Children’s play has been an enduring area of interest for both developmental psychologists and early childhood educators. During the past 50 years, significant advances have been made in defining this fascinating activity (cf. Garvey, 1977) and its correlates (e.g., Fein, 1981). However, questions have also been raised regarding the presumed contributions of play to children’s development and education. Recent exchanges about whether or how play promotes development can be witnessed in previous and recent reviews (e.g., Lillard, 2015; Lillard, Lerner, Hopkins, Dore, Smith, & Palmquist, 2013; Rubin, Fein, & Vandenberg, 1983), and the dialogue that has emerged in relation to these reviews (e.g., Berk & Meyers, 2013; Nicolopoulou & Ilgaz, 2013; Weisberg, Hirsh-Pasek, & Golinkoff, 2013). In our view, several tacit but untested agreements on the existing definitions of play combined with an equally unexamined conviction that “we recognize play when we see it” have resulted in misunderstandings and conflicts regarding the nature of play and its role in development.

Our goal in this article is to examine the existing criteria for childhood play in the allied fields of developmental psychology and early childhood education using the general principles of sociocultural perspectives. We focus our discussion on what has been variably referred to as imaginative, pretend, symbolic, or fantasy play. The distinguishing feature of imaginative play is one in which a signifier (e.g., a banana) is used to represent the meaning of the signified (e.g., a telephone). As described by Vygotsky (1933/1967), this feature exemplifies the separation of meaning from object and action: the banana becomes a phone, and the act of bringing the banana to one’s ear enables “talking on the phone.” While pretense is widespread in play, and potentially universal, cultural variations in the quantity and quality of pretense, as well as the role of pretense and imaginative play in cultural activities, exist (see Gaskins, 2013). Cultural differences—including the values of particular activities, ways of participating in activities, and opportunities to participate—shape the role of pretense and imaginative play in different cultures. Thus, the question of what is or is not “universal” in imaginative play requires more research.

In what follows, first, we briefly describe the basic tenets of sociocultural perspectives that constitute the basis for our claim that the existing criteria for imaginative play should be expanded. Afterwards, we describe the four features of imaginative play that are commonly assumed to be central definitional criteria—intrinsic motivation, the open-endedness of goals, pleasure seeking, and an escape from reality—and expand these features moving them from assumptions to research questions. First, we move from the assumption of intrinsic motivation to the query: What are the motives for imaginative play and how are they constructed in the relation between child and environment? Second, we move from the assumption of open, free, and goal-less action, to the query: What are the goals for imaginative play and how are they constructed in the relation between child and environment? Third, we move from the assumption that engagement in play is a pleasure-seeking activity, to the query: What affective dimensions emerge in imaginative play and how are they constructed in the relation between child and environment? Fourth, we move from the assumption that play is a paradox, to the query: What, how, and why do features of reality and imagination emerge in imaginative play? Expanding definitional criteria in this fashion allows researchers to remain open to cultural variations, rather than imposing dominant cultural assumptions as explanatory heuristics. In addition, we discuss why and how each criteria could be expanded to elucidate both existing areas of debate and ways to resolve debate in future research through a framework of expanded criteria.

Why sociocultural perspectives?

Sociocultural perspectives on the study of human development flourished as extensions of Vygotsky’s theory after the publication of Mind in Society (Vygotsky, 1978), and more comprehensively since the six-volume set of The Collected Works of L. S. Vygotsky was published between 1987 and 1998 (Vygotsky 1987–1998). In reviewing the existing literature, Göncü and Gauvain (2011) stated that despite some variations in their emphases, sociocultural perspectives share the conviction that human development takes place through historically situated activities that are mediated by children’s cultures. Through their participation in cultural activities, individuals engage in negotiations with others appropriating the “cultural tools” that are necessary for their current functioning. In addition, individuals transform the use of these same tools as they craft potential social futures (Vadeboncoeur, Perone, & Panina-Beard, 2016). Evidence exists that individuals appropriate a wide variety of cultural tools through their engagement, and these tools shape the development of psychological functioning. For example, young children learn how to guide their attention to what is expected of them in problem-solving situations by developing intersubjectivity through dialogue with their mothers (e.g., Wertsch & Stone, 1985).

