key: cord-0054464-lol6bncc
authors: Vestergaard, Anne; Uldam, Julie
title: Legitimacy and Cosmopolitanism: Online Public Debates on (Corporate) Responsibility
date: 2021-01-05
journal: J Bus Ethics
DOI: 10.1007/s10551-020-04703-1
sha: 5da5099afec6523680117909f51c9bfae3906c1d
doc_id: 54464
cord_uid: lol6bncc

Social media platforms have been vested with hope for their potential to enable ‘ordinary citizens’ to make their judgments public and contribute to pluralized discussions about organizations and their perceived legitimacy (Etter et al. in Bus Soc 57(1):60–97, 2018). This raises questions about how ordinary citizens make judgements and voice them in online spaces. This paper addresses these questions by examining how Western citizens ascribe responsibility and action in relation to corporate misconduct. Empirically, it focuses on modern slavery and analyses online debates in Denmark on child slavery in the cocoa industry. Conceptually, it introduces the notion of cosmopolitanism as a general disposition of care and responsibility towards distant others, conceived as a prerequisite for the critical evaluation of corporate (ir)responsibility in the Global South. The analysis of online debates shows that citizens debate child slavery in terms of individual consumer responsibility rather than corporate responsibility. Corporations are not considered potential agents of change. As a consequence, online citizen debates did not reflect a legitimacy crisis for the cocoa industry, as debates over responsibility were overwhelmingly concerned with the agency of the Western individual, the individual agency of the speakers themselves. Participants in debates understood their agency strictly as consumer agency.

Traditionally, the judgements of evaluators such as the legacy press, accreditation bodies, surveys and NGOs have been seen as key to the construction of organizations' legitimacy (Joutsenvirta and Vaara 2015; Palazzo and Scherer 2006; Schultz et al. 2013) . Social media platforms have been vested with hope for their potential to enable 'ordinary citizens' to make their judgments public and contribute to pluralized discussions about organizations and their perceived legitimacy (Etter et al. 2018) . Especially consumers as 'ordinary citizens' have been highlighted as empowered agents in this respect (Dauvergne and LeBaron 2014; Etter et al. 2018) . This conception of the pluralizing function of social media rests not only on the assumption that civil society actors are able to effectively influence the practices of corporations. It is also preconditioned by the existence of a commonly shared sense cosmopolitan agency, that is, a recognition of global responsibility and an ability to integrate this general concern into every day practices (Nussbaum 1996; Beck 2002; Calhoun 2002; Held 2003; Silverstone 2007; Beck and Grande 2007; Chouliaraki 2010) . This precondition remains underexplored in relation to corporate responsibility. Therefore, this paper asks how do citizens attribute responsibility and perceive possibilities to act against social injustice in social media platforms?

The paper addresses this question by examining the extent to which Western citizens respond with a sense of cosmopolitan agency to modern slavery. It takes its point of departure in the broadcast of the documentary "The dark side of chocolate" on Danish prime time television-a broadcast which received significant subsequent media coverage, with 256 news items mentioning it in the following weeks (see "Methods" section for an overview of the coverage). What makes this documentary a particularly pertinent case for exploring the potentialities of citizens for pushing companies towards effective action is that the documentary was animate in holding companies accountable for continued practices of illegal child labour in the cocoa industry in spite of its strong commitment to self-regulation as early as 2001 (the so-called Cocoa Protocol). Using thematic content analysis, the paper examines 949 comments made by ordinary citizens (non-expert, non-activist) during a 3-week period following the broadcasting of the documentary. The paper approaches these comments as part of the discursive construction of corporate legitimacy, and examines the ways in which citizens (i) ascribe responsibility for modern slavery, (ii) express expectations of and possibilities for action, and (iii) legitimate their views and expectations.

