Social Interaction

1. Introduction

Teasing is something we all have experience of: We know how it feels when someone is pulling our leg and we sense the moments that call for a little prank. While participants in interaction recognize teases and know how to react to them in an expected and acceptable way, defining teasing in exact terms is a challenging task. The definitions found in research literature point to the strategically ambiguous quality of teasing: It is a social action that is achieved by performing humorous actions in a po-faced manner (Drew, 1987; Haugh, 2017). The po-faced manner of delivery covers the jocularity of teasing so that it is not recognized right away. Because of this strategic ambiguity, teasing is an intriguing social action in terms of action ascription (Levinson, 2013): How and based on what cues are teases recognized in interaction?

This paper focuses on participants’ embodied work in teasing. We aim to show how participants’ embodied conduct works in the service of action ascription in teasing environments. We analyze two extracts in which the teasing activity is organized as an other-initiated repair sequence. In these sequences, the other-initiations of repair (OIRs) are used as vehicles for carrying out the teasing actions (see also Schegloff, 1997, 2007; Rossi, 2018). In this sense, the focal OIRs are examples of double-barreled (Schegloff, 2007, pp. 73–78) or layered (Rossi, 2018) social actions. Our analysis shows that the double-barreled OIRs are designed as multimodal action packages in which the verbal part is delivered in a serious manner and combined with embodied conduct that typically characterizes repair environments, such as head turns and tilts (Seo & Koshik, 2010; Mortensen, 2016), forward-leaning (e.g., Pajo & Klippi, 2013), and gestural holds (e.g., Floyd et al., 2016), but produced in a somewhat exaggerated or pretended manner. We will show that the subtle embodied exaggeration and performed character are the key elements in making the tease recognizable for the recipients.

We begin the paper by reviewing previous conversation analytic research on teasing and the role of repair in the accomplishment of teasing. The analytic section involves two extended teasing sequences that are examined in detail to show how embodied conduct works in the service of action ascription.

2. Teasing as a strategically ambiguous social action

Teasing has been defined as ”interactional provocation accompanied by playful markers that together comment on something of relevance to the target of the tease” (Keltner et al., 2001, p. 229). Teasing is thus a dualistic social action, containing both playful and serious elements. As social actions, teases are known to have specific sequential characteristics. First, they occupy a second position after a turn by the person who is teased and draw on these prior turns. In his article on po-faced receipts of teases Drew, (1987) suggests that interactants usually get teased if they have been “overdoing” something, for example complaining excessively. Speech errors provide other bases for teasing (see Glenn, 2003, p. 125; Norrick, 1993, p. 80). In drawing on their prior turns, teases often exaggerate or highlight something said in them. At the same time, they also make relevant some aspects of the teased participant’s speaker identity (Drew, 1987). In the extracts analyzed in this paper, minor linguistic mistakes in the second language user’s talk are exploited for teasing. This way, the speaker identity of a L2 speaker is occasioned in the tease. By initiating repair on the linguistic mistakes, the L1 participants put the second language speaker in a situation in which she is asked to repeat the erroneously formulated or pronounced word several times before she recognizes that the repair initiations are not done for the sake of restoring mutual understanding but rather in jest.

The second sequential characteristic of teases is that they are designed to involve subtle elements that make their humorous and playful aspects recognizable. Such elements may involve lexical exaggeration (e.g., Drew, 1987; Glenn, 2003; Haugh, 2016; Kontio, 2017), facial expressions (Looney & Kim, 2019), or prosodic cues that serve as off-record markers that help the recipient of the tease to recognize the playful and humorous aspects of it (Keltner et al., 2001, pp. 234–236; Mulkay, 1988, p. 47; Haakana,1995). In the analysis to follow, we will show that it is through embodied conduct typically drawn on to initiate repair that the participants in our data design their turns to be recognizable as teases. The embodied conduct, however, is slightly exaggerated or performed in a pretended way. This is connected to the strategically ambiguous quality of teasing as social action: The teases should not be too readily recognizable. However, at the same time, it is also important that the teases are in fact recognized as teases at some point because if their humorous side is not recognized, the actions that carry them may seem offensive.

