Abstract
This study examined the relationship between four common types of language play and their correlations with the verbal and social abilities of 3- to 5-year-old children. While observation has shown that children this age produce a range of play, research has not yet examined whether play is a measurable skill connected to preschoolers’ language development. In this study, we designed four language play elicitation games involving creating rhymes, word switching, word creation, and hyperbolic play. Children’s ability to produce novel play for each game was measured, and classroom teachers filled out assessments of children’s verbal and peer interaction skills. Results indicated that while children’s peer interaction scores were not related to their play scores, verbal skills scores were highly correlated with language play scores, highlighting that for preschoolers the ability to produce common types of language play may be a distinctly verbal skill.
While almost any 3-year-old child can be observed occasionally misusing a word or making one up entirely, often to the comic delight of the adults around them, formal research has largely ignored the implications behind children’s intentional play with their language. It is not long after children begin speaking in basic words and then sentences that they learn to subvert conventional structures to not only surprise and amuse people but also to convey meaning creatively and poetically. Previous research has pointed to children’s play with language as the beginning of their understandings of richer uses of language like humor and poetry (Crystal, 1996; Holmes, 2011; Howard, 2009), but children’s ability to produce verbal play in this age group has not been systematically explored. Play appears rapidly as children age, and children end up developing play in a variety of settings: from the kitchen table to the playground and with a range of partners: from parents and teachers to siblings and other children (Geller, 1985). But while all children seem to engage in verbal play to some degree, there are differences in the precocity, the amount, and the types of language play preschool age children seem to engage in. What do those individual differences connect to in the children’s development?
Verbal play takes many forms and is a regular part of most young children’s early communicative experience. Repetition as play is ubiquitous in games, books, and songs for young children and even within their self-generated language play, especially early on. Through interaction with parents, siblings, classmates, and teachers, children use repetition not only to practice and develop their linguistic knowledge but also to play with language in the most basic form (Howard, 2009). Unlike repetition of specific words to practice pronunciation or cement meaning, repetition in play is used to garner humorous responses or elicit play from play partners, such as repeating the punch line of a joke to provoke another laugh (Howard, 2009). However, studies show that children even as young as 30 months learn to implement more advanced types of play in their social lives such as rhyme and word switching (Crystal, 1996).
Children in preschool, specifically ages 3–5 years, begin to exhibit a number of fascinating playful verbal skills in the midst of rapid language and social developments. In home observations, research has shown that children as young as 2–3 years can mimic and produce their own forms of joking or word play (Cameron et al., 2008). Three- to five-year-olds in our target age range have been observed to produce novel jokes (e.g. “knock-knock” routines) and word play, especially when supported or encouraged by teachers or parents (Fitzgerald, 2008; Hoicka and Akhtar, 2011; Horgan, 1981). Similarly, in observations of 3- to 6-year-olds during preschool lunchtime, Holmes (2011) has shown that verbal play in the form of rhymes and jokes is the fourth most frequent discussion topic, behind topics including the food they are actually eating and daily routines. Thus, play is common and varied among preschool children. While theories suggest that children’s comprehension of the humorous intent behind play and older (6- to 12-year-old) children’s ability to comprehend the incongruities and ambiguities inherent in riddles and jokes may relate to their cognitive development (e.g. Bernstein, 1986; McGhee, 1971, 1979; Schultz and Horibe, 1974), to date, research has fallen just short of analyzing the potential relationship between children’s ability to produce language play in the preschool years and the development of their broader verbal and social skills. This leaves us with questions about the underlying cognitive and social abilities involved in “successful” verbal play at this young age.
As we explored these questions about whether play could be measured and connected to other developing skills in the preschool years, we also found it necessary to categorize the most common types of verbal play in this age range. In a recent survey, parents answered questions regarding different aspects of language play with their 2- to 5-year-olds, including how often the children used different forms of language play, and anecdotes of individual instances of language play (Read, 2014). Based on these observations as well as previous studies examining children’s use of rhyme (Crystal 1996; Dowker, 1989; Holmes, 2011), sound play (Ely and McCabe, 1994; Esposito, 1980), word play (Horgan, 1981; Johnson and Mervis, 1997), and hyperbolic play (Varga, 2000), we organized the types of verbal play children in this age range typically produce. While these categories have distinct aspects, occasionally individual instances of language play can be examples of more than one type and have overlapping categorization, such as a child creating a word in order to make a rhyme work (e.g. “dog, log, plog”) which would count as both phonological/poetic play and word creation. Even with some overlap, however, we have established five distinct categories 1 that occur frequently in this age range as follows:
Sound play;
Phonological/poetic play;
Word switching;
Word creation;
Hyperbolic/boastful play.