The analysis shows that instead of collective conceptions of agency vis-à-vis corporate power, participants in debates understand their agency strictly as individual, economic consumer agency and demonstrate little faith that even this agency can be productively applied to change corporate misconduct. The findings testify to an absence of a cosmopolitan outlook in social media publics, which implies that information about corporate irresponsibility, such as the use of child slaves, is not sufficient to mobilize such publics to challenge corporate practices even when they are presented with a call to do so. The paper thus contributes to the literature on discursive legitimation first by proposing that its scope of investigation should be widened to include not only those publics who engage in critical constructions of corporate legitimacy, but also the multitudes that do not. Secondly, the paper makes a theoretical contribution by introducing the notion of cosmopolitanism as an analytical and conceptual device for nuancing our understanding of the dispositions that underlie evaluations of corporate conduct. The empirical contribution of the paper, finally, lies in its suggestion that social media publics are less inclined to challenge corporate behaviour than what the literature has tended to assume.

The paper proceeds as follows: the first section frames the conception of corporate legitimacy as discursively constructed in relation to responsibility. The next section relates this perspective to the role of social media in affording ordinary citizens possibilities for contributing to discursive constructions of corporate legitimacy. The following section provides theoretical perspectives on the interrelations between responsibility in relation to distant others from the perspective of cosmopolitanism. Following this theoretical framing, the paper's research design and method is introduced. Subsequently, textual analyses of online debates on child slavery are presented, followed by a discussion of the implications of expectations that 'ordinary' citizens will play a significant role in constructing corporate legitimacy in relation to existing and further research, and, finally, the conclusion.

Legitimacy has generally been defined in terms of social acceptance (e.g. Crane 2013; Palazzo and Scherer 2006; Schultz et al. 2013) . This conception takes its point of departure in Suchman's (1995) contention that legitimacy is "a generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions" (p. 574). In their seminal call for a discursive approach to corporate legitimacy, Palazzo and Scherer (2006) argue that legitimacy has become a key issue for multinational companies, as they face increased levels of complexity, often with manifold contradictory requirements in relation to legal and moral responsibility (Palazzo and Scherer 2006) . Drawing on Habermas' notion of communicative ethics, Palazzo and Scherer (2006) approach corporate legitimacy as constructed through public debate. A key premise here is that "public acceptance can no longer be decoupled from public discourse" nor indeed from corporate responsibility (Palazzo and Scherer 2006, p. 79) . In this way, corporate legitimacy is intrinsically tied to corporate responsibility and, crucially, debate about corporate responsibility Colleoni 2013) . This approach has gained traction and is now well versed in business and society research on dialogue vis-à-vis a multiplicity of organizational stakeholders (e.g. Brennan et al. 2013; Christensen et al. 2015; Joutsenvirta and Vaara 2015; Rodrigue et al. 2015; Whelan et al. 2013) , and on criticism beyond corporate control (Author B 2016; Etter et al. 2019; Schultz et al. 2013) . With the proliferation of social media platforms also ordinary citizens have come to play an increasingly important role in contributing to public discourses on responsibility (Etter et al. 2019) .

Indeed, the possibilities for citizens to voice their opinions in social media have been celebrated for their pluralizing potential for public discourses (e.g. Uldam and Vestergaad 2015; Castells 2013; Etter et al. 2018) . NGOs and activist networks have been shown to rely on social media platforms in their attempts to influence public discourses (Author A 2015; Authors 2015; Bennett and Segerberg 2013; Böhm et al. 2010; De Bakker 2015) . Schultz et al. (2013) argue that the proliferation of social media platforms has changed the construction of corporate legitimacy, so that it "…is not only formed in separate spheres of society, within hierarchical orders of stable institutions or powerful rational elites" (Habermas 2001) , but co-constructed by the citizenry (p. 685). The role of ordinary citizen voices has been examined in relation to outcomes and influences. In relation to outcomes, for example, Castelló et al. (2016) have analysed tweets to assess citizens' reactions to CSR communication and Hunter et al. (2008) have analysed the use of websites dedicated to a call for consumers to respond to layoffs in the food company Danone by boycotting products from the company. In relation to influences, for example, Author A (2015) have analysed Facebook comments to assess the role of citizens' criticism in influencing dominant discourses on Nestle's legitimacy in relation to deforestation. Accepting the premise that corporate legitimacy is co-constructed by multiple stakeholders and that ordinary citizens are potentially empowered by social media to play a key role in this co-construction raises important questions about how citizens perceive a corporation's conduct as "desirable, proper, or appropriate" (Suchman 1995, p. 574) . Caruana and Crane (2008) have noted how citizens in their roles as consumers have been "charged with a responsibility to use their 'purchase votes' to effect positive social outcomes" (p. 1495). On the basis of this, they stress the importance of critically addressing the assumptions about citizens as consumers that underpin this responsibility. In this regard, they highlight a number of assumptions as problematic, notably the assumptions that citizens are concerned about issues such as climate change, social justice and human rights, and that citizens actually act on these concerns (Caruana and Crane 2008) . While research points to a clear concern with issues of climate change, social justice and human rights among many citizens (Della Porta and Diani 2020), the ways in which citizens ascribe responsibility for companies' social and environmental misconduct remain underexplored. If citizens do not assign responsibility to companies for social and environmental misconduct, the companies' legitimacy remains unchallenged. In relation to modern slavery in the corporate value chain, it warrants questions about citizens' cosmopolitan dispositions. In the following, we introduce notions of cosmopolitanism to capture the preconditions for citizens to assign responsibility to companies for their misconduct towards distant others.