Taken together, sociocultural perspectives share three specific claims that are essential to the argument advanced in this article to expand the existing criteria for imaginative play. First, cultures vary in terms of valued activities and, further, the kind and frequency of activities that are available for members’ participation. These activities may range from domestic labor and chores to leisure-time activities to labor outside the home (e.g., Goodnow, 1988; Larson & Verma, 1999). Thus, each culture implicitly and explicitly presents members with sets of values about the activities through which it is enacted. Sometimes these values are tacit and unanalyzed, and other times they are explicitly stated; sometimes they are shared by all segments of a society or only certain segments. The values include who should participate, how one participates, and when participation is desirable. As such, each culture serves as a guide about the kinds of activities in which children are expected to engage, as well as those they should avoid. Illustrations of this claim can be seen in many sources (e.g., Rogoff, 2003; Tudge, 2008). Research on variations in play across cultures is also evident (e.g., Edwards, 2000; Göncü, Mistry, & Mosier, 2000; Roopnarine & Johnson, 1994).

Second, human learning and development always occurs through participation in cultural activity with others, including culturally specific ways of participating in activities. Children learn to talk, remember, problem-solve, and tell right from wrong by living with others in environments saturated with cultural values for how to participate. This is important to keep in mind because even those so-called solitary activities of individuals, children, or adults are guided by their experiences with cultural others in the past and/or by their anticipation of engaging with others in the future. Moreover, each culture presents its members with norms and expectations of participation in cultural activity. For example, each culture has norms and expectations about whether or how young children can engage in conversation when adults are having a conversation among themselves (e.g., Rogoff, Mistry, Göncü, & Mosier, 1993), as well as pragmatic rules about whether or not what we are saying in a given moment is appropriate in given social relationships and context (e.g., Miller, 1982).

Third, a culture is not merely a context presenting activities and framing them with values and expectations, but each culture is a source of meaning and opportunities for meaning making. On one hand, it is a shared conviction among the sociocultural perspectives that children’s development is, in part, a process of socialization into existing systems of activities and meanings. In addition, and potentially more significant, however, development is a process of individuation; through development individuals contribute to culture and shape their own developmental contexts (Vygotsky, 1934/1987). This position is exemplified in Vygotsky’s discussion of word meaning and concept development: Any word a child learns is shaped by evolving meanings that become more complex and deeply felt through experience and participation in cultural activities. Experiences mean something in the moment, and this meaning evolves and changes over time and in relation to others. Subsequent research on concept development has provided evidence for how this contextual process unfolds in educational settings (e.g., Bakker & Derry, 2011; Wells, 2008).

In what follows, we argue that unless we study each culture as a system of valued activities, ways of participating, and opportunities for meaning making, we cannot understand how children construct imaginative play, the content of their imaginative play, or, and perhaps more important, why children engage in particular forms of imaginative play. Further, we argue that, without attending to culture, much of the research conducted regarding imaginative play to date has been narrowed to reflect Eurocentric assumptions. Expanding definitional criteria enables researchers to attend more comprehensively to the ways in which culture shapes imaginative play.

From intrinsic motivation to investigating motives for play

The commonly accepted motive for play is described as an intrinsic one; however, this criterion is often assumed, rather than examined, in play research. The distinction between intrinsic versus extrinsic motivation has been paramount in the study of achievement motivation (see Ryan & Deci, 2000). Intrinsic motivation describes a desire that comes from within the individual to engage in an activity while extrinsic motivation describes engagement in an activity given external rewards. Emphasized in this distinction is a sharp division between internal and external sources of motivation. In addition, neither in the case of intrinsic nor in the case of extrinsic motivation is the social or cultural context of the individual acknowledged. However, a closer look at the conceptualization of motive for play reveals important insights for consideration in future research.