The inherent dualism of teasing arises from the fact that the person who is teased usually recognises that the category that is attributed to her in the tease is in some respect true to life. Drew (1987) shows that this is also the reason why the teased persons are motivated to respond to the tease in a serious manner. Accordingly, the third sequential characteristic of teases is that the receipts of teases display recognition of the humorous side of teases while also orienting to the tease seriously and trying to set the record straight (Drew, 1987).

In the extracts analyzed here, OIRs are used as vehicles for accomplishing the teases. In the analysis of repair organization, repair initiation is usually distinguished from repair solution and the terms self and other are used to refer to the speakers involved. Self is the party whose actions are treated as the source or trouble, and other is the recipient of the treated-as-troublesome-talk (see, e.g., Schegloff, Jefferson & Sacks, 1977; Dingemanse, Blythe & Dirksmeyer, 2014; Kendrick, 2015). Previous research on repair phenomena has already shown how OIRs can be used to accomplish not only repair but also other actions at the same time (Schegloff, 1997, 2007). These actions can be either positively valenced, such as displays of surprise (e.g., Wilkinson & Kitzinger, 2006), or negatively valenced, such as pre-disagreement or challenges of previous turns (e.g., Wu, 2006). Our analysis adds to this line of research by showing how the participants initiate repair to tease in second language interactions.

This paper scrutinizes how teases become recognizable to the recipient by analyzing the participants’ embodied conduct performed together with the verbal OIRs. We will show that embodied conduct typically associated with initiating repair is slightly exaggerated or even ‘performed’ in the analyzed extracts. For example, the participants in our data perform exaggerated holds by which they display their orientation to the not-yet-resolved status of the repair sequence (see also Kamunen, 2019). By “hold” we refer to what Floyd et al. (2016, p. 183) defined as “any meaningful maintenance of a bodily configuration”. In other words, the participants freeze their ongoing bodily actions — e.g., hand gesture, gaze, facial expression, posture — to signal that there is something that needs to be solved before they can continue their actions.

3. Data and methods

Our analysis draws on a dataset of everyday L2 dinner and coffee table interactions in Finnish. The data was originally collected to analyze repair practices in L2 everyday interactions (see Lilja, 2010, 2014). As part of this analysis, the sequences analyzed in this paper were also identified. They are part of a larger collection of other-initiated repair sequences (c. 300 instances) in which they form a subcollection of cases where the OIRs are used as vehicles for accomplishing other actions (see also Schegloff, 1997).

Teasing through repair practices is not the only way of doing teasing in this data set. Instead, teasing is a recurrent activity that can be accomplished in many ways. In this paper, the focus is on teasing activity that is implemented through repair practices. The analytic focus is on the subtle embodied signals the participants draw on in action formation and ascription. The focal sequences have been transcribed according to the conventions for multimodal CA transcripts developed by Mondada (2018, n.d.). The conventions pay special attention to the detailed timing of actions. Because verbal language unfolds “linearly” in time, it forms the basic line of written transcripts. Participants’ embodied conduct is indicated in relation to that in a line below the speech line. The speaker’s embodied conduct is always presented first and the recipients’ possibly relevant embodied conduct in the line below that.

We present the sequences also as graphic transcripts (Laurier, 2014) based on the original videos to secure the anonymity of the participants. Their methodological advantage is that they highlight the progression of action, visualized in the drawings, rather than the progression of talk. The disadvantage is that the sequentiality is lost; this transcription standard is most apt for interactions in which talk is scarce or absent and therefore not the main organizational principle for the achievement of social order. We therefore use both Mondada’s multimodal transcription to capture and represent the sequentiality of the multimodal interaction and an application of Laurier’s (2014) graphic transcripts to further highlight the local material ecology of action.

4. Analysis: Initiating repair to tease in everyday L2 talk

The focus of our analysis is on the participants’ embodied conduct that plays into action ascription. We will show how the participants design teases by carefully drawing on slightly exaggerated enactments of bodily conduct that typically characterizes repair environments, including head-turns (see Seo & Koshik, 2010), forward-leaning positions (see Rasmussen, 2014, Mortensen, 2016), and gestural holds (Floyd et al, 2016).

