Emotions,Violence and Social Belonging: An Eliasian Analysis of Sports Spectatorship

Introduction

Following a figurational approach (Elias, 2000[1939]; Elias and Dunning, 2008[1986]), this article examines how aggressive impulses and actions by spectators at sporting encounters become more individualized and less spontaneous in the context of changing feelings of belonging towards the state. These changes also occur as people become less dependent on intergenerational inheritance and other forms of local dependence. Spectator violence shifts in structure over time in terms of situational dynamics. In the sport, period and place examined – Gaelic football ¹ from the late 19th century in Ireland – the socio-temporal nature of violence changes. Initially, games were prone to inter-player violence, which often produced almost immediate ‘spectator’ responses in the form of collective pitch invasions and assaults on players. Gradually, violence at matches shifted to collective violence after the match had ended, with the referee becoming the main target of spectator anger. The socio-temporal structure of the situation again changed to a more individualized form of violent action, though again typically directed at this neutral functionary.

We explain these relative changes in the context of broader social processes, including state breakdown, national distinction and the gradual effectiveness and legitimization of the new state. While figurational theory stresses the efficacy of state agencies of violence control, we also argue for the importance of feelings of belongingness towards the state as a symbol of group cohesion. In order to distinguish the figurational argument, we compare Elias’s theories against the interactional approach of Randall Collins (2004, 2008). There are many parallels (particularly in the focus on emotions in violent encounters), but Collins’ avoidance of habitus weakens the connection between broader social processes and those occurring within immediate interactions involving physical co-presence.

There have been numerous publications, including those within the figurational tradition, on the question of spectator violence, but these have tended to focus on more organized forms of violence. Here, we are not concerned with hooligan gangs and the conditions of their emergence and maintenance. We address emotional experiences encompassing spectator aggression and violence at the games themselves. As Bairner (2002: 128) notes, there is no tradition of organized hooliganism related to Gaelic football in Ireland. Dunning et al. (1988), following Elias, argue that hooligan groups are attracted to soccer because of its oppositional, masculine character and the opportunity it affords as a site to engage in violence. In other words, soccer serves the function of providing spaces for exciting violence. Violent encounters are often deliberately organized prior to the games themselves. Indeed, the violence can have a rather tenuous relationship to the game and sport. They do, however, find that spectator violence before the First World War was related more to the flow of the match, and referees were often assaulted. In Ireland, spectator violence at games has been, and remains, primarily related to game events themselves.

The article is organized as follows: the figurational (Elias) and interactional (Collins) perspectives are compared; the methods employed are briefly outlined; the changing situational dynamics of spectatorship are discussed; and such changes are explained in the context of processes of state breakdown, state reformation, uneven state legitimization, and broadening and multiplying social interdependencies beyond sporting situations themselves.

Civilizing Processes, Emotions and Forward Panics

Elias (2000[1939]) argues that, as people become enmeshed and enabled through expanding social interdependencies over many generations, codes of conduct become more demanding and differentiated. People have to learn how to attune their conduct in the navigation of multiple, overlapping and often contradictory social relationships. More demanding social controls induce emotions of varying intensities and forms, according to the relative success of meeting such demands. As social change is constant, though not uniform in pace or direction, encompassing new social relations and attendant compulsions and constraints, this is a never-ending process throughout life. But key social and cultural codes are learnt early in life – a process producing a particular habitus, or second nature (a learnt mode of behaviour, thinking and feeling that has been so ingrained in habits and routines that it feels natural or unlearned). Of particular concern here is the emotional dimension of habitus, the relative impulsiveness or control of emotional experience, and in particular the interweaving of feelings and actions that constitute emotional experience. A more self-steering habitus develops in the context of lengthening and multiplying links of interdependence, which both depend upon and further impel an increasing monopolization of violence within a given territory. The concept of habitus has been popularized by Bourdieu, and tends to be associated with self-reproducing mechanisms of social structure. Elias uses it in a more straightforward way, connoting the outcomes of the social constraints toward self-restraint. The emphasis on restraint suggests some Freudian repressive hypothesis, but habitus formation does not imply docility; rather it is the crystallization of fluid social relations. Recurring violent social relations tend to produce an emotional habitus predisposed to quick aggression.

Elias developed the concept of figuration to connote the changing network of mutual dependencies bonding people together. Figurations can be conceived on many levels, from the dyadic conversation to the relations between political states. The concept addresses the form, strength and extent of dependencies between people – who do we need for which purposes? The purposes themselves are relational outcomes from earlier figurational developments. For example, we increasingly need organized forms of excitement – like sport – as life loses some excitement and unpredictability due to the greater routinization of the occupational and domestic spheres (Elias and Dunning, 2008[1986]).

