Teasing and policing in a multilingual family-Negotiating and subverting norms and social hierarchies

This paper demonstrates how multilingual adolescents initiate language-directed teasing in family interaction and thus contribute to reinforcing or challenging social hierarchies and norms in the family. It investigates the case of a multilingual family living in Northern Norway (two parents and five children ranging from 3 to 18 years old). To a varying extent, and with varying degrees of competence, all family members use three languages in their daily lives: English, Spanish, and Norwegian. The data consists of self-recorded material of family interactions (9 h; 549 min) that were collected over the course of one year. A close interactional analysis shows how the siblings target linguistic production in teasing attacks, and use language-directed teasing as an interactional resource to position themselves and their family members. Drawing on Billig's (2005) theory of disciplinary humor, the article argues that playful corrections of perceived norm transgressions may be understood as situated (re)production and negotiation of social and linguistic norms, through which the young family members participate in the construction of the family as a community of practice. © 2020 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).


Introduction
By focusing on the interpersonal functions of teasing occasioned by prior linguistic transgressions (cf. Haugh, 2017), this article investigates how a multilingual family is constructed as a community of practice through teasing among siblings. The analyses show how the young family members exploit the whole range of the family's repertoire of linguistic resources to (playfully) engage in linguistic sanctioning, and to create links between language forms and social attributes. Basing on Billig's (2005) theories of humor and social order, the article demonstrates that teasing offers a glimpse into how linguistic resources and competences are negotiated in a multilingual family. Moreover, teasing is used by the children as an interactional resource in (re)producing, negotiating and subverting socialroles, hierarchies, ideologies and norms, and locally relevant identitycategories within the family. studied within the fields of (socio-)pragmatics, sociolinguistics and linguistic anthropology, and has been explored interdisciplinarily through a wide range of approaches (see Haugh, 2017 for a comprehensive overview). In language socialization and the family context, teasing has been shown to serve norm-controlling and disciplining functions. Aronsson (2011: 468e469) reviews verbal play in language socialization, and argues that much of the teasing games within the family can be seen as "borderwork play, in the interface of children's impotence and adults' authority." Parents may, for example, shame their children into behaving according to parents' expectations of appropriate social conduct through teasing (Eisenberg, 1987;Schieffelin, 1987). Eisenberg (1987) shows that parents tease their children to control their behavior, but also that teasing forms and reinforces social relationships within the family. Miller (1987) connects the socializing functions of teasing to the children learning the value of self-assertion, standing up for oneself, and fighting back. Teasing as a socialization practice involves the learning of a range of social and linguistic skills, such as how to produce, recognize and react to teasing, learning the appropriate contexts of teasing (e.g. who can be teased and when) and the (socially) unifying functions of laughter (Eisenberg, 1987).
Within anthropology, the research on teasing had a bias up until recently towards adult-initiated teasing and a preference for adultechild interactions, and not children or peer interactions (Lytra, 2007b). The bias of viewing the adults as the ones setting the rules has been criticized in language socialization research (Ochs and Schieffelin, 2011) and in the more recently emerged field of Family Language Policy (Fogle and King, 2013), and the agency and engagement of both novices and experts, children and adults are increasingly acknowledged. Moreover, several studies have investigated teasing practices among childrenwith the aim of remedying this bias (Eder,1993;Lytra, 2007a;Svendsen, 2004;Tholander and Aronsson, 2002;€ Ozkan et al., 2015). This line of work has provided insights into the multiple social function of teasing among children. Studies from linguistically and culturally mixed peer groups have shown how children, through teasing, monitor the social conduct and speech of their peers (Svendsen, 2004), where, for example, linguistic production or academic performances are targeted to ascribe particular (unwanted) identities to peers (Lytra 2007a(Lytra , 2007b€ Ozkan et al., 2015). Research on how children employ and use linguistic resources creatively and playfully in family contexts is rather scarce, especially within multilingual families. We find an exception in De Fina (2012), who reports that performances and metalinguistic comments (expressed through, for example, teasing and joking) are linguistic strategies through which family members in a three-generation Italian-American family demonstrate engagement in Italian, the heritage language. Teasing also works to display, test out and negotiate identities. Drew (1987) shows how teasing tend to ascribe (or at least implicate) an unwanted identity or deviant attributes to its victim. The social practice of teasing also opens interactional spaces were identities may be contested and resisted and alternative self-identifications are foregrounded (Lytra, 2007a). In addition, teasing directed at linguistic forms may involve the affirmation or negotiation of group identity, or similarly, marking a distance from other groups by mocking the way they speak (Lytra, 2007a, b). By engaging in teasing practices, participants can develop and confirm relational identities among speakers by e.g. displaying alignment (Boxer and Florencia,1997;Eisenberg,1987). Eisenberg (1987) argues, for instance, that teasing within the family reinforces the relationship between the participants involved. Similarly, Boxer and Florencia (1997) suggest that by teasing, people demonstrate personal knowledge of one another and confirm a relationship of closeness and intimacy, which is why hostile or offensive teasing requires social proximity in order for the tease to be interpreted humorously. One of the social skills acquired through teasing is to be able to be the butt of a joke without being offended, and insights from conversational analytical approaches to teasing show that second-teases or non-serious uptakes tend to be preferred responses of teasing victims (Drew, 1987). In sum, teasing is a versatile semiotic resource that can serve various social functions. By examining child-initiated teasing in a multilingual family, this article adds to this existing knowledge.