We will go through two examples to show how teasing is organized as extended other-initiated repair sequences. In these sequences, first language speakers produce several consecutive OIRs that target a minor mistake in the L2 speaker’s speech. The OIRs are designed to invite the L2 speaker to repeat the mistake. This way the OIRs work towards making the minor linguistic mistake salient in interaction. After the completion of the repair sequence, the L2 speaker shows orientation to the teasing activity by laughing and by either explicitly commenting on the teasing and counter-teasing (extract 2) or by returning to the main line of the conversation (extract 1).

In both of the following extracts, there are three participants: Ada, Simo and Siiri. Ada is an exchange student living in a family in Finland for a year. Simo is the father of the host family and Siiri is his daughter. The participants are eating dinner and discussing at the same time. Because teasing is a sequentially organized activity, we present the sequences as whole even if they are rather lengthy.

4.1 Doing teasing by exaggerating embodied actions

The sequence in extract 1 begins as Ada asks whether the meat that they are eating is turkey or chicken. Ada’s question is understandable but involves two minor mistakes: First, the pronunciation of the word kalkkuna (turkey) is non-standard, since the word-internal k-sound is too short. Second, the conjunction or should be vai and not tai in alternative questions. It is possible that in this context Ada is not asking whether the meat they are eating is either turkey or chicken but rather whether it is turkey or chicken. The minor mistake related to the length of the k-sound is used as a resource for teasing by the L1 participants, Simo and Siiri.

We focus our analysis on the OIRs that are used as vehicles for the teases and, in particular, the embodied conduct connected to them. In this sequence, the OIRs are produced by Siiri and accompanied by a pointing gesture towards the trouble source speaker, Ada, (l. 4), a forward lean combined with turning of the ear to Ada’s direction (l. 8) and turning an open palm toward the participants as if asking for an answer. Siiri uses her embodied conduct quite clearly to perform non-hearing in a situation in which a trouble in hearing is highly unlikely. The pretended embodied conduct thus works in the service of teasing. For easier reading, we have marked with arrows the OIRs through which also the teasing activity is accomplished.

Extract 1. Kalk(k)una

Simo’s first answer to Ada’s questions orients to the ambivalence in the design of Ada’s question. By answering just yes, he treats the question as a polar one (l. 3). Because of this, the answer could already be heard as humorous and teasing. However, it is not treated in any specific way by the other participants, since right after it, also Siiri reacts to Ada’s turn. She produces a turn that orients back to Ada’s previous turn and seeks clarification instead of offering an answer to her question. Siiri’s turn is thus recognizable as an OIR. She begins the turn with the question word “what” in nominative case. In Ada’s original question only the noun kalkkuna (turkey) is in nominative case (and the word chicken is in partitive; kana-a). Because of this, already the grammatical form of the question word indicates that it targets the first of the two alternatives in Ada’s turn (see also Haakana et al., 2016). After a micropause, she continues her turn with a turn-constructional unit that invites Ada to repeat specifically the first of the two alternatives in the trouble source turn (on similar alternative questions, see Koshik 2005). At the same time, she points towards Ada to indicate that Ada is expected to react to her turn next (l. 4, fig. 1). In sum, Siiri’s action is embellished and has a performed character because she uses more resources than are needed in the situation to mobilize Ada’s response. And as expected, in the following turn Ada repeats the first alternative. As she repeats the word, she also reiterates the mistake and pronounces a too brief version of the word-internal k-sound (l. 6, fig. 2).

