‘Something That Unites Us All’: Understandings of St. Patrick's Day Parades as Representing the Irish National Group

The present study investigates how attendees at national celebratory crowd events—specifically St. Patrick's Day parades—understand the role of such events in representing and uniting the national community. We conducted semi-structured interviews with people who attended St. Patrick's Day parades in either Dublin or Belfast. In year 1, full-length interviews were conducted before and after the events (N = 17), and in years 1 and 2, shorter interviews were conducted during the events (year 1 N = 170; year 2 N = 142). Interview data were analysed using thematic analysis, allowing the identification of three broad themes. Participants reported that (i) the events extend the boundary of the national group, using participation to define who counts as Irish; (ii) the events strategically represent the nature of the national group, maximising positive images and managing stereotypical representations; and (iii) symbolism serves to unify the group but can also disrupt already fragile unity and so must be managed. Overall, this points to a strategic identity dimension to these crowd events. We discuss the implications of these findings for future research in terms of the role of large-scale celebratory events in the strategic representation of everyday social identities. © 2015 The Authors. Journal of Community & Applied Social Psychology published by John Wiley & Sons Ltd.

2012; Fenton, 2007). In such settings, overt displays of 'hot' nationalism are something we associate with others (Billig, 1995)-for example, those that have a troubled national history or those within our own national group that have questionable national credentials (Stevenson & Muldoon, 2010). Nationalism is often seen as something exclusive and divisive, leading to discrimination against minorities within the nation and conflict between nations (Brubaker, 1999;Croucher, 2003;Fenton, 2007). As such, talking or acting up one's national authenticity can be counterproductive, allowing accusations of bigotry, of subscribing to 'the wrong kind' of nationalism and of displaying an 'over-heated reaction' (Billig, 1995, p. 37).
One interesting exception to the rule against overt displays of national identity pertains to when nations celebrate and commemorate through parades or rituals attended by large crowds-as nearly all countries do. Here, demonstrations of national identity are acceptable and have a 'taken for granted' quality. This acceptability may come from the fact that such events are assumed to unite the many diverse elements of the nation. They are often carefully managed by state organisations to produce unity, and the sociological and psychological literature suggests that they are successful in doing so (Edensor, 2002;Páez, Rimé, Basabe, Wlodarczyk, & Zumeta, 2013, as cited in Páez & Rimé, 2014;Pehrson, Stevenson, Muldoon, & Reicher, 2014). We argue that although collective events involve actively waving flags or painting faces in national colours, they do not fit with Billig's (1995, p. 44) notion of ' "hot" nationalist passion' and instead are seen as inclusive, because they involve national groups coming together to represent their nation to others and to themselves (Conway, 2010). They may also seem harmless because in established democratic nations, they are viewed as special occasions (Kong & Yeoh, 1997;Thompson, 2001). This is not to say that such events are unimportant; rather, we argue that they are a key site for the construction and representation of national identities. In fact, national identity research demonstrates the flexible, strategic nature of national categories (Reicher, Hopkins, & Condor, 1997). Rather than being fixed or objective, national identities are argued to be constructed and reconstructed (Gillis, 1996) such that identity claims can shift across situations (Abell, Condor, & Stevenson, 2006;Wallwork & Dixon, 2004). As such, the representation of national identity at collective events is worthy of study. Here, we consider whether national celebratory crowd events can be understood by participants as unifying for the national group while not being seen as nationalist in the divisive sense. To address this issue, we aim to investigate how participants understand the role of public collective displays of national identity in representing and uniting the national community, across contexts. We do this by examining views of those attending one particular type of event, the St. Patrick's Day parade, in two locations-one where Irish national identity is assumed and another where it is contested.

