Social Interaction

1. Introduction

One of the fundamentals in everyday family interaction is negotiation over deontic rights—that is, the right to decide what happens next (Kent, 2012a; Stevanovic, 2013). In the course of interaction, participants are in constant negotiation—either implicitly or explicitly—over the distribution of deontic authority; that is, the question of who can set the rules of what should be done (Kent, 2012a; Stevanovic & Peräkylä, 2012). During such negotiations, which can be made visible through a turn-by-turn micro analysis, children and their caregivers take a stance regarding the right of a child to do what they want and the right of an adult to tell the child what to do. In this study, we use multimodal conversation analysis to examine a situation recognizable to many parents and caregivers. In the video-recorded exchange, a five-year-old child is getting ready to go out with his younger brother and their au pair, who, together with the child’s mother, demands that the five-year-old go to the bathroom before going out. Our analytic focus is on the adult’s verbal and bodily directives, by which we mean actions that aim at making the recipient act in a way specified by the speaker (see, e.g., Couper-Kuhlen, 2014; Searle, 1976). We seek understanding of how different resources are used to form and ascribe directive actions as well as manage the deontic rights of the participants.

This paper seeks to answer the following questions:

1. What verbal and bodily elements the au pair’s directive turns consist of,
2. How the child responds to the directive turns, and
3. How participants orient to the child’s deontic autonomy.

It has been established that family members often use both speech and bodily resources to implement their interactional projects (Cekaite, 2015; Goodwin & Cekaite, 2018; Kent, 2012a). Our first aim is to describe the formation of directive actions by the au pair, drawing attention to both her verbal and bodily actions. Directives often aim at affecting the recipient’s physical actions, and they are responded to physically or multimodally (see, e.g., Kent, 2012a; Rauniomaa & Keisanen, 2012). Knowing this, our second aim is to examine action ascription of the directive turn by the recipient. By action ascription we mean the assignment of an action to a turn (see Levinson, 2012), which is revealed by the recipient’s next actions—in this case the child’s physical and verbal responses.

In this study, we will call a person’s right to determine their own actions their deontic autonomy (cf. Castelfranchi & Falcone, 2003). Stevanovic (2011, p. 15) has suggested that although parents often take deontic authority over their children, they also orient to the norm that everyone has deontic authority over their own actions. This is also seen in the majority of Goodwin and Cekaite’s (2018) examples. Based on this, our third aim is to investigate how the adults in the situation show orientation to the child’s deontic autonomy, even when telling him what to do.

The study at hand is a single case analysis that demonstrates the course of one specific (yet easily recognizable) everyday interactional situation. With this setting and a conversation analytic approach, we do not aim at a wide generalizability of the results: The verbal and bodily means used by the au pair in this situation may differ from the ones she or any other person uses in another situation. For the same reason, we cannot make assumptions about the socio-cultural reasons behind the participants’ conduct: We do not know, for instance, if the au pair acts the way she does because she is an au pair, Spanish, a non-native speaker of English and Finnish, a woman, or a certain age. Furthermore, our analysis of one extract cannot be used as evidence that the child’s actions are somehow typical for Finnish children. Nevertheless, this study contributes to the aims of this special issue by discussing multimodal action formation and ascription in a multilingual setting. It gives a detailed account of how a person can use different resources in order to accomplish their interactional project and seek mutual agreement on the participants’ near future actions. The study also contributes to the ongoing discussion on deontic rights, introduces the term deontic autonomy, and examines whether deontic autonomy is seen as a norm even in a situation where some participants (adults) have significantly higher deontic status compared to others (children).

2. Background

What is (social) action? In Couper-Kuhlen and Selting’s (2018, p. 212) terms, a social action carried out by a turn-at-talk is the “main job” of that turn. Directive turns, then, are turns whose "main job" is to make the recipient act in a way specified by the speaker (cf. Couper-Kuhlen, 2014; Searle, 1976). There are different verbal and bodily practices people use for carrying out directives (see, e.g., Ervin-Tripp, 1976; Goodwin & Cekaite, 2014; Kilani-Schoch, 2021). Some of them, such as second person imperatives, are relatively conventionalized for that purpose, although there is never a one-to-one relationship between an action and a practice (see, e.g., Couper-Kuhlen, 2014). Other turns are ascribed as directives according to their sequential position (Sorjonen et al., 2017, pp. 15–16). For instance, pointing a finger in a certain direction can be understood as a directive to move that way when the pointing happens in a context where a directive is expected.

