The construction of ‘tough’ masculinity: Negotiation, alignment and rejection1

Drawing on narrative data collected during a three-year ethnography of a Scottish high school, this article examines the construction of working-class adolescent masculinities. More specifically, the analysis focuses on how adolescent male speakers negotiate, reject and align themselves with the hegemonically dominant ideology of ‘tough’ masculinity, the role socially low-risk discourses of ‘tough’ masculinity play in interaction, and how speakers integrate a range of discursive strategies which help maintain homosociality when ‘tough’ masculinity is at stake. I argue that discourses which appear to be about ‘being tough’ do a great deal more social work than might be expected. keywords: glasgow; glaswegian; masculinity; critical discursive psychology; urban ethnography


Introduction
The city of Glasgow, Scotland has long been associated with criminality, violence and anti-social behaviour, with many representations of the city exploiting the image of Glasgow as Scotland's most violent city (Davies 2007). Such behaviour is strongly linked with working-class males and a specific form of 'tough' masculinity which is considered normative within Glasgow and post-industrial urban contexts more generally (Skelton 1997). 'Tough' masculinity has a long-standing social value in Glasgow, and in a city with a celebrated industrial history, 'being a man' has typically been identified with strength, toughness and physical skill. James Reid, a Scottish trade unionist involved in the River Clyde shipbuilding industry, even went so far as to say that 'we don't only build ships on the Clyde, we build men' (Johnston and McIvor 2007: 35).
The construction of 'tough' masculinity is perhaps best realised in the figure of the 'hard man', a working-class male who embodies toughness, a willingness to fight, a propensity towards physical violence, and a disregard for his own personal safety (Whyte 1998). Representations of the 'hard man' in Glasgow are wide ranging, from the razor gangs in McArthur and Kingsley Long's 1935 novel No Mean City, to the 'Big Man' from the Scottish television show Chewin' the Fat, who solves all his problems with violence. But as ubiquitous as the idea of the 'hard man' is in contemporary Glaswegian society, questions remain over how productive it is in discussing the lived reality of working-class men (both adult and adolescent) in Glasgow.
Indeed, despite the fact that the construction of identity among adolescents has been a central concern in recent sociolinguistic scholarship (e.g. Eckert 2000;Moore 2003;Bamberg 2004), how men in Glasgow discursively negotiate the 'hard man' ideology (and laterally the idea of 'tough' masculinity) has been almost entirely ignored. Moreover, while many studies of identity and masculinity have focused almost exclusively on middle-class speakers (Cameron 1997;Edley and Wetherell 1997;Bucholtz 1999;Kiesling 2004, although see Labov 1972, Cheshire 1982and Milroy 1987 for some important exceptions), very little contemporary sociolinguistic research has focused on identity construction among working-class males, Scottish or otherwise. Lastly, there is an assumption within the sociolinguistic literature that doing 'tough' masculinity is a relatively straightforward endeavour which primarily involves explicit acts of violence (cf. Kiesling 1998;Coates 2003). One consequence of this is that 'tough' masculinity is viewed as a homogeneous construct expressed mainly through physical action. This offers a particularly limited picture of 'toughness' as it relates to adolescent masculinities, and with moral panics about adolescent criminality and violence a pertinent issue in recent times, particularly following the riots in Birmingham, London and Manchester in August 2011, it appears to be a fruitful time to critically discuss the relationship between urban masculinities and 'toughness' and how language is implicated in this relationship.
Adopting a social constructionist approach to identity, where identity is viewed as something which dynamically emerges in interaction (Bucholtz and Hall 2005), this article draws on narrative data and ethnographic observations collected during three years of ethnographic fieldwork in a high school in Glasgow to address three main aims: first, to discuss the construction of masculinities among a group of young working-class male speakers, and in particular, how they negotiate, reject and align themselves with the hegemonically dominant ideology of 'tough' masculinity; second, to argue that alongside discourses of 'tough' masculinity, young men use low-risk conversational strategies which help them preserve the principles of homosociality; and third, to show how discourses of 'tough' masculinity can stand in for the deployment of inter-personal violence to establish oneself as 'hard' . As such, this article is a potentially valuable contribution to our understanding of the discursive construction of masculinity among young men.
In the next section of the article, I discuss the concept of 'hegemonic masculinity' , focusing on the concepts of the 'hard man' and 'tough' masculinity. In section three, I outline the methodology and fieldwork site before analysing three conversational narratives to examine how the speakers in these narratives construct and negotiate 'tough' masculinity. I conclude with some comments on the implications these findings have for language and masculinity research.

Hegemonic masculinity in an urban context
Working-class Glaswegian males have traditionally been accorded with a reputation of violence, aggression, and criminality (Patrick 1973;Davies 2007;Kintrea et al. 2011), with one of the most persistent themes in the social history of Glasgow being the 'hard man' (Johnston and McIvor 2007;Young 2007). Ubiquitous in post-industrial cities (Skelton 1997: 352-353), the 'hard man' is an important touchstone and an embedded cultural theme for men in Glasgow. Scholarly treatment of the 'hard man' is, however, almost non-existent, despite its ideological centrality with Glaswegian society.
