This item is only available as the following downloads:
EFFECT OF PROPS ON PRESCHOOL CHILDREN'S COMMUNICATIVE INTERACTIONS IN SCRIPTED PLAY EVENTS BY GABRIELLE YVONNE DEFIEBRE A Thesis Submitted to the Division of Psychology New College of Florida in partial fulfillment of the requirements for t he degree Bachelor of Arts Under the sponsorship of Dr. Michelle Barton Sarasota, Florida May, 2009
ii Acknowledgements Thanks to Dr. Barton for being such an understanding and helpful thesis sponsor during my thesis writing process and t hrough my health issues, and thanks to Dr. Baram and Dr. Bauer for also being so understanding. Special thanks to my family for always encouraging me to finish my thesis, as well as recover from transverse myelitis. Lastly, thanks to Jason for making it po ssible for me to finish this project by helping me with things I can no longer do by myself, and for staying by my side through this tough time.
iii Table of Contents Acknowledgements ii Table of Contents iii Abstract iv Pretend Play 1 Developmental Trends in American Children's Pretend Play 2 Theoretical Roles of Pretend Play 5 Play as Social 6 Communication and Shared Meaning in Pretend Play 6 Scripted Play Definitions 8 Culture and Families as Providers of Scripts and Communication 9 Objects as a Factor Influencing Social Interactions and Communication 18 How Children Use Scripts 22 Scripted Play With Props 31 Research Proposal 36 General Discus sion 40 Appendix A 42 Appendix B 43 References 44
iv EFFECT OF PROPS ON PRESCHOOL CHILDREN'S COMMUNICATIVE INTERACTIONS IN SCRIPTED PLAY EVENTS Gabrielle Yvonne deFiebre New College of Florida, 2009 ABSTRACT The curre nt literature review and research proposal investigated the role of props in preschool children's communicative exchanges during scripted play. It has been shown that several factors influence the physical environment that children are exposed to, and that the physical environment (including toys) influences how children play and communicate. The proposed study would try to explicate the role of abstract or realistic props during scripted play interactions. A total of sixty four preschool children (thirty t wo boys, thirty two girls) between the ages of four and five would participate in this study. Preschool children would play with a same sex peer and would be given either abstract props or realistic props for their play session. Children's conversation wou ld be analyzed for how the role of props would influence the duration of play episodes, the number of turns in an episode, the communicative strategies used during the play, the definition of roles, and children's violation of the script. Dr. Michelle Barton Division of Social Sciences
1 Pretend Play "Deep meaning lies often in childish play." Johann Friedrich von Schiller, German poet. Thinking about play creates an image of children participating in an activity for fun rather than for serious purposes. The Digital Oxford American dict ionary defines play as activity engaged in for enjoyment and recreation, esp. by children" (2006). This definition does not fully convey what children can accomplish during play. Children can communicate complex ideas and information to themselves and to one another. Children's play varies in form due to developmental, cultural, social and environmental influences. The physical environment surrounding children, which is culturally and developmentally variable, often includes toys. These objects are often a n important part of play sessions, but the influence of toys in communication has not yet been fully explored within the field of developmental psychology. This thesis proposes a way to close this gap in the research as well as bridge the separation betwe en psychology and anthropology by using literature from both fields. Starting with the developmental variations in the way children play, many factors that influence play and communication are discussed. These include cultural, social, and finally environm ental factors, which are the main focus of this research. The purposes of pretend play are also discussed, including the learning of cultural information. Culture is another focus of this thesis because one's environment is culturally mediated and this env ironment houses different objects for children to play with.
2 One way to study play is to look at how children communicate with each other during play bouts. Their communicative strategies can give information about what children are trying to accomplish in their play. Children often attempt to share meaning, or effectively communicate parallel ideas to one another through language (Farver, 1992). Developmental Trends in American Children's Pretend Play There are several different types of play that gener ally emerge at different stages of development, although the stages are not definite and play types can emerge at different times for different children. Sensorimotor play begins to occur in infants (Cole, Cole & Lightfoot, 2005). This type of play can be seen when children use their own body to feel different sensory experiences, such as splashing water during bath time. Children then participate in sensorimotor play that incorporates objects and other people. As children's mental abilities become more adv anced, different play types emerge. Children begin to participate in symbolic play which "is play in which one object stands for ( represents) another," such as when a child pretends a block is a cash register (Cole et al., 2005, p. 216). In these pretend p lay episodes, children can perform more developmentally advanced actions than they might otherwise be able to perform (Cole et al., 2005). Children learn how to pretend play by themselves, and also learn how to participate in social play in which they take on reciprocal roles with another person (Howes, Unger & Seidner, 1989). Once they are able to participate in both solitary pretend play and social play, children can use their skills to participate in social pretend play with another person in which they include nonliteral meanings in their play, such as when one child offers another child a block as "food" to eat (Howes, Unger & Seidner, 1989).
3 Changes in play type are developmental, and represent changing cognitive abilities. At approximately 12 15 mont hs of age, children participate in pretend actions in the presence of a partner, but without jointly playing with them, such as when one child pretends to feed his or herself and the play partner gives no response (Howes et al., 1989, p.78; Howes et al., 1 992). Children then begin to understand that the meaning behind pretend play actions can be communicated to another person. At approximately 16 20 months of age, children begin to perform pretend acts on objects. Also, two children may try to perform the s ame action at the same time in their play, and from this they may learn that meaning can be communicated through these acts. When a child is approximately 18 months old they participate in symbolic play, which is when one object stands for another (Cole et al., 2005). By the time a child is 24 months of age, their social pretend play is more coordinated, probably because of advancements in cognitive abilities. When children are 20 24 months old, they can attribute independent agency to inanimate objects (su ch as toys) when they are playing alone. For example, they can pretend that a teddy bear has the desire to bake cookies (Howes et al., 1992). This shows that children are able to include objects in their play in a meaningful way very early in life. A cent ral task for 2 and 3 year olds is to learn how to communicate meaning to others, and social pretend play allows them to practice this task (Howes et al., 1992). Social pretend play occurs when the meaning of something is interpreted in ways that differ fr om how the meaning is usually thought of; for example, children pretend that they are using pencils as phones to communicate with each other (Farver, 1992, p. 502). Children learn how to participate in social pretend play with their peers and with experts (adults or older children) (Howes et al., 1992). With peers, children develop skills in
4 social pretend play as they gradually integrate pretense and social play. When children play with experts, they are provided with a play structure by the expert that he lps them participate more fully in the play. Scripted play emerges at around 30 months of age (Howes et al., 1992). Scripts are "mental representations of routine events" (Furman & Walden, 1990, p.227). Scripts have a clear temporal order that can be mimi cked in play. People can share knowledge about common events through the use of scripts, making it easier to organize information. Within play, scripts are ordered sequences of actions within a play sequence that are organized around a particular goal (Fur man & Walden, 1990). Some examples of scripted knowledge are knowledge about going shopping, baking cookies and waking up in the morning. If children know the script for baking cookies, they will pretend to preheat the oven, measure the ingredients, mix th em, put them on a baking sheet and then bake them in the oven. When children are between 25 30 months old, they can use scripts in pretend play with peers and experts (Howes et al., 1992). Scripted play, which can include props, varies based on the social context. When a child uses scripted knowledge with a parent (an expert), the expert can check the script for accuracy or ask the child for more information. When a child is using scripted knowledge in play with a peer, there is no authority on what constit utes the "right" scripts and no one can check the accuracy of the scripted interaction. At the beginning of the use of scripted knowledge, scripts are not often articulated verbally but are represented through actions. At around 25 30 months however, child ren will use a verbal script (Howes et al., 1992). Children who have experienced similar life events or who are good friends may be more capable of engaging in scripted social pretend play at a younger age. Finally, at approximately 31 36 months
