Combatting sectarianism from the ground up: The Northern Irish Green and White Army and the football carnival

This paper draws upon participant observation data conducted before and after all 10 qualifying matches for the 2016 UEFA European Football Championship to explore the dynamics underpinning the Northern Irish Green and White Army (GAWA) football carnival. Supplemented by interviews with Northern Ireland fans themselves and informed by the dramaturgical perspectives of Erving Goffman, the paper contends that the anti-sectarian ‘norms’ of GAWA football fandom are dominant in public spaces before matches when the GAWA ‘performance team’ are visible to one another and their audience within a geographically circumscribed space. In a post-match evening context however, various ‘cliques’ of supporters enact their own social rules within a ‘back-stage’ environment of more relaxed peer-to-peer surveillance. These at times transgress the established norms of GAWA fan behaviour. The paper acknowledges that inappropriate supporter behaviour, even within small groups, can damage the wider image of football teams and their supporters in ‘impression management’ terms. But rather than imposing top-down ‘solutions’ which are often devised with limited input from supporters, we suggest that football governing bodies, associations and clubs should look to work in partnership with supporters to identify creative ways in which supporters can be resourced to proactively become ‘norm entrepreneurs’ and challenge inappropriate behaviour from within.

In August 2022, the Premier League in England, in partnership with all 20 Premier League clubs, announced that players would no longer 'take the knee' prior to every Premier League game (Hytner, 2022).Premier League captains instead agreed to 'select significant moments' during the season when their teams would deploy the knee 'to highlight our unity against all forms of racism' (Hytner, 2022).The decision brought to an end more than 2 years of the routine deployment of the symbolic gesture which emerged during Project Restart in the aftermath of the killing of George Floyd and the subsequent Black Lives Matter protests.The broader social impact of the symbolism remains unquantified, but it had become increasingly apparent that football fans themselves were not united in their support for taking the knee as a meaningful way in which to challenge racism (Ibbetson, 2021).Prior to several England warm-up matches ahead of Euro 2020, some supporters booed the players while taking the knee, apparently to register their distaste at some of the political aims of the Black Lives Matter movement rather than anti-racism work per se (BBC, 2021).In the aftermath of this fan behaviour, Show Racism the Red Card condemned the vocal minority of supporters who voiced their opposition to the practice and suggested that ongoing education work is required to continue to tackle racism and other forms of discrimination within the game (SRTRC, 2021).
However, it is important to note that while educational work can be important in the longer term in challenging prejudicial views, the effectiveness of educational programmes on their own among football fans has been questioned (Cleland and Cashmore, 2014;Dixon et al., 2016;Garland and Rowe, 2014).There is also a broader conceptual point whereby such educational approaches assume a logical and linear relationship between attitudes and behaviour (with the former viewed as always driving the latter), which is not necessarily the case (Ryle, 1949;Wittgenstein, 1953).This assumption is particularly problematic for conscious yet non-discursive forms of behaviour driven by a more subtle form of 'practical consciousness' (Giddens, 1984) which may underpin repetitive fan practices such as singing and shouting (see Bell and Bell, 2021).Attitudes are comprised of three related constituent elements, often referred to as Affective, Behavioural, Cognitive (ABC) (Cuadrado et al., 2022).The affective component refers to how one feels about the target.The behavioural aspect is one's predisposition to act in a certain way towards the target.The cognitive dimension refers to one's knowledge, thoughts or beliefs about the target.In this regard, to effect attitudes it can be more practical to focus efforts upon ending discriminatory fan behaviours than trying to stop all fans feeling or thinking negatively about the many 'others' in society (Penfold and Cleland, 2022).But it is also important that attempts to tackle discriminatory fan behaviour involve football supporters themselves (Burdsey, 2014;Stott et al., 2020).Not to do so can fuel resentment among supporters who then lack any sense of 'ownership' of the initiative and subsequently feel that they are marginal to the process (Doidge, 2017).
A focus on improving supporter behaviour (without necessarily changing all aspects of attitudes) has been referred to previously in the pages of this journal with regards to Northern Ireland football fans and their partnership with the Irish Football Association (IFA) through the Football for All campaign (Bell et al., 2020).However, much of this work has been centred upon in-stadium fan behaviour which can more easily be controlled as a result of the panoptical surveillance of 'disciplinary space' (Foucault, 1977).But how does the Green and White Army (GAWA) fan culture of Northern Ireland supporters manifest itself in the football carnival in public space before and after matches?In these instances where 'self-policing' is at a lower-level outside of the stadium, to what extent does this impact upon minority fan displays of sectarianism or racism?
This paper draws upon participant observation data of the fan practices of Northern Ireland supporters before and after all 10 matches of the 2016 UEFA European Football Championship qualifying campaign to explore the dynamics underpinning the GAWA football carnival.Supplemented by interviews with Northern Ireland fans themselves and drawing upon the work of Erving Goffman (1959), the paper contends that the anti-sectarian 'norms' (Sunstein, 1996) of GAWA football fandom are more difficult to enforce in the 'back-stage' area away from the stadium, particularly in bars and pubs.It is here where small 'cliques' of fans will tend to adhere to their own normative rules as they no longer feel the need to assume 'responsibility for the constraints of power' (Foucault, 1977: 202-203).These sub-groups then feel empowered to overtly express views that may be anathema to the mainstream support base.In the Irish context (both north and south) this often relates to support for one or other of the more extreme political perspectives (Irish republican or Ulster loyalist).