The status of being a 'hard man' relies a great deal on the intersection of several different practices, including physical strength, fearlessness, a willingness to engage in acts of violence (premeditated and reactive), aggression, toughness, social competitiveness, and (usually) violent reactions against perceived insults. Drawing these elements together, we can offer a definition of the 'hard man' as someone whose configuration of social practices demonstrates engagement with a culture of excessive aggression and violence. Indeed, for many people in Glasgow, the use of violence is a key component of being a 'hard man' and it is the case that violence is often considered to be a hallmark of masculinity (Kimmel 2001: 278), and a necessary part of being respected as a 'real man' (Quinn 2004: 111). As such, the 'hard man' is one substantiation of hegemonic masculinity within Glasgow since it '[embodies] the currently most honored way of being a man [and requires] all other men to position themselves in relation to it' (Connell and Messerschmidt 2005: 832). But like most forms of hegemonic masculinity, very few men are able to enact the practices required of being a hard man, due to the threat of personal attack, the potential legal ramifications of violence, and individual physical and psychological limitations. Additionally, while being a 'hard man' can facilitate social hierarchies, structures of domination and peer-group status (Phoenix et al. 2003: 180;Kenway and Fitzclarence 2005: 43-45), it can also result in a breakdown of social relations, peer marginalisation, peer rejection, and personal injury (Anderson 1997: 18-23;Hawley 2007: 4). Thus, there is a tension between being a 'hard man' and developing and sustaining robust friendship networks.
While the 'hard man' is an acute embodiment of 'tough' masculinity, the role of 'tough' masculinity more generally has been a recent focus in contemporary language and masculinity research. For example, in their analysis of data collected from a group of men undertaking foundational degrees at the Open University, Wetherell and Edley (1999: 342) discuss how men take on three types of imaginary positionings: 'heroic' (the most closely aligned with hegemonic masculinity), 'ordinary' (where speakers emphasise themselves as normal, moderate or average) and 'rebellious' (where men describe themselves in terms of non-normative discourses of masculinity). One of their important findings was that many of the men adopted the imaginary positions of 'ordinary' and 'rebellious' masculinity, rather than 'macho' or 'heroic' , as a way of reinforcing other hegemonic ideals such as individual autonomy and personal choice. This is alluded to by Bucholtz (1999: 444) who argues that 'physically based' masculinities are becoming subordinated in favour of more 'technically based' masculinities. Such distancing from the hegemonic ideals of 'macho' masculinity is surprising, especially given how far 'toughness' is assumed to be a key orientating point for men. Indeed, in their study on adolescent masculinities, Phoenix and Frosh (2001) outline how 'hardness' is an important predictor in determining not only a boy's popularity, but also their sense of self worth as normatively 'masculine' . As I show in the analysis below, however, while 'toughness' might be an important component of adoles-cent male life, it is certainly not the only, or even the most predominant, component.

Methodology
The ethnographic fieldwork on which this article is based began in 2005 after ethical approval from the high school, the University of Glasgow and Glasgow City Council had been obtained. In this section of the article, I outline the Communities of Practice encountered in the high school (CofP hereafter), the data collection process, and the approaches used in the analysis of the narrative data. I also briefly consider the notion of 'identity' as emergent in discourse.

Communities of Practice
As previous research has demonstrated (e.g. Eckert 2000;Mendoza-Denton 2008), language is only one of a range of social practices through which individuals signal their membership of a particular group and construct their social identities. Consequently, in order to investigate the range of practices which contribute towards the construction of identity, including language, the Community of Practice framework was used, rather than the speech community or social network approach. Eckert (2000: 35) defines a CofP as: an aggregate of people who come together around some enterprise. United by this common enterprise, people come to develop and share ways of doing things, ways of talking, beliefs, values -in short, practices -as a function of their joint engagement in activity. Particular kinds of knowledge, expertise and forms of participation become part of individuals' identities and places in the community. Importantly, the use of the CofP framework allows us to go beyond 'topdown' identity categories such as 'working-class adolescent male' towards identities which emerge as socially relevant for the speakers (I discuss this issue in more detail below). Membership of a particular CofP was decided by a process of 'triangulation' (Mendoza-Denton 2008: 240), informed by speakers' self-identification, other-identification, and ethnographic observations of shared social practices and mutual endeavours. Four CofPs emerged during the fieldwork which I named the Alternative, Sports, 'Ned', and Schoolie CofPs (although in the analysis section, I only discuss data collected from the Sports, 'Ned' and Schoolie CofPs). These four CofPs represented the broad social spectrum of the high school, with each group occupying a distinct position by virtue of their differentiated social practices (see Lawson 2011 for more detail on these practices).
