1

Play’s role in language development

RUNNING HEAD: Play’s role in language development

In: Weisberg, D., Zosh, J., Hirsh-Pasek, K., Golinkoff, R.M. (2013). Talking it up” Play, language development and the role of adult support. American Journal of Play. Special issue, 6, 39-54.

Talking it up: Play, language development, and the role of adult support

Deena Skolnick Weisberg (University of Pennsylvania)*

Jennifer M. Zosh (Penn State University, Brandywine)

Kathy Hirsh-Pasek (Temple University)

Roberta Michnick Golinkoff (University of Delaware)

*corresponding author

Department of Psychology, University of Pennsylvania

3720 Walnut St., Solomon Labs

Philadelphia, PA 19104

(215) 898-1712

To appear in American Journal of Play, special edition.

Abstract (word count = 116)

Play contextsfacilitate children’s language learning. This review evaluatesevidence for this claim, showing that play benefits children’s language development because playful environments incorporate many of the social-interactive and cognitive elements known to promote language growth. Although much of this evidence is correlational, a series of recent intervention studies offer evidence that a key variable relating play and language is the presence of adult support.In particular, guided play situations, in which adults scaffolding child-initiated learning, seem ideal for encouraging language development. Based on this argument, we conclude that understanding the efficacy of play for learning requires careful attention to the type of play and to the outcomes that result from these different types of play.

Talking it up: Play, language development, and the role of adult support

Play as a context for language development

Language is the currency of social interaction and school achievement. Thus, it is hardly surprising that thousands of pages have been devoted to understandingand encouraging optimal language acquisition in children (e.g., Clark, 2003; Clark& Clark, 1977; Dickinson, Griffith, Golinkoff, & Hirsh-Pasek, 2012; Frank, Goodman, & Tenenbaum, 2009; Golinkoff & Hirsh-Pasek, 1999; Golinkoff, et al., 2000; Harris, Golinkoff, & Hirsh-Pasek, 2011; Hoff & Naigles, 2002; Hollich, Hirsh-Pasek, & Golinkoff, 2000). Many of these investigations find that language thrives when children are interacting with adults and peers in a playful manner (e.g., Hirsh-Pasek & Golinkoff, 2003; P. Smith, 2010; Vygotsky, 1967; Zigler & Bishop-Josef, 2004), suggesting that play contexts can make important contributions to children’s language learning. Evaluating just how play might influence language, however, is a relatively new enterprise.

In a recent and influential paper, Lillard and her colleagues (2013) describe three categories of potential links between pretend play and other cognitive or social skills, including language, based on Smith (2010). The strongest potential link that one could posit isthat play is a unique and crucial component in the development of cognitive or social outcomes, such that theseoutcomeswould not develop without play. This causal view would claim that it is impossible for children to achieve, say, a certain kind of self-control if they do not engage in play; this skill would be impossible to acquire otherwise. An intermediate claim is that of equifinality: Play has a causal effect on the construct at hand, but is only one of many activities that can lead to gains in this outcome. In this case, play would lead children to develop self-control abilities, but other activities would do so equally well. The third and weakest claim holds that play has no role at all in the development of a given outcome but ismerely an epiphenomenon or a byproduct of the learning situation. This would mean that play is involved in children’s achievement of the kind of self-control in question, but not because of anything intrinsic to the play situation. Rather, something involvedin the play situation, such as increased social interaction, does the developmental work.

Although Lillard et al.’s (2013) categories clearly delineate several possible relationships that play could have to cognition and in particular to language development, phrasing the options in such stark terms may inadvertently lead researchers todiscountthe more complex relationship that exists between these two skills. Put simply, asking whether play causes language development may be the wrong question. Instead, this paper rephrases the question to ask: “What aspects of play contexts might promote language development?” Under this framework, it becomesclear that play contains many of the ingredients that are necessary for optimal language growth, even though there may be no single element of play that does the majority of the work. Although this means that we must accept an equifinality view on Lillard et al.’s (2013) account, this does not mean that play’s role in language development is unimportant. To the contrary: Even if no single aspect of play is a necessary or sufficient condition for a particular language outcome, various aspects of play when taken in the aggregate link play and language.

In this paper, we review several lines of evidence showing connections between play and children’s language development. We use this evidence to suggest possible mechanisms whereby play contributes to language development.

An issue of definitions

As a prerequisite to thinking about links between the two constructs, it is imperative that we operationally define language and play.We begin by describing the kind of language that we will focus on in this paper, and then turn to the issue of defining play for the current purposes.

Languages are communicative systems that encode meaning through combinations of arbitrary symbols. Children acquire the rudiments of their language (or languages) by the age of 3. At that time, they can converse with strangers, make their desires and opinions clear, ask questions, and discuss the past and the future. There are many different aspects of language that we could discuss, but the current paper focuses specifically on vocabulary and grammar development.We will thus not explore the role of play in learning how to produce smaller linguistic structures, such as phonemes,nor larger structures, such as narratives (but see Ilgaz & Nicolopoulou, this issue).Children who establish the fundamentals of their vocabularies and syntactic skills in their early years are wellequipped to enter school and to succeed socially and academically.Those who do not, such as a disproportionate number of children from lower socio-economic brackets, often fall behind (e.g., Dickinson & McCabe, 2003; Gershoff, 2003; Hart & Risley, 1995).

