Social Interaction

1. Introduction

Interpersonal touch is a central mode of embodied expression, experienced from the first moments of life, and essential to normative growth and healthy development (e.g., Montagu, 1971). An important purpose of touch, such as hugs, pats, stroking, and tickling, which are often used together with talk and other communicative modalities, is to convey positive affect and emotion. As Linden (2015) has argued, “touch can be used together with other sensory signals to communicate a broad range of emotional intentions including support, compliance, appreciation, dominance, attention getting, sexual interest, play, and inclusion” (p. 29). Recently, following the public outcry over the Trump administration’s separation and caging of migrant children at the southern U.S. border, the media has reminded us of the necessity, and even naturalness, of affectionate touch in children’s everyday lives (Romm, 2018).

Over the last few years, research on adult-child interaction in institutional settings has paid an increasing amount of attention to affectionate touch (e.g., Bergnehr & Cekaite, 2018; Burdelski, 2010; Burdelski & Mitsuhashi, 2010; Cekaite & Bergnehr, 2018; Cekaite & Holm Kvist, 2017). The present paper builds upon this research by focusing on ‘compassionate touch’ (Cekaite & Bergnehr, 2018) that is deployed in situations when a child is in distress caused by pain so as to soothe and relieve it (Cekaite, 2020). In comparison to psychological and philosophical perspectives that often view compassion as an emotional state arising from observing or hearing about another’s circumstances (e.g., Nussbaum, 1996; Snow, 1991), this study views compassion as a situated and negotiated practice that emerges and is enacted in social interaction (see Ruusuvuori, 2005 on empathy and sympathy). From this perspective, compassion is considered to be an ‘affective stance’—defined as “mood, attitude, feeling and disposition, as well as degrees of emotional intensity vis-á-vis some focus of concern” (Ochs, 1996: 410)—that is indexed through linguistic, paralinguistic, and embodied resources, including touch. As in other institutional settings, in preschool, teachers’ use of touch can be characterized as a ‘formal professionality’ (Mondada & Tekin, this issue) that is rooted within asymmetrical relationships and linked to the rights, obligations, and responsibilities to engage in (appropriate) acts of touch on another’s body (e.g., controlling, caring, displaying affection). Although preschool teachers’ compassionate touch emerges in various situations, in this paper I focus on situations of peer conflict and accidents, which have been either observed by a teacher or conveyed through children’s reports, and have (usually) resulted in children’s displays of distress caused by pain (e.g., crying). In these contexts, teachers’ compassionate touch was not only aimed at soothing and relieving the child’s distress, but also at restoring the moral and social order of peer relationships (e.g., encouraging the children to go back to playing together). The analysis is guided by the following questions: 1) What parts of the body are centrally deployed and received in compassionate touch and in what ways? 2) How is compassionate touch deployed with other communicative resources and practices, especially talk? 3) How is compassionate touch used in relation to larger sequences of interaction? 4) How do children respond?

2. Background

Research on adult-child interaction has shown how adults mobilize interpersonal touch (i.e., tactile and haptic acts on another’s body) for various purposes, such as gaining children’s attention, displaying affection, controlling, and caring (e.g., Burdelski, 2010, 2015; Cekaite, 2010, 2015, 2016; M.H. Goodwin, 2017; M.H. Goodwin & Cekaite, 2018; Guo, Katila & Streeck, this issue). A number of studies have discussed touch in families for displaying affection, such as ‘intimate touch,’ ‘comforting touch,’ ‘haptic soothing,’ ‘bodily endearment,’ and ‘emotional touch’ (e.g., Cekaite, 2020; M.H. Goodwin, 2017, 2020; M.H. Goodwin & Cekaite, 2018; Mondada, Monteiro & Tekin, 2020; Tahhan, 2014). Such touch may involve hugging, kissing, embracing, stroking, picking up or carrying a child, sitting a child in the adult’s lap, and physical play. It often co-occurs with affective talk (e.g., praise, terms of endearment). For instance, in Japanese families, Tahhan (2014) examined the enactment of the important cultural notion of sukinshippu ‘skinship,’ or ‘intimacy through touch’ (p. 11). In addition to the daily rituals of co-sleeping and co-bathing observed in the anthropological literature, Tahhan points out that skinship was also manifested in a range of actions and activities which foster connection and intimacy, including playing, holding hands, massage, onbu (transporting on the back or front in a sling), dakko (picking up and holding), and reading picture books.

