Content uploaded by Michelle Eskritt
Author content
All content in this area was uploaded by Michelle Eskritt
Content may be subject to copyright.
Journal of Nonverbal Behavior 27(1), Spring 2003 䊚 2003 Human Sciences Press, Inc. 25
DO ACTIONS SPEAK LOUDER THAN WORDS?
PRESCHOOL CHILDREN’S USE OF THE VERBAL-
NONVERBAL CONSISTENCY PRINCIPLE DURING
INCONSISTENT COMMUNICATIONS
Michelle Eskritt and Kang Lee
ABSTRACT: The present study investigated whether preschool children could use
the conventional “actions speak louder than words” principle (or the verbal-nonver-
bal consistency principle) to process information in situations where verbal cues
contradict nonverbal cues. Three-, 4-, and 5-year-olds were shown a video in
which an actor drank a beverage and made a verbal statement (e.g., “I like it”) that
was inconsistent with her emotional expression (e.g., frowning), and were asked
whether the actor liked or disliked the beverage. If children used the verbal-nonver-
bal consistency principle, they should respond according to the information con-
veyed by the actor’s emotional expression. Results showed that when the message
was more naturalistic, the majority of children tended to respond based on the
actor’s verbal message. However, when the inconsistency between the verbal and
nonverbal messages was made salient, more children appeared to rely on the non-
verbal cue. Younger children’s reliance on verbal cues reported in previous re-
search may be partly explained by the salience of the verbal message.
KEY WORDS: display rules; emotional expression; inconsistent communication;
nonverbal communication; preschool children; verbal-nonverbal consistency prin-
ciple.
Everyday social communication is a dynamic process in which various
messages are communicated either concurrently or across time and space
between a sender and a receiver. These messages are generally consistent
Michelle Eskritt, Mount St. Vincent University, Canada. Kang Lee, University of Califor-
nia, San Diego.
Address correspondence to Dr. Kang Lee, Department of Psychology University of Cali-
fornia, San Diego, La Jolla, CA 92093-0109; e-mail: kang噝psy.ucsd.edu.
This research was supported by two separate grants to the authors from the Social Sci-
ences and Humanities Research Council of Canada. The authors thank the children and their
daycares for participating in the present study, and Laura Cavanagh and Julie Board for their
assistance in data collection. The authors’ gratitude is also extended to the three anonymous
reviewers for their constructive comments on earlier versions of this article.
26
JOURNAL OF NONVERBAL BEHAVIOR
with each other. Sometimes, however, the sender conveys inconsistent
messages. In inconsistent communications, it is essential for the receiver to
discern the discrepancy between the messages and use the different mes-
sages to interpret their meaning. One of the major inconsistent communi-
cative situations involves the verbal message being different from, or con-
tradictory to, the nonverbal message. In this type of situation, adults tend to
use the so-called verbal-nonverbal consistency principle (Rotenberg, Si-
mourd, & Moore, 1989). This principle assumes that one’s verbal and non-
verbal behaviors normally convey the same message. When there is a dis-
crepancy, adults tend to treat the nonverbal rather than verbal message as
the more reliable source of information (Rotenberg et al., 1989). It has
been suggested that this principle is based on a commonsense assumption
that “actions speak louder than words” and the nonverbal message, not the
verbal message, reveals the true intentions of the sender (Demorest, Meyer,
Phelps, Gardner, & Winner, 1984).
Understanding the verbal-nonverbal consistency principle is critical to
successful social interaction because discrepant verbal and nonverbal mes-
sages are not atypical in everyday communication. Inconsistent verbal-
nonverbal communication occurs frequently and serves various useful in-
terpersonal functions, for example, to amuse (joke), to deceive (lie), or
when one is experiencing conflicted feelings. With appropriate use of the
verbal-nonverbal consistency principle, individuals can infer the truth of
information contained in the sender’s inconsistent messages and respond
accordingly. Inappropriate use of the verbal-nonverbal principle, on the
other hand, may lead to communication failure (e.g., a joke is taken se-
riously), or an undesirable consequence (e.g., being deceived by the
sender’s lie).
However, while it is well established that adults use the verbal-nonver-
bal consistency principle in inconsistent communications (for reviews, see
Blanck & Rosenthal, 1982; Volkmar & Siegel, 1982), there is little evidence
regarding when children learn to use this principle. Limited research to
date has found that children until 9 to 10 years of age are unable to use the
verbal-nonverbal consistency principle in communicative situations where
nonverbally expressed emotions are inconsistent with verbally expressed
emotions (Demorest et al., 1984; DePaulo & Jordan, 1982; DePaulo, Jor-
dan, Irvine, & Laser, 1982; Friend, 2000, 2001, in press; Friend & Bryant,
2000; Rotenberg et al., 1989; Volkmar & Siegel, 1982). They tend to rely
on verbal rather than nonverbal cues when encountering inconsistent com-
munication. In other words, they have a “lexical bias” (Friend & Bryant,
2000), or follow an unconventional assumption that “words speak louder
than actions” (Demorest et al., 1984). Although recent evidence shows that
children around 7 years of age begin to appreciate the contribution of non-
27
MICHELLE ESKRITT, KANG LEE
verbal cues in determining the meaning of an inconsistent message (Friend,
2000), most studies as cited above suggest that the appropriate use of the
verbal-nonverbal consistency principle begins after 9–10 years of age.
