Ini hanya catatan pribadi.
The intergroup dynamics of collective empowerment:
Substantiating the social identity model of crowd behaviour
University of Sussex
University of St. Andrews
Correspondence: John Drury, University of Sussex, Sociology and Social Psychology Division, Social Sciences (Arts E), Falmer, Brighton BN1 9QN, East Sussex, UK.
A number of recent accounts of collective action point to the importance of psychological empowerment, but conceptualize it merely as a precondition for such collective action. By contrast, the social identity model (Reicher, 1996; Stott, 1996) suggests that empowerment be considered as a product as well as a precondition of collective action. However, previous research on the social identity model has merely inferred the emergence of feelings of power theoretically rather than shown it empirically. This paper describes an analysis of a town hall anti-poll tax demonstration, using a variety of material, including interviews with demonstrators, councillors and police. Consistent with the social identity model, the analysis suggests that feelings of power emerged and increased within the crowd event; these feelings were attributable to expectations of support due to the more inclusive categorization among crowd members brought about by their perceived wholesale illegitimate exclusion from the town hall. Moreover, the empowered action of crowd members was limited by shared definitions of proper practice.
Psychological research on collective action has recently become increasingly concerned with the role of power. Subjective power has commonly been theorized in terms of the concept of ‘efficacy’, which Chase (1992) defines as the (perceived) ability of the individual to bring about a desired state of the world or to avert an undesired state of the world. Klandermans (1992) points to the widespread agreement among recent social movement theorists that collective protest takes place in the belief among participants that the experienced grievances can be eliminated by the challengers' collective action. A number of social psychological studies have provided evidence to support this argument (Abrams, 1993; Breakwell, 1992; Cocking, 1995; Kelly & Breinlinger, 1995).
Efficacy, whether group or individual, is conceptualised as a dispositional characteristic which explains who does or doesn't participate in collective action. By contrast there are what may loosely be termed situational accounts of collective empowerment. Le Bon's (1895) model of the crowd falls into this latter category. He argues that, by virtue of being submerged in a mass of people, crowd members lose all conscious sense of self and hence all rational control over their behaviour but they gain a sense of invincible power. Ironically, the modern translation of the submergence concept into deindividuation theory (e.g. Diener, 1977, 1980; Prentice-Dunn & Rogers, 1982, 1983, 1989; Zimbardo, 1970) tend to retain the notion of loss of self while tending to ignore the gain of empowerment (cf. Reicher, Spears & Postmes, 1995).
By contrast, outside the field of social psychology, resource mobilization approach theorists (e.g., Gamson, 1975; Gamson, Fireman & Rytina, 1982; McCarthy & Zald, 1977; Oberschall, 1973) have argued that mass action brings about empowerment but without engendering a descent into irrationality and loss of control. It is the provision to individuals of organizational resources such as communication and the ability to coordinate action which lies at the root of collective empowerment. Within social psychology, the social identity model of crowd action (Reicher, 1982, 1984, 1987; Reicher & Levine, 1994a,b; Reicher, Levine & Gordijn, in press; Reicher, Spears & Postmes, 1995) also seeks to retain the notion of collective empowerment while discarding the idea of loss of selfhood in the mass. However, whereas resource mobilization theory sees the self as staying constant in collective contexts, albeit with enhanced resources to attain its goals, the key premise of the social identity model is that the self is transformed by virtue of participation in collective action.
In the mass, personal identity becomes less salient and people act in terms of that social identity which is associated with the relevant social category. Control over behaviour is not lost but rather governed by the understandings and values that define social identity. However collective contexts don't just have cognitive implications in terms of identity salience, they also have strategic consequences for the expression of identity. By virtue of being both relatively anonymous to outgroups and hence relatively immune from their sanctions but also visible to ingroup members and hence able to coordinate with them, crowd members are enabled to enact their collective values and understandings even in the face of outgroup opposition. As Reicher, Spears & Postmes (1995) put it, far from crowd action being uncontrolled, perhaps it is uniquely in the crowd that people can fully express their social identities.
Following years in which social psychology has ignored the issue of power (Cartwright, 1959; Dépret & Fiske, 1993; Ng, 1980; Sachdev & Bourhis, 1985; Tedeschi, 1974), the recent turn to issues of empowerment in collective action and intergroup relations is surely to be welcomed. Yet it is arguable that accounts of empowerment which stress ‘efficacy’ or situational determinants are somewhat static. Either power is a state that pre-exists collective action or else it is automatically and fully in place as soon as the collective forms. In either case, there is no consideration of the possibility that a sense of power develops in the course of collective processes. However, documentary studies of social movements and crowd events commonly point to the occurrence of empowerment not just as a pre-condition, nor as a simple correlate of collective events, but also as an outcome. For example, the ghetto rioters of the 1960s who felt their action achieved something against a normally powerful outgroup subsequently felt satisfaction, pride and confidence in their identity as participants - personally as well as collectively (Boesel, Goldberg & Marx, 1971). Similarly, commentators on the French events of May 1968 describe how occupying students displayed increased confidence in their own abilities and capacities: ‘The occupants of Censier suddenly cease to be unconscious, passive objects shaped by particular combinations of social forces; they become conscious, active subjects who begin to shape their own social activity’ (Gregoire & Perlman, 1969, p. 37; emphasis in original). While an account of the preconditions of empowered collective action is necessary, this must not preclude an understanding of the emergence of empowerment in the collective; indeed, ideally what needs to be shown is the possible relations between these two aspects of power in collective action. What is necessary, in other words, is an account of the dynamics of empowerment processes in crowd events.
Recently, the social identity model has been elaborated precisely in order to account for developmental changes in the course of crowd action (Reicher, 1996, 1997; Stott, 1996). The key to these elaborations is the observation that crowd events are typically intergroup encounters and therefore the position of any one party must be understood in relation to the ongoing intergroup dynamic. In a number of studies (Reicher, 1996; Stott, 1996) it is shown how outgroup power may serve to create the context within which crowd members define themselves. In particular, where an outgroup are seen to act against all participants as if they were the same, this creates conditions under which those participants are likely to define themselves in terms of a single all inclusive categorization. Moreover, where the outgroup action is experienced as illegitimate, this creates conditions under which the inclusive ingroup will sanction resistance against that action. What is more, outgroup action does not simply provide the inclination for the ingroup to resist, it also provides the means. Whereas, previously, the crowd may have been made up of a plethora of smaller categories, each of which felt relatively powerless in the face of the police presence, the formation of a single large category along with the feelings of consensus and the expectations of mutual ingroup support that are created by common categorization empowers the group to oppose the police.