A number of studies have also examined interpersonal touch in preschools (e.g., Burdelski, 2010; Ben-Ari, 1996; Bergnehr & Cekaite, 2018; Burke & Duncan, 2015; Cekaite, 2010; Cekaite & Holm Kvist, 2017; Hayashi & Tobin, 2015). In their analysis of teacher-to-child touch in Sweden, Bergnehr and Cekaite (2018) identified five categories: 1) ‘control touch,’ 2) ‘affectionate touch,’ 3) ‘affectionate control touch,’ 4) ‘assisting touch,’ and 5) ‘educative touch.’ In particular, they observed that affective touch was used to “show fondness, to comfort, or to express praise and approval”; the forms deployed included “embracing, holding someone in one’s lap, patting, stroking, caressing and hugging” (p. 318). Among various kinds of affectionate touch, the researchers identified ‘comforting touch,’ which was “used in response to the child’s distress and involves hugs, strokes, caresses, for instance when soothing a crying child” (p. 321). They also found that affectionate touch could be combined with control touch, or what they called ‘affectionate-control touch,’ which was deployed to “control – in a mitigated way – the child’s bodily position or orientation”; the haptic forms “involved stroking a child’s arm, a half-embrace, or lifting the child gently and putting the child in one’s lap to control the child’s bodily conduct” (p. 218). They reported that about 25% of cases were affectionate touch (81 out of 322 tokens) and 20% were affective-controlling touch (67 out of 322 tokens). Their findings suggest that when teachers touched children, more than 46% of the time they did so to display affection towards children.

In Japanese preschools, Tahhan (2014), among others (Ben-Ari, 1996; Burke & Duncan, 2015; Hayashi & Tobin, 2015), observed affective touch at naptime, where caregivers often used onomatopoetic expressions in lulling them to sleep (also see Kuroshima, this issue) while softly patting or rubbing children on their back, stomach, legs, or head. In their cross-cultural study of New Zealand and Japanese preschools, Burke and Duncan (2015) reveal examples of teacher-to-child touch that are suggestive of Bergnehr and Cekaite’s (2018) category of affective-control touch. For instance, in describing a scene in which a boy was fidgeting and poking other children during a school assembly, a teacher came up to him and, rather than verbally reprimanding him—as the authors argued would be typical in New Zealand—she used a half-embrace by putting an arm around the child and holding his hand, while silently gazing at him until he calmed down (see also, Hayashi & Tobin, 2015: 32-34 for similar observations in a Japanese preschool). As related above, studies in preschools in various societies have observed the use of affectionate touch (often combined with control touch); however, few studies have detailed such touch within (and as constitutive of) social actions and sequences of actions and its potential import for child care and sociality.

3. Methods, Data and Setting

This study takes a multimodal conversation analysis approach (e.g., C. Goodwin, 2017; M. H. Goodwin & Cekaite, 2018; Mondada & Tekin, this issue) in examining interpersonal touch initiated by caregivers to children in preschool. It employs analytic tools furnished by conversation analysis (e.g., Sacks, Schegloff & Jefferson, 1974), which has increasingly focused on multimodal and multisensory interaction (e.g., Mondada, 2019). This approach entails closely observing and audio-visually recording naturally occurring interaction, reviewing and transcribing the recordings, and making collections of cases. In line with this approach, I pay particular attention to the embodied ‘participation frameworks’ (Goffman, 1981) being organized through touch in displaying stances and performing social actions with children.