While the members of each of these CofPs knew of, and sometimes informally socialised with, one another, the ethnographic fieldwork uncovered significant polarisation between the groups, a finding consistent with previous 'school ethnographies' conducted in the UK (e.g. Willis 1977;Skelton 1997). The primary distinction was between the Schoolie and the 'Ned' CofP who represented the extreme school and anti-school positions respectively. For example, the 'Ned' CofP were involved in the local subculture, including skipping school, participating in a range of age-restricted activities and low-level crime such as petty theft and minor vandalism. They also appeared to be well versed in the gang culture of Glasgow 2 and either knew of or informally socialised with individuals who were involved in gang-related violence (Lawson 2009: 365-367). The Schoolie CofP tended to reject such social practices and instead positioned themselves as pro-school by orientating positively towards the values promoted by the education system. By recognising (and accepting) the authority of the teachers, the members of this group were more fully aligned with the 'establishment'. The Alternative and Sports CofPs formed the 'grey area' between the Schoolie and 'Ned' CofPs, and although not as anti-school as the 'Ned' CofP, they were not as pro-school as the Schoolies. In terms of distinct social practices, the Alternative CofP listened to rock music and participated in nontraditional sports such as wrestling and BMX riding, while the Sports CofP participated in more mainstream activities such as football and rugby.
Over the course of the fieldwork, it became apparent that masculinity was constructed differently across the CofPs encountered. More specifically, members of the 'Ned' and Sports CofPs appeared to construct more 'tough' identities while the Schoolie CofP explicitly distanced themselves from such identities. Focusing on narrative data collected from members of these three CofPs, the analysis below suggests that, contrary to the positions outlined above, 'toughness' is not only (or always) about 'being tough' , and that conceptualising masculinity as static psychological categories of 'ordinary' , 'heroic' or 'rebellious' removes much of the complexity of the moment-by-moment unfolding of identity construction.

Data collection
Like many ethnographic studies in sociolinguistics, the main method of data collection was interviews. Participants were recorded (in conversational dyads or triads with myself present) once they had returned a permission form signed by a parent or guardian, and to ensure confidentiality and anonymity, all participants were given pseudonyms. the construction of 'tough' masculinity 375 Although there are a range of issues associated with the use of interviews in qualitative research (Potter and Hepburn 2005: 285), the difficulties of access and ethics associated with collecting 'naturalistic' data meant that interviews were the only possible method of data collection. Several steps were, however, taken in order to address some of the perceived weaknesses of interview approaches. First, the recordings were conducted after approximately six months of fieldwork to allow the informants to become comfortable with speaking about their lives with someone with no predefined role in the school. This 'lag' also meant that I had background information about the participants' social lives and was better able to draw on this knowledge during the recordings. Second, the recording context was relatively informal to encourage informants to be less self-conscious of their talk. This meant that the first few recording sessions were facilitated with drinks, sweets, playing cards and so on to reduce participants' degree of 'active monitoring' of their speech (the Observer's Paradox, Labov 1972). Participants were also informed that the research focused on 'how people spoke in different groups' (cf. Potter and Hepburn 2005: 290), although the participants were generally uninterested in the aims of the research. Third, I was wary about the recording sessions falling into a 'question and answer' session, so although it was necessary to ask direct questions of the participants to ensure that useful data was collected (for the purposes of the quantitative sociolinguistic analysis presented in Lawson 2009Lawson , 2011, an attempt was made to have the participants guide the conversations themselves, rather than the conversational agenda be established by me. Nevertheless, it is important to note that my presence during the recordings means that we should view the interviews as co-constructed speech events between the participants and myself, rather than simply co-constructed between the interviewees (Rapley 2001;Baker 2004). Last, in order to mitigate the effects of any perceived association with the authority of the school, I did not observe classes or interact with teachers (Eckert 2000: 72-73;Evaldsson 2002: 204).
By the end of the fieldwork, the dataset consisted of approximately 30 hours of fully transcribed conversations (250,000 words), following the conventions outlined in Atkinson and Heritage (1984). Although the speakers used Glaswegian Vernacular, the narratives I discuss have been rendered in Standard English. Distinct Scots lexical features have been retained where possible, and glosses have been provided. My turns are marked as 'RL' .

Analytical approach and 'emergent' identities
Following transcription, the data were coded for salient conversational themes, 3 including fighting, arguing, friendship, life after school and so on. During this process, several narratives emerged as interesting in terms of how the speakers seemed to enact 'tough' masculinity. Since narratives are the vehicles through which speakers perform their 'identity' work (Bamberg and Georgakopoulou 2008), it was decided that these data warranted further investigation using critical discursive psychology, where 'attention to micro-level detail is supplemented with a macro-level layer of analysis in order to focus on the historical, social and political contexts of identity construction' (Benwell and Stokoe 2006: 9). Importantly, within critical discursive psychology, identity is viewed as something socially constructed; as something speakers do rather than something that speakers have (this framework draws heavily on Judith Butler's theory of performativity).
