-regulation: A developmental perspective. Developmental Psychology
1982;18:199–214.
VALLOTTON Page 16
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Lagattuta KH. The development of young children’s understanding of the causes of emotions:
Experimental and natural language studies. Dissertation Abstracts International: Section B: The
Sciences and Engineering 1999;60(5B)
Lagattuta KH. Thinking about the past: Early knowledge about links between prior experience, thinking,
and emotion. Child Development 2001;72:82–102. [PubMed: 11280491]
Lagattuta KH, Wellman HM, Flavell JH. Preschoolers’ understanding of the link between thinking and
feeling: Cognitive cuing and emotional change. Child Development 1997;68:1081–1104. [PubMed:
9418227]
Murphy, LB.; Moriarty, AE. Vulnerability, coping, and growth: From infancy to adolescence. New
Haven, CT: Yale University Press; 1976.
Nelson CA. The recognition of facial expressions in the first two years of life: Mechanisms of
development. Child Development 1987;58:889–909. [PubMed: 3608661]
Piaget, J. Play, dreams and imitation in childhood. New York: Norton; 1952.
Pons F, Lawson J, Harris P, deRosney M. Individual differences in children’s emotion understanding:
Effects of age and language. Scandinavian Journal of Psychology 2003;44:347–353. [PubMed:
12887556]
Reich, PA. Language development. Englewood Cliffs, NJ: Prentice Hall; 1986.
Sorce JF, Emde RN, Campos JJ, Klinnert MD. Maternal emotional signaling: Its effects on the visual
cliff behavior of 1-year-olds. Developmental Psychology 1985;21:195–200.
Symons DK, Peterson CC, Slaughter V, Roche J, Doyle E. Theory of mind and mental state discourse
during book reading and story-telling tasks. British Journal of Developmental Psychology
2005;23:81–102.
Ungerer JA, Zelazo PR, Kearsley RB. Developmental changes in the representation of objects in symbolic
play from 18 to 34 months of age. Child Development 1981;52:186–195.
Vygotsky, LS. Thought and language. Kozulin, A., editor. Cambridge, MA: MIT Press; 19341986.
Werner, H.; Kaplan, B. Symbol formation. Oxford, England: Wiley; 1963.
Zeidner M, Matthews G, Roberts R, MacCann C. Development of emotional intelligence: Towards a
multi-level model. Human Development 2003;46:69–96.
Appendix
APPENDIX: Samples of Gestures in Each Category Used by Caregivers and
Children in the Study Population
Gesture Descriptions by Category
Category Gesture Description
Action/Event Diaper Change Pat hip
Play Closed fist with pinky and thumb sticking out, hand rotating
Wash Two hands rubbing together as if running after water
Emotion Happy Open hands, palm out, to frame sides of face
Mad Clawed hand running in front of face (accompanied by furrowed brows)
Sad Draw forefinger down cheek
Scared Open palm tapping chest
Feeling/Sensation Cold Arms to side, figts clenched, shaking as in “Brrr”
Gentle One hand stroking the opposite arm
Hurt Closed fist tapping chest
Loud Hands over ears
Sleepy Folded hands laid against cheek
VALLOTTON Page 17
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Gesture Descriptions by Category
Category Gesture Description
Nonsymbolic Point Using finger to point at something or someone
Wave Waving goodbye or hello
Object Ball Palm of hand down, motioning up and down as if bouncing a ball
Bird Arms or hands fluttering
Car Mimicking steering a wheel
Parent Dad Open palm, thumb tapping forehead
Mom Open palm, thumb tapping chin
Question Where? Palm of hands up next to shoulders
Request More Bunched fingers of both hands tapping together
Outside Fingers in claw shape (as if gripping doorknob), twisting
Snack Fingers of one hand together tapping mouth
Time Later Rotated right thumb/forefinger in open left hand
Popsicle Time Tapping back of palm to chin
Wait Right fist tapping open left hand
Yes/No No Head shaking from side to side as if saying “No”
Yes Nodding head up and down as if saying “Yes”
Note. All emotion, feeling, and time gestures are described rather than just a sample. Italicized gestures were used only by caregivers, as recorded either
in our data or in data collected by the program.
VALLOTTON Page 18
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Figure 1.
Order of appearance of gesture categories as measured by the average age at which each gesture
category was first observed to be used by each child.