Children’s difficulty with the verbal-nonverbal consistency principle
may be due to their lack of some skills that are necessary for the appropri-
ate use of this principle. To use the verbal-nonverbal consistency principle
appropriately, individuals must be able to recognize both verbal and non-
verbal cues (e.g., emotions). They must also detect the discrepancies be-
tween the two channels if such discrepancies indeed exist. In addition,
they need to understand the distinctions between appearance and reality,
in particular, the distinction between expressed emotions and felt emo-
tions. Only with these prerequisite skills, can individuals begin to use the
verbal-nonverbal consistency principle to make decisions about which
channel to rely on during inconsistent communication.
Existing research has consistently shown that most of these prerequi-
site skills have been acquired by preschool years. For example, the ability
to interpret others’ nonverbal expressions of emotions emerges in infancy
(Walker-Andrews & Dickson, 1997). Infants are not only able to recognize
and differentiate between different emotions but also can use another’s
emotions to guide their own actions (Mumme, Fernald, & Herrera, 1996;
Walden, 1991) and make predictions about others’ desires and preferences
(Repacholi & Gopnik, 1997). By the end of the preschool period, children
are able not only to recognize emotions but also to label them appro-
priately (Camras & Allison, 1985; Denham & Couchard, 1990; Field &
Walden, 1982). Preschoolers, furthermore, demonstrate an understanding
of the causes of emotions (Denham & Zoller, 1991; Dunn & Hughes, 1998;
Fabes, Eisenberg, Nyman, & Michealieau, 1991). By the end of preschool
years, children also have acquired the necessary language skills to compre-
hend verbally expressed emotions (Banerjee, 1997a).
The ability to detect the discrepancy between verbal and nonverbal
displays also emerges in infancy (for a review, see Walker-Andrews, 1997)
and reaches an adequate level of sophistication by preschool years
(Gnepp, 1983, 1989). With regard to the understanding of the appearance
and reality distinction, 3-year-olds under some conditions are already able
to appreciate what is real and what is apparent (Sapp, Lee, & Muir, 2000).
Specifically, with regard to the distinction between expressed and felt emo-
tions, Bannerjee (1997a, 1997b) found that even 3-year-olds could distin-
guish between how a person feels and the emotion that a person expresses
in certain situations. By 6 years of age, children are able to infer correctly
the true emotions felt by another individual in spite of the individual’s overt
expression and understand the individual’s reasons for doing so.
Given these findings in the literature, it is surprising that the existing
28
JOURNAL OF NONVERBAL BEHAVIOR
studies on inconsistent communication find children to have difficulty in
using the verbal-nonverbal consistency principle until 10 years of age. One
possibility of such delay may be due to the methodology used in these
studies. They may have underestimated young children’s ability to use the
verbal-nonverbal consistency principle due to several limitations. First, in
some of the studies, the inconsistent nonverbal information conveyed to
children might have been too subtle or artificial (e.g., Demorest et al.,
1984; DePaulo et al., 1982; Volkmar & Siegel, 1982). For example, emo-
tional expression has been presented in a storybook format. Because emo-
tional expression is dynamic, the story-telling method could severely re-
strict the delivery of emotional information, hindering children’s use of the
verbal-nonverbal consistency principle (see Rotenberg & Eisenberg, 1997).
Children have also been asked to identify whether or not a character is
lying as a measure of their understanding of the verbal-nonverbal consis-
tency rule (e.g., Rotenberg et al., 1989). However, as indicated earlier, in-
consistent verbal and nonverbal communication does not automatically
entail lying; jokes and sarcasm also involve verbal and nonverbal inconsis-
tencies. Furthermore, while many preschoolers understand deception as
the intent to instill false beliefs into another’s mind (Chandler, Fritz, &
Hala, 1989), research has consistently shown that they have a limited un-
derstanding of the concept of lying (Bussey, 1992; Peterson, 1995; Peter-
son, Peterson, & Seeto, 1983; Piaget, 1932/1965; Strichartz & Burton,
1990; Wimmer, Gruber, & Perner, 1984; for a review, see Lee, 2000). They
often mistakenly label verbal transgressions (e.g., swearing) or factually
false statements as lies, particularly when the speakers’ intentional states
are not highlighted for them (Siegal & Peterson, 1996, 1998). It is possible
that children in the Rotenberg et al. (1989) study could use the verbal-
nonverbal consistency principle but failed to demonstrate this ability be-
cause of their difficulty in labeling an incongruent communicative act as
lying.