The paper suggests that while in-stadium displays are often the focus of targeted efforts to challenge discriminatory fan behaviour, it remains important to monitor the practices of fans (and players) in the informal football carnival as in the era of social media and smartphones, even those who are physically absent can be 'present' (Cable et al., 2022).Inappropriate supporter or player behaviour, even within small groups, can damage the wider image of football clubs and their supporters in 'impression management' terms (Goffman, 1959); but this also has the potential to more negatively impact upon wider inter-communal relations.This is particularly pertinent for societies such as Northern Ireland and Scotland wherein the support base of football teams may have historic links with differing ethno-national allegiances.Within the context of the 'sectarian system' in such societies (Liechty and Clegg, 2001), the actions of a small minority of those from the out-group tend to be viewed as reflective of the majority; yet the in-group is judged much less severely for similar transgressions of behaviour.
When we refer to 'sectarianism' throughout this paper we are referring to those fan practices which directly insult/attack the personal/social identity of others by using derogatory language or symbolism within a Protestant/Unionist and Catholic/ Nationalist context (e.g.referring to fenians or huns). 1 This includes using 'add-ins' with abusive phrases which can transform the meaning of previously innocuous songs, including the Beach Boys song Sloop John B, the tune of which has been used by some fans of Rangers FC to the lyrics of the 'Famine Song' (Millar, 2016).Fan practices which are rooted within a particular ethno-political culture are not in and of themselves explicitly sectarian, but rather an expression of personal and communal identity (which if taken to the extreme may lead to direct sectarian outbursts).While Northern Ireland fans, via the Football for All campaign, sought to challenge some of the worst excesses of explicitly sectarian fan behaviour (particularly with regards to songs sung), that is not to ignore the fact that the majority of Northern Ireland fans still come from a Protestant and Unionist background, perhaps best evidenced by the continued playing of the British national anthem at Northern Ireland matches (and northern Catholics and Nationalists are still more inclined to support the Republic of Ireland despite the best efforts of the Football for All campaign, see Hassan, 2006).Given that sectarianism remains 'a relatively vague and elusive legal term' in Northern Ireland (Jarman, 2012: 2), it is important to bear in mind these definitional parameters when reading the remainder of this paper.

The football 'carnival'
Although Mikhail Bakhtin's (1965) theory of the carnivalesque has long been drawn upon for study in other fields, Richard Giulianotti (1991Giulianotti ( , 1995) ) was the first sociologist of sport to apply the theory to the behaviour of those football supporters who did not seem to fit neatly within the 'hooligan'/'new consumer' binary which dominated the literature until the mid-1990s.Until this point those supporters who wanted to drink heavily, sing raucously and generally engage in expressive bodily comportment at football matches to create an atmosphere through a cultural performance of fandom were ignored (Armstrong and Young, 1999).Within Bakhtin's medieaval carnival, social class or authority only exist to be mocked or ridiculed (Giulianotti, 1995), and the carnival involves the inversion of social boundaries and suspension of traditional norms on what is appropriate behaviour in a public setting (Giulianotti, 1991).As noted by Bakhtin, the carnival is not a spectacle observed by others, but is a lived-in experience in which everyone participates.Carnival behaviour involves eating, drinking (alcohol), singing, swearing, joking, wearing costumes and engaging in sexual innuendo which threatens to transgress 'authorised boundaries' (Giulianotti, 1995: 194; see also, Armstrong, 1998;Hughson, 2002;Jones, 2008).In his seminal analysis of Scottish international football fans, Giulianotti argued that from the 1980s onwards Scottish football fans changed their image and reputation to one of good natured, funny 'ambassadors' representing their country in contrast to English 'hooligans' (Giulianotti, 1995).Prior to 1981, Scottish fans had themselves been associated with fan violence and the attempt to redefine themselves as the opposite to English 'hooligans' led to the eventual creation of the 'good-natured' Tartan Army, with fans engaging in heavy drinking but showing open friendliness to their footballing hosts while 'self-policing' their own behaviour (Finn and Giulianotti, 1998;Giulianotti, 1991).
Other footballing nations have adopted a similar approach in terms of promoting themselves as carnival rather than violent or racist fans, including the Republic of Ireland Boys in Green (Giulianotti, 1996), the Danish Roligans (Peitersen, 2009), the Drillos in Norway (Goksøyr, 1998) and the Dutch Oranje fans ( Van der Brug, 1994).Carnival fans tend to be more prevalent in northern Europe and have been contrasted with the dominant Ultra fan culture which exists in southern Europe (or barras bravas in South America, see Kennedy and Kennedy, 2014).If the focus of the football carnival is on fun, then Ultra fans are primarily concerned with dedication to their team (Batuman, 2012).
Yet the concept of a solely good-natured carnival crowd devoid of many negative aspects has been critiqued as the carnival can also include the abuse of minority social groups (Stallybrass and White, 1986; see also Robson, 2000), paradoxically while including more minorities, such as women (Jones, 2008).It has also been suggested that removal from the public eye, when the carnival is not televised, can have an adverse effect on fan behaviour (Giulianotti, 1991).Indeed, even the positive social identity of the Scottish Tartan Army was primarily defined by anti-English sentiment, with chants such as 'Fuck off England' emerging at certain times (Giulianotti, 1995;see also, Bradley, 2002;Whigham, 2012).In his study of Millwall supporters, Robson (2000) found that the football carnival is neither utopian nor playful, but rather serves to reassert boundaries so that the Weberian 'ideal type' Millwall fan is white, male and working class.Robson suggests that Giulianotti 'glosses over historical and cultural continuity in class dialectics' and that certain manifestations of the carnivalesque may be less benign than others (Robson, 2000: 121).Based upon more than 2000 hours of participant observation of Blackpool, Manchester United and England fans, Pearson (2012: 36) similarly found that there was a complex nature to the make-up of the carnival crowd in which while some fans 'enjoyed chanting racist songs, others refused to join in and would object to the singing of them by others'.Like Brown (1998), Pearson found that the away carnival tended to be less well-behaved than at home.
