personalities in the crowd: those who would escalate a sports riot

AGGRESSIVE BEHAVIOR Volume 21, pages 91-100 (1995)

Personalities in the Crowd: Those Who Would Escalate a Sports Riot Gordon W. Russell

Department of Psychology, University of Lethbridge, Lethbridge, Alberta, Canada

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A series of four studies was conducted at ice hockey games with a view to establishing the correlates of spectators’ self-reported reasons for attending and their propensity for involvement in crowd disturbances. Spectators attending for the reason of ‘‘I like to watch the fights” and those most likely to join in a fight if one were to break out in the stands were young, single males. Ratings on the dependent variables were further related to individual differences measures of assaultiveness, psychopathy, self-esteem, and public self-consciousness. Anomy was unrelated to either dependent measure. Support for the false consensus effect was additionally forthcoming. @ 1995 Wiley-Liss, Inc.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Key words: riots, psychopathy, false consensus, sports, violence, personality

INTRODUCTION

Social scientists have shown a long-standing, albeit sporadic interest in crowd phe- nomena [e.g., Lang and Lang, 1961; Milgram and Toch, 19691. Of the diverse forms of crowd behaviors, violent outbursts at sports events have given rise to persistent and heightened levels of public concern, especially in Europe. Soccer crowds in particular have attracted media and government criticism for their disruptive behaviors both within and without the stadium. Historically, efforts to control riotous behavior have ranged from banning alcohol, to segregating rival factions, to raising admission prices to discourage the riffraff [Guttmann, 1986; Russell, 19931.

Retrospective analyses of outbursts typically attribute responsibility to one or more iden- tifiable groups, e.g., fans of the visiting team, known troublemakers, a minority of disaffected youths, etc. European investigations of soccer hooliganism have yielded a consistent demographic profile of those involved in disorders at sports sites. Instigators are variously

Received for publication October 13. 1993; accepted June 16, 1994.

Address reprint requests to Gordon W. Russell, Department of Psychology, University of Lethbridge, 4401 University Drive, Lethbridge, Alberta, Canada TI K 3M4.

0 1995 Wiley-Liss, Inc.

92 Russell

described as young, single, disaffected, and economically disadvantaged males [e.g.,Adang, 1992; Bakker et al., 1990; Murphy et al., 1990; Pilz, 1989; Roversi, 1991; Van der Brug, 1990; Van Limbergen et al., 1989; Zani and Kirchler, 199 1 1.

An alternative that goes beyond identifying the characteristics of aggregates as causes lies with a different level of analysis. It is suggested that a more fine-grained assessment of crowd members by means of an individual differences approach is likely to expand our understanding of those who would instigate or otherwise escalate disturbances.

Studies of crowd violence in which personality measures have served as independent variables are few in number [e.g., Forward and Williams, 1970; Meier et al., 1941; Ransford, 19681. As a consequence, explanations regarding the sorts of people involved tend to rely almost exclusively on a mix of speculation and generalizations from the social-experimental literature. Confirming evidence from participants or would-be participants who patronize disturbance-prone events is scarce.

Ransford’s [ 19681 study of the 1965 Watts riots makes a major contribution to an in vitro approach. Male heads of households (N = 3 12) were interviewed while buildings were still smoldering. When asked, “Would you be willing to use violence to get Negro rights?”, those with a belief in an external locus of control or, alternatively, men who felt powerless in their circumstances were more willing to resort to violence as a solution. Of the 16 men who admitted to being personally involved in violent action for Negro rights, 15 scored high on the powerless measure. Age was not a factor.

However, note should be taken of the Forward and Williams [ 19701 study of participants in the later Detroit uprising [see also Lachman and Waters, 19691. Black males with favorable attitudes toward violence instead had an internal orientation. The re- searchers attributed the discrepant findings to their use of a much younger sample (12-18 years) and a different measure of locus of control.

Finally, an early study in this tradition [Meier et al., 19411 asked subjects to choose a course of action following the presentation of a scenario describing a developing riot. Those who would take part in mob action were predominantly males who exhibited tendencies toward extroversion and lower intelligence.