Background
Early psychological accounts of crowd behaviour suggested that crowd participation led to impulsivity and a loss of identity (Le Bon, 1895Bon, /1947. However, within social anthropology, organised celebratory events have been termed 'social ritual' and have been argued to allow members of an 'imagined community' to come together and be represented (Durkheim, 1915), and more recently, to embody, reflect and sustain broader social relations (Jarman & Bryan, 1998;Partridge, 1977).
Recent social psychological accounts have developed to suggest that rather than losing identity in crowds, people's behaviour shifts from acting in terms of personal identity, to social identity. This behaviour is determined by the person's current position on the interpersonal-62 A. T. O'Donnell et al.
intergroup continuum, which in turn is influenced by social context and the inclusion of self in the relevant group (Tajfel & Turner, 1979;Turner, 1982). Two distinct but related bodies of research within the social identity tradition demonstrate that (i) collective participation in numerous settings can engender a sense of shared identity (e.g. Drury, Cocking, & Reicher, 2009) and (ii) immersion in a crowd of fellow ingroup members allows one to fully express that social identity and to strategically present a certain version of that identity (Reicher, Spears, & Postmes, 1995;Klein, Spears, & Reicher, 2007). So from this general perspective, collective events allow one to feel like an ingroup member and to express this to others.
While research on the emergence of identity at collective events often focuses on crowds coming together to effect social change or to assert identity against another force or threat (e.g. Drury et al., 2009), recent evidence demonstrates that numerous elements of collective experience, not just threat, can impact on identification-such as synchronicity (Novelli & Drury, 2011;Páez & Rimé, 2014). Indeed, the social identity approach has been applied to celebratory events such as festivals (Neville & Reicher, 2011) and an annual Indian pilgrimage event, the Magh Mela (Prayag Magh Mela Research Group, 2007), which have similarities with the type of event we focus on here.
In any case, even events that appear celebratory and harmonious may include subgroups that would view and represent their relevant identity differently to one another. For example, if, as social anthropologists have suggested, ritual events reflect everyday social relations, existing differences are unlikely to be erased in such a context. Nevertheless, at the Magh Mela, pilgrims of different castes pray and bathe alongside one another, which implies inclusion. Therefore, we suggest that ritual events, although allowing the recognition that the ingroup consists of several subgroups (Coleman, 2002), may also provide an opportunity to represent a unified group identity. Indeed, they may offer the chance to strategically present the group in this way, as might be predicted by the social identity model of deindividuation phenomena, discussed earlier (Klein et al., 2007). However, because superordinate identities do not necessarily erase within-group divisions, they can themselves become a source of contestation (Lowe & Muldoon, 2014). This raises the question of how events, which are built on shared identity, deal with identity issues in a way that sustains rather than undermines identity. Indeed, what has not yet been studied is how the events are understood by participants to relate to the groups they purport to represent in a broader sense.

The present paper
The data presented here relate to St. Patrick's Day events in Dublin and Belfast. We argue that St. Patrick's Day is an occasion on which Irish people may feel their (national) identity is displayed or even defined. Dublin is the capital of Republic of Ireland, where Irishness is the majority national identity and as such is the assumed national category. The Dublin parade forms part of the St. Patrick's Festival, which was established in 1995 by the Irish government. It is a major tourist attraction, aiming to showcase a modern and creative Ireland (St. Patrick's Festival's, 2014), including being inclusive and multicultural (Pehrson et al., 2014). Belfast is the provincial capital of Northern Ireland, a region of the UK where national identity is hotly contested, and a minority of the population views Irishness as its first preference national identity (Muldoon, Todd, Trew, Rougier, & McLaughlin, 2007). Belfast City Council has been involved with the St. Patrick's Day parade since 2006 and endeavours to produce an inclusive parade in a traditionally divided society. Here, inclusivity relates to the inclusion of Protestants, a group who often do not see themselves as Irish (Muldoon et al., 2007). During the Troubles (a period of conflict between Protestant and Catholic ethno-religious political groups over the constitutional status of Northern Ireland as part of the UK rather than the Irish Republic), St. Patrick's Day celebrations were confined to Irish nationalist/Catholic areas. Events now take place in the city centre, which is intended as a neutral or 'shared' space, and parade organisers are sensitive to the presence of divisive symbols including national flags, which are discouraged.
In this paper, we examine attendees' understandings of St. Patrick's Day parades. Our research question centres on how participants understand the role of collective displays of national identity in representing and uniting the national community. In order to investigate whether similar processes may be at play, albeit potentially manifested in various ways, we chose to conduct this research across two different contexts. Given that both events are socially sanctioned, often presented as inclusive (St. Patrick's Festival's, 20144) and are not viewed as 'hot' displays of national identity (Billig, 1995), we will address our research question by analysing qualitatively the extent to which both parades are seen to represent and unite the national community, irrespective of context.