However, a third explanation is also possible. One could argue that
emotional expressions displayed in previous research might not be as sa-
lient as the verbal statements. Therefore, if the nonverbal cues were made
more salient, children may be more likely to use them to interpret the
situation. Hence, the reason that the children relied on words instead of
actions might not be because they followed the principle of “words speak
louder than actions.” Rather, they might simply rely on the more salient
cue.
The purpose of the present study was to specifically address this possi-
bility. We examined whether 3-, 4-, and 5-year-olds would rely on nonver-
bal rather than verbal cues to determine the true state of affairs when the
29
MICHELLE ESKRITT, KANG LEE
inconsistency between the cues was made obvious to them. Children were
shown video vignettes in which an actor drank a beverage and made a
verbal statement (e.g., “I like it” or “I don’t like it”) that was inconsistent
with her emotional expression (e.g., frowning or smiling). One group of
children saw videos in which the actor expressed emotional expressions
naturalistically while making verbal statements about her liking or disliking
of a drink (the Naturalistic Condition). Another group of children viewed
videos in which the actor exaggerated their emotional expressions in order
to make it salient to the participants (the Exaggerated Condition). Unlike
Rotenberg et al. (1989), children were not required to label an inconsistent
communicative act as lying. Rather, they were asked directly whether the
actor liked or disliked the beverage.
We hypothesized that if young children can use the verbal-nonverbal
consistency principle appropriately, they should respond according to
whatever is conveyed by the actor’s emotional expression. In contrast, if
children have a lexical bias, they should respond according to the actor’s
verbal statement. Further, if children’s response bias, if one exists, is due to
their reliance on a more salient cue, children should be less likely to rely
on the actor’s verbal statements in the Exaggerated Condition than in the
Naturalistic Condition. Alternatively, if children indeed have a lexical bias
and are following the notion that “words speak louder than actions,” they
should rely on the actor’s verbal statement, as did the children in previous
research, regardless of the salience of the nonverbal cues in the present
study.
Method
Participants
Children were recruited from local day-cares. Thirty-three 3-year-olds
(M⳱3.5 years, SD ⳱0.3; 9 boys and 19 girls), twenty-eight 4-year-olds
(M⳱4.4 years, SD ⳱0.3; 9 boys and 19 girls), and thirty 5-year-olds
(M⳱5.5 years, SD ⳱0.4; 17 boys and 13 girls) participated. Children
were divided into two groups; those who saw the Naturalistic video and
those who saw the Exaggerated video.
Materials and Procedure
Children were seen individually and shown a video on a 24-inch tele-
vision of a vignette depicting a female actor tasting different colored drinks.
The vignette had an introduction and three experimental phases. The actor
30
JOURNAL OF NONVERBAL BEHAVIOR
first introduced herself and told the viewers that they were going to play a
game where they had to guess whether she liked or disliked a drink.
The One-Channel Phase always came first. In this phase, the actor had
four different-colored drinks in front of her. They were lined up side by side
and the actor tasted each of the drinks in turn. Drinks were made from
KoolAid䉸 powder mixes. Half of the drinks were sweetened according to
the instruction on the package and half of them were not sweetened and
therefore tasted very sour. For each drink, the actor either uttered a verbal
statement with a neutral facial expression or exhibited a facial expression
with no verbal statement. These verbal or facial reactions were either posi-
tive (when the drink was sweetened) or negative (when the drink was not
sweetened). Thus, for the sweetened drinks, the actor reacted with either a
positive verbal statement (i.e., “I like it”) or a positive nonverbal facial
expression (i.e., a smile); for the unsweetened drinks, the actor made either
a negative verbal statement (i.e., “I don’t like it”) or a negative facial ex-
pression (i.e., a frown). For each trial, after the actor tasted a drink, she
resumed a neutral expression. At this point, the video was paused, and the
children were asked: “Do you think she liked the drink or not?” The video-
cassette was played on a videocassette recorder with a still pause feature
so that the presentation could be paused to allow ample time for the child
to answer. After the actor finished tasting the four drinks, a blank screen
appeared for a few seconds, which was followed by the actor drinking
another four new drinks. The above four One-Channel trials were repeated
in a different, randomized order.
After the One-Channel Phase, two Two-Channel Phases followed in
which children were shown videos in which the actor displayed both ver-
bal and nonverbal behaviors. Children were assigned to either of two
groups. In the Naturalistic Group, the experimental phase consisted of two
within-subject conditions. In the Two-Channel Consistent condition, four
new drinks were in front of the actor (two sweetened and two unsweet-
ened). She reacted to two sweetened drinks with a smile and the statement
“I like it” (the positive valence trials), and two unsweetened drinks with a
frown and the statement “I don’t like it” (the negative valence trials). In the
Two-Channel Inconsistent condition, the actor drank another four new
drinks (two sweetened and two unsweetened). She reacted to the two
sweetened drinks with the statement “I don’t like it” paired with a smile
(the positive valence inconsistent trials), and the two unsweetened drinks
with the statement “I like it” with a frown (the negative valence inconsi-
stent trials). For all the trials, the actor was instructed not to mask her
genuine reaction to the drinks and to display her reaction as naturally as
possible (see Figure 1 for examples). The order of the four drinks for each
31
MICHELLE ESKRITT, KANG LEE
Figure 1. Examples of displays used for the Naturalistic Condition.
condition was randomized. Two versions of the video were used with each
video containing the drink tasting vignettes in a different randomized order.