The first foundation from which to question the assumption of internal motivation for play is a theoretical one. It derives from the claims of theorists, such as Piaget (1945) and Freud (1961), who noted that imaginative play has its origins in affectively significant social nonplay experiences in the world. Indeed, what leads children to play is a nonplay experience that takes place in children’s daily living involving interactions with others. For example, Piaget (1945) claimed that children engage in the world of imaginative play in order to test their mastery over what they experience in the nonplay world with others. His examples included his children pretending to be a dead duck after having been exposed to a dead duck to be cooked for the family meal, and pretending to be a church steeple after having been shown a church building. In addition, Freud (1961) claimed that children pretend in order to heal affective wounds caused by experiences that are threatening, such as going to a doctor. In these examples, what motivates children to move into the world of imaginative play is experiences with others. As such, conceptualizing motivation for play as simply intrinsic undermines the social origins of play.

Further, however, from a sociocultural perspective, play emerges from a dialectical relationship between the individual and the social environment, or what Vygotsky (1935/1994) called perezhivanie: the relationship between the individual and her environment, including the meaning of the experiences for the child. Vygotsky (1933/1967) stated that children pretend in order to engage in actions, realize tendencies, and “play at” becoming competent in practices that they are not able to perform in real life. A classic example involves children in an imaginary situation in which a piece of candy is labeled toxic and, given the rules of play, the children refuse to eat it; perhaps something that they would be less inclined to do outside of play. A second example highlights an increase in conscious awareness when two sisters pretend to be “sisters” in their imaginative play, thus, allowing them to reflect upon the rules for being sisters as a foundation for their performance of “sisters.”

Research from mother–infant/toddler pretend play illustrates that mothers actively engage in socializing children to the world of pretend play. For example, research shows that some mother–infant dyads from middle-class backgrounds introduce their children to imaginative agendas in their play and encourage and cajole their children to engage in it (Haight & Miller, 1993). Further, by introducing imaginary scenes, mothers serve as directors of play and motivate children to adopt imaginative roles for socialization (Haight, Wang, Fung, Williams, & Mintz, 1999). Similarly, Göncü and Jain (1999) found that in a middle-class community in Turkey, mothers actively entered into children’s nonplay activities and introduced pretend elements. In one instance, a child was holding an embroidery hoop given to the mother–infant dyad to be explored by them. While the infant was holding the hoop, the mother introduced an imaginary theme by saying, “Let’s pretend that this is a steering wheel and you are driving.” The child took up the suggestion and pretended to drive. On another occasion, when the mother of a toddler was asked why she thinks imaginative play is important for young children, she stated that her course in child development provided her with insights about the developmental benefits of imaginative play. Based on this and similar examples, it is plausible to argue that the motivation for imaginative play exists in social interaction shaped by cultural expectations and resources.

The relation between child and social environment is also extant in research on children’s peer play during preschool years. This line of research corroborates the theoretical claim that proposals for imaginative play are based on day-to-day social nonplay experiences. For example, Corsaro (1983) illustrated that in social pretend play, when the agenda introduced for pretend play by one child is not familiar to others, the proposal is not followed. He gave the example of a girl who wanted to pretend with her peers that they view her baby brother on a TV screen. However, since the other children were not familiar with this experience, the proposal was not taken up in play. Göncü and Kessel (1984) observed two girls pretending to be mothers with no children. After this experience, Göncü learned that a new sibling had arrived for one of the children, and the other child was expecting a sibling. According to the adults, being a mother with no children allowed the children to work through their difficulties and make sense of the feelings they were experiencing toward their new siblings. What these examples illustrate is that imaginative proposals adopted for a group of children in play have social origins and, thus, rely upon experiences of potential players in some way.