Participants and data gathering strategy
Data were collected at the St. Patrick's Day parades in Dublin and Belfast in 2009 and 2010 (four events in total) to ensure adequate data pertaining to each location. We used an interview methodology as we were interested in participants' understandings of the experience of the events in relation to their identity. Short onsite interviews (approximately 5-10 minutes) were conducted with attendees during the parades (year 1 N = 170; year 2 N = 142). These provided the opportunity to talk to a large number of participants and gave us access to participants' thoughts about the events while they were experiencing them. In addition, a small number of in-depth interviews, lasting around 45 minutes, were conducted with parade participants before and after the events (year 1 only; N = 17), providing more detailed accounts of participants' experiences. These participants were recruited from performance groups and community groups, using the parade organisers as gatekeepers. The collection of opportunistic data before, after and during crowd events allows the researcher to gain understanding of the variety of participants' experiences (Reicher, 1996;Stevenson & Abell, 2011). Similarly, our recruitment of both parade participants and attendees allowed us to access the views of those involved in the representation of Irishness and also those observing it.
Because of the busy, crowded environments where we collected most of our data, it was not always practical to obtain demographic information from our participants, although we collected this information where possible (all available information is given about participants who provided the extracts). We have included nationality information about each participant (Irish, Northern Irish or non-Irish 1 ) and gender information for most participants.
We sampled mainly Irish people but also 'new' Irish 2 and visitors. Rather than interviewers raising and defining the issue of Irishness, nationality was often emphasised by participants; so our focus was whether they raised it and how they defined Irishness.
All interviews were semi-structured. Thus, an interview guide provided a loose structure within which to explore the topics of interest, and participants were prompted to expand on relevant and interesting responses. The interview guide comprised the following topics: (i) why St. Patrick's Day is celebrated; (ii) what the parade says about Ireland; (iii) the experience of being in the crowd and the nature of relations between people in the crowd; and (iv) what an ideal St. Patrick's Day parade would be like.

Analytic approach and procedure
A thematic analysis was undertaken. Thematic analysis allows the identification and analysis of patterns or themes in linguistic data. This process may form part of many different types of qualitative analysis (Boyatzis, 1998) but is also a technique in and of itself, providing a purely qualitative account of a dataset. It is extremely flexible and thus facilitated a focus on experiential aspects of participants' accounts at the explicit (semantic) level. It has been used previously to analyse participant accounts of social identity-related experiences (Stevenson, McNamara, & Muldoon, 2014). Our analytic procedure followed Braun and Clarke's (2006) guidelines for thematic analysis.
In year 1, we open coded our data using the ATLAS.ti software (Scientific Software Development GmbH, Berlin, Germany), which enabled coding of utterances relating to similar points. This first level of coding stayed very close to the data, paraphrasing meanings of the utterances and marking any deviant cases relevant to the code. After re-reading the transcripts and reviewing the first level codes, the first and third authors considered how they would be amalgamated by attending to evidence of conceptual commonalities between codes. As a consequence, broader themes were proposed, under which some of the original 50 codes might be combined, but categories were not fixed. Of the original 50 codes, a small number were not included in our final analysis as they did not fit a meaningful pattern. On a final re-read of the amalgamated codes, we wrote detailed outlines of each theme, which became the building blocks for the findings presented here.
This synthesis of codes was guided by our theoretical interests into how identity and Irishness in particular was understood and interpreted by participants, the role of context in driving construction of identity at our two sites, participants' declared national identities and any identity-related talk. Given this broad interest, we did not constrain our analysis to focusing only on inclusive notions of the national group. Further, because our focus was on participants' understandings of how parades might represent and unite the national group regardless of context, we did not systematically compare the two locations but rather oriented to whether similar processes might play out in different ways in each context.