An orthogonal but integral dimension of the language play of young children is also the use of humor to initiate and perpetuate play with others around them, inherently creating social games. Humor can be used in play in many different ways; some types of language play are more likely to co-occur with humor than others, and individual differences occur frequently. Phonological/poetic play, including play on words, mockery, rhymes, and puns, is one play type in which there can be humorous and non-humorous intent (Howard, 2009; Lems, 2011). An example of this with poem creation is the difference in presentation by the child—whether they indicate their humorous intention with laughter or looking for a response, or if they are merely creating a poem for their individual enjoyment of the sound. However, two types of language play that almost always have humorous intent are word creation and word switching, as seen in the “playful nonsense variations of words … and deliberate mislabellings” observed by Cekaite and Aronsson (2005: p. 17) and incongruous labelings observed by Horgan (1981), as well as almost all examples of deliberate word switching or word creation given in parent anecdotes (Read, 2014). Humor in children’s language play is often easy to recognize, and researchers have used important social cues to understand when children are using language play with a humorous intent versus non-humorous intent by observing laughing, smiling, or seeking a reaction from a play partner (Hoicka and Akhtar, 2012). These social cues fuel the question of whether or not skill in language play is connected more to social abilities, verbal abilities, or a combination of the two.
Sound play
One very common type of language play is sound play. Sound play can be defined as children’s deliberate sound distortion, word repetition, singing, and random noises (Ely and McCabe, 1994). Children use sound play in a variety of settings: for entertainment on its own and to accompany action both as an identifying tag, such as onomatopoeia, and to add to physical play (Esposito, 1980). Sound play can serve as a dramatic feature of larger play, like the use of cat meows to identify that a child is role-playing (Ely and McCabe, 1994). In survey data, parents’ examples of their children’s sound play were among the most common types noted in anecdotes, especially within the youngest 2-year-old age group (Read, 2014). Parents specifically discussed the most common instances of sound play as singing, making up new sounds, and distorting words. One parent even stated that her child enjoyed stringing magnetic letters together on the fridge and having her read the resulting non-words, illustrating the intentional fun young children have subverting the regular sounds of language (as well as demonstrating an understanding of letter–sound correspondence).
Phonological/poetic play
Even in simple classroom observations, children can be heard playing with the pronunciation of words or using phonological parallelism by creating nonsense word variations in a more systematic way than mere sound play. For example, when a child hears the rhyme and begins to string words together such as “dog, frog, log, bog,” there is a patterning present that goes beyond simple noise making or unique phonotactics found in basic sound play. This more systematic play with phonology makes up phonological/poetic play. In Dowker’s (1989) experimental study, almost 60 percent of children between the ages of 2 and 6 were able to produce short original poems (either with rhyming or with alliteration) when they were given examples, and while there were no consistent age-related changes in the ability to produce poems in this broad group, it was still suggested that being able to use rhyme to make poems could indicate a level of phonological awareness that might help with early literacy. Researchers have also made naturalistic classroom observations of children implementing rhyme, metaphors, and verse to play with language in conversation with their peers, and parents can readily provide examples of rhyme play from their 2- to 5-year-olds. One parent surveyed described the following: We play rhyming games together frequently. He says a word and wants me to rhyme it, and then he tries to rhyme it, and so forth. So if we start out with “sky,” I might say “pie” and “fry” and “why,” and he might say “fly” and “pie” and “bye.” (Read, 2014)
Word switching
Through observations, researchers have found that what we have labeled word switching, or intentionally mislabeling or switching the labels for words, is another form of verbal play (Cekaite and Aronsson, 2005; Horgan, 1981; Johnson and Mervis, 1997). Children coordinate their deliberate mislabeling with humorous language play or cues to indicate a playful nature. Word switching in language play has been proposed as an aid to helping children learn new words and concepts; the argument goes that through word switching children can enjoy confirming their new knowledge of the meanings of new words by then constructing their opposites (Geller, 1985). In a sense, playing with alternative meanings or “wrong” words confirms a child’s knowledge of the “right” words or meanings. In McGhee’s (1979) stage model of humor development, this kind of incongruous labeling indicates that children have moved into stages 2 and 3 (typical of preschoolers) where verbal humor begins. For example, in the Read (2014) survey, one parent stated that when she asks her 3-year-old son what a word means he always replies with a smile and says “it’s a truck” even when he knows the correct label for the object and clearly knows truck, illustrating this kind of deliberate word switching used playfully, almost as a mock of the standard “what is this called?” vocabulary learning routine.
