Pretend play has long been understood as an important aspect of children’s development and a cornerstone of pedagogical practice, though playing during mealtimes in early childhood education and care (ECEC) has either been ignored or else explicitly treated as being the antithesis of good table manners (Bae, 2009). To date, we know little about the affordances of food pretend play during playtime (though see Sheldon, 2014), and even less about food play during mealtimes (though see Bae, 2009; Holmes, 2012; Mechling, 2000). In this paper, we examine a collection of pretend play activities during preschool meals, to offer insight into the affordances of food for generating imaginary worlds, and for sharing those worlds with others as part of everyday mealtime interaction.

Imagination, Pretence, and Social Interaction

It is generally held that pretend play allows children to process past experiences as well as try out new roles and scenarios (e.g. Mellou, 1994). Vygotsky (1967, 2004) further proposed that the creative capacity to generate new meanings out of experience and perception forms the bedrock of human development and hence should be promoted in education. Although variously articulated, the preoccupation with children’s imagination has resulted in an extensive body of work on pretend play, also known as make-believe or imaginary play. Numerous empirical studies have emerged from Corsaro’s work on peer cultures (Corsaro, 1979), Garvey and Berndt’s work on the organisation of pretend play (Garvey & Berndt, 1975), or Sawyer’s work on the improvisational nature of play and the use of ‘voicing’ to enact play roles (Sawyer, 1993). A common focus in these studies is on the transformative procedures, which entail adapting an “as if” attitude towards the surrounding world, that allow children to bring about alternative imaginary worlds (Garvey, 1974, 1977; Schwartzman, 1978).

Research in conversation analysis has begun to explicate how such transformations are accomplished—i.e. initiated and structured—within mutually oriented trajectories of social actions. For instance, a distinction can be made between an ‘assertion’ (where one child informs others about the pretence) or a ‘stipulation’ (where children mutually agree on the pretence; Sidnell, 2011). Pretence can be initiated as part of ‘play signalling’ (Bateson, 1976) in which participants communicate a playful stance to others as an invitation to play. Ventriloquising an object (Bateman, 2018) is one such example, and the involvement of others can then be encouraged through establishing a shared play connection and sustained co-participation (Pursi, 2019; Pursi et al., 2018). When participating in shared play, children simultaneously draw on several modalities (c.f. Mondada, 2007): various aspects of voice, gesture, and embodied moves are routinely used to act within the play, as well as to include or exclude other children from taking part in the activity (e.g. Cromdal, 2001; Bateman, 2012, 2015; Butler et al., 2016; Burdelski & Cekaite, 2022; deSouza, 2020). Joining in the play of others can, however, be a delicate business (Corsaro, 1979). It may need to be negotiated through entitlements to access the play world and its asymmetric roles (Kyratzis, 2007), such as playing families (Björk-Willén, 2012; Butler & Weatherall, 2006) or enacting a wedding (Breathnach et al., 2018).

Most of the conversation analytic research on pretend play has studied multiparty interaction in which children are playing with other children, with no adults present. There are a few exceptions that examine adult–child interaction during pretend play (Bateman, 2015; Breathnach et al., 2018; Pursi, 2019; Pursi & Lipponen, 2018; Pursi et al., 2018) such as when a teacher initiates (Bateman, 2015; Pursi, 2022) or becomes involved in (Breathnach et al., 2018) a play sequence. While adult involvement in children’s pretend play can lead to ‘teachable moments’, comforting children (Pursi, 2019), or alerting them to dangers (Bateman, 2015), it can also change the structure and quality of the play (Breathnach et al., 2018). Preschool teachers seeking to share children’s pretend worlds are therefore facing a challenge (Pursi, 2022). When we consider what happens with food play, the nature of this challenge becomes a little clearer; teachers need to balance children’s interests with the institutional nature of shared mealtimes.