While the elaborated social identity model provides a theoretical basis for understanding the emergence of empowerment in crowd events, and while the empirical studies show evidence of behavioural shifts which are consonant with the argument that people come to be empowered as a function of the enlarged ingroup category, there is virtually no direct evidence to show that people do indeed become empowered in this way. There is little material in their studies on crowd members' actual subjective feelings of empowerment, and no material at all on the attributions of any such feelings - whether to collective support or any other factor. All in all, the model of empowerment is inferred rather than demonstrated.
The aims of the present study are therefore threefold. First, and most generally, the aim is to examine whether and the extent to which empowerment can arise from collective action. Second, and more specifically, the study aims to investigate the impact of intergroup actions in altering self-categorizations and hence power relations. The third and final aim was related to the Le Bonian claim that power and loss of self control go together in the crowd. The study therefore aims to investigate whether crowd simply used their power to ‘do anything’ or whether (as the social identity model would suggest) empowerment enabled crowd members to express their shared identity - and hence meaningful limits to crowd action remained in place. To these ends, a detailed study of empowerment processes in a crowd event - an anti-poll tax demonstration at a town hall - was carried out. At this demonstration, a council meeting was partially disrupted and a crowd outside the building came into prolonged conflict with the police as they tried to push their way past officers and into the chambers. The analysis focuses on participants' feelings of empowerment and the possible factors contributing to these feelings both during and after the incidents.
Selection of the event
There were two main criteria for the selection of a crowd event for this study. First, the issue of power had to be central to the actions of participants and to their development. While power issues may be relevant to all crowd events, they will be most clearly revealed in situations of constraint where at least one party seeks to repress what at least one other seeks to do. Moreover, according to the theoretical model we have outlined, the development of power relations depends upon asymmetries in the way that different parties view what rights obtain in the situation and what forms of action are legitimate. For the purposes of our analysis, then, the event had to be contested.
Secondly, it was necessary to choose an event in which we could collect the data necessary for our analysis. This is not just an issue of access to subjects. In analysing contested events where crowd members are doing things that are opposed by police and local authorities and where the topic concerns acts that might be censored by these authorities or even be illegal, it is necessary to have the full trust of respondents. This is complicated by the fact that members of many groups in protest have a generic distrust of academics who they see as implicated in the system that is being opposed.
On these twin grounds, we chose the protest against the setting of the poll tax by Exeter City Council in March 1990. On the one hand, the event was clearly contested. Indeed it was possibly the most intense collective confrontation which had occurred in the city for many years. On the other hand, the researchers had good contacts both amongst the protestors and amongst the councillors. On the basis of these contacts a sizeable number of participants were prepared to discuss their perspective and their actions in some detail.
(a) Participants' accounts Twenty-nine of the approximately 200 demonstrators were interviewed, including five of the 20 or so people who got inside the town hall. Twenty of those interviewed were male, nine female. Ages ranged from early 20s to 50s, with late 20s being the modal age. Interviewees were an opportunity sample; over a period of around seven months, attempts were made to contact everyone who was present at the demonstration. The local anti-poll tax union initially supplied some names; each of these was asked if they knew others who were present who might be willing to be interviewed. No-one who was contacted refused to participate although in one case it was impossible to find a time to organize an interview. Four of the interviews took place with groups of two or three at a time. Such a group interview technique can be used to reproduce the social relations through which individuals experienced collective events. It may therefore not only aid memory for perceptions at the time but also produce some of the identity-based actions (shouts, feelings etc.) that occurred in the earlier context. (See Banks (1957) Burgess (1982) and Reicher (1996) for discussions of group interviews.) The standard interviewing technique was to show interviewees an edited version of the video of the events. This video, taken by a protestor who was outside the council meeting, was edited down from 60 minutes to 25 minutes for reasons of practicality. It covered all the phases of the event from initial crowd formation to crowd dispersal. The video both served as a further memory prompt, and was used as a stimulus for the interview questions, using a semi-structured format. Finally, all interviewees were shown a scrapbook of news cuttings about the event, and asked to comment on the accounts. (See Whyte (1984, p. 106) and Thompson (1978, p. 173) for use of a similar technique in interviewing.) Most of these interviews lasted about an hour.
Fourteen of the councillors who present at the meeting that evening were contacted. Seven of them agreed to be interviewed, two of them together. Three of those interviewed were Conservatives (including the then mayor), two were Labour members and two were Liberal Democrats. An additional two provided written accounts. Of the remaining five, one failed to reply and four were unavailable for interview for a variety of reasons. As well as the councillors, five police officers were interviewed: a superintendent , three constables and a sergeant. One of the two mace-sergeants (town hall security/guides) on duty that evening was also interviewed. The scrapbook was used on each occasion. However the video was only used when interviewing the mace sergeant and the former mayor. In the other cases the interviews were carried out, at the interviewees' request, in council or police offices without access to a video player. These interviews each lasted about half an hour.
The interview schedule for crowd participants covered the following issues: (i) background (e.g., ‘What was the poll tax issue all about?’); (ii) nature of the crowd participants (e.g., ‘What kind of people were there?); (iii) relationship with the rest of the crowd (e.g., ‘What did you think of the rest of the crowd?’); (iv) the police (e.g., ‘What was your relationship like with the police at the beginning? Physically, emotionally, socially.’); (v) the council (e.g., ‘How did you see the council?’); (vi) aims and expectations (e.g., ‘What did you hope to achieve?’); (vii) changes in behaviour (e.g., ‘At what point did people start pushing to get into the town hall?’); (viii) feelings of power (e.g., ‘Why did you feel able to do this (push and shove like this) [if applicable]?’; (ix) feelings afterwards (e.g., ‘How did you feel about the events afterwards?’). Versions of questions under these headings were asked at different points during the showing of the video to grasp any changes in participants' perceptions during the event. A similar schedule, with obvious modifications, was employed for the interviews with police, councillors and the mace sergeant.
Additionally, two demonstrators, two councillors and a police officer provided written accounts using questionnaire versions of the interview schedules.
All interviews were taped and transcribed. All interviewees and those providing written accounts were guaranteed anonymity. In the analysis, initials D = demonstrator and P = police officer. The reference at the bottom of each excerpt of interview text is to the date of the interview [dd-mm-yy]. Transcription conventions are based on those in Parker (1992) and Potter & Wetherell (1987):
When material has been edited out of the transcript, it is signalled with an empty pair of square brackets, thus [ ].
Where information has been supplied to the text, it is put in square brackets [like this].
Where material is unclear or inaudible, empty round brackets are used, like this ( ).
Where sound quality leads to doubts about the accuracy of material, it is put in round brackets (like this).
(b) Video footage Video footage taken by one participant was made available on request. The footage is of the crowd outside the town hall and lasts about an hour, covering the period during which the crowd built up until people started to drift away and the conflict ended.
(c) Newspaper accounts The analysis makes use of accounts of the event in three local newspapers, plus a more detailed account in an alternative publication (The Flying Post), written by a demonstrator. All accounts include a number of pictures of the event.
In the first place, a consensual account of the event was constructed, using different sources (demonstrators, police and press) (cf. Denzin, 1989). The rationale for constructing a consensual account is not only to guide the reader, but more importantly to identify some of the patterns of behaviour that need to be explained in the analysis proper: what were participants' feelings and perceptions during behaviours identified in the consensual account? Moreover, the extent to which the theoretical account offered on the basis of these feelings and perceptions helps explain the evidence in the consensual account can be taken as one index of its adequacy as an explanation (cf. Reicher, 1996).
Consensus was operationalized as follows. If a claim was made by different sides (demonstrators on the one hand and police, councillors, mace sergeants or press on the other), or was supported by the video and at least one other source, and if nobody disputed the claim, then it was considered consensual. Where accounts diverged or only one source existed for a particular claim, references are given (in brackets).
The analysis proper is broadly in the tradition of thematic analysis (cf. Kellehear, 1993). However, since the rationale for the study is to explore the model provided by Reicher (1996) and Stott (1996), analytic categories are based on previous work on social identity and collective action rather than derived entirely ‘bottom up’ from the data (cf. Glaser & Strauss, 1967). Material was read to identify: (i) any reasons participants gave for their actions; (ii) any talk of the legitimacy or otherwise of own or other's actions; (iii) any talk of feelings of power and/or powerlessness at different times during the event; (iv) any perceptions the various participants held of the crowd at different times during the event - in particular of mutual support and unity (or lack of it).
An account of the event
(a) Background to the protest: On Tuesday 6 March 1990, Exeter City Council met to set its community charge or poll tax for the first time. The Conservative government had introduced this new flat-rate local taxation scheme to replace rates, which were based on property values. Although many Labour-run councils voiced opposition to the tax (as in Exeter), none of them refused to set it. During 1989 and early 1990, a nationwide movement developed against the tax which critics argued was regressive and therefore unfair; hundreds of local anti-poll tax ‘unions’ were formed across the country in an attempt to co-ordinate local opposition to the tax and to encourage resistance through mass non-payment. In the weeks previous to the Exeter demonstration, council poll tax rate-setting meetings at town halls around the country had been met with protests, ranging from demonstrations outside the buildings to disruption inside (Burns, 1992). In Exeter, the local anti-poll tax union heard about the city council meeting at short notice and publicized by word of mouth, through street meetings in the city centre and by flyposting throughout the city.
(b) The initial period - crowd formation: The meeting was due to begin at 6 p.m. At 5.15, a crowd of about 50 were already gathered outside the doors of the town hall in Exeter High Street. A police public order van was stationed a few yards along the road. By about 5.30, the crowd had swelled to around 100-150 people. Three police officers stood at the door outside the building. Demonstrators at this stage were mostly standing around chatting in small groups.
When the councillors started to arrive there was some booing from the crowd and shouting of ‘Don’t pay’ (video). The crowd was packed tightly round the door at this point and councillors had to pass amongst the demonstrators to get into the building; people in the crowd spoke to the councillors or shouted at them. Some councillors avoided the crowd by going in at the back of the building. As the door closed on the councillors, the crowd gathered round the doors, facing the entrance to the building (video). There were shouts of ‘Let us in’, ‘Public meeting’ and other complaints. Then the first sustained collective chanting began - ‘We won’t pay the poll tax’; most of the crowd faced the doors and joined in (video). This was followed by other collective chants: ‘Can’t pay won’t pay’ and ‘They say poll tax we say fight back’. Shortly afterwards, some demonstrators were let in. There is some disagreement over the number (both within and between the groups of police, councillors and demonstrators), with estimates ranging from ten to thirty. There is no dispute, however, that those who went in were simply those nearest the front of the crowd, and, more importantly, that the rest of the crowd voiced displeasure that more were not let in.
(c) The middle period - conflict inside and outside the town hall; Demonstrators in the public gallery were quiet at first, but soon unfurled an anti-poll tax banner and disrupted the chaplain’s prayer and the councillors' speeches with shouts and heckles. The mayor called for order three times before suspending the meeting and asking that the demonstrators be removed. At this point, a volley of food was thrown by the demonstrators and struck some of the councillors. The demonstrators then made their exit.
By this point the crowd outside numbered about 200. The majority were close to the building facing the doors (video). There was loud and sustained chanting of ‘Let us in, Poll tax out’ (video). There were two police public order vans in the street by then, and the number of police on the door had increased to seven (video). For a period, crowd members made sustained and vigorous pushes against the doors. In the process, those police officers standing in front of them were pressed up against both the doors and adjacent railings. The pushing was accompanied by the chant ‘In In In’ (video). There were struggles between the police and members of the crowd: the video shows crowd members shouting police numbers and calling for help. It also shows heated arguments between police and crowd members. A number of interviewees mentioned crowd members being rescued from arrest by their fellows during this period. Some accounts on both sides describe blows being exchanged. Glass was broken and demonstrators threw some missiles, specifically eggs and flour, at the doors. Many accounts speak of tempers being raised. However, most interviewees agreed with both the claim in The Flying Post that ‘there were more helmets flying than fists, and the overall impression was of a lot of ritualised pushing and shoving punctuated by the occasional angry outburst’ (April 1990, p. 13).
When the pushing became more concerted outside, there were attempts by the demonstrators in the lobby to open the doors to let people in. A number of accounts from both mace sergeants and demonstrators agree that the doors nearly burst inwards with the force from both sides. Some of the demonstrators inside then went upstairs into other parts of the building. One of them set a fire alarm off. This was met with cheers (video). All those inside were finally ejected through the backdoor of the town hall.
(d) The final period - dispersal and aftermath: A fire engine arrived in response to the alarm, blocking traffic in the High Street. More police were stationed at the town hall entrance (about ten altogether). While the fire officers consulted with the police, the crowd spread about the area and became more relaxed - demonstrators sat in the road or stood around chatting, and less faced the town hall doors. However, shortly afterwards, demonstrators started chanting ‘Poll tax out, Let us in’ again and people approached the doors once more (video). The crowd had diminished by about 100, and there was a space now between the police and the crowd (video). The council resumed its meeting, meanwhile, and finally, just after 7.25pm, a council official announced to those left outside that the council had set the tax (demonstrators). At this point the demonstrators rapidly dispersed
The event made the front page of the local newspaper the next day, and there were some accusations in the press of ‘travelling outsiders’ being behind what happened. All the demonstrators interviewed scoffed at this, mostly saying instead that the crowd comprised a heterogeneous group of local people. Some of those councillors who were interviewed also commented that they recognized many of those present as local people. The police tended to describe those active in the protest as homogeneous (e.g., ‘lefties’) but local.
For reasons of space, the analysis will focus only on the behaviours and experiences of the largest section of the crowd: those outside, rather than inside, the town hall. The analysis will examine: (i) whether and to what extent demonstrators felt empowered during the event; (ii) features of the changing intergroup relationship seen to be responsible for any such feelings of empowerment; and (iii) any normative limits to behaviour within such empowerment. Accordingly, the analysis is divided into three sub-sections.
Evidence of empowerment
(a) Feelings and perceptions: Some demonstrators said that, during the early stages of the demonstration, the swelling numbers produced in them a growing sense of power:
D10: As more and more people got there there was a feeling of power and just feeling of erm being able to have our say, you know, really.
Such comments increased in frequency among interview responses when demonstrators were asked about their feelings when the crowd was chanting and pushing at the town hall doors:
D23: Yeah, because of the size of the crowd and the strength of feeling in the crowd, that we had a certain solidarity there
However, it is clear that it wasn't numbers per se that made demonstrators feel confident, but rather the feeling of unity and hence the perceived support for their actions that such numbers offered:
D22: It was a feeling of a group of people acting together (the sort of thing) that very few people there would have done on their own. I don’t know, a sort of crowd consciousness, we could do this, we had some sort of power together to actually do it, so we did
(b) Support for specific actions - pushing chanting and resisting arrest: Some demonstrators explicitly mentioned the encouragement they took from others' behaviour. They said that they began to shout or push because they had gained the confidence to do so within the crowd; the fact that others were already doing so made them feel supported in taking such action themselves:
D21: There’s always a certain amount of nervousness to start anything like that [chants or pushing] until you’re absolutely sure you’ve got enough people there that are also into it
Int: Right You felt that you felt that during this, did you?
D21: Oh definitely there was enough people to be into it during this
Moreover, this support took the form not simply of co-action, but also of verbal encouragement by others who were not necessarily themselves joining in:
Int: When you were pushing how many people were joining in did it seem like most people or
D22: Oh about two thirds of the crowd there, I think. There was a few people outside standing back but supporting. Yeah there was a lot of people who I wouldn’t normally have expected to join in, there there were some very respectable people [ ]
Int: [ ] how could you tell they were supporting you?
D22: ( ) cheering. Cheering or chanting from behind. People in buses cheering, bus drivers hooting
As we have indicated, one of the consequences of the pushing and shoving was that the police attempted to arrest some crowd members. The sense of empowerment which led people to participate in the pushing, also led them to resist the police reaction. What is more, once again the sense of empowerment was related to a sense of unity and the consequential expectation that any crowd member would gain support from others:
D19: There was two police officers on me had got both of my arms both sides up behind my back, got me about face and dragged me out obviously arresting me. [ ] I just had the feeling or the sense that if I shouted out ‘help’ no I’m not gonna passively be nicked; potentially the situation here where collectively we’ve got the strength to prevent myself being arrested or anyone else being arrested for that matter. In the same way that I felt that I had the strength to be able to move in and try and prevent someone from being physically hurt [ ] I didn’t see exactly who did it but, those two police officers were instantly removed. Two or three people leapt onto each police officer (and) literally physically dragged both police officers off me. [ ] I was free to move so I escaped like the person I helped earlier
(c): Ingroup power and outgroup vulnerability: This last extract raises a basic point about power and empowerment: these are not simple properties of any given group but rather relations (and perceptions of relations) between groups. Thus, the unity of the protesting crowd served to empower it in relation to the police and authorities. In the specific case, the strength of the ingroup is a strength to resist arrest by the outgroup. Hence, in considering empowerment, it is important to consider not only how strong ingroup members considered themselves to be, but also how weak they saw the outgroup to be.
Most of the demonstrators said they felt that the police were vulnerable once the pushing started. This vulnerability was a product both of their physical location and of their small numbers in relation to the crowd:
D8: I think there was a certain amount of vulnerability there [ ] I think they did feel vulnerable to the efforts of the crowd [ ] so the police were actually forming a line in front of an impregnable barrier and, you know, if I was one of them I wouldn’t fancy being in the situation having like a large degree of people pushing at me against what effectively is a wall
It is worth noting that this perception was shared between demonstrators and the police. Whereas D23 referred to "the police as severely under-resourced there", one officer commented that he felt the police lost control and had to send for reinforcements since "we didn't expect the volume of people there or the trouble" (P23). When it comes to explaining how new possibilities and new actions arose in the crowd, it is neither the sense of ingroup strength nor the sense of outgroup vulnerability alone which are important. It is the combination. It may be that sometimes demonstrators mention the one and leave the other implicit. However, sometimes they are explicit in invoking both:
Int: At what point did your expectation change about what you thought you could do then? I mean you said initially you went there just to sort of erm
D23: Well I think as the numbers built up, as we saw it was such a big demonstration and such an angry demonstration that there was a real feeling of passion about this issue, that was when I think perceptions changed, yeah.
Int: Did you join in with the pushing?
Int: And what did you hope to achieve by getting in?
D23: Stop the council meeting
D23: We saw our opportunity because we saw that the police were so undermanned for the occasion.
To summarise thus far, the analysis indicates that participants did indeed come to see themselves as more empowered in the course of events. This sense of empowerment was invoked in explaining the emergence of new forms of action whereby protestors challenged and resisted the police. It was also invoked in explaining how the goals of crowd members became more radical. Empowerment itself was explained as deriving from a sense of collective unity. This gave the crowd the strength of numbers and made the police seem vulnerable in comparison. Before going on to analyse how the sense of unity (and hence of empowerment) was itself produced, it is worth making one final point which makes such analysis all the more significant. From what our respondents said, the empowerment that emerged in the events and the new sense of possibility to which this gave rise did not simply disappear as soon as the crowd dispersed. As in the examples of ghetto rioters and the French students of 1968, empowered self-perceptions endured and formed a new starting point for subsequent forms of collective action:
D11: I mean we were still on a high for ages afterwards, after doing that we had a fucking great party that night, everyone was just on a high after doing this and congratulating ourselves for making a statement in Exeter for once in Exeter's life
Intergroup processes and empowerment
(a) Initial heterogeneity: To trace out the relation of the above evidence of empowerment to features of the intergroup situation, it is first necessary to indicate what the demonstrators brought to the event. Not surprisingly, those who had come to protest against the poll tax that day explained their decision in terms of the perceived iniquities of the tax. Many opposed the tax in universalist rather than partisan political terms. In one group discussion, D1 objected to "the unfairness of the system which was charging everybody the same amount regardless of the size of house". D2 and D3 assented to this view and D3 added "it seemed to be a moral issue really".
If the demonstrators agreed on opposition to the Poll tax, they also shared the belief that all had the right to let their representatives know of their opposition. This meant not only the right to protest outside but also the right to observe inside if they so desired:
D23: I think a lot of people felt very strongly it’s their council they’ve got a right to go and see it.
D9: There was perhaps a rather naive thought at the start that everyone outside would actually be allowed to go into the public gallery [ ] and the feeling was well there’s a constitutional right to go in, there’s a public gallery there, the meetings are open, we are allowed in [ ] that’s our right to go in
However, beyond their right, as an opposition, to be present, there was little agreement amongst demonstrators - certainly not when it came to the purpose of their presence. For some, this presence was an end in itself since it would draw attention to the issues involved around the poll tax. Thus, D8 defined the aim as "just an effort to get publicity for the cause really". For a small number of others, such as D9 the purpose was more radical. As he put it: "my aim was to go into the public gallery and, well, to voice disquiet to put it least and to disrupt the meeting".
However the heterogeneity of the crowd was not only a matter of ideas. A number of respondents stressed the initial lack of unity amongst crowd members: the fact that different types of people were present and the fact that they kept aloof from each other. D15, for instance, said of this early stage that "I think they saw themselves as individuals, I think". D16 gave an even stronger sense of the barriers that existed within the crowd:
D16: I think in the beginning people did tend to stand in their own little cliques if you like
The term 'clique' clearly signifies exclusivity and difference. What is more, the psychological distance is reflected in physical distance and lack of interaction between subgroups in the crowd - something that is confirmed by the video evidence. At the outset of events then, the protestors cannot be described as a single category either psychologically or behaviourally and even physically there was distance between them.
(b) The production of unity: If the demonstrators, despite their heterogeneity, believed that they had a right to attend the council meeting, then the act of excluding most people from the town hall was seen by all as illegitimate, irrespective of whether the individuals themselves had intended to go inside rather than stay outside. Thus D23, quoted above as saying that protestors felt strongly about their right to see 'their' council in action, continued by saying that "they were quite surprised and shocked to be kept out". In most of the interviews, respondents echoed the notion that 'rights' of attendance rendered exclusion illegitimate . D1, for instance, expressed his anger that, in a supposedly democratic country where "you should be able to observe the democratic process going on" he and fellow protestors had been prevented from doing so. Such a sense of outrage is also apparent in the contemporaneous comments of demonstrators, captured on video, as the doors were shut on them:
D4: I thought this was a council building, is that right?
D5: Public meeting! Public meeting!
D4: Out of order.
D6: What happened to democracy?
Most interviewees attributed the beginning of the conflict between police and demonstrators to what they saw as the illegitimate act of exclusion. These claims are supported by the video evidence which shows that the more intensive chanting and the pushing itself began only subsequent to the refusal to admit more than a small minority to the public gallery. What is more, all those interviewees who said they participated in the pushing mentioned the denial of their right to get in as justification for their actions:
Int: Why did you start pushing and shoving?
D15: Because they were locking us out. I was aware that some people had got in ( ) I didn’t know exactly what had happened, what number they’d stopped people going in, but they [ ] had stopped people going in. And my response to that was that it was that people had the right to go in. And I suspected that the gallery wasn’t full, they’d just got scared and I think meetings should be public, accountability ( ) and so on. So yeah I was having a little shove there.
From a position in which the crowd was divided into differing groups with different aims, the common sense of injustice caused by the actions of the authorities led to a common sense of purpose and therefore common forms of action. We have already shown how this psychological unity was reflected in forms of interaction and altruism within the crowd: individuals would even risk arrest to stop fellow crowd members from being taken by the police. However this interaction was reflected in more banal (and hence more pervasive) ways as well. People who hitherto had been in 'cliques' began to chat and joke together. In the process, psychological unity came to be reflected in the physical unity of the crowd as well:
D6: I suppose as well what wasn’t happening so much before was that people were talking to each other, not just the- I mean, you know, ( ) not just people who knew each other, but everyone was there together and you could chat to someone next door you’d never met before in your life and if someone made a joke everyone was going ‘yeah yeah really funny’, you know, whereas before you wouldn’t do that cos you were just standing around in your own little groups
Limits to action
So far we have established that at least some demonstrators came to feel empowered during the course of their protest and that this sense of empowerment can be related to the production of crowd unity caused by their exclusion from the council meeting. The aim of this final section is to challenge the Le Bonian assumption that the power of crowds is matched by their lack of control. Were the demonstrators prepared to do just anything once they discovered the vulnerability of the authorities in the face of their unity? Certainly, the authorities themselves perceived that control had been lost and that, if extreme behaviour had not occurred it could do so at any time. Thus one police officer (P3) described how "pushin' and shoving and shouting and kicking started and almost got out of hand". He went on to emphasise that: "someone sooner or later was bound to get hurt. Bound to". Other officers were more specific, expressing their personal fear of what the crowd might do to them:
P2: I can remember being crushed against the wall and people surgin' forward and I thought I’m gonna be squashed here because I could feel the wind being pushed out, it really was quite frightening. I would say that it was the most frightening situation I’ve ever been in. Yeah. [ ] Because there was only about six of us there [ ] it was mayhem that’s all I can describe it as, there was a lot of pushin' and shovin', they were totally out of control I believe
There are three things to note about these claims. The first is that there is some ambiguity about the concept 'out of control'. It is not always clear whether officers are referring to crowd members as being out of police control or as lacking control over their own behaviour. Indeed the two tend to be conflated in such a way that the loss of police dominance is equated with the danger of complete social breakdown. Secondly, we have previously noted that outgroups frequently perceive crowds as limitlessly violent (Reicher & Potter, 1985). Seeing them act aggressively without understanding the basis or the constraints upon that violence, how are outsiders to know if they too might not be targets? Hence their perceptions of the crowd are dominated by fear (Reicher & Potter, 1985). This sense of potential danger to the self is quite explicit in the police accounts. The third point is that, while the officers describe the imminent danger of destruction and injury (particularly to themselves), they do not give evidence that such events occurred. Indeed they did not. However powerful the crowd may have been, their aggression against the police and others was not taken to extremes.
Crowd members understood their own action to be operating within clear, collectively defined limits of legitimacy. Some of those who pushed against police lines mentioned that they felt the crowd was physically capable of getting in to the building if people were really determined. However they also explained that they would not entertain just any actions in order to achieve their ends: while they had a legitimate right to get in, it would be illegitimate to injure the police in the process. D22 referred to a collective sense that crowd members would not "go around hitting people there", while D23 used a similar claim to explain why, despite their power, the crowd didn't simply overwhelm the police and enter the building:
D3: I think it would have been quite difficult for the police to keep people out if people were really determined to get in. But that would have meant being quite quite heavy and physical with the police, and I think most people weren’t really thinking in those terms
This evidence suggests not only that there were limits to crowd action and that these behavioural limits match collective notions of legitimacy. If the police were targets of crowd aggression, there were constraints on the way in which that target would be attacked: pushing, shoving, shouting were permitted; hitting and trampling were not. In addition, there were obvious limits to which targets were chosen for aggression in the first place. The police and council authorities were the subject of concerted collective attack insofar as their presence and their actions were considered as illegitimate. No other targets were chosen: adjacent shops and passers by were completely left alone. This seemed so obvious to the protestors that it hardly rated a mention. However it is of theoretical interest in showing that crowds employ their power with considerable discrimination both in choosing what to attack and how to attack it.
The present analysis of the behaviours and perceptions of demonstrators in a town hall anti-poll tax protest has suggested how a sense of self-empowerment can be understood not simply as a pre-condition of collective action but also as an emergent product during such action. The sense of empowerment emerged through the changing nature of crowd members' collective relationship with outgroups (the police and/or council); the actions of these outgroups served to create a more inclusive self-categorization for crowd members, which in turn had implications for expectations of support.
The analysis suggests that many of the demonstrators came to feel increasingly empowered in the course of the event: they felt more able to carry out various forms of activity which would not have happened otherwise, including shouting, pushing against the town hall doors and mutually resisting arrest. Although the demonstrators began the demonstration as relatively disparate sub-groups, unity amongst them (beyond the fact of turning up together at the demonstration itself) was most evident after most of them had been excluded from the town hall. Demonstrators were excluded from the meeting of their council wholesale, irrespective of the different rationales that they had each brought to the demonstration. Since they now shared the same identity, demonstrators came to expect support from each other as they acted against this illegitimate exclusion. Moreover, the perceived impotence of the police qua an outgroup in the face of their unified action served to encourage the demonstrators to go further than they may have originally intended in their actions against the council meeting. Finally, although to outgroup members the behaviour of the crowd appeared out of control, there were clear, collectively defined limits to what crowd members did and considered appropriate (e.g., the reluctance to use violence).
Consistency among interview responses supports the claim here of a clear pattern in the fact of empowerment (its commonality within the crowd), the causes of it (support through an emergent common relationship to outgroups) and its consequences (acting within shared conceptions of legitimacy). However it is important to stress that, while evidence of empowerment was to be found in the case of most of the interviewees, our argument has not been that everyone present on the anti-poll tax demonstration came to feel more empowered or that all demonstrations lead to empowerment. Obviously, without knowing the distribution of prior beliefs about legitimacy, or indeed prior feelings of empowerment, it is not possible to state whether and how many people will change in any particular crowd event. The aim of the analysis has been to show that people may become empowered through participation in collective action and to suggest the conditions under which this can occur. In particular, individuals will only come to feel empowered where they perceive both that the ingroup category has become extended and where this is sufficient to negate the repressive power of the outgroup. These shifts in categorisation will only occur where the outgroup has different notions of acceptable action to those in the crowd and where that outgroup acts to impose its notions so as to produce a common sense of grievance amongst the ingroup. These conditions may be relatively infrequent. Most crowd events remain largely consensual and peaceful (Reicher, 1987). Our focus on a conflictual event is not to perpetuate the dangerous assumption that crowd events are inherently dangerous but rather because of the theoretical significance of such conflict.
Of course, as with the work by Reicher (1996), our theoretical claims are made largely on the basis of post hoc data and that means the study is open to certain criticisms. First, since all of the interviews took place after the event itself, memory lapses were inevitable. Perhaps the main problem with relying so heavily on retrospective interview data, even with the use of video material as a supplement, is that of reconstructing the sequence of empowerment - how the different elements of action and perception might fit together as a whole. Clearly, a further study of the sequential contributions of such factors as the outgroup inability to respond and the role of collective support is necessary, this time using more contemporaneous data. Second, there is the likelihood that participants' responses might be shaped by a number of strategic considerations, both in relation to the presentation of the events and in terms of the interpersonal setting of the interview itself: participants may be constructing rather than recalling (cf. Edwards & Potter, 1992).
While these are potentially serious epistemological objections, a number of points can be made in defence of this study. In the first place, the limits of the data must be judged in the light of the well-documented problems of collecting any data at all on crowd events (e.g., Adams, 1992; Milgram & Toch, 1969; Wright, 1978). Second, while there may well have been a number of memory limitations, cognitive biases or discursive strategies operating in the interviews, the plausibility of the interview responses as reconstructions of participants' feelings and motivations at the time is supported by both their consistency with the verbal material on the video tape and by the extent to which these accounts of perceptions make sense of the contours and dynamics of the crowd's behaviour. There is a match between interviewees' accounts of changes in feelings (of legitimacy and power) and the consensual evidence of ‘escalation’ of the protest (e.g., trying to push to get in only after being illegitimately excluded). The value of the present account, like any other, lies in part at least in its explanatory adequacy to the evidence.
An objection might also be raised regarding the representativeness of the interview sample. Given that the main contact source, initially at least, was the local anti-poll tax union, the people interviewed in this study may have been the more active participants in the crowd; others may have had a quite different experience. On the other hand, the relative size of the sample (around 15 per cent of the crowd) tells in the study's favour; not all the interview sample were involved in the pushing (for various reasons - some were at the demonstration with small children, for example). What is more it is worth reiterating that we are not seeking to establish that everybody became empowered in the crowd but rather that people can be empowered in the process of interaction and, when they are, this can be related to shifting intergroup dynamics. Even if 'changers' are over-represented in our sample, the fact that such change occurs and how it occurs is sufficient to ground our argument.
The social identity model of crowd behaviour defines identity as a conception of proper and possible action in social relations (Reicher, 1996). The model has been developed to show how the nature of self is not only a social product (as in the typical self-categorization study; e.g., Oakes, Haslam & Turner, 1994; Turner, Oakes, Haslam & McGarty, 1994), but is also a social producer. This duality flows from the fact that self exists in the relationships between groups - and sometimes in the overt struggles between them. It is particularly in situations of crowd conflict (whether ‘violent’ or not) that the created-creative nature of the self is most apparent - largely because 'context' is most clearly seen as the actions of the other and hence a fluid product of intergroup interactions. The present study has been consistent with the account of Reicher (1996) and Stott (1996) whereby collective action needs to be grasped in its historically developing intergroup context, where there may be conflict over what counts as legitimate behaviour (e.g., the right to get into the town hall) and where there is a difference in the power of each group to impose its definition. The interaction between conflicting groups may serve to change each other's social context; this changed context may form the basis of a changed identity.
This study has provided evidence that the issue of collective empowerment also needs to be analysed in such a historical and interactive manner. In so doing it both endorses the growing interest in power as a key mediator of collective action and simultaneously challenges the static manner in which power has been conceptualised. The study also provides further support for the elaborated social identity model of crowd behaviour while at the same time suggesting some further extensions. In particular, it further underlines the point that power must be conceptualised as a social relation rather than a group attribute. Thus what is important is not only the way in which unity provides strength to the ingroup but also how the outgroup is rendered vulnerable as a result. Power refers to the relative ability of each group to impose itself on the other and constitute the context in its own terms.
On a more general level, our aim here has been to add to the rehabilitation of crowds within psychology - to show that their action is not just meaningful but also creative. Crowds produce change. It is therefore worth stressing that the change brought about during the events we described was not just confined to those events. We have noted how at least some crowd members report that they joined subsequent protests with an enhanced sense of what they were capable of achieving and with a greater sense of the vulnerability of the 'authorities' rather than their own. This extended to quite different actions such as refusing to pay the poll tax and a willingness to challenge the law. As D15 put it: "it made me feel braver I suppose".
Theoretically, this only goes to underline the importance of our general approach. Not only might crowd members' sense of empowerment develop during crowd events but also the point which they start from may be the result of prior experience in crowds. Methodologically, it shows that studies of single events are inevitably limited. A properly historical study should look not only at developments within a single events but also the development over multiple level campaigns. In both senses, this study shows the necessity to make the investigation of social psychological constructs such as power more historical. It also shows the value of using the crowd as a domain in which to investigate that history.
Abrams, D. (1993). From social identity to action. Paper given at BPS Social Psychology Section Annual Conference, Jesus College, Oxford, September.
Adams, R. (1994). Prison Riots in Britain and the USA. (Second Edition). London: Macmillan.
Banks J.A. (1957). The group discussion as an interview technique, Sociological Review, 5, 75-84.
Boesel, D., Goldberg, L.C. & Marx, G.T. (1971). Rebellion in Plainfield. In D. Boesel & P.H. Rossi (eds.), Cities Under Siege: An Anatomy of the Ghetto Riots, 1964-1968, pp. 67-83 New York: Basic Books.
Breakwell, G.M. (1992). Processes of self-evaluation: Efficacy and estrangement. In G.M. Breakwell (ed.), Social Psychology of Identity and the Self Concept, pp. 35-55 London: Surrey University Press.
Burgess, R.G. (1982). The unstructured interview as a conversation. In R.G. Burgess (ed.), Field Research: A Sourcebook and Field Manual, pp. 107-110 London: Allen & Unwin.
Burns, D. (1992). Poll Tax Rebellion. Stirling: AK Press.
Cartwright, D. (1959). Power: A neglected variable in social psychology. In D. Cartwright (ed.), Studies in Social Power, pp. 1-14 Ann Arbor: Institute for Social Research, University of Michigan.
Chase, J. (1992). The self and collective action: Dilemmatic identities. In G.M. Breakwell (ed.), Social Psychology of Identity and Self Concept, pp. 101-127 London: Surrey University Press.
Cocking, C. (1995). Comparing attitudes towards the environment: Does direct action make a real difference? Unpublished M.Sc. dissertation, University of Surrey.
Denzin, N.K. (1989). The Research Act: A Theoretical Introduction to Sociological Methods. (Third edition). New Jersey: Prentice Hall.
Dépret, E. & Fiske, S.T. (1993). Social cognition and power: Some cognitive consequences of social structure as a source of control deprivation. In G. Weary, F. Gleicher & K.L. Marsh (eds.), Control Motivation and Social Cognition, pp. 176-202 New York: Springer-Verlag.
Diener, E. (1977). Deindividuation: Causes and consequences. Social Behavior and Personality, 5, 143-155.
Diener, E. (1980). Deindividuation: The absence of self-awareness and self-regulation in group members. In P.B. Paulus (ed.), Psychology of Group Influence, pp. 209-242 Hillsdale, New Jersey: Lawrence Erlbaum.
Edwards, D. & Potter, J. (1992). Discursive Psychology. London: Sage.