Although the wider international literature on football fandom has explored the concept of the football carnival and the autotelic aspects of soccer fandom, the literature to date on Northern Irish sport has tended to focus to such an extent on political identity and sectarianism that it has tended to overlook the potentially fun aspects of fandom.While some fans actually enjoy being able to engage in sectarian behaviour, it is of course important to remove sectarianism from football and so the majority of fans, who neither relish nor engage in sectarian displays, should have a central role to play in eliminating it from the game.Emerging evidence recently published within the pages of this journal (Bell et al., 2020) would suggest that the development of the antisectarian Football for All campaign by the IFA and Northern Ireland supporters from the year 2000 onwards led to the creation of the GAWA, which has certain parallels with the creation of the Scottish Tartan Army some 20 years previously.Both can be viewed as fan-led attempts to alter their collective image to an external audience (Goffman, 1959), with a focus on promoting new behavioural 'rules' associated with a football carnival which retain the 'hedonistic excess and psycho-social jouissance' of previous forms of fandom (Giulianotti, 1995: 194), but try to avoid the negative connotations of either football hooliganism or sectarianism and racism.Whether Northern Ireland supporters have managed to achieve this or not (outside of the stadium) is one of the areas explored in this paper.

Northern Ireland fans and Football for All
The 1998 Good Friday/Belfast Agreement signalled an end to 30 years of ethno-national conflict in Northern Ireland between Irish republicans, the British state and Ulster loyalists which left almost 4000 people dead and more than 40,000 injured (McKittrick et al., 1999).While the Agreement itself facilitated a much more peaceful political dispensation in Northern Ireland, it did little to address the core issue of differing political and constitutional allegiances which continues to have ramifications for society and sport today (Bairner, 2013).
Yet as Sugden and Bairner (1993) have highlighted, it was the signing of the Anglo-Irish Agreement more than 10 years previous to the Good Friday/Belfast Agreement in 1985 which was a watershed moment in terms of changing the atmosphere at Northern Ireland football matches.The football team became increasingly important for a section of Ulster Protestant supporters as symbolic of their Britishness and Northern Ireland's continued presence in the United Kingdom (Bairner, 1997).To demonstrate this allegiance to the British state, a contingent of supporters began to sing anti-Catholic songs and started wearing red, white and blue to matches, the colours of the Union flag (Bairner, 2005).Fan sectarianism at Northern Ireland matches reached its zenith in 2001 when Celtic FC player Neil Lennon was booed by a section of the home crowd every time he touched the ball in a match against Norway (Reid, 2008).Unfortunately, Lennon was forced into early retirement from international football in August 2002 after receiving a death threat, purported to have come from the extremist paramilitary grouping the Loyalist Volunteer Force (Reid, 2008).
The treatment of Neil Lennon acted as a catalyst to encourage renewed vigour on antisectarian work between the Irish Football Association and the Amalgamation of Official Northern Ireland Supporters' Clubs (AONISC).Supporters continued with their Football for All campaign to challenge sectarianism in the stadium and created new non-political songs, such as the rather tongue-in-cheek 'We're not Brazil, We're Northern Ireland' and 'Green and White Army' to replace more sectarian songs such as the 'Billy Boys', or more overtly political songs such as 'The Sash'.To challenge the fan practice of wearing the colours of the Union flag to matches, fans were encouraged to wear the team colours of green and white.In 2006, Northern Ireland fans, or the GAWA as they had otherwise become known by then, were awarded the Brussels International Supporters Award by the governing body for football in Europe (UEFA) in recognition of their efforts to challenge sectarianism at matches (UEFA, 2009).
It is important to note that the top-down efforts from the football authorities did not appear to have a lasting impact on the overt sectarianism of fans (see also Bairner, 2009).Rather, the main behavioural changes occurred when fans became involved in attempting to change the accepted norms of behaviour, working creatively alongside then Community Relations Officer of the IFA Michael Boyd and his colleagues, including the late Maura Muldoon.This is an example of real and sustained change happening from the ground up via norm entrepreneurship (Finnemore and Sikkink, 1998; see also Bairner, 2009).Norm entrepreneurship can be understood as a socialisation process occurring when 'principled ideas held by individuals become norms in the sense of collective understanding about appropriate behaviour which then lead to change in identities, interests and behaviour' (Risse and Sikkink, 1999: 11).The transformation of the norms of Northern Ireland fan culture away from sectarianism has been documented within the pages of this journal (Bell et al., 2020;Bell and Bell, 2021).However, despite seeming to adopt elements of wider carnival practices to challenge sectarian and discriminatory fan behaviour, the GAWA football carnival outside of the stadium remains underexplored.It is the purpose of this paper to attempt to rectify this knowledge gap.
The remainder of the paper is structured as follows.The following section discusses some of the key elements of Erving Goffman's (1959) work, particularly the concepts of focused interaction, front/back-stage and the differentiation between 'teams' and 'cliques'.This is followed by a presentation of the research methodology which involved both participant observation and semi-structured interviews with Northern Ireland supporters and IFA staff.The paper then proceeds with a presentation and analysis of the empirical data.Drawing upon Goffman's work, we suggest that the GAWA pre-match public football carnival (in the front-stage) is based upon greater levels of 'focused interaction' (Goffman, 1959) and higher levels of self-policing among fans which limits the potential for violation of the non-sectarian norms of GAWA fandom.However, in the 'back-stage' of bars and pubs where smaller 'cliques' of fans dominate and there is less coordination of fan practices, there is more likely to be a challenge to the dominant GAWA norms.We conclude the paper by offering some thoughts on the wider relevance of the findings for other contexts, with particular regard to the need to pay attention to sub-group fan behaviour outside of the football stadium.
Football fandom and the dramaturgical perspectives of Erving Goffman Goffman (1959) suggested that there are two main forms of interaction.Unfocused interaction relates to all forms of possible communication between actors due to their close bodily proximity in a given location.This includes being aware of individuals with whom you are not directly engaging with but are cognisant of their presence.Focused interaction is very specifically related to the coordination of the activities of two or more people, usually when the individuals concerned are looking at one another and reflexively monitoring their physical comportment, facial expressions and vocal utterances (Giddens, 1984).