5 of age, ch ildren's pretend play is influenced by their gradually more developed understanding of social roles that they may use in their play, such as pretending to be a parent and a baby (Howes et al., 1992). Theoretical Roles of Pretend Play Play often helps chi ldren practice actions that will be helpful later in childhood and in their adult life, and allows them to use their newly developing mental capabilities (Briggs, 1998; Lancy, 1996). It has been shown that early social pretend play is related to the abilit y of a child to develop understanding of how other people feel and what they think (Youngblade & Dunn, 1995). This indicates that a child's experience with participating in social pretense helps them master "the relation between mental life and real life" (Youngblade & Dunn, 1995, p. 1472). Vygotsky (1978) posited this idea in his book Mind in society by stating that play is extremely important for children because it helps them separate the immediate visual realm from their internal motives. It is importa nt to note that what Vygotsky (1978) stated was theoretical, but his ideas have been highly influential in developmental psychology. Vygotsky asserts that action in these imaginary situations help children guide their behavior in other contexts and during play. Also, in play, thought begins to be separated from objects, and actions begin to arise from ideas and not objects. Therefore, at first the object dominates over meaning in the child's perspective, but once an object becomes "a pivot for detaching mea ning," the child then begins to allow meaning to predominate their perspective in play (p.98). Vygotsky also argued "Play creates a zone of proximal development of the child. In play a child also behaves beyond his average age, above his daily behavior; in play it is as though he were a head taller than himself" (p. 102). This means that play gives children a
6 setting in which they can behave at a more advanced level than during other parts of their life. Vygotsky (1978) also stated that play should not be described as an activity that solely gives pleasure to children, but as an activity that fulfills children's needs. Play begins to develop once children experience desires that they cannot gratify immediately. They try to resolve the tension between the fa ct that they want something and the fact that they cannot have it right away by joining an imaginary world that he deems "play." Vygotsky (1978) argued that in play a child creates an imaginary situation, and that later on in childhood and adolescence chil dren engage in play without action, which is imagination. He also argued that play always has rules, even if they are not explicitly written down in advance. For example, a child pretending to be a parent of a doll uses the role of parental behavior to gui de his or her play. Play as Social Communication and Shared Meaning in Pretend Play. The play Vygotsky (1978) discussed often occurs in a social environment. One way to measure differences in how children play is to look at their communication within a so cial setting. It has been shown that certain variables influence the way children play and how they communicate with one another during these play episodes. For example, children use different communicative strategies when they are trying to create shared meaning with one another, and these strategies change as children get older (Farver, 1992). Shared meaning occurs when children learn how to effectively communicate parallel ideas to one another (Farver, 1992). When children begin to communicate with one a nother in these play events, they realize that words can create ideas in other people's minds that match the
7 ideas in their minds. This means they "...enter into a system of shared meaning" (Nelson, 1985, as cited in Farver, p.501). In one study with chil dren between the ages of 2 and 5 years, Farver (1992) investigated how young children used communicative strategies to structure and create shared meaning during social pretend play. The children in Farver's study regularly attended a full time day care c enter and knew each other before the research was conducted. The children were all paired by age and sex, and were audio recorded for 20 minutes while playing with a Fischer Price camping set. Farver (1992) coded for the social pretend play complexity that ranged from simple social pretend play, associative social pretend play, to cooperative pretend play. Simple social pretend play was coded if children participated in social play in which both children pretended, but there was no script that ordered the p lay sequence into a meaningful sequence (e.g. children both pretended to talk in a telephone to one another). Associative social pretend play was coded if the children participated in social play with a script (e.g. children made cookies). The most complex form of pretend play was cooperative pretend play, which was coded if children engaged in social play with a script and complementary pretend roles, which are roles that are related to each other (e.g. being a passenger and a driver). Farver (1992) found that when children were engaged in cooperative social pretend play, they used descriptions of actions and semantic ties (adding new semantic elements to their partner's previous utterance) more frequently than in simple social pretend play, whereas when t hey were engaged in simple social pretend play, they used more directives and repetitions. Children also used tags, which are placed at the end of a conversational turn in order to get a response from the other partner, more often in
8 associative and cooper ative pretend play than in simple social pretend. When children were engaged in long episodes of play, they used more descriptions of action, semantic ties and tags than in short or medium episodes. In general, 2 year olds had short play episodes and used calls for attention and repetition often, while 3 year olds also had short episodes and used frequent calls for attention, but they also used more paralinguistic cues and semantic ties than the younger children. The 4 and 5 year old children used more des criptions of action, semantic ties and tags to communicate play in their long play episodes than the younger children. Their results indicate that younger children use unsophisticated conversational strategies that allow them to communicate simple preten d play but do not require much cognitive strength from either partner. Therefore, the 2 year olds used much repetition to create a common reference point from which to play from, and to indicate that they knew the other partner was speaking to them. The 3 year olds used similar strategies as the 2 year olds, but also used changes in their intonation to show they were pretending. The 3 year olds also related their utterance with their partners' previous utterance, which was similar to the 4 and 5 year olds. Language during scripted play has not often been studied using this coding scheme. These communicative strategies can be a useful way to understand what children are trying to communicate in their scripted play interactions. Scripted Play Definitions W hile play and communication during play is influenced by cognitive developmental factors, social experience influences the way children play and communicate. Children's knowledge of particular events can influence the way they play when they are acting out these events. Scripted play occurs when children enact an event that has a temporal order with specific roles, props, and scenes
9 (Nelson & Seidman, 1984). These events are extended over time, and there is a temporal order of subevents that people are expe cted to follow. Two examples of scripted events would be eating dinner at McDonald's versus eating dinner at a restaurant. If a person is getting dinner at McDonald's, they are expected to wait in line until the cashier is available, and then order food fr om the menu. They pay first and then wait until their food is ready and then walk away to a table to sit and eat the food. At a restaurant, the same person would be expected to wait to be seated by someone, and then a server comes and gives them a menu. Th ey then order food from the menu and sit at the table and wait until the server brings it to them and pay when they are done eating. These two scripts involve eating dinner at a location that serves food, but the script is very different for each of them b ecause temporal order is key to each. Culture and Families as Providers of Scripts and Communication. Culture provides children with culturally appropriate settings and scripts that can be incorporated into their play. The scripts and language children wi ll use is mediated by their families and their families' culture. Research has shown that the way children play and communicate differs from culture to culture, and can even differ between subcultures. As Tomasello (1999) argued, children learn about their cultures' norms for behaving through observation and at times direct teaching from their parents, as well as other adults and children. Play and the way children communicate can influence what they learn about their culture and their social setting. In th is way play may help in enculturating and socializing children, or teaching them about the norms and values of their culture and society. One of the ways that play and communication is studied in culture is through
10 ethnographic studies. In these studies, t he ethnographer goes to a certain location and studies the people there for a long period of time. For example, in the book Inuit morality play Jean L. Briggs (1998) gives rich descriptions of different interactions between a little girl and members of her community. Briggs lived with a group of Inuit who were living in a small hunting camp with about 60 residents located in the eastern Canadian Arctic. The little girl Briggs discusses is named "Chubby Maata," a nickname given to her by Briggs. Chubby Maata was approximately three years old when Briggs was observing her in her hunting camp. Briggs acted as a participant observer for approximately six months, but she had previously conducted research with this group of people. The assertions that Briggs makes about Chubby Maata's play experiences are based on observation and inference, and not on any experimental data. Regardless of this, the observations Briggs made are important for considering issues of culture and learning through play. When Briggs studied the Inuit, school was not required while the children lived at this camp, but the Inuit educated their children in order to, in their words, "increase thought." Briggs stated that the Inuit "increase thought" in children by exposing them to emotional dram as that the child is unable to ignore. Even though the dramas raise dangerous questions and sometimes include violent ideas, these dramas are seen as "humored, benign, and playful" (p. 5). These dramas seemed like a type of scripted social play. For exampl e, Chubby Maata would play the role of a baby with her mother and sister, exaggerating the roles they were playing out. Even though Briggs claimed that these playful interactions are a way to socialize their children, the Inuit people do not discuss this b ehavior as a way they raise their children.
