Opportunities knock: Mediation of peer-relations during meal-time in toddler groups

According to socio-cultural perspectives, adults are seen as mediators of cultural believes, values and practices. Qualitative analyses of teachers’ mediation of peer relations based on video-recordings in 9 toddlergroups indicate that meal-time represents opportunities for teachers to facilitate togetherness and peer-relations between toddlers. The teachers might facilitate sharing, passing food, routinized practices such as singing, and conversations. The results indicate variations between childand group-oriented approaches, and accomplishing meals in an effective way. The childand group-oriented approaches are marked by encouragement of toddlers’ initiatives and teachers supporting peer-interactions. Raising teachers’ consciousness about their contributions to children’s development of peer-relations and togetherness in group settings might contribute to enhance reflected practices in working with toddlers in groups.

Today we have considerable knowledge about toddlers' social competencies in interactions with their caretakers as well as with their peers. But both in literature and research these relationships are often treated as if they belonged to different worlds.
A tendency in research concerning peer relations is to exclude caretakers from research focus and not discuss their role in development of relationships between children (cf. File, 1994;Hay, Payne, & Chadwick, 2004;Howes & Tonyan, 1999;Kontos, 1999;Nash & Hay, 2003;Os, 2007;Williams, Mastergeorge, & Ontai, 2010). But as Marianne Gullestad (1994) claims: There is not, and should not be, a Chi-nese wall between the world of peers and world of adults. According to Corsaro and Johannesen (2007) peer cultures are created, shared and developed through everyday activities where children and caretakers take part. Seeing peer relations and relations between children and caretakers as belonging to divided worlds might have unwanted consequences like exclusion from the group, lack of understanding and respect between peers followed by feelings of loneliness and lost opportunities for learning and common joy.
For the very young, for toddlers, establishing, developing and maintaining relations to peers might be challenging and they might need some support from their caretakers (cf. Howes & Phillipsen, 1998;Pramling Samuelsson & Fleer, 2010;Schaffer, 1984, p. 127). And opportunities knock: Everyday life in day care contains opportunities for working with toddlers' peerrelations.
But it might be challenging for caretakers because knowledge about how to work with peerrelations is restricted (Girolametto & Weitzman, 2007;Girolametto, Weitzman, & Greenberg, 2004). In order to illuminate some aspects connected to caretakers' role in peer relations in toddler groups during mealtime, I will present some tendencies in the results from my ongoing PhD-project 1 .

METHOD
The research questions in my PhD-project concerns how caretakers mediate peer relations during mealtime and in free play situations, and how quality of the mediation seems to affect immediate interactions between the toddlers. Based on video-data from toddler-groups, qualitative analyses of teachers' mediation of peer relations are accomplished. Are teachers initiating and supporting activities that include groups of children? Do they make efforts to help all children to be active participants in the group and active agents in their interactions with peers? The project is an observational study in nine toddler-groups in nine different day care centres 2 . There are 82 children from 1 to 3;6 year, and 30 caretakers participating. The analyses are based on 32 hours of video-recordings. All sequences during meals and free play containing mediation are transcribed (about 500).
One child in each group is chosen to guide us through everyday life in the day care centre. Letting children guide us into their world, we aim for approaching children's perspective. But focus in the analyses is not on individual children. The guiding child is leading our attention and the research interest is connected to whatever is going on in the group where the guiding child participates.