Gamson, W.A. (1975). The Strategy of Social Protest. Homewood, Illinois: Dorsey.
Gamson, W.A., Fireman, B. & Rytina, S. (1982). Encounters with Unjust Authority. Homewood, Illinois: Dorsey.
Glaser, B.G. & Strauss, A.L. (1967). The Discovery of Grounded Theory. Chicago: Aldine.
Gregoire, R. & Perlman, F. (1969). Worker-Student Action Committees: France May '68. Detroit: Black and Red.
Kellehear, A. (1993). The Unobtrusive Researcher. Allen & Unwin: St Leonards.
Kelly, C. & Breinlinger, S. (1995). Identity and injustice: Exploring women's participation in collective action. Journal of Community and Applied Social Psychology, 5, 41-57.
Klandermans, B. (1992). The social construction of protest and multiorganizational fields. In A.D. Morris & C.M. Mueller (eds.), Frontiers in Social Movement Theory, pp. New Haven: Yale University Press.
Le Bon, G. (1947). The Crowd: A Study of the Popular Mind. London: Ernest Benn. (Originally published 1895.)
McCarthy, J. & Zald, M.N. (1977). Resource mobilization and social movements: A partial theory. American Journal of Sociology, 82, 1212-41.
Milgram, S. & Toch, H. (1969). Collective behavior: Crowds and social movements. In G. Lindzey & E. Aronson (eds.), The Handbook of Social Psychology. Volume 4. (Second edition), pp. Reading, Mass: Addison Wesley.
Ng, S.H. (1980). The Social Psychology of Power. London: Academic Press.
Oakes, P.J., Haslam, S.A. & Turner, J.C. (1994). Stereotyping and Social Reality. Oxford: Blackwell.
Oberschall, A. (1973). Social Conflict and Social Movements. Englewood Cliff, NJ: Prentice Hall.
Parker, I. (1992). Discourse Dynamics: Critical Analysis for Social and Individual Psychology. London: Routledge.
Potter, J. & Wetherell, M. (1987). Discourse and Social Psychology: Beyond Attitudes and Behaviour. London: Sage.
Prentice-Dunn, S. & Rogers, R.W. (1982). Effects of public and private self-awareness on deindividuation and aggression. Journal of Personality and Social Psychology, 39, 104-13.
Prentice-Dunn, S. & Rogers, R.W. (1983). Deindividuation in aggression. In R.G. Green & E. Donnerstein (eds.), Aggression: Theoretical and empirical reviews. Volume 2. Issues in Research, pp. 155-177 New York: Academic Press.
Prentice-Dunn, S. & Rogers, R.W. (1989). Deindividuation and the self-regulation of behavior. In P.B. Paulus (ed.), Psychology of Group Influence. (Second edition.), pp. 87-109 Hillsdale, NJ: Erlbaum.
Reicher, S.D. (1982) The determination of collective behaviour. In H. Tajfel (Ed.) Social Psychology and Intergroup Relations. Cambridge: CUP and Paris: Maison des Sciences de l'Homme.
Reicher, S.D. (1984). Social influence in the crowd: Attitudinal and behavioural effects of de-individuation in conditions of high and low group salience. British Journal of Social Psychology, 23, 341-350.
Reicher, S.D. (1987). Crowd behaviour as social action. In J.C. Turner, M.A. Hogg, P.J. Oakes, S.D. Reicher & M.S. Wetherell, Rediscovering the Social Group: A Self-Categorization Theory, 171-202 Oxford: Blackwell.
Reicher, S.D. (1996). ‘The Battle of Westminster’: Developing the social identity model of crowd behaviour in order to explain the initiation and development of collective conflict. European Journal of Social Psychology, 26, 115-34.
Reicher, S.D. (1997) Collective psychology and the psychology of the self. Newsletter of the Social Section of the BPS, 36, 3-15.
Reicher, S.D. & Levine, M. (1994a) On the consequences of deindividuation manipulations for the strategic communication of self: identifiability and the self-presentation of social identity. European Journal of Social Psychology, 24, 511-24.
Reicher, S.D. & Levine, M. (1994b) Deindividuation, power relations between groups and the expression of social identity 1: the effects of visibility to the outgroup. British Journal of Social Psychology, 33, 145-63.
Reicher, S.D., Levine, M. & Gordijn, E. (in press). More on deindividuation, power relations between groups and the expression of social identity: Three studies on the effects of visibility to the ingroup. British Journal of Social Psychology.
Reicher, S.D. & Potter, J. (1985) Psychological theory as intergroup perspective: a comparative analysis of 'scientific' and 'lay' accounts of crowd events. Human Relations, 38, 167-89.
Reicher, S., Spears, R. & Postmes, T. (1995). A social identity model of deindividuation phenomena. In W. Stroebe & M. Hewstone (eds.), European Review of Social Psychology, 6, 161-98.
Sachdev, I. & Bourhis, R.Y. (1985). Social categorization and power differentials in group relations. European Journal of Social Psychology, 15, 415-34.
Stott, C.J. (1996). The intergroup dynamics of crowd behaviour. Unpublished Ph.D. thesis, University of Exeter.
Tedeschi, J.T. (1974). Introduction and overview. In J.T. Tedeschi (ed.), Perspectives on Social Power, pp. 1-15 Chicago: Aldine Publishing Co.
Thompson, P. (1978). The Voice of the Past: Oral History. Oxford: Oxford University Press.
Turner, J.C., Oakes, P.J., Haslam, S.J. & McGarty, C. (1994). Self and collective: Cognition and social context. Personality and Social Psychology Bulletin, 20, 454-63.
Whyte, W.F. (1984). Learning from the Field. Beverly Hills: Sage.
Wright, S. (1978). Crowds and Riots: A Study in Social Organization. Beverly Hills: Sage.
Zimbardo, P.G. (1970). The human choice: Individuation, reason and order versus de-individuation, impulse and chaos. In W.J. Arnold & D. Levine (eds.), Nebraska Symposium on Motivation 1969. Lincoln: University of Nebraska.
 A full version of the interview schedule is available from the authors.
 The fact that at least some of this food was vegetarian seemed particularly to enrage councillors and indicate the disreputable nature of the protestors. One councillor made specific reference to the fact that the Cornish pasties had no meat in them and therefore were as inauthentic and alien to the area (Cornwall being adjacent to Devon of which Exeter is the county town) as those who threw them.