Indeed, there is an inconsistency in play research regarding the claim of intrinsic motives: The claim denies the ways that adults have shaped the environments in which children live, particularly at home, in early childhood education and K–12 education settings. It is undeniable that research and classroom training interventions have themselves shaped the definitional criteria for the features of imaginative play in constructing these environments (see Freyberg, 1973; Saltz & Johnson, 1974). Here, cultural expectations for valued activities and ways of participating shape what play looks like and the meanings made in play. The relation between a researcher’s, or an educator’s, criteria for play and the opportunities for imaginative play in any given context shape the motives for imaginative play with significant implications for how to study it.

From an open-ended activity to investigating the goals of play

The second feature of play that appears to be accepted without question is that play is an open-ended activity (e.g., Brown, 2009). As a corollary to this, play is described as an activity where children pay attention to the process rather than to the end result. This feature renders play as a free-flowing activity often layered with improvisations, yet without a specific plan (e.g., Sawyer, 1997). For example, when children pretend to play house or imagine to be superheroes, they focus on how to conduct their actions as their play evolves without concern for where play ends, or its goal(s). At first glance, this feature appears to describe young children’s imaginative play fully; however, assuming this without examination misses the ways in which play can be a goal-directed activity. We propose five insights for expanding this definitional criteria for imaginative play.

First, sociocultural theoretical insights quickly reveal that play emerges through the child’s relation to the social environment, the meanings she has made of her experiences, and her interpretations of the rules that are entailed by taking up a particular role in play (Vygotsky, 1933/1967). Indeed, children engage in the illusory world of imaginative play to realize tendencies that they cannot realize in a nonplay world. For example, when children pretend to be sisters, they do so to practice roles and behaviors that are consistent with being sisters. This becomes possible because the roles adopted for play bring with themselves rules that structure actions, ways of being, and speech patterns that children will draw from and recombine as they play (Vygotsky, 1930/2004). There are individual and cultural variations involved in the enactment of these roles, but the roles themselves are fashioned through experience, and this undermines the assertion that play is a totally open-ended activity. Indeed, Vygotsky (1933/1967) argued that compliance with the rules of the roles adopted for play enables children to prepare for future experiences. By practicing role-appropriate rules in play, and surfacing inappropriate rules in play, children become able to adapt and transform them later in life.

Second, that play is not always as open-ended as it appears becomes evident when we consider adult–child play. While features, such as flexibility, repetition, and variation, appear open-ended, a close look at adult–child play often shows the adult and/or child agendas that exist under the guise of ostensible openness. In fact, parent–child or teacher–child play involves great attention both to the means and the ends. In our own research, we witnessed that Western parents and teachers are often goal-oriented and create a means of engaging in specific forms of play with young children (Göncü, Jain, & Tuermer, 2007). For example, Göncü, Abel, and Boshans (2010) documented that in pretending to be a patient of a child who was pretending to be the doctor, the teacher engaged the child in practicing language and literacy lessons. Similar observations are made in parent–child play when parents use the activity to offer language, literacy, and math lessons, as well as socialization (Göncü, Tuermer, Jain, & Johnson, 1999; Haight et al., 1999; Roopnarine & Davidson, 2015). Although the focus of this research has been on ascertaining the negotiation of adult goals, additional research has highlighted the bidirectional relationship between the goals of both parents and children (Vadeboncoeur, 2017).

Third, it is important to remember that play does not always occur independently of income producing activities or chores (e.g., Goodnow, 1988; Kağitçibaşi, 1996; Larson & Verma, 1999). It is often possible to see play embedded in work in cultures where children’s participation in income-producing activities is not prohibited. Existing research indicates that in many countries, such as India, Brazil, and Turkey, children work as street vendors, earning money. In such contexts, it is possible to see children play while they work. Clearly, ours is not an argument in support of child labor, but a recognition that childhoods differ and that imaginative play can occur in settings and in ways that are not immediately apparent to researchers (see, on childhoods, Lancy, 2008; Wells, 2009). It is in this sense that we can argue that play is not an open-ended activity since it occurs within the context of other goal-directed work activities.

