Children's verbal and nonverbal communication in a conflict situation

Children's verbal and nonverbal communication in a conflict situation

Children's Verbal and Nonverbal Communication in a Conflict Situation Linda A. Camras Department of Psychology, DePaul University Chicago, Illinois S...

990KB Sizes 16 Downloads 79 Views

Children's Verbal and Nonverbal Communication in a Conflict Situation Linda A. Camras Department of Psychology, DePaul University Chicago, Illinois

Same-sexed pairs of children (N = 144) were videotaped during brief play sessions involving a limited but desirable resource. Pairs varied in sex, age (preschool, kindergarten, second grade), friendship (friends vs. nonfriends), and perceived dominance (dominant vs. subordinate status within the pair). Videotape coding focused on verbal and nonverbal communication during negotiations for access to the resource. Results showed that dominant children played longer with the desirable object. Negotiation strategies varied with interactional role (e.g., attempting to gain vs. retain access to object) as well as age, sex, and social relationship. The politeness of children's language varied with their affective state as indexed by use of facial expressions. Key Words: Verbal communication; Nonverbal communication; Dominance.

In the classical ethological literature, two concepts often have been invoked to explain conflict negotiation and resource distribution among members of a social group. These are the concepts of dominance and social display. According to the classical view (Jolly 1972), dominance status is based on fighting ability and determines priority of access to scarce and desired resources. Communication involving the use of displays has been considered one means by which dominance is both established and expressed in social interactions. For example, subordinate individuals are thought to respond to threat displays produced by dominant group members as well as to actual attacks (Collias 1970; Jolly 1972). Acce!6ted April 9, 1984. Address reprint requests to: Linda A. Camras, Department of Psychology, DePaul University, 2323 North Seminary, Chicago, IL 60614.

Although classical views of both dominance and threat displays have been challenged by some researchers (Bernstein 1970, 1980; Blurton Jones 1968; Chance 1976; Gartlan 1968; Rowell 1972), there is evidence that these notions (at least in modified form) may be usefully applied to the study of children's social behavior. Using various conceptual and methodological approaches, a number of investigators (Abramovitch 1976, 1980; Barner-Barry 1980; McGrew 1972; Missakian 1980; Savin-Williams 1976, 1980; Sluckin 1980; Sluckin and Smith 1977; Vaughn and Waters 1980; Weisfeld, Omark, and Cronin 1980) have constructed dominance hierarchies for children ranging in age from preschool to adolescence. Furthermore, several studies (McGrew 1972; Strayer and Strayer 1976) have found that such hierarchies predict the outcome of observed struggles over objects or position. Regarding display communication, research strongly suggests that despite cross-cultural variation in use (Birdwhistell 1970: Eibl-Eibesfeldt 1970), human facial expressions of emotion comprise a species-characteristic set of signals that have evolved from primate display systems (Chevalier-Skolnikov 1973; Darwin 1872/1965; Ekman and Friesen, 1971). Furthermore, previous investigations of children's facial expressions (Blurton Jones 1972; Camras 1977, 1980; Grant 1969) suggest that some expressions are morphologically and functionally analogous to primate threat displays and that these expressions are used in conflict encounters between unacquainted kindergarten children (Camras 1977). The present study examined the roles of both dominance relations and display communication in conflict encounters between acquainted chil257

Ethology and Sociobiology 5:257-268 (1984) © Elsevier Science Publishing Co.. Inc., 1984 52 Vanderbilt Ave., New York, New York 10017

0162-3095/84/$03.00

258 dren. In addition, it focused on two other aspects of social behavior and relationship: verbal communication and friendship. These latter variables have received relatively little attention from human ethologists (but see LaFreniere and Charlesworth 1983; Strayer 1978; Weigel and Johnson 1981). In the present investigation, the verbal and nonverbal communication of children who varied in age, sex, dominance, and friendship relations were examined as the children engaged in play with a limited but desirable resource. One major purpose of the study was to determine whether conflict negotiation and resource distribution varies with age, sex, and social relationship. A second major goal was to determine whether some forms of verbal and nonverbal communication (herein considered communication strategies) are more effective than others when used during negotiations over object possession. Several specific questions regarding communication and social relationship were addressed. 1. Is resource distribution related to dominance as would be predicted by classical conceptualizations of dominance relations? Does resource distribution also vary depending on age, sex, and friendship relations? 2. Do conflict negotiation strategies (i.e., verbal and nonverbal communication) differ depending on children's age, sex, dominance, and friendship relations? 3. Do conflict negotiation strategies differ in effectiveness? 4. Are there relationships between verbal and nonverbal behaviors employed in conflict situations? In particular, does children's language vary with their affective state as indexed by an analysis of their facial displays?

METHOD Subjects The subjects for this study were 144 children drawn from several middle-class nursery schools and elementary schools. The sample included 22 preschool children (10 boys and 12 girls), 48 kindergarten children (26 boys and 22 girls), and 74 second-grade children (38 boys and 36 girls). Participating subjects were selected from a larger sample of children whose parents had given permission for them to take part. Subject selection was based on responses to a sociom-

Linda A. Camras etric interview and was made so as to maximize as far as possible the number of subjects in each cell of the experimental design.