ANALYSIS
Three broad themes were identified in the data. They do not represent a numeric measure of how common a topic was but rather represent part of the pattern of ways that 2 'New' Irish refers to immigrants who have plans to stay in Ireland long term, become citizens, and so on.
Representing the Irish national group 65 participants spoke about identity in relation to the events. Each extract is an example of a different way that identity was discussed by participants.

Extending the boundaries of the national group
Participants made the basic point that St. Patrick's Day was for Irish people. In so doing, they defined-implicitly or explicitly-the boundaries of the national category. The event was presented as appealing to all and the national category as highly inclusive, such that shared participation in the events unified the nation. This was counterposed to a state of division and diversity, which characterises the everyday life of the nation: The emphasis here is on the way the parade reifies an inclusive and united group. Whilst there is an acknowledgement that unity is fragile, the event and the day itself are seen as having power to unify the diversity within the national category.
The usual boundaries of national category inclusion were stretched. Descent, normally a necessary if not sufficient marker of being Irish, was not viewed as a necessary criterion for inclusion (Joyce, Stevenson, & Muldoon, 2013). In the extract in the succeeding text, whilst the entire crowd is presented as Irish, and 'Irish blood' is being celebrated, the interviewee does not go so far as to state that the crowd share Irish blood.

Extract 2: Dublin 2009, onsite interview, Irish male
Interviewer: Do you feel you have much in common with others here in the crowd?
Participant: Yeah, there's an excitement. We're all Irish here you know, just celebrating Irish blood, like, you know.
The St. Patrick's Day parade in Dublin is a huge tourist attraction, so attendees cannot fail to see the many visitors from Europe and beyond in attendance. Suggesting that all attendees were of Irish descent is probably a claim too far; therefore, including all attendees in the national category must include those of Irish and non-Irish blood.
Evidence for the complexity of inclusiveness comes from a participant in Belfast. The participant, who is Protestant-a religious category traditionally associated with Britishness, which is often positioned as oppositional to Irish in Northern Ireland (Muldoon et al., 2007)-presents the event as something that can represent him. Two things are striking here. On one hand, the participant is concerned about whether the nature of the parade gives the right impression of his identity. On the other hand, he is aware that he has different identities with different meanings. Consequently, it makes no sense to ask him if he is represented by an event without asking which of these identities is being represented. However, we see an insistence here that the event transcends communal boundaries. As such, it represents him as someone who lives in Northern Ireland but not as a member of one of the religiously divided communities therein (namely Protestant). In short, even in this troubled identity landscape, this event is seen as something that can act to represent everyone, irrespective of religious tradition.

Strategically representing the national group
This theme encompasses the ways the event is used to represent national identity. Participants were aware of the importance of using the occasion strategically, viewing it as an opportunity to project a positive image of the nation and expressing concern with managing the way the event represented their national group. Although different concerns were evidenced in Belfast and Dublin, in both contexts, a representation of the national category that was singular and positive was believed to yield significant social and political capital, allowing this small nation to 'punch above its weight'. In the succeeding text, when asked what St. Patrick's Day meant to him, this participant made a light-hearted remark about dancing and fun but went on to say:

Extract 4: Dublin 2009, onsite interview, Irish male
Participant: The best thing about St. Patrick's Day is it's a world festival, I mean there's not many leaders of any country that get to spend the day in the White House every year. It's a great opportunity, it's something that we've built as well you know, how much money are we going to make today across the world.
Participant: I mean its good, for a country that's four and a half, five million, we have a brand that no one else has, it's incredible how well known we are.
The parade is seen to represent Irishness to the world, and indeed, it is a rare opportunity because the spotlight of the world is on Ireland. This arguably is seen as important in the Dublin of 2009-2010, which was suffering the ill effects of the global financial crisis.
Participants reported pleasure at the way the events depicted the nature of Irishness. In Belfast, the issue of Irishness has been and remains (to some extent) highly contested. Despite this division and contestation, the event and Irishness are presented as being 'for everybody'. It is remarkable that merely mentioning the possibility that the parade may be for one community (by implication, Catholics more than Protestants) leads to four successive rebuttals involving both participants. They vehemently reject the notion of division or exclusivity. The inclusion of 'all cultures' allows everyone to have ownership and participation in the event, resulting in a shared sense of community and cohesion. The cultures of the divided communities are presented as varieties of an inclusive Irish culture.
Indeed, the awareness that the nation was represented by the event meant that participants became concerned with managing the way the event represented their national group. These strategic concerns were linked to the ambient concerns of the context, North and South. We argue that the most problematic representations are those that may reinforce negative stereotypes of the Irish; thus, these are policed most carefully. So in Dublin, for example, we see concerns about drunk and disorderly behaviour. In the following extract, the participant is content with the representation of Irishness that is communicated by the parade-but less so with those watching it. We're so much more than that.
The participant's claim that the perpetrators were not Irish is in many ways central to the narrative. Contrary to the notion that everyone in the crowd shares Irishness, these crowd members are Australian. From her perspective, they had failed to understand the normative constraints on their behaviour at the event and consequently presented a danger that was entirely representational. The need to avoid consolidation of negative representations is writ large, and indeed, the final sentence could be read as the representational rationale for the parades: 'We're so much more than that'.
In Belfast, representational concerns centre around the possibility that disorder will arise as a consequence of sectarian animosities. The extract in the succeeding text comes just after the participant has spontaneously mentioned that those attending the parade are 'all Interviewer: Now they were saying all the people are here for the one thing, they also say that there are different representations of Ireland here and some they agree with and some they don't agree with, is that a sentiment that you would go along with? Participant: Um, that everyone is out to enjoy the patron Saint's day and shake off all our worries and cares. Absolutely.
The participant speaks in highly coded terms-indicating that there are normative constraints around what should be spoken in this context. Recalling Extract 5, we might infer that the 'serious things', which remain undefined, are the sectarian divides and conflict associated with Belfast.
Taking these several representations together, the important thing is the way in which the parades challenge negative representations of Ireland and replace them with positive images. It is clear that there is strategic social and political value in attuning to the positive, unifying elements of the same national identity and that this can be used to combat negative representations associated with both contexts.

Symbolising the national group
Symbolism, particularly flags, played an important role in communicating shared identity within the group, as in the extract in the succeeding text, which mentions the Irish tricolour and the use of the national colour, green, in the Dublin parade: Extract 11: Dublin 2009, onsite interview, Irish female Interviewer: Do you feel you have much in common with others in this crowd?
Participant: Yeah because we're all here to celebrate the one thing, which would be, you know, St. Patrick and to be Irish. When you look around and you see so many people wearing green and tricolours and all of the rest of it, you know. You know that they're proud to be wearing that sort of stuff.
Here, the common and uncontroversial use of symbols leads to the assumption of shared purpose (everyone is celebrating Irishness) but also shared experience (everyone is feeling Representing the Irish national group 69 the same thing-pride). Indeed, shared symbolic practices are sufficiently powerful to draw non-nationals into the group and allow them to partake of collective unity. When one participant was asked during the Dublin parade 'do you feel included in the day?', he replied, 'Yeah for sure, I am wearing the Irish flag, although I am not Irish and these guys [gestures towards his friends] all wear Irish stuff so I think we are yeah'.
The role of symbolism in relation to unity was much more controversial in Belfast. Here, there seemed the ever-present possibility of division, which could undermine the experience of unity. Because religious and national identity in Northern Ireland are often conflated (Muldoon et al., 2007), Irish nationhood and its symbols-particularly the tricolour flag-become highly political (Bryan & McIntosh, 2005).
Extract 10: Belfast 2009, post-event interview, Northern Irish female Participant: A lot of the groups went all out in the parade ehm to fly, both the tricolours or have themselves wrapped in them, and… I just felt that we were the outsiders in it all, and that, we really didn't belong there.
Displaying the flag serves to make Protestants feel excluded, and here, we see a Protestant member of the crowd reporting this. Issues around flags and symbols remain a political hot potato in Northern Ireland, and here, we see their power to undermine a fragile unity that emerges at this event.