11 Furthermore, she believes that culture is a "bag of ingredients" that is used by people when they create their world, and that the elements of this bag are imbued with meaning for the people who use them (p. 14) This "bag" can include objects that help define their world. Briggs (1998) described these Inuit dramas as "testing, experimenting and seducing the child" into indirectly learning social rules (p. 106). Briggs (1998) stated that playing allows children t o practice social rules in a way that gives them a "safe transitional space" (p. 142). The dramas and interactions noted by Briggs are all social. Children do play alone, but play is often social, and this social environment is culturally mediated. Simila r to Briggs' (1998) argument, Lancy (1996) stated that children can use play as a way to try on social roles and experiment with these roles in a context in which they are safe to make mistakes. In his ethnography, Playing on the mother ground Lancy (1996 ) attempted to create an image of Kpelle society in which he highlighted what children needed to learn in order to be successful adult members of society. He gathered data in a small interior village called Gbarngasuakwelle in Liberia where he was a partic ipant observer. He observed most often, but at times he conducted structured interviews with the people of the village, or created quasi experiments in which he asked children or adults to perform certain activities and he recorded the data. Lancy stated t hat his work was consistent with previous work done with the Kpelle. Lancy's aim in the book was to discuss how culture influences child development. He defined culture as an information economy in which information is given and received. In Kpelle society there was not much evidence of active parenting in the way Western parents tend to act towards their children. For example, Kpelle children of all
12 ages were often playing with other children in the open spaces between houses called the "mother ground." A dults rarely intervened in their proceedings in this space. Rather than using direct teaching, Kpelle adults tended to influence their children by setting limits and by being an example to their children. Kpelle culture was somewhat "public" in that it was open to all, and children actively watched the proceedings that occurred around them. Children watched adults when they were playing in the mother ground, and were therefore offered many situations they could imitate in their make believe play. Lancy (199 6) stressed the need for observation before imitation, and children were provided plenty of opportunity to observe the daily lives of the adults around them. The make believe play that Lancy observed allows children to begin to understand adult work, and r ehearse social scenes that they saw on a day to day basis. This type of imitation seems to be a form of scripted play, and was acted out by children between the ages of 4 to 11. Children also use props during this form of play, such as pretend machetes, ho es and hammers. Even though some vocational skills were too advanced or specific for children to acquire via pretend play, they could use the correct language and try to understand appropriate social roles in the situation they were mimicking. Along with make believe play, games provided a good stage for Kpelle children to learn social information because children can afford to be wrong in these contexts without suffering big social consequences. Lancy called games clever because "they are fun to play and, thereby seduce the child into learning things society thinks are important" (p. 94). Kpelle games taught children information about the social order, as well as ways of counting and other basic mathematical skills. Play functions in this society in a simi lar way as Inuit (Briggs,
13 1998) and American children's play. Play is social, and it allows children to play in a more advanced way than they would otherwise (Vygotsky, 1978). Children learn skills necessary for their adult life and use objects to do this. Like play, language and communication can differ greatly between cultures as well as between subcultures. In her book Ways with words Shirley Brice Heath (1983) discussed the types and uses of language in two different communities located in the Piedmon t Carolinas: Roadville and Trackton (both pseudonyms). Roadville was the home to a white working class community working in the textile mills in the region, while residents in Trackton were black working class people who also worked in the mills. Heath des cribed these two communities while focusing on the different types of language, both oral and written, used by adults. She stressed that the way the adults of the community used language influences the way the children of each place used language. For exam ple, she stated that the way children are socialized into their communities, and the way they learn to use language is based on the ways families are structured, the ways the adults of the community define certain roles, and the way the community members u nderstand ideas of childhood. Heath focused on the idea of culture as "...learned behavior and language habits..." that are part of the shared learning of each community (p. 11). Heath argued that because of the different goals and uses of language in the different communities in the book, children developed their own culturally appropriate uses of language. They learned what was socially appropriate and what was not, and based how they communicated and interacted with others on what they learned from the adults around them. Also, children in each community played with different kinds of objects. Roadville children played with educational toys that were made for play, while
14 Trackton children played with objects that were found around the house. This ethnogr aphic study shows that cultural variations can occur within subcultures (Trackton and Roadville) of a larger culture (American culture). It is important to consider that different cultures may communicate differently because when we study children's play a nd communication cross culturally we need to understand that one style of play or communication is not necessarily better than the other, but that these styles may vary based on the circumstances within the culture. Both Briggs (1998) and Lancy (1996) sho w that culture can influence the social and physical environment, which influences the way children play. Heath (1983) gives information that shows that the culturally variable environment can influence how children communicate and play. Therefore children in one place will have different toys and scripts for their play than children somewhere else. Several other studies have been conducted involving children from different cultural backgrounds. Farver and Wimbarti (1995) measured how Indonesian children i nteracted in a play setting with their mothers and older siblings. Children were first observed in mother and sibling child pairs as they played with a bag of wooden objects. These objects were human, animal and tree figures, arched and rectangular pieces blocks and a wooden train. Wooden toys were chosen in order to avoid introducing American toys to the Indonesian play setting. Children and their families were then observed for 4 hours in their homes, and the observer took notes. Mothers were then inter viewed about their feelings of the importance of play. They found that mothers used more teaching behaviors than siblings. Also, mothers who believed that play was valuable for making their children happy participated in more cooperative social pretend pla y than those who
15 believed it was for developing social skills and intelligence. Also, the mothers who believed that play developed intelligence used more teaching behaviors. There was more cooperative social pretend play and pretend play with objects obser ved with older children when playing with their siblings than with their mothers, and there was less exploratory object play with the older children than with the younger children. Overall, they stated that the developmental progress of play for the child ren in this study was similar to Western samples in terms of the type of play with objects and their mutual involvement in cooperative social pretend play. Sibling caretaking is relatively common in Indonesian culture, which is important for understanding some of the results. Sibling interaction increased with age in this study, and older siblings used scaffolding behaviors often with their younger siblings. This relates to Vygotsky's (1978) idea of the zone of proximal development. Based on this and previo us research with Kenyan children, Farver and Wimbarti (1995) concluded that siblings can be "effective agents of socialization and facilitators of pretend play" (p. 1501). The mothers in this study did not participate in long and complex pretend play with their children. Instead, it seems as though siblings are the family members that participate in this type of play. It is important to consider the results of this study because it discusses members of the social environment (mothers and siblings) who influ ence children's play. There have also been several studies conducted that involve immigrants from other countries who come to the United States. Farver, Kim and Lee (1995) examined how culture influenced social interaction and play behaviors in both Korea n American and Anglo American preschool children. In this study, Farver et al. (1995) predicted that there would be differences in values, developmental goals and the beliefs adults had