Word creation
Another area of language play that appears frequently in natural observation, but has not been studied systematically, is what Johnson and Mervis (1997) observed as “word play” and what we have categorized as the use of word creation for language play or to elicit a humorous reaction from playmates. Again, parents gave numerous examples of “silly words” that their child had recently made up and provided anecdotes in Read’s (2014) survey. One parent perfectly described the use of word creation for language play purposes in her twin 4-year-old sons: My sons make up new words all the time, but not because of a story. They discovered they could get a rise out of their Grandfather by changing the endings of words to become their own, new words is fun for them. “Backhoe dodo” or “backhoe lodo” instead of “backhoe loader.” Making up new words for things and changing words to be silly and make us laugh is one of their favorite games.
While word creation can be combined with other types of play, it is a unique category in that it requires children to create novel and humorous multi-syllabic phonotactically correct sounds that function like words even while the child is fully aware that they are not words. This indicates a greater level of complexity than the simple novel noises of sound play, and a seemingly greater level of creativity than word switching with familiar labels while having some elements of each.
Hyperbolic play
Finally, hyperbolic play and boasting offer a deeper look into the nuances of children’s language play and the cognitive skill that may be involved in keeping the play going. Children co-establish tones of speaking around their boastful games, and are conscious of enforcing perceived rules for these games, using facial expressions and differences in intonation to indicate play. One example of hyperbolic play is a game in which one child says, “I’m so hungry I could eat a whole chicken!” and another child says, “Well I’m so hungry I could eat this whole table!” or the whole room, or the whole world, or the whole universe. As each child contributes a turn to the play, they must not only find something larger to “eat” but also something more outlandish or creative. Hyperbolic play of this kind is especially common when children are in groups and illustrates the humorous and social aspects of language play, while also highlighting the differences in children’s comprehension of this form of play. While some children are successfully able to come up with greater boasts to outdo one another, other children struggle with the concept and cannot create successful hyperbolic interactions (Varga, 2000) and will contribute statements such as “I’m so hungry I could eat this apple!” or “You can’t eat a table that’s not possible” to the conversation. These individual differences in children’s comprehension and ability to play along with the inherent rules of the play indicate that this form of language play may illuminate a spectrum of children’s cognitive and social development, especially in the preschool age range before most children are understanding more complex riddles and jokes.
Language play in children has been explored from a handful of vantages, but there is a need for clarification of children’s potential differences in ability to produce different types of language play as well as the relationship of children’s language play abilities to the development of other cognitive and social abilities. Different studies, both observational and experimental, have defined language play in different ways, most looking at only one specific type of play (Crystal, 1996; Cumming, 2007; Dowker, 1989; Varga, 2000). And, while some research breaks down the types of language play into broad categories (e.g. word play and sound play) as we have seen, there seem to be more distinct types of verbal play by children in the preschool age range. We believe it is especially important to focus on the categories of hyperbolic play, along with word creation, word switching, and rhyme creation, and the ways in which all four might relate to children’s developing verbal and social skills at large. Also, while sound play is a category of children’s play emerging and continuing during this age range, it is of less interest here because it is the simplest form of language play for children to produce and may be less advanced than the other four requiring little understanding of phonetics, semantics, patterning, or pragmatic rules. We specifically wanted to investigate the four play types that do rely on some linguistic knowledge in children 3–5 years of age because of the wide range of verbal, social, and play-specific abilities that are emerging during the preschool period and evaluate the possible relationships among all of these overlapping and co-developing abilities.