Sometimes, actions ride on other actions (Couper-Kuhlen & Selting, 2018, p. 213; Levinson, 2012, p. 110; Niemetmaa, 2021; Raevaara, 2016, p. 159; Schegloff, 1997, p. 505). For instance, ultimatums of the format [X will happen if you don’t Y] are formulated as conditional announcements, but they seem to be rather conventionally used as directives in directive sequences. In such cases, one action becomes a practice used to accomplish another action (see also Niemetmaa, 2021), and the line between actions and practices becomes blurred (see also Schegloff, 1997). In this study, we analyze the directive turns in terms of action categories that have been established in prior studies (such as proposal), in terms of other action categories that are used as practices to form directives (such as ultimatum), and in terms of morphosyntactic and bodily resource categories (such as imperative or pointing) that are used as practices to form directives. Before getting to the analysis, however, we will introduce some of the central theories, concepts, and earlier studies relevant to the discussion. Section 2.1 is about deontic rights in child-directed directive sequences and 2.2 about directive actions and multilingual family interaction.

2.1 Deontic rights in child-directed directive sequences

Researchers of professional decision making have distinguished between deontic rights and deontic stance, as well as deontic authority and deontic status. By deontic rights we mean a person’s right to determine future actions. Deontic stance is a person’s locally and interactionally positioned expression regarding deontic rights. Through deontic stancetaking, a person can demonstrate orientation to either their own or another participant’s deontic authority (i.e., higher deontic status in a given situation). A person’s deontic status is the relation of their deontic rights to that of other participants in the interactional situation. (Stevanovic, 2013.) These notions have been used in the study of workplace interaction (Stevanovic, 2013; Stevanovic & Peräkylä, 2014), classroom interaction (Stephenson, 2020), as well as studies of psychotherapists’ (Ekberg & LeCoutour, 2020; Muntig et al., 2017) and doctors’ appointments (Weidner, 2015). So far, they have not been widely used in the study of family interaction (see, however, Henderson, 2021; Kent, 2012b).

Directive actions are socially problematic, because asking (or telling) someone to do something breaks the normative expectation that people have the right to determine their own actions (Schegloff, 2007, p. 83; Stevanovic, 2011). This is why requests are often mitigated and marked as dispreferred (Schegloff, 2007). In this study, we shall call people’s deontic authority over their own actions their deontic autonomy.

Stevanovic (2018, p. 374) introduces the term deontic concern, which refers to a person’s vulnerability regarding their right to determine actions. Stevanovic (ibid.) suggests that by the age of 4–5 years, children may already be sensitive to deontic concerns. She argues that such concerns are apparent in not only the children’s noncompliant responses to parental control, which indicate their growth as autonomous agents (see Kuczynski et al., 1990), but also in negotiations over deontic rights during compliant responses (see Kent, 2012a). Conversation analytic studies take the point of view that directive actions are “practices through which parents and children together build important events in their lives” (Goodwin, 2006, p. 538). Parents have different styles in how they formulate directives (ibid.), but in the end, directives are interactionally occasioned. For instance, when a child does not comply, parents may repeat or reiterate the directive (Goodwin, 2006). In their conversation analytic studies of family mealtimes, Craven and Potter (2010) found that children’s noncompliant responses to parental directives could lead to upgraded directives. Kent (2012b), however, found that even noncompliant responses to parental directives can be treated as legitimate if they are accounted for; for instance, by stating contingencies. Kent (2012a) also found responses that she named incipient compliance. These were verbal or bodily actions the child performed to buy time without fully complying with the parent’s directive.

A significant contribution to the study of parental directives has been made by Goodwin and Cekaite (2013, 2018). They view the family as a social form, in which children are socialized into giving and receiving directives as a part of everyday tasks (Goodwin & Cekaite, 2018, pp. 40–41). Goodwin and Cekaite (2014) show that family directive sequences can take very different trajectories depending on the timing, embodiment, and linguistic formulation of the parental directive. They demonstrate this with an example of a parent issuing a question-form request for the child to go the bathroom (Goodwin & Cekaite, 2018, pp. 50–51). This request is followed by a negotiation and the parent giving in to the child’s will to postpone going to the bathroom (ibid.). All in all, the authors show that children are independent agents who can show considerable resistance to parental directives (see also Aronsson & Cekaite, 2011).

Henderson (2021) discusses a family interaction which in many ways resembles the one in this paper. Hers is a case study of a mother trying to get her six-year-old son, who is on the autism spectrum, to go to the bathroom. It is noteworthy that, in Henderson’s study (ibid.), the mother displays orientation to the child’s deontic autonomy in all of her turns, even after considerable verbal and physical resistance from the child. The mother also makes this orientation explicit by saying it’s up to you (Henderson, 2021, pp. 179–180). Despite the mother’s constant orientation to the child’s deontic autonomy and the child’s initial resistance, the child finally complies (Henderson, 2021, pp. 183–184). Thus, as Henderson (2021, p. 185) concludes, the child “is doing what he is told but not because he was told but because he chose to”.