A key debate about constructionist approaches has, however, emerged in recent years, centring on the extent to which the researcher predetermines the categories speakers occupy. For example, Benwell and Stokoe (2006: 56-57) argue that constructionist or 'gender-as-performance' studies 'rely heavily on analysts' rather than participants' categories' , leading to a tautology where researchers start out already 'knowing' the identities of the speakers whose identity constructions they are supposed to be investigating (Stokoe and Smithson 2002: 81;Benwell and Stokoe 2006: 57). In qualitative research, then, it is important to outline under what categories speakers are recruited (Potter and Hepburn 2005: 290).
Since one of the aims of the research was to investigate quantitative patterns of linguistic variation among young working-class Glaswegian males (Lawson 2009(Lawson , 2011, the ethnographic fieldwork focused on speakers who fit this profile (although only speakers who belonged to one of the four CofPs outlined above were interviewed). Importantly, however, the ethnographic fieldwork (outlined above) uncovered socially meaningful and locally embedded 'ways of being' which went beyond the homogeneous category of 'working-class adolescent male' , moving away from identity categories such as 'working-class' and 'male' towards identities which were informed through 'bottom-up' processes. This article, therefore, does not investigate how 'working-class male' identity is constructed through an analysis of 'working-class male' language, but instead, how salient cultural discourses such as 'tough' masculinity emerge in interaction and how these discourses function as part of a wider set of identity strategies (cf. Kiesling 2006).
The (ir)relevance of 'extra-discursive' features has, however, also been disputed in discourse studies (Wetherell 1998). In her discussion of hegemonic masculinity, for example, Speer (2001aSpeer ( , 2001b argues that extra-discursive issues which are not directly orientated to by participants should not form part of an analytical account. In response, Edley (2001b) notes that it is not enough to focus only on the data, and the construction of 'tough' masculinity 377 although 'hegemonic masculinity' may not be explicitly named as such by speakers, 'it is a mistake to imagine that what it describes is entirely absent from everyday talk' (Edley 2001b: 137). Additionally, the use of ethnography helps us to develop 'detailed insight into the concepts and processes that underlie what people do -but that they are often unaware of' (Forsythe 1999: 129). Indeed, given the ideological centrality of the 'hard man' identity within Glasgow, it would make little sense to suggest that this culturally valued way of 'being a man' in Glasgow would not be a relevant issue.

Narrative I: negotiating 'tough' masculinity
In the analysis of the first narrative, I discuss how Nathan and Phil (two members of the Sports CofP) collaboratively construct and negotiate social identities which align with 'tough' masculinity over the course of a co-constructed narrative. The two speakers discuss a key event in the collective memory of their social group (what Georgakopoulou 2007 calls a 'shared story'): a fight between Nathan and Mark (another Sports CofP member). Although Phil, Nathan and Mark were friends at the time of the fieldwork, there had been a fall out between Nathan and Mark which led to a fight between them. Phil attempted to intervene to protect Mark from injury, but was prevented by others from doing so. In the first part of the analysis, I present the opening excerpt of the transcript and discuss how the 'looking good principle' (Ochs et al. 1989) can help us illuminate the importance of self-presentation in the narrative. In the second part, I outline some of the ways in which 'tough' masculinity is constructed collaboratively and negotiated by the speakers. The 'looking good principle' states that speakers 'present narrated events in a way that portrays themselves in the most complimentary light' (Ochs et al. 1989: 244). In following the 'looking good principle' , speakers attempt to present a positive image of themselves to their interlocutors. In Excerpt 1 of the conversation, Phil and Nathan observe the 'looking good principle' by downplaying the negative aspects of their character as being 'fighters' and position themselves as unwilling participants in the event (although as I argue below, both speakers use discursive means to demonstrate alignment with 'tough' masculinity). Phil uses the conditional modal verb would (line 21 and line 25) to suggest that attacking Nathan is something he considered but did not do (a claim immediately countered by Nathan), while Nathan argues that he 'keeps himself to himself ' (line 41). His use of the tag question 'don't I' (line 41) is a way of seeking affirmation and agreement from Phil to bolster his claims. As Ochs and Capps (2001: 137) point out, however, 'there are risks… whenever recounting… a narrative to an intimate: the moral glow may be dashed when someone recalls a rather discrediting background detail' and after being invited to respond, Phil's dispreferred response is prefigured by an almost two second pause (line 42) before he rejects Nathan's statement, pointing out that Nathan 'sometimes causes fights as well' (line 43-44). Taken together, Phil undermines Nathan's attempt to justify his lack of culpability in and responsibility for the fight. Nathan then rejects the idea that he started the fight by claiming that he only fought Mark because he was forced to (line 56-57). There is a degree of similarity here with Andersson's (2008) study of narratives of violence in which Salim, a young man who had been sent to a youth detention centre for assault, explains away his use of violence as 'self-defence' . Such techniques of neutralisation are often an attempt to justify one's behaviour and place the blame on a second party, and we can see this technique deployed in Nathan's contribution to the narrative. The opening sequence of the narrative is also important in that both speakers use this opportunity to initially construct