Sociometric Interviews All subjects who were permitted to participate were given a sociometric interview designed to assess perceived dominance and friendship relations with classmates. The interview was conducted with the aid of a bulletin board displaying a photograph of each participating member of the child's class. To assess perceived dominance relations, the experimenter used a paired-comparison procedure similar to that used in previous sociometric studies of dominance (Omark, Omark, and Edelman 1975; Sluckin and Smith 1977). Pointing to the picture of the first classmate, she asked, " I f you and X (naming the classmate) were in a fight, w h o ' d win, you or X ? " She proceeded to do this for each member of the child's class. To assess perceived friendship, the experimenter then asked the subject to look at all the pictures and to name his or her three best same-sexed friends in the class.

Pairing the Subjects Using the interview data, subjects were paired for participation in a play session conducted within 2 weeks of the interviews. Same-sexed pairs of children were formed according to a factorial design with grade, sex, and friendship (friends vs. nonfriends) as between-subject-pair factors and perceived dominance (dominant vs. subordinate) as a within-subject-pair factor. Pair members who were friends had each chosen the other during the sociometric interview. Pair members who were nonfriends had both failed to choose their partner as a friend during the interview. All pairs had an established dominance relationship. That is, both members agreed about who would win in a fight (the dominant member) and who would lose (the subordinate member). To ensure the reliability of perceived dominance and friendship within each pair, pair members were reexamined regarding their relationship immediately before engaging in the play session, which took place about 2 weeks after the interview. Each child was asked (in private) whether he or she or his or her partner would win in a fight and whether his or her partner was

Children's Communication

259

one of his or her best friends. Because all children tended to claim their partner was a best friend when questioned in this manner, reliability of friendship choices could not be determined. However, the mutuality of choice (or nonchoice) during the initial interview was considered to indicate the validity of the friendship classification. Regarding dominance, the reassessment indicated considerable instability in perceived dominance relations. Consistency in responses during the initial interview and the reassessment was found for about half the pairs. Pairs were included in the study only if their responses were consistent in both interviews. This resulted in a considerable reduction in sample size, especially for the mutual friend pairs. That is, mutual friend pairs with well-established dominance relations appear to be relatively rare among young children. The final number of subject pairs in each cell of the experimental design is shown in Table 1. Because of the particularly small cell sizes for friendship pairs, the data for nursery school and kindergarten children was collapsed for the analyses report below. The friendship factor was not eliminated because of its conceptual importance for this study.

Play Sessions

Each pair of children was covertly videotaped during a brief play session designed to generate occasional conflicts between the interactants. For these play sessions, the children were seated at a table and allowed to play with an object with which only one child could play at a time. This limited and desirable resource was a pair of gerbil's in a box. The children could pass the box back and forth across the table through a opening in the clear plexiglass divider that bisected the table. When a child had the box on his side

of the table, he could play with the gerbils in a number of ways. The play apparatus had several features that ensured that each child could always reach a rope handle enabling him to pull or hold onto the box. However, only one child could play with the gerbils at a time. Instructions to the children were minimal during this seminaturalistic experiment. The children were told that they could play with the gerbils for 10 to 12 minutes while the experimenter left the room to " d o some w o r k . " No instructions were given about sharing or competing for the gerbil box. However, the children were asked to remain in their seats and pull the box over to their side of the table when they wanted to play. CODING OF VIDEOTAPES Videotape coding focused on processes involved in negotiating the exchange of the gerbil box. Coding involved identification and classification of attempts to obtain the box, responses to these attempts, and outcome with respect to the exchange of the box. Conflicts between the children were considered to occur when one child made an attempt to get the gerbil box and his partner responded by resisting the attempt. F o r each attempt and each resistance, the performing child's verbal behavior, facial behavior, and relevant physical acts (e.g., pulling or holding the box) were recorded on transcripts on which sequential information was preserved. Verbal statements were classified according to a system that was largely data derived but also included important distinctions discussed by linguists and psychologists studying social use of language by children and adults (Bates 1976; Clark and Schunk 1980; Ervin-Tripp 1976a, 1976b, 1977; James 1975, 1978; Searle 1969, 1975, 1976). Facial behavior was coded using a

Table 1. Design and Number of Subject Pairs

Sex Grade Preschool and kindergarten Second grade TOTAL

Friendship

Male

Female

Total

Friends Nonfriends

6 12

4 13

10 25

Friends Nonfriends

8 11

8 10

16 21

37

35

72

260 modified version of Ekman and Friesen's (1978) Facial Action Coding System (FACS), an anatomically based system in which facial behavior is coded in terms of action units (muscle or muscle group contractions producing the facial movements). Based on the investigator's past research (Camras 1977), certain distinctions made by FACS were hypothesized to be noncritical for this study and were eliminated to facilitate the laborious facial coding process. On the basis of the modified FACS coding, for each attempt and resistance, the performing child was scored as (1) using a target facial expression, or (2) using only nontarget facial expressions. The target expressions were those expressions hypothesized to be functionally equivalent to threat displays on the basis of previous work by the author (Camras 1977, 1980a) and by other ethologically oriented investigators of children's facial behavior (Blurton Jones 1972; Grant 1969). Briefly described, the target expressions were lowered brows, stare, lips pressed together (several forms), face thrust, nose wrinkle, and lower lip drop accompanied by jaw thrust. All coding was done by two coders who were naive with respect to the sociometric relationship between pair members. Reliabilities between coders were assessed for identification and classification of attempts, responses, and outcomes and for the coding of verbal, facial, and physical behavior performed during each of these interaction units. Reliabilities ranged from 85% to 90% for identification and classification of nonfacial acts. Reliabilities for the coding of facial behavior were somewhat lower. The reliability for identification of attempts and resistances containing a target expression was 72%, and the overall reliability for coding facial actions was 73%. These figures are comparable to those usually reported for FACS scoring (Ekman and Friesen 1978). RESULTS AND DISCUSSION Presentation of the data analyses and results will be organized in terms of the questions raised in the introductory section.