In her next turn, Siiri initiates repair again (l. 8, fig. 3). This time, she produces a hm-sound with rising intonation. This type of a sound is a typical way to accomplish an open-class repair initiation (Drew 1997) in many different languages (Dingemanse et al. 2015). While producing the OIR, Siiri turns her right ear towards Ada as if indicating that she has trouble in hearing what Ada said (e.g., Mortensen, 2016). At this point, she has already initiated repair twice on the same trouble source. It has been shown that repair sequences are rarely extended beyond two successive OIRs on the same trouble source (see Schegloff 2000; Dingemanse 2015). Because of this, Siiri’s next contribution is quite unexpected and clearly exaggerated. In the turn following Ada’s second repetition, she initiates repair on the same item once again and asks Ada to repeat the focal word (l. 12). While producing the OIR, Siiri leans forward towards Ada. Leaning forward is one of the embodied resources that have been identified in previous research as a frequent resource in repair sequences (see, e.g., Pajo & Klippi, 2013; Rasmussen, 2014). Siiri is thus drawing on these specific embodied resources, such as leaning forward and the turning of the right ear towards Ada, to signal that she has trouble in hearing what her coparticipant is saying. In this context, however, the embodied resources have an air of pretense also because it is implausible that there would be any real trouble in hearing in this situation in which the participants are sitting close to each other with unobstructed visual access and with no other sounds that would make hearing difficult. Both the continued repeating of the OIRs and the performed quality of the embodied resources thus indicate that what Siiri is doing is not done entirely seriously.

Ada reacts to this third repair initiation by laughter (l. 14). She also lifts her chin and moves her head backwards and laughs in a rather quiet voice. In his article on po-faced responses to teases, Drew (1987) notices that quite often the recipients of teases first react to the tease by laughing. Laughter shows that the recipient of the tease acknowledges the humorous element in the tease even if she would not take the substance of the tease seriously (Drew, 1987; Glenn, 2003, p. 123). Emphasizing the linguistic errors of one’s fellow conversationalists’ speech in this manner could possibly be taken as an act that is face-threatening (see also Kurhila, 2006). However, if such behavior is treated as non-serious, it does not lead to as “serious” consequences as it might otherwise (see Glenn, 2003). In this way, Ada’s laughter also normalizes the situation and releases the tension caused by the explicit focus on her mistakes (see also Haakana, 1999). However, in her next turn, Ada still also orients to the sequential expectations created by Siiri’s third repair initiation and repeats the trouble source word once again. She also moves her upper body towards Siiri while articulating the word, this way orienting to the claimed trouble in hearing (l. 15, fig. 6).

After the third repetition by Ada, Siiri corrects her (l. 17). The correction is very explicit and done without any hesitations (Haakana & Kurhila, 2009). In it, Siiri stresses the length of the word-internal k-sound, making it even too long and prominent. The straightforward correction at this point in the sequence has interactional consequences. First, producing the correction in this manner shows that Siiri adopts the speaker identity of a linguistic expert and simultaneously positions Ada as someone who is linguistically less proficient (see also Theodórsdóttir, 2018, on teaching sequences in the wild). Second, the correction indicates that Siiri has not initiated repair because of a hearing problem but rather because she wanted to focus on the problem in Ada’s pronunciation.

Next, Ada takes a turn which shows that she does not treat the correction as a legitimate action at this point (l. 18). This shows in the beginning of her turn as she produces the dialog particle “niin” that signals that she thinks that “kalkkuna” is exactly the word she has been repeating all along. As evidence of this, she articulates the word again, and this time the word-internal k-sound is hearable as long enough. Here Ada is thus setting the record straight (Drew 1987).

The teasing quality of Siiri’s actions become even more clearly observable as the sequence continues. In the next turn, Simo explicates that that there are three k’s in the word (i.e., one at the beginning and two in the middle) (l. 19). Siiri then goes on explaining further what the problem has been. This time she imitates Ada’s pronunciation and at the same time, reproduces the “mistake” Ada has done but in a clearly exaggerated manner saying that “you said kaakuna” (l. 21). Siiri thus not only produces the k-sound too briefly but also prolongs the a-sound in the beginning of the word. This is something that Ada did not do in her own pronunciation. Here she is thus teasing Ada by clearly exaggerating the minor mistakes Ada has made in her pronunciation.