their identities as 'tough' , albeit in slightly different ways. Phil's first contribution (line 24) positions himself as 'heroic' through his attempted intervention in the fight to protect Mark, while his second contribution (line 27-28) furthers an idea of 'tough' masculinity by virtue of the fact that he had to be held back by other people in the group, suggesting that if this had not happened, Phil would have caused serious harm to Nathan. Nathan's construction of 'tough' identity is more straightforward in that he opens with the claim that he 'battered Mark' (line 5), and although the remainder of his contribution in Excerpt 1 is an attempt to explain his actions, in Excerpt 2, Nathan jettisons his attempt at 'looking good' , which up until now has been based largely on the rejection of violence. Instead, he states that the only possible solution to the situation in which he found himself was to resort to physical aggression. When the narrative arrives at its climax and culminates in physical blows, we revert to a presentation of 'tough' masculinity by Nathan which is not mitigated in any way. Unlike earlier parts of the narrative where Nathan attempts to deflect responsibility for the fight, here he emphasises the agency of his actions. Syntactically, Mark occupies the object slot in the utterance (line 1 and line 2 in Excerpt 2) and is the one towards whom action is directed. Moreover, Mark's position as 'object' is highlighted by the fact that in Nathan's narrative, Mark does not attempt to fight back. From my observations during the ethnographic fieldwork, it is unlikely that Mark would have passively accepted being attacked by Nathan since doing so would have resulted in social censure and a potential loss of status. Nevertheless, by glossing over Mark's participation in the fight in this narrative, Nathan attempts to cement his own position as 'tough' , placing Mark in the undesirable position of being considered an ineffective and inept fighter. Nathan also takes up the earlier point from Excerpt 1 that Phil had to be restrained from intervening, adding the detail that it was 'big Peter' (line 11-13) who ultimately stopped Phil. The repetition of this point solidifies the co-construction of 'tough' masculinity for both speakers: Phil's attempts to intervene and the fact that Nathan's behaviour required intervention. Following Nathan's account of him fighting Mark, he questions Phil's attempts to 'look good' , and in Excerpt 3, offers an (implicit) moral evaluation of Phil's actions. Excerpt 3 includes features which would normally be indicative of a cooperative speech style, such as the repetition of the verb 'see' by both speakers and the presence of simultaneous speech (lines 11-14). In the case presented here, however, the conversation is anything but co-operative, with both speakers vying for control of the conversational floor to contest the issue of Phil crying. Nathan's claim is hedged by the fact that he says 'looked like you were crying' as opposed to 'you were crying' , but nevertheless, Nathan calls into question Phil's claims to a 'tough' masculinity, since crying is often seen as an antithetical masculine quality. It is expected (if not demanded in certain communities) that men should not cry, since doing so belies emotional fragility (Migliaccio 2011: 229). Nathan's comments are an attempt to foreground Phil's breaking of social norms and function as a face-threatening attack on Phil's construction of a 'tough' masculinity.
What is interesting about this excerpt is that Nathan appears to contribute two very conflicting statements (lines 2-3 and lines 7-8). He initially states he did not see Phil crying (lines 2-3), a claim strengthened by the adverb 'honestly'. In line 7-8, however, this statement is contradicted when he says 'I did see tears of water dripping from your eyes'. This claim is boosted by an appeal to the external group of peers observing the fight, a tactic Nathan attempts four times (lines 9, 11, 13 and 16). The commentary on Phil's supposed crying episode is further developed in Excerpt 4. At the start of Excerpt 4, I offer a supportive alignment with Phil (lines 1-4), a comment which Phil rejects by pointing out that he was not fighting, the implication being that since he was not fighting, he had no need to cry. Phil also explicitly positions himself as 'protector' when he says that he was 'just going to stick up for [Mark] (line 6). In line 10, Phil challenges Nathan's claim, upgrading his position that he was not crying through the repeated use of 'really' , a fact that Nathan agrees with (line 14). Nathan's agreement here is positioned as a co-operative speech act which shows alignment with Phil's own version of the event. Nathan then restates his two contradictory claims from Excerpt 3: the first, that it did not look as though Phil was crying (line 15), and the second, that it did look like Phil was crying (line 16). The contest between Nathan and Phil on who is 'right' becomes more apparent from line 19 onwards, during which both participants seek to convince the other of their version of events, highlighted through the use of disruptive overlap throughout the excerpt.
Ultimately, however, we have to ask why Nathan produces the contradiction he does. I suggest that it happens because Nathan has to simultaneously manage a critique of Phil's claim to 'tough' masculinity and maintain the relationship. If he had decided to not mitigate his claim that he saw Phil crying, then it is entirely possible that his comments would have been taken more seriously and with potentially dangerous repercussions. Both participants here are collaboratively defending their sense of 'tough' masculinities and in the process, they use the conversation as a way to explore what constitutes 'tough' masculinity and what does not. Importantly, Nathan's mitigating comments offer Phil a safe way of contesting the claims that he cried (thus to counter accusations that he is not a 'real man'), while allowing Nathan an opportunity to further his own sense of 'tough' masculinity, primarily by positioning himself as an arbiter of acceptable masculine behaviour.