VALLOTTON Page 19
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
VALLOTTON Page 20
TABLE 1
Number of Children Who Used Emotion, Feeling, and Time Gestures
No. of Children Using Gesture
Gesture Total Infants Toddlers Earliest Age of Observed Use
Emotion Happy 1 1 0 12.9 months
Mada0a0a09.0 monthsa
Sad 6 3 3 10.9 months
Scared 2 2 0 11.1 months
Feeling Sleepy 1 1 0 14.7 months
Cold 4 0 4 18.7 months
Gentle 1 0 1 19.2 months
Hurt 2 1 1 16.7 months
Loud 1 1 0 17.5 months
Time Popsicle Time 9 4 5 12.8 months
Wait 3 1 2 17.2 months
Later 5 1 4 15.2 months
aThough the mad gesture did not appear in our quantitative data, infants’ use of the mad gesture has appeared in systematically collected transcripts of child behavior in this cohort of children.
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
VALLOTTON Page 21
TABLE 2
Number of Episodes in Which Each Child Used Symbolic Gestures to Describe Internal States of Self and Other
Emotion Word Episodes Feeling Word Episodes
Child ID About Self About Other Insufficient Information About Self About Other Insufficient Information
Infant I-3 1 – 2 – – –
Infant I-9 – – 1 – – –
Infant I-10 4 3 4 2 – –
Toddler T-1 – – – 1 – –
Toddler T-3 – – – 6 – 2
Toddler T-4 – – – 1 – –
Toddler T-7 – 3 – – – –
Toddler T-9 1 – – 1 1 2
Toddler T-12 2 – – – – –
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
VALLOTTON Page 22
TABLE 3
Samples of Transcribed Observations of Children Communicating Emotion Concepts Through Symbolic Gesturesa
Transcription Category/Interpretation
Cathy (11.13 months) picked up a small spider stuffed animal on the floor and
looked at it for a while. She looked at her caregiver and said, “Me!” “Yeah, you are
holding a spider, Cathy,” her caregiver said. Cathy looked at the spider with her
fist pounding on her chest (the gesture for scared). Then, she looked back at the
caregiver. “It seems that you are telling me that you are sacred of the spider,” the
caregiver said. Cathy nodded without a smile.
1. Reflecting on own internal state in past experiences. In
this observation, Cathy spontaneously expresses her fear of
spiders to her caregiver, using the prop of a stuffed animal
spider and the “scared” gesture. Though many successful
communication exchanges end with a smile, this one does
not, which is congruent with the child’s affective state.
Alana (15.33 months) was lying on the changing table and made the gesture for
sleepy/nap. Her caregiver asked if she was tired; then Alana repeated the sleepy/
nap gesture. Another child’s caregiver was setting up a diaper table nearby and
oversaw Alana make the gesture for nap. The second caregiver asked, “Are you
tired, Alana? It’s Monday, and I’m always tired on Mondays. Are you tired on
Mondays?” Alana smiled and said, “Me?” in the tone of a question while pointing
to herself. The second caregiver said, “Yes, that’s right. You’re pointing to yourself
and saying, ‘Me.”’ Alana smiled and repeated “Me,” and pointed to herself,
followed again by the gesture for sleepy/nap.
2. Expressing own current internal state. In this
observation, Alana expresses her internal state, “sleepy,”
clarifying that it is herself that is sleepy by adding the word
“me.” There may be some question as to whether the sleep/
nap gesture describes an action or a feeling. Here the child
is talking about herself, and since she is not napping, she is
not saying “I’m napping,” using the gesture as a verb/action
(unless she were engaging in pretend or imaginative play);
thus, she must be saying “I’m sleepy.” She is not reflecting
on a behavior but rather an internal state.
The children and caregivers were in the outdoor play area. Elsie’s caregiver was
sitting on the edge of the sand area, watching some of the infants sitting in the sand.
Elsie (14.9 months) was walking around the infant yard by herself. Her mouth was
curled down and her eyes pointed downward. She had her right hand up near her
ear in a relaxed fist position (a relaxed version of the mom/dad gesture). She walked
slowly over toward the sandbox where her caregiver was sitting. When she came
within about 3 feet of her caregiver, she made eye contact, and very quickly she put
her thumb to her forehead with the other four fingers extended and gestured the
sign for dad. She turned her mouth down and pushed her eyebrows together. Her
caregiver said, “You look like you are thinking of your mom and dad. You look
sad.” Elsie turned and walked away.