2.2 Directive actions and multilingual family interaction

The participants in the data under examination here use language in a way that could be called translanguaging—“a practice that involves dynamic and functionally integrated use of different languages and language varieties” (see Li, 2018, p. 15). Bi- and multilingual parents and other caretakers have the advantage of using a larger set of linguistic resources as well as the possibility to rely on codeswitching (switching between languages or linguistic varieties) as a contextualization cue (see, e.g., Auer, 1984; Gumperz, 1982). For instance, in directive sequences, bi- and multilingual caretakers can reiterate their directive in another language if the child does not respond (Gal, 1979, p. 112; Gumperz, 1982, pp. 78–79). When there is more than one caretaker present, directives can be uttered by different people in different languages (Hiratsuka & Pennycook, 2020). Codeswitching can be used to indicate a change of footing when a caretaker moves from one action to a directive action (e.g., Shin, 2010). In these cases, the codeswitching acts as a contextualization cue which makes the directive action more salient.

Translanguaging practices vary between speakers, situations, and communities. Although codeswitching can act as a pragmatically meaningful contextualization cue as described above, not all instances of switching between different named languages do so. Many times, bilingual speakers make language choices rather sporadically, uttering a word in whichever language comes to mind first (see, e.g., Ervin-Tripp & Reyes, 2005). The family studied here has created a familylect (Søndergaard, 1991) that involves translanguaging between English, Finnish, and occasional Spanish elements (see Section 3). Codeswitching in the directive utterances studied here may or may not be locally significant, and while it is often impossible for an outside researcher to know whether codeswitching played a role in how recipients interpreted a turn-in-interaction, we will refer to the participants’ language choices where it seems relevant for the analysis.

3. Data and methodology²

The video recording we analyze in this study was made in the 2010s of a family residing in northern Finland. The family has seven children, two of whom are present in the recording: two-year-old Eetu and five-year-old Olli, who is the focus of this study. The family hosted a 19-year-old Spanish au pair who had been staying with them for seven months at the time of recording. The family video recorded their interactions sporadically during a period of five months, with an emphasis on the au pair’s interactions with the two youngest children. The family members mainly interact in Finnish, which is the first language of the children, and English, which they started using after the au pair arrived. The au pair studied and learnt Finnish during her stay. Codeswitching is extremely frequent in the interactions between the au pair and children. Even if we look at intrasentential code-switching only, 15% of the au pair’s and 14% of Olli’s multiword turns in the collection contain both English and Finnish³. Frequent codeswitching is a part of the familylect that formed during the au pair’s stay. Codeswitching seems to be a neutral choice for the family members, and it does not bear as much pragmatic significance as codeswitching in situations where monolingualism is the norm.

For this study, we chose a clip showing a prolonged directive sequence which includes several directives within a single interactional project of getting the child to go to the bathroom. A clip taken from a single interactional project enables us to investigate the variety of ways in which the au pair displays orientation towards the child’s deontic autonomy without having to address the possible effect different projects might have on the au pair’s conduct. The data clip includes the first three minutes of a recording titled Pukeminen1, which was recorded in the family home during a situation where the youngest children were getting ready to go out with the au pair. The data was transcribed according to the Mondada conventions for multimodal transcriptions (Mondada, 2016). Translations are provided for turns that include Finnish words.

The data was analyzed using conversation analysis (Sacks, 1992). Conversation analysis is a data-driven method that allows us to examine the sequential unfolding of a stretch of conversation, showing how each turn is motivated by prior turns while, at the same time, creating a context for following actions. When uttering turns in interaction, speakers form social actions. At the same time, recipients ascribe actions to the speaker’s turn while planning their response accordingly. Responsive turns—whether verbal on embodied—are, for the analyst, proof of how the recipient interpreted the turn in question. (Levinson, 2012.) For instance, if a caretaker utters you should go to the bathroom, and the child responds by getting up and going to the bathroom, we can take this reaction as proof of him having interpreted the caretaker’s turn as a directive (and not, for instance, an announcement). The same kind of next-turn proof procedure (Sacks et al., 1974, p. 728) applies for determining people’s deontic concerns: If, for instance, a child does not comply with a caretaker’s directive but, instead, protests against it, it can be taken as proof that he considered the matter as something he himself has deontic authority over.