I draw upon a corpus of audiovisual data in a Japanese preschool (48 hours). In Japan, the majority of children (more than 80%) attend preschool. In many preschools, including the one where I did fieldwork, the primary aim is to care for children and encourage their sociality, including their development of empathy and the ability to get along with others as part of a social group (Burdelski, 2010). This sociality is fostered within group activities, including play, storybook reading, walks outdoors, crafts, eating, and naptime. In these activities, teachers employed various kinds of touch for tactile (e.g., affectionate) and haptic (e.g., controlling) acts. The data for this analysis was assembled as part of a larger study on teacher-child and peer interactions. The analysis of compassionate touch was carried out by reviewing a collection of examples (40 in total) that were made for examining adult-to-child touch in preschool from a cross-cultural perspective (Burdelski & Cekaite, 2020). The examples identified are thus not exhaustive, but nonetheless represent what I deemed “typical” examples. Due to space limitations, I will provide a qualitative analysis of three representative excerpts, drawing upon transcription conventions described in Mondada (2018) (see Appendix).

4. Responding to children’s distress in peer conflict and accidents

As noted above, the teachers recurrently deployed compassionate touch (along with other kinds of touch, such as control touch) in responding to children’s (displayed and/or imagined) distress arising out of peer conflict or accidents that occurred during play. This touch was used along with verbal resources and practices that conveyed an affective stance of compassion (i.e., concern for the [in this case, physical] suffering of another) within a sequential organization. Previously, we have characterized teachers’ intervention in peer conflict as having two to three phases: 1) intervention, 2) investigation, and 3) reparation (Burdelski & Cekaite, 2020). Here, I will employ this organization as a heuristic, and note that all three phases were not always observed (e.g., when a teacher has witnessed a conflict or accident she may skip the investigation of what happened). Although compassionate touch could occur in any of these phases, below I will primarily focus on the first and third, as these two were where compassionate touch was most often deployed.

4.1 Compassionate touch in the intervention phase

An illustration of compassionate touch within the intervention phase is shown in Excerpt 1. Here, a boy, Anik (ANI), has been playing with a toy train on the floor as a nearby teacher (TEA) was addressing other children who were also engaged in play nearby. When a girl, Mao (MAO), suddenly came up behind Anik and pushed his back (without warning or evidence of a prior conflict between them), causing Anik to fall face forward to the floor, the teacher, who ostensibly observed the push, ceased her interaction with the other children and intervened by quickly moving Mao away from Anik in order to attend to him. As we join the excerpt, this intervention continues, as the teacher comes over to Anik who has already lifted himself halfway up off the floor onto his hands/knees and engages him in compassionate talk and touch (at the beginning of each excerpt, each child’s age, sex, and nationality are provided; children’s ages are shown in years and months [2;4 = 2 years and 4 months]).

Video 1. Surprised. (Mao: Japan, girl, 2;4; Anik: India, boy, 2;6). May 17, 9:32am

Excerpt 1. Surprised. (Mao: Japan, girl, 2;4; Anik: India, boy, 2;6). May 17, 9:32am

In this excerpt, the teacher engages in a series of touches that display an affective stance of compassion, by first putting her arms around Anik’s waist (line 01 and Figure 1) and then lifting him up off the floor and turning him around to face her while lightly gripping both of his forearms (Figure 2-a). These touches and accompanying talk are ‘mutually informing’ (Goodwin, 2000): The talk elaborates on the touches and the touches elaborate on the talk in congruent but different modalities. More specifically, after placing her hands on Anik’s waist in order to lift him up off the floor (line 01 and Figure 1), the teacher addresses him by posing a question that inquires about his current condition (line 01: >daijobu? Aniiku< ‘Are you okay, Anik?’). As the teacher stands Anik upright and turns him toward her, she utters a ‘response cry’ (Goffman, 1981), which co-occurs with vowel elongation (line 02: a:ra::::::::: ‘Oh my’). In these ways, the teacher’s verbal resources, together with touch, display a heightened affective stance of compassion towards the child who is presumed to be in distress.