For citizens to engage with companies' transnational practices in relation to climate change, social justice and human rights, a cosmopolitan disposition is required. The notion of cosmopolitanism is widely used to focus attention on the conditions that generate cross-national ties, variously defined in articulations of support and responsibility towards members of other societies, especially of course, when these suffer some misfortune (Nussbaum 1996; Beck 2002; Calhoun 2002; Held 2003; Silverstone 2007; Beck and Grande 2007; Chouliaraki 2010) . Different cosmopolitanisms share the idea of a figure with a cultural disposition, which recognizes global responsibility and can integrate this broader concern into everyday life practices. Beck's (2006) distinction between cosmopolitanization and "cosmopolitan outlook" is important in this respect, because it highlights the difference between "latent", "unconscious" and "passive" cosmopolitanism (p. 19) and "its self-conscious political affirmation, its reflection and recognition" (p. 21). The "selfconscious political affirmation" of a cosmopolitan outlook is key to a sense of responsibility and agency and therefore analytically useful for examining the ways in which citizens accept and ascribe responsibility and expectations for action to different societal actors (Kyriakidou 2015) . Also of note is Beck's later (2012) qualification of a cosmopolitan outlook as part of a bottom-up rather than top-down process (Beck 2012) . In this view, "the global public impinges upon political communication in other kinds of public discourse, creating as a result new visions of social order in which codification of both Self and Other undergo transformations" (p. 12). This resonates with the conception of corporate legitimacy as discursively constructed by multiple societal actors, including ordinary citizens. Where conceptions of cosmopolitanism differ is on the problem of distance, the question of how individuals' realm of responsibility may be extended from the proximal to the distant (Hannerz 1990; Berking 1996; Tomlinson 1999; Bauman 2001; Silverstone 2007; Chouliaraki 2010) . The issue of distance is particularly pertinent to questions of Western citizens' sense of responsibility with respect to climate change, social justice and human rights in relation to corporate conduct in the Global South.

The fundamental condition for generating cross-national ties, extending individuals' realm of responsibility from the proximal to the distant is visibility, which is thus a prerequisite for any cosmopolitan agency. A cosmopolitan disposition requires visibility, because there is no immediate contact between the Western citizens and those populations whose suffering they could respond to. Therefore, western publics rely on media and experts for knowledge about global reality and equally on media and experts for action (Author A 2014). The increasing diffusion of information and images of distant events through the media opens up the local world of the spectator to the sight of the other and to non-local experiences. Visibility may help to stimulate and deepen a sense of responsibility, which may translate into a public-political consciousness (Giddens 1990; Thompson 1995; Tomlinson 1999) . However, critics of this view contest the idea that mere co-presence is sufficient to contain the promise of a global sense of responsibility (Baudrillard 1988; Robins 1994; Ignatieff 2001; Sontag 2003) . Rather than responsibility, it is argued, the visibility of misfortune may foster apathy and acceptance of the banality of suffering, or "compassion fatigue" (Tester 2001; Chouliaraki 2006; Author A 2008) . The intervention of the media, in this view, causes the mediated reality to be sanitized. The embeddedness of scenes of misfortune in the materiality of both technology and the domestic milieu severs the reality of suffering from its own nexus of sensations and reinserts suffering into another nexus of sensations-the spectator's own immediate states and moods (Chouliaraki 2010; Tester 2012) . The post-modern absence of grand narratives and shared morality to organize dispositions vis-à-vis suffering populations, it is argued, produces uncertainty and selfreflexivity-a tendency to engage with one's own moral doubts, rather than with the condition of the other (Author A 2011; Chouliaraki 2013).