Ada, however, stands by her opinion. She articulates the word once more in the next turn and stresses the two k-sounds. She produces her turn smilingly, which again shows that she has identified the humorous tone of the interaction (l. 22). After this, Siiri initiates a spontaneous song, “kaakuna kaakuna kaakuna”, and flutters her arms as if they were bird wings (fig. 9). Simultaneously, Ada reiterates the original question by asking whether the meat they eat is turkey and Simo confirms that it is indeed turkey.

4.2 Setting the record straight and counter-teasing

In extract 2, the teasing sequence is initiated by Simo. Again, the repair is initiated on a minor linguistic mistake in Ada’s talk. As Simo initiates repair, he freezes his bodily actions (see Floyd et al., 2016). This freeze emphasizes the importance of the repair activity: All the other ongoing activities (such as reaching for the butter) are frozen and postponed until the trouble of understanding will be solved. Given the minimal mistake that the repair targets, hardly a cause for serious understanding trouble, Simo’s bodily behavior is exaggerated.

The extract starts in the middle of a topic concerning a phone call Simo has had with the researcher who has asked them to videorecord their interactions. Ada has asked several questions about the phone call and in the beginning of this extract, Simo reports that the researcher had asked whether they have already recorded many conversations (i.e., “if there are many things on tape”, l. 1-2). He then goes on reporting what he has replied to this question (l. 5–7).

Extract 2. Kassetta

In the beginning of the sequence, Simo reports what he has answered to the researcher’s question. His turn in line 5 begins with the reporting clause (“I said”) that is followed by a direct quotation of his own previous speech. This quotation is formulated in a rather peculiar way: Simo says that he has answered the researcher’s question by saying “what things would there be now that the girl has been at home only in passing, i.e. popping in and out”. The fact that Ada has not been at home very much is consequential to the taping of the conversations as Ada is the second language speaker in this group who is supposed to be present as they videotape their interactions. Simo’s comment could thus be heard as criticizing Ada for not having been at home enough. However, the wording of (“in most urgent situations”) makes it sound not entirely serious.

In her reaction, Ada disagrees with Simo’s stance and states that they have indeed recorded a lot as the recordings already fill two cassettes (l. 10). The form that Ada uses for the word cassette contains a small error since she adds the partitive case ending (-A) to an incorrect stem of the word and therefore ends up with an erroneous form. In any case, the form seems to be perfectly understandable in this context: The form she uses (kasettaa) is not so different from the correct form (kasettia) and the context of the turn should help in interpreting the meaning of the erroneous form.

At the same time with Ada’s turn, Simo takes a sip of his drink (line 10) and therefore is unable to react to Ada’s turn verbally immediately. Ada takes the turn back and starts to increment something to the previous turn with the conjunction or (l. 12). This beginning is, however, followed by a rather long gap during which Simo finishes drinking and turns his gaze towards Ada. He then produces a turn that is recognizable as a repair initiation. The initiation is constructed so that it clearly targets the erroneous word kasettaa (cassettes): It consists of two words, the first of which is the word kaksi (two) that was the word preceding the problematic word in the trouble-source turn. The repeated word kaksi (two) is followed by a question word mitä (what) (l. 14, fig. 1). The question word is in the partitive case, just like the problematic word kasettaa. Therefore, it seems clear that the target of the repair initiator is the word form kasettaa.

While Simo produces the OIR, Siiri gazes at him and smiles (l. 14). The smile shows that Siiri has recognized the non-serious quality of Simo’s turn. Simo, however, articulates the repair initiation entirely po-faced and gazes toward Ada. His right hand is positioned above the glass that he has just put on the table after having drunk from it (fig. 2). While producing the OIR, he freezes his hand in this position above the glass and stays motionless until the end of the repair sequence (l. 22). This is thus an example of a hold (see, e.g., Floyd 2016, Oloff, 2018; Kamunen 2019). It is noteworthy here that Simo makes use of this embodied practice even though it is unlikely that he has any genuine trouble in understanding or hearing. Simo is here thus drawing on the semiotics of “a hold” to perform non-understanding.

As the repair initiation so clearly targets the word kasettaa, it projects a repetition of the trouble source or an explanation of it as the relevant next. Ada chooses to repeat the word that was picked up as a trouble source. As she repeats the word kasettaa she simultaneously repeats the grammatical mistake (l. 17, fig. 2).