Ultimately, we wanted to examine further the relationship between types of language play and verbal and social development in children 3–5 years old. Within our research design, we implemented a systematic play elicitation task (see Table 1 for examples of play from the children’s responses to our tasks) to create a method of measuring their play ability. We hypothesized that young children who are “better” at playing with language within the parameters of our elicitation tasks in our four categories would also have more advanced verbal and social abilities as assessed by their teachers. While we know of no previous research that has directly examined the relationship between children’s developing ability to produce language play and other measures of children’s development during the preschool years, we felt it was essential to try to measure these possible connections. Children’s verbal play offers a unique way of understanding how children develop the cognitive skills necessary to manipulate and maximize the richness of language in both social and linguistic domains.
Examples of children’s language play for each game type from the current sample.
Methods
Participants
In total, 26 children (12 females, 14 males) aged 3–5 years (Mage = 4.0 years, range: 2.11–5.1) took part in our study. The children were from three different preschool classrooms at a local on-campus preschool serving faculty, staff, students, and alumni of the university. Children were all typically developing, using English as their primary language, and largely from educated, middle-class, ethnically diverse Northern California homes. Every child participated in all four games, except one child who declined to participate in any and was then removed from the sample, resulting in a total of 25 children who completed the entire experiment.
Materials and procedure
Before beginning our observations and experimental games, with the help of the preschool director and classroom teachers, we obtained permission from parents for their child to be observed and tested while in his or her usual preschool classroom. Our experimental procedure began with one of the participating children going with the researcher to a separate play area within the classroom. On the first day of testing, each participating child played two of the four games, one in the morning and one in the afternoon. Children then played the remaining two games the next day, following the same pattern as before. Researchers left time between morning and afternoon games intentionally so that children would not perseverate the rules from one game to the next. Games were played with one participating child and a researcher, each using a puppet as an interlocutor (friendly but unfamiliar “muppets” meant to be reasonably perceived as “silly” by children). Games generally began with the researcher introducing their puppet as a silly partner, who wants to play some silly games with the child and the child’s puppet.
Games consisted of (1) the rhyme game, in which the researcher had one of the puppets and a picture book full of familiar objects for the child to look at during the game. The researcher instructed the child, “This is my silly friend Wembley who wants to play a little game with you. Wembley, what words does this picture make you think of?” Using the picture book, the researcher’s puppet then rhymed three different words with the target picture word. The researcher then laughed and said, “Wow, good job! (turning the page to the next picture) Now it’s your turn! What words does this make your silly puppet think of?” The child was then prompted to participate in the game (e.g. to think of as many words that rhyme with the picture), and after their turn the researcher’s puppet repeated the same game with the next picture, repeating this process for three pictures each (child and researcher). After the game was over, the researcher thanked the children, presented them with a small sticker, and asked whether they had fun and would want to play another silly game later in the day. (2) The word switching game involved very similar procedures. The researcher and the child used the silly puppets for the game and a flip book again. The researcher began the game by reiterating, “These are some silly puppets. They like to say some silly things sometimes. When my puppet sees a cat, like this one, he goes, ‘woof-woof!’ Isn’t that silly? What does your puppet say when he sees a cow?” The game continued on like this, with the researcher and child alternating different pictures of familiar animals from the flip book with different, incorrect sounds for three pictures each, ending the game in the same manner as described above. (3) The word creation game followed the same procedure as the word switching game except the researcher instructed the child that “When Wembley sees this he calls it a ‘blick’! Isn’t that silly? What does your silly puppet say when he sees this?” Pictures of novel items (also used in Frank et al., 2015; see Figure 1 for example image) were used with this game so that the child would not already have a label for the picture and would be more readily able to invent their own “silly word.” As before, the researcher and child alternated turns for three rounds each. (4) The final language play game that the children played was the hyperbolic game in which the researcher began the game in the same way but without the picture book and made a statement that “Wembley is so hungry! He’s so hungry that he could eat a fork, actually he’s so hungry he could eat a whole plate! How hungry is your silly puppet?” The researcher and child alternated saying how hungry their puppets were for three turns each. At the end of the final game, the children were given a sticker for their participation and told that they did a great job playing these silly games with the puppets. Each of the four play games was short, lasting between 2 and 5 minutes for each child, thus keeping children engaged and excited to continue playing.