Our fourth objection to play being defined as an open-ended activity has its roots in historical, social, and cultural conceptions of childhood that are differentiated by social class and race (see Lareau, 2011). Cultural communities have tacitly understood notions about what defines acceptable play behavior. Children are not always allowed to do everything they want because the activity is labeled as play; cultural notions of “acceptable play behaviors” establish boundaries for what is tolerated. If play were, indeed, completely free, we would not have any need to provide explanations about behavior in play at times of conflict among children: “No, no, he is not fighting, he is only playing.”

Fifth, although there have been debates about how children construct their imaginative play narratives, an examination of them indicates thematic and structural integrity. One group of scholars claim that children’s play narratives are organized around an action plan or scripts (e.g., Garvey & Berndt, 1975; Göncü, 2017; Seidman, Nelson, & Gruendel, 1986), while others claim that they are constructed through free-flowing improvisations (e.g., Fein, 1981; Sawyer, 1997). We do not know for certain how children construct their imaginative narratives. To answer this question, we would need to talk to children themselves. However, examinations of children’s play texts (i.e., the transcription of their play sessions) appear to show a structure that moves between experience and imagination. Regardless of whether play activities are constructed with a priori plans or the structure emerges, play seems to be shaped by the ongoing negotiation of shared goals. If we examine children’s play in relation to their social and cultural context, we will gain a clearer idea of the goals adopted for play and why certain performances are preferred and embellished in variation while others remain almost prescribed and limited to certain ways of enactment.

From the pleasure of play to the affective dimensions of play

Play is often understood as a pleasurable activity (e.g., Freud, 1961; Piaget, 1945). Consistent with the view that play is intrinsically motivated are descriptions of the predominant feeling in play as joy. Children engage in play because of the joy experienced; joy becomes a rationale for and a result of play. This is evident in the repetitive and exaggerated play behaviors that are often coupled with smiles and laughter. Examples of joy experienced in play can be found in many accounts, including autobiographies (de Beauvoir, 1994), scientific literature (Sutton-Smith, 1979), and psychiatry (Brown, 2009). Cultural notions of play as unrestrained activity and the opposite of work seem to reflect play as a universally joyous activity; as such, we do not question the potential for play to elicit the full range of emotions, from joy to conflict to frustration.

Against this background, however, there are insights, beginning with Vygotsky’s (1933/1967) and reflected in more recent developmental literature, that play may not always or even primarily be pleasurable. Three different sets of recent research provide evidence for an extension of this feature of play. The first set of evidence reviewed in Lillard (2015) is that human children play under conditions of stress imposed on them. For example, we have evidence that children in concentration camps engaged in pretend play. Also, young children represent their experiences of sadness, disappointment, and loss in their play (Clark, 2003). This is true for both typically developing children as well as those who have experienced emotional trauma and/or are diagnosed with emotional difficulties (Gordon, 1993).

Second, depending on the lens of the researcher, this kind of play is sometimes referred to as “cruel play,” “illicit play,” or “evil play,” based on the observed content of the activity (cf. Sutton-Smith, 1979). In this vein, there are two views on why children’s play may not always involve pleasure for all the participants. One such view is offered by Sutton-Smith (1979) who argued that children’s engagement in cruel play is the result of their desire to gain power and construct identity. For Sutton-Smith, children of the Western/industrialized world grow under oppression from all social institutions: family, school, places of worship, and social clubs. In these institutions, adults see their role as guiding children’s socialization into cultures and contexts and, therefore, they offer a limited range of activities and often under close supervision. Therefore, when they have the opportunity to do so, children use play as a breathing space, and to practice ways of being on their own terms. This process reflects children’s perceptions of adult domination and their responses to it in the form of play, such as teasing, insults, mocking, and representations of adult roles. For Sutton-Smith, this type of play appears more frequently after preschool years, and although it may occur in the classroom, it usually occurs behind closed doors or outside the immediate view of educators and/or adults. While it may be pleasurable for the players, it is degrading to those who are the subject of children’s play.