Simo reacts to the turn by repeating it and stresses the first syllable in an exaggerated manner. Because of the stress the repetition appears to be prosodically overdone — it seems almost harsh. This overdone prosodic design of the repetition serves as a contextualization cue (Gumperz, 1992) to indicate that Simo is not acting in an entirely serious manner anymore. It is also noteworthy that as Simo repeats the word, he also repeats the error. This is noteworthy since it has been observed that in second language interactions the first language speakers usually do not repeat the grammatical mistakes that the second language speakers make but rather either let them pass or correct them “en passant” in specific sequential environments (see Brouwer et al., 2004; Kurhila, 2001, 2006; Theodórsdóttir, 2018). Also Norrick (1993) has analyzed the interpersonal dimension of conversational joking and observed that one strategy of introducing a joke in conversation is to repeat a slip of the tongue or an error produced by the previous speaker. This kind of repetition might also be done in a mocking way. He (1993, p. 77) points out that a polite thing to do would be to ignore the errors and slips of the tongue that other speakers do. Drawing attention to an error by repeating it might be interpreted as offensive.

After this repetition, Ada produces a turn in which she orients to clarifying the focal word. She adds the adjective little in front of it, corrects the form and depicts the small size of the cassettes by both hands (l. 20, fig. 3). After this, Simo acts as if he only now understood what they are talking about. He produces a dialog particle jaa, indicating understanding (see Koivisto 2017), repeats the word kasettia again and stresses the final part of the word which was the cause of the trouble in the previous turns. He also disengages form the hold and reaches for the butter (fig. 4). All in all, Simo acts as if he only now understood what Ada was saying.

It is only at this point that Ada laughs for the first time (l. 23). By her laughter, she shows that she has identified the non-serious quality of Simo’s actions and gives support to the interpretation that Simo’s repair initiations have not been done to signal a genuine problem of understanding but rather to highlight the error Ada made. As in extract 1, here the laughter also serves to show Ada’s recognition of Simo’s teasing and to normalize the possible tension in the situation.

In extract 1, Ada set the record straight after the repair sequence. The same happens here. After the laughter, Ada points out that Simo understands anyway (l. 25, fig. 5). She thus explicitly shows that she understands that Simo’s actions have not been genuine indications of trouble in understanding and sets the record straight (see Drew, 1987). Toward the end of her utterance, she starts to laugh again, and this time Siiri joins in (l. 25–26).

From here on, the sequence continues with a humorous discussion about the word “kasetta”, i.e., the original erroneous form, and whether it belongs to the Finnish language or not. Simo argues that he could not possibly have understood what Ada meant by kasettaa (l. 28–29), and Siiri suggests that the word form Ada used was possibly Hungarian (l. 32). Ada denies this, and her denial inspires Simo to declare that in any case the form was certainly not Finnish (l. 36–37). By emphasizing that the word kasettaa is not Finnish, Simo acts as an expert of Finnish language: as someone who has the right and the required knowledge to say what forms belong to the Finnish language and what forms certainly do not. By underlining his own expertise, Simo at the same time makes relevant the fact that Ada is the one who lacks this kind of expertise, as she has not been competent enough to produce forms that are understandable to proficient speakers of Finnish. Accordingly, at this point in the conversation, the asymmetrical linguistic roles of the participants are evident.

The prosody of Simo’s turn in lines 36–37 is again exaggerated: he changes the quality of his voice and produces the turn in a theatrical manner. The changes in voice quality and in the speech style are a possible indication of the non-serious mode of the interaction (Keltner et al. 2001: 234–236; Mulkay, 1988). The story continues as Simo repeats the word kasettaa again and informs that in Italy there is a paper called “Kasetta dello sport” (l. 42–43).¹ Simo thus implies that the word form kasettaa belongs to a name of an Italian newspaper rather than to the Finnish language.