4. The multimodal organization of going to the bathroom

The question we asked in this study was what linguistic and bodily resources the family au pair uses when asking five-year-old Olli to go to the bathroom. We also wanted to know how Olli reacts to the au pair’s directive turns. In this section, we will examine how the three-minute episode unfolds, paying special attention to the participants’ orientation to the child’s deontic autonomy. The whole episode can be seen on the attached video clip, while the transcript has been divided into nine extracts.

The episode under study here begins when five-year-old Olli and two-year-old Eetu are getting ready to go out with their au pair Maria. It is cold outside, and some of the children’s winter clothing is lying on the floor ready for them to wear. The boys’ mother is also present. In Extract (1), Olli and Mom are finishing a negotiation about which overalls to wear (lines 01–03). In this extract, the mother and children speak Finnish to each other, while the exchange between the au pair and the mother is in English. In Extract (1), line 08, we can see the mother addressing the au pair regarding the child’s need to go to the bathroom. This extract shows how the adults establish an activity contract (Aronsson & Cekaite, 2011, p. 3) that sets the bathroom episode going.

Studies of middle-class Swedish family interactions show that parents’ directive sequences often include or are preceded by establishing an activity contract (Aronsson & Cekaite, 2011, p. 3). Activity contracts are “intergenerational agreements to comply with the directive” (Goodwin & Cekaite, 2013, p. 124). A typical format is “declarative statements in a deontic modality encoding obligation (for instance, about what needs to be accomplished within a particular time frame, i.e., ‘You gotta be in bed in twenty minutes.’” (ibid.) In the case at hand, an activity contract regarding Olli’s going to the bathroom is established between two adults—Olli’s mother and the au pair.

During the pause in line 04, Olli is sitting on the floor alone while everyone else is in other rooms. Two-year-old Eetu enters the room, followed by Mom, whose attention Eetu is trying to get (lines 05, 07, 09). Mom walks past Olli, meets Maria, and announces that Olli has to go to the bathroom first (line 08). In her turn, Mom, who is facing Maria, uses a third person pronoun he and points towards Olli. This, and the language choice of the turn, indicates that the turn is directed at the au pair. Maria immediately responds to the turn with an okay. Mom’s announcement shows that she takes deontic authority over Olli and treats Maria as an involved party in the matter. Maria’s response can be treated as a sealing of an activity contract with the mother. What comes next, Extract (2), shows that Maria has taken Mom’s turn as a request for her to execute the “bathroom project”. Olli, however, has not yet sealed the contract. At this point Olli has not done anything but glance at Maria (Extract (1), line 10), which cannot be understood as an agreement to do what Mom wanted. Extract (2), line 01 shows how, immediately after this, Maria proposes going to the bathroom and Olli refuses. In Extract (2), the au pair’s orientation to Olli’s deontic autonomy can be seen in the formulation of her turns. Olli relies on the norm of deontic autonomy when refusing to do as Maria proposes.

In Extract (2), line 01, Maria utters a directive using the Finnish verb mennään, which is morphologically a “unipersonal” (Helasvuo & Vilkuna, 2008) passive that is used here as a first-person plural equivalent (cf. “let’s go” in English). Such a turn can be understood as a proposal (Couper-Kuhlen, 2014) indicating that the proposed activity is a joint effort of the speaker and recipient. In such use, the Finnish passive can be meant either as inclusive or exclusive of the speaker (VISK, § 1325, § 1654); that is, the same format would be used regardless of whether Maria actually intends to go to the bathroom with Olli or not. In fact, the first-person plural form is often used when the speaker strongly encourages someone to do something (Goodwin & Cekaite, 2018, p. 40). Nevertheless, the action is shared in the sense that it is highly relevant for the speaker as well (VISK, § 1325). Since the speaker has deontic authority over her own actions, a proposal is also less of a breach of the norm of deontic autonomy than a second person request would be.

Maria’s directive is multimodal: While uttering the proposal, she also points at the bathroom door. Olli turns to look at the direction of the pointing, then turns away (line 02), places his hand on his hip, and shakes his head. This is interpreted as a refusal by Maria (see also Kendon, 2002). By refusing to go to the bathroom, Olli orients to his deontic autonomy: He is the one who decides what he will do. Maria responds in line 03 by giving an ultimatum: There will be no going to the park if Olli does not go to the bathroom. Ultimatums, which are also called conditional threats (Church & Hester, 2012), are announcements of negative outcomes should the recipient not comply (cf. Llewellyn & Butler, 2011, p. 48; Sikveland & Stokoe, 2020, p. 334). While uttering her ultimatum, Maria spreads her arms, a gesture that is associated with resignation, helplessness, and the meaning “I don’t know” (see, e.g., Eskildsen & Cadierno, 2020, p. 536; Sielski, 1979, p. 239). Her doing so could possibly indicate that Olli’s not going to the bathroom would be something Maria would not know how to handle—other than cancel the whole trip. With this turn, Maria accepts Olli’s deontic autonomy but shows him what the outcome of it would be. The ultimatum is thus based on a simple logic: You get to decide what happens next, but the outcome will not be something you would like.