Does Resource Distribution Vary with Children's Age, Sex, Dominance, and/or Friendship Relations? An analysis of variance with three between-subjects factors (sex, grade, and friendship relation)

Linda A. Carnras and one within-subjects factor (dominance status of pair members) was conducted using length of playing time as the dependent variable. The analysis yielded a main effect for dominance, F(1,64) = 13,52, p < 0.0005, and no other main effects or interactions. Thus perceived dominance was related to the distribution of a desired resource, as would be predicted by classical dominance theory. However, it should be noted that the difference in playing time between dominant and subordinate children was relatively small; dominant children played an average of 5.4 minutes while subordinate children played 4.6 minutes. This failure of dominant children to completely monopolize the resource suggests that other factors (e.g., norms of fairness and sharing) also play an important role in determining resource distribution.

Do Conflict Negotiation Strategies (i.e., Verbal and Nonverbal Communication) Vary with Age, Sex, and Social Relationship? Separate MANOVAs were conducted to analyze behavior produced during attempts to get the gerbil box and during resistance to the other child's attempts. In each MANOVA, sex, grade, friendship, and dominance status served as independent variables. For the dependent measures, a number of variables were selected representing important and/or frequent strategy elements used by the children. These included pulls on the box, smiles, target expressions, use of several verbal statement categories, and use of no verbal statement at all. For both attempts and resistances, the six verbal statement codes most frequently appearing in the data were identified and those that could be meaningfully interpreted were used in the relevant MANOVA. Thus, for attempts, the analysis included four verbal statement categories: (1) nonspecific imperative demands (e.g., "Come on," "Hurry"); (2) imperative demands for permission to play ("Let me have it"); (3) announcements of turns ("It's my turn"); and (4) interrogative requests for permission ("Can I have a turn?"). Although appearing frequently in the data, miscellaneous attention-getters (e.g., "Hey") and fragments ("Yes," " N o " ) were not included in the analysis of attempts. For resistances, the verbal statements included in the MANOVA were: (1) simple "No"s; (2) imperative demands to give up the

Children's Communication

attempt (e.g., " D o n ' t pull," " S t o p it"); (3) temporizing imperatives ( " W a i t a minute"); (4) declarative statements giving reasons for resistance ( " I t ' s still my turn," " T h a t ' s not fair," " M y gerbil isn't through playing"); and (5) miscellaneous nonspecific imperatives (e.g., " C o m e o n , " " W a t c h it"). Although appearing frequently in the resistances, again attentiongetters (e.g., " H e y " ) were not included. The M A N O V A for attempts yielded significant main effects for sex and grade and a significant sex x grade interaction, Fse×(8,121) = 2.51 p < 0.02: Fgr~d~(1,121) = 2.15, p < 0.04; Fsex,grade(8,121) = 2.00, p = 0.0513. Univariate analyses indicated a significant effect of sex on children's tendency to pull on the box and to make imperative demands for permission to play (e.g., " L e t me have a turn"), Fpu,(l, 128) = 5.78, p < 0.02; Fimperative(1,128) = 5.12, p < 0.03. A significant effect for grade was found for children's tendency to use: (1) nonspecific imperative demands (e.g., " H u r r y " ) , F ( I , 128) = 8.42, p < 0.005; (2) imperative demands for permission to play, F(1,128) -- 3.83, p -- 0.0525; (3) pulls on the box, F(1,128) = 5.56, p < 0.02; and (4) no verbal statement at all, F(1,128) = 9.27, p < 0.003. Inspection of the data showed that boys tended to use pulls more than girls; pulls occurred in 66% of the b o y s ' attempts but only 53% of the girls' attempts. In contrast, girls made more permission demands than boys (in 9% vs. 5% of their attempts, respectively). Older children used both permission demands and nonspecific imperatives more often than younger children (9% vs. 5% for permission demands; 12% vs. 5% for nonspecific imperatives) Younger children tended to pull more than older children (64% vs. 56%) and more often used no verbal statement at all (48% vs. 38%). The univariate analyses also showed significant sex x grade interactions for: (1) interrogative requests for permission, F(1,128) = 5.69, p < 0.02; (2) announcements of turn, F(1,128) = 6.38, p < 0.02; (3) pulls on the box, F(1,128) = 3.69, p = 0.057; and (4) no verbal statement at all, F(1,128) = 7.67, p < 0.007. Inspection of the data (Table 2) showed that girls tended to increase their use of both verbal statements as they became older, whereas boys tended to decrease their use of both statement types. Boys' tendency to use pulls and to use no verbal statement with their attempts did not change with

261 age. However, older girls used fewer pulls and more verbalizations than did younger girls. Thus the main effects of grade on pulls and no verbalizations were due largely to changes in the girls' behavior. In addition, the overall sex effect for pulls was due largely to decrease in pulling on the box by older girls. These results suggest that attempt strategies become sex differentiated as children grow older. Whereas boys remain unchanged in their use of verbal statements and/or physical pulls, girls become increasingly verbal and decreasingly physical. Regarding the use of particular statement types or categories, girls increased their use of both imperative statements ( " L e t m e , " " H u r r y " ) and nonimperative statements ("Can I . . . ? " , " M y turn"), whereas boys increased only in the use of imperatives. Since imperatives generally are considered less polite than nonimperatives (James 1975, 1978; Lakoff 1973a), older boys may be hypothesized to be less polite on the average than older girls. The M A N O V A for resistances yielded significant main effects for grade and friendship and a significant sex × grade interaction, Fgrade (9,106) = 3.71,p < 0.0005; Ff,~nd(9,106) = 2.40, p < 0.02; Fs~x.grade(9,106) = 2.02, p < 0.05. Univariate analyses indicated a significant effect of grade on: (1) tempor&ing imperatives, F(1,114) = 15.62, p < 0.0001; (2) nonspecific miscellaneous imperatives, F(1,114) = 10.50,p < 0.002); and (3) the no-verbalization strategy, F(1,114) = 6.31, p < 0.02. A significant effect for friendship was found for declarative statements giving reasons for resistance, F ( I , 114) = 11.30, p < 0.002. Inspection of the data showed that younger children used the no verbalization strategy more often than older children (in 20% vs. 14% of their resistances, respectively). Older children used both temporizing imperatives and nonspecific imperatives more than younger children (39% vs. 17% for the former; 9% vs. 4% for the latter). In addition, reasons for resistance were used more often by nonfriends (34%) than by friends

(18%). The significant sex × grade interaction reflected differences in the use of pulls, F(1,114) = 8.58, p < 0.005, and the no-verbalization strategy, F(1,114) = 11.46, p < 0.001. As age increased, the girls tended to use both strategies less. F o r boys, use of pulls increased slightly with age, while use of the no-verbalization strategy remained about the same (Table 3).

262

Linda A. Camras

Table 2. Sex x Grade Interactions in Use of Attempt Strategies: Mean Proportion of Attempts in Which Strategy Was Used

Strategy Sex

Male

Female

Grade

Pulls on box

No verbal statement

Permission interrogative

Turn announcement

0.67

0.47

0.09

0.10

0.66

0.45

0.04

0.06

0.61

0.49

0.03

0.05

0.46

0.31

0.11

0.08

Preschool and kindergarten (N = 36) Second grade (N = 38) Preschool and kindergarten (N = 34) Second grade (N = 36)

The findings for resistance strategies in part parallel and in part diverge from the findings for attempts. As for attempts, girls become more verbal and less physical with age, whereas boys show little (if any) change in their tendency to resist using pulls and/or verbal statements. In addition, there is an increase with age in use of imperative statements by both sexes during resistances as well as during attempts. However, as opposed to the attempt analysis, the findings for resistances did not show that girls also increase their use of nonimperatives with age. In fact, use of the nonimperative examined in the resistances MANOVA (i.e., declarative giving reason for resistance) was related to friendship rather than to sex and/or grade. Thus, use of polite verbal resistance statements may not simply become increasingly sex differentiated with age as may use of polite verbal statements during attempts.

The politeness of children's verbal attempts and resistances was further explored in an analysis that examined the full range of children's verbal statements (rather than just the statement categories used in the MANOVAs described above). For each subject, three verbal attempts and three verbal resistances were randomly selected. These were presented in written form to undergraduate psychology students (three males and three females) who rated each one for "politeness" on a seven-point scale. The politeness scores for attempts and resistances were analyzed separately, each in an analysis of variance with sex, grade, friendship, and dominance as independent variables. Results for the analysis of verbal attempts showed a virtually significant significant sex × grade interaction, F(1,120) = 3.64, p = 0.059. Inspection of the data (Table 4) showed that girls did indeed become increasingly polite with age while boys became increas-

Table 3. Sex x Grade Interactions in Use of Resistance Strategies: Mean Proportion of Resistances in Which Strategy Was Used

Strategy Sex

Males

Females

Grade Preschool and kindergarten (N = 34) Second grade (N = 36) Preschool and kindergarten (N - 32) Second grade ( N - - 33)

Pulls on box

No verbal statements

0.60

O.19

0.67

0.19

0.55

0.21

0.35

0.09

Children's Communication

263

Table 4. Mean Politeness Scores for Attempts a Sex

Grade

Male

Female

Preschool and kindergarten Second grade

3.91 4.18

4.33 3.86

a Possible scores ranged from 1 (Very Polite) to 7 (Very Impolite).

ingly impolite. Results for the analysis of verbal resistances yielded two significant interaction effects: sex x dominance, F ( I , I I 6 ) = 4.69, p < 0.04; and sex x dominance × friendship x grade, F(1,116) = 5.86, p < 0.02. Inspection of the mean politeness scores relevant to the sex x dominance interaction (Table 5) indicated that dominant boys were considerably less polite than the other sociometrically defined groups. Dominant girls, subordinate girls, and subordinate boys were about equal in politeness. Inspection of the mean scores relevant to the sex x dominance x friendship x grade interaction showed that in most cases dominant girls were equally polite or even more polite than their subordinate partners. Only within the younger nonfriend pairs did dominant girls actually use less polite language than subordinates. One possible interpretation of this pattern of results is that dominant girls are gradually socialized to mask their exercise of power during conflicts with use of polite language. Thus younger dominant girls are polite to their friends but not their nonfriends. Older dominant girls are polite to both friends and nonfriends. Therefore the politeness of verbal resistances appears to depend upon factors such as dominance and friendship as well as age and sex. Regarding the use of facial displays, the data analyses (MANOVAs) showed no significant differences among children in their use of those expressions hypothesized to be functionally analogous to primate threat displays (i.e., the target expressions). However, a distribution pattern observed in an earlier investigation (Camras 1979) was replicated to some degree in the analysis of resistance strategies. In the previous study, dominant children had been observed to use more target expressions when paired with a nonfriend as opposed to a friend. In contrast, subordinate children used more target expressions when paired with a friend as opposed to a nonfriend. In the present study, this dominance

x friendship interaction was weakly replicated for female interactants. That is, a significant (p < 0.066)sex x dominance x friendship interaction was found when the univariate analysis of target expressions was examined (although the overall F for the M A N O V A was nonsignificant). Inspection of the data showed that girls followed the dominance x friendship pattern of expression usage described above. However, among boys, dominant children used more target expressions when paired with friends as opposed to nonfflends whereas subordinate children used these expressions equally often with friend and nonfriend partners. Possibly a difference between the two studies i n criteria for defining dominance produced the difference in facial expression usage that was observed.1 However, it also is possible that any relationship between social status and facial expression usage is fragile and will not be consistently observed no matter how dominance is defined. Furthermore, it is significant that neither study showed a main effect of dominance on facial expression usage. Thus dominant children (defined by either cflteflon) did not invariably use more facial threats than subordinates. Nonetheless, as shown in two past investigations (Camras 1977, 1979), using a target expression while defending possession of the object is related to greater success in retaining it relative to a child's own baseline for success. Thus using a facial threat may aid a child to some extent but its power in this study may have been overshadowed by overall baseline differences in the success of dominant versus subordinate children.

Do Conflict Negotiation Strategies (i.e., Verbal and Nonverbal Communication) Differ in Effectiveness?

Separate two-way A N O V A s were performed to analyze the effectiveness of attempt strategies and the effectiveness of resistance strategies: For each A N O V A , the two fixed factors were: physical component (pull vs. no pull on the box) and verbal component. In both analyses, the yerbal component factor had several levels corresponding to those verbal statement categories utilized in the M A N O V A s described above. F o r J In the previous study, dominance status was defined according to the overall ratings of classmates but pair members sometimes did not agree between themselves (with both children claiming dominance).

264

Linda A. Camras

Table 5. Mean Politeness Scores for Resistances Sex x Dominance interaction Sex Dominance

Male

Female

Dominant Subordinate

4.61 4.28

4.27 4.22

Sex x Dominance x Friendship × Grade interaction Sex Male Preschool and kindergarten Second grade

Female

Dominant

Subordinate

Dominant

Subordinate

Friend Nonfriends

4.60 4.62

4.25 4.27

4.06 4.60

4.84 3.88

Friends Nonfriends

4.61 4.61

4.43 4.20

4.07 4.04

4.13 4.55

the attempt analysis, " e f f e c t i v e n e s s " was defined as " m e e t i n g no resistance." F o r each attempt strategy examined, the proportion of attempts meeting no resistance was determined for each subject using the strategy. Arcsine transformations of these proportions served as the dependent measures. For the resistance analysis, effectiveness was defined as "resulting in the resister keeping the object at the end of the conflict." An effectiveness score for each resistance strategy was computed in a manner analogous to that described for attempts. 2 The A N O V A for attempt strategies failed to show significant differences among strategies in terms of effectiveness. However, the A N O V A for resistance strategies yielded a significant effect for the verbal component, F(4,393) = 2.91, p < 0.02, indicating that some strategies were more effective than others. Post hoc analyses (Dunn Multiple Comparisons, n = 15, p < 0.05) showed that the most successful strategy (declarative giving reason for resistance) was significantly more effective than the least successful strategy (imperative demand to give up the attempt), with the other verbal statement categories falling somewhere in between (see Table 6). The relative effectiveness of the reason declaratives and the relative ineffectiveness of the imperative demands does not appear to be attributable to preferential use of these strategies by children who tended to be characteristically 2 Although technically inappropriate, use of any strategy by any subject was treated as an independent data point. The more appropriate within-subjects analysis was not performed because few subjects used all of the strategiesbeingexamined.

successful or unsuccessful overall. That is, the M A N O V A for resistances described in the preceding section did not show these statements to be differentially associated with either the dominant children (who tended to keep the box longer) or subordinate children (who tended to keep it for a shorter period of time).