While habitus refers to psychic processes ‘within’ the person, figuration connotes social processes between people. These psychic and social processes are inseparable. The habitus is formed primarily in social relations at home, school and play spaces. But as the person grows and adopts new roles and functions, the individual habitus enters new situations and sets of dependencies. In periods of rapid social change, the habitus can be ill-suited to new forms of social conduct expected in different relations. Habitus is always in process, but not always in harmony with emerging social relations. While Collins’ (e.g. 1981, 2004) work has consistently engaged the micro-macro connection in sociological theory, he resists the use of habitus as a linking process between different levels of social integration. For Collins, people bring ‘emotional energy’ and cultural capital (resources) from one situation to the next (constituting an ‘interaction ritual chain’). However, the habitus as an organizing principle of translation from one network of social dependencies to another (or one level to another) is avoided. This is because Collins implicitly sees the concept as too close to the idea of the individual as the unit of analysis and agency (2004: 6). Also, based on his interpretation of Bourdieu, he sees it as a static concept reproducing the social order (2004: 42, 288–9). Collins interprets habitus as stasis, when it actually helps to explain the dynamics of social situations as well as other figurations at higher levels of social integration, e.g. nation-states. This is not a question of micro versus macro, and certainly not agency versus structure (on this both Collins and Elias are very close), as habitus is formed through social dependencies.

Alongside broadening and expanding interdependencies comes an increasing ‘scope of identification’ between people over broader social networks across class, ethnicity and nationality (Elias, 1991: 155–237). The fate of individuals becomes more intertwined as they increasingly depend on each other for socially generated needs. Within states, the scope of identification is facilitated unintentionally by the state monopoly of physical violence and the eventual transfer of ownership of this monopoly into public hands (a process of relative democratization in state politics). Thus, a feeling of belonging and caring for others is part of a general process of ‘functional democratization’ or relative equalization. The state becomes a nation-state. Elias stresses the twin monopolies of violence and taxation in state formation processes. Unlike Weber, he implicitly dismisses the significance of legitimacy. Mennell (2001: 34) argues that the legitimacy of the monopoly of violence is a ‘red herring’; what matters is the effectiveness of state control of the means of violence.

But state legitimacy is significant once shorn of its normative connotations (from the perspective of the sociologist), and recast from the perspective of the people living in the state territory. Then, legitimacy relates to the valence and intensity of the emotional identification between the state agents of violence control and the rest of the population. Where certain groups do not recognize the monarch or government as their head of state, then perceived legitimacy can affect the willingness of such groups to respect the commands of state forces. In his more ‘macrohistorical’ work, Collins (1986: 155–60, 1999) emphasizes state legitimacy as significant for both inter- and intra-state relations. Elias (1991: 178) identifies the weakness of the emotional identification between the state at the more general level and its people as new states are formed. However, the emotional relations between state agencies of violence and the rest of the population in terms of perceived legitimacy are less emphasized. Collins (1999), following Weber, argues that military mobilization permeating throughout the population fosters nationalism and state legitimacy. Following Durkheim, he also stresses the importance of rituals and symbols for the construction of emotional devotion to the idea of the nation-state. These processes vary in their success of course, and we argue that the historically contested territory of Ireland produces varying degrees of state legitimacy within this space. This in turn produces divergent forms of habitus, particularly in relation to emotional responses to certain political and state symbols of authority. But the habitus is enacted in different social situations and produces a more generalized mode of emotional action beyond the conditions of its initial generation. The habitus is not merely reproduced from place to place; such places constitute different figurations of people which can shape the habitus to some extent. It remains in process, but the perceived legitimacy of state agencies affects the readiness to depend upon and trust such agencies. This in turn affects the propensity for self-reliance and collective action in situations involving inter-group conflict and emotional tension.

