The performance of publicness in social media: tracing patterns in tweets after a disaster

Abstract

This article sets out to contribute to the critical understanding of public communication in social media by studying the use of Twitter after a severe earthquake in Aotearoa New Zealand in 2011. It also sets out to contribute to methodologies for studying this particular kind of publicness. It argues that the contours of the ‘social imaginary’ of the public, which are usually so hard to delineate and can be approached only in fragments or typical form, can be identified a little more clearly in the traces that people leave behind in their social media communication at critical, reflexive moments such as in the aftermath of disaster. The article draws on computer-assisted discourse analysis, specifically a corpus-linguistic-informed analysis of half a million tweets, in order to describe four main public discursive moves that were prevalent in this form of public communication. This is not to claim to describe a stable set of norms, but in fact the reverse. The article suggests that empirical, large-scale analysis of public communication in different situations, media and places opens up a project in which the varying norms of public communication are described and critiqued as they emerge in a range of discursive situations.

Keywords

Aotearoa New Zealand computer-assisted discourse analysis disaster communication public publicness

Introduction

This article sets out to contribute to the critical understanding of public communication in social media by studying the use of Twitter after a severe earthquake in Aotearoa New Zealand in 2011. It also sets out to contribute to methodologies for studying this particular kind of publicness. It argues that the contours of the ‘social imaginary’ of the public (Taylor, 2004), which are usually so hard to delineate and can be approached only in fragments or typical form, can be identified a little more clearly in the traces that people leave behind in their social media communication at critical, reflexive moments, such as in the aftermath of disaster. The article suggests that empirical, large-scale analysis of public communication in different situations, different media and different places opens up a project in which the varying norms of public communication are described and critiqued as they emerge in a range of discursive situations.

The public is understood here as a phenomenon of collective understanding, an idea that people in a society can use to make sense of their social existence. It is not a group or a thing; rather, its reification into ‘public opinion’ or the ‘public interest’ never adequately encompasses the idea. In Laclau’s (1996) terms, the public can be something of an empty signifier, a term subject to different claims – contested, colonised and therefore always in crisis – with a limited degree of stable, base meaning. On one level, this makes the public difficult to study empirically because it is hard to know what to measure and what expectations of the public to measure the phenomenon against. On another level, this way of thinking opens up the possibility of studying how people do public communication as they negotiate social norms in interaction with limited critical preconceptions about those norms. In this article, public communication will be defined as communication taking place in publicly available media and in relation to matters of collective importance. There is a risk that a definition of this kind becomes too wide to be of value, but the four widespread and distinct discursive moves identified in the analysis below suggest that specific communicative phenomena have been identified.

A fairly open approach to the idea of the public is particularly valuable in studying the massive and rapid shifts in the ‘architecture of political communication’ (Marcinkowski, 2008) that are taking place in many contexts. In 15 years, if the birth dates of instant messaging platforms such as QQ (1999), Twitter (2006) and the public version of Facebook (2006) can be used as milestones, the production of public talk has suddenly come within the grasp of nearly every cell phone and Internet user, the boundaries between public and interpersonal communication have weakened and the identity of the institutions shaping that communication has changed. The critical apparatus for conceptualising public communication has also come under pressure, with analysts reaching for new terms such as ‘citizen journalism’ (Allan and Thorsen, 2009) and ‘produsage’ (Bruns, 2007). These often still depend on philosophical assumptions about how people come together in publicness. In the west, classic liberalism has given us the spatial notion of the public as comprising individuals gathering in public spaces or in the virtual spaces of public media to discuss matters of common concern. This deliberative democratic ideal was imaginable in early industrial, patriarchal Western society, where the numbers of individuals in any one society were smaller than today and where a few institutions, such as the legislature, universities and the Church, were able to embody the idea of bringing together the public or embodying the public interest. Hampton (2010) describes, for example, an ‘educational’ ideal of the press in 19th-century British society, in which the press was analogous to parliament itself:

In other words, the same debates about public affairs that took place within Parliament also took place among the wider political nation in the pages of daily newspapers and serious periodicals, enabling what might be called a ‘politics by public discussion’. (p. 4)

As Hampton notes, this imaginary was less easily sustainable as the franchise, the state and mass media expanded. Dominant discourse of the public shifted to a ‘representative’ model, in which the idea of the public was embodied by institutions that represented it and were answerable to the individual members of the public in various ways. Yet, that version has lost some of its authority too, at least in Western democracies, for a wide range of mutually entangled reasons beyond the scope of this article (see, for example, Castells, 2000). Markers of traditional participation in public life, such as voting and consuming news, have declined. At the same time, public communication has become subsumed within privatised spaces, such as Facebook. As scholars in this field note, it is a mistake, though, to conclude simplistically that the public is in crisis. There is more access to means of public communication than ever before, as cases such as the enormous social media response to earthquake discussed below illustrate. The problem is partly one of how public communication and participation are theorised, measured and evaluated. At the broadest level, this article’s contribution is to that question.

Performative publics

The public is a rich object, embedded in history, structures of power and people’s experience of it. In communication and media studies, it is, though, too rarely studied as a phenomenon in itself, perhaps because it tends to form the grounds of analysis. The ideal public is assumed to be rational, civil and universally available within a society, with the media then evaluated against those criteria. So, Splichal (2006) notes that the invocation of an Enlightenment ideal of publicness always has a ‘critical sting’ (p. 696), aiming to critique structures of power and sustain human freedom. This normative basis to research on publicness is important, but runs the risk of grounding research in particular political problems and historical trajectories. A key research problem is that the political culture of a few dominant sites of publicness, whether the European project, the United States or the United Kingdom, produced these norms. Lowe and Nissen (2011) point out that these places are in fact atypical of public life elsewhere, providing a poor model for analysis or policy in smaller, less wealthy, non-Western or post-colonial contexts. For example, the political culture in Aotearoa New Zealand (population: 4.6 million people) is shaped by relatively low social distance between people, where political leaders may literally share the same buses with citizens and where norms of egalitarianism are deeply embedded. Fischer (2012; cited in Wendel, 2012) argues that, in contrast to the United States, fairness is privileged over freedom. The idea of the public therefore needs to be ‘provincialised’ (Barton Scott and Ingram, 2015), ‘approaching it less as a normative model for modern society than as a culturally peculiar notion caught up with the particular history of the North Atlantic region (i.e. “the West”)’ (p. 358).

Similarly, there may be value in conceptualising the public less as an entity or a space than as a performance, as people acting out public cultural practices. Heikkilä and Kunelius (2006) note that describing the public as a ‘sphere’, as if it were a space, can divert researchers from thinking about the action that happens there. Seeing the public as action allows for the complex meanings of the public for people, as well as critique of the way that it is dominated or organised by certain interests, within an awareness of the technologies that shape public interaction. The particular ways that these come together in any one context can also be explored. In Calhoun’s (1992) terms, the public is a field of discursive actions. This conceptualisation of the public, therefore, depends upon understanding the way that its cultural parameters are reproduced and oriented to in meaningful action, and these may be complex and contradictory.

A cultural and performative view of publicness gives us an unstable and not always definable object. Taylor (2004) assumes that the public sphere is metatopical, that is, it does not depend for its existence on the topic of today’s debate but will be there in similar form tomorrow. But imagining it as something that people do allows the ebb and flow of publicness to be captured a little better. In Dewey’s (1946) terms – while acknowledging he was thinking of a particular early 20th-century context – the public may be thought of as inchoate and loosely organised (p. 131). Dewey called for people to become reconnected with politics through a system where there was access to rich understanding of the social world through education and news media and where community-level communication allowed common concerns to be talked through, building the public out of community. This can be seen as a constant process of re-forming and renegotiating norms, rather than in such deficit terms. Contemporary social media provide examples of publics that remain loose coalitions of interests, rising to accommodate needs and disappearing again.

Methodologies for studying publics

The huge datasets that digitised social activity produces provide an opportunity to study the public in these performative terms. The old texts of the public – official and institutional texts such as government statements, textbooks, news media reports and other forms that represent the public – lent themselves to assumptions of the public as a collection of people and by metonymy the spaces where they could gather. Digital archives of large numbers of public interactions are, by contrast, well suited to test out and refine theories of the public as something that differs according to time and place, that takes form in communication and that need not be imagined as a stable entity.

Of particular interest here is the form that public communication takes. Early research on social media made large claims about its potential to step outside existing constraints on public debate. In particular, researchers sought to establish that blogging, social media and other interactive practices allowed public debate to become more deliberative, available to a much wider range of people, freed from the problems of the commodified debate to be found in mass media and freed from the power of an elite to control and suppress (e.g. Gillmor, 2006). Quite quickly, those claims were found to match actual activity only poorly. Most political blogging, for example, draws upon and interacts with professional news media (see Haas, 2005). Still more problematically, most people’s public activities online are not deliberations on public themes but mix the personal and the public, the rational and the emotional in ways that make it hard to pin down. What is much clearer is that online public debate operates in ways that are not easily accounted for by classic theories of public debate.

One major finding of research on participation in social media publicness has been that participation is highly individualised. Schmidt (2014), for example, argues that discussion on Twitter operates according to what he calls ‘personal publics’. When someone posts about something of significance beyond their personal life, they take part in the public in ways that are often referenced according to their own selves. He sets out three aspects that are partly related to structural aspects of Twitter and partly to convention. First, information is selected and displayed according to criteria of personal relevance. Second, unlike in mass media forms where the audience is unknown and weakly bounded, it is addressed to an audience of network ties made explicit. A person posting on Twitter will know at least some of those who read his or her tweet because he or she can track followers and because the responses of some of them will appear underneath the tweet as comments. If a frequent Twitter user, he or she will orient to that known audience when tweeting. Third, the communication is generally in a conversational mode, operating at the level of personal opinion, personal relationships and, to an extent, the common stock of ideas that arise in conversational talk. Twitter therefore enables a form of networked individualism, where people operate ‘more as connected individuals and less as embedded group members’ (Rainie and Wellman, 2012: 12).

Bruns and Burgess’ (2011) research presents a slightly different picture. Their research finds that Twitter users frequently step into a different mode when they need to move beyond their immediate social circle, using hashtags. Hashtags allow anyone, not just a Twitter user’s followers, to read the tweets, and at times of major events, such as disastrous floods in Queensland in 2011, a few hashtags quickly become established as ways for large numbers of people to communicate efficiently about the event. In these situations, the notion of relevance of material drifts towards the individual’s sense of what is relevant to a much wider group than the self or the personal network; that network becomes relatively unbounded and the modes of communication combine the personal; a community-oriented voice; formal public announcement; and other forms. In relation to all three aspects of Schmidt’s analysis, it is hard to see this communication in purely personal terms. Bruns (2008) argues,

What we see emerging … is not simply a fragmented society composed of isolated individuals, but instead a patchwork of overlapping public spheres centred around specific themes and communities which through their overlap nonetheless form a network of issue publics that is able to act as an effective substitute for the conventional, universal public sphere of the mass media age. (p. 69)

Bruns’ work suggests a form of publicness that is emergent, performative and dependent both on the media being used and on the conventions that emerge around those media. It is, importantly, not an aggregate of individual views but a collective imaginary.

Much of Bruns’ work cited above has focused on moments of disaster to study Twitter’s public uses. As McAfee (2008) notes, individuals may become particularly aware at moments of crisis that their experience is not only in public but constitutes a public act. When people have recognised each other as part of the same public, they begin ‘to identify themselves as a public created in this moment of recognition’ (p. 116). The public does not exist only in crisis, but such moments provide a high degree of reflexivity (i.e. when people are conscious that they are participating in mediated events), a scale to the public communication and a recognisable connection of publicness to collective social action. Publicness is brought to the surface in ways that are methodologically valuable for studying how it is performed.

This study combines quantitative and qualitative analysis of tweets, using a corpus linguistic concordancing tool, Antconc. Publicness, as defined above, is a matter of how people interact and orient towards shared norms, something that a qualitative methodology is most suitable to track, particularly a discourse analysis that analyses how people perform certain norms of public interaction such as inclusiveness. Yet, qualitative studies can only study small amounts of data and can make only limited claims about the patterns of public communication that people themselves experience through their social participation. The large datasets of social media communication now available provide an opportunity to process wide streams of material so as to identify significant patterns. This project combines the two approaches, cycling from the big picture of overall patterns to close analysis. As Koller and Mautner (2004) and others have argued, a concordancer ‘effectively helps to break down the quantitative/qualitative distinction, providing as it does the basis for quantitative analysis without “deverbalizing” the data’ (p. 225). Concordancers enable lexis and phrasing to be studied both as frequencies and in the context of the other textual items around them. These recurrent patterns of phrasing can be studied as shared communicative acts and their roles in building social institutions traced (Stubbs, 2010). In this study, the following analytical steps were taken. First, after dividing the data into chunks of roughly equal size according to date, word lists were generated for each chunk to identify the main topics of discussion. Second, more sophisticated keyword lists were generated by comparing these word lists with a frequency list of ordinary written English (the million-word Brown corpus of ordinary American English was used) so as to identify words that were, on statistical measures, of particular salience in the data (Baker et al., 2006: 97). Tweets containing common words and keywords were then closely read to identify how the language was used. Third, n-grams of common words and keywords were identified manually, that is, phrasal units that these words commonly built into, with close reading then done of the full tweets that they occurred in. Finally, and following on from the third step, frequently retweeted or replied-to tweets that appeared in the n-gram analysis were analysed closely. This analysis was done in an iterative and exploratory way, and no claim is made that all the important patterns have been found. It was assumed that these patterns would be meaningful public participation and would allow some norms of usage that shaped the public communication on Twitter to be uncovered. As much as possible, the discourse was studied in its context of meaning, as part of the flow of tweets that users would have seen if they had followed people tweeting about the quake or looked up the #eqnz and other hashtags.

A collection of tweets from the 2011 Christchurch earthquake was studied. On 22 February 2011, the city of Christchurch (350,000 people) was hit by a large and shallow earthquake, killing 185 people, leaving the city centre uninhabitable and damaging about 100,000 homes. The event fits the criteria noted above, a large-scale event that public institutions struggled to deal with and where a wide set of people were, therefore, consciously performing public roles. There was indeed considerable agency shown in local social media use, particularly tweeting, for example, widespread linking to websites where displaced people could find accommodation with residents in other cities or find out basic information, or linking up students who had time on their hands as their study had been interrupted into a ‘Student Volunteer Army’ which helped clean up the city (see Matheson, 2014). Twitter became understandable in the aftermath of this disaster as a mode for people to get things done collectively. The dataset used comprised nearly a million tweets, which either contained hashtags or keywords related to this and other earthquakes in the region in the year before and after. Of those, 420,000 were produced in the 18 days after the major earthquake, from 22 February to 11 March 2011, making up an immediate post-disaster dataset of 7.6 million words.¹ About 112,000 tweets within this subset were explicitly public through their use of the common hashtag #eqnz, 181,000 had http links and 189,000 were retweets, allowing some confidence that the set contained much communication that was public in character. The analysis left to one side the many news reports whose headlines were posted on Twitter, as they operated by a particular understanding of the public that was not so relevant to this article’s interest and because many of these tweets were automated postings of news headlines from news organisations around the world.

There are methodological problems with this approach. The ‘Twitterverse’ is not spatially bounded, meaning that without a reliable network analysis of what public communication different Twitter users had access to, there is some risk that the dataset gathered many disparate social networks of Twitter users into a mass. This leads to two problems. First, the analysis is of a constructed phenomenon to an extent. No one could actually experience the set of 420,000 tweets analysed here, and all users’ experiences of the tweets around the earthquake would have been a little different. Combining corpus analysis with network analysis is the next stage of the project. The current analysis follows the affordance of Twitter in allowing the horizons of public communication to widen and narrow enormously in unstable ways over the course of the period studied. In that, it is perhaps successful at charting Twitter as experienced by individuals. However, that does make it harder to be confident that a particular tweet was meaningfully performed in front of others as public communication. The article cannot claim to describe what New Zealand Twitter users would have read about the quake. For that reason, the analysis below restricts itself to exploring broader patterns of language use, specifically the discursive resources for public communication that people drew on at this moment, rather than particular topics or speakers.

Analysis

The analysis below is divided into four parts. The first finds a specific form of the self-organising often attributed to networked communication media, yet one which did not operate as an alternative public infrastructure. The second discusses by far the largest pattern in the data, a vast global response made up of expressions of solidarity in the first few days post-quake, which constituted an individualised network of care for unknown others. The third part discusses negotiated statements, a more tentative kind of communication in which individuals and organisational voices shared knowledge among themselves and offered help to a public made up of more knowable others. The fourth analyses a less linguistically coherent collection of expressions of shared experience, through which a specific local public emerged.

Twitter as self-organising infrastructure

The public communication observed in the data points in a slightly different direction from two of the central threads of the growing post-disaster Twitter literature, one of which focuses on the reliability of information produced via Twitter and the other on how social media works as a self-organising system that takes over certain government and fourth estate roles. Klein et al. (2012), for example, found two major roles for Twitter during disaster and risk events: ‘the early diffusion of emergency information and the potential to organize mutual help within neighborhoods’ (p. 462). Similarly, Acar and Muraki (2011) found that Twitter functions to enable information seeking and providing. In both of these roles, communication on Twitter is imagined as if it were a public institution – as an alternative media form, as a disaster response mechanism. Such a functional analysis applies weakly to the Christchurch dataset, particularly when close attention is paid to the way people carried out informational activities. The data provided little evidence of people using Twitter as a public forum, traditionally defined. Little of the tweeting deployed language that spoke on behalf of the city or its people. Below is an exception, where the tweeter took on the kind of authoritative public voices that a leader or professional might:

CHCH does not need advice on how to rebuild, it needs time to clear up and wait to see if this fault is permanent or not #eqnz

Rare tweets of this kind were, however, still more rarely retweeted by others, tending to be met by silence. These tweets presumably did not ‘resonate’ (as Zhou et al., 2010 put it) with local Twitter users’ ideas of public talk on Twitter at this moment. The platform was not predominantly a platform for the clash of individual views here. A collective rationality was more in evidence than an individual rationality, perhaps reflecting collectivist aspects of New Zealand’s political culture (Wendel, 2012).

There was more evidence of Twitter being used by people to organise horizontally, bypassing or augmenting some of the usual hierarchies and institutions. This conception of Twitter was prominent in one widely discussed anecdote. On the day of the earthquake, an older couple, who were trapped in a hotel tower (room 2302 of the Grand Chancellor) and unable to telephone authorities via the 111 emergency number, instead called their son in India, who sent requests out on Twitter. A New Zealander Rob Thompson called 111 in another city, Auckland, and then tweeted about it, and his tweet was picked up by various people, including a television reporter who was with rescuers in Christchurch. The authorities then sent rescuers back into the hotel (Bussmann, 2011). That story, while most likely unparalleled, is retold because it feeds an understanding of Twitter’s value in which individuals reach out to other individuals across the network, building on similar stories from previous crises. And indeed people in Christchurch made sense of Twitter in this way, using it to search for missing individuals, to alert people of tools to find others or to find accommodation and above all to pass on information. One widely shared tweet, for example, read,

RT @[user]: Bob Parker: Schools closed for the week, as are workplaces. Stay at your homes, or help in your communities. Don’t drive …

A number of users with larger followings, such as celebrities and technology innovators, took on such a role of retweeting public announcements. As already noted, nearly half of all tweets in the sample are retweets, and a similar proportion contains Internet links; a major use of these functionalities of Twitter was to pass on information. Among the more common keywords in the first 3 days are practical words about phones, airports and Google People Finder.

Close analysis of the data, however, shows that there was little apparatus of formal public talk or metatalk about Twitter, which might be expected when individuals are usurping public functions that typically belong to institutions. The talk seemed to be a practical performance of helping and informing, shorn of any politicisation or self-consciousness or the sometimes strident language of blogging. This is partly to do with the medium: Twitter’s syntax lacks hierarchy and favours terseness. There is therefore less scope than in other media for framing and overt self-presentation. But there was also a casual informality in some of the tweets that suggested an orientation to modes where the interpersonal, the understated and the pragmatic carried their own kind of commonsensical authority. For example, a common element in tweets seeking people who were missing was the informal phrase, ‘Anyone know …?’ The degree to which people did not seek to take on authoritative forms of talking has its own politics, grounded in what Palmer (2008: 278) calls a political culture of egalitarianism and pragmatism.² Thus, the structural flattening of social media works not so much to take back power as to provide a platform in which that political culture can be asserted in a performance of pragmatic collaboration. The self-organising of Twitter appears inflected here within a particular local set of norms of public life.

Individuals expressing concern in public

The single most common textual form comprised expressions of care and offers of help from individuals from around the world. In the first 2 days after the earthquake, a large number of individuals expressed their solidarity with victims in this way − 5% of all tweets on the first day included the word ‘thoughts’, often in the formulaic phrase, ‘our thoughts and prayers’. Some of these tweets came from celebrities and sports stars, such as the Formula One driver Jenson Button or the actor Alyssa Milano, and such tweets were often retweeted hundreds of times. These tweets tended to invoke a public of ‘us all’ and a rather generalised object of care of Christchurch and its people, usually using the abstract term, ‘people’, as in ‘the people affected’. This phase of response was quickly joined by a huge wave of formulaic calls to donate funds, which were again widely retweeted, and tweets aimed at celebrities asking them to retweet calls for donations:

RT @[user] Please can you RT? Christchurch NZ needs your help! Please help quake victims http://bit.ly/hpufB or …

In these ways, an instant public was widely imagined by Twitter users, which was both globalised and abstract and which operated on humanitarian, emotional and popular cultural terms. This kind of public discourse was almost universally unidirectional: people did not reply to these solidaristic tweets. In Chouliaraki’s (2014) terms, there is a risk of this discourse becoming self-referential, a performance of ‘ironic solidarity’ in front of people’s social networks, which was dissociated from the people being talked about and which operated on a consumerist, lifestyle logic. It is hard to read how self-referential the ritual of ‘thoughts and prayers’ was, but certainly it constituted a set of statements made in the presence of a wide public rather than talk between actual people. Indeed, when individuals who had more personal relationships with the earthquake victims expressed solidarity, they sent their thoughts and prayers to ‘family’ and ‘friends’. The norm underpinning the more general call to people could be termed, following Scannell’s (2000) nomenclature, a ‘for-everyone structure’, in that it encompasses everyone relevant to its address. Scannell described mass production as tending to produce ‘for-anyone’ structures of consumption, for example, toasters provide a service for anyone caring to use them. British broadcasting, by contrast, produced ‘for-anyone-as-someone’ communicative structures, addressed to anyone who wanted to be personally addressed, thereby invoking a ‘public, shared and sociable world-in-common between human beings’ (Scannell, 2000: 12). Twitter here produces something else, in which people are not differentiated as individuals and yet directly addressed. An explicit example would be,

Tell everyone to change their voicemail to let people know they are okay, if their battery is running low! #eqnz / Great idea! #chch

The structure operates in a similar way to how Durham Peters (1999) describes broadcasting, as an act of imagination and faith in an unheard and unseen public, and yet it is an individualistic act of care as well. There is a risk of the communication collapsing, as Chouliaraki’s (2014) critique notes, for the performance inevitably contains more of the speaker, constructing the public in her or his own self-conception, as Schmidt (2014) also suggests in his analysis of ‘personal publics’.

Negotiated statements

These tweets dwindled very rapidly after the initial news of the disaster. By day 3 of the Twitter sample, and particularly among those in Aotearoa New Zealand, a common way of talking emerged which are here called negotiated statements. These were tweets where users’ address to others was more complex, anticipating some interaction. They did not assume the tweet’s relevance or importance, presenting a less self-assured tweeter; they did not assume a universal addressee; and they used less formal language. An emblematic syntax among these would be the conditional ‘If anyone’ or ‘If you’, as in

If you’re in Chch and you have no toilet or portaloo I now have 300 chemical toilets in my garage #eqnz drop me a tweet

It is noteworthy that the tweet does not simply inform by telling people, ‘I have 300 toilets’, or call on people to act, ‘Contact me to pick up a portable toilet’, but offers information conditional on the listeners’ interest. Tweets in this category often linguistically marked participation in a group of those affected, and thus were insider messages, affiliating to the group, rather than claiming to speak on behalf of the public or addressing an aggregate group. For example,

Awesome, the combined effort of everyone has just tipped the NZ Red Cross 2011 Earthquake Appeal over the $1 million mark. #EQNZ :-) ^ME

which was written by a public relations account belonging to a large business, chose phrasing, ‘the combined effort of everyone’, that assumed shared membership of a collective act, as well as signalling that the group had an informal national identity through the common New Zealand colloquialism, ‘Awesome’. The tweets were widely retweeted within the country.

They can be characterised as forming a ‘for-anyone-as-someone’ structure (Scannell, 2000) for they recognise the realness and specificity of the individuals listening. In fact, these tweets do that more strongly than broadcasting can, for they recognise the unseen people’s specificity through the negotiated quality of the claim to be relevant to them. They offer a relationship rather than assume one. In Benhabib’s (1992) terms, people are acknowledged as ‘concrete others’ while being indefinite still. This kind of address makes sense in a small country, where people are likely to know some of their addressees or are likely to be later held accountable by people they know. But it is also culturally appropriate in the country. In Aotearoa New Zealand, it can be seen as arrogant or bossy for an individual to take on too much public authority or to impose on others, symptoms for Belich and Wevers (2008) of a ‘knocking culture’ and low self-confidence. The claim to knowledge or the emotional intensity of the message is often negotiated down. For example, language such as ‘wow’ and ‘awesome’ was widely taken up by official and corporate voices later, which recognised it as a culturally appropriate way of doing publicness on Twitter because it took claims down register and expressed polite thanks, forms that are more comfortable in a public made up of specifiable others:

Wow! We’ve just hit $10.3m for the people of Christchurch! Thank you to every person, family and business for your incredible support #eqnz

Shared experiences

Some of the tweets that were shared most widely in the post-crisis period referenced shared experiences. These tweets were not so strongly coherent linguistically, other than a few humorous memes that built on a set phrase, ‘You know you’re from Christchurch when …’ They tended to cohere more through their requirement that the reader bring specific local knowledge to bear. This included deixis (such as ‘that’ to refer to one of the many aftershocks):

That quake was not much fun. Back under the table. #eqnz

There were also more subtle manifestations, including humour, that carried a claim to capture the mood of people in the city or referring to individuals such as the city’s mayor by name and without explanation. They acted as a collective self-representation by people in the city. They also acted out a sense of togetherness, according to dominant cultural tropes, including self-deprecating humour, understatement and casualness (Belich and Wevers, 2008). A widely observed phenomenon after the disaster was a rapid increase in residents’ sense of community, as neighbours helped each other or provided services to others for free (TEPHRA, 2012). This was clearly replicated in Twitter, but in doing so it produced a public discourse that tended to favour a particular version of community. The shared references brought people to the same level by privileging the experience and not differentiating between politicians, celebrities, corporates and citizens. ‘Please RT when you can’, a New Zealand sportsperson tweeted to the Prime Minister about a fundraising auction of his own sportsgear. The city’s mayor quickly became a symbol of the city’s resilience, but in a way that celebrated his capacity to articulate a shared experience. He was mentioned on Twitter often in celebration of this capacity:

Uh oh. :-O RT @ChristchurchCC – #Christchurch Mayor Bob Parker demonstrates the use of the chemical toilets. http://ow.ly/48ZFS #eqnz #chch

As the text above also illustrates, these texts were often pleasurable, celebrating survival in a subversive kind of public togetherness that also tended to level social differences and emphasise the shared experience. Wagner (2011) notes a similar response in post-disaster Haitian popular culture:

Joking allows one to assert one’s humanity in what would seem to be impossibly dehumanizing conditions – of saying that despite everything, the speaker is still here, still a person, and still telling a story rather than being dissolved and absorbed into the story.

Twitter allowed individuals to assert their survival and also for a pleasurable collectivity to be expressed.

Final thoughts

So what social imaginary of the public emerges in people’s communication in this post-disaster moment? As we have come to expect on social media, the public was not performed primarily by leaders or through publishing factual information or through rational debate. These traditional markers of a mass-mediated public sphere do not match well onto the ways that individuals use Twitter. Instead, there were multiple overlapping ways of being public, which tended to be both individualised and oriented towards mutual regard for other individuals, but which differed widely in their global or local focus. These manifestations of publicness could be regarded as inadequate or confused, were they to be judged according to traditional norms. This article has instead suggested that they be read as the way publicness is done in the weakly institutionalised space of Twitter and as governed also by local cultural norms of pragmatism, humour, informality and low social distance. They should also be read as connected to the particular moment of the disaster. At a later moment in the disaster response, it is likely that there was more political discussion and less mutual regard, as consensus gave way to frustration and different visions of the city (see Hayward, 2012). At other times, Aotearoa New Zealand public discourse in online fora such as news and blog comment sections appears dominated by the same aggressive culture as is found in those fora in other countries (Hopkins, 2009). The analysis also shows that the Twitter public was not autonomous. Assumptions that Twitter is used by people as a self-organising system after disaster were only borne out to a small degree. Similarly, the hashtag #eqnz was an important connector, but it was only part of the logic of this public. The four forms of public discourse described in the analysis should be understood then as multiply determined, by norms of publicness, by Twitter’s affordance, by the disaster and by the local political culture.

Although there was rich evidence of people orienting towards others and expressing care, the data cannot be read as tending towards a collective response or as a unified communicative structure of any kind. Instead, what emerges is an unstable, fragmented and multiple public. Because of their multiple determinants, the tactics of publicness here were not a collective act, not a coherent set and not specific to Twitter, but a range of ways of negotiating different latent ideas of how public talk works. It is a mistake to reduce this diversity to an idea of a self-organising system or a personal public or to any other attempt to unify the multiplicity of the public communication. Some of the communication here was on a global scale, some on a local; some individualised and some collective. The data make a case for resisting the assumption that the performance of publicness and the social imaginary of shared public life have a centre point of relative stability. This article suggests instead that we conceptualise an often inchoate public that coheres as a set of tactics.

The questioning of assumptions about how public communication works is enabled by computer-assisted analysis of a large dataset such as this and particularly by the combination of quantitative analysis of large quantities of text (allowing frequencies and distinctive usages of language in this dataset to be identified) and computer-assisted close analysis of instances through concordancing. Some of the complexities and what McAfee (2008) calls the mystery of how people come together in public life in contemporary societies become visible. Interpretation of those acts depends upon an appreciation of the norms of public life as they are instantiated and necessarily remains an interpretation. For publicness exists at the level of people’s imaginations, motivations and interactions in particular situations, qualities which do not reduce easily to objective measures. But the analysis here was moved beyond evaluating how far norms of public communication are followed or breached to tracing how norms are lived. This kind of work opens up a wider empirical challenge. If publicness, particularly social media publicness, is as malleable as described here, then scholars need to chart the digital traces of ways of doing publicness in a range of settings and cultures, provincialising it (Barton Scott and Ingram, 2015) and observing its lived forms. Methodologies for studying public communication need to be further refined, but the article sets out the value of combining analysis on a large scale that captures a range of forms and close analysis of how those forms are made meaningful.

Footnotes

Funding

The author(s) received no financial support for the research, authorship and/or publication of this article.

Notes

References

Acar

Muraki

(2011) Twitter for crisis communication: lessons learned from Japan’s tsunami disaster. International Journal of Web Based Communities 7(3): 392–402.

Allan

Thorsen

(2009) Citizen Journalism: Global Perspectives. New York: Peter Lang.

Baker

Hardie

McEnery

(2006) A Glossary of Corpus Linguistics. Edinburgh: Edinburgh University Press.

Barton Scott

Ingram

(2015) What is a public? Notes from South Asia. South Asia: Journal of South Asian Studies 38(3): 357–370.

Belich

Wevers

(2008) Understanding New Zealand Cultural Identities. Wellington, New Zealand: Stout Research Centre for New Zealand Studies, Victoria University of Wellington.

Benhabib

(1992) Situating the Self: Gender, Community, and Postmodernism in Contemporary Ethics. New York: Routledge.

Bruns

(2007) Produsage. In: Shneiderman

(ed.) Proceedings of the 6th ACM SIGCHI Conference on Creativity & Cognition. New York: ACM, pp. 99–106.

Bruns

(2008) Life beyond the public sphere: towards a networked model for political deliberation. Information Polity 13(1–2): 65–79.

Bruns

Burgess

(2011) The use of Twitter hashtags in the formation of ad hoc publics. In: Proceedings of the 6th European consortium for political research conference, Reykjavik, 25–27 August.

10.

Bussmann

(2011) A Twitter Year: 365 Days in 140 Characters. London: Bloomsbury.

11.

Calhoun

(1992) Introduction: Habermas and the public sphere. In: Calhoun

(ed.) Habermas and the Public Sphere. Cambridge, MA: The MIT Press, pp. 1–50.

12.

Castells

(2000) Toward a sociology of the network society. Contemporary Sociology 29(5): 693–699.

13.

Chouliaraki

(2014) The Ironic Spectator: Solidarity in the Age of Post-humanitarianism. Cambridge: John Wiley & Sons.

14.

Dewey

(1946) The Public and Its Problems. New York: Henry Holt and Company.

15.

Durham Peters

(1999) Speaking into the Air: A History of the Idea of Communication. Chicago, IL: The University of Chicago Press.

16.

Fischer

(2012) Fairness and Freedom: A History of Two Open Societies, New Zealand and the United States. Oxford: Oxford University Press.

17.

Gillmor

(2006) We the Media: Grassroots Journalism by the People, for the People. Sebastopol, CA: O’Reilly Media.

18.

Haas

(2005) From ‘public journalism’ to the ‘public’s journalism’? Rhetoric and reality in the discourse on weblogs. Journalism Studies 6: 387–396.

19.

Hampton

(2010) The fourth estate ideal in journalism history. In: Allan

(ed.) The Routledge Companion to News and Journalism. Abingdon: Routledge, pp. 3–12.

20.

Hayward

(2012) Canterbury’s political quake. The Press, 30 March, p. 7.

21.

Heikkilä

Kunelius

(2006) Journalists imagining the European public sphere: professional discourses about the EU news practices in ten countries. Javnost: The Public 13(4): 63–80.

22.

Hopkins

(2009) Blogs, political discussion and the 2005 New Zealand general election. PhD Thesis, Massey University, Wellington, New Zealand.

23.

Klein

Laiseca

Casado-Mansilla

et al . (2012) Detection and extracting of emergency knowledge from Twitter streams. In: Proceedings of the ubiquitous computing and ambient intelligence: the 6th international conference (UCAmI 2012) (ed Bravo

Lopez-de-Ipiña

Moya

), Vitoria-Gasteiz, 3–5 December. Berlin: Springer.

24.

Koller

Mautner

(2004) Computer applications in critical discourse analysis. In: Coffin

Hewings

O’Halloran

(eds) Applying English Grammar: Functional and Corpus Approaches. London: Hodder & Stoughton, pp. 216–228.

25.

Laclau

(1996) Emancipation(s). London: Verso Books.

26.

Lowe

Nissen

(2011) Preface. In: Lowe

Nissen

(eds) Small among Giants: Television Broadcasting in Smaller Countries. Göteborg: Nordicom, pp. 7–20.

27.

McAfee

(2008) Democracy and the Political Unconscious. New York: Columbia University Press.

28.

Marcinkowski

(2008) Public sphere, fragmentation of. In: Donsbach

(ed.) International Encyclopedia of Communication. Blackwell Reference Online. Available at: http://www.communicationencyclopedia.com (Accessed 1 May 2016).

29.

Matheson

(2014) Tools in their pockets: How personal media were used during the Christchurch earthquakes. In: Thorsen

Allan

(eds) Citizen Journalism: Global Perspectives, Vol 2. New York: Peter Lang, pp. 99-112.

30.

Palmer

MSR

(2008) The Treaty of Waitangi in New Zealand’s Law and Constitution. Wellington, New Zealand: Victoria University Press.

31.

Rainie

Wellman

(2012) Networked: The New Social Operating System. Cambridge, MA: The MIT Press.

32.

Scannell

(2000) For-anyone-as-someone structures. Media, Culture & Society 22(1): 5–24.

33.

Schmidt

(2014) Twitter and the rise of the personal publics. In: Weller

Bruns

Burgess

et al . (eds) Twitter and Society. New York: Peter Lang, pp. 3–14.

34.

Splichal

(2006) In search of a strong European public sphere: some critical observations on conceptualizations of publicness and the (European) public sphere. Media, Culture & Society 28(5): 695–714.

35.

Stubbs

(2010) Three concepts of keywords. In: Bondi

Scott

(eds) Keyness in Texts. Amsterdam: John Benjamins, pp. 21–42.

36.

Taylor

(2004) Modern Social Imaginaries. Durham, NC: Duke University Press.

37.

TEPHRA (2012) Community Resilience: Case Studies from the Canterbury Earthquakes, vol. 23. Wellington, New Zealand: Ministry of Civil Defence & Emergency Management.

38.

Wagner

(2011) Something to laugh about: a few thoughts on humor in post-earthquake Haiti [blog post]. In: Savage minds: notes and queries in anthropology, 13 January. Available at: http://savageminds.org/2011/01/13/something-to-laugh-about-a-few-thoughts-on-humor-in-post-earthquake-haiti/(Accessed 17 September 2015).

39.

Wendel

(2012) Political culture and the rule of law: comparing the United States and New Zealand. Otago Law Review 3: 663–702.

40.

Zhou

Bandari

Kong

et al , et al. (2010) Information resonance on Twitter: watching Iran. In: Proceedings of the 1st workshop on social media analytics, Washington, DC, 25–28 July, pp. 123–131. New York: ACM.