3. Expressing internal thoughts that have emotional
salience. This child shows us that she knows what she’s
feeling (sad), and why (missing dad). In addition, she uses
a gesture, the relaxed “mom/dad” gesture, when she is
alone whereas most gestures observed are in a person-to-
person conversation (Is she using the gesture to think out
loud?)
Elsie also demonstrates that she knows that the caregiver
will need a more clear version of her gesture to understand
her—she articulates her gesture more clearly when she
makes eye contact with the caregiver.
Sophie (9 months) sat in the outdoor infant garden on a mat; a few manipulative
toys lay close, but she showed no interest in them. Sophie sat slumped at the
shoulders, crying with a low constant hum in her voice. One of her caregivers sat
down across from Sophie, and she paused in her sobs and looked at the caregiver’s
face with an open mouth and eyes drawn down at the sides. The caregiver said
“Sophie, I hear you crying and I am wondering if you are sad?” while she gestured
hear (index finger tap at the ear) and sad (finger tracing a tear down cheek from
eye). While focusing her gaze on the caregiver’s face, Sophie brought her right hand
with fingers slightly separated into a clawlike position up to her face, the gesture
for anger. The caregiver responded “I see you are angry, what can I do to help?”
while using the gesture for angry (claw hand over face drawing down) and what
(turned palms up with hands on top of knees).
4. Clarifying own internal state after caregiver
misinterpretation. This child distinguishes between “sad”
and “mad,” revealing the infants’ capacity not only to be
aware of internal states but to differentiate between two
different negative emotions.
Note. This observation did not come from a videotaped
transcript but was recounted by the Academic Child
Development Specialist at the CCFS rather than transcribed
from videotapes as were the other observations.
Allen (24 months) stood and watched Cathy as she cried after she had just fallen
down. He walked a little closer to her and then hesitated for a moment, still watching
Cathy with her caregiver. He squinted his eyes a little as he stood with his mouth
open a bit. Then he looked over at his caregiver and walked over to where she was
sitting. He pointed over to where Cathy was and made the gesture for sad, with his
finger under his eye. He also had a sad look on his face as his eyes drooped a little
and his mouth was shut, showing no sign of a smile. His caregiver said to him,
“Allen, you see that Cathy is sad, don’t you?” He looked at the caregiver with the
same expression on his face and slowly nodded his head. She said back to him, “I
think she fell down and bumped her head,” as she brought her hand to her head.
He looked at his caregiver and did the same gesture, bringing his hand to his
head and said, “bumped head.”
5. Noticing others’ emotion-related behaviors. In this
observation, you see that the toddler shows an intense
interest in the other child’s apparent emotional state, and
inquires about it with the caregiver. He uses the “sad”
gesture to initiate a conversation with his caregiver about
the causes of being sad.
Ellie (15.5 months) stopped as she was walking across the room, and made the
gesture for hear. Her caregiver commented that she heard Billy crying. Ellie then
made the gesture for sad. The caregiver said, “Yes, Ellie, I think Billy is sad. Emily
is going to hold him and make him feel better.” Ellie looked at her caregiver and
made the signs for bottle and sleepy/nap. The caregiver said, “I think you are right.
Maybe Billy needs a bottle and a nap,” while repeating the gestures bottle and
sleepy/nap. Ellie looked at the caregiver and pointed at Billy, as she again gestured
sad, bottle, and sleepy/nap directly following one another.
6. Reflecting on cause or solution for the emotions of
another. In this observation, Ellie demonstrates an
understanding of the reasons someone might feel bad, or
what would make him feel better.
aAll gestures used by children in the following transcripts are bolded and underlined for easy reference. Names of children have been changed, but ages
are accurate.
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
VALLOTTON Page 23
TABLE 4
Results of Binomial Statistics Testing Whether Children’s Emotion, Feeling, and Time Gestures Were Imitations of
Adults’ Gestures
Binomial Test Results
Gesture Category Conversational Context Observed Expected p (two-tailed)
Emotion Response/Imitate 1 6.5 <.01
Initiate, Continue, Reply 12 6.5
Total 13
Feeling Response/Imitate 0 2 .125a
Initiate, Continue, Reply 4 2
Total 4
Time Response/Imitate 0 4 <.01
Initiate, Continue, Reply 8 4
Total 8
aThere were too few instances of feeling/sensation gestures to obtain a valid estimate of statistical significance. Minimum observations to detect statistical
significance at the p < .05 level using binomial test is 6.
Infant Ment Health J. Author manuscript; available in PMC 2009 September 11.