Mealtimes: Rule-Following vs. Playfulness

Research into eating practices in ECEC settings has often focused on what has been termed the meal environment, including the rules and structure of the mealtime as well as socialisation into specific norms or behaviours (Willemsen et al., 2023; Dodson et al., 2015; Glaser, 2019). This is captured in the concept of the ‘pedagogical meal’ in Nordic countries (Osowski et al., 2013) and ‘family-style’ dining service (e.g. Sisson et al., 2017). What is becoming clear from research into ECEC mealtimes is that teachers often struggle to find the balance between a structured routine and an enjoyable social atmosphere (Hansen et al., 2016). In colloquial terms, “playing with one’s food” is synonymous with “not eating”, furnishing adults with a ready-made accusation against tardy eaters (see Bae, 2009). Balancing the demands of the institution with those of the child might therefore mean that play has been overlooked in favour of creating a mealtime that is a suitable environment for socialisation.

The apparent dichotomy between playfulness and rule-following during mealtimes is also evident in the scant literature on play during mealtimes, which has been primarily theoretical (Mechling, 2000) or based solely on observations and audio recordings (Bae, 2009; Dotson, et al., 2015; Holmes, 2012). Alcock (2007) showed how children exploit verbal mealtime routines—such as polite requests to pass food—to collaboratively create playful activities, occasionally subverting adult-imposed mealtime orders, while Dotson et al., (2015) discuss children’s resistance to teachers’ systematic strategies for promoting savoury foods over sweet ones.

Studies have also discussed the affordances of the food being offered for boosting children’s agency. Mechling’s work (2000), for instance, considers how children might play with food to exercise power, subverting the rules of the mealtime to ‘take food seriously’, or to deconstruct cultural restrictions on what we are allowed to eat (e.g. we can eat spaghetti ‘worms’ but not real ones). Holmes (2012) describes an example of a child using a banana as if it were a telephone, and then peeled to resemble a slide; or a piece of ice being treated as a ring when placed on a finger. Bae’s (2009) work gives an example of a child pretending that a piece of orange peel is a turtle to initiate a playful interchange with a teacher and which subsequently enabled the child to become more involved in the mealtime interaction. Each of these examples hints at the value of pretend play during mealtimes though we have limited empirical evidence of how such sequences begin and unfold or how they have relevance for the mealtime environment.

Given the pedagogical importance of play and the interactional nuances and affordances that have been evidenced in children’s pretend play in other settings, as well as the tensions between playfulness and rule-following in mealtimes, there is a pressing need to understand what happens when children play with their food during ECEC mealtimes. The aim of this paper is therefore to examine the sequential and multimodal features of food pretend play as these unfold in preschool lunches and to explicate the affordances of this play for constructing and sharing imaginary worlds, as well as for eating.


Data Collection and Participants

This paper draws on a corpus of 97 video-recorded lunches (circa 68 hours of data) from four preschools in two large cities in Sweden. The data were collected between October 2021 and November 2022, with one or two tables being recorded at each preschool every week for 16 weeks, divided over three time periods. Three of the preschools involved were situated in suburban areas and the fourth was in a city centre location. The children participating in the study ranged from 1–6 years old, and each lunch comprised two to seven children of varying ages, and their teacher(s). 22 teachers and 92 children were involved in the study overall. The lunches were situated in various rooms that were used for other preschool activities throughout the day. As is usual for preschool lunches in Sweden, a warm meal was provided free of charge to all children, comprising a meat, fish, or vegetable main item with accompanying starch (e.g. potatoes or rice) and a selection of vegetables or salad. Alternative food options were provided for those children with special dietary requirements or allergies.