Yet Goffman asserted that the individual should not be taken as the sole unit of human interaction.To do so would neglect the relational nature and mutual dependence of individuals who come together to constitute a 'performance team' ('team'), which is defined as an 'action group' and 'any set of individuals who cooperate in staging a single routine' (Goffman, 1959: 85).Goffman devotes chapter 2 of his classic text 'The Presentation of Self in Everyday Life' to 'teams' and 'cliques' in terms of collective 'performances' before an 'audience'.He suggests that a core dynamic of the 'team' is that while a 'teamperformance' is in progress, it tends to serve as a source of cohesion among members who would otherwise hold different views from one another.As a result, any one member of the 'team' may potentially 'give the game away' through inappropriate conduct (Goffman, 1959: 88).Team performances may be guided by a 'Director' who has two functions, to bring back in to line any member whose conduct becomes inappropriate and to allocate roles to personnel and the 'sign-equipment' which is to be used in a 'performance'.Goffman refers to 'maintaining the line' during a 'performance' wherein the 'team' try to not openly disagree with one another in front of an 'audience'.In order for this to occur, 'team-mates need to know and be made aware of what the "official word" is to be part of it' (Goffman, 1959: 93).
Goffman further distinguishes between 'front' and 'back' regions as the settings for human interaction.The 'front' region is usually associated with the face and the front of the body and is often a 'façade' wherein actors play the 'role' and give the 'correct performance' which they feel is expected of them in a given situation (Giddens, 1984).Any deviation from widely held norms in the 'front' region may be sanctioned by others.But in the 'back' region the agent can 'be themselves' and swear, smoke, relax their bodily stance, and not have to uphold the same standards of behaviour as in the 'front' region.This back region is therefore that which is 'hidden away' because it is deemed inappropriate in certain contexts (Giddens, 1984: 124).This paper will explore these concepts of 'front' (public squares) and 'back' regions (bars and pubs) and any possible differential there may be between fan practices in either region.
It is important to note that usage of Goffman's dramaturgical metaphors of 'team', 'performance' and 'audience' in this paper does not imply that fans are somehow not being genuine or are otherwise being deceptive when putting on a 'performance' in support of their team.Rather, they are used to suggest that there undoubtedly is a performative aspect to football fandom generally (Guschwan, 2016), and that fans tend to take in to account the behaviour of their peers and others when acting.

Methods
This paper is based upon data from two sources; participant observation of Northern Ireland fans in public spaces on match-days before and after all 10 UEFA 2016 European Football Championship qualification matches held between September 2014 and October 2015; and interviews with Northern Ireland fans and IFA staff conducted in 2016 to explore the emerging findings in greater detail (n = 25).The research upon which this paper is based was part of a larger research programme which aimed to investigate contemporary patterns of Northern Irish international football fandom (Bell, 2017).The specific aims of this wider research were to explore the meaning and significance of supporting Northern Ireland, to consider the views of fans on issues such as flags, songs, anthems and player eligibility and assess what challenges or opportunities there are to further promote a fully inclusive Northern Ireland team.
Participant observation for home matches typically involved the lead author going to the general area near the home stadium at approximately 2 pm on match-days, about 6 hours before kick-off.This included mingling with fans inside and outside bars and restaurants (the 'hanging about' strategy, see Armstrong, 1993).After matches, the stream of fans were followed into the various pubs in Belfast city centre which they tended to frequent until closing time in the early hours of the morning (although the researcher refrained from drinking alcohol).The extent of researcher participation could be located within Bryman's (2012) six-fold participant observation role typology, somewhere between the 'minimally participating observer' and the 'non-participating observer with interaction'.While the lead researcher was a 'partial insider' with regards to the group under study in terms of being a Northern Ireland supporter (historically a sporadic match attender with one family member, but never a member of a supporters' club), it is also important to bear in mind Aristotle's sentiment that, 'Plato is dear to me, but dearer still is truth' (Flyvberg, 2001: 1).This meant that it was important to attempt to reflexively ensure some distance from supporters, as Elias (1987) suggests, to attempt a 'detour via detachment' and as part of this process the researcher wore 'ordinary' clothes such as jeans and a t-shirt (or shirt/jumper) in the field, did not engage in fan practices and did not socialise with or try and befriend supporters (which is a common practice among those researching football fans, see Pearson, 2012).In this regard it was helpful to view the participant observation element of the research in Gans' (1999) terms as a means of assessing if fans did (and sang) as they said they did when supporting the team on match-days.
The 'Twelfth Man' guide which the Northern Ireland supporters and IFA produce was a very useful resource for away matches as it provides the location where fans will gather in the host city on match-days (in Budapest, Athens, Bucharest, Tórshavn and Helsinki).A similar process unfolded as occurred in the build-up and aftermath of home games, which involved being in these settings, usually public squares, where Northern Ireland fans were beginning to gather in large numbers from the early afternoon onwardswhile afterwards attending the main pubs being frequented by them.
It would have been impossible to inform all fans attending a Northern Ireland match (or those in public squares and bars abroad or in Belfast) of the research, whether that be all 120 fans who travelled to the Faroe Islands, or the thousands of supporters who gather in Belfast before home matches.In this regard as not all Northern Ireland fans were aware that they were being studied, the observation strand of the research could be said to be 'covert'.Given that the observations were focused upon fan culture and behaviour in open settings and television cameras and other media outlets (as well as supporters themselves) were recording fans for a wider audience at home, the public nature of such observations operates at an acceptable level of extraction (Spradley, 1980).The observations were also overt to the extent that when talking and engaging with supporters 'in the field', they were informed of the research taking place.It must however be noted with regards to the limitations of participant observation that there is only so much that one researcher can observe and observations were limited to fan behaviour in public space pre and postmatch in and around the stadium as well as a small number of bars and pubs where fans had gatheredbut there were numerous other establishments where fans had gathered at the same time which could not be observed (although the research endeavoured to include different pub 'types' as far as practically possible including more 'neutral' bars along the Lisburn Road and more loyalist pub types within single identity locations of South Belfast).