Example stimuli for word creation play game.
Each child participated in the four games in the same order (i.e. the games were not counterbalanced). This design choice was made intentionally because past research on children’s verbal play has indicated that rhyme is very prevalent and easily understood by children in this age range (Crystal, 1996; Holmes, 2011). Thus, the rhyme game was placed first and was intended to function as an easier entry point into the experimental procedure for children. Word switching and word creation being medium–hard games were deliberately given next and hyperbolic play, which we hypothesized would be the most difficult game for children because of the multi-layer rules implicit in the game, was placed last for all children to maximize their “warm-up” to the procedure and the experimenter. While this choice meant that we could not later rule out order effects when comparing children’s play abilities on individual game types, it did ensure that comparing individual children with one another within a game type was not extraneously influenced by how comfortable they were or how much practice they had had playing these games—a tradeoff that was important to us given the small sample available.
In addition to experimental games, researchers casually observed children over the course of 2 to 3 weeks and took notes of spontaneous language play by the children participating in the study that were witnessed during this time frame. An audio recorder was placed on the children’s snack table to record their exchanges for the purpose of later replay and confirmation of note taking. Observation notes and transcriptions of spontaneous play were used for the purpose of confirming these types of play had a place in this age group, and in order to have some natural examples of children’s different play types and strategies from our own sample.
For each child that participated, researchers also requested that their primary classroom teacher fill out a short 20-question survey on the child’s verbal skills (10 questions) and peer interaction skills (10 questions) which is described below. Teachers were given the survey for each individual child and took about a week to fill out and return the form.
Measures
Play game coding
Children’s answers in all four games were scored on a 0- to 3-point scale, with 0 indicating no coherent response, 1 indicating a logical/non-silly response (such as “the dog says ‘woof woof’” in the word switching game), 2 indicating some understanding of the game’s rules but not full mastery (such as “my puppet’s so hungry it could eat a whole apple” in the hyperbolic game), and 3 indicating a successful answer that follows the pattern and rules implicitly set by researchers. Two researchers independently coded an overlapping 25 percent of the participants’ responses from transcriptions made of each child’s sessions in order to assess interrater reliability. Reliability between coders was greater than .95.
Teacher surveys
In order to assess teachers’ perceptions of children’s social skills, we used 10 play interaction subscale items from the teacher version of the Penn Interactive Peer Play Scale (Fantuzzo et al., 1995, 1998), which asks teachers to rate the frequency of positive peer-focused behaviors on a 4-point Likert scale (see Appendix 1 for questionnaire). And, in order to assess teachers’ perceptions of children’s verbal skills, we developed our own parallel verbal skills measure of children with ten 4-point Likert scale questions assessing how often children demonstrated comprehension and use of difficult vocabulary, figurative language, and instances of language play in the classroom (also see Appendix 1 for verbal skills questionnaire). Classroom teachers were asked to make these evaluations because having the same teachers, who were familiar with all the children, assess each child provided consistency in ratings in the social context of the classroom.
Results
Play types
Children were scored for each of the four games that they played and those scores were tallied within each of the four different play types yielding a possible score between 0 and 12 for each type of play game. Children’s scores on all four play types were compared with a repeated measures analysis of variance (ANOVA), which uncovered a main effect of play type, F(3, 22) = 11.79, p < .001, with a moderate effect size, eta squared = .62. While in Figure 2 the graph seems to indicate a stepwise increase in scores as children participated in the four different play types, planned pairwise comparisons revealed that this main effect was actually driven by children’s lower scores on the rhyme game alone compared to the other three. Mean scores for word switching, word creation, and hyperbolic play did not differ significantly from one another (all ps > .15). However, the rhyme game scores (M = 4.46, standard deviation (SD) = 2.54) were significantly different from the word switching game (M = 6.54, SD = 3.34), t(24) = –3.19, p < .01; the word creation game (M = 6.46, SD = 2.67), t(24) = –4.37, p < .001; and the hyperbolic game (M = 7.46, SD = 3.48), t(24) = –4.81, p < .001. These results indicate that the rhyme game may have been the most challenging type of play for the children; however, because all children participated in the rhyme game first, we cannot rule out the possibility that the novelty of the testing situation was also having an impact on children’s initial play scores.

Children’s mean scores for each play game type.
Play score correlations with other measures
Table 2 provides the bivariate intercorrelations among the four play types. As can be seen from Table 2, all six possible pairwise correlations were positive and greater than r = .30, and while four are significant at the .01 or .05 level, the other two are marginally significant. From this we concluded that in general, children who were adept at any one play type were more likely to be successful at all four, and thus, children’s scores for each individual type of play game were compiled into a sum total play score that could have ranged from 0 to 48 (M = 25.00, SD = 9.16). We also collected measures based on teachers’ responses to the 10-question-each peer play interaction and verbal behaviors questionnaires which could range from 0 to 40 of (1) their peer interaction skills (M = 15.88, SD = 4.22) and (2) their verbal ability (M = 14.52, SD = 3.79). And, for each child we also recorded (3) their age at testing (M = 4.02 years, SD = 0.64). In order to analyze the relationships each of these three latter variables has with children’s play abilities, we entered them as predictors of total play score in a multiple regression. Table 3 provides the bivariate correlations among each of these variables and indicates initially that children’s total play scores were related to both their age and their verbal ability scores and that verbal ability and age were related as well. However, peer interaction scores did not correlate directly with any of the other three variables. When age, verbal ability, and peer interaction were all entered into the multiple regression analysis, the model was significant, F(2, 21) = 7.29, p < .01, with all three factors combining to account for 51 percent of the variance in children’s total play scores. However, when holding the other variables constant, only children’s verbal ability scores were a significant predictor of total play scores, p < .05, with the largest Beta value of the three at .59. This indicated that although children’s age was correlated with their play scores, it was their verbal ability that is the best predictor of their language play abilities.
Bivariate correlations among children’s individual play type scores.
Correlation is significant at the 0.05 level (two-tailed).
Correlation is significant at the 0.01 level (two-tailed).
Bivariate correlations among children’s total play scores and external variables.
Correlation is significant at the 0.01 level (two-tailed).
Discussion
This study is the first to measure preschool children’s play abilities and analyze the relationship between multiple language play types and children’s overall verbal and social skills. Our results indicate that while age and verbal ability were highly correlated with play abilities, as might be expected, verbal ability was the strongest predictor of children’s play scores. Our results also indicated that while we hypothesized that children with higher peer interaction scores would also be more competent at verbal play, our findings did not indicate a strong relationship between the two variables. In terms of measuring children’s language development, our findings lend support for the relationship between children’s development of verbal skills and their quantifiable ability to produce language play in a game-like setting.
While our findings are an important step toward understanding the progression and relationship of language play to children’s broader language development during these preschool years, there are still questions left that warrant further research. First, we intentionally did not counterbalance the four different play types for children because we believed rhyme would serve as a warm-up game for children and then progress easiest to hardest based on previous research findings (e.g. Crystal, 1996; Dowker, 1989; Esposito, 1980). This lack of counterbalancing may have been a confound in our results. While we found that rhyme creation was the most difficult play type for children, the order effect makes it impossible to determine whether the rhyme game was actually more difficult or whether children simply performed better on later play types because they had experienced some learning or familiarity effects. We also found that there were no significant differences between word switching, word creation, and hyperbolic play types for children. Our results indicated that children who had higher verbal scores and were more successful at the play games were equally good at all three types of play (excluding rhyme). This result was counter to what we originally hypothesized in that we believed the hyperbolic game would be the most challenging. However, this leaves the question of whether or not all three of these play types and potentially other play types all assess the same, broader verbal play ability. Further research should examine the interaction between these four types and others to evaluate children’s abilities across different language play games and to potentially develop new methods of assessing verbal ability in young children.