Dunning (in Elias and Dunning, 2008[1986]: 230–1) attributes the declining violence in sport over generations to the receding influence of family and residence ties in favour of occupational ties beyond the local territory. He likens these lengthening chains of interdependence to Durkheim’s analysis of the movement from mechanical to organic solidarity. The scope of identification remains quite local such that relations with outsiders are sporadic and hostile. Dunning (1994: 152) also argues that part of the lack of incorporation of some sections of the working class into dominant values concerning the use of violence is due to their relations with state agencies of violence control. They are often less protected by the state, and are thus sometimes less constrained in the use of violence. They are more likely to experience state violence. Dunning et al. (1988: 89–90) also note that prior to the incorporation of most of the working class, attacks on police were common and so attacks on referees can be understood as an extension of the contempt for authority. Though Dunning does not use the term, this points to the significance of legitimacy as an important factor in the maintenance of state monopolization of violence. Some people in state jurisdictions may not identify with police and other defence forces because they are believed to act on behalf of other ethnic, colonial, national or religious groups (these may overlap with each other and with class), or even of different versions of the territory and of the scope of ostensibly the same nation. These social divisions sustain ‘them’ and ‘us’ modes of orientation. While one boundary can recede, another can emerge.

Collins’ (2008) recent analysis of violence from a micro-sociological perspective parallels the figurational concern with social relations, but he sees them in immediate, face-to-face interactions rather than long-term processes. Collins stresses the situational, micro-interactional accomplishment of violence, and sees it as difficult, incompetent and unusual. Most conflict encounters do not lead to violence at all, but mild forms of verbal aggression that quickly fade in intensity as antagonists engage in bluster, face-saving presentations of self and somewhat respectable retreat rituals. His concept of ‘forward panic’ conveys the phenomenon of people overcoming natural inclinations towards positive emotional ‘entrainment’. Violent outbursts become uncontrolled, assuming a life of their own as individuals get caught up in a tunnel of anger and aggression. While Collins’ analysis is persuasive, his neglect of an historical understanding of the development of emotional habitus means he cannot explain changing moral orders surrounding violence.

Collins does allude to the escalation of tension over time. This history can vary from the wins or losses during warfare to the ‘series’ of matches between baseball teams. Similarly, Collins (2008: 475n2) refers to a ‘type of individual’ who had worked in dangerous manual labour as a factor explaining his vocal aggression towards a baseball player. This is close to Elias’s use of habitus in describing the ‘second nature’ aspects of individual dispositions. Both interactions and situations are shaped by wider chains of interdependencies which converge and intersect as the individual habitus. Just as Goffman neglects this broader dimension of emotions (see Kuzmics, 1991), so Collins leaves this broader dimension for the time being. While interpretations and definitions of the situation occur as each social situation unfolds, situations themselves are interpreted differently over long time horizons in accordance with ‘the total context of social relationships’ (Kuzmics, 1991: 13).

We argue that situational dynamics are important in understanding violent displays, but the capacity to overcome confrontational tension itself exhibits socio-historical variability. Generally, more conflict-prone, and hence more violent, societies develop more extreme emotional dispositions and the dam of self-control can be breached more easily and quickly. This is not simply a case of inner impulses surmounting weak egos and superegos, but socially generated tensions advancing more quickly than the socially generated self-steering capacity to control emotions in tension situations. As Cooney (2009) notes, Collins offers little by way of explanation for the variability of violence across time and space. We argue that such variability is in part due to the structure of conflicts within society and the associated ambivalent ideals and morals relating to violence and aggression. However, sports rivalries also develop their own dynamics, or relative autonomy, once set in motion.

Methods

The methods for generating data included newspaper reports of Gaelic football matches from the late 19th century up to 2010. Identifying relevant newspaper reports was facilitated through the use of online archives: the Irish Times Digital Archive (specific to that newspaper) and the Irish Newspaper Archive (which covers a range of national, provincial and local newspapers). Both archives are databases of digitized newspapers, so that the articles generated are precise replicas of the original newspaper reports. Our aim was not to quantify the increase or decrease in reporting of incidents of spectator violence. This would serve little purpose, as increases are as likely to signify growing moral concern over violence as to signify an actual increase in violent encounters. Our focus was on the changing socio-temporal structure of the situations entailing spectator violence. Reports were analysed for the form of violence depicted. Finding such incidents required flexible use of keywords in online searches, as the words used to describe ‘violent’ incidents changed over time. This too is an indication of the varying moralities of violence and expected impulse controls among the population. While ‘wild’ or ‘exciting’ scenes proved useful in the earlier historical period, words like ‘disgraceful’ and ‘violent’ proved more fruitful in the later periods. We also used evidence from other sources, such as referee reports, which were accessed from searching manually through the archives in the Gaelic Athletic Association (GAA) headquarters at Croke Park, Dublin.