A second view is described by Schousboe (2013), who stated that “evil play,” defined as “an antisocial act through which a person—for his or her own personal gain and without having been attacked—causes somebody else to suffer,” does occur in the activities of some young children. Giving examples from her own observations with appropriate caveats, she discussed the potential uses and abuses of evil play. One of Schousboe’s examples for evil play involved two 4-year-old girls, one being a “mother” and the other being a “doctor,” treating a younger girl as the “sick child.” The older girls urinated in a soda bottle and pretended to use the urine as “medicine” to treat the sick child. The younger girl refused to drink the “medicine,” and as a result, was excluded from the play group for being “too young.” In another example, a group of “armed” boys pretended to murder another boy. As part of their play, they cornered the target child, struck him in the stomach, and put a “knife” on his throat. When an adult intervened, the target child was out of breath and pale. When the pack was questioned about why they treated the target boy as such, they said that it was a part of their play. In the end, all involved appeared bewildered. After describing these examples, Schousboe provided the caveat that “evil games” do not have to be allowed (see also Berk, Mann, & Ogan, 2006). However, she also raised a significant question for research: Do all evil games have harmful effects, or is it possible that, by engaging in evil games, children learn to distinguish good and evil in play without having to confront it in “real life”? These questions remain to be addressed in future work.

Third, affect in play is sometimes described in terms of children’s relationships, yet these observations still require us to question the assumption that the dominant feeling observed in play is always joy. The common denominator of the insights from the three studies we report next is that power hierarchy among children may shape the feelings associated with play. The first example comes from the classic work of Schwartzman’s (1978) ethnographic study of kindergarten children. In her extensive work, Schwartzman observed that children’s relative roles in the organization of their play were related to their status in the classroom. In this study, a girl named Linda was a dominant child in the peer group, and she also took this role in play by assigning roles and offering directions about the course of pretend events.

Consistent with this, Göncü and Weber (2000) obtained informal insights from a preschool teacher regarding a stable relation between children’s peer status and the roles adopted in their imaginative play. The issue emerged when the teacher interrupted an imaginative play episode as it was being undertaken by a group of children. They were in the process of deciding who was going to be the house pet, and it was at this time the teacher intervened reminding children of a classroom rule: They had agreed not to make peers play the role of pet when they played house. Intrigued by the observation, when the teacher was asked why this kind of play was not allowed, she stated that there was a relation between children’s status in the peer system and in the play group. Usually, children who had lower status were given the role of house pet in play, further establishing their low status and powerlessness. In turn, when children became house pets in their imaginative play, their status in the peer group was lowered.

According to Gaskins (1999; Gaskins & Miller, 2009), who reported consistent findings from her research based on Mayan children’s play in the Yucatán, children’s play is devoid of intense affect of any kind since children’s lives and play are highly regulated by older peers and siblings. On the one hand, children do not experience psychological conflicts that require working through in symbolic play. Therefore, the occurrence of imaginative play is not as frequent as it is in the Western world. On the other hand, when children do play, what they are allowed to do is often controlled by a script overseen by older children. As such, Mayan children do not experience pleasure, or lack thereof, in play, as ascribed to Western children.

In summary, rather than assume pleasure as the rationale and result, play includes a range of affective experiences and representations of both enjoyable and troubling experiences. In certain cases, children’s play may not be experienced as pleasure due to cultural events, like wars or conflict, or due to interactional differences, like having a lower status imposed by older children; both are beyond a child’s control. Further, children’s play may be regulated either by adults, older siblings, or peers who control the representation and experience of affect. And finally, in situations where children can express their feelings, those feelings may vary both in intensity and in kind. From a sociocultural perspective, cultural events and interactional differences must be taken into account not simply as what Vygotsky (1935/1994) called “objective factors,” but in relation to the meaning individual children make of these factors.