What occurs in this conversation is slightly different to what Goodwin (1990: 248-256) and Evaldsson (2002: 218) find in their analyses of boys' story-telling. Both argue that counter-narratives offered by a boy who is under attack generate further counters from the peer-group. In the case of this data, however, the rejection of Nathan's claims by Phil does not entrench Nathan's viewpoint or generate stronger and more insistent claims. Instead, Nathan utilises strategies which mitigate the strength of his claim, even going so far as to contradict himself. The collaborative nature of the conversation becomes even more apparent when we consider Excerpt 5 where Nathan appears to offer a supportive comment that it too sometimes looks as though he is crying. Nathan's comments about crying (line 10-17) appear to be an attempt to validate Phil's earlier claim in the narrative and shows how judgements about apparent 'weak' emotionality can be reintegrated, refashioned and reinterpreted for the purposes of maintaining homosociality between interlocutors. Indeed, the negotiation of 'tough' masculinity in this narrative relies a great deal on indirection and delicacy between the two interlocutors. Both participants are aware that prototypical expressions of 'tough' masculinity (i.e. fighting) could potentially alienate them from their social group (as was the case in other examples where individuals in the high school had fought with one another). Without collaboratively negotiating in the 'game' of 'tough' masculinity, the narrative could have developed in a radically different direction, particularly if both speakers were truly committed to the notion of 'overt competition' and 'one-upsmanship' . For example, Farrington (1998: 19) suggests that many altercations between adolescent males begin with arguments or disputes. Nathan's contributions could have been interpreted by Phil as insulting, resulting in potentially more confrontational strategies which would have run the risk of threatening the friendship. The way the conversation is framed, however, provides both parties with an opportunity to perform 'tough' masculinity without the 'game' going too far.

Narrative II: Personal histories of 'tough' masculinity
The next narrative was collected during a conversation with two members of the 'Ned' CofP, Danny and Will. 4 As mentioned in section 3, of the four CofPs I encountered, the members of the 'Ned' CofP were the most integrated into the local subculture of Glasgow. Their social practices included a range of age-restricted activities such as smoking and drinking, illegal activities such as drug taking, and a knowledge of local gangs and gang-related activity (Lawson 2009: 152-162). As such, members of the 'Ned' CofP were seen by many as the 'hardest' in the school, leading to some pupils avoiding any interaction with them. 5 In particular, a knowledge of gangs and gang-related activity were important indices of group membership, even though I saw limited evidence that members I spoke to were actively involved in any of the gangs surrounding the local area. Nevertheless, gangs remained an important conversational point for a number of reasons. First, gangs in Glasgow are transient, mobile and changeable, so knowing the best fighters, what fights had happened, who the 'hardest' members were, and other demonstrations of 'gang knowledge' conferred a degree of insider status. Second, because knowledge claims about gangs and gang-violence were difficult to verify, status could be negotiated by claiming to 'know the right things' without serious worry of other people showing this knowledge to be demonstrably false. And last, since gangs in Glasgow are generally organised around physical violence and other antisocial acts, members could vicariously attain 'hard man' status through claiming even peripheral membership. The main speaker, Danny, was identified by many people as a prototypical 'hard man' , a status he maintained through outright rebellion against teachers, claims of 'running' with local gangs, and the retelling of a range of fight narratives (recorded both on and off-tape). Prompted by a discussion on Glasgow gang culture, Danny's narrative focuses on his participation in a gang fight. Throughout the narrative, Danny draws on dominant discourses of 'tough' masculinity, but whereas we might expect the narrative to display elements of 'heroic' masculinity (cf. Wetherell and Edley 1999) and to clearly foreground his skills and abilities as a fighter (cf. Coates 2003: 110), he uses the narrative as a way of distancing himself from dominant expressions of 'tough' masculinity. I suggest, however, that he uses his historical involvement with gangs to also reify his identity as 'tough' . Immediately, Danny distances himself from participation in gang-related fighting, claiming that it was something that he 'used to' do (lines 2-4). When asked about why he stopped, he initially does not complete his first response (line 6). Instead, in the following line, he self-repairs to claim that his involvement in gang violence was only restricted to one evening (line 7) and that it was only after this that he stopped. When questioned about why he stopped, we are faced with a complex interweaving of multi-faceted orientations towards 'tough' masculinity. First, Danny states that the reason he stopped fighting was because he 'didn't like it' (line 11), a claim which, on the surface, appears to be a rejection of 'tough' masculinity since 'real men' are expected to enjoy violence and fighting (Lewis 1983). This expla-nation is then rejected for one where he stopped because he could have been seriously injured during the fight by someone wielding a bottle (line 13-14). Danny appears to be searching for an 'acceptable' reason as to why his involvement in fighting ceased. Nevertheless, his two opening contributions suggest an apparent rejection of 'tough' masculinity along two potential axes; a lack of enjoyment and fear for one's own personal safety, both of which contradict the 'hard man' ideology. In line 16-17, however, a sense of 'tough' masculinity is re-established when he admits that during the fight, he 'smashed the bottle and fucking shoved it right into some cunt there' . Here, Danny presents a stark reframing of the situation in which he engages with a form of extreme 'tough' masculinity. The utterance also alters the dynamic of the event to place Danny in the dominant position and his foe to the subordinate position (acutely marked through his use of the insult term cunt). In lines 19-20, he comments that he did not go back to the scene because he thought he had 'almost killed' his opponent, relating this back to a previous occasion where he had been 'done' (charged) with attempted murder. 6 Finally, towards the end of the narrative, Danny alters his presentation of 'tough' masculinity again by admitting that he would run away from a fight (line 41), reverting back to his original stance of rejecting 'tough' masculinity.