The potential of media visibility for fostering global responsibility is thus subject to academic debate. Rather than assuming an optimistic or pessimistic perspective, this paper examines empirically the extent to which Western citizens respond with cosmopolitan dispositions to modern slavery made visible through the media.

In the following we detail the contextual, empirical and analytical basis of our study.

In order to investigate how Western publics contribute to the construction of corporate legitimacy, the paper investigates how this notion is articulated in discourses of ordinarynon-expert, non-activist-citizens in Denmark. The case is a public issue associated with a global production industry, namely the cocoa industry. In the spring of 2010, a documentary on child trafficking and slavery was broadcast on Danish national TV. 'The Dark Side of Chocolate' was produced by Danish journalist Miki Mistrati and Italian human rights photographer Roberto Romano, with support from Danida, 1 as well as a range of European national broadcasting agencies. Through undercover visits to cocoa plantations in the Ivory Coast, the documentary claims to uncover 'brutal trafficking and illegal child labour in the chocolate industry', in spite of the 2001 Cocoa Protocol by which total eradication of child labour in the cocoa sector should have taken place by 2008. The documentary holds global chocolate manufacturers, such as Nestlé and Tom's, accountable for the exploitation of children in their supply chain and the issue received substantial attention from the Danish mass media in response to the broadcast, with 256 news items mentioning either in the following weeks. As introduced in the section on cosmopolitanism above, this mediated visibility of the use of modern slavery in the chocolate industry has the potential to elicit both a sense of responsibility (Giddens 1990; Thompson 1995) or "compassion fatigue" (Chouliaraki 2006; Author A 2008). Rather than assuming either this optimistic or pessimistic perspective, this paper examines this potential in the analysis below. Due to renegotiations of the failed Cocoa Protocol, 2010 is a critical year for examining public perceptions of legitimacy and responsibility and the potential of public opinion to drive change in cases where an industry does not deliver on commitments made.

Comments from ordinary citizens on social media platforms are used as a source of unsolicited data. The advantage of social media data is that these provide access to a multiplicity of voices and discourses, as they unfold in a pre-existing communicative environment (Castelló et al. 2016) , and that there is a widely held expectation that social media hold an emancipatory potential which may amplify discourses of critique with the potential to hold corporations accountable for their actions (Bennett and Segerberg 2013; Author A 2015) . In a Danish context, comments on social media platforms have the potential to provide access to a multiplicity of voices and discourses, as the country had an internet penetration of 87% when the documentary was released, and news and social media platforms such as Facebook, blogs and news sites were accessed at least weekly by users between 16 and 74, across demographic characteristics (Statistics Denmark 2018). However, debates in social media also have characteristics, which restrict their generalizability and representativeness. While social media platforms may facilitate open exchanges of political ideas (Castells 2013; Papacharissi 2012) , they have also been shown to create echo-chambers with merely like-minded citizens debating, and facilitate the circulation of misinformation, thus polarizing and intensifying opinions, and contributing to widening gaps between those on opposite sides of public issues (Askanius and Mylonas 2015; Fenton 2016; Tufekci 2015) . As such, the comments in our dataset may reflect polarized debate, with commenters more prone to venting unpopular or controversial positions rather than engaging with the substantive content of the messages. Indeed, most commenters comment only once or twice rather than engaging in longer debates (see Author B 2013 for a discussion of this tendency in relation to societal issues) and 762 individuals are represented in the dataset. While our focus is on non-expert, nonactivist citizens in Denmark, some comments in our dataset were made by experts (communications professionals and NGO staff), but, to the extent that it is verifiable, only on the popular liberal blog, Punditokraterne.dk, and representing a minor part of the entire dataset (six comments). Other comments may, of course, have been made by experts using the anonymity of online commenting to let their expert position remain undisclosed. However, the comments were retrieved from popular rather than extremist platforms, and, with these important reservations in mind, can be taken to represent a window onto the existence of a commonly shared sense of global responsibility and cosmopolitan agency.