The focus of this article is on the event structure so reports were analysed in terms of the interdependent people or groups involved – primarily players, spectators and referees – especially in terms of the timing of the initial event, the instigators and the related counter-movements by other interdependent groups. The article identifies changes in the structure of these social encounters over the long term, and seeks to relate these changes to broader figurational changes that account for shifting forms of the national habitus in relation to emotional impulsiveness and preparedness for violent action, along a collective-individual continuum. As Maguire (2011: 880) states: ‘The task is to trace and analyse the significance which specific events have in time and their conjunction with other events.’

The quotes from newspaper reports selected here are illustrative of the general direction of social change. These can only be a selection, but are chosen for their typicality in a given period. The periodization of change cannot be fixed in stone, as there is inevitable overlap in prevailing event structures. We argue that there is a general change in the typicality of encounters involving violence at these sports events.

Collective Impulses and Violent Actions

Although earlier folk versions of football were played in Ireland (Connolly and Dolan, 2010), from the foundation of the GAA in 1884 up until the 1950s the typical socio-temporal structure of violent encounters involved a rapid collective mobilization of physical force against offending players.

As matches between local communities, villages and townships were organized, generating the need for standardized games on a national level, the existing habitus functioned according to the expectations of local belonging. The duty to protect group members from violence was keenly felt in the context of the incomplete and increasingly illegitimate state monopoly of violence, as the power chances of tenant farmers and administrative functionaries rose relative to those of Anglo-Irish landlords (see Dolan, 2005, 2009). The GAA was an outcome of this process, and subsequently reinforced the tendency towards ethnic nationalism (Dolan and Connolly, 2009). The GAA also standardized a more uniform structure for the sport, and enabled more frequent games between local parishes and towns, thereby facilitating the escalation of rivalries between local groups. As a result, the games resembled rituals producing symbols of moral solidarity (Collins, 1990, 2004). Following Durkheim, Collins (1990: 44) argues that ‘righteous anger’ is directed at violators of sacred symbols; such anger is highest where a group has a local focus. This is similar to Dunning et al.’s interpretation above, although Collins does not explain the transition of groups towards cosmopolitan values. During the phase we discuss here, anger seems to have been directed at players on the opposing team who assault players from the group’s team. Assaults emanate from a habitus formed under conditions of limited and illegitimate state control of violence, which predisposes violent reactions to assaults on fellow group members.

During a match between two districts less than 10 miles apart, the ‘spectators’ entered the field and participated in the play (Southern Star, 14 May 1892). Two players fell during play, and one struck the other as he got back on his knees:

The spectators in the vicinity witnessed the action, and with loud shouts a number of them surrounded the two men; two or three sticks were raised, and while Neagle [the offending player] was still on his knees, he received a violent blow of a stick from a respectable-looking man, whose hat was off, and who seemed filled with passion and excitement.

According to the journalist, but for the interference of four priests, ‘a serious disturbance might have ensued’. The police arrived but the ‘crowd were now all over the field, and the greatest disorder prevailed’. Though there is ostensibly a division of functions and a figuration of social interdependencies comprising interlocking role specializations, the prevailing habitus has not yet attuned to this social differentiation. Spectators behave as players in attempting to interfere with the game, and also as referees in reacting to rule infringements.

Although there are brief attempts to control emotions and aggressive conduct during phases of escalating violence, they prove unsuccessful as a form of forward panic (Collins, 2008) ensues. The game itself builds up emotional tensions within and between spectators and players. The social bonds between players and spectators from the same district, as well as the fragility of functional specialization in the performance of the game, including its regulation and spectacle, contribute to cavalier attitudes toward spatial boundaries containing these ostensibly different functions. Players enter the space for spectators and vice versa at moments of physical assault. In the absence of a legitimate state monopoly of violence or a mutually acceptable apparatus of regulation concerning the match itself, members of the same group exert their own retribution as part of the emotional experience of anger and excitement brought on by attacks to the social group.

Attacks on referees could be precipitated by his decisions, or occur as part of a general riot-like process. In 1928 (Irish Times, 13 October), for example, the crowd invaded the pitch immediately after one player was struck by another. The referee was assaulted first by a spectator, and then by a player. At a match in 1931 the referee decided to end the game following his ‘dispute’ with one set of supporters (Irish Times, 21 July). He was immediately pursued ‘by an angry crowd’ over the field and was repeatedly struck, eventually finding refuge in a house in the local village.