Although there are similarities to the narrative discussed in section 4 (i.e. self-defence against a perceived or actual threat), some crucial differences emerge. Unlike Nathan's narrative, which segues into a negotiation of both his and Phil's claims to 'tough' masculinity, Danny's narrative is, I suggest, a sophisticated and dynamic negotiation of 'tough' masculinity which cannot be read as a straightforward substantiation of 'heroic' , 'ordinary' or even 'rebellious' masculinity. Danny states that he never wants to be involved in a fight of that scale again (line 24), that he does not want to go to jail for murder or assault (line 26-27), and that he is more likely to run away from a fight than to confront an attacker (line 41), allowing him to distance himself from 'tough' masculinity. But his association with gang violence, as brief as it was, also allows him to claim a 'hard man' identity. Danny's story here is a complex personal narrative which shows that he is capable of being a 'hard man' , and as such, it is an advertisement of his ability to embody an extreme 'tough' masculinity. The subsequent telling and retelling of the story serves as a 'pre-emptive strike' against anyone who might bother him, with the words standing in and removing the need for similar actions in the future (cf. Anderson 1997: 19). He is able to reject the hard man identity now because he has 'proven' himself in the past.

Narrative III: The construction of alternative 'tough' masculinity
The last narrative shows how Victor, a member of the Schoolie CofP, rejects 'tough' masculinity while simultaneously orientating towards certain aspects of it. The Schoolie CofP was by far the most integrated into the educational system, recognising and acceding to the authority of the school and the teachers (Lawson 2011: 249). None of them, to my knowledge, engaged in any age-restricted activities and were more likely to meet up with one another outside of school to play computer games or practise guitar playing. As such, the members of the Schoolie CofP existed almost completely outside the sub-cultural context of the high school and were considered by many within the school to be 'model students' . While it was certainly the case that many of the Schoolie CofP members rejected the discourse of 'tough' masculinity, Victor's narrative shows a passing familiarity with some of these discourses, and an implicit agreement with others.
The narrative was elicited through a conversation about gang activity in the local area, during which Victor related how he had been involved in an altercation with a group of young men while he was out with Gary, one of his friends and another member of the Schoolie CofP. The previous narrative (not presented here) focused on an event where Victor and his friends were beaten up by a group of boys and Victor did not attempt to fight back. Excerpt 7 follows on after Victor relates the first encounter. masculinity. Yet this positioning as 'victim' is also done in parallel with a partial engagement with ideologies of hegemonic 'tough' masculinity. For example, he states that he was 'trying to make up' for letting his friends be attacked (line 18), an implicit acknowledgement that he is lacking in some way and that he needs to prove himself. Towards the end of the narrative, Victor's negotiation of 'tough' masculinity is further developed when he states that although he was beaten up (line 23), it 'wasn't that bad' (line 25) and 'it didn't actually hurt' (line 29), a claim that is accompanied by laughter (line 28), apparently trivialising the event. His defeat is reformulated in a positive light by a rejection of weakness and vulnerability, and his subsequent reworking of 'tough' masculinity is achieved through a discourse of being able to stand the pain, rather than deal it out. This aligns with previous research which shows that not only is the denial of pain a typical characteristic of 'tough' masculinity (Courtenay 2000(Courtenay : 1389, but also that being able to endure and withstand pain without complaint is reconfigured as a positive character trait (Zeeland 1997: 119).

Discussion and conclusions
My main point in the analysis of the preceding narratives has been that discourses which appear to be about 'being tough' do a great deal more social work than might be expected. More specifically, the article demon-strates how the speakers' narratives do not focus on heroic, against-theodds achievements, but instead contain a great deal of delicacy, nuance and indirection which allows them to maintain homosociality, distance themselves from their past behaviour, or demonstrate an awareness of what it means to be 'tough' . We also have some evidence that 'tough' masculinity is at least partially rejected by some of the speakers. For example, Danny rejects 'tough' masculinity through a discourse of 'I was a hard man, but I'm not any more' , while Victor does so through a discourse of 'I've never been a hard man' . In contrast, Nathan and Phil positively align themselves with 'tough' masculinity in their narrative more explicitly. None of the speakers, however, offer a more general rejection of 'tough' masculinity (to wit, 'it's not good to be a hard man'), suggesting that such an identity is accepted as the hegemonic one for young men in the city.