Ada, however, does not easily accept the position of a teased person. Throughout the whole sequence she has been resisting this role. This is especially evident in how she reacts to the information about the name of an Italian paper. In fact, she turns the focus from the word kasetta to the word dello and asks how it is possible that Simo knows that the combination of an article and preposition in Italian in such a context is dello (l. 48). It is possible to analyze Ada’s behavior also as counter-teasing: She is making use of the available resources for questioning Simo’s behavior.

Next, the topic of the conversation shifts to the food that they are eating. After a couple of turns, the counter-teasing becomes observable again as Ada revisits the topic of the newspapers and asks in line 59 what a similar kind of a newspaper might be called in Spain. Simo provides “La Repubblica” as his answer. Ada’s question of the Spanish newspaper could again be heard as challenging Simo’s expertise, but this punch from Ada leaves no holes in Simo’s armor (even if La Repubblica is in fact another Italian newspaper). After a short recourse to a discussion about ice hockey, the language expertise issue emerges once again as Ada states that she drinks apple juice and asks if the partitive form she used was right. Here, again, Simo argues that the word Ada uses (tsuissi) is not Finnish, but he admits that the partitive form is nonetheless right. After this the topic of the conversation changes.

6. Summary and discussion

Teasing is an action that has two sides: It is humorous and amusing, but at the same time, there is also a cutting element to it. In formulating teasing actions, teasers need to balance between the two sides. The analysis presented in this paper has focused on participants’ embodied conduct in the service of action ascription in teasing sequences. The analysis was motivated by the observation that as a strategically ambiguous social action, teasing is intriguing in terms of action ascription. We hope to have shown that in the analyzed sequences, teasing was accomplished through the enactment of embodied conduct that typically characterizes repair environments. The teases were implemented by initiating repair and combining the verbal initiations with forward-leaning (extract 1, see also, e.g., Pajo & Klippi, 2013, Mortensen, 2016), head turns and tilts (extract 1, see also Seo and Koshik, 2010), and gestural holds (extract 2, see also Floyd et al., 2016). All of the embodied actions were performed in manner that was somewhat overdone in their situated contexts. The teasing thus emerges from the local discrepancy between the rather minimal problems that the OIRs targeted and grandiose embodied quality of the repeated OIRs that were launched to signal the problem in mutual understanding. The repetition of OIRs targeting the same trouble source led to accumulating tension, which is then alleviated by the non-seriousness in the analyzed sequences.

In the analyzed sequences, the OIRs were used as vehicles for carrying out the teasing actions. The OIRs were used as vehicles for carrying out the teasing actions. In this sense, the analyzed OIR are examples of double-barreled (Schegloff, 2007) social actions. Our analyses of the “kalkkuna” and “kassetta” examples illustrated that the double-barreled OIRs were designed as multimodal action packages in which the verbal part is delivered in a serious manner and combined with performed and exaggerated embodied conduct. In the accomplishment of these teasing sequences, the participants were shown to perform embodied conduct typically found in repair environments and through this, to manage action ascription in ways that are recognizable to their coparticipants. In this sense, they bring in their embodied interactional competence and knowledge of the meaning of embodied resources for action accomplishment.

The analyses illustrated that in teasing sequences, in which the non-standard language use is targeted and used as a resource for teasing, the participants’ orientation to the speaker identities of an expert language user and language learner become clearly observable. Previous research on second language interaction in different contexts has shown that emphasizing linguistic mistakes in this way is rather uncommon (see, e.g., Kurhila, 2006). The sequences analyzed here showcase that while linguistic mistakes or deficiencies are not commonly dealt with in interaction explicitly; they can, however, be brought into interactional focus with the help of humor. In the analyzed sequences, the negotiation of linguistic expertise was a complex activity since the teased participant always set the record straight and showed that she did not treat the repeated repair initiations as legitimate in situations in which mutual understanding was not really at stake. In this sense, our analysis also illustrated how the linguistic status of participants can be negotiated, defined, and redefined in the course of the interaction.

Acknowledgements

We thank Nathalie Schümchen for her help with the figures.

^{1 The name of the Italian newspaper is "La Gazetta dello Sport" but simo pronounces the word "Gazetta" as "kasetta", and it sounds similar to the word that Ada has used to refer to cassettes.↩}