Fieldnotes were made on a smartphone.This was a much less obtrusive method of writing down details such as the songs fans were singing, the flags on display and the general 'atmosphere' and the nature of interactions between supporters (Spradley, 1980).When working with fieldnotes, one has to bear in mind that they are always partial and selective reconstructions made by the observer (Emerson et al., 2011).As Schutz (1967) has reasoned, an experience and the reflection upon an experience are not the same thing; indeed, reflections and interpretations can change over time.But despite the subjective nature of fieldnotes and their liminality, their writing both aided the beginning of the process of data analysis and acted as an 'aide-mémoire' when fieldwork was complete (Bryman, 2012: 457).
The strategy underpinning the recruitment of interviewees was purposive and involved maximum variation sampling (Denzin and Lincoln, 2005).While interviewees were predominantly male and from a Protestant community background (which is representative of the main demographic of the support base), the study also recruited female fans (2), Catholic fans (3), a wide range of ages (21-67), fans from differing parts of Northern Ireland, those who were members of supporters' clubs and those who were not, and those who had been directly involved in the Football for All campaign alongside those with no involvement.Participants were recruited in a variety of ways, including obtaining access to those involved with Football for All by staff at the IFA, contacting supporters' clubs directly, and posting online information on the IFA website.Snowball sampling was also used.For example, some members of supporters' clubs after their interviews were able to suggest access to other fans that they knew, who were not members of clubs.Interviews were conducted with 21 supporters between January and February 2016 in the aftermath of the EURO 2016 qualifying campaign.During this period, interviews were also conducted with four IFA staff members.All interviews were digitally recorded with the prior informed and written consent of interviewees.Pseudonyms have been assigned to protect the identity of interviewees.
NVivo 12 was utilised to code and analyse the interview data.It must be recognised that case-study data is context dependent and this impacts upon external validity (Yin, 2009).But while the specific context of Northern Ireland must be kept in mind, there are enough similarities in the workings of the football carnival across national borders to suggest that the findings may be of some relevance for other jurisdictions.

Research findings
The pre-match public carnival Like the Scottish Tartan Army and carnival fans more generally (Giulianotti, 1995;Pearson, 2012), the pre-match ritual for many Northern Ireland fans in public space is to drink copious amounts of alcohol; yet generally behave gregariously and interact with one another, locals and the police in such a way as to project a positive social identity (Tajfel and Turner, 1979).The ludic aspects of the football carnival appear to be even more prevalent away from home among Northern Ireland fans, which previous research has suggested is the case in other contexts (Brown, 1998;King, 1998;Pearson, 2012).Examples include the young 20-something Northern Ireland fans dressed as 'Green Spidermen' in Budapest; the four young fans dressed as Spartan warriors who entered the square in Athens with the Acropolis as a backdrop; the fans who travelled wearing green and white high-visibility workmen jackets to make light of the 'falling down' of the Kop stand at Windsor Park in Belfast; the numerous fans who dressed as green and white 'Santas' in Helsinki as well as the older fans who had dressed as 'Green Beatles' (John, Ringo, Paul and George) complete with inflatable guitars.This is also not to forget the numerous others who wore green wigs, had their faces (and bodies at times) painted green and white, brought with them inflatable animals or insisted on wearing other colourful costumes: Away… there tends to be a bit more of a sort of carnival atmosphere about it, people getting dressed up and wearing silly hats and wigs and costumes and all.There's a bit of a buzz and a bit of banter.I think people make the effort when they go to away matches…there's probably a lot more alcohol being consumed and there's more of that sort of party feeling.(Gareth) Such behaviour is certainly in line with Bakhtin's (1965: 5) own conceptualisation of the medieval public carnival as a form of 'ritual spectacle' which is based upon 'laughter' and is also consistent with Robson's (2000) assertion that such displays are often attempts to present the fan base as positive 'cultural ambassadors' in an attempt to make up for (traditionally) having a poor football team.This colour, noise and spectacle created by the GAWA in the public squares away from home often invoked interest (and at times bewilderment) from both locals and tourists alike who regularly stopped to take photos and videos of fans.Such 'audience' (Goffman, 1959) responses only seemed to encourage fans to sing louder and dance harder: I love chatting to the locals as well like…it's great when they come up to you and sort of say like, in Finland, locals just couldn't believe it.They were like, 'We can't believe how fanatical you are' (Laughs).Like that was really cool too, cos you sort of felt like you are getting a bit of credit for the atmosphere or, you know, that kind of thing.Like the Finland fans were nowhere near as passionate, so…(pause), it made you feel good like I suppose.(Josh) But it should be noted that there is a crucial difference between home and away matches which impacts upon the pre-match 'atmosphere'.Travelling Northern Ireland fans are provided with a guide which informs them where to gather on the day of the match when they are abroad.By organising themselves in such a defined locale and geographically circumscribed space in circumstances of 'co-presence' with one another (Goffman, 1959), fans are able to cater for more 'focused interaction' and better coordinate their activities.To 'perform' as a 'team' the GAWA need to be visible and to 'see' (other fans and locals) and 'be seen' and the spatial dynamics of a public square in a city centre are conducive to this high level of visibility (see also Foucault, 1977): …there's obviously more togetherness built up at away matches before the match.Cos you are all together maybe in a designated square and then you have this sort of traditional walk to the stadium.And then you've guys have got drums and various things and the locals get involved sometimes… the build-up is somewhat different at away matches.(Tony) The tendency for the majority of fans with regards to match-day is to 'exhibit presence' (Goffman, 1959) and their membership of the GAWA by wearing the team colours.This is in contrast to references within the literature to fans predominantly wearing red, white and blue to symbolise their continuing support for the Union with Great Britain (Sugden and Bairner, 1993;Bairner, 1997;Magee, 2005).If green and white are the dominant colours worn by fans on match-day, then the dominant emblems on display are those of the Northern Ireland flag (rather than Union flags) and other supporters' club emblems, often with the name of a supporters' club or the area where the supporters are from emblazoned across it.