After this, Olli turns to his mother—the person who first announced that he should go to the bathroom—and tries to negotiate his way out of the situation (Extract (3)). In Extract (3), we see Olli orienting to the family norm that children do not have full deontic autonomy but can negotiate matters concerning themselves. The mother’s turns in this extract do not show any orientation to Olli’s deontic autonomy.

In Extract (3), Olli tries to negotiate about the need to go to the bathroom. In so doing, he takes the stance that his going to the bathroom is something that he and his mother should decide together. He orients to the mother as a higher deontic authority than the au pair. In negotiating, Olli does not position himself as the sole deontic authority in the matter but does not accept the adults’ authority either. This shows his stance that while he may not have full deontic autonomy, he can negotiate matters that concern him. Olli poses the argument that he does not need to go to the bathroom (line 05) and that he could go outside (line 10).

Mom firmly rejects both Olli’s arguments (lines 06 and 13). Her declarative sä käyt nyt vessassa (line 13) conveys full deontic authority in the matter and leaves no room for negotiation (cf. Goodwin & Cekaite, 2018, pp. 47–48). The turn does not contain any mitigating elements (cf. Stevanovic, 2011, p. 15) that would indicate Mom’s orientation towards Olli’s deontic autonomy. Mom then tells Olli to listen to (lit. “obey”) the au pair (line 17). The clitic -pA in the imperative tottele+pa is often used in parental directives (VISK, § 1672) and it marks the requested action as something that should be non-problematic and easy to comply with (VISK, § 835, § 1672). According to Stevanovic (2017), -pa is commonly used in imperatives immediately after the recipient’s failures that need to be remedied before a joint activity can proceed. They also demonstrate the speaker’s deontic authority (ibid.). With her turn, Mom lets Olli know that Maria is the deontic authority in the situation. Maria accepts this role as she takes the next turn (Extract (4)). Extract (4) shows how the au pair engages in a negotiation with Olli, showing her stance that Olli has deontic autonomy in the matter, but might be persuaded to make the desired decision if given sufficiently convincing arguments.

Extract (4) shows how Maria, in line 01, produces a turn that has no finite verb, but is fully understandable as a directive (cf. Marttila, 2020, pp. 35–39; VISK, § 1462). Ensin vessaan kulta is a free NP (see, e.g., Helasvuo, 2001) accompanied by a temporal adverbial (VISK, § 650). The turn states the time ensin “first”, agent kulta “darling”, and the place vessaan “to the bathroom” of the suggested activity. The illative case of the word vessaan indicates a movement towards a place; it implies a change (VISK, § 1235) and is therefore well suited for a directive turn that aims at changing the course of actions. Maria utters the turn with hands placed on her hips, a posture that, according to Sielski (1979), signals readiness.

Olli continues negotiating with Maria (line 03), arguing that he could do it outside, but Maria rejects this idea (line 05). She also presents an argument of her own (lines 08–11), that animals might drink Olli’s urine if he went to the bathroom outside. In this family, animals are often regarded as parties in discussions and negotiations as if they actually had a role in the family life. Maria’s argument is based on this habit, and it creates a shared narrative space that is designed to appeal to Olli’s love for animals and sense of humor. Through her argument in lines 08–11, Maria aligns (see Stivers, 2008) with Olli by joining the negotiation activity he started. Maria’s reaction is thus different from Mom’s, who, in Extract (3), did not join in the negotiation with Olli. The stance Maria still takes is that Olli has deontic autonomy, but that he could be persuaded into making the desired decision himself by providing him with convincing arguments. Olli does not respond. Maria’s smile and raised eyebrows in line 12 can be understood as seeking affiliation and evaluating her own turn as non-serious (see Tabacaru & Lemmens, 2014), but Olli does not show affiliative emotive reactions to the non-verbal cues either. In Extract (5), Maria continues in the non-serious mode, using singsong prosody, raised eyebrows, and smiling to mitigate her directive.