Collective violence was still occasionally occurring by the 1970s. A referee stated in his report to the county board of a club game between local teams in 1976 that fights between players and spectators ensued following a series of actions, including a spectator kicking a player in the head as he lay on the ground (Irish Independent, 11 August). The game was labelled a disgrace to the GAA by the county board chairman. Another club game in 1977 between local rivals in Tyrone (county within the jurisdiction of Northern Ireland) led to the death of one of the spectators following an injury sustained during a spectator invasion of the pitch (Irish Times, 10 August). Again the invasion occurred before the end of the match. A subsequent account of the match (Grimes, 2009) reveals that there had been trouble at matches between the clubs preceding this particular event. So a narrative of conflict built emotional tension between rival supporters. This was also the case in the 1976 match.

The prevalence and relative tolerance of interpersonal violence in the late 19th century and much of the first half of the 20th century is explained by the recurring and shifting social conflicts over this period. The Irish agrarian class structure was realigned along clear antagonistic lines from the mid-19th century such that the conflicts between landowners and tenant farmers came into sharper focus (Clark, 1979). International (but intra-state) interdependencies between Irish farmers and British consumers advanced with the intertwined industrialization of Britain and commercialization of Irish agriculture (see Dolan, 2005, 2009). The consequent rising power chances of farmers together with the declining power chances of the aristocracy and gentry through British industrialization provided the context for agrarian agitation for improved conditions from the 1870s. According to Elias (1996: 235), ‘human groups usually revolt against what they experience as oppression not when the oppression is at its strongest, but precisely when it begins to weaken’.

In the Irish case, the increasing economic interdependencies fostered by the political union between Ireland and Great Britain since 1801 weakened the position of the established landed class in Ireland. This encouraged violence against perceived oppressors. However, this was moderated by middle-class groups and minor gentry relatively excluded from aristocratic circles. A continuum between actual physical violence and violent rhetoric emerged, though some groups eschewed violence on moral grounds (Curtis Jr, 1988). The escalation of conflict and violence along class, religious and national lines varied throughout the country, but this inter-class conflict also produced considerable intra-class conflict and intimidation. Boycotts and other reprisals aimed at landlords became even more vociferous against group members who broke boycotts. Subsequent conflicts emerged between large and small farmers, between employers and working classes, between adherents to nationalism and unionism (loyalty to the United Kingdom), and between exponents of divergent versions of nationalism, which lasted well into the 20th century (Dolan, 2005). While state forces of violence control gradually attained more legitimacy in the Republic of Ireland, the legitimacy of the state in Northern Ireland declined from the perspective of the minority group with positive feelings towards Irish nationalism. This group was (and still is) the main proponent and supporter of Gaelic football in Northern Ireland. This has led to multiple violent attacks on GAA premises and supporters (Hassan and O’Kane, 2012). There, state de-monopolization (for most) and de-legitimization (from the perspective of sections of the nationalist community) of violence control encouraged some to develop self-reliance for defence functions. These changing, though recurring, social conflicts provide the conditions for a habitus predisposed towards greater social distances and emotional antipathy between ‘them’ and ‘us’.

Throughout this, neither Gaelic football nor its organization and propagation through the GAA were mere epiphenomena of independent social processes. The foundation and activities of the GAA were bound up with various projects of national self-determination and in Northern Ireland served as a counterweight to a state and society dominated by opposing groups (Hassan and O’Kane, 2012). The GAA quickly attracted rising groups in Irish society, but its functions from the start were not simply the organization of mimetic leisure activities to generate pleasurable emotions deemed necessary through the relative routinization of social life. Unlike England, Ireland did not experience a gradual and steady period of pacification and parliamentarization (see Elias in Elias and Dunning, 2008[1986]: 3–43). Sporting organizations were formed partly in response to social conflicts, not to generate them in pacified form. The GAA was formed as part of a dynamic figuration involving various sporting organizations, some of which were oriented towards conserving the Union of Great Britain and Ireland as well as the established position of nobility, gentry and closely interdependent sections of the professions and bourgeoisie.

Deferred Satisfaction and Social Protection

From the 1950s another form of social interaction developed more strongly which involved shifting dependencies between the sport’s various participants and functionaries. Players increasingly began to protect referees from assault by spectators.

For their part, spectators became more likely to control their emotional displays by deferring them until the end of the match. This was partly a rationalization of conduct, as their preferred outcome – avoiding the defeat of their team – would be undermined by early intervention. However, this is not entirely rational thought interposing on separate emotional processes. Emotions and reason cannot be compartmentalized in social action (Barbalet, 2001; Collins, 2004; Elias, 2000[1939]). Although spectators feel anger at the actions of interdependent others, this is also counterbalanced with hope and anticipation until the referee ends the game.