16 about play that would shape the social interaction and play interactio n between the adults and their children. They stated that Korean culture is similar to Japanese culture in that people value social harmony and are more group oriented than Anglo Americans who tend to be more individualist. An acculturation measure that wa s given to the mothers of the target children indicated that the Korean American families were not extremely assimilated into American culture, meaning that the results about the Korean American children can probably be extended to Korean children in gener al. Farver et al. (1995) found that the Anglo children participated in more pretend play when they were at home than other types of play, while the Korean children participated in more educational play. They related this to the fact that 47 percent of Kore an mothers and only 10 percent of Anglo mothers thought play was solely for amusement. This indicates that it is probable that the Anglo mothers encouraged pretend play more than the Korean mothers. Furthermore, the Korean children scored higher on a pictu re vocabulary test than the Anglo children. The Korean children engaged in less pretend play and engaged in more parallel play than American children. Farver et al. argued that this may be the case because the Korean preschool setting did not have many toy s that could be used for pretend play, such as unstructured toys like clothes or dolls, while the Anglo preschool did have these types of toys. They also argued that children have to "abandon self consciousness and express inner creativity" during pretend play, and that this type of individuality is discouraged in Korean culture (p. 1997). Therefore the Korean children are less likely to participate in this kind of activity. This study shows that parental beliefs about children's play, the environment in wh ich children play, including materials, and the culture of the children influence the type of play the children participate in. The
17 Korean children had fewer toys in their school than the American children, which possibly made the Korean children participa te in less pretend play than the American children. This shows that the toys and objects children are able to access has some effect on how they play. In another study, Farver and Lee Shin (1997) looked at how culture influenced the frequency and expressi on of pretend play in Anglo American and Korean American preschoolers. More specifically they looked at how children used specific communicative strategies during play and the thematic content of their play sessions. Children participated in free play sess ions with other children at their preschool, and in a quasi experimental toy session with a same age, same sex partner using a Fisher Price castle toy. Familiar themes, such as playing house, seem to be common in children's social pretend play, and this ma y be because children are familiar with these themes and they can share their knowledge with each other without much effort or argument. Farver and Lee Shin (1997) found that the Anglo American children participated in more social pretend play than the Ko rean children. Furthermore, the Anglo children participated in more social pretend play when they were playing with a peer with a Fisher Price castle toy than when they were in a free play session with other children. The Korean American preschool had few objects for pretend play, and mostly had toys for gross motor play, such as bikes, while the American preschool was set up so children could move from one play station to another. This shows that different objects can have different influences on pretend play. For both populations, though, almost half of the play with the castle involved social pretend play, meaning the two groups used the same amount of social pretend play when playing with the castle. Furthermore, family roles
18 and other everyday themes w ere more common for Korean children, whereas American children participated in more fantastic play. Farver and Lee Shin (1997) argued that this may indicate that American children have more advanced ideational skills because fantastical play requires more pretense, but they state that this finding may actually indicate that American children watched television shows that had fantastical themes. Therefore Anglo American children's play may be reflective of what they are used to watching on television. The Korean children were less confrontational and less assertive than the American children during their play, and they less often directed or rejected their partner's play. Korean children described their partners' actions, while American children described t heir own actions and rejected and directed their partner's play. Both groups of children were able to adapt their speech to accommodate for pretend play. The Korean children used more formality when they were pretending to be high status people, such as ki ngs or queens, whereas American children changed the pitch of their voice in these instances. This study also shows how different cultural contexts may shape play in different ways, and how things that children are exposed to, such as television, may influ ence they way they play. Farver and Lee Shin (1997) stated, "The results suggest that there were two distinct culturally defined social environments which in part determined children's opportunities for social interaction and pretend play" (p. 551). Also, both groups of children reflect their knowledge of themes in their play with each other, which seems to allow them to engage in the play and to sustain the bout. Objects as a Factor Influencing Social Interactions and Communication. Children are born into culturally defined settings, as Farver and Lee Shin (1997) described, and in
19 these settings children use play to learn cultural knowledge from their parents and other children. Even though children are born into a cultural environment, Tomasello (1999) ar gued that until around nine months, children cannot use knowledge about others to understand and utilize things within their cultural context. This cultural environment includes culturally variable objects that children can use in play. At around nine mont hs of age, infants can imitate adult's actions that are performed on objects such as tools and artifacts, and Tomasello (1999) calls this the "first truly cultural learning" (p. 81). With this type of learning, babies learn that people use an object for a particular purpose, and in turn learn the sensory motor affordances as well as the intentional affordances of the object, or what people plan to accomplish with the object. Children learn this and use it in early symbolic play, which is when children play with objects and pretend that these objects are other objects. For example, a child might pick up a pencil and pretend that it is a hammer (Tomasello, 1999). In order to accomplish this task, the child has to understand that adults intend to do something w ith hammers. Also, the child has to detach the normal intentional affordance of the pencil, and attach a new affordance of "hammer ness." Tomasello (1999) also discussed how children use gestures to communicate. At around 11 or 12 months, children learn th at pointing is a cultural convention, and use it to communicate that they want the adult to attend to whatever they are pointing at. Children learn to take the perspective of another person in order to use an artifact in a culturally specified way, and thi s allows them to "culturally mediate their understanding of the world through that of other persons" (Tomasello, 1999, p. 91 92). In a cross cultural study, Haight et al. (1999) tried to "...articulate a culture sensitive theory of development, one that s eeks to understand how child and culture are
20 co created" (p.1477). They used data from the pretend play of preschoolers from five Irish American families and nine Chinese families in Taiwan to identify some universal, developmental, and culturally variable dimensions of pretend play. They proposed that some universal dimensions of play are the use of objects for play (physical ecology), that pretend play occurs in interaction with others most of the time within the first years of life and that interactive p lay is more complex and sustained than individual play (interpersonal context). Some developmental dimensions of pretend play are that as children develop social and communicative competency, they become more active and successful in elaborating upon partn ers' initiations and initiating pretend play (social interaction) and that pretend play serves multiple social functions (social functions). Some variable dimensions (meaning they differ by culture) include the variability in the physical ecology of the pl ay context, who is available for play (interpersonal context), how people interact with children in the play context, various uses of play for diverse socialization goals (social functions), and the content of the play. This last one is very important beca use cross cultural research seems to indicate that children use culture specific themes in their play. These themes include activities and values that are important in their culture. To test some of these proposals of different aspects of pretend play, Ha ight et al. (1999) recruited five Irish American families living in Chicago and nine Mandarin speaking families living in Taipei. The families had one to three children and the target children were 2.5 years old. The first part of the study consisted of et hnographic fieldwork. A Chinese researcher worked with the Chinese families and a European American researcher worked with the other families. These researchers took fieldwork
21 notes about the homes and play areas of the subjects. In the second part of the study, data were coded from recordings of the children at home at 2.5, 3 and 4 years of age. The children were observed twice for 2 hours each time. Finally, parents were administered formal interviews to get information about the family. The play episodes were coded for the presence of pretend play, social pretend play, and the content of parent child play. From the ethnographic research, it was found that the Irish American families lived in large homes with miniature toys, such as dolls, stuffed animals pretend food and kitchen sets, and furniture for children. Their homes usually had playrooms or family rooms. The Chinese children lived in compact apartments and had fewer toys than the American children, but they did have a stuffed animal, toy cars or a doll. From the observations of pretend play, the universal dimensions of play proposed by the authors were found to be accurate. They found that all of the children pretended with objects, including toys, and that objects are important because they serve as reference points for children to expand their play. Furthermore, they found that pretend play in these families was primarily a social activity, with family members and friends participating. Also, in terms of the developmental dimensions that they pro posed, it was found that these children became more active in their initiations of pretend play and that caregivers became less playful over time. In terms of the variable dimensions, they found that Chinese children played more with their caregivers, whil e Irish American children played with other children more. Furthermore, a greater proportion of Chinese pretend play was social than Irish American play. Also, the pretend play in Irish American families was more often initiated by children, while many of the Chinese caregiver's initiations were a