In Extract (5), which is a direct continuation of Extract (4), Maria produces a directive (line 01), this time in a clausal form, so please come on go to vessa. Gal (1979, p. 112) and Gumperz (1982, pp. 78–79) were probably the first to report that bilingual parents often reiterate their turns in the other language if a child does not respond. This is what the au pair does as well: She and the children share both Finnish and English resources, and although neither is the au pair’s native language, she can resort to them both. Her turn initial so links the utterance to the preceding turns, formulating the turn as a conclusion of Maria’s arguments. Please and come on in the turn are lexical indications of the requesting action, and the singsong prosody and smiling give the imperative turn a non-threatening tone that is in line with the non-serious mode Maria started in her preceding turns (cf. Goodwin, 2006, pp. 522–523, 536–537; Stevanovic, 2011, p. 15). The singsong prosody can be seen as a mitigation of the problematic action and an indication that Maria acknowledges Olli’s deontic autonomy.

According to Goodwin and Cekaite (2018, p. 43), immobility and inaction is one of the possible responses children give to directives. In Extract (5), Olli remains lying on the floor, only slightly moving his leg and hand in line 02. He responds to Maria’s directive with a growl (line 03). Maria raises her eyebrows and recycles a part of her turn come on, adding the child’s first name (line 04). Lappalainen and Mononen (2017) have found that, in Finnish conversations, one of the contexts in which first names are used are problematic sequential trajectories (see also Butler et al., 2011). Olli does not take the next turn, and Maria continues with a new strategy (Extract (6)). In Extract (6), the au pair “bribes” Olli with the promise of taking a ball to the park on the condition that Olli goes to the bathroom first, and Olli pushes the limits of his deontic rights by negotiating for more.

In Extract (6), line 01, Maria proposes that they take a ball with them to the park. Olli proposes a volleyball (line 07), which Maria agrees to (line 09). At this point we can see Olli moving closer to Maria and squatting down, facing her. This can be seen as an aligning and affiliating (see Stivers, 2008) action: At this point in the conversation, Olli and Maria have found something they can agree on. Olli has gotten the chance to actually decide on something, and he pushes his deontic boundaries a bit further by proposing they take two balls.

Maria agrees, but poses a condition for taking two balls: Olli has to go to the bathroom first. This happens in lines 12, 14, and 17, where Maria utters Okay we take kaksi, mutta ensin vessaan. The condition mutta ensin vessaan is a free NP, similar to the one in Extract (4), line 01. Maria’s turn shows that she acknowledges Olli’s deontic authority regarding taking the balls—but only conditionally. While producing the directive turn, Maria taps on the bathroom door. With this turn Maria takes the stance that she has deontic authority over Olli regarding taking the balls and that her agreement to do so is conditional on Olli going to the bathroom. This narrows Olli’s deontic autonomy, but still leaves him the chance to decide what to do.

Olli, who had been in a squatting position, now sits back down and continues negotiating for more balls (lines 16 and 19). In Olli’s physical reaction we can see disaffiliation and non-compliance to what Maria is requesting, but at the same time, his verbal turn suggests an openness for negotiation. Maria agrees to what Olli is proposing by nodding (line 20). In Extract (7), lines 02–05, Maria poses a condition for her agreement and uses verbal and grammatical means to mitigate her turn, thus displaying orientation to Olli’s deontic autonomy, and Olli shows incipient compliance.

In Extract (7), lines 02–05, Maria verbalizes her agreement to Olli’s proposal of taking three balls and reiterates her own terms to the contract with a declarative clause first we go to vessa. The clause is a slightly modified bilingual reiteration of the au pair’s Finnish turn in Extract (6), lines 12, 14, and 17. Statements regarding future actions are often used as directives (Goodwin, 2006; Stevanovic, 2011). Compared to Mom’s statement in Extract (3) (sä käyt nyt vessassa “you are going to the bathroom now”), Maria’s turn in Extract (7), lines 02–05 contains mitigating elements, namely the use of the first-person plural pronoun we and the temporal adverbial first which indicates that going to the bathroom is (just) a condition for taking more balls. Like in Extract (6), such a conditional formatting of the directive leaves Olli the right to decide for himself and, at the same time, offers him an award for doing what Maria is asking. While uttering the directive, Maria taps on the bathroom door like she did in Extract (6).

Olli starts complying by standing up already before Maria has uttered her directive, and by walking towards the bathroom door. At this point, in line 06, Maria looks at Mom, smiling, thus seeking affiliation after seemingly having completed the task Mom had given her (that is, getting Olli to go the bathroom). The mother’s reaction is not visible on the recording. Olli utters huh huh, which is used in Finnish as a sign of relief after strenuous or otherwise challenging efforts (see Pehkonen, 2020). This, too, suggests that the negotiation is now finished. Right after this, however, Olli continues negotiating taking four balls to the park (line 08). Maria utters minimal agreement tokens (lines 10 and 14) while Olli walks towards the door and even opens it (line 12). As we move on to look at Extract (8), it becomes clear that Olli’s actions were only what Kent (2012a) calls incipient compliance: The child is seemingly complying to the adult’s directive, but is, in fact, only buying time. In Extract (8), lines 47–49 and 59, Maria utters imperatives which are, for the first time in the whole episode, produced almost without mitigating elements.