Over time a gradual, though partial, acceptance of specialized functions occurs. Players do not necessarily accept the referee’s decisions, but they generally learn to exert greater, and more even, emotional self-control. ² This is also encouraged through social observation by the sport’s administrators within the multi-level organization (local to national), and consequent punishment in the form of playing bans. Players learn that the sport itself is unsustainable without impartial controllers and rule enforcers. Increasingly, spectators also learn their role within the figuration and refrain from intervention during the game itself. Social constraints on spectators include surveillance by police, stewards and journalists. The advancing threshold of moral repugnance towards violence, including calls for greater social control of spectators, became more prominent in newspaper and referee reports (e.g. Nenagh Guardian, 12 May 1951; Irish Times, 23 January 1952; GAA Central Council minutes, 2 December 1955, 9 November 1962). This is always an unstable process because the structure of interdependencies is complex and shifting between players, spectators and referees over the course of each game. The growing acceptance of functional specialization depends also on the acceptance of a spatial specialization within the sporting arena for a specific time period. Spectators must refrain from entering the field of play during the course of the game.

A partial social fragmentation on each side advances. Players protect referees from their own supporters. Emotions come under more effective social and self-control. Spectators’ attachment to the social group (whether club, parish, town or county), especially as a symbol of the group, remains high while the feelings of solidarity towards specific players recedes somewhat. The responsibility felt to defend them from assault is superseded by the desire for group charisma and pride through victory. As relative detachment towards players advances, the emotional involvement towards referees heightens. This is because the partial acceptance of functional specialization within sporting encounters places increased pressure on referees. Increased dependence on functional specialists of rule compliance does not necessarily lead to the moderation of anger in the short term. Rather, the object of anger shifts as the acceptance of roles proceeds. The deferred spectator violence illustrates less immediate dependence on specific player actions and conflicts, and a heightened sense of belonging more oriented towards the symbolic functions of the team and the need for status through victory, rather than a code of honour to come to players’ defence in violent encounters. Van Bottenburg and Heilbron (2006) identify a changing balance of power between practitioners, spectators, organizers and television viewers to account for increasing violence in no-holds-barred fighting. García and Malcolm (2010) counter that the apparent violence is spectacularized, and reflects informalization processes (Wouters, 2007) and the quest for excitement (Elias and Dunning, 2008[1986]). We argue that the shifting object of violence in Gaelic football is due to changing forms, rather than levels, of interdependency.

There is overlap between different forms of spectator violence. In an inter-county match in 1955, some spectators encroached on the pitch during the game, and an umpire was assaulted. At the final whistle the referee and umpire were struck by some spectators. The referee reports: ‘Due to the intervention of members and officials of both teams – to whom I am grateful – the incident did not reach more serious proportions’ (GAA Central Council minutes, 2 December 1955). The referee notes that the match ‘was played in a very sporting manner, the conduct of both teams was exemplary and they did not in any way contribute to the actions of the spectators mentioned’. After a national league final in 1966, spectators attempted to attack the referee, and players ‘appealed to the supporters to leave the field’ (Irish Independent, 10 October).

As this culture of players protecting referees developed, it constituted another form of dependency between them. This expressed and accelerated a functional distance between players and spectators, representing the internalization by players of the prohibition against violence towards referees. Players and spectators remained highly interdependent; but their emotional experiences did not necessarily converge in time and space. Players appealed for calm on the basis of group solidarity and belonging, so their emotional experience of disappointment, shame or frustration would be supplanted by the need to manage the emotions of others when neutral functional specialists were threatened.

Another match between Derry (a county in Northern Ireland) and Kerry in 1973 produced ‘disgraceful scenes’ when ‘a large group of young men and youths waged a pitch battle with gardai [police] on the field’ (Irish Times, 9 April). Some 200–300 Derry spectators invaded the pitch on the final whistle and the referee suffered ‘several blows’ before being protected by officials and police. As stated earlier, while pacification became more secure in most of the Republic of Ireland, it became more precarious in Northern Ireland from the late 1960s. Gaelic games were sometimes followed as a symbol of national belonging across the contested border between the two Irish states, but this solidarity could also form the basis of animosity based on divergent expectations in relation to the national ideal.

These collective, deferred social actions on the basis of emotions aroused through observing representative teams did not entirely disappear; they occurred occasionally in areas where habitus formed within a social figuration lacking state legitimacy, or where local rivalries in club games escalated and intensified through a highly emotional narrative of recurring opposition (see Collins, 2008: 305 for a similar point on ‘emotional memories’). However, from the 1980s these actions began to be overtaken by more individual emotional responses to actions on the field of play.