While linguists have provided relatively clear definitions of the characteristics of language, those working in the domain of play research still wrestle with how best to define thisconstruct, which has blurry edges and takes many forms.

“Play” can refer to just about any activity that children do that meets a number of criteria and can range fromwordplay (e.g., crib speech, children’s spontaneous riffing on phrases and making rhymes; Nelson, 2006; Weir, 1962) toworldplay (e.g., paracosms, elaborate and extended imaginary worlds created by a single child or group of children; see Root-Bernstein, 2013). In line with previous research (e.g., Burghardt, 2011; Fisher, Hirsh-Pasek, Golinkoff, Singer, & Berk, 2011; Garvey, 1990; Hirsh-Pasek, Golinkoff, Berk, & Singer, 2009; Hirsh-Pasek & Golinkoff, 2003; Johnson, Christie, & Yawkey, 1999; Pellegrini, 2009; Sutton-Smith, 2001), our definition of play emphasizes several specificfeatures that can be used to distinguish play from other activities. First, play has no specific purpose and it is not linked to survival. Second, playful activities are often exaggerated such that an action done in pretense often takes longer or involves a wider range of motion than the same action done in reality. Finally, play is joyful and voluntary.

Our concern here is with language development, so in this paper we will focus on those types of play where we will be most likely to observe play’s impact on language development. Specifically, we will consideronly play in the early years, up to around age 6. Although both play and language develop past this point(see Roskos & Christie, 2001), play likely has its greatest impact when children learn the fundamentalsof language. An additional limit on our current definition of play is that we will only consider instances of play that have some degree of sociality, because children are particularly likely to use and hearlanguage when social partners are present.

Finally, we emphasize that play is child-led. While it is notoriously difficult to make a sharp delineation between play and other kinds of activities, a key way to distinguish between these two categories is by looking at who has the locus of control. In play, children’s interests determine how the interaction moves forward, not adults’. This is a crucial distinguishing feature of play, and one of the main ways that can help us as researchers to determine whether a given interaction qualifies as play or as merely playful. This is also one of the features that characterizes guided play,which is a form of play wherein adults scaffold children’s active exploration in service of a learning goal (Fisher, Hirsh-Pasek, Golinkoff, Singer, & Berk, 2011; Hirsh-Pasek, Golinkoff, Berk, & Singer, 2009; Weisberg, Hirsh-Pasek, & Golinkoff, 2013b). In such situations, children are free to play and interact however they like, but a sensitive adult prepares the environment and joins in the play to help subtly respond to and guide children to focus on specific constructs. This situation still counts as play because the adult’s role is primarily to follow the child’s lead. When the adultsare in control, the situation is more like work dressed in play clothes --- what Bruckman (1999) termed “chocolate covered broccoli.” Such activities might be playful, and even fun, but they are not technically play.

Our definitions in place, we next ask: Do play contexts offer important support for language growth?

The role of play in early language development

Four characteristicsof play potentiallylinkplay and language outcomes. First,many forms of play enlist symbolic thinking. In object-substitution pretense, for example, props are symbols that stand in for real objects, as in the classic banana-as-telephone scenario. This relationship between a prop and the object it represents is similar to the relationship that a word has to its referent. Because both play and linguistic communication share a representational character, play provides children with the opportunity to practice forming symbolic relationships (e.g., McCune, 1995; Miller & Almon, 2009; Piaget, 1962; Vygotsky 1967).Indeed, Tamis-LeMonda and Bornstein (1994) found that the frequency of symbolic play was related to children’s language comprehension at 13 months and to the diversity of a child’s speech at 20 months. In a similar investigation, Laasko, Poikkeus, Eklund and Lyytinen (1999) observed solo symbolic play in 14-month-olds and found that children’s level of language comprehension was associated with their symbolic play competence at 18 months, as measured by a version of the Symbolic Play Test (SPT; Lowe & Costello, 1976). However, it could be that children who played more gained better language skills, or that children who had better language skills played more. Future studies should work to unravel the direction of these correlations. In addition, children can learn generic knowledge about animals from a symbolic pretend context.For example, when presented with a pretend scenario that illustrated one novel animal’s behavior, children readily generalized that behavior to new instances of that animal in real life (Sutherland & Friedman, 2012, in press).These studies suggest that the connection between play and language development may rely on theshared symbolic nature ofboth activities.Creating object-symbol relationships during play may allow children to practice distancing themselves from the here-and-now, preparing them to use words as abstract referential symbols. But this argument is still largely theoretical, and would benefit from studies focusing on whether and how play develops children’s symbolic understanding.

A secondcharacteristic of play that may feed language development is the social interaction inherent in many play situations. Insocio-dramatic play, children establish a play frame (e.g., going to the store) and collaborate to carry it out. Such situations often involvethe need to negotiate roles and coordinate actions, whichtypically recruits the use of mental-state verbs (e.g., “want”, “explain”) (Pellegrini & Galda, 1990). Further, children often switch back and forth between speaking within the play frame (“Can you give me some milk?”) and providing commentary or stage directions (“I’m the baby; you be the mommy”) (see Bretherton, 1989). These two factors --- taking on roles and negotiating --- point to aspects of social play situations that appear toencourage children to practice more advanced linguistic forms than they would use in non-play interactions.As Bruner wrote, “the most complicated grammatical and pragmatic forms of language appear first in play activity” (1983, p. 65). The evidence suggests that children are indeed more likely to use complex language in play situations than in non-play situations (Bergen & Mauer, 2000; Christie & Enz, 1992; Christie & Roskos, 2006; Singer, 1998), although future work should more carefully document such instances.

A third characteristic of play that contributes to language development concerns the sheer amount of language input available in many play situations. It is well known that the amount of language children hear is strongly related to their overall linguistic skills (e.g., Hart & Risley, 1995; Hoff, 2006; Hoff & Naigles, 2002; Hurtado, Marchman & Fernald, 2008; Tamis-LeMonda & Bornstein, 2002).Play with adults and peersbolsters language development because it encourages greater language use. For instance, the amount of talk children use with their peers during play in preschool is positively related to their vocabulary size in kindergarten (Dickinson & Moreton, 1991), as well as to measures of early literacy (see Roskos & Christie, this issue). Additionally, Dickinson and Tabors (2001) examined the relationship between talk during pretend play and language skills among low-SES children from 3 years of age through kindergarten.Increased amounts of talk during play with peers were associated with better comprehension and production. Indeed, a recent clinical report rightly advocates for the importance of play, especially for children of low-SES backgrounds (Milteer, Ginsburg, & Mulligan, 2012).

Finally, play might be important for language development because, when children are in control of an interaction, they are engaged: They produce and hear language about what they are particularly focused on.When talk refers tothings that the child is concerned with, he or she does not have to switch attention from his or her own focus of interest to that of another. In line with this argument, a child ismore likely to learn novel vocabulary items when an adult followsthe child’s interest as opposed to makingthe child to follow the adult’s interest (Dunham, Dunham, & Curwin, 1993). Even toddlers recognize the role of social cues in vocabulary learning.They only learn new words when a speaker is clearly labeling an object for their own benefit, and not when the speaker mentions the label outside of that kind of one-to-one interaction (Baldwin, Markman, Bill, Desjardins, Irwin, & Tidball, 1996). Thus, social play with other children and adults sets the child up to learn new words and sentence structures because the child is deeply involved in the common play scenario. A direct test of this claim should measure both children’s engagement and their degree of learning in play and non-play situations.

These are just four of many possible characteristics that could explain whyplay has been positivelyrelated tolanguage development in the literature. However, the studies reviewed above, and many others that show associations between more play and more advanced language skills or larger vocabularies (see citations in Lillard et al., 2013, pg. 18 and Table 9), are largely correlational. It thus remains unclear whether playing causes children to develop better language skills, or whether children with better language skills play more, or whether a third variable is responsible for both better language skills and increased play.Indeed, some researchers have found that children with larger vocabularies are more likely to engage in symbolic play (L. Smith & Pereira, 2009), suggesting the reverse direction of causation from the one we describe above.Thestrongestcase for play’s contribution to language development thus comes fromintervention studies.

Intervention studies of play and language development

Lillard et al.’s (2013) reviewnotedfour intervention studies that directly examine links between play and language:Christakis, Zimmerman, and Garrison (2007), Levy, Schaefer and Phelps (1986); Lovinger, (1974); and Smilansky (1968), and there are other studies that address the same issue. In general, these studies increased the amount of play, by providing children with more time to play or with some play training, and reported subsequent increases in language.These studies bolster the claim above that play scenarios serve to facilitate language development. For example, Lovinger (1974) found that a group of 4-year-olds who engaged in one hour of pretend play every day for 25 weeks showed increases in language relative to a control group, who received no intervention. A similar study(Bellin & Singer, 2006)involvedperforming a video-based intervention with low-SES preschoolers. Parents, teachers, and other caretakers were provided with engaging video materials that instructed the participants how to guide a “Magic Story Car” through a number of activities to improve emergent literacy skills, such as phonemic awareness and letter recognition. A key part of this intervention was that children were not simply passively watching a program on TV. Rather, the video served as a spark to encourage beyond-the-screen activity. Following the video, caretakers and children created their own Magic Story Cars and engaged in pretend games about their own adventures. In justtwo weeks, children in the experimental group showed significant gains in vocabulary and emergent literacy skills compared to a control group.