In Extract (8), we see how Olli walks away from the bathroom, to the closet, and starts a lengthy discussion with Maria about which balls to take to the park. Maria takes out the balls, drops them, and Olli starts kicking them. After scolding him, Maria utters a directive in the imperative mood (line 47–49) And Olli, go ensin vessa. Butler et al. (2011) have found that address terms are used in turn beginnings when the turn is not fitted to prior talk: when it opens a new action sequence or is disaffiliative and disaligned with the prior speaker. Maria’s turn in Extract (8) line 47 is both disaligned and disaffiliative to Olli and opens up a new action sequence. Studies in family interaction have shown that children’s non-compliance can lead to upgraded directives (see, e.g., Craven & Potter, 2010). Using the imperative mood with no other mitigating factors but the word ensin can be seen as such. On the other hand, the use of the imperative mood can be motivated by the fact that Olli has, at this point, already shown (incipient) compliance by going once to the bathroom door. Olli’s incipient compliance, which we saw in Extract (7), may have been taken by Maria as an indication of Olli sealing the activity contract regarding going to the bathroom. In that case, Maria would be entitled to demand compliance.

Olli starts complying by walking towards the bathroom door (lines 50, 52). During this, Maria says come on twice. The come ons appear to be reactions to the fact that, instead of fully orienting towards going to the bathroom, Olli first kicks one of the balls (line 49) on his way to the bathroom and then faces Maria (and not the bathroom) with his body twisted when opening the bathroom door (line 52). Maria utters another imperative turn Mutta ensin go in line 59. This turn does not include a word to indicate where Olli should go, but this is quite clear from Maria’s pointing to the bathroom door—and from the fact that this is already the 11th time Maria tells Olli to go to the bathroom. Extract (9) shows Maria’s last directive turns before Olli finally complies: an ultimatum in line 01 and Olli’s first name in line 05.

In Extract (9), line 01, Olli closes the bathroom door while Maria is uttering an ultimatum that she will not take more balls unless Olli goes to the bathroom. After this, Olli opens the door, looks at something on the floor (possibly balls rolling on it) and makes a whistling/puffing sound (line 03). This can be seen as Olli’s final demonstration of his deontic autonomy. Finally, in line 05, Maria utters just the first name of the child, using a “scolding” prosody and lengthening of the final vowel. Olli then shuts the door and proceeds with going to the bathroom.

5. Discussion

The multimodal analysis conducted in this paper shows several ways a caretaker can combine verbal and bodily resources in formulating child-directed directive turns. When urging the five-year-old to go to the bathroom, the au pair uses a variety of practices to form directives: a mennään-proposal, two ultimatums, a free NP, negotiation, a bribe, conditions, imperatives, three come ons and the child’s first name twice. Many of the directive turns are accompanied by bodily actions such as smiling, raised eyebrows, pointing at or tapping the bathroom door, or assuming different body postures. Although the participants use both English and Finnish resources in many of their turns, they do not seem to orient very much to the switching of languages. This kind of translanguaging is a typical trait of their familylect that could be studied in more detail in the future.

What makes our study significantly different from earlier studies on family directives is the fact that there is an au pair in the family. Child-directed directive sequences differ when there is one adult present versus two, and if the second adult is a babysitter, the deontic status of the adults is unequal, which may have an effect on the formulation of the directives (cf. Gaskins & Frick, 2021). In the interactional episode examined in this study, both the child’s mother and au pair were present. In the beginning of the extract the child’s mother announces that the child needs to go to the bathroom, and the au pair interprets this as a request for her to implement the plan. This “activity contract” that the two adults make obliges the au pair to follow through with the directive trajectory and fortifies her deontic rights, as there is now another adult who grants her deontic authority over the child. Regardless of this, the child’s resistance shows that he is aware of deontic concerns and orients strongly towards his deontic autonomy.

We can see in the data that the au pair orients to the child’s deontic autonomy. This orientation is seen in all the au pair’s directives except the very last ones. The au pair mitigates her turns, for instance, by using a first-person plural equivalent and indications of a non-serious mode, such as smiling, raised eyebrows, and singsong prosody. She also formulates an ultimatum expressing the outcomes of what would happen if the child did not comply, but, at the same time, leaving the right to decide to the child. The au pair also engages in a negotiation with the child, which shows her stance that Olli has the right to make the decision but could be persuaded with the right arguments. Furthermore, Maria ties Olli’s going to the bathroom with her decision to take a ball to the park (a “bribe”), thus creating a condition that should appeal to Olli.