For a 3-week period following the broadcast in Denmark, social media platforms were monitored for responses to the documentary. The monitoring was based on daily Google searches using the documentary title (Danish and English) as keywords as well as related terms [chocolate OR cocoa AND (child slavery OR child labor in Danish and English)]. The latter were subsequently manually verified to relate to the documentary. Following Bennett et al. (2014) social networking sites, blogs and online news media platforms were included so as to capture citizen voices as they appeared in diverse media ecologies. Data selection was based on a "tracking" strategy (Marcus 1995) , which entails following the struggle or debate and helps counter selection bias in relation to discursive struggles (Author B 2010). In this way, the study follows the debate as it appeared across social media platforms rather than pre-selecting specific platforms.

During the three-week period, a total of 949 texts by members of the general public were produced and these constitute the data set for analysis. The texts derive from 12 political debate blogs (with no explicit organizational affiliation), 87 blog comments, 591 comments on online news sites and 261 entries in debate forums. During the data collection period, there was no Facebook commentary. At the time of release in 2010, different media practices (Couldry 2012) characterized Facebook commentary, with a focus on socializing rather than public debate (Enli and Thumim 2012) . Facebook commentary emerged subsequently as a consequence of the documentary showing outside of Denmark. For example, the documentary's director Miki Mistrati created a Facebook group for the documentary in 2010. The only two posts (both by Mistrati, both about the documentary's release internationally) during the data collection period were not commented on.

By far the majority of public text (82%) was produced during the first 4 days after the documentary was broadcast and derives from four sources: (1) the 'society and politics' debate forum of the widely popular debate platform Jubii (dindebat.dk), (2) the blog Zeitgeistmovement.dk, which was the Danish platform of the international Zeitgeist Movement, a self-proclaimed "sustainability and public health organization" (thezeitgeistmovement.com, nd), (3) the popular liberal blog, Punditokraterne.dk, which since 2005 has been co-authored by a collective of journalists, academics and practitioners in the fields of law, economics and international relations, and (4) the national tabloid Ekstrabladet's online news platform (ekstrabladet.dk), which was the most popular online news platform in Denmark with most comments per article in the early 2010s (Schrøder and Kobbernagel 2012). Most of the comments (613) derive from Ekstrabladet, which covered the broadcast of the film most ardently (with 12 articles) and the director of the film, Mistrati, at the time was Head of News. Debates in these platforms are generally characterized by numerous iterations and turns from citizens, as reflections of users offering points, counterpoints and elaborations. The remaining 18% of comments are from a wide range of news platforms, including the (now discontinued) left-wing Modkraft and the online platform of the (now discontinued) free daily newspaper Urban. They resemble the other 82% of the comments in terms of placement of responsibility and cosmopolitan dispositions, but vary in terms of iterations and turns from citizens, as each site only includes few (1-3) comments. As such, the vast majority of comments are from platforms that are generally considered popular rather than extremist. The commentators can thus be seen as ordinary citizens in the sense that they are non-expert, non-activist commentators in popular platforms. Comments from all platforms are included in the analysis. However, even though the majority of comments are from what can be considered mainstream, popular platforms, they may be characterized by polarized debate, as is often the case in online platforms (Author B 2013) . Comparing the online comments to the media coverage of the documentary shows that the media coverage in some ways reflects the general sentiments of the comments. A total of 256 news media mentioned the documentary during the 3-week period of data collection (the media coverage was collected through the Danish news database Infomedia). The majority of mentions were from the tabloid Ekstrabladet (12) and their online platform (12) and the centre-left broadsheet Politiken (12). The remainder are mainly from Ritzau and regional newspapers. Overall, the media coverage deplores the use of child labour and child slavery and assigns responsibility to politicians or companies, calling for fair-trade schemes and a voluntary code of conduct (40%). However, 29% articulate child labour (often not mentioning child slavery) simply as unavoidable or as more complex than represented in the documentary, urging consumers to continue buying chocolate so that the children can keep their jobs. This position is mainly informed by press releases from a major Danish supermarket chain (the co-op FDB, now COOP) and the Danish chocolate manufacturer Toms, also mentioned in the documentary. In this way, the media coverage reflects the comments in our study on key points, although it differs in terms of the assignment of responsibility, as our analysis below will show.