From play as paradox to investigating the real and imagined in play

Perhaps the most definitive feature of play resides in its paradoxical nature regarding the assumed incompatibility between what is real and what is imagined. Regardless of the form in which it occurs, a play behavior connotes nonliteralness. Actions in play range from simply throwing a ball with a smile on one’s face to indicate the action is an invitation to play, rather than a chore, to a player telling a partner, “Let’s do beauty shop!” when they are outside in the backyard of their house. In both cases, the messages inform the partner that the actions are expected to be interpreted for what they represent, rather than what they actually mean. In his well-known work, Bateson (1955) made this point succinctly in relation to research with animals. Working with dogs, Bateson stated that play actions are framed by metacommunication that indicates to the partners that initiations, such as a playful nip, do not denote what they would denote outside of play (i.e., aggression). Rather, a nip is an invitation to practice this activity in a play context.

Bateson’s (1955) ideas made their entrance into the psychological literature most prominently in the 1970s (cf. Garvey, 1977; Schwartzman, 1978; Sutton-Smith, 1979), and their foundation in empirical work appeared in the 1980s (e.g., Giffin, 1984; Göncü & Kessel, 1984; Howes, Unger, & Seidner, 1989). What was paramount in this early research was that metacommunicative frames can be established both by nonverbal (e.g., smiles), as well as verbal messages, such as, “Let’s pretend.” Moreover, verbal messages can be direct, as in saying, “Let’s play house,” and they can be indirect, as in announcing “baby is hungry,” referring to the doll the child holds in his hands. However, Western (e.g., U.S.) children appeared to use explicit invitations rarely. They devoted much of their attention to the enactments of activities. Children often initiated their pretend play by moving directly into the activity through their enactments or planning statements.

Despite these advances, however, after the 1990s, play research almost abandoned examining important dimensions of children’s communication in and about imaginative play. To date, we do not know how children at different age levels and in different cultural, social class, and gendered contexts initiate, maintain, and terminate their play. Although some significant ethnographic work in sociolinguistic and literacy studies remain as guides (cf. Goodwin, 1990; Lee, 2000), developmental researchers have not taken advantage of such advances.

Children’s transformations in play serve at least three purposes simultaneously. As they establish intersubjectivity (Göncü, 1993), they inform partners about the changing nature of their activity from nonplay to play as well as the content or topic of their play. Western theories have mainly focused on object substitutions and role-play in examining both the communicative and the representational power of play transformations. However, in support of the thesis that play is a social and a cultural activity, recent work has illustrated that symbolic representations in pretend play are not limited to object substitutions or role-play. Following Bateson (1955), some researchers have conceptualized play as an expressive activity and hypothesized that pretend play representations are extensions of the symbolic traditions of children’s communities. As part of this thesis, Göncü et al. (1999) claimed that in communities with strong oral traditions, children’s symbolic representations are likely to include examples of those traditions in addition to role and object transformations. And, in support of this expectation, results indicated that children in low-income communities in Chicago, both African American and European American, and in a rural community in Turkey manifested examples of pretend play that were distinctly different from object substitutions or role-play. For example, in three communities, play included sound and rhythm play, whereby children make variations of a tune. Fun-filled teasing was another example. Finally, there were different examples of language play. Based on these pieces of evidence, Göncü et al. (1999) argued that communication and representation in symbolic play are shaped by broad cultural traditions of communities.

Given these findings, there is a need for research that examines not only the kind of realities reflected in children’s play, as well as the interweaving of experiences and imaginings, but also how children do so. Our call explicitly includes the play of children from diverse backgrounds, in non-Western, urban, and rural contexts, and in communities that reflect different ways of being and engaging with the world. Unfortunately, much of the play research has been conducted in a laboratory setting, as an evaluation of cultural groups, and/or has imposed a narrow framework for definitional criteria.