In comparison to many other examples in these data, here the child who is the recipient of the teacher’s compassionate touch did not display overt verbal signs of distress, such as crying; rather, the teacher’s having witnessed the child being forcefully pushed and falling face forward to the floor has seemed to invite her to ostensibly “imagine” that he is in distress and in need of assistance and compassion. As Nussbaum (1996) has suggested, feelings of compassion towards another always involve an element of imagining the other’s situation. In these data, imagination was a public display (rather than solely an internal state or feeling). Here, along with touch, this display involved “glossing” (Burdelski, 2015), or putting into words, what the child could or should be thinking or feeling in relation to the untoward act. More specifically, in saying, ‘You were surprised, right’ (line 03), the teacher verbalizes the child’s reaction to being pushed, while rubbing Anik’s side and back on the site of the child’s body that was the target of Mao’s push. In this situation, this gloss somewhat downplays the magnitude of what has just happened (as in English, in Japanese bikkuri ‘surprise’ can have either positive or negative connotations). Thus, the teacher has glossed the child’s reaction to having been pushed and falling face forward on the floor as a rather “mild” feeling, rather than, for instance, as being angry (as in Miller and Sperry, 1987 on mothers’ reactions to children being pushed/hit in a U.S. working-class community). This response, which is laminated with compassionate touch, may prevent the conflict from escalating, and shows the teacher’s preference for responding first to the child in distress before disciplining the child who caused it (the disciplining is not shown here).

The final point to make about teacher’s compassionate touch (and talk) is that, in such sequences, children are agents who align with and on occasion resist it. At first the child aligns, but, as the teacher continues to employ compassionate touch (and control touch), he turns his head slightly away from the facing formation (Figure 2-b), displaying a possible desire to return to playing with the trains (which are behind him). This desire is further hinted at in line 04, as the child quickly turns his head to his left (Figure 2-c) and then about 180 degrees around. These embodied actions co-occur as the teacher begins to bring the intervention phase to a close by stopping her rubbing of Anik’s back, while turning towards Mao who has walked away from the scene in order to begin the reparatory phase where she will demand an apology from Mao (not shown, but see Burdelski & Cekaite, 2020). Here, the child’s gaze and head and torso turning also has the effect of collaboratively bringing the compassionate touch and the intervention phase to closure.

4.2 Compassionate touch in the reparatory phase

As mentioned above, following intervention in peer conflict and accidents, teachers recurrently engaged children in a reparatory phase in order to restore the social and moral order between the children involved. During this phase, they often used compassionate touch, especially with children who continued to display distress, such as crying.

An illustration of this is shown in Excerpt 2, which is divided into two parts (2-a and 2-b). As a teacher (TEA, different from Excerpt 1) is straightening up a rack of futons after the children’s naptime, a Butanese boy, Danan (DAN) (3;1), and a Japanese boy, Kazu (KAZ) (5;7), begin walking towards her (TEA’s gaze is towards the futons, or in the opposite direction of the children). The older boy, Kazu, has his hand to his forehead as Danan is crying. After the teacher initiates an investigative phase (begun with a typical question, doo shita no? ‘What happened?’), the older boy conveys that the boys accidentally bumped heads during play. During this investigation, which lasts for 17 seconds, the teacher continues to straighten up the futons and does not deploy any touch, despite the fact that Danan is crying from the beginning. We join the excerpt as Danan starts to cry more loudly. In response, the teacher puts her futon-straightening activity on hold and attends to him.

Video 2-a. I’m sorry. (Danan: Butan, boy, 3;1; Kazu: Japan, boy, 5;7). December 19, 3:04pm

Excerpt 2-a. I’m sorry. (Danan: Butan, boy, 3;1; Kazu: Japan, boy, 5;7). December 19, 3:04pm

Here the teacher uses talk to mediate between the two children while engaging in compassionate touch on the crying child (Danan). More specifically, after placing her hands on Danan’s head, she starts rubbing the back of his head and forehead with her hands (line 03 and Figure 3), and uses reported speech to convey to him a version of Kazu’s speech (line 03: ‘He says both of you got hurt’), which was originally addressed to the teacher during the investigative phase. In Japanese caregiver-child triadic interaction, adults often use reported speech to convey a message from one child to another in mediating a conflict or for other concerns (Burdelski, in press). Here, this reported speech also functions as a mediating practice, but one that is laminated with compassionate touch. This mediation continues as the teacher glosses the precipitating event that led to Danan’s crying (line 04: ‘It can’t be helped’). This utterance evokes the bumping as accidental, which ostensibly attempts to bring the reparatory phase to closure.