The physical settings of the environment and 'sign-equipment' (Goffman, 1959) were drawn upon to set the context for the fan interactions which were to unfold on match-day.This involved flags and banners being placed anywhere fans were able to display them; lampposts, bus-stops, railings, on the ground, even draped over statues.Such symbolic displays are significant both home and away, although there is a particular focus at away matches to publicly display a collective identity before the hosts (Brown, 1998): …like you don't have to have it out at the home games (banner) cos you are not really there to sort of put across who you are, everyone knows they are coming to Northern Ireland to play, you know?(Josh) Alongside spontaneous irruptions of songs, musical instruments (such as drums and saxophones) were also significant in encouraging fans to sing.In Budapest, a group of Northern Ireland fans sitting at an outside bar in the middle of the square appeared to take it upon themselves to act as Goffman's (1959) 'Directors' of the team performance, promptly handing a music cd to the barman which he subsequently played on a loop for the next several hours leading up to the match.The cd contained specific Northern Ireland fan songs such as 'We're not Brazil' and Neil Diamond's 'Sweet Caroline' alongside pop songs such as the Tony Christie/Peter Kay hit '(Is this the way to) Amarillo?' But despite the centrality of the fun and ludic aspects of football fandom, Giulianotti's (1995: 194) focus upon the 'functionally integrative need in the strongest Parsonian sense of the football "carnival"' is not entirely applicable to Northern Ireland fans.Such a position downplays divisions within the fan base and the fact that the football carnival can be both fun and political.While the majority of fans prioritised a GAWA or Northern Ireland specific identity by wearing the team colours and predominantly singing GAWA specific songs such as 'We're not Brazil', 'Sweet Caroline' or 'Green and White Army', a minority of fans continue to focus upon British cultural 'signifiers' (Barthes, 1957) and more ethnic cultural and vocal markers of identity.This pertains to both home and away matches.
Giulianotti makes a similar inference to Parsons in assuming that adherence to general collective norms means an acceptance of them.Yet adherence to a norm can depend more upon fear of external sanction due to the dynamics of power relations than it does on believing a particular course of action is right or morally justifiable.Given that power resides in the social body (Giddens, 1979), particularly away from home in public spaces when the GAWA is numerically at its greatest strength, fans are most visible to one another as a collective (Foucault, 1977) and it is here where there is the greatest level of the 'moral coordination of action' (Giddens, 1993: 116).In these instances, the dominant norms of behaviour are those ordered around the idea of the GAWA and more civic forms of fandom and Northern Irish identity (Smith, 2009).This is also more or less the case in the mini-carnivals which emerge outside pubs along the Lisburn Road in Belfast in the build-up to the match.Yet even within these parameters, some fans still attempt to display symbols more predominantly associated with Ulster unionism or loyalism in Northern Ireland.The bandstand in the square in Budapest is a case in point.Although numerous Northern Ireland and supporters' flags dominated the displays, a Union Flag and an Ulster Defence Regiment (a Northern Ireland based regiment of the British Army which existed between 1970 and 1992) flag were also hung from opposite corners of the bandstand.Such flags may have been in the minority, but nevertheless indicate that a section of supporters retain more than a residual attachment to a sense of Britishness which the Northern Irish specific literature suggests (Bairner, 1997;Sugden and Bairner, 1993).
That Union flags at times sat alongside Northern Ireland or other fan flags, or subgroups of supporters on occasion sang differing songs in close proximity to one another suggests that political displays can sit very much within the wider fun football carnival rather than being antithetical to it.Several examples immediately come to mind of such diversity within the carnival.Although the majority of fans in Helsinki focused upon collectively singing 'Green and White Army' while kicking a football around the square, two fans took this opportunity to ask locals to take photos of them posing with several flags, one of a loyalist flute band and the other displaying the sentiment 'British and proud'.Back in Budapest, as the fans around the outside bar sang 'We're Not Brazil' along with the cd, at the bandstand less than 100 yards away another sub-set of supporters first sang 'Sweet Caroline' before following this up with a rousing rendition of 'Ten German Bombers' which includes the line 'And the RAF from Ulster shot them down'.Similarly, in Athens while awaiting further instruction from the Greek police as to the travel arrangements to the stadium, one group of fans sang 'Stand up for the Ulstermen' in an attempt to 'drown out' another group who were vociferously singing 'God Save the Queen'.Indeed, despite the progress which has been made in promoting a positive form of football fandom through the GAWA, having 'God Save the Queen' (or 'God Save the King' as it has been after the death of Queen Elizabeth in September 2022) as the anthem remains problematic for many fans and potential future fans, as well as players and potential future players.It is important to note however that football is not alone in this regard, problems with anthems also occur in the other two main sports, rugby and Gaelic Games (Hargie et al., 2022).
Fan attitudes to such songs as 'Ten German Bombers' and 'God Save the Queen' are certainly divided, but these examples highlight the complexity within the football carnival, and that fun and politics are not mutually exclusive.The norms of behaviour with regards to supporting Northern Ireland and being a member of the GAWA must continually be (re)produced in praxis, and as such, are continually open to challenge depending upon how those sub-groups of fans who are more inclined towards ethnic forms of fandom can mobilise the rules and resources in their favour (Giddens, 1984).Where the GAWA 'team' cannot gather en masse as a collective and there is less coordination of action, there is more likely to be a challenge to the dominant GAWA norms.Such challenges are more likely to emerge in pubs in private, rather than in public, and will be discussed in detail shortly.However, before turning to this task it is first necessary to assess the significance of the public walk for the GAWA 'team'.

The walk
As fans gather in increasing numbers around the squares or at a particular meeting point as kick-off approaches, fan leaders inform them of the time they are to be escorted to the stadium (or to buses or trains taking them there).Usually as fans finish up their last drinks in the square a musical instrument will strike up a tune to start the procession, and following the local police, the countdown to the match begins.In his role as 'Director' of the GAWA team performance (Goffman, 1959), George noted, 'We would bring the drums and the mega-phone up to the front and then everybody would follow'.The sheer corporeal experience of fandom and sense of being part of the crowd heightens as, typically, two very loud thuds of the bass drum serve as a 'mnemonic device' (Robson, 2000) to which Northern Ireland fans vocally respond with 'Green and White Army! Green and White Army!', while simultaneously clapping and raising their hands to the skies.While small-group conversations and sub-group songs continue sporadically, focused interaction as a collective assumes dominance once again as the GAWA come together in full voice periodically to sing a song from their repertoire.
One of the main rationales behind the Football for All campaign and creation of the GAWA was the desire to change fan practices to make them more inclusive.Part of this has been the internalisation of a positive GAWA 'social identity' (Archer, 2000) which focuses, like the Tartan Army did before them (Giulianotti, 1995), on reinforcing positive norms, being 'friendly', 'good natured' and 'great craic', regardless of results on the pitch.Numerous examples emerged wherein Northern Ireland fans used the public walk to present themselves in such a way.In Finland, a fan ran up a steep hill to give a local child and her mother a Northern Ireland scarf in exchange for a Finnish flag; a gesture cheered both by locals and fellow fans.In Hungary, local and Northern Ireland fans applauded one another while collectively 'booing' (whilst laughing) a passer-by who was wearing an England football shirt.In the Faroe Islands a small group of younger fans broke away from the main crowd and danced with locals while singing 'Stand up for the Faroese' (rather than the traditional 'Stand up for the Ulstermen'). 2Certainly, if Millwall fans were renowned for their 'No one likes us we don't care' attitude (Robson, 2000), part of the collective identity of the GAWA now relies, in contrast to previous perceptions of them in the aftermath of historic episodes of sectarianism, of being liked; of considering themselves to be among the 'best fans' who support their team and enjoy themselves, regardless of results on the pitch.
One incident in particular typified this collective desire to be viewed in a positive light in the aftermath of Northern Ireland's 2-1 victory over Hungary in Budapest.As Northern Ireland fans celebrated a rare away win and drank into the early hours of the morning back at the outside bar in Vorosmarty Square, they once again listened and sang along with the fan cd playing GAWA and other songs with a Northern Irish connection.When Snow Patrol's hit 'Chasing Cars' was heard, more than 50 fans lay down on the ground and sang along with the line 'If I lay here, if I just lay here, would you lie with me and just forget the world?' Yet not long after this public spectacle, a 'false note' was struck in the GAWA team 'performance' at around 1am (Goffman, 1959: 92) when several fans in the square began fighting among themselves, knocking over tables and chairs in the process.In response, a number of onlooking fans intervened to separate the fighting fans telling them to 'Cut that shit out'.Other supporters shook their heads in disgust and responded with a rendition of 'We're the best-behaved supporters in the world' to restore the new norms and rules of behaviour.Winch (1958: 32) argues in his discussion of 'rules' in a sociological context, that '…the notion of following a rule is logically inseparable from the notion of making a mistake', which implies that one must take into account the reactions of others when assessing whether or not social behaviour is in conformity with a 'rule' or not (Winch, 1958).Clearly for most of the watching Northern Ireland fans, these fighting fans had breached the socially accepted 'rules' of GAWA fandom and they were being challenged accordingly: We are kind of good fans.And we know that.So you see a bit of policing when anybody does start getting a bit daft you will see somebody else, you know, pulling him up (sic) and going kind of 'Stop that'… I'd say less so at the game.But certainly more so in the environment in the city when we are wandering round, you know, there is something about emphasising to each other, we are all representatives, we are all ambassadors.(Irene) This example aside, such challenges to the socially accepted 'rules' of GAWA fandom were less likely to occur within the public displays of the football carnival, the 'frontstage' when there is an external audience present to judge the performance of the GAWA 'team' (Goffman, 1959).But what about where there is less of an external audience?The next section of the paper focuses specifically on the 'back-stage' in Goffman's terms, the private carnival in bars and pubs and seeks to provide some answers to these questions.

The post-match carnival
Where the public pre-match carnival (and the stadium itself) provides a focus for the 'regulating gaze' (Foucault, 1977) of the GAWA collective given the organisation of supporters into a highly de-limited space, after the match they tend to (re)associate in smaller groups and 'do their own thing' to an extent.Goffman (1959) refers to these smaller groups as 'cliques' and suggests that they differ from larger 'teams'.A 'clique' of several people usually joins together for 'informal amusement' which is significantly different from a 'team' which attempts to maintain a collective 'definition of a situation'.'Cliques' may 'stage a show' to highlight their distance from the practices of the larger 'team' and those individuals who are in status terms 'above' or 'below' them within the 'team' (Goffman, 1959).