Each child was randomly assigned to one of the two versions of the video
as well as to one of two orders of two channel experimental conditions,
either the Inconsistent condition first or the Inconsistent condition second.
For the Exaggerated group, the content of the video presentation was
identical to that used for the Naturalistic group, except that the emotional
behavior was exaggerated. The positive emotional display was modified
such that the actor smiled, rubbed her stomach, licked her lips, and mak-
ing a sound “mmmm.” The negative emotional display included the actor
frowning, shuddering, wiping her mouth, and making a sound resembling
“blech.” Once again, two versions of the video were used with each video
containing the drink tasting vignettes in a different randomized order. Each
child was randomly assigned to one of the two versions of the video as
well as to one of two orders of two channel experimental conditions, either
the Inconsistent condition first or the Inconsistent condition second. The
entire session took approximately 10 minutes. Upon completion of the ex-
periment, participants were given a small gift and thanked for their help.
To ensure that the difference between the Exaggerated and Naturalistic
videos was due to the salience of the nonverbal channel, a control group
of 24 adults (M⳱29.5 years, SD ⳱9.2) was asked to rate the intensity of
the verbal and nonverbal messages. Half of the participants watched a ver-
sion of the Naturalistic video, while the other half watched a version of the
Exaggerated video. Adults were asked to rate the intensity of positive or
negative affect associated with the verbal and nonverbal messages of the
actor for each of the test trials. They rated the intensity of emotion on a 7
point scale, with a rating of 1 for “strongly negative,” 7 for “strongly posi-
tive,” and 4 for “neutral.” For the verbal message, the television screen was
32
JOURNAL OF NONVERBAL BEHAVIOR
covered so participants heard the actor but could not see her nonverbal
behavior. In rating the nonverbal behavior, the television screen was un-
covered but the sound was muted so that participants could not hear what
the actor was saying.
The intensity ratings were transformed so they were on the same scale
of 0 for a neutral rating to 3 for very strong intensity. Analyses of Variance
(ANOVAs) were then conducted to compare the intensity ratings by each
group for the different valence trials for the verbal and nonverbal channels.
There were no significant differences between the Naturalistic and Exagger-
ated groups in terms of their rating of the intensity of the verbal message
regardless of whether the valence was positive or negative. However, there
was a significant difference between the two groups for the nonverbal
channel during both the positive and negative valence trials, F (1, 18) ⳱
9.67, p⬍.01. Therefore, participants did find the nonverbal message to be
more salient in the exaggerated condition.
Results
Preliminary analyses showed no effect for order of conditions and therefore
the data were collapsed across this variable in subsequent analyses. For the
One-Channel trials and the trials in the Two-Channel Consistent conditions
where the actor’s verbal and nonverbal cues conveyed a consistent mes-
sage, most children in both groups performed perfectly (see Table 1). In
cases where children did make mistakes, chi square analyses indicated that
the children as a group still performed significantly above chance (p⬍
.05). Therefore, children across all three age groups were able to use both
the emotional and verbal cues separately or together to interpret whether
the actor liked or disliked the drink.
To examine the performance of children in the Inconsistent conditions,
children were divided based on which cues they chose to rely on across
the two trials (see Figure 2). Children’s responses were divided into those
who relied on verbal cues for at least one trial, and those that relied en-
tirely on nonverbal cues to solve the task. This division of response types in
the present study allowed us to examine the pattern of cue types children
at different ages relied on.
Two logistic regression analyses were conducted to compare the per-
formance of children across age and condition (Naturalistic versus Exagger-
ated) for the positive and negative valence inconsistent trial types. The pre-
dicted variable was whether a child relied on verbal cues for at least one
trial or relied entirely on nonverbal cues. In the regression model for the
33
MICHELLE ESKRITT, KANG LEE
TABLE 1
The Mean Scores and Standard Deviations for the One and Two Channel Positive and Negative Valence
Conditions for Each Age and Group
One channel Two channels
Year
Verbal-
positive
Verbal-
negative
Nonverbal-
positive
Nonverbal-
negative
Consistent-
positive
Consistent-
negative
3-year-olds
Naturalistic 2.0 (0) 2.0 (0) 1.6 (.5) 1.3 (.7) 2.0 (0) 1.9 (.5)
Exaggerated 2.0 (0) 2.0 (0) 2.0 (0) 1.9 (.4) 2.0 (0) 2.0 (0)
4-year-olds
Naturalistic 2.0 (0) 2.0 (0) 1.9 (.4) 2.0 (0) 2.0 (0) 2.0 (0)
Exaggerated 2.0 (0) 1.9 (.4) 2.0 (0) 1.9 (.3) 2.0 (0) 2.0 (0)
5-year-olds
Naturalistic 2.0 (0) 2.0 (0) 2.0 (0) 2.0 (0) 2.0 (0) 2.0 (0)
Exaggerated 2.0 (0) 2.0 (0) 2.0 (0) 2.0 (0) 2.0 (0) 2.0 (0)
34
JOURNAL OF NONVERBAL BEHAVIOR
positive valence inconsistent trials, only the condition effect was signifi-
cant,
2
(N⳱86, df ⳱3) ⳱13.25, p⬍.005, Nagelkerke R
2
⳱.19. For
the negative valence inconsistent trials, there was also a significant effect
for condition,
2
(N⳱86, df ⳱3) ⳱12.94, p⬍.01, Nagelkerke R
2
⳱.20,
and a significant effect of age,
2
(N⳱86, df ⳱3) ⳱12.94, p⬍.01, Na-
gelkerke R
2
⳱.20.