22 way to get the child to practice proper behavior. Finally, the centrality of themes varied across cultures. For example, American children enacted fantasy themes more often. 1 Haight et al. (1999) also argued that the centrality of objects within play seems to vary from culture to culture. This is due to the variation in the social as well as the physical environment. The Irish American children often used toys that were commercially marketed and related to childre n's movies. This could be why the Irish American children used fantasy themes in their play. Also, the parents of these children often purchased toys for pretending. Haight et al. (1999) stated that the Irish American children were able to use shared scrip t knowledge based on the toys used in their play. The Chinese children had few personal possessions, and therefore did not use as many objects as the American children, but instead relied upon shared knowledge of social activities. This is important to con sider because even though both groups had different objects available to them, they used objects in their play and used scripts in their play based on their toys. How Children Use Scripts. Several studies have shown that scripts are culturally variable, b ut they do not discuss how children use them (Farver, Kim & Lee, 1995; Farver & Lee Shin, 1997; Farver & Wimbarti, 1995; Haight et. al, 1999). Other studies using American children have shown that scripts allow children to maintain longer communicative and play interactions (Carelli, 1999; Furman & Walden, 1990; Nelson & Seidman, 1984). It has been proposed that scripts are extremely important for young children because they likely help organize information in a way that is easy for them to 1 This may be because of the emphasis on toys from movies within American culture, which relates to Farver and Lee Shin's (1997) study. Farver and Lee Shin (1997) argued that Anglo American children's play may be reflective of what they are used to watching on television.
23 understand and d oes not require them to stay extremely aware of what is going on in the social situation (Furman & Walden, 1990). Children are not as proficient as adults at communicating, but scripts probably allow them to play out social scenes at a higher level of comp lexity than scenes for which they do not have a script. This scripted practice then allows them to begin to represent other actions and events in a more abstract and complicated manner. One of the first studies that involved scripted play interactions was conducted by Nelson and Gruendel (1979). This study showed that preschoolers were able to engage in "true communication" when they shared script knowledge. This "true communication" consisted of advanced communication skills, such as turn taking, elaborat ion, and coordination. Without a script, children were more likely to describe their own actions rather than elaborate on another's actions (Nelson & Gruendel, 1979). Also, children with incomplete knowledge of scripts were more likely to have unstructured and undirected play, and did not have much agreement about the goals of the play compared to children with more complete knowledge of scripts (Nelson & Gruendel, 1979). Another early study defined scripts as "composed of a sequence of acts organized arou nd goals and specifying actor roles, props, and scenes" (Nelson & Seidman, 1984, p.47). Nelson and Seidman argued that scripts help structure play, but that a script alone is not enough to maintain a coherent dialogue between children. In their study, they hypothesized that the presence of objects, a shared topic, and shared script knowledge would contribute to the maintenance of dialogue and sustained play bouts. They argued that children's play that is focused on objects themselves would mainly consist of requests for objects, possession claims, or talk about the child's own actions, because
24 each child would focus on the object independent of the other child, but if objects were used as props, they may help children sustain coherent play episodes. In thei r first study, Nelson and Seidman (1984) looked at the play of six white, low SES children who were 4 years old. They were from the same classroom and matched to a peer by their teacher. There were two dyads of girls and one mixed sex dyad. Their play was audiotaped and observed in three sessions. In the first session, the children were shown a housekeeping area and were given props for baking. In the second session, they were given a sandbox with shovels and other props, and in the final session they were given two fake telephones. These three sessions were picked because they believed children would use the objects as props and that they would help them maintain their play episodes. The observer sat in the corner while the children were playing. Two out of the three pairs invoked a common script for their extended verbal discourse in each context. In the housekeeping session and the sandbox session, all the dyads responded contingently to one another and used scripts to guide their actions. In the phone ses sion, two of the three dyads planned visits when they were pretending, and one dyad did not use the telephones in their pretend play. In all of these sessions, the shortest conversations were usually requests for the play objects, and these requests were u sually made outside of the scripted topic, as predicted by Nelson and Seidman (1984). These requests tended to inhibit the establishment of joint play. For example, one child said, "Give me!" and the other child said "No" (p. 58). The children also express ed emotion as a topic in their shorter play episodes. Overall, most of the longest segments of continuous talk were in scripted contexts, and these scripted conversations were much longer than had previously
25 been reported for children of this age. Therefor e these types of props did contribute to sustaining the play conversation, as hypothesized by the experimenters. Nelson and Seidman (1984) then recruited 20 low SES children for their second study. The teachers in three classrooms paired each child with a same aged peer. These children were between 3 and 4 years of age. They were placed in a room with a large sand filled tub with shovels, a plastic bucket, a measuring cup and a large spoon. The goal of the study was to compare the number of utterances in s cripted versus unscripted pretend play, which included utterances that were fantasy based, and to determine the structure of play in physical object play. This type of play included utterances that were reality based and that included the use of play props as objects in themselves. Scripted utterances were those used to "...create a play scene of sequential elaborated activity, based on knowledge of a familiar event" (Nelson & Seidman, 1984, p.61). Out of the ten dyads, eight used episodes with scripts. The se episodes included making egg salad, Jell O and cake. Four out of the five dyads who produced at least two or more pretend play episodes repeated and developed the same event discourse in all of their play episodes. The fact that eight out of the ten dya ds used episodes with scripts indicates that scripted play interactions are common for this age group. Most of the utterances in pretend play were part of a script. Earlier studies about scripted play did not investigate the role of knowledge of scripted events, but merely showed that scripts influenced children's play. They also did not go into much depth about the role of props. Furman and Walden (1990) went on to investigate the communicative performance of preschool children as a function of their know ledge of scripts of routine events. Furman and Walden (1990) argued that "shared
26 script knowledge may provide a bridge between solitary and interactive play by providing a context within which each partner can make contributions that are understandable to the other" (p. 228). They expected children playing out less familiar scripted events to stray from the topic more frequently than children playing out the more familiar scripted event. They also expected children using an unfamiliar script to produce fewe r turns (continuous talking by one child, followed by the other child's conversation or silence), fewer statements that required a response from their partner, fewer responses directed towards their partner, more communicative failures (when one child fail ed to fulfill a conversational obligation) and a lower proportion of pairs of turns that were related to the conversation. The investigators' sample consisted of 15 dyads of 3, 4, and 5 year olds. Children were paired with a same age, same sex classmate w ho had approximately equal knowledge of the script, as was determined by a script knowledge pretest. The dyads played for approximately 15 20 minutes with one familiar script (either going to the grocery store or to McDonald's) and one unfamiliar script (e ither going on a train or a plane trip). The children were given miniature replicas of adult items, such as a shopping cart, for each event, and Furman and Walden (1990) provided approximately the same number of props in each event. They were then asked to pretend as if they were at the script location and were told to use the toys. The investigators looked at how the children maintained the rules of the game (by using role assigning statements, material assigning statements and behavior constraining statem ents), proportions of script violations (deviations from the script), turns per session, proportion of adjacent turns that were dependent on the previous statement, proportion of obligating turns, proportion of