DISCUSSION
The current study investigated how participants understand the role of public collective displays of national identity in representing and uniting the national community. It explored understandings of how Irish national identity is represented at St Patrick's Day parades in both Dublin and Belfast, in order to determine how such understandings are expressed, across two contexts. Our findings evidenced three main ways in which participants spoke about issues of national identity in relation to St. Patrick's Day events. First, the events are seen as having the ability to set the boundaries of the national group at their widest and most inclusive, having a unifying function. Second, the events are understood to offer an opportunity to strategically represent the national group to the wider world. This was evidenced both in terms of the opportunity to present a positive image of the national group and managing potentially negative stereotypical representations. Third, the presence of symbols such as flags plays an important role in communicating the relevant identity to those present, serving to either unify or divide attendees, depending on whether Irish identity is obvious and uncontroversial (as in Dublin) or at odds with the official State identity (as in Belfast).
Participants' understandings of the events as functioning to extend the boundaries of the national group and unite those who ordinarily might not feel united speaks to the existing literatures on understandings of national identity and on the way collective events represent existing identities more generally. That the exact meaning and composition of national identity can be considered to shift is in line with previous research demonstrating that national identities are constructed and reconstructed through interaction and negotiation (Abell et al., 2006;Gillis, 1996;Reicher et al., 1997;Wallwork & Dixon, 2004). The events were seen as an opportunity to unify the national group, while still recognising that 70 A. T. O'Donnell et al.
on an ordinary day, divisions might exist within the group, or indeed, those from outside the national group might not be invited in. Interestingly, it was acceptable to 'claim' Irishness, despite the fact that this has previously been shown to undermine perceived authenticity (Stevenson & Muldoon, 2010;Joyce et al., 2013). As such, our analysis complements earlier findings showing that collective events not only have a unifying function (Páez & Rimé, 2014) but also extend recent important work by social identity researchers (e.g. Drury et al., 2009;Neville & Reicher, 2011;Novelli & Drury, 2011) and from other social science disciplines (e.g. Jarman & Bryan, 1998) by further elaborating our understanding of the relation between everyday understandings of identities, the way they function at 'special' events such as parades and indeed how this might relate to existing social relations and indeed divisions. It may be that celebratory events can be regarded as 'embodied arguments' about identity; domains where the question of 'who we are' is contested and produced. This is clearly an intriguing issue for further research.
Further building on these findings, our analysis suggests that the events are also seen to strategically represent the national group to the wider world. The existing literature describes national collective events as highly specified, rehearsed performances intended to convey a particular image of the nation (Edensor, 2002;Kong & Yeoh, 1997), and official communications illustrate that parade organisers have a particular vision in mind (St. Patrick's Festival's, 2014). It is interesting to note that those attending the parade also consider it to have this strategic function-perhaps fitting with previous claims that those attending this type of event are actively involving themselves in group representation (Conway, 2010). That is, in a country where there is a history of violent nationalism, participants consider that performances must be managed to ensure that the divisive and difficult elements of nationalism do not materialise. The desire to avoid appearing divisive fits with literature suggesting that exclusive or ethnic nationalism is viewed as unpalatable in Western democratic nations (Billig, 1995;Binks & Ferguson, 2012;Brubaker, 1999). Moreover, the aversion to confirming negative national stereotypes is in line with previous findings suggesting that British people find negative national stereotypes embarrassing and dislike the notion of potentially confirming them (Fenton, 2007). The current finding extends this to the strategic representation of the national group in collective events. In demonstrating this empirically, we also build upon theoretical work from the social identity model of deindividuation phenomena perspective suggesting that collective events allow the opportunity for crowd members to strategically represent their group in particular ways (Klein et al., 2007).
Of course, the events were not seen as representing the national group only to those outside. In particular, the use of symbolism was seen as playing an important role in communicating the identity of the national group to those present. Our findings fit with existing theorising, which suggests that national groups provide a sense of community that is dependent on the use of symbols (Croucher, 2003). Indeed, the use of such symbols was experienced as unifying in Dublin, a context where Irish national identity was uncontroversial, even for those not usually included within the national group, which relates back to our first main finding. However, the very same symbolism could also serve to indicate difference and undermine unity in Belfast, a context where national identity is less taken for granted (Muldoon et al., 2007).
Broadly, then, our analysis addressed our research question by evidencing how participants understood these collective national events as offering an opportunity to celebrate the national group in a way that was inclusive and therefore not frowned upon and to strategically present this type of representation to those present and those further afield. Moreover, Representing the Irish national group 71 it demonstrated how the representation of the national group at the events was seen as relating to understandings of national identity more generally. Although these parades are examples of mundane celebration, the potential for national identity to become 'hot' is always present (Billig, 1995). In light of this, each event functions to portray a particular version of national identity, and the appropriateness (or not) of identity performances are noted and policed throughout by those taking part.