The analysis shows that although the au pair and child disagreed on whether the child should go to the bathroom, they both orient to the same norms of interaction and the norm of deontic autonomy more specifically. Only later in the episode when the child displays incipient compliance (see Kent, 2012a), but does not follow through with going to the bathroom and starts to play with a ball instead, does the au pair start mitigating her directives less and using imperatives instead. The mother’s orientation to the child’s deontic autonomy is quite different in the few turns she takes. The mother makes statements regarding the child’s future actions, using the indicative mood and no mitigating elements (“You are going to the bathroom now”), thus demonstrating strong deontic authority in the situation and not leaving room for negotiation. We know from the bachelor’s thesis of Lehtola (2019) that, soon following the extract studied here, the au pair continues to use a variety of directive turns with and without mitigating elements in a lengthy attempt to make Olli wash and dry his hands. We can only speculate how the situation might have turned out had there been no mitigations in the au pair’s turns in the first place—perhaps the mother’s “firmer” strategy was aimed at resolving the situation quickly while the au pair’s actions led to a prolongation of the situation. This is one of the questions that could be addressed in future studies.

Further studies are needed to find out how deontic autonomy is manifested by children and adults in everyday settings, and how co-participants orient to a person’s deontic autonomy. Our study gives reason to suspect that there are instances in which deontic autonomy is not oriented towards as a norm—such as the mother’s turns in this exchange. We do not know, however, whether and how adults’ orientations towards a child’s deontic autonomy can change in the course of interaction or whether they depend on what the activity in question is.

We also know from earlier studies on child-directed directives (Aronsson & Cekaite, 2011; Goodwin, 2006; Goodwin & Cekaite, 2018; Henderson, 2021; Kent, 2012b) that primary caregivers and first language speakers orient to children’s deontic autonomy, as does the au pair in our study. Nevertheless, it would be worthwhile to conduct further research on the nuances present in the orientation of primary vs. secondary caregivers and native vs. non-native speakers to children’s deontic autonomy. All in all, the social patterns and regularities governing deontic autonomy are yet to be fully discovered.

Further conversation analytic studies are needed to investigate the distribution and manifestation of deontic rights in family interactions in which there are au pairs or other participants who are not the children’s primary caregivers. Some researchers have also suggested that there are cultural differences regarding whether or how much children’s deontic autonomy is a norm. Demuth’s (2013) study shows that middle-class mothers from Münster, Germany use more mitigated directives than mothers from a farming community in Kikaikelaki, Cameroon. Goodwin and Cekaite (2018) studied Swedish and U.S. family directives in a conversation analytic framework, and although their approach was not comparative, they found some practices to occur mainly in one or the other country. The current study was a single-case study and as such, could not address the question of the adults’ roles as primary vs. secondary caregiver or cultural specificity in the participants’ conduct.

What the data demonstrated, though, was that in this multicultural setting, the participants use a large variety of linguistic (both English and Finnish) and bodily resources to form and ascribe directive actions—and that they use these different resources to manage the deontic nuances of the situation.

Transcription conventions

Data source

University of Oulu Kikosa Collection https://etsin.fairdata.fi/dataset/0891e1ce-5365-49fa-b9a0-aa5e20cab0b5 (January 3, 2021)

¹ The study was conducted as a part of the Academy of Finland project Linguistic and Bodily Involvement in Multicultural Interactions. We would like to thank Kaarina Hippi for her comments on an earlier version of this paper and Ivana Leinonen, Samu Pehkonen, and Anna Vatanen for viewing and discussing the data, as well as the anonymous reviewers and editors of this special issue of Social Interaction for their invaluable help in finalizing this paper. ↩

² The “Au pair” dataset is a part of the University of Oulu Kikosa collection and consists of 54 recordings made with this family. Earlier studies using the au pair dataset include Gaskins et al. (2021), Poutiainen (2019), Quick et al. (2021) and Tolonen (2020). ↩

³ We transcribed 4 h 13 min of the family’s conversations recorded during a period of five months. The transcriptions contain 3,485 turns (intonation units) by the au pair, 2,417 of which could be coded for language (the rest were coded ambiguous, because they contained either doubts in transcription or were language neutral). 1,762 of the unambiguous turns contained two or more words. Out of these multiword turns, 65 % were monolingual English, 18 % monolingual Finnish, 15 % English-Finnish mixed, and 2 % monolingual Spanish. Single word turns contained only single occasions of codeswitching. Out of Olli’s 466 unambiguous multiword turns, 45 % were monolingual English, 41 % monolingual Finnish, and 14 % English-Finnish mixed. ↩