27 communicative failures and the conceptual com plexity of their contributions to the topic. This last dependent measure was rated on a four level scale, with Level One being the least complex and Level Four the most complex. Furman and Walden (1990) found that there were fewer script violations and mo re turns in the familiar script condition than the unfamiliar condition. This means that children in the familiar script condition used their topical knowledge to follow the conversational topic effectively. They found that age was not predictive of the nu mber of interactive turns in the familiar condition, but was predictive in the unfamiliar condition. This means that in the events in which script knowledge was used by the children, the difference between the children's age and the number of turns they us ed was no longer visible. For older children, Furman and Walden found that there were more communicative failures in the unfamiliar script. Complexity was not found to differ by age or by script condition. They also found that older children had generally high numbers of interactive turns, even in the unfamiliar condition, meaning script knowledge was not very necessary for them in turn taking. This means that the unfamiliar condition did not influence the amount of talk, but rather the quality of talk as evidenced by their communicative failures in this condition. They also found that "...unscripted interactions did not produce fewer communications that initiated obligating sequences, [but that] children were less effective in responding to these communica tions when initiated by their partners" (Furman & Walden, 1990, p. 231, emphasis in original). They stated that this might show that children's ability to initiate conversation is not as influenced by script knowledge as the ability to respond to these com munications. They argue that since this occurred solely
28 with older children, older children may be more aware of the idea of the appropriateness of responses, and therefore they do not answer questions or statements by their partners when they do not have an appropriate response to give in unfamiliar situations. This means that younger children may respond to obligating statements with more inappropriate responses than older children. Furthermore, the proportion of turns in pairs of statements was not influ enced by script familiarity, but was influenced by age, with older children engaging in more turns. Also, Furman and Walden (1990) note that the props used in the study may have given the children enough external context for them to effectively communicate within the play sessions, as even the younger children were fairly effective communicators. Overall, the investigators found that only one of the measures of responsivity (communicative failures) occurred more often when children were in the unfamiliar co ndition. They argue that it is possible that the children were not responding to one another, but were responding appropriately to the objects that were given as props instead. Furman and Walden (1990) matched children by their script knowledge, meaning a child played with another child who demonstrated the same level of knowledge of the script. In another study involving scripted play interactions, Short Meyerson and Abbeduto (1997) wanted to see how script knowledge influenced conversation when children were matched or mismatched on their knowledge of scripts, and therefore whether they could accommodate the partner's level of expertise into the conversation. In one condition of this study, children were matched by script familiarity, meaning the script was familiar to both children. In the other condition, children were mismatched, meaning the script was familiar to only one child. They wanted to see how
29 children sustained the conversation, made topically relevant conversation, were responsive to one an other and understood each other. They hypothesized that children's strategies for determining their partner's level of expertise would be used less often in the matched condition than the mismatched condition. These strategies included assessing level of e xpertise, indicating level of expertise, acquiring expertise and supplying expertise. A total of 30 children participated in this study. They played in same age, same sex dyads with unfamiliar peers. Their knowledge about four scripts (an airplane trip, ba king cookies, doing laundry and going to the dentist) was assessed prior to the play session. They were given a picture recognition and a picture sequencing task to test their knowledge of these scripts. Short Meyerson and Abbeduto (1997) found that match ed script knowledge allowed children to stay on topic more often. Children also used fewer clarification requests when both dyad members had knowledge of the script than when only one child had knowledge of the script, possibly indicating that mismatched s cript knowledge created more misunderstandings. They also found that children tried to establish mutual knowledge ("assess and adapt to their discourse partner's level of expertise") more often when they were in the mismatched condition than in the matched condition (Short Meyerson & Abbeduto, 1997, p. 469). Moreover, they found that establishing mutual knowledge was negatively related with talk that was unrelated to the script in both conditions. They argue that this means that establishing mutual knowledg e is an effective way for children to maintain engagement in a scripted play session. In contrast, when both children had knowledge of the script, establishing mutual knowledge was negatively associated with the amount of talk and their maintenance on the topic. They argue that
30 this may indicate that if both partners know the script, using communicative means to establish mutual knowledge may actually interfere with the play session. Furthermore, Short Meyerson and Abbeduto (1997) found several results tha t were in contrast to previous research. For instance, the number of turns produced by a dyad did not vary based on script knowledge, which, as Short Meyerson and Abbeduto (1997) stated, is in contrast to what Furman and Walden (1990) found. This indicates that if only one child knows about the script, the children will still be likely to have some sort of discourse with one another. Another result was that communicative failures did not vary based on script knowledge, so the children were able to respond a ppropriately even if they did not share the knowledge of the script. The final contrast found to Furman and Walden (1990) was that script knowledge did not reduce the amount of talk that was unrelated to the script, but that there was more off script talk when the children both knew the script. This may be the case because if two children understood that they both know the script they might play without solely discussing the script. Short Meyerson and Abbeduto (1997) conclude by saying that script knowledge influences communicative effectiveness in some way, but not as much as would be expected from Furman and Walden and other researchers who have investigated similar issues (Nelson & Gruendel, 1979). For example, script knowledge did influence children's co mmunicative effectiveness and communicative strategies to establish mutual knowledge, but the difference between the conditions was small, even though it was statistically significant, which may indicate that scripts are not as central to organizing childr en's behavior as previously thought.
31 Scripted Play With Props. Although Short Meyerson and Abbeduto (1997) investigated the role of mismatched script knowledge in their study and argue that there are many factors to consider when looking at children's sc ripted play interactions, they did not discuss the role of props in their study. They gave the children in their study toys, but did not discuss the role of these toys in their play. The environment has an influence on children's play, so it's important to see how toys influence play. Furman and Walden (1990) noted that the role of props in the play interactions was unclear and that "...investigating children's ability to play out events with major props missing, thus testing their ability to fill in the bl anks with their script knowledge, may illuminate the role of props in facilitating interaction" (p. 232). Furman and Walden (1990) and Short Meyerson and Abbeduto (1997) had provided the children in their study with props, therefore the role of props has n ot yet been clearly explicated. Other studies have indicated that toys influence play (Farver, Kim & Lee, 1995; Farver & Lee Shin, 1997; Farver & Wimbarti, 1995; Haight et. al, 1999), but Carelli (1999) was the first to attempt to explicitly answer questio ns about the role of props in communication during scripted play interactions. She argued that the internal context (script knowledge) and the external context (physical setting) of a play episode might ease communication between two preschoolers. For th is study, Carelli (1999) manipulated the level of script knowledge and the type of physical setting in two different experiments. In Experiment 1, Carelli had two conditions, a free play condition, and a familiar scripted event condition. Carelli chose a f amiliar event (getting up in the morning) based on 48 children's ratings of their familiarity with eight common events. In each condition, she told 10 pairs of same age
32 same sex (4 and 5 year old) children that they could play together in a room while she observed them. In the free play condition, the children were brought to a room and were asked to talk and play together using toys she had provided (wooden blocks, dolls, cars, trucks and Lego blocks). In the morning condition, Carelli explained that they were to play in a room that was similar to a room in a house, and that they were to play together and pretend they were waking up in the morning and getting ready for school. The room contained realistic props for the script (bed with a pillow and blanket, sink with a towel, soap, toothpaste etc). The number of props in each condition was the same. Carelli hypothesized that shared script knowledge in the bedroom context would help facilitate conversation more than in the free play context. Children's conve rsations were analyzed by the number of turns, number of conversations, and the length of conversation. Turns were coded when there was continuous talking by one child that was then followed by talk from the other child or silence, and this was coded as an individual measure. Conversation was defined as four or more turns that were focused on a particular topic, and this occurred when the children demonstrated semantic and communicative contingency. Semantic contingency occurs when the discourse shares a co mmon topic, and communicative contingency occurs when the utterances of one child are directed at the utterances of another child. The length of conversation was based on the number of turns between topic shifts. Carelli (1999) found that there were higher levels of performance in the number of turns, conversation, and length of conversation in the scripted condition, but the scripted condition produced only marginally longer conversations than the free play session.