The present investigations sought to extend an individual differences strategy to a different culture and social setting. To this end, a battery of aggression-related personality measures was administered in srages to spectators attending ice hockey games. This piecemeal approach was necessitated by the brief, 15 min intermission between periods of play. Study 1 was intended to establish at the outset that a sufficient propor- tion of spectators was attracted by the prospect of player fights to justify use of the motive as a dependent variable. Thereafter, those personality measures emerging as leading predictors in studies 2 and 3 were to provide the content for a systematic replication in study 4. Finally, a test of the false consensus effect [Ross et al., 19771 was introduced in study 4 to provide additional insights into the cognitive perspective of fans.

STUDY 1 Subjects

The subjects were males (N =57) and females (N = 46) found in attendance at a Western Hockey League (WHL) game in Lethbridge. The males had a mean age of 37.5 years (SD = 14.5) and the females had a mean age of 36.8 years (SD = 14.5).

Personality and Sports Riots 93

Procedure Senior students (N = 20) were rehearsed in their roles as interviewers and randomly

assigned to occupied sections of the arena. Uninvolved and sober spectators, 18 and over, were approached individually using a procedure that approximated a random sampling of each assigned section. The interviews were conducted before the game began and in the 15 min intermissions between periods. Potential subjects were asked for their cooperation in completing a short, 10 min survey for a class project at the university (approximately 6% declined). The interviewers described the rationale for the study as an interest in identifying “the reasons why different types of people attend hockey games.”

Independent Measures In addition to the biographical items of sex, age, and marital status, subjects were

also asked to indicate the frequency with which they attend games. A 3-point scale labeled “Rarely,” “Sometimes.” and “Frequently” accompanied the item.

Dependent Measures Initially, it was necessary to establish the importance of ‘‘I like to watch the fights”

relative to seven other plausible reasons for spectators’ attendance. Subjects rated the strength of each reason on accompanying 6-point scales (0-5). The remaining reasons for attendance were “To meet new people,” “To meet people I know at the Sportsplex,” “To support my team,” “I like to watch good, skillful hockey,” “To learn some of the finer points of the game,” “Nothing better to do this evening,” and “To please someone else.”

Results and Discussion A 2 (sex) x 8 (reasons) mixed design, repeated measures analysis of variance was

conducted on subjects’ ratings of the reasons for their attendance. The results are presented in Figure 1 and show a significant main effect for reasons, F(7,707) = 70.1 1, P < 0.0001, and a significant sex x reasons interaction, F(7,707) = 2.51, P < 0.02. Fisher’s protected t-test was applied and revealed that men (M = 2.51, SD = 1.58) were more likely than women (M = 1.61, SD = 1.20) to attend because they “like to watch the fights,” t(95) = 3.19, P < 0.005.

The principal findings are presented in Table I. In the case of study 1, the correla- tional analyses indicated that it is primarily young males whose motive for attending is that of liking to watch the fights. However, the frequency of attendance measure was unrelated to the dependent variable.

STUDY 2 Subjects

Their mean age was 33. I years (SD = 11 .O>.

Procedure

outlined in study 1.

The subjects were males (N = 1 12) found in attendance at a WHLgame in Lethbridge.

A group of senior students (N = 22) followed the same interview procedures as those

94 Russell

a c 0 a a 3 Q) a r 0

2 c,

t ZI C

6 1

0

Males 1 3 Enjoy the skill 2 I Support the team 3 = I like to watch the fights 4 I To learn the finer polnts 5 = To meet people I know 6 t To meet new people 7 3: Nothing better to do 8 t To please someone else

1 2 3 4 6 8 7 8

Reasons Fig. I . Spectators’ reasons for attendance.

Independent Measures In addition to several biographical items, i.e., age, marital status, and frequency of

attendance, subjects were also asked to complete three personality measures. Specta- tors completed a 7-item measure of public self-consciousness [Buss, 19801. A sample

TABLE 1. Biographic and Personality Correlates of Spectators’ Attraction to, and Propensity for, Aggression

‘‘I like to watch the fights” “I would join in a disturbance” Medicine Medicine

Lethbridge Hat Lethbridge Hat

sa 1 s 2 s 3 s4 s 2 s3 s 4

Biographic - - - - - Sex 0.30** -

Age -0.55*** -0.48*** -0.36*** -0.46*** -0.23* -0.43*** -0.21* Marital status - -0.30** -0.36*** -0.30** 0.06 -0.36*** -0.35** Attendance - 0 . 1 1 0.10 0.1 I 0.16 0.00 0.12 0.20

Self-esteem - - 4.19* - - 0.10 - Anoniy -

Public self- - 0.12 - - -0.20* -

Psychopathy - 0.30** - 0.40*** 0.27** - 0.31**

Individual differences - -0. l9* - - -0.0 1 -

- consciousness

Assaultiveness - 0.29** - 0.39*** 0.41***- 0.57***

5 = study. One-tailed tests: * P < 0.05; * * P < 0.01; ***P < 0.0005.


item from the scale is “I’m concerned about what other people think of me.” Addi- tionally, Levenson’s [ 19901 13-item psychopathy scale based on clinical criteria provided by Cleckley [ 19761 and the I0-item assault subscale of the Buss Durkee Hostility Inventory [Buss and Durkee, 19571 were administered. An item from the psychopathy scale is “I often do things just for the hell of it.”’

Dependent Measures

Subjects were again asked to rate the importance of “I like to watch the fights” as a motive for their attendance. Additionally, spectators were asked to rate the likelihood that they would join in a fight or disturbance were one to break out in the stands.

Results and Discussion

As noted in Table I, it was the young, single males who were drawn to the game by the prospect of watching fights. They also expressed the greatest willingness to join in a disturbance.

With respect to the personality measures, both psychopathy and assaultiveness had strong positive associations with an attraction to fights and subjects’ self-reported likelihood of involving themselves in a disturbance. Moreover, while public self-consciousness was negatively associated with joining in an altercation, it was unrelated to enjoying the fights.

The relative strengths of the sets of biographic and personality measures in predicting the likelihood of spectators escalating a disturbance were assessed. Hierarchical regression analysis revealed that the combination of public self-consciousness, psychopathy, and assaultiveness accounted for the dependent variable over and above that accounted for by the set of biographic items, F(3,100) = 8.81, P < 0.005.

Intercorrelations among the individual differences measures administered in study 2 yielded nonsignificant (NS) relationships between psychopathy and assaultiveness, r( 11 0) = 0.148, NS, psychopathy and public self-consciousness, r( 110) = 0.138, NS, and assaultiveness and public self-consciousness, r( 110) = 0.029, NS. A further check made on the possibility that men’s ratings on the dependent measures were influenced by the sex of the interviewers yielded NS differences, F( 1,107) = 1.68, NS; F(1,108) = 1.45, NS.

STUDY 3

Subjects

the Lethbridge Sportsplex. They had a mean age of 36.2 years (SD = 14.8). The subjects were male spectators (N = 109) found in attendance at a WHL game at

Procedure

study 1. The interviewers ( N = 22) followed the same procedures as those described in

‘In the individual case, psychopathy is typically diagnosed by trained clinicians using the Psychopathy Checklist [Hare et al., 19901.

96 Russell

Independent Measures The same biographical items asked in study 2 were again administered. Subjects

were additionally asked to complete Bown’s [ 19611 6-item measure of self-esteem along with the 9-item anomy scale of McClosky and Schaar [ 19651. An example from the self-esteem scale is “In almost every respect, I’m very glad to be the person I am.”An anomy item is “With everything in such a state of disorder, it’s hard for a person to know where he stands from one day to the next.”

Dependent Measures

their joining in a disturbance.

Results and Discussion As shown in Table I, age and marital status were negatively related to “I like to

watch the fights” and subjects’ self-reported likelihood of involvement in a disturbance. Furthermore, self-esteem was (negatively) related to the dependent variables whereas anomy was not.’

A planned systematic replication was undertaken using the leading predictors from the foregoing studies. For this purpose, a different WHL franchise city was selected. Also, the age range of subjects was restricted so as to exclude those who, while they might report feeling compelled to join a disturbance, usually lack the physical where- withal to actually participate effectively.

Again, spectators were asked to rate their degree of liking fights and the likelihood of

STUDY 4 Subjects

Hat, Alberta. The men had a mean age of 3 1.2 years (SD = 7.4).

Procedure The procedures were the same as those previously followed in Lethbridge with the

exception that the interviewers ( N = 17) sampled a more homogeneous group of spectators, i.e., men between the ages of 18 and 40 years.

Independent Measures The Levenson [ 19901 measure of psychopathy and the assault subscale of the Buss

Durkee Hostility Inventory [Buss and Durkee, 19571 were again administered along with the biographical items. Additionally, a preliminary test of the false consensus effect [Ross et al., 19771 was introduced. Subjects were asked to estimate the percentage of other fans who were in attendance for the reason of liking to watch the fights.

Dependent Measures

served as the principal dependent measures.

The subjects were males (N = 71) found in attendance at a WHL game in Medicine

Again, the attraction to fights and the likelihood of involvement in a disturbance

‘Self-esteem and anomy were negatively related, r( 107) = 4.245, P < 0.01.


Results and Discussion As noted in Table I, the biographic and personality measures were again strongly

related to the dependent variables. The only exception was the frequency of attendance measures. which has consistently yielded NS relationships.3

Cultural and sports differences notwithstanding, age and marital status proved to be strong predictors, generally confirming the findings of European investigators [Adang, 1992; Murphy et al., 1990; Pilz, 1989; Roversi, 1991; van der Brug, 1992; Van Limbergen et al., 1989; Zani and Kirchler, 19911. However, age may not be a defining characteristic of participants in other types of riots. For example, Ransford’s [ 19681 study of ghetto rioters demonstrates that a history of racial oppression and unfair dis- crimination is suffered equally and can prompt residents of a/ / ages to take to the streets in violent protest.

The variable of age is certainly of preliminary importance in our attempts to under- stand the dynamics of crowd violence. However, any number of covarying, causal variables is likely subsumed by age, from a need for excitement [e.g., Apter, 19921, to testosterone, to economic frustrations. Parenthetically, in the current economic cli- mates of most nations, to be a young male is often to be unemployed, disillusioned, and bored. Nevertheless, age is a starting point in describing those who would involve themselves in public disorders.

STUDY COMPARISONS

Studies 2 and 4 allowed for comparisons of the relative importance of the biographic variables vis-A-vis the personality measures in predicting spectators’ likelihood of joining in a disturbance. As in study 2, a hierarchical regression analysis again revealed that the personality measures of psychopathy and assaultiveness accounted for the dependent measure over and above that of the biographic items, F(2,63) = 14.03, P < 0.005. Indeed, in these particular analyses the demographic variables added nothing to that predicted by the sets of personality measures, F(3,100) < 1, F(2,63) = 2.58, NS. Thus, while the personality measures contribute uniquely to our understanding of the dependent measure, the demographic variables are redundant.

Inspection of the intercorrelation matrices of reasons for men’s attendance revealed the singular importance of ‘‘I like to watch the fights.” Without exception, it was unrelated to any of the remaining reasons for attending the game across all four studies. Men attracted to hockey games in anticipation of seeing fights attend for that reason and that reason only. Watching skillful playmaking, supporting their team, and the prospect of striking up or renewing friendships play no part in their being present in the stands.

Study 4 provided an especially robust set of findings in systematically replicating the results of study 2 in a different WHL franchise city. As a result, a clearer picture of those spectators most likely to become embroiled in disturbances begins to emerge. To no one’s surprise, these men are physically assaultive. The importance of this lies in the ease with which high trait aggression manifests itself in actual physical aggression when

‘The dependent measures in studies 2 4 were correlated, r( 109) = 0.456, P < 0,0005; r( 107) = 0.405, P < 0.0005; and r(69) = 0.390, P < 0.0005. respectively.

98 Russell

opportunities are provided even in relatively everyday circumstances [Wilkins et al., 19741.

In addition to admitting to strong assaultive inclinations, those attracted to violence also exhibited psychopathic (or, alternatively, antisocial personality) tendencies.4 In addition to a proclivity for aggression, psychopathy has been variously seen to include “a lack of empathy, guilt, remorse, and fear of punishment, an extreme degree of self- ishness, impulsivity, and irresponsibility, a callous disregard for the feelings and wel- fare of others, weak inhibitory controls. . . .” [Williamson et al., 1987: p. 454; see also Hare, 19941. Add to this the strong tendency of antisocial youths to perceive hostile intent in others [Sarason, 19781 and the result is that public disorders become an even greater likelihood.5

Blackburn [ 19781 offers evidence that would account for the apparent attraction of psychopaths to the prospect of witnessing fights. He reported that psychopaths sub- jected to noxious stimulation show heightened cortical arousal. In his view, inputs from the external environment are processed at a more rapid rate. As a consequence, psychopaths may “seek out stimulating events not to increase arousal, but to maintain a high level of information flow” [Blackburn, 1978: p. 1631.

The results of study 2 also revealed a somewhat greater likelihood of involvement in crowd disturbances among those low on public self-consciousness. These persons generally express less concern than others with public evaluations that might accom- pany their actions.The prospect of public condemnation that might normally deter people from engaging in antisocial behaviors consequently may be less effective among spectators having few concerns with what others will think of them.

Previous suggestions [e.g., Van Limbergen et al., 19891 that those involved in riotous behavior are disaffected and have poor self-concepts received mixed support in the present context. Those attracted to the game by the prospect of watching player fights and those willing to join in a disturbance scored somewhat lower on self-esteem. On the other hand, anomy bore no relationships to the dependent measures.

A cognitive phenomenon has been shown to provide a further indirect link with those who would join in a disturbance. Subjects’ attraction to the fights was positively related to their estimates of the percentage of other fans who were in attendance for the same reason, r(69) = 0.415, P < 0.0005. Thus, the false consensus effect [Ross et al., 19771 was shown to operate among hockey spectators insofar as a major motivation for their attendance is disproportionately attributed to others. One implication is that those who enjoy watching fights and who estimate greater numbers of like-minded spectators to be in attendance may as a consequence assume there is general support for aggression. Moreover, in a climate of acceptance, those men predisposed to join in a crowd disturbance may be further emboldened by the perception that their actions would be favorably received by other fans who would applaud their joining in the fray.

Several methodological points warrant comment. Subjects were drawn from spectators in actual attendance at hockey games. At the same time, it must be acknowledged that the present analysis rests on spectators’ self-reported likelihood of involvement in

‘Assaultiveness and psychopathy were positively related, r(69) = 0.340, P < 0.005.

SRecent unpublished data indicate that the present measure of psychopathic tendencies also distinguishes male Dutch soccer fans from same-age nonfans [Russell and Goldstein, 19941.

Peimnality and Sports Riots 99

a hypothetical disturbance rather than a behavioral criterion of actual involvement. Williams [ 199 I : p. 2.51 captures the essence of this limitation in delicately observing that “there are invariably more ‘bullshitters’ than fighters.” Certainly, self-reports of previous involvement in disturbances [e.g., Roversi, 19911 or, better still, official arrest reports of disorderly conduct at sports sites are preferable.

Finally, future studies might profitably assess the relationships between additional personality variables, e.g., a need for excitement [Apter, 19921, macho tendencies [Rus- sell, 19921, and spectators’ likely (preferably, actual) involvement in disorders. Cer- tainly, the present listing is incomplete. Thereafter, the resulting battery of promising predictors could be simultaneously administered on-site and analyzed by multivariate techniques.

To summarize, young, single males are disproportionately attracted by the prospect of witnessing violence and attend hockey games solely for that reason. Also, a small number appears more than willing to involve themselves in crowd disturbances. Among the leading personality traits thus far associated with these men are strong assaultive and antisocial tendencies and a lack of concern with how others might view their actions.

ACKNOWLEDGMENTS

This paper was written under the auspices of the Vakgroep Sociale en Organisatie

The author thanks Jeffrey H. Goldstein, John R. Vokey, and an anonymous reviewer Psychologie, Universiteit Utrecht, The Netherlands, during a 1922-93 study leave.

for their helpful comments on an earlier version of this paper.

REFERENCES Adang OMJ (1992): Crowds, riots and the police:

An observational study of collective violence. Presented at the meeting of the International Society for Research onAggression, Siena. Italy, September.

Apter MJ ( 1992): “The Dangerous Edge: The Psy- chology of Excitement.” New York: The Free Press.

Bakker F, Whiting HTA. van der Brug HH (1990): “Sport Psychology.” Chichester: John Wiley & Sons.

Blackburn R (1978): Psychopathy, arousal, and the need for stimulation. In Hare RD, Schalling D (eds): “Psychopathic Behavior: Approaches to Research.” New York: John Wiley & Sons, pp 157-164.

Brown D (1961): The development of a self-report inventory and its function in a mental health assessment battery. American Psychologist l6:402.

Buss AH (1980): “Self-Consciousness and Social Anxiety.” San Francisco: W.H. Freeman.

Buss AH, Durkee A (1957): A n inventory for as- sessing different kinds of hostility. Journal of Consulting Psychology 21 :343-349.

Cleckley H ( 1976): “The Mask of Sanity.” 5th Ed. St. Louis: C.V. Mosby.

Forward JR, Williams JR (1970): Intemal-external control and black militancy. Journal of Social Is- sues 26:75-92.

GuttmannA ( 1986): “Sports Spectators.” New York: Columbia University Press.

Hare RD ( 1994): Predators: The disturbing world of the psychopaths among us. Psychology Today, pp 54-56,58.60-63, JanuaryFebruary.

Hare RD, Harpur TJ, Hakstian AR, Forth AE, Hart SD, Newman JP (1990): The revised psychopathy checklist: Reliability, and factor structure. Psychological Assessment: A Journal of Consulting and Clinical Psychol-

Lachman SJ, Waters TF ( 1969): Psychosocial profile of riot arrestees. Psychological Reports 24: 171-181.

Lang K. Lang GE (1961): “Collective Dynamics.” New York: Thomas Y. Crowell.

Levenson M R (1990): Risk taking and personality. Journal of Personality and Social Psychology 58:1073-1080.

O ~ Y 21338-341.

100 Russell

McClosky H, Schaar JH (1965): Psychological di- mensions of anomy. American Sociological Re- view 30:1440.

Meier NC, Mennenga GH, Stoltz HJ ( 1941):An experimental approach to the study of mob behavior. Journal of Abnormal and Social Psychology 36506-524.

Milgram S, Toch H (1969): Collective behavior: Crowds and social movements. In Lindzey G. Aronson E (eds): “The Handbook of Social Psy- chology.“Vol4. Reading, MA: Addison-Wesley,

Murphy PJ. Williams JM, Dunning EG (1990): “Football on Trial: Spectator Violence and De- velopment in the Football World.” London: Routledge.

Pilz GA (1989): Social factors influencing sport and violence: On the “problem” of football fans in West Germany. Concilium-International Re- view of Theology 532-43.

Ransford HE (1968): Isolation, powerlessness. and violence: A study of attitudes and participation in the Watts riot. American Journal of Sociology 7358 1-59 1.

Ross L, Greene D, House P ( 1977): The “false consensus effect”: An egocentric bias in social perception and attribution processes. Journal of Experimental Social Psychology I3:279-30 1.

Roversi A (1991): Football violence in Italy. Inter- national Review for Sociology of Sport 26:

Russell GW (1992): Response ofrhe macho male to viewing a combatant sport. Journal of Social Behavior and Personality 7:631-638.

pp 507-61 0.

3 11-332.

Russell GW (1993): “The Social Psychology of Sport.” New York: Springer-Verlag.

Russell GW, Goldstein JH (1994): Personality differences between Dutch football fans and nonfans. Manuscript submitted for publication.

Sarason IG (1978): A cognitive social learning approach to juvenile delinquency. In Hare RD, Schalling D (eds): “Psychopathic Behavior: Ap- proaches to Research.” New York: John Wiley and Sons, pp 299-3 17.

van der Brug HH (1992): Football hooliganism in The Netherlands. Unpublished manuscript. Department of Political and Social Sciences, University of Amsterdam, Amsterdam, The Netherlands.

Van Limbergen K, Colaers C, Walgrave L (1989): The societal and psycho-social background of football hooliganism. Current Psychology: Re- search and Reviews 8:4-14.

Wilkins JL, Scharff WH, Schlottmann RS (1974): Personality type, reports of violence, and aggressive behavior. Journal of Personality and Social Psychology 30:243-247.

Williams JM (199 I ) : Commentary. Journal of Community and Applied Social Psychology

Williamson S, Hare RD, Wong S (1987): Criminal psychopaths and their victims. Canadian Jour- nal of Behavioural Science 19:454462.

Zani B, Kirchler E (199 I ) : When violence over- shadows the spirit of sporting competition: Italian football fans and their clubs. Journal of Community and Applied Social Psychol-

1~23-28.

ogy 1 :5-?1.

personalities in the crowd: those who would escalate a sports riot

Documents