MEDIATION -A CULTURAL PRACTICE
The project is based on social-cultural perspectives within a social-cultural frame of reference (Rosa & Valsiner, 2007;Valsiner, 2000). According to socio-cultural perspectives, adults are seen as mediators of culture (Hasan, 2002;Rosa & Valsiner, 2007) at the same time as toddlers are active in constructing their relationships with peers. Peer-relations and togetherness are parts of cultural processes. The way everyday life is organized with child-rearing practices and direct interactions with children, will mediate cultural values and practices. Since culture is not homogeneous, different cultural values and practices might be parts of teachers' practices.
I have chosen the concept 'mediation', originally from Vygotsky's work, to describe caretakers' involvement in peer relations, instead of other concepts often used in research, like support, facilitate, scaffold etc. By using the concept of mediation, I want to underline that children's development and learning are social and hence cannot be divided from the cultural contexts children are embedded in. Caretakers communicate or mediate cultural knowledge, skills, and attitudes which children actively appropriate, and among these are how to relate to peers. The mediating styles of caretakers do not mirror the caretakers' personality, competence or personal believes, but must be seen as a part of a common cultural ideology and practice. Because mediation is a less value loaded concept than support, scaffold and facilitate, the choice of mediation as a core concept, also implies that mediation might, independent of caretakers' intentions, not only strengthen peer relations between toddlers, it might as well weaken them. (1999) claims that togetherness is a manifestation of day care life. I would like to add that feeling of togetherness and belonging to the group, are important aims for pedagogical work in day care settings. Just being physically together does not imply togetherness. Ac-1. The PhD-project is a part of a larger research project "Norwegian day-care centers as a link in the chain of care for children under the age of three", a cooperation between Høgskolen i Oslo (HiO) og Norsk institutt for forskning om oppvekst, velferd og aldring (NOVA). The project was supported by Norges Forskningsråd. For further reading, see the following publications: Eide & Winger, 2008;J.-E. Johansson & Winger, 2007;Os, 2007;Winger, 2008Winger, , 2007. 2. The data construction was accomplished in cooperation with associate professor Brit Eide and associate professor Nina Winger. cording to Hännikäinen (1999Hännikäinen ( , 2001 and van Oers and Hännikäinen (2001) togetherness is characterized by a feeling of being a part of a group and affective relationships between the people involved. Day care groups comprise not only one group showing signs of togetherness, there will be several subgroups for example groups of friends. The concept of togetherness does not imply a conflict between emphasizing group life and emphasizing individual needs and caring for the individual child. According to Hännikäinen (1999Hännikäinen ( , 2001 feeling of togetherness helps people to see and treat each other as subjects; and I will claim, that togetherness must be built upon respect and consideration for the individual child. A kind of togetherness based upon oppression of individuals, can hardly be considered as togetherness.

Hännikäinen
For caretakers in day care centres, it might be demanding to distribute their awareness and care to all the children in the group (Winger & Os, 2010), maybe because of the predominating idea of the dyadic nature of ideal relationships between caretakers and toddlers. But now it is time to recognize that relational processes between caretakers and children might be different in day care settings compared with home settings. In home settings the interactions between caretakers and children often are dyadic. In group settings, the ratio between children and caretakers is not one-to-one, and the challenges the caretakers are facing are how to be aware of, and take care of several children at the same time. In recent attachment research there are implications that concern for the individual and concern for the group not necessarily are conflicting aims in pedagogical practices. Several researchers emphasize that group-related sensitivity rather than individual child-focused sensitivity, is the key to secure attachment patterns in day care groups (Ahnert, Pinquart, & Lamb, 2006;De Schipper, Tavecchio, & Van Ijzendoorn, 2008;Howes, Galinsky, & Kontos, 1998). De Schipper and co-workers (2008, p. 468) say that: "In center day care, children not only need sensitive caregivers, but more importantly, they need sensitive caregivers who find the time to display their sensitivity frequently enough to create a sense of confidence in their availability as a safe haven and a secure base." To find the time to display sensitivity to all the children in groups, presupposes development of pedagogical principles tailored for group care (Winger & Os, 2010). And a core principle is obviously to make use of the possibilities that are inherent in group settings since many children have to share the attention of quite a few adults. Feelings of togetherness do not arise as a result of just putting toddlers together. Physical proximity is a necessary but not sufficient condition for the development of togetherness in groups. My question is how teachers work with peer-relations as a vital part of care for children in group care settings.