33 In the second experiment, Carelli (1999 ) manipulated both event familiarity and degree of contextual support. She conducted a pilot test with adults (teachers and parents) to choose a highly familiar event and a less familiar event for children for use in the experiment. Going to sleep was chos en as the most familiar event, while going to an exhibition of paintings was chosen as the least familiar event. The bedroom was created using realistic props (bed with a pillow, bedside table, lamp, pajamas), while the exhibition setting consisted of pain tings hanging on a wall. The number of items in each context was the same. None of the 16 children between 4 and 5 years of age had ever been to an art exhibition, but they knew what an art exhibition was, although Carelli (1999) does not note how she obta ined this information about their familiarity. Each dyad played in the familiar context and the unfamiliar context, with half the dyads using props, and the other half using no props in an empty room. They were instructed to talk and act out what they woul d do in each situation. The same analysis was completed involving the number of turns, number of conversations and the length of conversations as in Experiment 1. Replicating the findings of Experiment 1, children in Experiment 2 exhibited better performa nce with the familiar event than with the unfamiliar event, and conversation was more advanced when there were objects in the setting than when they were provided no objects. The effect of event familiarity was limited to the condition in which children we re given props to play with, meaning conversation was more advanced in the familiar script condition where there were props, but not where there were no props. Carelli's (1999) design had some problems that make the results hard to interpret. For example, the children in the no prop condition were placed in an empty room with
34 nothing to play with. The strangeness of this condition was probably something the children had never experienced, and they may not have been sure how to act in the situation. Further more, in the prop condition at the art gallery, children were not able to physically touch or handle the props (paintings hanging on the wall), whereas they were able to play with the items in the bedroom setting. This difference in the ability of children to handle the materials might have influenced the results she obtained, and this makes it hard to conclude much from her study. The effect of props is not knowable from this manipulation because the children were not able to handle the props, and the abil ity to touch props should be held constant across conditions. To remedy this problem, Carelli (1999) could have given the children art they could handle, or she could have chosen a more normal unfamiliar condition for children of this age group, such as go ing on a train (like Furman and Walden's (1990) experiment), that would allow usable props to be provided. Although Carelli (1999) tried to investigate the role of props in scripted play events, her design had flaws and she measured only three dependent v ariables. Therefore there is still a gap in the literature about how objects influence how children communicate in play. The physical environment, which includes toys, clearly has some effect on how children play and communicate during play. Children with access to less pretend play toys participated in less pretend play than children with those types of toys, also, it seems as though the developmental progress of play for the children in other cultures is similar to Western samples in terms of the type of play with objects, and that centrality of objects within play seems to vary between cultures (Farver, Kim & Lee,
35 1995; Farver & Lee Shin, 1997; Farver & Wimbarti, 1995; Haight et. al, 1999). What effect different props have on communication in scripted pla y is still not entirely clear. The proposed study would investigate children's communicative interactions in a scripted play event, and would try to explicate the role of props in these interactions. Preschool children would play with a same sex peer. Chi ldren would be given either abstract props or realistic props for their play interaction. Abstract props are objects that can be used in play in multiple ways, such as boxes or blocks. Realistic props are more specific objects, such as fake money or a shop ping cart. Looking at how communication differs with these two types of props can add to knowledge about how one aspect of the physical environment (props) can influence play. Children's conversation would be analyzed for how the role of props would influe nce the duration of play episodes, the number of turns in an episode, the communicative strategies used during the play, the definition of roles, and children's violation of the script. Children who would be placed in the abstract prop condition would be expected to use more definition of roles statements than children in the realistic prop condition. If abstract toys are similar to having no toys, as Carelli (1999) found, children with abstract props would be expected to have shorter duration of episodes On the other hand, if Carelli's (1999) data reflect the strangeness of her manipulation (children were unable to handle some of the props), abstract props may be as conducive to play and communication as realistic props. Children in the abstract prop con dition would be expected to use paralinguistic cues, descriptions of actions and directives more often than children in the realistic prop condition. Children in the abstract prop condition would likely have to establish mutual knowledge about the props an d their actions since the
36 props would not be clearly defined at the start of the play session. They would likely do this by defining roles, and using paralinguistic cues, descriptions of actions and directives (Short Meyerson & Abbeduto, 1997). Research Pr oposal Participants A total of 64 preschool children (32 boys, 32 girls) would participate in this study. They would be between the ages of 4 and 5 years. Children would be recruited from preschools in Western Florida. All children would be attending the selected preschools at the time of the study. Materials Children would be given props to play with while they participated in a scripted play event. The scripted play event would be going shopping at the grocery store. Half of the children would be given abstract props, which are nesting boxes that fit one inside the other (See Appendix A). Half of the children would be given realistic props (grocery cart, cash register, fake food, See Appendix B). The number of props would be kept equal for both groups an d the distribution of large, medium and small props would also be identical. Procedure Pretest. Children's script knowledge of "going to the grocery store" would be assessed in a pretest with the experimenter. Each child would be tested individually in a private room at the preschools with two tasks adapted from Short Meyerson and Abbedutto (1997). In the picture recognition task there would be 7 script components that the children would be tested to recognize (Furman & Walden, 1990). The children would
37 be presented line drawings one at a time, half of which would be script components and half of which would be nonscript components. The experimenter would label each action as the drawings were presented in random order and would ask the child if it was part of the event of going grocery shopping, and they would be asked to say yes or no. In the picture sequencing task, the experimenter would put the script component pictures in front of the child in random order, with the first component placed in front of a nd to the left of the child and with all children seeing the same order. The child would be asked to identify what happens next in the sequence (for example, "Find the picture that the person would do next and point to the picture that she does next."). Th e picture selected by the child would be placed to the right of the first picture. These tasks would be used to assess whether the child had knowledge of the components of the script and the correct temporal sequence. A child's score for the picture sequen cing task would be the proportion of pairs of pictures in the correct sequence, except for the first continuous pair. Based on Short Meyerson and Abbedutto's study (1997), a child would need to answer correctly (respond yes for script components and no for non script components) 79% of the time, and would need to get a temporal sequencing score of at least 80% for the picture sequencing task to be included in the next part of the study. Play Session After script knowledge would be assessed, children would be randomly placed in one of two conditions: a realistic prop condition and an abstract prop condition and paired with a same sex partner. This experiment would place children in dyads because it has been shown that pretend play occurs in interaction with others most of the time within the first years of life and that interactive play is more complex and sustained than individual play (Haight et al., 1999). All the children would know each
38 other. The play session would occur in a preschool classroom. They would be familiarized with the video camera used for recording their play session, and would be told the experimenter was going to observe them while they played. To begin the session, they would be introduced to a puppet. The experimenter would ask them to pretend the puppet had never been to a grocery store and did not know how to grocery shop. They would be told to play as if they were going grocery shopping to show the puppet how people do this activity and to use the objects they were given in their play. They would be instructed not to talk to the puppet or the experimenter, as adapted from Furman and Walden (1990). They would be video recorded for 15 minutes while they played or until they stopped playing. They would be given a book in return for th eir participation. Dependent Measures Children's play sessions would be transcribed using the CHILDES formatting standards ( MacWhinney, 2000) Data would be coded according to several coding schemes. First, play episodes would be identified following Farv er's (1992) method. Play Episode. An episode is defined as an interactive sequence occurring between the two children containing three or more exchanges of continuous discourse with a shared theme or topic. It would begin when one child verbalized and end when (a) the partner did not address the topic of the previous turn, (b) if any child's attention was not on the play activity for more than 30 seconds, or (c) either child moved away from the play activity. This would be a dyadic measure. Turn. Each chi ld's turn is defined as all or part of that child's utterances before the partner took the floor. This would be an individual measure.