Concluding thoughts

There is still much to learn about how socially and culturally valued activities, ways of participating in activities, including imaginative play, and opportunities for meaning making shape how children engage in imaginative play. Indeed, the definitional criteria of imaginative play—as intrinsically motivated, open, and pleasure-seeking behavior that reflects a paradox between real and imagined worlds—are consistent with dominant cultural perspectives and, as such, they need to be investigated rather than assumed. Sociocultural research has contributed to highlighting the limitations of these assumptions with research within and across diverse cultures (e.g., Göncü et al., 2000, 2007, 2009), in addition to conceptually challenging Eurocentric perspectives in play research (Göncü & Vadeboncoeur, 2015) and calling for a recognition of imaginative play as a culturally constructed activity (Gaskins, 1999, 2013). In the present effort, we have provided evidence that counters the definition of imaginative play as an intrinsically motivated, open-ended, pleasurable and a paradoxical activity and called for expanding the definitional criteria to shift dominant cultural assumptions to research questions. This effort reveals three significant insights that are foundational for future research.

First, research on children’s imaginative play must examine and retain the social and cultural contexts of children’s lives through observations of children in their daily routines and over an extended period of time. This sort of study requires ethnographic research rather than laboratory research. Research that captures imaginative play as it occurs spontaneously in naturalistic settings will allow researchers to gain information about the kinds and frequency of imaginative play activities of children who grow into different cultural contexts. In addition, studying children over time will enable us to establish connections between their nonplay experiences and their resultant representation in imaginative play. Although ethnographic and observational methods are time intensive, systematically and naturalistically observing imaginative play in relation to children’s lives will shed direct light on the kinds of social and cultural experiences that motivate them to engage in the world of imaginative play, and perhaps guide us to see various, potentially competing, motives for participants in play. Further, this sort of research will make visible goals as they emerge, change, and are negotiated between participants, along with affective differences over time and between participants. Rather than perpetuating dichotomies—intrinsic versus extrinsic, real versus imagined—ethnographic research may make visible the blending of these categories along a continuum.

Second, if, as theory holds, representations in imaginative play emerge from children’s own experiences, any request to engage in representational activity in laboratory research is likely to reveal more information about children’s capacity to follow an experimenter’s or educator’s directions than about a child’s symbolic development in play. Acknowledging that imaginative play emerges from the relation between child and social environment, including the meaning of the activity for the child, not only highlights the conduct of research and provides a cautionary tale for potential biases in educational efforts, but it also speaks to the significance of learning from the participants in play, the children themselves. While researchers would need to engage with participants in play in ways appropriate for development, much could be learned from children on the role of play in their lives, for example, on the objects, actions, and narratives that shape their play. In addition, it is important to discuss with children what motivates them to engage in play, how their thinking and feeling shapes specific engagements in play, and how their play builds from and extends their lived experiences. A sociocultural perspective provides a counter to the laboratory research in which children are asked to engage in specific pretend activities by the researchers; it also has significant implications for how play is incorporated and/or supported in the classroom.

Third, imaginative play emerges, develops, and is continually negotiated as a cultural activity in relation to other cultural activities that shape the lives of children in different communities. While outside the scope of this article, cultural activities are influenced by macrolevel factors—including political structures, economic resources, oppression, discrimination, and privilege—that are interpreted in different ways in microlevel interactions. More research that links microlevel ethnographic interactions with these sorts of macrolevel factors is sorely needed. We see a recognition of the cultural biases evident in play research that equate a dominant perspective with an objective perspective as a small step to remediate this complex issue. However, researchers themselves exist within political and economic systems that produce and reproduce oppression, discrimination, and privilege. Recognizing, acknowledging, and remaining open to the ways in which cultural perspectives shape our research questions, preferred methodologies, methods, and forms of analysis—as well as the research outcomes invariably produced by these decisions—requires reflexivity on the part of researchers. In addition, conducting research on imaginative play in relation to political and economic systems must cautiously avoid the location of identified “differences” inside the participants themselves.

A significant contribution of a sociocultural perspective, then, is moving the location of imaginative play from a locus inside individuals to the relation between individuals and their social and cultural environments. We see this as enabling a crucial shift in research on imaginative play: one that may eventually allow us to go beyond describing to explaining imaginative play. As a result, we conclude that not only will we develop a more complete understanding about imaginative play but, based on this understanding, we will also have a more robust foundation from which to examine its correlates, benefits, and developmental outcomes.