Sandwiches and subversion: Teachers’ mealtime strategies and preschoolers’ agency

Data collection

Data emerged through ethnographic observations conducted by the first author (H.M.D.) in two preschools in a city in the southeastern United States. This study was reviewed and approved by a university Institutional Review Board. To initiate the study, H.M.D. emailed the directors of both facilities to enquire about conducting the research at their sites. Both directors met with her in person, at their respective schools. Parents and school associates provided written informed consent to the current study. One family did not permit their child to participate in this study. H.M.D. did not take notes or analyze interactions involving this child. We have referred to all study participants and schools by pseudonyms.

H.M.D. conducted weekly ethnographic observations at both schools in Spring 2010. She arrived at both schools prior to morning snack time and left the schools after lunch. Most visits lasted between 3 and 4 hours. Although we observed snack time and lunchtime at the schools, we have limited our discussion to lunchtime in the interest of space. Throughout the observations, she took notes in a small notebook to assist with the creation of full ethnographic field notes within 48 hours of each observation (Emerson et al., 1995). We terminated data collection once we achieved saturation.

The decision to participate and observe or only observe interactions was dependent on whether children were engaged in structured activities (e.g. “circle time,” lessons, meals) or unstructured activities (e.g. free play). H.M.D. observed children’s actions during structured activities, but did not actively participate. During most structured activities, children were seated either on the floor or at tables. She sat on the floor near the children during structured activities. When children were seated at tables, she sat next to the table rather than at the table due to space constraints. The research team decided that she would not actively participate during meals. Similar to other structured activities, she observed mealtime interactions while sitting near the children, but she did not eat nor did she sit at the tables. This was done to be consistent with H.M.D.’s behavior during other structured activities and to minimize influencing children’s meal routines.

During unstructured activities, H.M.D. would sit near groups of children and would play with children when invited. In an attempt to take on the “least adult role” (Mandell, 1988), during free play, She played with children when invited and actively engaged with them on their own terms. Moreover, she actively rejected engaging in adult-authority throughout the observations. For example, when teachers or children called her “Miss Hilary,” she would ask them to call her simply “Hilary.” When children asked her to intervene during conflicts, she would remind them that she was not a teacher.

Analytic strategy

In line with grounded theory methods, we did not select theories or analytic codes prior to data collection. We should note that our decision to complete this project was informed by our previous interests, knowledge, and research on children, children in schools, and childhood obesity (e.g. Cunningham et al., 2012). Consistent with grounded theory methods, analytic coding began while data were still being collected (Corbin and Strauss, 1990). Our analysis emerged through line-by-line coding of our completed field notes followed by focused coding strategies of dominant themes (Charmaz, 2001).

Setting

Two preschool classrooms for 3-year-olds were observed in two separate schools in one large city in the southeastern United States. Three-year-olds’ classrooms were selected because children without developmental delays should be able to speak in full sentences, convey ideas, use their imagination, replicate others’ behaviors, use utensils, and play with other children (Centers for Disease Control, 2012a, 2012b). The first site, Greenwich School, was part of a nationwide chain of early childhood education centers. The franchise we observed was located in an upper-middle-class suburb. The classroom we observed had a maximum of 23 children present, but most days, 20 children were present. Typically, two teachers were present except during lunchtime, when only one teacher was present. The second site, Lakefront School, was an independent preschool affiliated with a college. Lakefront School was located on the periphery of the college campus. Residential dwellings for college students, lower-income families, and middle-income families surround the campus, along with fast food restaurants and grocery stores. The classroom we observed had a maximum of 15 students, but most days, 12 children were present. Two teachers were present at all times. At both schools, meals were supervised by children’s regular classroom teacher(s). Children at each school were permitted 15–20 minutes for lunch and then began preparing for naps.

Both classrooms observed were racially and ethnically diverse, similar to the city in which both schools were located. At Greenwich School, 11 children were White, 5 children were African American, 2 were Hispanic, 3 were East Asian, and 2 were South Asian. One Latina and one White woman taught the class. At Lakefront School, nine children were White, two were African American, two were East Asian, one was Hispanic, and one was South Asian; the teachers were two African American women.