39 Duration of Episode. The duration of the dyad's play episode is the number of conversational turns within an episode. Th is would be a dyadic measure. Communicative Strategies Next, communicative styles and strategies used by the children would be coded as identified by Corsaro's coding scheme (1986, as cited in Farver, 1992, p.505). Seven communicative strategies would b e scored based on the number of times each strategy occurred during a specific turn within each episode. These measures would be coded as frequencies for each child. 1. Paralinguistic Cues are changes in intonation and pitch to mark fantasy and animation of objects ("Cha Ching" for the sound of a cash register). 2. Descriptions of Action are declarative statements with ongoing activity or describing past or future action ("I'm going to the checkout line"). 3. Repetitions are when a child repeats prior ut terance(s) of their partner (A says, "I like to shop" and B repeats it). 4. Semantic Tying is when children add new semantic elements to their partner's previous contribution (A says, "The line is long!" B replies, "The line is long because people are buy ing food"). 5. Calls for attention are when a child utters something that is used to get their partner's attention. ("Hey!" "Look!"). 6. Directives are declaratives used to control a partner's action ("Do this or else you can't bring your food home!"). 7. Tags are verbal devices put at the end of a conversational turn as a way to get a response from the other partner ("It's time to check out, right?").
40 Rules of the Gam e. The frequency of "Rules of the Game" statements would be calculated for each dyad adapted from the coding scheme from Furman and Walden (1990). These statements would indicate children's knowledge of the roles played by people in certain contexts, the behavior that makes sense for these roles, and the materials that make sense for the s cripted context (Furman & Walden, 1990). 1 Role assigning statements would be coded when children assigned roles to one another ("You be the cashier and I'll be the shopper"). 2. Material assigning statements would be coded when children assigned roles to materials ("These blocks are cookies"). 3. Behavior constraining statements would be coded when children limited the behavior of themselves or their partner ("You're the shopper so you can't use the cash register") Script Violations The proportion of script violations would be coded as deviations from the script (Furman & Walden, 1990). For example, the statements "Do you have a dog at home?" or "I want to go to the airport" are violations. This would be a dyadic measure and would be created by tak ing the number of script violations and dividing by the total number of utterances. General Discussion The proposed study would contribute knowledge about how different types of props influence children's play and communication. Carelli's (1999) study sho wed that children do not communicate well when they do not have toys to play with, but she did
41 not investigate the influence different types of props have on communication. The current study would give us a better idea of how different types of toys influe nce children's communication and play. It would expand the communication variables measured to give a broader view of children's scripted play. Other types of play should be studied to more fully understand the way objects, or props, influence communicatio n during play. It would also fix the problems in Carelli's study by giving all the children props they could physically handle. Even though the proposed study only includes American children, it would be important to study children from other cultures to compare results. As evidenced by the cross cultural research, culture influences the physical environment. The proposed study is limited to American children, but a similar procedure would likely work for children in other places.
42 Appendix A Abs tract Props (Nesting Boxes)
43 Appendix B Realistic Props (Shopping cart with food)
44 References Briggs, J.L. (1998). Inuit Morality Play New Haven: Yale University Press. Carelli, M.G. (1999). Event representation and contextual support Communication and Cognition, 32 (3/4), 173 188. Cole, M., Cole, S.R., & Lightfoot, C. (2005). The development of children (5th ed.). New York: Worth Publishers. Farver, J.M. (1992). Communicating shared meaning in social pretend play. Early Childhood Research Quarterly, 7 501 516. Farver, J.M., Kim, Y.K., & Lee, Y. (1995). Cultural differences in Korean and Anglo American preschoolers' social interaction and play behaviors. Child Development, 66 1088 1099. Farver, J.M., & Lee Shin, Y. (1997). Socia l pretend play in Korean and Anglo American preschoolers. Child Development, 68 544 556. Farver, J.M. & Wimbarti, S. (1995). Indonesian children's play with their mothers and older siblings. Child Development, 66 1493 1503. Furman, L.N., & Walden, T.A (1990). Effect of script knowledge on preschool children's communicative interactions. Developmental Psychology, 26 (2), 227 233. Haight, W.L., Wang, X., Fung, H.H., Williams, K. & Mintz, J. (1999). Universal, developmental, and variable aspects of youn g children's play: A cross cultural comparison of pretending at home. Child Development, 70 (6), 1477 1488. Heath, S.B. (1983). Ways with words New York: Cambridge University Press. Howes, C., Unger, O., & Matheson, C.C. (1992). Mastery of communication o f meaning
45 in social pretend play. In C. Howes, O. Unger, & C.C. Matheson (Eds.), The collaborative construction of pretend (pp. 13 24). Albany, NY: State University of New York Press. Howes, C., Unger, O. & Seidner, L.B. (1989). Social pretend play in t oddlers: Parallels with social play and solitary pretend. Child Development, 60, 77 84. Lancy, D.F. (1996). Playing on the mother ground New York: The Guilford Press. Nelson, K. & Gruendel, J.M. (1979). At morning it's lunchtime: A scriptal view of chil dren's dialogues. Discourse Process, 2 73 94. Nelson, K., & Seidman, S. (1984). Playing with scripts. In I. Bretherton (Ed.), Symbolic play: The development of social understanding (pp.45 71). Orlando, FL: Academic Press, Inc. Short Meyerson, K.J. & Abb eduto, L.J. (1997). Preschoolers' communication during scripted interactions. Journal of Child Language, 24, 469 493. Tomasello, M. (1999). The cultural origins of human cognition. Cambridge, MA: Harvard University Press. Vygotsky, L.S. (1978). Mind in s ociety: The role of play in development. In M. Cole, V. John Steiner, S. Scribner, & E. Souberman (Eds.), Mind in Society (pp.92 104). Cambridge, MA: Harvard University Press. Youngblade, L.M., & Dunn, J. (1995). Individual differences in young children' s pretend play with mother and sibling: Links to relationships and understanding of other people's feelings and beliefs. Child Development, 66 1472 1492.