As shown in Figure 2, the majority of children in the Naturalistic con-
dition, across age groups, relied predominately on verbal cues to solve the
task regardless of the valence of the trial. Children in the Exaggerated con-
dition, on the other hand were more likely to use to nonverbal cues. Fur-
thermore, a priori contrasts revealed that 5-year-olds were more likely to
rely on nonverbal cues compared to 3-year-olds during the negative val-
ence inconsistent trials.
The number of children in the Exaggerated condition relying on non-
verbal cues appears to be greater in the negative valence inconsistent trials
compared to the positive valence inconsistent trials. A sign test was there-
fore conducted to examine whether the valence of the trial had any effect
on children’s performance. Children were assigned one point for each trial
they relied on nonverbal cues to determine whether or not the actor liked
the beverage. Children’s scores for the positive and negative valence in-
consistent trials were then compared. No significant differences were
found for valence type for any age for the Naturalistic condition. However,
the distribution of scores for 3-year-olds was significant different from
chance (p⬍.05) for the Exaggerated condition. As shown in Figure 2,
3-year-olds were more likely to rely on the nonverbal cue in the negative
valence inconsistent trials than in the positive valence inconsistent trials.
However, no significant differences were found for 4- and 5-year-olds.
Older children tended to continue using the nonverbal cue to respond re-
gardless of the valence of the cue.
Discussion
The purpose of the present experiment was to determine if young children
would use the verbal-nonverbal consistency principle to interpret an ac-
tor’s mixed message. We used dynamic displays of nonverbal behavior and
did not require children to recognize and label a lie in order to make the
task easier for children. Despite these modifications, in the Naturalistic
condition, where the nonverbal cues were more subtle, our findings appear
to be consistent with the results from previous research that suggest youn-
ger children have a lexical bias (Demorest et al., 1984; DePaulo & Jordan,
35
MICHELLE ESKRITT, KANG LEE
Figures 2a and 2b. The number of 3-, 4-, and 5-year-olds in the Naturalistic and
Exaggerated Conditions that relied on verbal or nonverbal cues on the positive and
negative valence inconsistent trials.
1982; DePaulo et al., 1982; Friend, 2000, 2001, in press; Rotenberg et al.,
1989). Children in the Naturalistic condition, regardless of age, relied
strongly on the actor’s verbal statement rather than her emotional expres-
sion to determine the true state of affairs. The children appeared to follow a
rule opposite to the verbal-nonverbal consistency principle, that is, that
“words speak louder than actions” (Demorest et al., 1984; Rotenberg et al.,
1989). This lexical bias was clearly not due to young children’s inability to
decode the actor’s emotional expressions. When the same emotional ex-
pressions were displayed alone in the one-channel condition, children in-
36
JOURNAL OF NONVERBAL BEHAVIOR
ferred correctly the actor’s liking or disliking of a given drink based on her
emotional expressions. The problem also cannot be attributed to children’s
problem with identifying a lie as children were only asked to decide
whether the actor liked or disliked the drink.
However, the results from the Exaggerated condition revealed that the
salience of an actor’s nonverbal displays had a significant impact on young
children’s reliance on nonverbal cues for truth information. When the actor
exaggerated her emotional expressions to make the nonverbal cues more
salient in the Exaggerated condition, children of all ages started to rely
more on the actor’s nonverbal cues to determine whether the actor truly
liked the drinks. In comparison with the children in the Naturalistic condi-
tion, those in the Exaggerated group were less reliant on the actor’s verbal
statements.
The use of nonverbal cues by preschoolers as young as 3 years of age
to interpret inconsistent messages is much earlier than suggested by pre-
vious research that did not find this ability until 9 to 10 years of age (De-
morest et al., 1984; DePaulo & Jordan, 1982; DePaulo et al., 1982; Roten-
berg et al., 1989). This finding suggests that preschool children might not
always have a lexical bias, although they still did not rely categorically on
the “actions speak louder than words” principle. Except for some 3-year-
olds’ performance in the positive valence inconsistent trials, the salience of
the nonverbal cues had a significant impact on most children’s decisions
about the actor’s true state of liking or disliking. When the nonverbal cues
are salient, they can use nonverbal cues for making such decision. Thus,
one of the reasons for young children’s reliance on the verbal cues during
inconsistent communication reported in previous studies might be because
verbal messages were more saliently conveyed than the nonverbal cues.