MEDIATION OF PEER-RELATIONS
In the research field concerning caretakers' involvement in peer relations some questions are raised. I will pay attention to three of them: The first one concerns whether caretakers involve themselves in peer-relations. The second question bring into focus if caretakers' involvements affect the relations between children, and the last one concerns whether caretakers should mediate peer relations between toddlers or not.
To the first question: Do caretakers involve in children's peer relations? Empirical research indicates that they do not. Howes and Clements (1994) comment that it is noteworthy how little teachers attempt to mediate peer-contacts, even if the teachers are sensitive and responsive in their own interactions with individual children. According to File (1994) children are expected to pick up social skills on their own. Johansson and Pramling Samuelsson (2006, p. 196) claim that there seems to be a general agreement among caretakers that they should not intervene in children's playing activities. This attitude, they say, can result in a neglect of key aspects of children's experiences and learning. NCKO in the Netherlands (Riksen-Walraven, 2011) finds; using their Caregiver Interaction Profile Scales; that caregivers' fostering of positive peer-relations represents the lowest score of all items in their scales, since it almost never happens.
My results indicate that caretakers do mediate peer relations. In my material I find 491 mediating sequences, each consisting of several mediating utterances (about 2600). The frequency of mediation is 1.6 times each minute. But there are major differences between the nine day care centres. If the mediating practice in my material is high or low frequent compared to other aspects of communication is uncertain, but the occurrence tells me that it probably is an important aspect of interactions between adults and children that calls for attention.
The second question is how caretakers' involvement affects peer-relations. Empirical results differ substantially. Some researchers summarize improvements (Girolametto, et al., 2004;Hay, et al., 2004) and some summarize no effects (Howes & Clements, 1994;Howes, Hamilton, & Matheson, 1994;Howes & Tonyan, 1999). But several researchers agree that the effect depends on how caretakers involve themselves in relationships between toddlers (Girolametto, et al., 2004;Pramling Samuelsson & Fleer, 2010;Williams, et al., 2010). A brief summery is that mediation that seems to work well for toddler's peer-relations is characterized by sensitivity combined with a positive and supervising approach. Restrictive mediation and firefighter approaches seem to be associated with poorer and short-lived peer-interactions.
These conclusions are to a large degree consistent with findings in my material. But in addition my preliminary analyses indicate that sensitive involvement combined with attentive distance, lasting over time seem to be associated with richer and more lasting peer-interactions. The third and last question concerns whether caretakers should involve in peer-relations? This question will be postponed to the end of this article.

MEAL-TIMES
Why are meals important settings for mediating peer-relations between toddlers? We could twist a well-known saying from Norwood Russell Hanson (1958), philosopher of science: "It is more to seeing than what hits the eyeballs" and say "it is more to meal-time than what hits the mouth".
First of all, meal-time is a group situation and one of few situations during the day in Norwegian day care centres where all the children are gathered together. It is a situation that gives opportunities for feelings of togetherness for the group as a unit at the same time as sitting together around a table gives opportunities for children to make themselves visible in the group and to be made visible for their peers.
A meal also offers several learning opportunities. Sharing a meal seems to be a situation where conversations are central. Cote (2001) claims that mealtimes provide good opportunities for conversations because then teachers are not engaged in working with overt educational goals. To participate in conversations with peers and caretakers might contribute to toddlers' ability to express themselves in a clear and understandable way to peers. According to Hay (2006) the ability to communicate clear and understandable is important for the development of peer-relations. Aukrust and Snow (1998) emphasize that meals seem to create culturally specific ways of talking that constitutes opportunities for socialization of children. Bae (2009) focus implicit learning, the unplanned and not articulated, embedded in interactions between caretakers and children and between peers around the table. Ehrlich and Blum-Kulka (2010) have in particular focused peer-talk and claim that peer-talk represents a double space opportunity: Children create shared meaning at the same time as they learn to use language to express themselves. This is what Ødegaard (2007a) finds in her study of children's narratives in toddler groups during meal-time in day care. Shared meaning is a presumption for shared play and especially for pretend play (Bondioli, 2001;de Haan & Singer, 2003;Singer, 2002).

RESULTS: MEDIATION OF PEER-RELATIONS DURING MEAL-TIMES
I will present some aspects of mediation of peerrelations that seem to appear during mealtimes in my material based on caretakers' verbal mediation. So far the analyses indicate that there are some opportunities for sensitive, active, attentive and lasting mediation of peer-relations during meal-times. I will focus on conversations and mediation of togetherness in the group, and also illuminate some challenges that call for attention. The children participating in this conversation are about two years old. First Anna takes an initiative to talk about Maria and her dad and Andrew follows up. After a while, Andrew draws their attention to Peter's dad passing outside. In this conversation the teacher is sensitive and responsive. Inspired by Lobman's (2005Lobman's ( , 2006 theory about improvisation in interactions we could say that she is responding in a way that invites to continuation of the conversation. She elaborates the children's contributions. She is sharing her attention between Andrew and Anna so that both of them get opportunities to take turns and participate in the conversation.