In relation to home games there appears to be a difference between pub types.Within the first pub type, while one may encounter GAWA or political fan behaviour one is less likely to encounter direct sectarianism or Ulster loyalist songs.In these more neutral pubs, fan practices post-match tended to be similar to those during the day; sub-group singing of GAWA songs.However, in the second loyalist pub type, which was attended in South Belfast after the 2-0 home victory over the Faroe Islands, despite the playing of regular pop songs, the playing of the Tina Turner song 'The Best' contained the add in of 'Fuck the Pope and the IRA' by a section of fans in attendance (see also Millar, 2016).The physical environment of the pub was also one which contained British cultural 'signifiers' (Barthes, 1957) and 'sign-equipment' (Goffman, 1959) such as pictures of the Royal Family, the Battle of the Somme, Royal Ulster Constabulary (RUC) plaques and 'historical' Ulster Volunteer Force (UVF) memorabilia.Northern Ireland fans and IFA representatives themselves were also cognisant of the fact that post-match, and particularly within some pubs at night, issues remained with regards to small groups or 'cliques' engaging in sectarianism or fan practices which were not particularly in keeping with the ethos of the Football for All campaign: Like after the Greece game, me and my brother went to the South Belfast NISC and there was loads of people there we knew and it was a nice atmosphere, but it just took -four fellas came in half-cut and started singing some sectarian stuff.And we left, you know what I mean?(IFA 3) In terms of typology of sports spectators, fans such as these who drink heavily and sing offensive songs would be categorised as 'dysfunctional fans' (Lebed and Morgulev, 2022), and if sectarian sub-group fan behaviour is to occur within the GAWA, it is most likely to be found away from home at night in pubs or bars, where the 'field of visibility' (Foucault, 1977) is reduced.Of the five away matches attended during the course of the research, post-match sub-group sectarian behaviour was directly encountered on four of the trips; Hungary, Greece, Romania and the Faroe Islands.Such behaviour took various forms, though on several occasions involved the same sub-groups of supporters, who interestingly, did not behave in such a manner either during the public carnival or at the match itself.
The difference between the public pre-match and more private post-match carnival was most visibly highlighted in Bucharest (albeit before the match as darkness fell) by a large section of fans inside the pub singing along with 'The Best' and other loyalist songs, while television pictures inside the bar showed the GAWA in full carnival mode outside the same pub at the same time singing in full voice 'Green and White Army'.These fans outside the bar were being filmed live for Romanian television while a number of those inside the bar watching them on television were singing loyalist songs.Yet once they re-joined the other fans and the larger 'team' outside the pub to be escorted by Romanian police to the stadium for the match, they returned to singing GAWA songs along with the rest of the collective: Cos they think they can get away with it…because they know it's not accepted at the match… and it will be drowned out if they do it.But when they are in the pubs by themselves they feel like they are in their clubs back in Belfast or wherever they are… It's very difficult to control what someone is doing in various pubs all around Europe.(George) George's point here about 'control' is an interesting one as it indicates that the main tactics which were used in the Football for All campaign to create the GAWA (and of which he played an active part), relied upon the power of the collective, the spatial distribution of fans in the stadium, and the use of instruments to guide the crowd singing, all of which cannot be utilised in pubs in a similar manner (see Bell et al., 2020).Those fans who wish to engage in the singing of such songs appear to be aware of the 'external sanctions' (Giddens, 1984) which they will face in public if they attempt to sing them, and therefore it seems, they rarely tend to do so.For a minority of fans therefore adherence to GAWA behavioural standards is contextual, based upon power relations, and does not necessarily require believing in or morally accepting the dominant norms of behaviour associated with the GAWA (Giddens, 1984).

Conclusion
This paper has drawn upon Goffman's dramaturgical concepts to suggest that performance 'teams' are more predominant in larger fan collectives in public space prior to matches and 'cliques' more predominant in bars at night after games.The significance of the distinction lies in the fact that 'teams' are more likely to conform to certain standards of behaviour before a perceived 'audience' while cliques tend to adhere more to their own norms and standards of conduct (Goffman, 1959).
Yet despite the significance of the fun and ludic aspects of the GAWA football carnival, it does not serve a socially integrative function in a fully Parsonian sense (cf.Giulianotti, 1995).This ignores the fact that fun and politics can co-exist in the carnival and that divisions persist between differing sub-groups of Northern Ireland fans as to whether or not fan practices should focus upon civic or ethnic displays of identity.This paper has identified that it is in the multiplicity of smaller pub-based carnivals where challenges to the established civic norms of GAWA fandom are more likely to emerge.Those fans who seek to transgress the norms of the GAWA tend to do so in such locations as there is a lack of a collective 'normalising gaze' (Foucault, 1977) and they are less likely to receive the same external sanctions for breaching such norms as they would incur should they do so in public space (Giddens, 1984).
The success of Football for All and the Northern Irish GAWA has been in encouraging a positive collective social identity for Northern Ireland fans, which as our interviewee Irene suggested, means that many fans see themselves as 'cultural ambassadors' (Robson, 2000) who will challenge other fans when they are perceived to be transgressing the acceptable norms of behaviour, particularly in public space.The fact that Northern Ireland supporters were listened to and empowered by their national association to challenge inappropriate supporter behaviour themselves (Bell et al., 2020) meant that they felt a sense of ownership over the campaign from the beginning.Had the IFA imposed symbolic stadium displays such as insisting fans raise 'Say no to sectarianism' placards before matches, it is unlikely to have had the same cultural and practical impact.It is clear that this active engagement by norm leaders on the ground to change fan norms has had a very significant impact in relation to behaviour in public spaces, particularly in stadiums and the town squares.However behind the scenes in more 'private spaces' smaller numbers of fans at times continue to display sectarian behaviour, although the fact that this activity has become unacceptable in more public spaces would appear to indicate a successful embedding of the new norms.
But if further lessons can be learnt from the GAWA and Football for All for other contexts, it is that complacency must be avoided.As part of an internal IFA restructuring process, in 2014 several community relations staff lost their jobs due to the end of European Union funding, and the few staff left who were responsible for the delivery of Football for All were incorporated into the technical department.As this paper has demonstrated, for there to no longer be a specific community relations department at the IFA, even if community relations work continues with fewer staff and in another guise, appears to be a somewhat premature step.Issues with regards to improving community relations and tackling sectarianism within a fan base are an ongoing challenge, particularly in the Northern Irish context of the ethno-national divide and high levels