Each lunch table was recorded with two cameras positioned on walls, shelves, or windows nearby and a 360° camera placed on the centre of the table to capture as much of the visual space as possible. Two small microphones were positioned at opposite ends of the table and attached to a voice-recorder that was secured underneath the table, for increased audio quality. Ethical approval for the study was provided by the Swedish Ethical Review Authority (diary number 2020-00496) and thereafter informed written consent was given by caregivers on behalf of their children, and by staff members who were present during mealtimes. Ethical concerns were, however, treated as a continual process to ensure that the children and staff felt comfortable with the recording. The researchers often talked about the study with the teachers and children while setting up or dismantling the equipment before and after the mealtime. Children's assent was confirmed before recording took place and the children were often curious about the cameras and, at times, eager to be recorded. Those who were not taking part in the study ate their meals in other rooms, as per their usual routines. No researcher was present during mealtimes. All names, locations, and faces have been anonymised in this paper.

Analytic Procedure

The current paper focuses on instances in which pretend play was initiated by a child during the meals. Following Garvey and Berndt (1975), pretend play is defined as those moments when food is observably used—through talk, gesture, or other multimodal means—as something other than just food, in contrast to other forms of play, such as mixing up potato and sauce to make ‘soup’. The instances of food pretend play analysed here are taken from three out of four of the preschools, and from the first round of data collection, circa 35 mealtimes. Over 20 sequences of pretend play were identified in this section of the corpus, and three examples are shown in this paper. Multimodal interaction analysis (Goodwin, 2000; Mondada, 2007) was used to analyse the data, focusing on the organisation of facial, embodied, vocal, and lexical turns in the mealtime and on the ways in which food was oriented to as a particular form of playful object. Detailed transcripts were produced using DOTE software (McIlvenny et al., 2022) to enable a sequential and multimodal analysis of the play instances; simplified versions of the transcript have been provided here for ease of readability. The transcripts are in the original Swedish and translated into English, translations were done by all authors who are fluent in both languages.


The analysis presents three examples of food pretend play sequences initiated by children and details the interactional methods by which imaginary worlds were displayed and offered for sharing. In each example, the pretence is asserted by a single child through various embodied and vocal elements as play signalling. The analysis provides further evidence for the multimodal aspects of pretend play noted in other settings (e.g. Bateman, 2018; Pursi, 2019) and additionally demonstrates how children establish a play connection with their teachers, how other children may become involved, and how eating and playing can co-exist forming instances of multiactivity (Haddington, et al., 2014) within routine institutional practice.

Imaginary Worlds Co-constructed

The first example illustrates how play was initiated while placing food onto the plate and how an imaginary world, in which peas are prickly and can sting, was jointly constructed between a child and the teacher. There are just two children (Jennifer, 5;9, Fiona, 3;5) and their teacher present at this meal. Jennifer has just made a small puddle of gravy in the centre of her plate and is now carefully scooping up peas from a bowl, building a small pile in the gravy. As she brings the spoon to her plate, a single pea rolls off and she exclaims “wooh!” then, as she slowly tilts the spoon she adds, "I will make it really nice” (l. 1).

Extract 1.
figure afigure a

Prickly peas (MNP1_X_02_A_26m_37s; Jennifer; Fiona; and Teacher)

Jennifer’s meticulous arrangement of the peas on her plate, paired with the verbal announcement that she wants to “make it really nice” solicits the teacher’s engagement with the pretence by asking what she will make (l. 2). The answer, “en taggis”/”a prickling” (a made-up noun constructed out of the adjective “prickly”), causes the teacher to ask what that is (l. 5). Jennifer’s responding enactment of hurting her finger on a pea (ls. 6–8, Fig. 1.1) solicits a token of recognition (“does it sting?”) from the teacher. The existence of imaginary prickly peas on Jennifer’s plate is thus jointly oriented to by both child and teacher.