When Danan begins to cry even louder, the teacher shifts ‘footing’ (Goffman, 1981) by crouching down to his eye level and apologizing to him (line 07), while continuing to rub his head in a compassionate way. In comparison to the English “I’m sorry” as an expression of sympathy (e.g., in hearing of the death of a friend’s family member: “I’m sorry for your loss”), the Japanese expression gomen ne ‘sorry’ is not used in that way. Rather, there is always a degree of responsibility (self-responsibility or on behalf of a group member) bound up with and displayed through the use of this expression (Burdelski, 2013; also see Björk-Willén, 2018 on Swedish caregiving). Thus, the teacher’s gomen ne, together with her touch, not only conveys compassion, but also responsibility as a caregiver (e.g., for not doing something that could have prevented the boys from bumping heads). As in other examples in these data involving accidents, although both children were clearly in distress, the teacher provides compassionate touch and directs this apology to the child who is displaying heightened distress by crying.

Moreover, as in other examples in this data (e.g., Excerpt 1), the teacher’s compassionate touch towards the child in distress (here, the crying child) is deployed for multiple purposes, as illustrated in the continuation of Excerpt 2-a.

Video 2-b. (continuation of Video 2-a). (Danan: Butan, boy, 3;1; Kazu: Japan, boy, 5;7). December 19, 3:04pm

Excerpt 2-b. (continuation of Excerpt 2-a). (Danan: Butan, boy, 3;1; Kazu: Japan, boy, 5;7). December 19, 3:04pm

As the teacher and Danan are now in face-to-face alignment (Figure 4), the teacher’s compassionate touch is performed in a slightly different manner from before. More particularly, while using one hand to rub Danan’s head, she uses her other hand to briefly move Danan’s hair away from his forehead, which enables her to quickly inspect it for any visible injury that may need first aid (usually applying ice in cases of visible redness or a bump). At the same time, she poses a question to him to inquire about the location of his pain (line 09: ‘Where did it hurt?). As other researchers have shown (e.g., Raia, M. H. Goodwin & Deng, this issue), touch employed in diagnostic ways can also have a soothing effect on the recipient. Thus, the teacher’s current touch can also be considered a kind of ‘assisting touch’ (i.e., used to attend to the child’s bodily well-being) (Bergnehr & Cekaite, 2018), which emerges within a prior episode, unfolding sequence of interaction, and an accompanying utterance (‘Where does it hurt?’)—the utterance coming in response to the child’s continued crying and after the teacher’s previous displays of compassionate touch.

When no visible injury is discovered, the teacher ostensibly attempts to close down the reparatory phase by using an ‘assessment’ (e.g., C. Goodwin & M. H. Goodwin, 1992) that is positively reassuring (line 11: ‘You’re fine, You’re fine’), which hints that his crying is no longer needed. I will not have space to detail the remainder of this interaction but suffice to say that here the teacher engaged in compassionate touch for seven more seconds. During this time, she recycled her earlier utterances: reporting Kazu’s speech to Danan, inviting Danan to confirm that he got hurt too, and showing compassion with the expression gomen ne again. Finally, although the teacher returned to her straightening up of the futons, when Danan’s crying began to subside, she again halted her futon activity, and re-engaged in touch but this time only for a brief moment, while asking Danan if he was ‘fine.’ When no answer was forthcoming (signaling to her that he was now indeed fine), she praised him (for stopping crying), which brought the reparatory phase to closure as the boys when off to play again.

In these ways, the teacher used compassionate touch in the reparatory phase, aimed at calming and soothing a crying child while attempting to reestablish the social and moral order of the peer relationship. In contrast to Excerpt 1, compassionate touch was not deployed immediately, but was manifested as a delayed touch, 17 seconds after the children came over to the teacher and reported what happened (even though Danan was crying from the beginning). Once the teacher temporarily disengaged from her prior activity and turned her full attention to the children, she used compassionate touch for a lengthy time over a number of utterances in response to the younger child’s continued and upgraded crying, demonstrating once again how children are active agents in co-constructing these compassionate touch sequences, though this time by inviting the teacher’s compassionate touch rather than resisting it.

4.3 Compassionate touch in the post-reparatory phase

As I suggested earlier, once the reparatory phase has been ostensibly brought to completion, children may continue to display distress (e.g., crying). This may invite further compassionate touch from the teacher. In this section, I will examine this use of touch in what can be called a post-reparatory phase. This sub-phase, moreover, usually occurred while the teacher simultaneously engaged in competing activities, such as addressing surrounding children (who were not involved in the prior conflict or accident).