Greenwich School employed a cook, and all lunches and snacks were provided for children. Children ate lunch in the classroom. Lunch usually consisted of a meat and carbohydrate entrée (e.g. lasagna with ground beef or chicken with yellow rice), warmed frozen or canned vegetables, canned fruit, and a cup of milk. All of the children were served the same lunch, unless they had documented restrictions or allergies. Children with food restrictions (e.g. vegetarian, lactose-intolerant) were offered foods similar to their peers, but their meal excluded the prohibited foods (e.g. vegetable lasagna, yellow rice without chicken, soymilk). Plates were prepared and placed at the tables before children were seated. Children were not permitted to talk to one another during meals at Greenwich School. The teacher supervising lunch sat at a separate table and ate food she brought from home; the other teacher ate lunch in the teachers’ lounge.

Lakefront School provided snacks but children brought their own lunches. There were no explicit restrictions or requirements on the types of food children could bring. There were institutional constraints, however. For example, the children did not have access to a refrigerator or microwave. Before lunch began, children retrieved their lunch bags from their cubbies. Children at Lakefront School ate their lunch outside on picnic tables, which were covered by a large awning. Teachers sat with the children at the same tables. Teachers opened containers of savory foods for children at the beginning of the meal. Children spoke to one another and teachers throughout meals.

“Only If You Eat Your Sandwich First”: Subverting gatekeeping

At Lakefront School, where students brought their own packed meals from home, teachers exercised authority by acting as gatekeepers of specific foods at every meal by refusing to open containers or packages that children were unable to open themselves. Restricted foods typically included juice, cookies, sweetened yogurts, and candies. Gatekeeping occurred even though sweets and juices were permitted and many children brought at least one of these items to school. At the beginning of the meal, when teachers would walk among the picnic tables to help children open up their foods’ containers, they would not open packages of sweet foods. Teachers would only open packages of savory foods, including items like sandwiches, vegetables, fruit, and crackers while failing to verbally acknowledge sweet foods. Once teachers opened up one child’s savory foods, they would move on to the next child.

When teachers would not open packages of sweet foods and juices, children tried to subvert teacher’s gatekeeping by attempting to open the containers themselves. Sometimes, children would also ask the teacher to open their sweet foods, at which point the teachers would use directives to instruct children to eat their savory foods first. Oftentimes, children would still attempt to access the restricted food. In order to successfully access restricted foods, children had to be dexterous enough to open the containers themselves, which, for the 3-year-olds we observed, was not common. Children appeared to have the most difficulty opening up manufactured packages and plastic food storage containers. In the rare cases when children were able to open their sweet foods’ packages on their own, the teachers would not take them away. Instead, teachers responded by using directives—which we discuss in the next section.

Teachers would open children’s packages of sweets and juices when two conditions were met. First, children had to eat most or all of their savory foods. Second, children also had to request that these foods be opened. When (and if) children finished their savory foods, teachers would open packages of sweets and juices if children asked. Importantly, the teachers at Lakefront School did not express to the children why they were not permitting them to have their sweeter foods, nor did they verbalize moral judgments on the healthfulness of foods, unlike findings from a previous study on Danish kindergarteners’ meals (Karrebæk, 2013). Even though judgments on the healthfulness of foods were not spoken outright, we believe judgments of healthfulness were salient in shaping which foods were restricted.

In one case, Brook asked Ms Maureen if she would open her package of cookies. The cookies were sealed in a manufactured package. Ms Maureen responded by saying that she would not open the cookies for her because Brook had not eaten her sandwich. Ms Maureen explained that “if Brook did eat her sandwich, she would open the cookies” [excerpt from field notes]. After this interaction, Brook did not eat her sandwich and Ms Maureen did not open the package of cookies. Ms Maureen acted as a gatekeeper by using the cookies as a reward for finishing the sandwich. In other similar scenarios, after children ate their savory foods, the teacher would open containers of sweets or juices. Both before asking and after Ms Maureen instructed her to eat her sandwich first, Brook unsuccessfully tried to open the package. At the conclusion of the meal, the unopened package of cookies was returned to her lunch bag.