Ultimately, our findings indicate that while language play may seem peripheral to cognitive development, engaged in simply to make other children or adults laugh, actually understanding how to play with language in common ways appears to relate strongly to children’s linguistic development, suggesting that language play may be more a verbal than social skill. Our results indicate that these short, light-hearted language play games with preschoolers were able to tap into their verbal abilities more than to their social abilities. However, this kind of elicited language play may differ slightly from children’s spontaneous play and further research should evaluate more closely children’s spontaneous, naturalistic language play in the preschool classroom and how that compares to their verbal and social skills as well. When considering children’s abilities with language play, the question of children’s humorous intent and the reactions that they get from peers could still influence the amount of and success of children’s play—encouraging more or less of it, or more or less linguistic complexity. And, since our elicitation games were played in one-on-one situations without any measurement of humorous effect, we hesitate to completely rule out the social skill indicator that verbal play might still be within a more social peer group context.
Our study created four subtypes of play that were all set up to be understood as humorous to the children as indicated by the silly puppets and the script used by researchers (e.g. Crystal, 1996; Dowker, 1989; Howard, 2009). However, we did not assess children’s comprehension of humor or give any differential feedback on how “funny” their various playful responses were within the games. Perhaps children with better comprehension of humor or children who are “funnier” would also be better at language play because there is an integral humorous intent behind their responses. For example, in the word creation game, some children gave answers that were completely original and humorous, very similar to the example stimuli (e.g. “this is a gigi!”), as opposed to other children who simply called the object something it looked like (e.g. “this is an airplane.”). This could potentially illustrate an instance in which children who comprehend humor to a greater extent provide “better” language play answers. Previous research has found that children in this age range can identify humorous versus sincere intentions by the social cues given by their play partners (Hoicka and Akhtar, 2011, 2012), and so future research should examine more closely the relationship between children’s productive language play skill and their comprehension of humor. Children’s ability to comprehend humorous cues as well as produce more humorous play could potentially be more strongly related to their peer interaction and social development than we have found here.
Furthermore, there is a question of the power dynamics at work in language play interactions, specifically between children and older peers or adults. Children clearly use play types such as word switching, sound play, and word creation to get a rise out of, and occasionally frustrate, adults. One could interpret this result as an intended consequence of play, and an attempt by young children to gain power at an early age. Children also use hyperbolic forms of play to boast and outdo one another in ways that create power structures without taboo use of force or physical dominance over one another. We experienced some of this anecdotally as children interacted with us before and after games or during snack time. Children took the liberty of calling us silly names (e.g. “Mr. Sillyface” or “Mr. Nononono”) but also in one case joining in together almost competitively to create play with us as one child said to the researcher, “Your head is as hot as a volcano!” and then a second child piped up, “Your head is as fiery as a volcano!” It was clear that when the children were allowed to act silly with us—adult play partners holding less authority than their teachers—they enjoyed the social and subversive nature of the play. Further research may investigate the role of power dynamics in playful encounters, and whether this social aspect relates to play prevalence.
Finally, our sample was relatively small and came from a highly educated preschool for faculty and staff of our university; thus, the children were more likely to have higher verbal skills than other children in the 3–5 age range. While ultimately our sample did have a range of verbal abilities as scored by their teachers and we were able to parse children’s age from their verbal abilities, a more diverse sample may provide even greater understanding of language play as a verbal skill for this age range. Additionally, we did not assess the interaction of children’s potential bilingual home environments on their language play abilities; thus, in future studies, the effect of bilingualism should be examined.
There are still many questions surrounding the differences between multiple types of language play and the relationship of these play types to children’s development. However, our research has indicated that across rhyme, word switching, word creation, and hyperbolic play forms, children with more advanced verbal skills are better at language play games, regardless of their peer play skills. This finding not only provides interesting insight into language play in children 3–5 years old but also aids parents and teachers. Our results support the idea that as children are practicing their verbal skills in conversation, reading, and writing, they can also practice and have fun by playing language games with their peers, their parents, or their teachers. In the meantime, whether or not researchers are tracking them, children will continue to play with their language at every stage of its development—upon finishing the games in our study, a child approached researchers and declared with a smile, “Ms. Andrew don’t go! Mr. Sarah don’t leave!”—an instance of novel and humorous language play all their own.
Footnotes
Appendix 1
Acknowledgements
This research was not financially supported, but we are indebted to the many children and parents who freely volunteered and participated in this study and the partnership of Kids on Campus Preschool at Santa Clara University. We would also like to acknowledge the help and feedback received from the Read Lab.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