In terms of the contribution this article makes to a more general understanding of masculinities in Glasgow (and Scotland more broadly), I would suggest that while the 'hard man' is an important cultural concept within the city, it is of relatively limited power insofar as it encapsulates young men's articulations of masculine identities in the city. Indeed, the picture of the 'hard man' as established by the mainstream media appears to be at odds with the kinds of accounts presented in this article. Although the article focuses on a specific set of speakers in a particular location, it nevertheless provides some substance to how young men in the city construct their social identities as men against a backdrop of a hegemonically dominant ideology of 'tough' masculinity.
Moving beyond Glasgow, this article has several implications for how we approach the study of language and masculinity. First, we should reconsider the usefulness of static identity categories such as 'heroic' and 'rebellious' masculinity, particularly since this implies that speakers deploy only one identity over the course of any given interaction (cf. Wetherell and Edley 1999). As the analysis above shows, identity is a dynamic entity which shifts on a moment-by-moment basis, and any analysis of language and masculinity should be sensitive to these shifts. It may be the case that speakers sometimes foreground certain facets of identity, but even in such cases, we should not focus on the foreground at the expense of the other identity work speakers undertake. Second, we have seen that the use of ethnography permits an additional layer of description in the narratives under analysis. Indeed, integrating insights garnered from ethnographic fieldwork means a more fully formed account of the social context the speakers inhabit can be developed. A third related point is that the use of ethnography also allows us to see the relevance of issues which might not be immediately retrievable from the conversational context (cf. Baker 2004: the construction of 'tough' masculinity 391 163; Benwell and Stokoe 2010: 95). While extra-discursive features such as 'tough' masculinity and the 'hard man' ideology might not be named explicitly by speakers, they are nevertheless important in our account of what speakers do (cf. Kiesling 2006: 268).
The research presented here has, of course, its limitations. Of particular concern, briefly alluded to above, is how far the analysis can be generalised to other men in Glasgow. Indeed, generalisation is an acute concern for most ethnographic work (O'Reilly 2009: 82-86), yet it is important to recognise that ethnography helps us bridge the gap between ideological constructs and how these might be embedded in everyday interaction. By investigating the 'local' , we can start to understand how speakers exploit more 'global' resources for interactional purposes and how the same resources might be deployed across different groups. The concomitant use of interviews to investigate the construction of social identity is also a potential area of weakness (cf. Potter and Hepburn 2005), but it is important to note that the interview data formed only one part of the study and that ethnography facilitated an investigation of the kinds of identities socially relevant to the speakers, going beyond the category of 'working-class adolescent male' . As such, the integration of ethnography with critical discursive psychology has helped to develop a more nuanced account of the role of 'tough' masculinity among adolescent male speakers and has shown how 'tough' masculinity is about much more than just being tough.

About the author
Robert Lawson is lecturer in linguistics at Birmingham City University. [[ Simultaneous utterances [ Overlapping speech which does not start simultaneously = Contiguous utterance [info] Contextual information added (e.g. names) (gloss) Gloss of lexical item (( )) Paralinguistic item (.)

Transcription conventions
Pause less than one second (sec) Pause timed in seconds -Speech stops abruptly : Sound is prolonged .
High rising pitch intonation Notes 1 I would like to thank Paul Baker, Scott Kiesling, Ursula Lutzky, Ruth Page, Nicolai Pharao, and Elizabeth Stokoe for their extensive feedback, friendly support and sagely advice as this article moved from the germ of an idea to final publication. I am particularly indebted to the anonymous reviewers who commented on several versions of this article as it made its way through the peer-reviewing process. I would also like to thank audiences at iClave, iGala, University of Lancaster, UC Santa Barbara, and Stanford University for their questions, suggestions and discussion, all of which have helped make this article stronger. Lastly, this research would not have been possible without the cooperation and involvement of the pupils who shared their stories. Thank you. 2 'Gangs' in Glasgow do not follow a hierarchical structure as that which characterises many urban gangs in North America. Instead, 'gangs' tend to be horizontally distributed and established around territorial areas, including local housing estates, parks, and other important boundary markers (Kintrea et al. 2011). 3 The coding process involved, among other things, a close reading of the transcripts and deciding what the topic of conversation was for each speaker turn. 4 Will's turns are all labelled 'inaudible' because his microphone was not properly attached. 5 They also advised me against trying to get to know anyone they considered a 'ned' . 6 Although I was never able to determine the veracity of this statement, it is a substantiation of my point that it is difficult, if not impossible, to confirm or deny the kinds of events Danny narrates here.