However, the finding that the valence type (i.e., positive versus nega-
tive expressions) affected how 3-year-olds responded in the task suggests
that salience cannot explain the pattern of results entirely. In the Exagger-
ated group, 3-year-olds’ reliance on the actor’s emotional expressions over
words was most pronounced in the negative valence inconsistent trials
(where the actor stated that she liked the drink but displayed negative emo-
tional responses). One possible explanation is that children’s acquisition of
display rules may have played a role. Display rules are social conventions
about the appropriateness of emotional expressions for different situations
(Saarni, 1979). They need to learn in which social context one needs to use
a particular display rule and adopt apparent emotional expressions that
differ from their truly felt emotions (e.g., faking pleasantry when receiving
an unwanted gift). Banerjee (1997a) found that even 3-year-olds appreci-
ated some basic reasons for inconsistencies in individuals’ displayed emo-
37
MICHELLE ESKRITT, KANG LEE
tional expressions and their real emotions, though they did not perform as
well as the 4- and 5-year-olds. Similar to our findings, several studies (Ban-
erjee, 1997a; Friend & Bryant, 2000; Gnepp & Hess, 1986) have shown
that children understand situations in which one must conceal negative
emotions (i.e., negative valence situations) at an earlier age than the situa-
tion where one conceals positive emotions, likely due to the influence of
socialization of cultural conventions.
The negative valence inconsistent trials in the present study might re-
semble the politeness situations that children often encounter in their daily
interactions. The politeness situation requires individuals to make state-
ments that are incongruent with their true feelings and beliefs to avoid
harming interpersonal relations. More often than not, the felt emotion tends
to be negative but the verbal statement is positive. Children may be ex-
posed to this type of inconsistent communication as displayed by their
parents. They are often encouraged and even taught to engage in this type
of communication themselves (e.g., to fake pleasantry for receiving an un-
desirable Christmas gift from grandparents; Gnepp & Hess, 1986; Talwar &
Lee, 2002). This suggestion is, however, rather speculative given a limita-
tion of the present study. The present inconsistent condition (negative val-
ence inconsistent trials) may not be perceived as a politeness situation be-
cause no necessary contextual information was given to suggest that the
situation depicted in the video called for politeness. This is an issue that
needs to be addressed in future studies. Also, gender differences have been
found in the use of display rules (Cole, 1986; Saarni, 1984; Underwood,
Hurley, Johanson, & Mosley, 1999; Zeaman & Garber, 1996). We could
not examine gender differences due to the uneven number of males and
females in each group, an issue that should also be addressed in the future.
It should be noted that 3- to 5-year-olds have yet to learn to use the
verbal-nonverbal consistency principle appropriately. To do so, they must
focus, and rely on nonverbal cues regardless of their salience during incon-
sistent communication. Our results suggest that children under 5 years of
age do not yet have this ability, which may be acquired during the elemen-
tary school years. As the present study showed, many preschoolers still
have a strong lexical bias under some conditions.
Exactly what factors contribute to children’s lexical bias and their
eventual loss of it is largely unknown and yet to be determined by future
studies. Nevertheless, existing research on this issue, and related areas of
development provides sufficient information for informed speculations. A
number of factors may play an important role. For example, level of cogni-
tive development may affect how young children handle inconsistent mes-
sages. Part of their difficulty could be in trying to cope with two different
38
JOURNAL OF NONVERBAL BEHAVIOR
representations at one time, a problem with metarepresentation (Perner,
1991). Young children may have difficulty holding the different information
from the verbal and nonverbal channels simultaneously. They would,
therefore, be forced to choose one representation to focus on. With in-
creased cognitive capacity after preschool years (Case, 1992), children
may become capable of metarepresentation, which in turn would allow
them to assess and integrate inconsistent verbal and noverbal messages
and make appropriate decisions about which message to believe.
The ability for metarepresentation is necessary but not sufficient for
children to overcome their over-reliance on the verbal channel. Children
also need to understand that the true meaning of a message is not solely
determined by what is said. Rather, one must take the sender’s intentions
into consideration when determining the meaning of another’s communi-
cative acts (Lee, 2000). In other words, children’s knowledge about inten-
tionality or theory of mind understanding needs to reach an adequate level.
Research has found that preschoolers have just begun to recognize the
importance of the speaker’s intention in communication (e.g., Lee & Cam-
eron, 2000). With increased theory of mind understanding beyond the pre-
school years, children may develop the ability to infer a speaker’s commu-
nicative intention when the speaker communicates inconsistently. With this
intentional information, children can then make decisions about whether
to believe or disbelieve an individual’s verbal statement.