These points are illustrated in Excerpt 3. Prior to the excerpt, Reo (REO) (2;8) and Erhi (5;8, not shown here) had been playing roughly when they both fell down onto the floor. When Reo began to cry, a teacher (TEA, same as excerpt 1), who had witnessed the children fall to the floor (but ostensibly had not seen the precipitating event that led to this fall), initiated an investigative phase in which she asked the children what happened and determined that they ‘bumped into each other’ (i.e., framing it as an accident). She then prompted Erhi to apologize to Reo—it can be noted that older children are often prompted to apologize to younger children for accidents especially when the younger child is crying (Excerpt 2 above was an exception). After Erhi has apologized to Reo and left the scene, the teacher picked up Reo (who continues to cry) and placed him in her lap, telling him ‘don’t cry.’ During this display of compassion in the post-reparatory phase, when a male child (C-1) has come over to the teacher, she told him, ‘please sit down’ (suwattete kudasai), as preparation for an upcoming group activity. As we join the excerpt, the teacher continues to embrace Reo while addressing another child (C-2) who is also not sitting down on the floor as expected (off camera). In order to highlight in the transcript the competing participation frameworks in which compassionate touch on the crying child is embedded, this excerpt is divided into two columns: On the left is the teacher’s interaction with Reo, and on the right is her interaction with the other child who is not sitting down (C-2).

Video 3. (Reo: Japan, boy, 2;8; C-2: unidentified child Kazu). December 22, 12:20pm

Excerpt 3. (Reo: Japan, boy, 2;8; C-2: unidentified child Kazu). December 22, 12:20pm

In the post-reparatory phase, teachers’ compassionate touch was primarily aimed at soothing a child who was still crying into order to relieve the child’s distress so that he or she can re-enter a prior activity or begin a new one. Here, having positioned the child (Reo) in her lap (Figure 5) in a kind of ‘nested alignment’ (Ochs, Solomon & Sterponi, 2005), the teacher rubs Reo’s head with her hand. As she continues to embrace him, she adjusts his leg by pulling it closer in to achieve a tighter intercorporeal alignment. Within this alignment, she issues a ‘directive’ (e.g., Cekaite, 2015; Takada, 2013) to Reo to not cry (line 05). In this directive, her tone of voice including rapid speech (>nakanaide< ‘don’t cry’) indexes a slight irritation with his continued crying. Although this directive can be heard as a mild upbraiding, the teacher’s touch mitigates it with a display of compassionate touch, in order to soothe him and stop his crying. Earlier we have observed how touch and talk are often mutually elaborating (Excerpt 1), but here there is a hint of incongruency between the two modalities: talk primarily as control and touch primarily as soothing.

As mentioned above, in the post-reparatory phase of conflict resolution, the caregivers’ attention to the crying child through compassionate touch often occurred within competing participation frameworks. Here, the teacher uses touch, gesture, and talk to oscillate between two frameworks: She uses compassionate touch towards the child in distress while interacting with one or more other children who were not connected to the precipitating event. This can be observed in the teacher’s talk and body: While she continues rubbing Reo’s head, she orients her gaze to a child who is not sitting down as expected, in preparation for an impending group activity (a teacher-led kamishibai ‘story using paper cutouts of characters’). The teacher then addresses this child by name (here, C-2), and issues a directive to him to ‘please sit down’ (line 02). As she issues this directive, she momentarily takes her hand off Reo’s head and uses it to produce a gesture within a different participation framework: a deictic point towards the floor that co-occurs with talk (i.e., directive to C-2 to sit down). Thus, in order to produce this embodied directive (verbal + gesture), the teacher briefly puts the rubbing of Reo’s head (i.e., compassionate touch) “on hold.” However, she does not entirely place this touch on hold, as she continues to embrace Reo in her lap with the other hand (Figure 5). Once her pointing gesture is completed, the teacher immediately activates the same hand to adjust Reo’s leg in a way that pulls him closer to her, achieving a tighter intercorporeal alignment: thus reestablishing the compassionate touch with both hands. In these ways, novel aspects of touch revealed in this excerpt are that it can occur within multiple and competing participation frameworks and be partly placed on hold in order to engage in interaction with other children who invite or demand the teacher’s attention. Here, the teacher’s verbal directive towards the other children was “preparatory” in the sense that she attempted to spatially organize their bodies in setting up an activity (Mondada & Burak, this issue, on touch and talk for this purpose). After several more seconds of her compassionate touch on the crying child, the teacher set Reo down on the floor in order to prepare the materials for this activity, bringing the post-reparatory phase to closure.