Individualization of Violence

Elias (1991: 197) notes that one of the bulwarks against individualism in earlier times was the hereditary character of social positions and opportunities. This also implies that the hereditary structure of local interdependencies limits the scope of mutual identification beyond the village, parish or town. In much of rural Ireland up to the 1960s, sons were dependent on fathers to inherit the farm and become ‘men’ in the eyes of the community. It was difficult for men to marry until they inherited property so their life chances and means of attaining status remained dependent on very close, local ties. Such ties did not as a matter of course produce feelings of solidarity and mutual care, but also, and intertwined with these, feelings of shame, frustration and resentment due to the often delayed pathway to status-attainment. As the state expanded, and employment opportunities increased beyond the confines of close relatives, habitus formation became more individualized as people were addressed as individuals by state systems of education and social welfare, and as people had to make their own occupational choices (Elias, 1991).

The more individualized form of spectator violence emerges as a result of this more individualized habitus, so that spectators become less likely to act in unison in response to perceived wrongdoing against fellow members of the group. This weakens emotional homogeneity within the spectating crowd; though, in Collins’ terms, emotional energy flows through the crowd, the experiences of joy and anger become more differentiated. The immediate situational dynamics produce emotional experiences, but these are anchored somewhat by the more individualized habitus, increasing the variety of emotional experiences. The greater acceptance of the social taboo against violence, the legitimization of violence-control specialists, and the ‘mutually-expected self restraint’ (Goudsblom cited in Wouters, 2007: 188) characteristic of more individualized societies, means people are less prepared to stop the conduct of others.

Reports of individual assaults on players and referees became more dominant from the 1980s. These attacks were still related to events during the game itself (either player fouls or referee decisions), but collective responses were increasingly rare. Individual players were sometimes attacked by lone supporters at the conclusion of matches (Irish Times, 1 August 1983; Irish Times, 4 December 1984). Violence towards players tended to involve ‘slaps’, while assaults on referees were sometimes very violent. For example, during a club game in 1987 the referee ‘was beaten about the head and kicked in the groin’ and later hospitalized (Irish Times, 19 August). After the 1988 All-Ireland final replay a player was assaulted by a spectator in the context of a brief ‘scuffle’ between that player and one from the opposing team (Irish Times, 11 October; Irish Times, 13 October; Irish Times, 21 October). What characterizes these incidents is both a sustained media commentary and responses by the senior administrators condemning such moral transgressions.

In 2010, a televised match between neighbouring counties produced extensive comment across television and newspaper outlets lasting several weeks. The match was decided by a goal ‘scored’ in injury time, which the referee later accepted should have been disallowed. Following the final whistle, a single spectator entered the pitch and shoved the referee, who was immediately protected by another spectator and several players; subsequently, several other spectators shoved or attempted to push the referee, who was then protected by the manager of the losing team and several police officers as he made his way off the pitch: ‘A man in a red shirt rushed onto the field and shoved the referee in the chest … he was jostled twice more by angry football fans and pushed in the back by another’ (Irish Times, 12 July 2010). That man was later convicted of assault on the referee, along with one other, and both received €1,000 fines (Irish Times, 1 March 2012). Although the television images show several other men confronting the referee, charges were only sought for these two men, who acted individually and shoved the referee at separate moments. The referee was shoved, pushed and jostled by several spectators, and some others remonstrated with him and engaged in finger-wagging gestures to indicate disapproval of his performance. However, these actions did not exhibit the levels of collective physical violence of earlier historical periods.

In one sense this could be interpreted in Collins’ terms as people unable to surmount the confrontational tension and fear produced in conflict situations. A few individuals appeared prepared for violence, but their actions were half-hearted and unsuccessful. They appeared to be largely incompetent, just as Collins would predict. But from a figurational perspective, this is precisely the point; the social conditions of habitus formation have changed to the extent that aggressive impulses brought on by conflict do not, by and large, spontaneously produce violent actions. Collins (2009) misinterprets Elias’s theory of civilizing processes by criticizing the notion that lengthening interdependencies repress the natural urge for violence. Elias makes no such claim, and instead sees violence produced in competitive figurations (Mennell, 1998: 249).