Remaining in the play world, the teacher recommends eating the peas with caution (l. 13), thereby orienting to the continued progression of the meal. The advice is unpacked further by Jennifer who suggests that they might be spiky in her mouth (l. 16) or even her stomach (l. 22–23) and they both elaborate the safest way to eat them (l. 19, 25). When Jennifer declares she is ready to eat the peas (l. 27), the teacher praises her courage (l. 32), after which Jennifer takes a bite and treats the peas as genuinely spiky (l. 33–34, 36). The interaction thus features eating as a central element to the pretence. Numerous gestures and non-lexical vocalisations—sounds that are not conventional words but which accomplish social functions (Keevallik & Ogden, 2020)—are used in this sequence to rehearse and then perform the eating. The teacher moves between observing and verbally sustaining the pretence world (Pursi, 2022), as well as inviting Fiona to act within this frame (l. 37), while also promoting consumption of the peas. In other words, rather than being merely responsive, the teacher’s actions enable several prospective action trajectories (cf. Wallerstedt, et al., 2021).

The extract thus shows that food pretend play can be constructed collaboratively by the teacher and the children, with the teacher actively engaging and initiating next steps in the play. Such teacher moves—known as “triggering”—have been shown to be prolific in pretend activities in that they “open up for fantasizing, engaging in exploring what is to a large extent unexpected, unpredictable, open” (Pramling, el al., 2019: 174). Our example above shows an additional feature in that teacher’s actions also support the progression of eating. The result is a multiactivity where a shared joyful excursion into an alternative imagined reality is productively intertwined with the practical task of ensuring that everyone eats their lunch.

In the next example, we witness the development of a narrative play sequence around one child’s creative configuration of the food on her plate, with all children at the table actively taking part. As the transcript begins, Emil (4;5) has just commented on how soft the potatoes are, while Maria (5;1) has been silently manipulating her food with her fingers, assembling a collection of peas surrounded by potatoes and breaded fish. When the teacher responds to Emil, Maria announces that her peas are stuck (ls. 3–4; Fig. 2.1). Also present is Emma (5;7).

Extract 2.
figure bfigure b

Potato wall (MNP1_Y_01_A_14m_06s; Maria, Emil, Emma, and Teacher)

The imaginary world is illustrated in Maria’s invitation to treat the peas as agentic objects (cf. Lowe, 1974) that should be able to move, and, indeed, should want to do so (l. 3–4). The scenic intelligibility (Jayyussi, 1992) of the peas as trapped by potatoes and fish is augmented by the prearranged spatial configuration of food on her plate (Fig. 2.2). Agreeing with Maria’s proposal (“nhho way, they are stuck”, l. 5), the teacher aligns with the proffered understanding of the events taking place on Maria’s plate, thereby validating the pretence. Furthermore, in line 9 she offers a creative formulation (“potato wall”) that transforms (Garvey, 1974; Schwartzman, 1978) the food into a building structure. The play connection is now established around the table, and Emil, who has been studying Maria’s plate (Figs. 2.1 & 2.2), proposes that the fish is a door. Though categorially congruent with the teacher’s transformative construction (both walls and doors are building structures), the teacher mishears itFootnote 1 and treats it as irrelevant to the current play theme.

Maria subsequently engages in a dialogue with the trapped peas while Emma—drawing on an object from a children’s television show “Summer Shadow” (Swe: Sommarskuggan)—claims that the peas are stuck in shadow slime. Although this may be seen as a constructive bid to develop the play scenario, her introduction of a different fictional world is not picked up by the others. Instead, the locally constructed world of the peas being trapped by a potato wall is held intact by Maria (ls. 23–26) and the teacher (l. 29).

It is notable that the children do not join in by playing with their own food but rather by connecting with the play world that exists on Maria’s plate. Narratively, the play connections offered by the other children are produced as continuations of Maria’s storyline: both Emil and Emma begin their turn with ‘and’ (l. 11, 12, 19), fitting their turns into transition relevant places (Sacks, et al., 1974). As the locus of action is a single plate, the other participants are visually oriented to Maria’s food, through gaze and upper body posture, and coordinate their narrative contributions with relevant gestures. There are also various multimodal features that enable the food to be brought to life, such as Maria’s non-lexical vocalisations combined with her fork movements as she enacts the peas’ escape (l. 26, 30).