CONVERSATIONS
She is also participating to joint attention and is chaining the children's contributions. In this part of the conversation the teacher is active, but later she sometimes withdraws for a while and leaves the floor to the children (attentive distance). This example is a small part of a lasting conversation which moves from one topic to another in a smooth flow. Sometimes the children introduce new topics and sometimes the teacher does. They talk about respecting each other's limits, rules for how to use their voice inside and outside, plans for a birthday-celebration and the birthday child's privileges is negotiated. Different seasons and present weather conditions also become subjects for discussion.
During this conversation the children learn something about the world, but they also practice and develop their social skills like turn-tak-ing, listening, responding, follow-up and they practice negotiations and learn something about how to handle different perspectives. Minor differences are sorted out. This make the children move from each having their own meaning to shared meanings built on understanding of each other's perspectives. And when the teacher leaves the room for a while the children continue the conversation on their own.
But it is worth noting that even if the children are active participators in conversations, the teachers' contributions often are dominant. In general they are chaining and elaborating children's input, but they rarely encourage the children to communicate directly with each other. They do not, to a great degree, provide for peers talking to peers. Instead they are intermediaries between the children. Often the conversations seem to be a kind of serial-dyads (cf. Schaffer & Liddell, 1984). Teachers' dominating position in conversations in day care is documented earlier (see e.g. Cote, 2001;Durden & Dangel, 2008;Girolametto, Weitzman, Lieshout, & Duff, 2000;Wood, McMahon, & Cranstoun, 1980). The fact that the children in the example presented above, at two years of age, are able to continue the conversations when the teacher leaves the table for some time or she holds herself back for a while (attentive distance), indicates the promising opportunities embedded in toddlers' peer talk.
Example 2: Conversation 2; okay, but … I will present a contrasting example. Rasmus (2;11 years) initiates a conversation about his stomach and tries to draw attention to other children in the group:

Rasmus: This is my stomach (pointing at his stomach).
Caretaker: Yes, it is your stomach, but now you have to eat.

Rasmus: That it his stomach (pointing at another child's stomach).
Caretaker: His stomach yes … mmm. Take a bite (gives Rasmus a piece of bread).

(Rasmus arise and stand on his chair and points at the children seated around the table)
Rasmus: His stomach too, his stomach and his stomach. nordisk barnehageforskning 2013 6(30), 1-9 issn 1890-9167 www.nordiskbarnehageforskning.no Caretaker: Yes, everyone has a stomach. Now you have to eat.
The caretaker in this group gives brief responses to Rasmus. However, she does not follow up even if his initiatives might have led to conversations about sizes and form, big stomachs and small stomachs etc. It could even have been a starting point for interactions between the children about similarities and differences. According to de Haan and Singer (2001) comparing seems to be typical for young children's own language for expressions of togetherness. But the teacher's focus is on accomplishing the meal without any disturbances: Meals are for eating. So later, when two children start talking to each other she tells them to stop and concentrate on eating.
Both these teachers are mediating peer-relations, but their mediating strategies differ. Different mediating styles form the basis for various possibilities for children's peer relations. The first teacher encourages joint attention, turn-taking between the children and development of shared meaning. The second one misses opportunities to strengthen the relations between peers that might have been a contribution to feelings of togetherness between the children in the group.