As people develop bonds over longer links and more varied and multiple types of dependencies based on the functions people perform or fulfil for one another, conduct becomes more emotionally controlled and differentiated. The emotional attachment to one’s local community and territory recedes to the extent that broader interdependencies provide meaning, opportunities and purpose. New layers of identity emerge. Depending on one’s position in the figuration and the degree of active participation in those longer links, some individuals find multiple sources of emotional involvement, while others experience resentment at new links that are seen as unequal. In other words, longer chains of interdependences produce heterogeneous emotions that diminish the potential for collective actions in the name of a specific town or parish. However, this very heterogeneity facilitates individual transgression of national norms of spectator conduct.

Conclusion

Collins and Elias emphasize different aspects of emotions and emotional control in the context of social encounters involving conflict. For Collins, the resort to violence is a difficult transition due to the presence of confrontational fear and tension in micro situations. According to Collins, people are generally very incompetent in conflict situations, so that anger and rage generally subside through an initial escalation of bluster and verbal abuse, before social withdrawal amid efforts to save face. This is because people are deemed to be positively, even naturally, disposed towards building rhythms of positive emotional entrainment through conversation and other bodily interactions.

Following Elias, we argue that such social situations generating emotions and emotional controls should also be seen in the context of the more long-term development of the habitus, itself formed through the intersection of multiple, often contradictory, dependencies between many people in diverse places over several generations. There is little doubt that Collins’ concept of emotional energy is useful for exploring the flow and narrative of situations, and his stress on interaction ritual chains mirrors to some extent the concern for lengthening chains of interdependencies in Elias’s account of social change. Collins sees ritual chains more as repeated sequences of similar events, allowing individuals to build up emotional energy and cultural capital. It is the tone rather than the content of symbolic rituals that build up moral solidarity. Although Collins acknowledges the two-way flow between micro and macro dimensions of social life, he tends to accord primacy to the micro-situational experience (Münch, 2005). The habitus is of course constructed in social situations, primarily through social relations between children and parents and teachers, but the content of symbols used in such encounters conveys the prevailing understanding of current and anticipated social dependencies and their associated power balances. Relations at higher levels of social integration, such as those between nations, even though embodied in social situations, play a significant part in forming the content and tone of emotional experience in immediate face-to-face encounters.

The importance of habitus for the outcome of tense, exciting encounters between oppositional groups can be seen in the changing structure of situational dynamics involving spectators, players and referees at sports matches. From the more micro aspect, to use Collins’ terminology, the ostensible game structure changes little – people come to a specific place to watch other people play in their name (at various levels of social integration, from local parish to county and province). But the figurational dynamics of this type of social situation does change – people learn to accept the functional and spatial specialization of the encounter. Spectators are there to observe and become emotionally invested in the flow of the game (building excitement and emotional energy). The broader social figuration develops a more emotionally controlled habitus, such that feelings generated by events on the field of play become less likely to spontaneously produce violent action. Over time, similar situations produce different outcomes, and this is due to changing standards of emotional control becoming more like a second nature, more securely and evenly built into the individual habitus. But this is an uneven process, not only because of the uneven effectiveness of state monopolies of violence control, but also because of the variable strength of emotional belonging between members of the population, representatives of the state and the associated political symbols of statehood.

The more stable and even emotional self-control, characterized by Elias as the social constraint towards self-restraint, also depends upon the degree of state legitimacy in terms of the scope of emotional identification between various levels of social integration – from the person to the state. In Ireland’s case, it was this lack of legitimacy that led to the formation of the GAA in the first instance. National distinction in the context of rising power chances for upwardly mobile middle-class groups fostered the organization of games to symbolize difference. Ironically, this effort at boundary maintenance at the international level helped to foment rivalries at the intra-national, inter-parish level, as more frequent games between contiguous groups escalated emotional tensions as narratives of defeat and vengeance. Under conditions of state breakdown and re-parliamentarization following intertwined inter- and intra-national conflicts (between national-level ‘us’ and ‘them’ and then between a lower level of ‘us’ and ‘them’ in terms of the imagined emotional belonging to the contested nation-space), the prevailing habitus is ill-suited to mere spectatorship.

As social interdependencies advance at the national and international levels, and the state gradually develops legitimacy, the habitus becomes less disposed towards emotions encompassing spontaneous, collective action and anger. The socio-temporal structure of spectatorship changes indirectly and unevenly in relation to structural changes at higher levels of integration. The bonds between people shape emotional experiences mediated through the contingent construction of the individual habitus. Interaction rituals remain significant, but the interactions change according to the changing ways habitus is formed through developing dependencies far beyond the immediate situation.