Are There Relationships Between Verbal and Nonverbal Behaviors Employed in Conflict Situations? In Particular, does Children's Language Vary with Their Affective State as Indexed by an Analysis of Their Facial Displays? Again, separate analyses were conducted to analyze data for attempts and resistances. For each analysis, only subjects who used target expressions during some episodes and smiles during other episodes were examined. For these subjects, politeness ratings of verbal statements were obtained by presenting the statements (in written form) to undergraduate psychology students, who were asked to rate each statement on a seven-point scale. These ratings were analyzed by means of a correlated t-test, which compared the politeness of statements accompanying target expressions to the politeness of statements accompanying smiles. Results for both attempt and resistance episodes showed that verbal statements accompanying smiles were more polite than verbal statements accompanying target expressions, F a n e m p t s ( 3 0 ) : 1.99, p = 0.055; tresist~(20) = 2.25, p < 0.04. These findings demonstrate a relationship between children's use of language and their use

Children's Communication

265

Table 6. Effectiveness of Verbal Resistance Strategies

Verbal statement category

Number of episodes Involving students

Proportion of episodes in which resistance was successful

Declarative giving reason Miscellaneous Temporizing imperative Simple "No" Imperative demand

97 36 79 60 45

0.77 0.74 0.74 0.73 0.65

of affect-related facial expressions (i.e., target expressions and smiles). Thus it seems possible that some aspects of language (e.g., variations in politeness) may function in part as an affect display system. That is, interactants sometimes may make inferences about a speaker's affect based on his choice of language as well as his nonverbal behavior. In fact, if use of facial displays declines with age (as has been reported by Zivin 1977) and if use of language increases with age (as shown in this study), then language variation may gradually become an increasingly important means of affect communication for children. GENERAL DISCUSSION The data analyses conducted so far suggest that dominant children may have greater access to a desired resource but not necessarily because they employ different and more effective negotiation strategies than subordinate children. Dominant children indeed played longer with the gerbil box than subordinates in this study. However, there were no systematic differences between dominants and subordinates in terms of the strategy elements they used when they attempted to obtain the object from their partner. Similarly, there were no significant main effects of dominance when use of resistance strategies was examined. However, there was a tendency for some dominant children (in particular, the boys) to use less polite language than subordinates during resistance. The results regarding resource distribution are consistent with findings of several previous studies of children's dominance relations. Other investigations (Charlesworth and LaFreniere 1983; McGrew 1972; Strayer and Strayer 1976, 1980) also have reported that dominant children have greater access to desired resources and/or win more object disputes than subordinates.

However, the present study differs from these past investigations in that dominance was assessed through sociometric interviews rather than direct observation of behavior. Previous sociometric studies of dominance (Sluckin 1980; Sluckin and Smith, 1977; Strayer, Strayer, and Chapeskie 1980) have not reported strong relationships between perceived dominance and dominance assessed through behavioral observations. Possibly this is because no attempt was made in these studies to utilize only reliable judgments of dominance relations. In the present study, perceived dominance was found to be fairly unstable, with children's judgments often changing between assessments made about 2 weeks apart. However, when only stably perceived relations were examined, these perceived relationships did predict resource distribution with significant accuracy. The patterns observed in this study for use of language, pulls, and facial expressions support a theoretical conception of dominance that emphasizes the relationship between individuals rather than the coercive behavior of some group members. That is, dominant children were not distinguished by their attempts to " d o m i n a t e " via impolite language, physical pulls, or target expressions (facial " t h r e a t s " ) . However, though dominant children did not consistently attempt to exert authority by these means, they did consistently have greater access to the gerbil box. This suggests that any form of negotiation strategy may be more effective when employed by a dominant as opposed to a subordinate individual. F o r example (see Rowell 1974), it is possible that the outcome of many interactions depends largely on the subordinate's tendency to defer to any message produced by a dominant partner. Further data analyses comparing the responses of dominant vs. subordinate children to the same negotiation strategy would serve to examine this possibility more directly.

266 The findings of this study also serve to demonstrate that children's communication involves a sophisticated process of stimulus integration and interpretation rather than the simple "release" of appropriate responses to social displays. Subjects' responses to communication strategies appeared to depend both on the message and on contextual factors, in particular the interactants' social relationship. Although ethologists traditionally have described communication using simple " r e l e a s e r " models, a greater emphasis on context-dependent responding may be appropriate. At least one model that emphasizes the role of context has been described in the ethological literature and applied to both human and nonhuman animal communication (Smith 1965, 1977). The present study suggests that social relationship constitutes an important aspect of contextual background for message interpretation by children. Other contextual factors, such as social situation, may also play an important role (for a more detailed discussion see Camras 1980a; Smith 1977). With regard to language use by children in this study, it is interesting to consider the results in light of nonethological investigations of verbal politeness. Several linguists and psycholinguists (Key 1975; Lakoff 1973; Thorne and Henley 1975) have reported general sex differences in language use, with females being more polite overall than males. In the present study, this report was confirmed in the analysis of verbal attempts. In addition, the sex x grade interaction found for the politeness of attempts suggests that a gradual socialization process is taking place with a significant degree of sex-differentiation occurring between kindergarten and second grade. Regarding politeness during resistances, sex differences also were observed, but the effect of sex interacted with the effects of dominance and friendship as well as age. Dominant boys were consistently less polite than subordinate boys. However, older dominant girls were equal in politeness to their subordinate partners and younger dominant girls also were polite as long as they were paired with a friend. Thus sex differences in language may vary depending upon a number of factors including the speaker's age, social relationship with the listener and interactional role (e.g., attempter vs. resister in an object dispute). However, the general pattern of sex differences observed in studies of nonconflict interactions was also found to

Linda A. Camras occur during certain aspects of children's conflict negotiations. A last interesting finding for this study is the relationship between children's use of language and facial expression. When both attempts and resistances were examined, target expressions were found to be accompanied by language that was less polite than the language accompanying smiles. If the target expressions are assumed to reflect either an attack tendency or feelings of anger and/or determination (Camras 1980a, 1980b), then impolite language similarly might reflect these feelings when it is used in a conflict situation. Consistent with this notion, judgment studies conducted with children and adults (Camras, Pristo, and Brown, 1983) show that subjects attribute relatively impolite language to angry speakers and more polite language to happy speakers. In real life situations, perhaps children and adults also will sometimes make inferences about the speaker's emotions based on the language he produces. Thus language may serve (in part) as an affect-display system related to facial expressions, a display system considered to have phylogenetic roots. If use of facial displays declines with age (as reported by Zivin 1977) and use of language increases with age (as found in this study), then language may become increasingly important for emotion communication over the course of development. In conclusion, the present study indicates that the conflict negotiation strategies employed by children are complexly related to factors such as age, sex, and social relationship. Although traditional ethological concepts of dominance and nonverbal displays can explain some aspects of conflict resolution and resource distribution, other variables also may play an important role. In particular, this study has shown that an analysis of children's language may be fruitful when performed in the context of an ethological study. In fact, if use of nonverbal displays and physical means of conflict resolution decrease with age, the analysis of language may become increasingly critical to our attempts to understand children's and adults' behavior in conflict situations.

This research was funded by a grant from the Spencer Foundation. The author wishes to thank the administration, teachers, children and parents of the following schools for their generous cooperation: Niles Township Jewish Congregation Preschool, St. Benedict's Elementary School, St. John's Elementary School,

Children's Communication

School for Little Children• In addition, the author appreciated the help of Kevin Allison, Mari Brown, Stephen Billingham, and Steven Spaccarelli in coding and analyzing the data. Special thanks are due to Daniel J. Conti and Toni M. Pristo who participated with enthusiasm and devotion in all aspects of data collection, coding, and analysis•

REFERENCES Abramovitch, R. The relation of attention and proximity to rank in preschool children. In The Social Structure of Attention, M. Chance and R. Larsen (Eds.), New York: Wiley, 1976. • Attention structures in hierarchically organized groups. In Dominance Relations: An Ethological View of Human Conflict and Social Interaction. D. Omark, F. Strayer and D. Freedman (Eds.). New York: Garland STPM Press, 1980. Barner-Barry, C. The structure of young children's authority relationships. In Dominance Relations: An Ethological View of Human Conflict and Social Interaction, D. Omark, F. Strayer, and D. Freedman (Eds.), New York: Garland STPM Press, 1980. Bates, E. Language and Context• New York: Academic Press, 1976. Bernstein, I. Primate status hierarchies• In Primate Behavior: Developments in Field and Laboratory Research, L.A. Rosenblum (Ed.). New York: Academic 1970, Vol. 1. --. Dominance: A Theoretical Perspective for Ethologists. In Dominance Relations: An Ethological View of Human Conflict and Social Interaction, D. Omark, F. Strayer, and D. Freedman (Eds.). New York: Garland STPM Press, 1980. Birdwhistell, R.L. Kinesics and context. Philadelphia: University of Pennsylvania Press, 1970. BlurtonJones, N.G. Observations and experiments on causation of threat displays of the great tit (Parus major). Animal Behavior Monographs 1(2): 75-158 (1968). --. Categories of child-child interaction. In Ethological Studies of Child Behavior, N.G. Blurton Jones (Ed.). London: Cambridge University Press, 1972. Camras, L. Facial expressions used by children in a conflict situation• Child Development 48: 14311435 (1977)• --. Peer Relations and Facial Communication in a Conflict Situation• Paper presented at the meeting for the Society for Research in Child Development, San Francisco, 1979. • Animal threat displays and children's facial expressions, in Dominance Relations: An Ethological View of Human Conflict and Social Interaction, D. Omark, F. Strayer, and D. Freedman (Eds.), New York: Garland STPM Press, 1980a.

267

Camras, L. Children's understanding of facial expressions used during conflict encounters• Child Development, 51:879-885 (1980b). --, Pristo, T.M., Brown, M. Social Use of Language by Children and Adults: Affect, Situation and Directive Choice. Unpublished manuscript, 1983. Chance, M.R.A. Attention structure as the basis of primate social orders• In The Social Structure of Attention, M. Chance and R• Larsen (Eds.). New York: Wiley, 1976. Charlesowrth, W., LaFreniere, P. Cooperative and Competition in Preschoolers: The Influence of Dominance and Friendship• Paper presented at the meeting of the Society for Research in Child Development, Detroit, 1983. Chevalier-Skolnikov, S. Facial expression of emotion in nonhuman primates• In Darwin and Facial Expression, P. Ekman (Ed.). New York: Academic, 1973. Clark, H., Schunk, D. Polite responses to polite requests• Cognition 8:111-143 (1980). Collias, N. Aggressive behavior among vertebrate animals. In Animal Aggression, C. Southwick (Ed.). New York: Van Nostrand Reinhold, 1970. Darwin, C. The Expression of the Emotions in Man and Animals• London: John Murray, 1872. Reprinted Chicago: University of Chicago Press, 1965. Eibl-Eibesfeldt, I. Ethology, the Biology of Behavior. New York: Holt, Rinehart, and Winston, 1970. Ekman, P., Friesen, W. Constants across cultures in the face and emotion. Journal of Personality and Social Psychology 17(2): 124-129 (1971)• --. Unmasking the Face. Englewood Cliffs, NJ: Prentice-Hall, 1975. , - - . Facial Action Coding System•" A Technique for the Measurement of Facial Movement. Palo Alto: Consulting Psychologists Press, 1978. Ervin-Tripp, S. Is Sybil there?: The structure of some American English directives. Language in Society 5:25-66 11976a). • Speech acts and social learning• In Meaning in Anthropology, K. Basso and H. Selby (Eds.). Albuquerque, NM: University of New Mexico Press, 1976b. • Wait for me, rollerskate! Child Discourse, In S. Ervin-Tripp and C. Mitchell-Kernan (Eds.). New York: Academic, 1977. Gartlan, J.S. Structure and function in primate society. Folio Primatologica 8:89-120 (1968)• Grant, E.C. Human facial expression. Man 4:525-536 (1969). James, S. The Effect of Listener and Situation on the Politeness of Preschool Children's Directive Speech• Unpublished Ph.D. dissertation, University of Wisconsin, 1975. • Effect of listener age and situation on the politeness of children's directives. Journal of Psycholinguistic Research 7:307-317 (1978).

268

Jolly, A. The Evolution of Primate Behavior. New York: Macmillan, 1972. Key, M. Male~Female Language. Metuchen, N J: Scarecrow Press, 1975. LaFreniere, P., Charlesworth, W. Dominance, attention and affiliation in a preschool group: A ninemonth longitudinal study. Ethology and Sociobiology 4:55-67 (1983). Lakoff, R. The logic of politeness. In Papers from the Ninth Regional Meeting, Chicago Linguistic Society, C. Corum, T.C. Smith-Stark, and A. Weiser (Eds.). Chicago: Chicago Linguistic Society, 1973a. --. Language and woman's place. Language in Society 2 : 4 5 - 8 0 (1973b). McGrew, W.C. An Ethological Study of Children's Behavior. New York: Academic, 1972. Missakian, E. Gender differences in agonistic behavior and dominance relations in Synanon communally reared children. In Dominance Relations: An Ethological View of Human Conflict and Social Interaction, D. Omark, F. Strayer, and D. Freedman (Eds.). New York: Garland STPM Press, 1980. Omark, D., Omark, M., Edelman, M.S. Formation of dominance hierarchies in young children: Action and perception. In Psychological Anthropology, T. Williams (Ed.). The Hague: Mouton, 1975. Rowell, T. The concept of social dominance. Behavioral Biology 11:131-154 (1974). Savin-Williams, R. An ethological study of dominance formation and maintenance in a group of human adolescents. Child Development 47: 972-979 (1976). • Dominance and submission among adolescent boys. In Dominance Relations: An Ethological View of Human Conflict and Social Interaction, D. Omark, F. Strayer, (Eds.). New York: Garland STPM Press, 1980. Searle, J. Speech Acts. Cambridge: Cambridge University Press, 1969. • Indirect speech acts. In Syntax and Semantics, Vol. 3, Speech Acts, P. Cole and J. Morgan (Eds.). New York: Academic, 1975. --. A classification of illocutionary acts. Language in Society 5 : 1 - 2 3 (1976).

L i n d a A. C a m r a s

Sluckin, A. Dominance relationships in preschool children. In Dominance Relations: An Ethological View of Human Conflict and Social Interaction. D. Omark, F. Strayer, and D. Freedman (Eds.). New York: Garland STPM Press, 1980. Sluckin, A , Smith, P.K. Two approaches to the concept of dominance in preschool children. Child Development 48:917-923 (1977). Strayer, F.F. An ethological analysis of preschool social ecology• W.A. Collins (Ed.). In The Minnesota Symposium on Child Psychology, New York: Erlbaum, 1978. , Strayer, J. An ethological analysis of sacial agonism and dominance relations among preschool children. Child Development 47:980-989 (1976). --, Chapeskie, T. The perception of social power relations among preschool children. In Dominance Relations: An Ethological View of Human Conflict and Social Interaction, D. Omark, F. Strayer, and D. Freedman (Eds.). New York: Garland STPM Press, 1980. Thorne, B., Henley, N. (Eds). Language and Sex: Difference and Dominance. Rowley, MA: Newbury House, 1975. Vaughn, B., Waters, E. Social organization among preschool peers: Dominance, attention, and sociometric correlates. In Dominance Relations: An Ethological View of Human Conflict and Social Interaction, D. Omark, F. Strayer, and D. Freedman (Eds.). New York: Garland STPM Press, 1980. Weigel, R., Johnson, R.P. An ethological classification system for verbal behavior. Ethology and Sociobiology 2:55-66 (1981) Weisfield, G., Omark, D., and Cronin, C. A longitudinal and cross-sectional study of dominance in boys. In Dominance Relations: An Ethological View of Human Conflict and Social Interaction, D. Omark, F. Strayer, and D. Freedman (Eds.). New York: Garland STPM Press, 1980. Zivin, G. On becoming subtle: Age and social rank changes in the use of a facial gesture. Child Development 48:1314-1321 (1977).