MEDIATION OF TOGETHERNESS
Unlike mediation in free play situations, meals seem to be a context where the mediation of peer relations to a certain degree explicitly refers to the group as a community, often expressed as routinized practices. These routinized practices might lead to a feeling of togetherness in the group (Os & Eide, 2013). They are characterized by conventional, fixed and repetitive patterns of action which makes it easy for small children even they who have not a well-developed verbal language, to take part (Corsaro, 2002, p. 183;Corsaro & Johannesen, 2007;Eckerman & Peterman, 2003;Singer, 2002).
Before starting the meal, the toddler groups sing and in some groups they are holding hands while singing. They have name-games where all the present children's names are mentioned, but not the names of he absent children. Not paying attention to the absent children can be seen as a missed opportunity for mediation of peer-relations (cf. Os & Eide, 2013). According to van Oers and Hännikäinen (2001), talking about absent children is a way to mediate togetherness in a group. However, these games around the table have repetitive patterns and most of the time all children are participating even if it sometimes happens that children resist to hold hands or sing. The teachers tell their peers that the resisting children are tired or give other reasons for their resistance, explanations that the children seem to accept.
When the day care center provides meals the children are encouraged to pass the food to each other and sharing whatever is on the table is strongly emphasized. But when the parents provide the food and children bring their lunch-boxes, the rules change. The children are not obliged to share, and sometimes they are not allowed to share. Due to the reduction of day care provided meals in Norwegian day care centres, this happens quite often. It is obvious for observers that the situation makes the teachers feel rather uncomfortable, and that they have difficulties explaining the new rules to the children. Anyway, the lunch-box practice reduces opportunities to work with togetherness within the groups.

CHALLENGES
In my material, it seems like different teachers and different day care centres have different challenges when it comes to working with peer relations and togetherness in the group during meal time. Some teachers seem to have strategies for focusing on relations between the children, while others seem to have the priority in feeding the children. Nevertheless even if there are differences, there seem to be some challenges that are common, to different degrees, for all toddler groups in my material.
First the teachers should engage themselves in conversations, but they also have to be willing to redraw and leave the floor to the children. It might be potentials for peer talk in encouraging the children to address their talk to each other. The serial-dyads the teachers engage the children in seem to lead to fewer children, often just one or two, participating in conversations at the same time. This is in accordance with Ødegaard's (2007b) findings.
The routinized practices during meal time seem to represent joy to the children. In all the centres they sing a song before the meal starts, but there are considerable differences between the centres, when it comes to how to cope with routinized games and conversations. Often these games and conversations, like talking about all the children, their names, what they are eating etc., are initiated by the children as in the example presented earlier, where a boy tries to initiate a conversation about stomachs. Some teachers encourage and follow up these initiatives, while others ignore or actively forbid them. To find a balance where both adult order, and the chaos that children's initiatives might lead to, are parts of everyday life in child care and represents challenges for the teachers. Probably these kinds of games and conversations might lead to joy and a feeling of togetherness between the children (cf. Berentzen, 1994). Grindland (2012) deliver quite convincing arguments for giving children opportunities to influence what goes on during meals, even if it sometimes brings elements of chaos into the situation. Berentzen (1994) argues for the potential of joy and learning, and Grindland (2012) for the feeling of togetherness that might emerge when the teachers loosen some of their control in their interactions with children. And not being in total control all the time, might represent a challenge for the teachers (cf. Bae, 1996), but as Løkken's (1990Løkken's ( , 1996 research on group glee in toddler groups indicate, there is a potential for interactions and shared joy between the children when the teachers reduce their control in interactions.
The last challenge is connected to provision of meals; whether the meal is prepared in the day care centres or the parents are responsible for bringing lunch-boxes. This is not an issue that can be solved by the teachers alone. It is connected to decisions by the authorities and to owners of day care centres. By offering center provided meals, there certainly will be increased opportunities for the teachers to facilitate peer relations and togetherness in toddler groups, but I will leave this discussion to another forum.

CONCLUSION
To end this article, I will go back to the question: Should caretakers involve themselves in peer-relations?
Well -that is not a relevant question: The question should not be if caretakers should mediate peer-relations. Independent of intentions, caretakers are involved in children's peer relations. Just by being present together with the children caretakers do mediate something about how to relate to peers, and are together with the children creating the atmosphere in the group.
To avoid unreflected practice, which might have serious consequences for the individual child and for the development of the group atmosphere, it is important to expand knowledge and awareness about mediation of peer-relations. Hence the relevant question is how caretakers mediate peer relations between toddlers in day care centres.