In lines 31–33, the teacher suspends the play activity to deal with an issue of order at the table, which Emma immediately replicates to Emil (l. 34). As Maria complies and corrects her sitting position, she scoops up food with her fork, announcing that the peas are escaping their trap. In response, the teacher seamlessly re-enters the imaginary action, celebrating Maria’s successful rescue operation (l. 36). As some of the food is now on Maria’s fork, the teacher suggests that she eat it to see how it tastes. In this way, the food on the fork functions within two semiotic fields that provide for distinct, but compatible, interpretations: an imaginary one, where the trapped pea has been airlifted from its captivity, and a realistic one, where a forkful of food is halfway between the plate and Maria’s mouth. ‘Saving’ the peas can thus potentially be achieved through eating them, and in this way the teacher exploits the dual contexts to promote the progression of the meal.

In this example, we have seen how all participants at the table align with, and propose substantial contributions to, an imaginary drama that emerges from one of the children having arranged the food on her plate in a certain way. The role of the teacher needs stressing here: moving seamlessly between the imaginary world and the mealtime setting, her actions are instrumental in the maintaining and unfolding of the imagined events and in instructing and correcting the table manners of the children, as well as promoting their eating. Again, the result is an instance of multiactivity where the drama transpiring on Maria’s plate engages all the parties at the table, without interfering with the primary purpose of the mealtime.

A Misfired Invite to an Imaginary World

The examples above have shown how food, if creatively handled, may afford the construction and sharing of imaginary worlds. The play opportunities during preschool meals are, however, not without limit. In our final example, we show how Johan’s (5;0) creative modification of crackerbread fails to elicit a joint play activity. The teacher is eating lunch with seven children, including Clara (5;9) and Hilda (4;8). Most children have finished their main course, when Johan suddenly announces that he has made a pistol out of his crackerbread.

Extract 3.
figure c

Crackerbread pistol (MSL1_X_03)

While the teacher is verbally responding to Clara’s request (l. 1) and waiting for Hilda’s decision (l. 3–4), her eye gaze follows Johan’s hand movements as he picks up his crackerbread. Figures 3.1–3.4 clearly show that the crackerbread becomes a pistol by being repositioned in his hand: Johan holds it so that his fingers are around the ‘grip’ with the barrel pointing forwards. The physical manipulation of the cracker thus becomes part of the transformation process by which Johan orients to the food as a different kind of object. While finalising his grip, he gazes at the teacher and verbally announces ‘I made a pistol!’ (l. 5) and ventriloquizes the object (Bateman, 2018) with a non-lexical vocalisation (p’yoo) of a single shot being fired. Through the coordination of these embodied and vocal elements, the play emerges from spontaneous orientation to the food being an imaginary firearm and is signalled as play with the teacher as the recipient.

The response to the play initiation is, however, rather muted. Although the teacher had previously been looking at Johan, she soon turns her gaze back to Hilda. Johan then pursues the teacher’s response, this time by combining the naming of the imaginary object with an imperative verb form, bidding for her attention (‘a pistol, look’), followed by a vocalisation of a series of four shots in quick succession (line 7). It is halfway through the enactment of shooting that the teacher makes a small, embodied response: an eyebrow raise and a barely perceptible nod (see Fig. 3.5). As she turns back to Hilda, Johan gives up on his gun and takes a bite from his cracker, bringing the imaginary world to an end.

Despite Johan’s relatively elaborate enactment of a shooting, this did not engender the communal playfulness that we observed in the first two examples. From this we might tentatively conclude that there is more to mealtime imaginary play than mere transformation of food. As Garvey (1974, p. 173–174) pointed out: “While objects are freely transformed, for example, stool to milk carton, to conform to the needs of the episode, consistency is maintained in respect to motives and appropriate actions for the roles adopted”. It is also noteworthy that in this example, the teacher is in the midst of an exchange with another child and hence less available to engage in the pretend play; in Extracts 1 and 2 there was potentially more interactional space to allow a play narrative to develop. One way or the other, the roles on offer in Johan’s imaginary event failed to attract the co-participation of the others.