“Eat Your Peaches”: Mealtime directives and children’s agency

One of the most common strategies teachers at both schools used to shape children’s consumption was the use of directives, defined as verbal commands and pressure to eat. The purpose of directives was to encourage consumption. At Greenwich School, teachers commonly told the entire class they “needed” or “had” to eat their fruits or vegetables. For example, halfway through the meal one day, Ms Barnes noticed that few children had not eaten any of their peas and carrots. She announced to the class that the children “need to eat [their] peas and carrots.” The teachers at Lakefront School also used directives, but they tended to frame the directives using softer language (e.g. “should” instead of “have” or “need”).

In addition to the aforementioned directives, teachers at both schools sometimes avoided using auxiliary verbs (e.g. “should,” “need,” and “have”) in their directives. Indeed, directives such as these were articulated as formal commands. On one afternoon, Esther, a child at Lakefront School, was given a small salad with a bread roll for lunch. Esther pulled very small pieces from the roll and began to eat them. Ms Wanda sat across from Esther at the table and said to her, “take a big big bite,” in reference to the roll. In this directive, and others following the same pattern, auxiliary verbs were not included. We considered statements such as these to be directives, in line with both Lareau’s (2003) conceptualization and the nutrition literature on the use of directives (e.g. Hughes et al., 2007; Murashima et al., 2012).

Children at both schools engaged in various agentic subversive tactics to avoid conformity with teachers’ directives. Children’s attempts at subversion were communicated either verbally or silently. Subversive verbal and silent responses conveyed either feigned assent or dissent. Feigned assents were indications to teachers that children would comply with the directive, but children did not eventually consume the food. Dissents included either verbal protests against the directive or silent responses where children would not acknowledge the teacher or the food they were directed toward. The use of verbal or silent techniques with either feigned assent or dissent varied between the schools. Children at both schools used each of the aforementioned techniques, but children at Greenwich School usually used silence to convey feigned assent or dissent, while children at Lakefront School regularly used both silence and verbal communication in their subversion.

Silent subversion, although not unique to the children at Greenwich School, was most commonly deployed there. One could interpret silence in Greenwich School as children obeying the rules against talking during meals. However, because children at Greenwich School would break the rules on silence occasionally during meals to communicate with other children, it is unlikely that children were particularly fearful of speaking during the meal. It is also feasible that children did not understand the directives, but their use of nonverbal cues and other behavior throughout the day suggest that the children we observed were competent and understood the directives. Children’s nonverbal communication indicated that they heard, understood, and chose to oppose the directives. Silent responses to directives also occurred in Lakefront School, but not as frequently as in Greenwich School. Our observations suggest that children’s silence was deliberate and defiant. Prior research has also identified children’s silence as defiance (Markström and Halldén, 2009; Silverman et al., 1998).

One indication that silence was used as subversion in Greenwich School, rather than demonstrating ignorance or conformity with mealtime regulations, is rooted in the timing and nature of rule breaking. Children at Greenwich School broke the rule on silence during meals to communicate with one another when it suited them. Children would attempt to speak to one another during the meal about various food-related and nonfood-related topics, but they were quickly hushed, even when they were speaking very quietly. In one example, Alexia broke the silence rule to chastise Ethan for failing to eat his food in ways deemed proper. Ethan had repeatedly picked up his fork carrying peas, brought the fork up to his lips, and then dropped the peas back on his plate. Alexia, upset with Ethan’s behavior, told him that he should “eat the right way!” Ms Barnes made a “hush” sound. Alexia proceeded to demonstrate “the right way” by placing her fork with peas in her mouth, chewing the peas, and then swallowing them. Children’s pushback against the rule on silence when children wanted to communicate with one another was deliberate and agentic.

A second indication that silence was used as subversion is that silent responses to directives also occurred at Lakefront School, even though there were no rules prohibiting conversations. When Ms Wanda told Esther to “take a big big bite” of her bread roll, as described earlier, Esther continued to pick at the bread and only eat portions of the roll that she had pulled off it. Esther did not verbally respond to Ms Wanda, nor did she put the entire roll up to her mouth. She continued to eat the bread roll as she saw fit. Although Esther was eating, she was not eating her bread in the manner deemed appropriate by Ms Wanda.

Children’s body language during silent responses indicated their intended behavior and the image they sought to present to teachers. Sometimes children would simply remain silent without using body language, while other times they would use facial expressions, head nods, or head shakes to indicate whether they would eat their food, for example, an exaggerated frown when directed to consume broccoli. Children engaged in silent feigned assent when they would nonverbally signal they would comply with a directive, but then did not eat the food they were directed toward. Two common tactics for indicating silent feigned assent included nodding their heads and touching their food after directives. However, the children would not actually eat the food after these cues.

There were other instances when teachers used directives and the child would not visibly respond nor acquiesce. We refer to this type of response as silent dissent. Children in these scenarios would remain quiet and mostly still. They would not touch the food. If teachers told an individual child she or he, specifically, “needed” or “should” eat her or his food, and the child utilized silent dissent, sometimes the teacher would continue deploying directives to the child for a brief period. Eventually, the teacher would give up. When teachers use directives, they are seeking a specific response: for children to eat. This does not necessitate speaking. Children exhibited defiance by refusing to eat and through appearing to not acknowledge the directives.

Children at both schools used verbal responses to convey feigned assents and dissents to directives, although these were much more common at Lakefront School. Verbal feigned assents occurred when children would tell teachers that they intended to eat the food they were directed toward. Usually, these were one or two word statements, such as “okay” or “I will” after a directive. Importantly though, children did not proceed to eat the food.

Occasionally, children would verbally dissent to the directives. An example of this includes a child telling a teacher she or he “did not want” to eat something, or “did not like” it. Verbal dissent happened quite rarely, as children understood that when they protested, the teacher would continue to direct them to eat their food.

Hyperbolic justifications: Attempts at reasoning and children’s responses

Teachers justified directives with hyperboles. Hyperbolic justifications were used to encourage consumption of fruits and vegetables. We observed two types of hyperbolic justifications: omens and embellishments. Omens were exaggerated consequences of failure to consume a specific food, while embellishments were exaggerated health benefits of consuming a particular food. Omens were a form of pressuring teachers used to attempt to convince children to consume specific foods. Embellishments, in contrast, were a form of reasoning to justify why someone should consume specific foods. Omens were mainly found at Greenwich School and embellishments were primarily associated with Lakefront School.

The most common omen observed were statements on the consequences of failing to consume one’s vegetables and vision decline. Teachers at Greenwich School told the children that they need to eat their vegetables, and if they did not, they would experience vision problems later in life. Teachers would often remark, “if you don’t eat your vegetables, you’ll have to wear glasses.” None of the children in the class wore glasses at the time, but Ms Lopez and H.M.D. did. Ms Barnes did not wear glasses, but she told the children she wore contact lenses. Ms Lopez explained that her own failure to eat vegetables when she was a child was the reason she had to wear glasses. Occasionally, teachers followed omens with questions like, “you don’t want to wear glasses, do you?” In most circumstances, children did not appear to respond to omens at all, thus engaging in silent dissent. Some children responded to omens by engaging in silent feigned assent, generally by touching their food with their forks. The children seemed to believe that if the teachers saw that they were “eating” their vegetables, they would not experience an ominous fate.

Embellishments were used to convey that consumption of fruits and vegetables would cause healthful physical growth. In one scenario, teachers at Lakefront School used embellishments to convey exaggerated benefits of consuming apples. At the beginning of the meal, Jessie opened her lunch bag and saw a cup of applesauce. Several children began to talk about how much they enjoyed apples. Ms Wanda praised their responses by telling the children that she was glad they liked apples because “apples make you big and strong.” The children became excited and continued to converse about how much they liked apples and how they too would become “big and strong.” The use of embellishments excited children, who became visibly happy and gregarious. Unlike children’s responses to omens, when embellishments were used, children would start their own conversations on food preferences and link the embellished food to the exaggerated health benefit. In addition to children responding favorably to embellishments during mealtimes, children sometimes used embellishments when playing at the kitchen station or when making play-food with clay or blocks.

Dialogues of reasoning and children’s mealtimes

There were some differences in the strategies employed by teachers and the children’s subversive tactics. These differences often occurred between the schools themselves, resulting from conflicting school and mealtime structures. At Greenwich School, all of the children were given the same meal and they were not permitted to talk to one another during the meal. At Lakefront School, children brought their lunches from home. The first strategy, gatekeeping, was predicated on children bringing foods from home. Teachers at Lakefront School acted as gatekeepers by restricting children’s access to desirable foods by refusing to open containers of desirable foods. Even more interesting is that teachers did not take restricted foods away from them. Children tried to avoid gatekeeping by trying to open restricted foods, although they were seldom successful.

Directives, the second strategy we identified, occurred at both schools; however, there were some key differences in the ways teachers framed directives in each school. In particular, the teachers at Greenwich School often used the auxiliary verbs “need” and “have” in their directives. The teachers at Lakefront School tended to use “should” in their directives, or omitted auxiliary verbs altogether. Children used agency in subverting directives through silent and verbal indications of feigned assent and dissent. The children at Greenwich School tended to use silence in their demonstrations of agency, while the children at Lakefront School utilized silent and verbal agentic responses.

The third mealtime strategy discussed was hyperbolic justifications, used to attempt to persuade children to consume healthful foods. Both schools utilized these, but at Greenwich School, hyperbolic justifications were framed as omens, while embellishments were used at Lakefront School. The underlying message of omens is that something negative would happen to the child if they failed to consume healthful foods, while embellishments exaggerated the health benefits of fruits and vegetables. Children used agency against hyperbolic justifications in a similar manner to their resistance of directives.

With the exception of hyperbolic justifications, teachers seldom reasoned with children during meals. On few occasions, teachers attached brief statements of reason to directives, such as by telling children a specific food was “yummy,” but these were quite rare. Childhood nutrition research has found that reasoning and probing questions are effective strategies to promote consumption of healthful foods (e.g. Hughes et al., 2007). With this in mind, it is particularly interesting how infrequently teachers reasoned with children in the schools observed. One potential explanation as to why teachers refrained from reasoning with children during meals is that both schools allocated children only 15–20 minutes to eat. Creating a substantive dialogue about food takes considerable time. With only a few minutes, meaningful dialogues about food and nutrition are unlikely to occur.

The strategies teachers in our study used, including gatekeeping and using directives, have been found ineffective at increasing children’s fruit and vegetable consumption. For example, gatekeeping of desirable foods is associated with greater desire for these foods (Fisher and Birch, 1999). Moreover, requiring children to consume an undesirable food in order to receive a reward is associated with further dislike for the foods adults are trying to convince children to eat (Birch et al., 1982). Pressuring children to consume is also associated with less consumption (Galloway et al., 2006) and further dislike of the pressured foods (Batsell et al., 2002; Birch et al., 1982). Although we know little about the direct effects of omens and embellishments on children’s consumption, we would expect omens to be similarly counterproductive in increasing children’s healthful food consumption, as omens are tantamount to pressuring. In contrast, we suspect that embellishments might have beneficial effects on children’s consumption because they are a form of encouragement.

Limitations

We note some limitations and challenges of this research. Our study involved only two schools and we do not presume that our findings are representative of meals in preschools more broadly. However, this study is a first step to deepen our understanding of adult–child interactions about food during mealtimes. Expanding the study to more schools and in different regions of the United States and other countries might uncover other types of mealtime strategies and subversion tactics. Second, as we were unable to record at the sites, it is plausible that some interactions were not included in our notes and subsequently in this manuscript. We were careful to take detailed notes of verbal and nonverbal interactions throughout the study as the interactions were occurring, although we acknowledge that some of the more subtle verbal and nonverbal interactions may not have been included in our discussion. Some children’s voices, especially those who were shier or less verbal, may have been inadvertently left out of the discussion, as a result. Our efforts to observe nonverbal interactions sought to address problems with relying entirely on verbal communication to understand children’s interactions with adults and one another. Finally, as we did not conduct formal interviews with the teachers, we cannot speak about how teachers come to understand mealtimes within their schools.

Future research

Future research should consider observing preschoolers mealtimes in other regions of the United States and in other countries. The nutritional consequences or benefits of using omens and embellishments on children’s consumption is another avenue nutrition scholars ought to pursue as, from our observations, teachers utilize these to attempt to increase consumption. Researchers interested in childcare practice should investigate how practitioners and schools come to understand school meals and their own place in shaping children’s consumption. More social science research on peer cultures during meals is another important and understudied area. Similarly, the co-construction of meals should be highlighted in future research on food, consumption, and meals in schools.