Teasing as metalinguistic, rebellious and disciplinary
As indicated above, teasing as a conversational resource serves several social functions, one of which is that of social control. In Billig's (2005) theoretical framework for the relationship between humor and social order, he distinguishes between two kinds of humor; disciplinary and rebellious. Whereas disciplinary humor "mocks those who break social rules" and "can be seen to aid the maintenance of those rules," rebellious humor "mocks the social rules, and […] can be seen to challenge, or rebel against, the rules" (Billig, 2005: 182). As Billig (2005) notes, however, teasing these categories apart might be difficult in practice. For example, rebellious humor may have disciplinary functions that confirms the hierarchies it seeks to transgress (see also Franz en and Aronsson, 2013). Verbal humor may be considered a metalinguistic activity in itself. Teasing is, for example, often characterized by performances (Bauman and Briggs, 1990): Tholander and Aronsson (2002) show how teasers impersonate others, parody, appropriate and stylize others' voices as resources in teasing attacks. All kinds of linguistic features that speakers have awareness of may, in theory, be subject to joking or teases. Joking and teasing about language is loaded with metalinguistic significance, such as speakers' attitudes towards language and linguistic norms (Norrick, 1993) and keeps the interactional participants attentive towards linguistic form (Poveda, 2005;Åhlund and Aronsson, 2015).
In interaction, teasing often emerges as locally occasioned opportunities to address minor conversational transgressions in ongoing talk (Drew, 1987;Haugh, 2017), such as breaking linguistic rules (intentionally or unintentionally). Correcting others in interaction is, however, much less frequent than self-correction, partly because of its interruptive nature, which may become a source of disagreement (Schegloff et al., 1977). Despite this, targeting linguistic competence in teasing is common. Norrick (1993) argues that jokes and teases that comment on linguistic form are metalingual (in Jacobson's,1960, terms), and may define and sort out the linguistic features appropriate for the interaction. Explicitly targeting linguistic production to tease may be interpreted as "overt linguistic power struggles," where interactants link together linguistic features and social values, and use semiotic resources to attribute others'/co-participants' unfavorable positions ( € Ozkan et al., 2015: 467). Such practices are documented in interactions in multilingual school and classroom contexts (Svendsen, 2004;Lytra, 2007a;€ Ozkan et al., 2015), but only to a minor extent within a multilingual family context (but see De Fina, 2012, or Norrick, 1993. Billig (2005: 184) argues that there is a duplicity in the disciplinary teaching of children; "the overt message of the adult, laying down the rules" on the one hand, and "the covert teaching that indicates how the adult world of rules can be disrupted" on the other, and points at how this pattern can be observed in, for example, language learning. By learning how to speak "appropriately", children also implicitly learn what is not appropriate. However, in a multilingual family setting, with diverse and unevenly distributed linguistic resources, children may possess linguistic competences that their parents do not, and vice versa. In Ag's (2016: 110) study of negotiations of authority in family communication, some parents reported that their children corrected their (linguistic) behavior, and she interprets such instances as a "jocular ritual among the family members." Moreover, as the examples in this article show, the young family members implicitly position themselves as authorized to linguistically discipline their siblings and their parents, and thereby contribute to form, reinforce and subvert linguistic and social norms and hierarchies within the family.
2. Methodology, data, and participants 2.1. Methodology and analytical framework Self-recorded interactions constitute the main data source, but the analyses and discussions are also informed by information gathered through semi-structured interviews and ethnographic observations. Methodologically, the article draws on linguistic ethnography and its attention to how communication functions in peoples' everyday routines (Copland and Creese, 2011;Lytra, 2007a), and the close, intensive analysis of linguistic signs of everyday interaction adopted in interactional sociolinguistics (cf. Rampton, 2017). In line with interactional sociolinguistics, the analysis focuses on how the interactants create social and linguistic norms, social relationships, identities and contexts in interaction (cf. Lytra, 2007a;Rampton, 2017;Gumperz, 1982).

Participants
The family in focus 1 consists of two parents (mother and father) and their five children; David (18), Samuel (16), Adrian (12), Erick (8) and Sofia (3). The father has lived in several different countries, one of them in Central America, where he met the mother. He learned English at an early age, and Spanish after he met her. The mother grew up in Central America, and learnt English through schools and travels. The family has lived in the Northern Norwegian city where the research took place for about four years. The four oldest children grew up in the Central American country, and speak Spanish, albeit to a varying degree. They attend or have attended an international school where English is the language of instruction, and where Norwegian is taught as a second language for pupils who do not have Norwegian as their first language. The youngest, Sofia, goes to a Norwegian-speaking kindergarten. After graduating middle school, David started in a public, Norwegian upper secondary school. In addition, he works in a Norwegian-speaking environment. Thus, within the family, various linguistic resources are in use: Spanish and English, and, to some extent, Norwegian.
The data from this family consists of 31 audio recordings of self-recorded interactional data (9 h; 549 min). During the data-collecting period, semi-structured interviews and follow-up conversations with the parents and the three oldest children were conducted. I also visited the family several times and had coffee or dinner with them when leaving and picking up the audio recorder. The recordings were made over a period of around 12 months. Throughout this period, the family had the recorder for three periods of a few weeks each. They were instructed to record family interactions, and were given the opportunity to delete interactions or part of interactions if they wished to.

Recordings and coding procedure
The data stems from a larger corpus of data collected to study language practices among multilingual (Spanish-speaking) families in Northern Norway (Johnsen, n.d.). Thus, the material was not collected with the aim of studying teasing and humor, but as examples of everyday language practices within multilingual families. The self-recordings were transcribed in ELAN (ELAN, 2018) and coded manually in nVivo. In the recordings, joint laughter, telling jokes and stories as well as teasing each other were recurrent verbal actions.
The analyzed excerpts where selected based on a series of criteria. Initially, the data was coded according to the overarching categories 'jokes and teases' (39 instances), 'metalinguistic comments' (35 instances), and 'language plays and metalinguistic humor' (25 instances). The overlaps between the different categories were further explored, and children's role in teasing episodes were given particular attention. The four analyzed examples were chosen because they were examples of language directed teasing (11 instances), where the children in the family exploited what they perceived to be linguistic norm transgressions in teasing. The teasing-episodes were identified through a combination of prosodic and intonational cues associated with humor or sarcasm, such as laughter, smiley voices, exaggeration, rhythm and pitch or stylizations (see € Ozkan et al., 2015), and were directed present participants (cf. Eder, 1993).

Findings: children as agents in negotiating social and linguistic norms and hierarchies
In this section, I discuss five examples of how teasing directed at linguistic form works as a means of reproducing and subverting social hierarchies between the family members. All the excerpts illustrate how the children exploit stylistic and linguistic mixing to initiate teasing attacks. The perceived norm transgressions involve interactionally unexpected language choices, lexical usage and pronunciation, and stylistic features. Throughout the section, I discuss the social and metalinguistic functions of the teasing episodes.

Disciplinary teasing and reproducing social roles and hierarchies
In the data, social hierarchies are often reproduced along an age-dimension. In Excerpts 1e3, we see examples of how such hierarchies are reinforced through teasing and linguistic sanctioning: The older siblings corrected and teased the younger ones and consequently invoked local authority positions and local rules of conduct.
"La playa" e performing accented speech The excerpt below starts after Adrian (12) initiated a verbal activity in a prior turn (Adrian: "okay, one thing you miss about (Central American country), mum"), and challenges the mother to name things she misses about their former home country. Both the mother and David contribute to the activity, and the turn eventually comes to Adrian, which is the start of Excerpt 1.
Excerpt 1* (Participants: mother, David (18), Samuel (16), Adrian (12), Erick (8) The teasing sequence starts off as other-correction, where the older brother, Samuel (line 2), identifies and repeats a trouble source in Adrian's turn (the determiner the in English, and not the Spanish la). The mother (line 3) also repeats the same phrase as an uptake, correcting the determiner into Spanish. She attempts to continue with the topic of conversation (line 3), but is interrupted by Samuel, who exaggerates Adrian's language choice error with a mocking intonation (line 4). Thus Samuel "parasitically exploits material in the prior turn" (Tholander and Aronsson, 2002: 584) to direct a teasing attack at Adrian. By targeting his little brother, Samuel claims authority and a position as "superior" in the family's social hierarchy. David responds to Samuel's teasing attack by laughing, seemingly displaying alignment with Samuel, acknowledging his social position. Adrian loudly self-corrects (line 6), perhaps to demonstrate his knowledge of the correct language choice and standard form of the definite article, la (in Spanish), not the (English). From line 7, Samuel performs the chorus of a pop song in accented Spanish (e.g. producing a postalveolar, "Americanized" /r/, [ɹ], in bailar), and takes the mocking of his brother's Spanish skills one step further from merely pointing out a mistake. The song could in itself be an index of 'poor Spanish,' 2 as the lyrics are based on popular 'Latino' stereotypes of going to the beach, partying, and dancing, and is performed in a mix of English and 'foreign-accented' Spanish. The original song Samuel refers to may be interpreted as Mock Spanish (Hill, 1998). However, in this instance, the Mock Spanish is used by a Spanish speaker (Samuel), who draws a link between Mock Spanish lyrics and his little brother's poor Spanish skills, to socially position Adrian as less competent. This may be viewed as a process of "othering" that renders Adrian's linguistic production deviant compared to the expected linguistic norms of the family (Coupland, 2010). This process of "othering" does not necessarily divide between "in-group" and "out-group" repertoires. Rather, by stylizing Mock Spanish, Samuel plays with the linguistic and social boundaries within the group in a way that retains a sense of in-group solidarity: Spanish is still a linguistic resource that links the family members together, despite their different proficiencies. Samuel's performance is taken up playfully by the brothers, as the giggles in lines 8 and 9 indicate. Samuel ♪ ♪ vamos a la playa ♪ Let's go to the beach 8 Adrian ((giggles)) 9 Erick ((giggles)) 10 Samuel ♪ $a mi me gusta bailar $ ♪ ((in heavy accent Spanish)) I like to dance 11 David ha ha ha ha *For transcription key, see below.
During one of my visits to the family, I had a conversation with Samuel where he took a similar metalinguistic stance towards Adrian's use of Spanish: disapproving of his accented Spanish, implicating that he should rather speak English: Excerpt 2 (Participants: Researcher and Samuel) Prior to this, Adrian directed a request towards Samuel in Spanish, and Samuel responded in English. I followed up by asking whether they speak Spanish to each other, as the self-recorded material indicated that Spanish was not the preferred language choice between the brothers. In this question-answer sequence, Samuel positions his younger brother as a poor Spanish-speaker by imitating him in an accented manner, and discredits his younger brother's attempt at speaking Spanish: Accented Spanish (or Mock Spanish) is identified as deviant by Samuel. Suggesting that he should speak English rather than 'accented' Spanish, he implicates that Spanish is not a 'real' part of his brother's repertoire. The alternative, as suggested by Samuel, is not to improve in Spanish, but to rather speak English (i.e. 'stick to what you know').
"Dad, not dude"e exploiting stylistic transgressions Styles and highly indexical features (such as slang words) were also subject to correction, as we see in Excerpt 3. Here, Samuel needs appropriate equipment for a school skiing excursion. In this interaction, Samuel and his father are discussing whether he should borrow from a family friend, or rent: Excerpt 3 (Participants: Father, David (18) and Samuel (16)) In line 5, Samuel suggests that they stop by the rental to check the prizes, and employs a lexeme that sounds like dude. In this immediate linguistic context, due to Samuel's intonational pattern the utterance is ambiguous, and dude could in fact be interpreted both as a vocative expression, referring to the father, and also to the person working in the rental ("the store dude"). Both the father and David seem to have interpreted it as a vocative expression referring to their father. The slang word dude may convey numerous meanings, especially among young male speakers. Dude is often used as a term of address among young men, and indexes a cool, casual stance and masculine solidarity (Kiesling, 2004). These indexical values that signal informality, coolness, and youth seem to be recognized by David, who explicitly corrects his younger brother in line 7. Samuel responds to the correction seriously by denying that he said dude. David's correction points out a norm break, and displays David's perceptions of what counts as appropriate ways of speaking to their father. By correcting his little brother, David may be assigning Samuel with an unwanted identity of being rude or disrespectful. Correcting this linguistic form seems to allow David to construct alignment with his father, while also demonstrating authority over his younger brother and (re)produces interactional norms by doing linguistic normative work.
What follows is a small dispute about whether Samuel actually uttered dude or not, which Samuel repeatedly denies. By denying it, Samuel also denies the ascribed identity as being disrespectful. However, the father aligns with David in line 10, and consequently authorizes the correction of his brother's language. In line 16, the father attempts to change the framing of the tease into a more playful one by laughing lightly, as the correction has led to a quarrel between the two brothers. In line 18, David insists on Samuel's mis-speech by stating that everybody heard him, and also takes up on the father's playful framing by stylizing the utterance in accented English (pronouncing /r/ as an alveolar trill [r]). The father affirms David's claim that because he goes like "Samuel pod es venir a mi cuarto?" (("Samuel can you come to my room", stylizing his brother in a heavy English-accented Spanish)) 5 And I like "why don't you just say it in English" everybody heard Samuel say dude (line 19), and follows up with another tease on Samuel's behalf suggesting that he said it unconsciously. Raising his voice, Samuel strongly signalizes 'unlaughter' (i.e. the rhetorical opposition to 'laughter' cf. Billig, 2005) through an attempt to move forward with the conversation, and repeats his initial request to go and check out the skiing store. By exploiting a perceived error in Samuel's turn, David manages to demonstrate a range of locally relevant sociocultural skills, such as loyalty and respect towards the father, and his own authority. He also demonstrates metalinguistic knowledge of the indexical values of dude, and the local norms of conduct and speech within the family context. In sum, the corrections provide a window into the acceptable linguistic norms of language use within the family (i.e. how to address parents 'adequately'), which in turn relate to their local family values and social hierarchies (i.e. perceived ways of acting respectfully).
The two excerpts demonstrate that there seems to be a certain age hierarchy in which the older siblings (Samuel in Excerpt 1, David in Excerpt 3) correct and tease the younger ones. More specifically, and following Billig (2005), the examples may be interpreted as disciplinary teasing: They illustrate how older siblings use language-directed teasing to construct alignments between family members, and to claim competence and authority. By identifying a norm transgression, the boundaries of the local, linguistic norms are also being negotiated. In claiming authority over linguistic norms, the older brothers also claim a superior position and reproduce and reinforce the social (age-scaled) hierarchy. In sum, the examples show how othercorrections directed at language may, on the one hand, be employed in a playful frame as a means of teasing in which social roles and positions can be reinforced and authority and power positions can be claimed and questioned. On the other hand, teasing directed at linguistic form may contribute to controlling, reinforcing or confirming the linguistic norms within the family through which linguistic norms are being negotiated. Through linguistic corrections, the teaser ascribes an unwanted identity to the target of the verbal correction. At the same time, the younger ones may resist being positioned with an unwanted identity ascription. As the following excerpt will demonstrate, the local hierarchies are by no means stable or static structures.

Subverting social hierarchies: correcting older siblings and parents
The following examples show how teasing attacks may subvert the social hierarchies in the family, and how children exert their agencies in negotiating authority and competence. In addition, both Excerpts 4 and 5 (similar to Excerpts 1e3) draw attention to how the family members orient towards various linguistic norms and how this metalinguistic awareness works to construct a multilingual family identity.
"Good characters" e undermining sibling authority through perceived errors This sequence is part of a longer recording of a family dinner, and the brothers in the family (David, Samuel, Adrian and Erick) have been teasing each other back and forth for a while. The younger siblings, Erick and Adrian, ascribe David with the identity "old" as it is soon his birthday, when he will turn nineteen. This leads to the following sequence in which David and Samuel start to discuss their personal achievements.
Excerpt 4 (Participants: Father, David Samuel, Adrian, and Erick) In line 6, David states his achievement of graduating school with good grades, but employs the word "characters" (with an English pronunciation) but with the lexical content from Norwegian (where karakterer means grades). David's use of a purported inappropriate word triggers a correction from his father and younger brother Adrian (lines 8 and 9), and, not least, material for Adrian to design a tease directed at his older brother. In line 8, the father offers a corrective feedback by repeating the error with a rising pitch, identifying the trouble source in David's turn, and possibly offering David the opportunity to self-repair (Schegloff  et al., 1977). Adrian (line 9) initiates a second-repair by repeating the same trouble source once again in a stylized manner, employing exaggerated prosody and volume, which emphasizes his intention of making fun of David's error. As the error has already been spotted and pointed out by the father, it is unlikely that he repeats this because he misheard his father. Repetitions recurrently reflect metalinguistic awareness and have a range of functions e didactic, playful, emotional and expressive, to mention only a few (Johnstone,1996). In this case, Adrian builds a teasing attack by way of the repetition, and succeeds in changing the footing of the interaction into a teasing frame. Adrian does not let the teasing opportunity slip away, and we see how the initial discussion of age and achievements becomes connected with the correction (line 15,19, and 21). The tease can be interpreted as a misrecognition of David's claimed achievement: In lines 4 and 6, David is claiming a status and identity of a 'good student' by highlighting his achievements, and by displaying his age as a positive resource that gives him a certain authority and status. David's utterance may have been particularly vulnerable to being teased about, as he (unconsciously) contradicted his bragging of good school results by producing a linguistic error. By stylizing David's error in a mocking tone, Adrian assigns an unwanted identity to David that is in stark contrast to the claims David has just made about himself, and succeeds in turning his brother's achievements against him: Whereas David presented his age and experience as a resource, Adrian succeeds in inverting this into a negative attribute. From line 13, Erick also joins in and aligns with Adrian and draws a link between the linguistic error and the fact that David is "getting old." As the transcription indicates, the prolongation of the vowels in /o::ld/, pronounced in an emphatic voice, makes Erick's performance seem like an attempt to stylize an 'old man,' or a way to exaggerate the teasing frame. By turning the advantage of being the older sibling into material for a teasing attack, Adrian and Erick succeed in challenging and subverting the relational hierarchy between the younger and the oldest sibling, and inverting the status of the category "old" into something utterly negative (as a cognitive decline with reduced capacity of lexical retrieval). David responds to the tease directed at him without laughing (see Drew, 1987 on 'po-faced' responses). From line 12, he offers an explanatory account for the error ("I was thinking in norsk"). However, he also manages to present himself as multilingually competent, first by self-repairing the utterance in Norwegian (line 12), and later also in line 14 in English. Expanding on this account, David states that "there are too many languages" in his head, and initiates a story of how he forgot the language in which he spoke with a friend (lines 18 and 20). The issue of 'linguistic confusion' is in fact taken up in the following turns with an anecdote of how he spoke Norwegian to an English-speaking friend, and followed up by the mother with a similar story of being 'linguistically confused'. David positions himself as a victim of his multilinguistic competence and implies an idea of linguistic confusion caused by his knowledge of several languages. The explanation in line 18 related to the many languages in his head implicitly displays biographical information that is shared by several of the family members. By conveying this information, he also engages in constructing a group identity as multilingual, i.e. doing 'being a multilingual family.' Regarding the mixing of different linguistic features, we see that David's Norwegian utterances (line 12) are not corrected, and are therefore in some sense unmarked, while the English lexeme with Norwegian semantic content is identified as an object of ridicule. The correction that takes place in lines 8e12 is not towards mixing language per se, but is perceived as a norm break, and more specifically, it is the norms of English language use that are transgressed. By sanctioning incorrect language, the younger brother, Adrian, can claim competence and knowledge of the use of English in the family, which enables him to act as an authority towards his older brother and play with, and perhaps subvert, the power structure between them. Implicitly, the act of sanctioning the language use may contribute to assigning indexical values to the family's linguistic resources: In this example, English is positioned as a language in which the family members are expected to have good competence, and errors are objects of laughter and ridicule. Similarly, in Excerpts 1 and 2, "Mock Spanish" and accented Spanish is made fun of.

«Potato sour» e exploiting mixing and divergent competences in teasing
The previous excerpts have demonstrated how teasing directed at linguistic forms is used as a resource among the siblings to display or claim authority and competence in interactional 'power struggles.' The adolescent children may also claim such roles by correcting their parents' speech, as illustrated in the excerpt below. In this sequence, the family is gathered around the dinner table. There are, in fact, two parallel correction sequences going on in this interaction: one between Samuel and the mother, and the other between Adrian and the father (and David). Both concern the parents' pronunciation of Norwegian features. The sequence starts a few turns earlier, where a fake plant in the living room is the topic of conversation, which turns into a rather long discussion of the pronunciation of gummy, and a clarification of what is meant by a 'gummi plant' (pronounced / gʉmi plaent/, i.e. a mix of Norwegian and English pronunciation). In line 1, the father asks for a clarification of what is meant by "gummiplant," and the mother responds in Spanish by explaining that it is a plant that is not real, while David repeats the word in Norwegian with emphasized intonation. Thus, the father explicitly corrects the utterance in accordance with English phonology (i.e. /gʌmi/) in line 4, whereas the mother responds (combining features from English and Norwegian) that David is speaking in Norwegian, not English (lines 5 and 6). The mother's response may be interpreted as an explanatory account that accounts for the confusion in meaning and pronunciation. By doing this, she modifies the father's correction in line 4, implying that a correction was not necessary. At the same time, through her response, she positions David as a speaker of Norwegian.
In line 8, Samuel questions his mother's pronunciation of Norwegian, and claims that her pronunciation "enjelsk" (English) was incorrect. More specifically, he identifies an /ñ/ in her utterance, a Spanish consonant (line 9). The laughing particles in line 9, along with a mild sarcastic tone of voice, indicate that he is mocking her, and he positions her as less competent in Norwegian than himself. Interestingly, he claims competence and authority in Norwegian through the use of English features only. The mother does not, however, seem to have a problem with being identified as a learner or less competent Norwegian speaker, and she asks Samuel to provide her with the right form and pronunciation, consequently confirming Samuel's position as the Norwegian expert. However, the imperative verb form and her slightly up-speed intonation give the impression of slight irritation.
In lines 14 and 15, the target word ('engelsk' English) is repeated by both Samuel and his younger brother Erick, who, based on his very low repetition of the word, almost seem to be testing out whether or not he can pronounce it the same way as Samuel. At this point, Samuel repeats the target word with a distinguishable Oslo intonation, and not the local dialect intonation that would have been expected based on the location and previous utterances in Norwegian. This might be an expression of metalinguistic awareness of a conceived standard pronunciation that adds a layer of authority to his correction.
In line 19, the mother states that she did not recognize any difference between her and Samuel's pronunciation. Adrian immediately aligns with the mother by saying that there was no difference (line 21), and that the difference is in Samuel's head. Samuel corrects the word twice more, and explicitly identifies the error of the mother again. Adrian continues to take advantage of the tension between Samuel and the mother to tease Samuel, and states again that the linguistic difference he perceives is all in his head (line 24). Thus, Samuel's claim of linguistic competence is challenged by Adrian, who aligns with his mother and discredits Samuel's position as the knowledgeable one by suggesting that the difference he perceived was a product of his imagination.
In parallel, perhaps as a meta(linguistic) comment on the discussion between the mother and Samuel, the father utters what seems to be a deliberate mistake in Norwegian (line 16): Combining features from Spanish, English and Norwegian, he asks what a 'potetsur' is ("Que es a potetsur"). The word 'potetsur' does not mean anything in Norwegian, while 'sur potet' (sour potato) is a silly, pejorative term for someone who is in a bad mood. Structurally, this question can be interpreted as contrastive language play invoking the placement of adjectives in Spanish versus Norwegian. Spanish adjectives are (usually) placed after the noun, while in Norwegian they are placed before. This alleged 'deliberate' error seems to be recognized by Adrian, who explicitly corrects his father's mistake (line 17), adding an interjection that makes it seem as though he has corrected this mistake before. Judging by Adrian's uptake on this mistake, it seems that the father utters this phrase as a friendly tease to which he expects reactions. David also repeats the error produced by the father (line 18) and laughs, seeming to recognize the 'deliberate' mistake. By the end of the sequence, the two correction sequences seem to interweave, as the father responds to Samuel's last correction by teasing him, asking him to not be a "potetsur" (line 26). At this point, there is little doubt that the mistake is deliberate, as he has been corrected once before. However, Adrian repeats the correction (line 27) and the father starts to laugh (line 28), in what seems like a humorous ritual. This excerpt shows various aspects of the interactional dynamics in this family: The sequence demonstrates through two correction sequences how the children actively engage in reproducing and challenging social norm hierarchies by linguistically disciplining their parents. Additionally, the children challenge their parents' authority positions by claiming (linguistic) knowledge authority (cf. Ag, 2016): The children identify norm breaks, and correct their parents' Norwegian according to a lexical and phonological standard and thus claim (linguistic) authority and competence. The parents also reinforce the children's position as experts, by explicitly positioning them as the more knowledgeable in Norwegian (e.g. the mother in line 5, the father through his 'deliberate' mistake in 27).
Lastly, the sequence demonstrates how different linguistic resources are used within the family. The mother (lines 3 and 5) employs Spanish, English and Norwegian features in the course of two turns to convey her message. Similarly, the father employs features from Spanish, English and Norwegian to engage in a verbal game. These practices suggest that polylingual norms (cf. Jørgensen, 2008) exist side by side with norms oriented towards a standard. Similar to Excerpt 4 we see that while linguistic features may be corrected, mixing per se, are not corrected. Rather, the correction of linguistic norm transgressions seems to be related to how language use is perceived within a particular interactional context.
Whereas teasing in other contexts can often be characterized as a way to exhibit power, directed from authority persons onto less powerful individuals (Franz en and Aronsson, 2013;Holmes, 2000), Excerpts 4 and 5 show that teasing within the family may be directed at family authorities (or claimed authorities), such as older siblings. Following Billig (2005: 182), the examples illustrate how humorous teasing may challenge social norms and be rebellious: Through teasing, younger siblings may question, discredit, or subvert social hierarchies. Such practices challenge established social relationships and roles within the family (e.g. the often-reproduced age hierarchy between younger and older siblings). Thus, authority positions and power are continuously negotiated, reconstructed and subverted in interactions. At the same time, these fleeting, conversational 'power struggles' also work as a way to produce and affirm family bonds: In the case of this multilingual family, the teases directed at linguistic form contribute to the creation of a shared family identity by playing on common reference points, such as their multilingual biographies and multilingual repertoires, that contribute to the formation of the family as a community of practice.

Concluding discussion
This paper shows how multilingual, adolescent children, through teasing, contribute to reproduce, negotiate, challenge and subvert social roles, identities and hierarchies in the family, and consequently participate in the formation of the family as a community of practice (Eckert and McConnell-Ginet, 1999;Wenger, 2001). Prior work on teasing in the family has focused on the socializing effects of teasing (Aronsson, 2011;Eisenberg, 1987;Miller, 1987;Schieffelin, 1987). While many of these functions are also present in this material, this paper also documents how teasing can be viewed as a peek into the negotiation of social and linguistic norms of a multilingual family. A close analysis of instances of teasing directed at linguistic form shows that the teasing attacks often start off as corrections (playful or serious), but turn into teasing attacks. Following Billig (2005), the examples show how teasing, discipline and authority become connected: In exploiting linguistic errors in teasing attacks, the children in the family regulate each other's linguistic productions, and use metalinguistic knowledge to claim (or challenge) positions in local social hierarchies (Excerpts 1e5).
The interactional achievements of teasing directed at linguistic form are both linguistic and social: On the one hand, the siblings monitor linguistic norm transgressions by correcting (perceived) errors, but by relating linguistic symbols to social attributes, they also reinforce, challenge and even subvert power structures and social hierarchies. In their work on verbal teasing in Køge and Eş kehir, € Ozkan et al. (2015) found that positioning others according to their linguistic competence or linguistic productions were common ways of teasing among the Køge children. By targeting linguistic competence and production, teasers ascribe unfavorable identities and assign negative attributes to others. The excerpts presented here show the acute metalinguistic awareness of the participants, and demonstrate how language negotiations in the family take place in mundane interactional activities, such as teasing. More specifically, they also show that linguistic competence and production are (efficient) targets in teasing practices.
Further, the examples recurrently show how teases directed at linguistic form create temporary alliances where the young family members play with the power hierarchies in the family. Language-directed teasing within the family may be rebellious (Billig, 2005), as it provides the teaser with semiotic material to challenge, or rebel against, other family members' social positions (as shown in Excerpts 4 and 5). Thus, linguistic resources play an important role in social negotiations of the family.
Between the siblings, the position as the linguistically competent one seems to carry a certain value (cf. Excerpts 1, 3, 4, and 5), by providing a possibility to exhibit power and display multilingual competence: For the teaser, correcting a linguistic transgression is a means of displaying competence or knowledge authority (Stapleton and Wilson, 2010;Ag, 2016). By targeting linguistic production in a teasing attack, the young family members display knowledge of locally relevant norms of speech and conduct, as well as family values (e.g. loyalty, speaking respectfully). Additionally, through teasing, the children may claim the right to use, or discredit each other's right to use particular linguistic resources and languages (Excerpts 1, 3 and 4), and further question or challenge social hierarchies. Teasing directed at linguistic form has the potential to work normatively, creating, reinforcing, or negotiating the relationship between language and social hierarchies within the family. In line with previous research on teasing in linguistically diverse peer groups (Lytra, 2007a;€ Ozkan et al., 2015), targeting linguistic competence and production are efficient ways of assigning negative attributes to others and of negotiating social alliances, also within multilingual families.
Regarding the distribution of multilingual resources within families, there is a possibility that certain linguistic norms and ideologies are sustained by the "disciplinary laughter" (Billig, 2005: 205): Siblings who mock the transgression of established linguistic rules display certain normative beliefs. These dynamics shed light on the role of siblings in negotiating and talking language choices and practices (or family language policies) into being in the family, and might even contribute to the understanding of differences in language competence between older and younger siblings (as described by e.g. Yamamoto, 2001;Barron-Hauwaert, 2011). Future, longitudinal and ethnographic research, for example within the fields of language socialization or Family Language Policy, could look further into how children may be socialized into not using a particular language through such interactions. However, the data presented does not provide evidence that teasing necessarily leads to a reluctance to use a language. Rather, the data show that language-directed teasing in this family is not driven by only one linguistic norm: What is sanctioned in one sequence (e.g. mixing features from English, Spanish and/or Norwegian) is not necessarily sanctioned in another, and similarly, the one who is teased for a linguistic error in one case may take on the role of teaser in another (e.g. David in Excerpts 3 and 4). The family members may opportunistically sanction perceived norm transgressions to comment or rebel against, claimed social positions (and particular language use per se). These dynamics tell us that different views of language are negotiated within the interactional course and the setting (cf. Ag and Jørgensen 2013), and that what is perceived as incorrect, and thereby sanctioned through teasing, is locally, contextually, and interactionally bounded. In other contexts, humor and playfulness may constitute frames that facilitate the use of particular languages (cf.De Fina, 2012 Johnsen, n.d.).
As Poveda (2005) argues on the basis of studies of metalinguistic activities in a classroom context, verbal play is used creatively to construct and reconstruct complex social identities and alignments. The present analysis has shown that linguistic resources, metalinguistic awareness, and multilingual competence are important resources in negotiating social relationships between the family members. The display of multilingual competence, metalinguistic awareness, and particular biographical information related to being a transnational family (cf. Zhu and Wei, 2016) contribute to the construction of a multilingual group identity within the family. Åhlund and Aronsson (2015) find that students use stylizations and verbal improvisations to display reflexive attitudes to language and language usage, and to mobilize their peers' metalinguistic awareness and reflexivity towards reflexive aspects of language. Similarly, as shown in the analyses, the children of this multilingual family also identify and assess the linguistic conduct of their co-family members. Through these metalinguistic activities, they participate in the formation and reproduction of linguistic norms and ideologies in the family, create bonds between them, and consequently (re)produce the family as a multilingual community of practice. Whether this metalinguistic awareness in teasing is typical of multilingual families, or if we might see the same attention to language in teasing in monolingual families, is a topic to be explored in future studies.

Declaration of competing interest
The author declare no conflict of interest in preparing this article.