The analysis provides some of the first empirical evidence to show how pretend play with food emerges spontaneously during children’s shared mealtimes. Some common findings were noted: the pretend play is signalled by a single child, preceded by manipulations of the food with hands or utensils before presenting them verbally and visually to the teacher. It is notable that it is a teacher who is the recipient: the children do not actively seek other children’s attention to join in with their play but rather first show their creations to the teacher. In doing so, the children also demonstrate an awareness that such play signalling is allowed, at least in the contexts in which it was observed in the data. The teacher then played an important role through actively engaging in the play and ‘triggering’ next steps, such as the inclusion of other children and the enabling of multiactivity through both playing and eating. In sum, the food offers affordances for generating imaginary worlds, and for sharing those worlds with others during everyday ECEC mealtimes.

The analysis of these sequences supports existing conversation analytic work on pretend play in other settings and provides further evidence for the importance of play in developing children’s inclusion of teachers and other children into their imaginary and social worlds. Multimodal resources were used to signal play stances and invite others into these play worlds. The inclusion of teachers offers a unique opportunity to balance the institutional requirements of socialisation into eating practices while also providing a sociable and relaxed meal environment. As seen in Extract 2, the teacher was also able to seamlessly incorporate a request for Maria to sit appropriately without disturbing the play narrative. The two examples of ‘successful’ play initiation also involved eating as a central part of the play narrative, demonstrating that playing need not be the antithesis of eating. Instead, pretend play could provide affordances for children to actively engage with their eating while also participating in the meal as a social and rule-bound event.

There are some noteworthy aspects of the examples examined here. The play episodes occurred when most children were already eating, so there may not have been concerns about the progression of the meal. The children also had some freedom to use their hands or cutlery to manipulate their food: while not encouraged, this type of behaviour was not reprimanded in those instances seen in our data corpus. This may be specific to the child-centred practices used in Swedish preschools, and further examples from other contexts would enable such comparisons to be made. Furthermore, the study took place in areas where food poverty was not a pressing issue, though food waste is still an important and relevant concern. The analysis here suggests, however, that playing with food is not synonymous with wasting food. The children who initiated the play were all over 5 years old and one of the older children at the table; this may have contributed to their willingness or capacity to engage in play and to interact with the teacher in such a direct manner. The teachers themselves interacted with the children during mealtimes both to assist eating but also as conversational partners, thus it was typical for them to chat together freely while they were eating. Finally, the food served was malleable enough to allow for multiple interpretations, such as squashable potatoes being like the affordances of clay in terms of its playability, and peas that are small enough to serve multiple functions. As Mechling (2000) observed, play items that are open to children’s imagination are much more engaging than those that are formalised.

We recall, in closing, Vygotsky’s (2004; p.88) insistence that educational institutions should dedicate their work to cultivating the creativity of their students by promoting “the development and exercise of the imagination”. While the ECEC mealtime has been identified as a natural venue for playful togetherness (Alcock, 2007), it has also been shown that teachers’ concerns with adult-imposed mealtime routines may offset children’s agency, playful involvement, and creativity (Murud-Riser, 2012; Dotson, et al., 2015; Glaser, 2019). The present study, however, provides empirical evidence that pretend play with food during preschool meals offers abundant affordances for generating imaginary worlds and for sharing these worlds with their teachers and peers during everyday mealtime interaction. Moreover, we have demonstrated how the teachers managed to seamlessly support, contribute to, as well as share the imaginative play worlds around the table, while sustaining an orientation to the children’s eating progression, sitting behaviour, and other routine features of institutional order. These observations carry profound implications for the ECEC profession, suggesting that teachers’ mindful engagement in children’s pretend play is not necessarily at odds with the institutional demands surrounding the preschool meal and may even promote the social atmosphere as well as the children’s development.