Language development also plays an important role in determining
which cues children use in interpreting inconsistent communication. Re-
cent research (e.g., Friend, 2000, 2001, in press) has found that infants
tend to rely on emotional expression to interpret the situation where verbal
and nonverbal messages are inconsistent. However, by the end of infancy
young children begin to gradually shift to relying more on verbal informa-
tion. Friend (2001) found that this shift was due to young children’s in-
creasing knowledge about their first language. The more advanced young
children are in language acquisition (e.g., vocabulary), the more likely they
are to rely on the verbal cue for truth information during inconsistent com-
munication. However, as children’s language ability further develops, their
over-reliance on the verbal channel weakens, which begins around 7 years
of age (Friend, 2000).
The difference in performance between the different valence condi-
tions in the present study suggests that cognitive and language factors are
not the sole influences in the acquisition of the verbal-nonverbal consis-
tency principle. Knowledge of cultural conventions may play an important
role as well (Lee, 2000). In addition, children’s understanding of display
rules (Banarjee, 1997a, 1997b; Josephs, 1994), and their familiarity with in-
39
MICHELLE ESKRITT, KANG LEE
consistent communicative situations (e.g., Halberstadt, Denham, & Duns-
more, 2001) could also affect children’s interpretation of inconsistent mes-
sages. As these studies have shown, such knowledge often develops gradu-
ally beyond preschool years. However, relying on salience rather than on
lexical information exclusively by some of the children in the present study
suggests that young children have already acquired the cognitive flexibility
to overcome the lexical bias. This flexibility is needed in order for the
factors identified above to exert their influence on children, which in turn
will lead children to overcome the lexical bias and to develop the ability to
use appropriately the verbal-nonverbal consistency principle.
References
Banerjee, M. (1997a). Hidden emotions: Preschoolers’ knowledge of appearance-reality and
emotion display rules. Social Cognition, 15, 107–132.
Banerjee, M. (1997b). Peeling the onion: A multi-layered view of children’s emotional devel-
opment. In S. Hala (Ed.), The development of social cognition (pp. 241–272). London:
Psychology Press.
Blanck, P. D., & Rosenthal, R. (1982). Developing strategies for decoding “leaky” messages:
On learning how and when to decode discrepant and consistent social communications.
In R. S. Feldman (Ed.), Development of nonverbal behavior in children (pp. 203–229).
New York: Springer-Verlag.
Bussey, K. (1992). Lying and truthfulness: Children’s definitions, standards, and evaluative
reactions. Child Development, 63, 129–137.
Camras, L. A., & Allison, K. (1985). Children’s understanding of emotional facial expressions
and verbal labels. Journal of Nonverbal Behavior, 9, 84–94.
Case, R. (1992). The mind’s staircase: Exploring the conceptual understandings of children’s
thought and knowledge. Hillsdale, NJ: Erlbaum.
Chandler, M., Fritz, A. S., & Hala, S. (1989). Small-scale deceit: Deception as a marker of
two-, three-, and four-year-olds’ early theories of mind. Child Development. 60, 1263–
1277.
Cole, P. M. (1986). Children’s spontaneous control of facial expressions. Child Development,
57, 1309–1321.
Demorest, A., Meyer, C., Phelps, E., Gardner, H., & Winner, E. (1984). Words speak louder
than actions: Understanding deliberately false remarks. Child Development, 55, 1527–
1534.
Denham, S. A., & Couchoud, E. A. (1990). Young preschoolers’ understanding of equivocal
emotion situations. Child Study Journal, 20, 171–192.
Denham, S. A., & Zoller, D. (1991). “When my hamster died, I cried”: Preschoolers’ attribu-
tions of the causes of emotions. Journal of Genetic Psychology, 152, 371–373.
DePaulo, B. M., & Jordan, A. (1982). Age changes in deceiving and detecting deceit. In R. S.
Feldman (Ed.), Development of nonverbal behavior in children (pp. 151–180). New
York: Springer-Verlag.
DePaulo, B. M., Jordan, A., Irvine, A., & Laser, P. S. (1982). Age changes in the detection of
deception. Child Development, 53, 701–709.
Dunn, J., & Hughes, C. (1998). Young children’s understanding of emotions within close
relationships. Cognition and Emotion, 12 171–190.
Fabes, R. A., Eisenberg, N., Nyman, M., & Michealieau, Q. (1991). Young children’s appraisal
of others’ spontaneous emotional reactions. Developmental Psychology, 27, 858–866.
40
JOURNAL OF NONVERBAL BEHAVIOR
Field, T. M., & Walden, T. A. (1982). Production and discrimination of facial expressions by
preschool children. Child Development, 53, 1299–1311.
Friend, M. (2000). Developmental changes in sensitivity to vocal language. Developmental
Science, 2, 148–162.
Friend, M. (2001). From affect to language. First Language, 21, 219–243.
Friend, M. (in press). What should I do? Behavior regulation by language and paralanguage in
early childhood. Journal of Cognitive Development.
Friend, M., & Bryant, J. B. (2000). A developmental lexical bias in the interpretation of dis-
crepant messages. Merrill-Palmer Quarterly, 46, 342–369.
Gnepp, J. (1983). Children’s social sensitivity: Inferring emotions from conflicting cues. Devel-
opmental Psychology, 19, 805–814.
Gnepp, J. (1989). Personalized inferences of emotions and appraisals: Component processes
and correlates. Developmental Psychology, 25, 277–288.
Gnepp, J., & Hess, D. L. R. (1986). Children’s understanding of verbal and facial display rules.
Developmental Psychology, 22, 103–108.
Halberstadt, A. G., Denham, S. A., & Dunsmore, J. C. (2001). Affective social competence.
Social Development, 10, 79–119.
Josephs, I. (1994). Display rule behavior and understanding in preschool children. Journal of
Nonverbal Behavior, 18, 301–326.
Lee, K. (2000). Lying as doing deceptive things with words: A speech act theoretical perspec-
tive. In J. W. Astington (Ed.), Minds in the making (pp. 177–196). Malden, MA: Blackwell
Publishers.
Lee, K., & Cameron, C. A. (2000). Extracting truthful information from lies: Emergence of the
expression-representation distinction. Merrill-Palmer Quarterly, 46, 1–20.
Mumme, D. L., Fernald, A., & Herrera, C. (1996). Infants’ responses to facial and vocal emo-
tional signals in a social referencing paradigm. Child Development, 67, 3219–3237.
Perner, J. (1991). Understanding the representational mind. Cambridge, MA: MIT Press.
Peterson, C. C. (1995). The role of perceived intention to deceive in children’s and adults’
concepts of lying. British Journal of Developmental Psychology, 13, 237–260.
Peterson, C. C., Peterson, J. L., & Seeto, D. (1983). Developmental changes in ideas about
lying. Child Development, 54, 1529–1535.
Piaget, J. (1932/1965). The moral judgment of the child. New York: Macmillan Publishing
Co., Inc.
Repacholi, B. M, & Gopnik, A. (1997). Early reasoning about desires: Evidence from 14- and
18-month-olds. Developmental Psychology, 33, 12–21.
Rotenberg, K. J., & Eisenberg, N. (1997). Developmental differences in the understanding of
and reaction to others’ inhibition of emotional expression. Developmental Psychology,
33, 526–537.
Rotenberg, K. J., Simourd, L., & Moore, D. (1989). Children’s use of the verbal-nonverbal
consistency principle to infer truth and lying. Child Development, 60, 309–322.
Saarni, C. (1979). Children’s understanding of display rules for expressive behavior. Develop-
mental Psychology, 15, 424–429.
Saarni, C. (1984). An observational study of children’s attempts to monitor their expressive
behavior. Child Development, 55, 1504–1513.
Sapp, F., Lee, K., & Muir, D. (2000). Three-year-olds’ difficulty with the appearance-reality
distinction: Is it real or is it apparent? Developmental Psychology, 36, 547–560.
Siegal, M., & Peterson, C. C. (1996). Breaking the mold: A fresh look at children’s understand-
ing of questions about lies and mistakes. Developmental Psychology, 32, 322–334.
Siegal, M., & Peterson, C. C. (1998). Preschoolers’ understanding of lies and innocent and
negligent mistakes. Developmental Psychology, 34, 332–341.
Strichartz, A. F., & Burton, R. V. (1990). Lies and truth: A study of the development of the
concept. Child Development, 61, 211–220.
Talwar, V., & Lee, K. (2002). Emergence of white lie telling in children between 3 and 7 years
of age. Merrill-Palmer Quarterly, 48, 160–181.
41
MICHELLE ESKRITT, KANG LEE
Underwood, M. K., Hurley, J. C., Johanson, C. A., & Mosley, J. E. (1999). An experimental,
observational investigation of children’s responses to peer provocation: Developmental
and gender differences in middle childhood. Child Development, 70, 1428–1446.
Volkmar, F. R., & Siegel, A. E. (1982). Responses to consistent and discrepant social communi-
cations. In R. S. Feldman (Ed.), Development of nonverbal behavior in children (pp. 231–
255). New York: Springer-Verlag.
Walden, T. A. (1991). Infant social referencing. In J. Garber, & K. A. Dodge (Eds), The devel-
opment of emotion regulation and dysregulation (pp. 69–88). New York: Cambridge Uni-
versity Press.
Walker-Andrews, A. S. (1997). Infants’ perception of expressive behaviors: Differentiation of
mulitmodal information. Psychological Bulletin, 121, 437–456.
Walker-Andrews, A. S., & Dickson, L. R. (1997). Infants’ understanding of affect. In S. Hala
(Ed.), The development of social cognition. Studies in developmental psychology (pp.
161–186). Hove, England: Psychology Press/Erlbaum (UK) Taylor & Francis.
Wimmer, H., Gruber, S., & Perner, J. (1984). Young children’s conception of lying: Lexical
realism- moral subjectivism. Journal of Experimental Child Psychology, 37, 1–30.
Zeaman, J., & Garber, J. (1996). Display rules for anger, sadness, and pain: It depends on who
is watching. Child Development, 67, 957–973.