5. Conclusion

This paper has examined teachers’ compassionate touch in a Japanese preschool, deployed in situations of peer conflict and accidents during play. This touch responded to different situations that the teacher had either witnessed (partially) or not witnessed. Here, I will address the research questions posed in the introduction.

First, in deploying compassionate touch, teachers recurrently used their hands to rub or stroke a child’s head/forehead, back/side, or other body parts, which had usually been indicated or reported by the children or observed by the teacher to have been the site of a physical act. Other body parts did different kinds of work in this service. For instance, as observed in other research on Japanese preschools (Hayashi & Tobin, 2015), teachers often squatted down to the child’s eye-level to achieve a face-to-face formation and arrange children’s bodies (see Figures 3 and 5). They also placed a child in their lap to achieve a ‘nesting formation’ (Ochs, Solomon & Sterponi, 2005) (Figure 5). Such touch was laminated with other kinds of touch, such as to control or assist a child.

Second, in terms of the accompanying communicative resources and practices, compassionate touch often occurred with affect words, pragmatic particles, and response cries that indexed a heightened affective stance; in addition, it often occurred with apologies and statements/questions about the child’s affective state or reaction to the event. Such resources and practices were often laminated with changes in prosody, such as high pitch voice or vowel elongation, which are also associated with heightened affective stance (Ochs, 1996). On many occasions, touch and talk were mutually informing (C. Goodwin, 2000), as they elaborated on each other in situated interaction. On other occasions, however, compassionate touch occurred with talk that was somewhat incongruent or at least less mutually elaborating, such as directives (produced with an exasperated tone of voice), reported speech (to inform what another child had said) or statements such as, ‘It can’t be helped.’ In these cases, teachers used touch and talk to do different kinds of interactional work while showing compassion (e.g., suggesting that, for her, there was no need for so much crying).

Third, compassionate touch was lodged within sequences of actions or what was referred to here as phases, including intervention, reparatory, and post-reparatory. Such touch became a resource that was deployed not only to soothe and comfort a child who was ostensibly (or imagined to be) in distress, but also in relation to larger projects of care and sociality.

Fourth, children’s actions and responses to teachers’ compassionate touch (often laminated with other kinds of touch) were variable, ranging from inviting and aligning with it to resisting it. This reveals children’s agency in episodes of compassionate touch and their abilities to co-construct this touch, even when they are primarily the recipients of it (i.e., they did not typically initiate compassionate touch to teachers).

In conclusion, particularly in institutional interactions, compassionate touch (and talk) competes with multiple demands, especially the imperative to move on to initiate, carry out, and complete other tasks and activities that may have been interrupted by an episode that commands the teacher’s attention. Within this broader context, this paper has shown how Japanese caregivers used their bodies and talk to display compassionate touch towards children in relation to events the caregivers had seen or those that were reported to them. This touch was part of the multimodal resources employed in performing childcare and encouraging young children’s sociality, which is variably organized in different institutions and societies. In the observed Japanese preschool, the analysis of compassionate touch in situations of peer conflict and accidents invites us to consider ways in which these practices might be similar in other preschools and societies as well as ways in which they might be culturally specific: Empirical research on this question remains for future research. As compassionate touch is a necessary aspect of institutional childcare, and one that has its root within the family, it surely is a tragedy when children in some situations (US border) do not receive compassionate touch from adults.

Acknowledgements

I am grateful to a number of people: the children, families, teachers and principal, Dr. Keiichi Yamazaki and his students, and the two anonymous reviewers who provided valuable comments on an earlier draft. Any misinterpretations and errors are my responsibility.

Appendix

Interlinear gloss abbreviations:

Transcription conventions: