Native and non-native speaker identities in interaction: Trajectories of power

1 Introduction

The native speaker (NS) as a category in Applied Linguistics has frequently been challenged (e. g. Ballmer 1981; Berruto 2003; Coulmas 1981; Davies 1991; Escudero and Sharwood Smith 2001; Gadet 2012; Gadet and Varro 2006; Leung et al. 1997; Muni 2013; Myhill 2003; Rampton 1990; Renaud 1998). Criticisms of the idea of the NS have often pointed to problems in identifying exactly who is a NS as the construct involves variously how languages are acquired, language abilities, language use and identity which conflate linguistic and non-linguistic features. Such critiques make it apparent that the category native speaker is an ideological product. While the idea of the NS may have been problematized by many researchers working in Applied Linguistics, it is nonetheless a widely dispersed notion that has particular resonance as a folk category that has evolved in parallel with the use of the term in Applied Linguistics and which shapes conventional ways of thinking about language users (Hackert 2012).

In many ways the ideology of the native speaker is not a strictly linguistic one. Muni (2013) argues that the NS is a political rather than a linguistic concept – that is it is located with frameworks of power. These frameworks of power are influenced by ideologies of language users’ legitimacy. Myhill (2003) argues that the ideology of the NS involves with it an ideology of authenticity of language use that functions in complex ways with identities and that authenticity implies a hierarchical arrangement of speakers and their affiliations with languages. Much study of the power relationships involved around NS-NNS identities have focused on educational contexts and on the ways that different groups of teachers are constructed politically and ideologically (e. g. Derivry-Plard 2011; Derivry 2006; Holliday 2006; Houghton and Rivers 2013; Kabel 2009; Modiano 2009; Tanaka 2010; Vassallo 2004; Waters 2007). Other studies have pointed out that the NS is often understood in racialized terms, with physical appearance counting as much, or even more, in the attribution of native speaker status than language capabilities (e. g. Amin 1997; Kubota and Lin 2009; Kumaravadivelu 2003, Zhu Hua this volume). Although language education is a highly salient context for understanding the ideological construction of the NS and the consequential construction of power relationships, NS and NNS identities have a much more widespread impact that reaches into aspects of everyday interactions between language users.

This article seeks to investigate the social production of the relative positions of NS and NNS as they are performed in online interactions and trajectories of language use in interactions in which power asymmetries related to control over language are performed by participants. At particular points in interaction, participants may orient to the perceived asymmetries that exist between their language capabilities in ways that bring the ideological construction of NS as legitimate speaker and NNS as less legitimate speaker into play as an interactional reality and make such constructions procedurally relevant to the unfolding of the interaction (Park 2007). This is particularly the case when talk becomes focused on the language of its production rather than on content of the talk (Liddicoat and Tudini 2013). When NS and NNS identities come to be performed in interaction, they create asymmetries as momentary social phenomena that result from the taking up of identity positions in interaction in which the ideological construction of the native speaker and the power relationships that this ideology implies lie at the base of the asymmetry. To examine how power is enacted as a feature of interactions between NS and NNS, this article examines instances of online interaction in which NS and NNS identities become interactionally salient.

2 Data

The aim of this article is to examine a small collection of excerpts from interactions that have occurred as spontaneous events in online interactions in which the learning of the language itself was not the main focus of the online interaction. Online contexts are very fruitful areas for investigating such phenomena as technologies bring speakers into contact across geographic and linguistic boundaries and so allow possibilities for interactions between NS and NNS that are spontaneous and purpose oriented and in which language repertoires are used for meaningful communication. The data comes from interactions on blogs, forums and chat rooms with foci as diverse as technology, holidays and general social interaction. Participation in such interactions is theoretically open to all and participants with a range of different language backgrounds come into contact. In participating in such interactions, all participants come with similar aims and goals regardless of the linguistic affiliations – that is participation in an online interaction dedicated to a particular topic or purpose – and so these interactions are indicative of the many ways that NS and NNS come into contact as language users. Each excerpt therefore represents a moment of lived reality which is captured as it occurs in the course of online communication – it represents a reality as experienced by participants in the interactions. The collection of examples can be understood as a ‘collection’ of interactions (Liddicoat 2011; ten Have 1999) assembled because they reveal facets of a similar social and interactional behaviour. The data are therefore not intended to be typical, representative or exhaustive but rather to serve to explore some of the possible ways that NS power is enacted interactionally.

Much communication on the internet occurs in a social space in which social presence is modified through the removal of live face-to-face communication by the technology being used as participants have the ability to remain anonymous while interacting (Thorne et al. 2015). This means that identities are to a certain extent unknowable, although they can be expressed when participants adopt subject positions in which certain identities are presented as interactionally salient. This article seeks to examine instances when subject positions related to authenticity or otherwise of language use are expressed through interactions.

The anonymity of the internet provides a problem for the identification of individuals as native speakers by participants and analysts, and a certain level of interpretation is required to construct identities as native-speaker or non-native speaker. When these arise in interaction, they are not usually self-ascribed in terms of NS or NNS status but rather are invoked in terms of legitimacy as speakers – that is in contexts where one speaker (NS) claims legitimacy that another (NNS) does not have. In many instances, it is the quality of a participants’ linguistic production that leads to the interactions being studied. In addition, non-native speaker positions can be invoked by participants themselves in discussions of their own personal language learning or comments on their linguistic productions (and reasons for it). The cues for native speaker identity are much less certain as the forms of linguistic production do not indicate particular affiliations to a language. Native speaker identities are not usually explicitly expressed by native speaker themselves and so have to be inferred, and in some cases participants do attribute such identities to others usually in some dichotomous position in relation to their own identities as speakers. In some cases, user names can give indications of language related identities through personal names (although these are rare in the data) or through pseudonyms that invoke particular linguistic or national identities. In reality, however, the native speaker and the highly capable non-native speaker are not distinguishable in any objective way in most internet interactions (unless language learning is the explicit focus of the interaction). I have elected to use the term native speaker for particular identities recognising the problems of such attribution, because the contributions I am discussing position the speaker as an authentic speaker. In some cases, however, it may be the case that a non-native speaker is adopting a native-speaker-like subject position in the interaction.

The anonymity of the internet also provides a particular resource for examining interactions that would not be available in face-to-face interaction. The anonymity of online environments means that participants have less need to attend to social relations because they may feel freed from social constraints by their anonymity on line (Cabiria 2011). Anonymity leads to more freer forms of change with fewer social or interpersonal barriers (Gimenez et al. 2003). This means that otherwise interpersonally consequential or even aggressive actions which may occur less frequently in face-to-face interaction, or which may occur in ways that are more sensitive to face threats may be performed more directly and with less redressive action online (Christopherson 2007).

3 Native speakers’ didactic voice

Liddicoat and Tudini (2013) have argued that the power of NSs is manifested through the enactment by NS of a ‘didactic voice’, that is by adopting ways of speaking that reflect teacherly activities. The special salience of the didactic voice is that it occurs when NS move outside the purposes of the interaction to perform meta-communicative talk. That is, the linguistic form used by the NNS becomes a focus for talk and replaces the content of the talk as the basis for developing the interaction. In so doing, the NS positions him/herself as having both the capacity and the right to comment on linguistic forms. One participant therefore takes up a position as an expert (NS), while the other is positioned as a novice (NNS). In Liddicoat and Tudini’s (2013) data the didactic voice is found in interactions in which NNS participants position themselves as language learners using social interaction with NS to practice the language (and so interventions of a teacherly sort may be invoked by the focus on learning), but such interventions can appear in other contexts where such learning is not the expressed aim of the interactions.

Extract 1 is taken from a blog for people with a general interest in Celtic languages. It is not primarily a language learning blog, although many of the participants identify as learners of various Celtic languages. The discussion at this point is about participants’ favourite Celtic languages. The blog is predominantly conducted in English, although there is occasional use of various Celtic languages. In this extract the language of interaction moves to French and the focus of the interaction shifts the language focus from English and Celtic languages to French. At this point of shift the following interaction occurs between the participants who position themselves in different ways in relation to the languages involved. These participants in the blog are not identified by name ^[1] and it appears to be the first interaction between these participants.

In Extract 1, the first participant presents himself as a NS of French in his chosen name, which is an adjectival form denoting a male inhabitant of a major French city. He enters into the online discussion about favourite Celtic languages with a bilingual posting first in English and then in French. In this way, he introduces French as a possible language of communication in the blog, although without cancelling the relevance of English. His posting receives a response written entirely in French, by a participant who identifies as a learner of French. In switching to French, the NNS is accommodating to the language of his interlocutor and further opening the space for French in the discussion, electing to use it as a communicative possibility. In electing to post in French, the NNS has moved out of a discussion in a language of which he has greater expertise to one in which he is less capable. His posting has a number of errors typical of learners of French – problems with gender, tense, etc. In introducing the reference to his French class, the speaker is bringing into the conversation a category bound activity (Sacks 1992) – language classes are activities associated with the category language learner/NNS – and further constructs himself as a member of this category with an apology for his errors, thus presenting himself as a NNS of the language.

In the NS’s response, the discussion of Celtic languages is suspended and the focus moves to the NNS’s linguistic production of French. The post has moved from talk on a topic to metacommunicative talk about the NNS’s linguistic performance. This change of focus is initiated by the NS, although the post is constructed as a response to the NNS’s down-playing of his own French abilities. In his post, the NS performs three actions that Liddicoat and Tudini (2013) cite as indicative of the didactic voice. First, the NS evaluates the language production of the other. In this case, the evaluation is framed as a rejection of the NNS’s negative self-evaluation (Mais non, ton français est très bon) but with a qualification (Juste 2 détails). In this part of the post, the NS passes judgment on the linguistic production of the NS – a judgment that is both positive and negative. His ability to judge the linguistic production of others is predicated on the ideologically constructed positions of NS and NNS that have become salient at this moment of the conversation and which are brought into being by the NS’s shift from topical talk to metacommunicative comment. Such interactions allocate to the NS the ability, and in some sense the right, to evaluate the language of the NNS simply because of his status as a NS of French.

The NS follows his evaluation of the NSS’s language with correction of some of the NNS’s errors – another teacher-like activity – and associates this with explanations of correct language use – the third type of didactic voice identified by Liddicoat and Tudini (2013). In offering correction, the NS focuses on two problems: the use of pardon in the apology and problems with tense use (present instead of past). Each of these is accompanied by a lengthy grammatical explanation in which the NS claims for himself declarative knowledge of the language that is not available to the NNS, including an account of exceptions to the use of past tense to refer to past time. In making his corrections and providing his explanations, the NS speaks from the position that his identifications of errors is authoritative and that his interpretation of the learner’s linguistic problems is valid. His interpretation of the NNS’s apology presupposes that the intended target is something like Pardon[nez] mon français. [Il] n’est pas parfait maintenant [Pardon my French. It isn’t good now], however the problem could have other origins, for example punctuation: Pardon, mon français n’est pas parfait maintenant [Sorry, my French isn’t perfect now]. In providing correction and explanation, however, the NS assumes an interpretation of the NNS’s meaning and provides correction and explanation based on that assumption. The NNS is thus not positioned as a valid creator of meanings but as an imperfect executor of the meanings that the NS has inferred.

The NNS responds to the NS didactic voice with an appreciation – Merci pour l’information:) – and in so doing ratifies the NS’s right to make metacommunicative comment on his linguistic production. The NS and NNS positions are therefore enacted as relational ones in which language serves as the point of distinction. In this way, the power of the NS to comment on NNS’s language and to adopt a teacherly voice is accepted as legitimate by both participants in the interaction. At the same time, the NNS positions himself as someone who has done the best he can to communicate with the linguistic resources available – he does not use the past tense because it is not part of his current curriculum. In so doing he takes up an identity for himself that is not as a faulty speaker of the language, but as someone who is in the process of learning and thus resists to some extent the NS’s characterisation of his linguistic production.

Extract 2 is taken from an online forum that exclusively uses English. In this posting the first participant is making a point about the shortcomings of technology through an anecdote about an experience he had while trying to get to an English test.

Extract 2: Computer forum

NNS:	Is the contemporary man can do without a computer, and maybe we are such that nothing can be done without a computer. I’m afraid that soon even procreation can not handle without the help of a computer It’s a little joke I love computers but everything has its limits.
	Every had bad days.
	Today such day happened to me
	I live in Wroclaw, is a very beautiful city in the southern Poland, approaching the end of the semester. Today I had a test of English language level B2 2.
	Unfortunately, today was the problem of urban transport, normally I use the trams are fast and do not stand in traffic jams, but today was the problem I reached only half the route, Than tram stopped, and before us another stop 5 trams and not going on while track was free and it was no traffic jam at all. It turned out that the computer system which manages public transport is crashed.
	I had to take a taxi to be on time for the exam, it cost me 9 zł and i was late, and a lot of nerves, I was really pissed off, but i have really good English teacher, he examined me while he really don’t have to do at all:-)
	I’m wondering are our Polish motorman and urban bus drivers are such idiots that they can not ride without a computer? Does your computer have them all show?
	After all in communism time (though I’m too young to remember it, I was born in 1986) drivers and tram drivers, do your job without a computer. Even 10 years ago did not have this system and it did just fine. These were funds from the European Regional Development Fund for infrastructure development Very Happy And today was opposite. Thus i must admit that this system really improved our mass transit, the bus and trams, are more reliable and better keeping schedule than before, especially the electronic display showing actual arrival on every tram on bus stop are really cool
NS:	I have been taking courses in transportation engineering for my masters degree. Your English is very good, although you tend to leave out what are called “direct objects” and “indirect objects”, but what you are saying is clear enough….

By raising the topic of the English test, the participant has positioned himself as a NNS of English, as taking such tests in a category bound activity associated with learners/NSSs, although this is incidental to the point he is making. Like the NNS in Extract 1, he too produces a number of errors, but says nothing about his self-evaluation as an English speaker – that is, his positioning as a NNS is discernible in both his story and in his linguistic production, but is not something that he foregrounds. The NS responds initially by continued topic talk but then moves into a didactic voice, evaluating the NNS’s production (Your English is very good). This move to evaluative languages is outside the established purpose of the forum – the discussion of computing – and so the NS creates a space in the interaction to comment on the linguistic production of the NNS. He then indicates that there are problems with the NNS’s linguistic production but in this case does not offer specific corrections. Instead, he provides an explanation of the nature of the problem – leaving out objects. This explanation contrasts with the more learned grammatical explanations of the NS in Extract 1. It is not so much an explanation as a labelling of the problem. Moreover as a characterisation of the NNS’s production is not particularly accurate as his linguistic production not primarily affected by problems with objects. What is interesting here is that the NS claims declarative knowledge of the language but appears not to have the type of scholarly language knowledge he claims. Rather he makes his analysis on the basis of the ideological assumptions of legitimacy that are associated with the NS position he has adopted and his relative identity in relation to the NNS.

Extract 3 is taken from an Italian language chatroom in which the a learner of Italian is participating as a language practice requirement for her course, although the chatroom itself is not designed as a language learning site. That is she is participating in a chatroom designed for social interaction among speakers of Italian. The two participants have moved into a private conversation in which the learner’s life as a student of Italian has been discussed, thus the discursive positons of NS and NNS have been introduced as a feature of the prior interaction.

In Extract 3, the NS compliments the NNS. The compliment is personal and marked as sincere (veramente). In responding to the compliment, the NNS chooses to downplay it and specifically rejects the designation ‘intelligent’ because of her poor level of Italian. In this way, she constructs her positioning in relation to Italian as one of deficit orienting to the ideological construction of NNS as less legitimate speakers. The NS however rejects this downplaying but stating positive assessments of the NNS’s spoken and written language. In Extract 3, the evaluation is not initiated by the NS but by the NNS, who rejects a previous positive assessment because of her language performance. In this way, the NNS can be understood as inviting the compliment from the NS and so indicates that for her positive assessments of her language capacity by a NS is welcome and affirming. In this way, she positions herself as a communicative inferior to the NS and constructs her linguistic performance as something that can legitimately discussed.

In these three extracts, NSs in different contexts employ a didactic voice and take on activities that are regularly associated with teacher talk and expert to novice communication through three metacommunicative activities:

1. NSs evaluate language production of NNSs (Extract 1, Extract 2, Extract 3)

2. NSs provide explanations about language use (Extract 1, Extract 2)

3. NSs provide correction of NNSs’ language (Extract 1)

The assumption of a didactic voice by NSs is assertion of NS power over the linguistic code as it invokes their legitimacy as speakers and correspondingly the less legitimate production of the NNS. That is, the NS can only legitimately adopt a teacher-like position in the interaction because of a perception of distinction between the participants in the interaction. It represents a particular form of control over interaction in which the NS can suspend topic talk (Extract 1 and Extract 2) – that is the interactional action underway in the talk between the participants – in order to comment upon the linguistic form of the communication. This intervention in the interaction is different from an instance of repair (Tudini 2010), in which topic talk is interrupted to deal with problems of comprehension and so allow for talk that is required for topic talk to be resumed. Instead, the NS refocus the talk away from its current interactional focus in order to perform actions in relation to the NNS’s linguistic production that presuppose a differential relationship between the two and construct the NNS’s linguistic production as something that can be legitimately discussed.

The legitimacy of NSs’ interventions in talk to comment on NNSs’ linguistic production is also seen in the ways that NNSs act in such contexts. In both Extract 1 and Extract 3, the NNSs orient to their status as less legitimate language users either by introducing category bound activities that relate to their NNS status in the communication and/or by apologising for their NNS production in the communication (Extract 1). That is, NNSs, through their language practices may also suspend topic talk to emphasise their problematic presence in the interaction and account for their linguistic performance. When NSs intervene to suspend topic talk and enact a didactic voice learners can ratify this as a legitimate contribution for example by accepting or thanking the NSs for these interventions (Extract 1 and Extract 3) thereby constructing the intervention as both needed and desired by the NNSs. In so doing, the NNSs are orienting to the ideology of the NSs’ greater legitimacy as speakers. In the instances above, the NNSs’ responses do not appear to perceive the correction as a face-threatening act or as in any way unusual and thus treat the intervention as legitimate activities within the context in which they occur. In this sense the power of the NS is hegemonic in that it shapes the understanding of both NSs’ and NSSs’ views of each other as language users. The didactic voice adopted by the NS is not contested and the move from topic talk in relation to the purpose of the internet groups’ primary focus (most notably in the Celtic languages and computing sites) is not treated as irrelevant to what is happening. This can be seen as an enactment of the perceived “complete and possibly innate competence” of the NS (Pennycook 1994: 175).

4 Consequences of NS power in interaction

The didactic voice is not simply an interactional accomplishment of NS-NNS interactions. It is a consequential structuring of the relationships between NSs and NNSs as they exist in interaction that allocates to the NS a power over and a legitimacy in language use. This NS power over language can have consequences for NS participation beyond aspects of the talk itself. In Extract 4, which is taken from a forum in which international students studying in English language universities discuss their experiences. The topic of the interaction at this point is speaking in class and the category NS is explicitly introduced by the participant. In the post, the NNS highlights how the awareness of the NSs’ power in interaction affects how she participates in class.

Extract 4 shows that events such as those performed explicitly in the interactions above may be perceived as potentially salient in interactions more generally. ^[3] Here the NNS explains how she is silenced in classroom interactions by the presence of NSs. She is not responding to overt behaviours by NSs in her class but rather to the possibility that they may enact internally aspects of the didactic voice that were discussed above. In this case, she is responding to the possibility that NSs will evaluate her linguistic productions and, in so doing, will discount her legitimacy as a speaker of English. She does not question the right of NSs to make such evaluations, thereby seeming to accept the inequalities inherent in interactions between NSs and NNSs. This inequality for her is disempowering as it prevents her from speaking – her fear of making mistakes makes participation in the class and the expression of her ideas impossible. Moreover, NSs’ evaluation of her is a threat to her own sense of identity as a teacher of English and so her NNS status throws into question her professional identity, indicating that a native-speakerist ideology (Holliday 2006) pervades her thinking both professionally and personally.

Extract 5, which is taken from the same forum expresses a similar feeling of being silenced by the potential for NSs to evaluate NNSs’ linguistic production.

In this extract, the NNS specifically invokes the difference between speaking with other NNSs and speaking with NSs. In interactions with NNSs, he feels a sense of equality because there is no feeling that one speaker has greater legitimacy than the other or that one speaker will evaluate the language of the other in unfavourable ways. For this student then the possibility that NSs have the power to evaluate linguistic performance and draw inferences from it about speakers by virtue of their status as NSs becomes a source of anxiety that makes the NNS less willing to speak.

This anxiety is expressed even more strongly in Extract 6, which is drawn from a discussion forum for learners of Japanese.

Here the participant expresses her feelings of using Japanese in front of NSs in very strong terms using words such as ‘terrified’ and ‘afraid’. She contrasts her ability in relation to others (I do know more than the average American) with her inability when viewed from a NS perspective (I’m not fluent) and her sense of lack effectively discounts any ability she may have. The source of the problem is again not the actual behaviour of the NSs involved but rather her internalised perspective of the NS as an evaluator of linguistic production and thus of legitimacy in using the language that expresses itself as embarrassment. This sense of embarrassment in face of the NSs’ potential power to evaluate language has therefore restricted her to a single one-off production of a widely known formulaic utterance ‘arigatou’, which hides her ability the speak more Japanese ‘than the average American’.

These extracts show again the hegemonic nature of the NS ideology and reveal that the power of NSs over language is potentially always present for NNS and that the possible deployment of the didactic voice by NS can have significant impacts on communication that lead to the silencing of NNS voices and anxiety about communication. This indicates that aspects of learners’ language behaviour understood as ‘willingness to speak’ or ‘willingness to communicate’ (Peng 2012; Yashima 2012) is situated in a complex power relationship which has the potential to emerge into interactions as a salient way of understanding participation that influences the nature of communicative practices.

5 NNS resistance to NS power

In some cases, NNSs may resist the exercise of NS power over language by rejecting NS versions of their own communication. In resisting power, NNSs enact positions in which they reject the NSs’ claimed right to comment on their linguistic productions particular ways. In Extract 7, which is taken from an online forum in which participants are discussing holidays, and in a sequence of posts, the linguistic form of one participant’s contribution becomes the focus of talk.

In the first post, the participant responds to the photographs that have been contributed and then moves into the didactic voice, positioning herself as a knowing and legitimate speaker (NS) and assuming both the capacity and right to comment on another’s language by making a negative evaluation of the other’s (NNS’s) English. The NNS explicitly rejects the validity of the NS’s criticism and the positioning that allows this to be done. She normalises the status of NNS as language users and the linguistic elements of her production that go with her status as both learner and user of English. She positions herself as a multilingual individual and in so doing asserts an identity for herself that is superior to that of her (imagined) critic – the monolingual NS. In responding to her critic, she valorises a multilingual identity over a NS identity and uses the additional cultural capital multilingualism provides her when compared to a monolingual to reject the NS’s assumed right to offer criticism of her language. Rather than accepting the NS’s competence to comment, she recasts the NS as less competent. The NNS counterattack shows that the behaviour of the NS was perceived as a face threatening act and as socially problematic and challenges the didactic voices as the natural prerogative of a NS. This contrasts with Extracts 1–3 in which NNS collaborate in the act of the NS and even express appreciation of the intervention.

In Extract 8 participants are discussing alternative ways of using software programs to resolve particular computer problems in a technology oriented forum.

Here, the first speaker has produced a statement that is not grammatically correct and the other adopts a position of greater expertise and corrects the language in his post. In so doing, he adopts a position of the NS as legitimate and authoritative speaker and has moved the focus of the interaction on computing issues, the official focus of the site, to a focus on the language production of the NNS. In offering the correction, he not only presents a corrected version of the NNS’s language but also an interpretation of her meaning. The NNS rejects the NS’s version of her own communication, although she acknowledges that her English is potentially problematic. The disagreement seems to relate to how she understands the way the negative marker works in the utterance. She rejects the idea that the negative applies to her thinking process but rather to the action she is thinking about (I think I’ll not do it like that). Her resistance here is to the accuracy of the NS’s formulation of her intended meaning and to the NS’s assumption of the role of interpreter of her meanings in English. She attributes the problem to the NS’s comprehension of her meaning not simply to her own problems in articulating the meaning and indicates that the correct response to problems of comprehension is not to propose understandings of her talk but to ask for clarification of the meaning (if you don’t understand you ask me).

A further example of how NNSs can resist NSs’ taking on a didactic voice be seen in Extract 9, which is taken from the same chatroom context as Extract 3, although with different participants. In this extract the participants are closing a conversation by discussing their plans after their chat.

Extract 9: Italian language chatroom

NNS:	dove vai stasera? Where are you going this evening?
NS:	non so… a rompere le scatole a qualche signora I don’t know… to hassle some woman
NNS:	rompere le scatole? che significa? Romprere le scatole? What does that mean?
NS:	a disturbare To bother
NS:	in senso buono In a positive sense
NNS:	ah
NNS:	allora, in boca al lupo! Well, good luck!
NS:
NS:	bocca
NSS:	lol si ho conociuto mio sbaglio. sono stanca, devo alzarmi in quattro ore I knew my mistake. I’m tired. I have to get up in four hours.
NS:	ok allora ti salute Ok then I’ll say goodbye
NS:	ci sentiamo domani forse We’ll catch up tomorrow perhaps.
NNS:	grazie molto per aiutarmi Thanks very much for helping me.
NS:	niente It’s nothing
NNS:	si forse! buona sera Yes perhaps! Have a good night.
NS:	ciao
NS:	ciao

In responding to the NS’s description of his plans for the evening, the NNS wishes him luck (in bocca al lupo!), but misspells bocca as boca. The NS corrects this, providing the correct spelling, and acknowledges humour in doing so with an emoticon. This correction constructs the NNS as someone who has less competence in the language and this characterisation is rejected by the NNS. She knows she made a mistake and therefore her spelling cannot be understood as a lack of competence, but as a performance error motivated by tiredness. That is, she constructs her misspelling not as a NNS’s behaviour but as the regular flawed behaviour that is possible for all speakers because of circumstances. She therefore characterises her mistake as conveying no less legitimacy for her as a language user than a similar mistake that could be made equally by a NS under the same circumstances. In Extract 1 above the NNS also rejects to some extent the validity of the NS’s assumption of the didactic voice when he states that the correction is not appropriate because it represents something that he is not expected to know.

Thus, NSs’ power as exercised through the use of a didactic voice in interaction can be resisted by NNS as not being a legitimate action in the interaction underway. They also reject the NS’s assumption of a right to be an arbiter of form or meaning in interactions. In making such rejections they invoke alternative understandings of their competence in the interaction: as a multilingual individual with a greater language capability than the NS offering correction (Extract 7), as a more legitimate interpreter of personal meanings (Extract 8) or as a language user who makes mistakes but whose mistakes are not the result of NNS status (Extract 9).

6 Conclusions

This article has argued that in NS-NNS interactions there is a possible interactional trajectory through which the ideological underpinnings of NS and NNS positions in interaction are brought into saliency and through which the relative power of NSs and NNSs as language users comes to shape the nature of interactions. In interactions with NSs, there are two potentially salient positionings for NNSs: language user and language learner (c. f. Kern and Liddicoat 2008). As a communicator the NNS acts as a language user. When, however, the focus of interaction moves from the content of communication to the form of communication, this reconstructs the user as a learner and as a deficient communicator. When this is the case, the NS and NNS are brought into an inequality as speakers of the language that is based on the ideology of the NS as the ideal, competent language user:

Those born into the matrix of nation and language can often invoke, in conversations with someone foreign to that matrix, the notion of a birthright of linguistic authority, an authority that is configured as an infallible innate sense of the acceptable utterance.

(Bonfiglio 2013: 29)

In these interactions, the NS takes up a position as having greater cultural capital (Bourdieu 1982) than the NNS by right of his/her knowledge of the language as NS. In such instances, linguistic affiliation is used as a basis for differentiation and to construct an asymmetry that confers status on the NS on the basis of the NS having something which the NNS as learner is presumed to want and to lack – linguistic competence. The power is manifested when the participants take up NS and NNS positions, which become salient as a result of a focus on the grounds of distinction between the participants – language – through a recasting of the communication as being about linguistic performance. Although it is when linguistic performance becomes the focus of talk that NS power is most clearly visible, nonetheless, this power may also shape ways of interacting that are not brought so obviously to the surface as Extract 4, Extract 5 and Extract 6 show. In these extracts participants are reporting the impact of the possibility of NS taking on a didactic voice, even internally, while to online interventions by NNS are self-selecting interventions into interactions that recast them as evaluations of language. The interactions suggest that NS may assume a right to intervene on the linguistic productions of NNS because of their greater legitimacy as language users and language knowers and that NNS may collaborate with them in such interventions, or at least consider these interventions to be warranted in some way. In this way, the power of the NS in such interactions is hegemonic in the sense that both NS and NNS assent to the same ideological construction of their interactional worlds (c. f. Gramsci 1975) – that is both NS and NNS act in a way that constructs the relative status of NS and NNS in the same way and understand the grounds of distinction as being a natural component of the language biographies and the identities they bring to the interaction. This does not mean, however, that such hegemonies cannot be challenged as Extracts 7–9 show.

These identities come to relate to legitimacy as speaker and the relative power that each has as custodian of the language and have consequences for the organisation of participation in interactions that can silence the NNS as participant in interaction. Joseph (2013) argues that the NS idea is inherently silencing and that the power of the NS is an historically developed habitus that has prevented speaking because of the power associated with the emergence of the Chomskyan idea of the NS as the ultimate arbiter of language – the ideal speaker-listener. Joseph argues that the idea of the NS has worked to silence even those who would claim entitlement to be considered as such. The idea of the NS as linguistically all-capable can have an increasingly silencing effect on NNS, who may not only feel most strongly their distance from the ideal but may also experience the didactic voice of the NS interaction as a legitimate interactional intervention. Abouchaa (2012) argues further that the ideal speaker-listener is a monolingual construct that has constituted multilingualism as an abnormal linguistic repertoire and thus renders the NNS invisible within linguistic and social discourses about language except in terms of their deficit position in relation to the monolingualised NS. The NNS’s act of resistance to the NS’s correction in Extract 7 can be seen as an interactional response to such a deficit position.

This article has discussed a practice through which NS power is performed at the micro-level in local contexts of interaction. In doing this, it has aimed to demonstrate that the power of the NS is not simply something that is relevant to understanding relationships at the macro-level of society, for example in the form of policy and practice in language education, but is something that can be relevant and shape social relations in many contexts. As such, academic discussions about the validity of the concept of the NS represent only a small part of the ways that the ideology of the NS influences language use (c. f. Hackert 2012).