Limitations and future directions
Given the restricted nature of the events we have studied, our opportunity sample and mode of analysis, we are in a position to make qualitative statements about the links between parades and identity. The use of onsite interviews allowed us powerful access to participants' accounts of their experiences in the moment. However, these interviews were necessarily shorter than typical research interviews, were sometimes conducted with more than one person at a time and may have allowed participants to be distracted by the crowd, which could have reduced the level of detail in their answers. These data were complemented by the longer more detailed pre-event and post-event interviews with parade participants.
Overall, then, our data are rich and allow us to show the way St. Patrick's Day parades are strategically used to portray a positive and forward-looking image of Irishness-but we cannot extrapolate from St. Patrick's Day to all collective celebrations. Nor can we assume that the incidence of identity-related talk reflects the nature of the event rather than, say, the nature of our interviews (which addressed what the parades say about Irishness) or the nature of our interviewers (some of whom were not indigenous). We are also aware that our participants' accounts of the national group as inclusive are not necessarily reflective of people's more general understandings of the national group. Indeed, the relevance of the collective events themselves to the construction of the national group is vital to our arguments. Nevertheless, we must acknowledge that more exclusive accounts of Irish national identity are of course also possible, even if not represented in our dataset.
The contribution of this work is that it moves the issue of strategic representation of identity up the agenda in studies of crowd psychology. Moreover, if indeed crowd events are viewed as a way to construct a particular version of identity, the next question that arises is 'are they effective'? Do they affect how outgroup audiences view the category? Do they affect how members of the category view themselves-not only actual participants but also the wider population?

Conclusions
These findings show that crowd events are not the exotic and atypical phenomena that occasionally erupt in the midst of our ordinary lives, as many social scientists have assumed. Rather, they are part of everyday social life, and they only stand out in terms of their visibility and hence their ability to tell us who we are and for us, both as individuals and as group members, to tell others what we are like (Edensor, 2002;Partridge, 1977). While they do not erase existing divisions (Lowe & Muldoon, 2014), they allow for the unifying of diverse groups under a common banner. As such, we suggest that those who are interested in issues of identity and social action need to look far more to crowd events as critical elements in the formation, elaboration, presentation and contestation of identity definitions.