Collecting data from online platforms must inevitably entail ethical considerations (Author B, 2013 (Author B, , 2014 . In this case, informed consent was not sought. This decision was reached, because all comments were posted on publicly accessible platforms . Moreover, the popularity of the platforms means that they are not just publicly accessible, but also accessed by publics on a daily basis. Nonetheless, we have chosen to anonymize all commenters to protect them-also when they have used seemingly anonymous profiles-as many comments are politically charged (Markham and Buchanan 2017; Kozinets 2002; Author B 2013) . No biographical data or information that could lead to source identification was collected (Markham and Buchanan 2017) and pseudonyms were used for all participants (created using an online random name generator).

In analysing the role of ordinary citizens in constructing corporate legitimacy through online debates, the analysis employs a thematic content analysis, in order to draw out the reoccurring discursive patterns in our relatively large dataset of 949 texts and enable rich description hereof (Chouliaraki and Zaborowski 2017) . Drawing on Chouliaraki and Zaborowski (2017), our analytical categories were first developed from the overarching conceptual focus of the study: assignment of responsibility and action orientation in the discursive construction of corporate legitimacy (Barros 2014; Castelló et al. 2016; Etter et al. 2018; Joutsenvirta and Vaara 2009; Joutsenvirta 2011) . These strategies were separated into constructive and destructive legitimation, where the former works by building legitimacy, the latter works by destructing the legitimacy of the opposition (Fig. 1) .

In this way, themes in the first stage were identified through an iterative process, moving between data and theory. In the second and third stages, comments were then coded and combined in NVivo, leading to the identification of three overall articulations of legitimation: Rationalization, deauthorization and demoralization through a combination of open coding (Meyer 2001 ) and, following Vaara et al. (2006) , an adaptation of these themes from Van Leeuwen's (2007) categories of legitimation, rationalization, authorization and moralization. These articulations were separated into constructive and destructive legitimation, where the former works by building legitimacy, while the latter works by destructing the legitimacy of the opposition:

(i) Constructive legitimation: Rationalization. Legitimation that articulates through two types of rationalization: systemic rationalization and agency rationalization. Systemic rationalizations articulate child labour and slavery in general socio-economic terms, e.g. with reference to economic growth. Agentic rationalizations are oriented around the self and question the ability of Western publics to function as agents of change. (ii) Destructive legitimation: Deauthorization and demoralization. Deauthorization questions the authenticity, sincerity or relevance of communication about child labour and slavery, attempting to weaken the authority of key texts and arguments. Demoralization is articulated through a refusal of morality is a common devise used for legitimation, e.g. by insisting on the right to fulfil consumption desires taking precedence over moral obligation (Table 1) .

This led to a set of codes, which are exhaustive in the sense that all comments in the dataset belong to at least one of the coded categories. Through this iterative process, Van Leeuwen's (2007) category of mythopoesis, as legitimation conveyed through articulations of reward for legitimate actions and punishment for non-legitimate actions did not emerge as prominent in the comments. Also, in this iterative process responsibility, legitimation and authorization departed from their theoretical bases to empirically informed constructs, i.e. responsibility from corporate to individual, legitimation from corporate to consumer, and authorization from institutional to individual. We return to this in the discussion.

An intercoder reliability test generated a reliability score of between 0.88 and 0.96 across sub-theme categories. This iterative process enabled a conceptually guided data-driven approach. To reiterate, our analytical categories seek to classify how, in the framework of discursive legitimation, citizens view corporate legitimacy from the perspective of different cosmopolitan dispositions, and how they ascribe responsibility and action in relation to responsibility.

The following analysis shows that instead of collective conceptions of agency vis-à-vis corporate power, participants in debates understand their agency strictly as individual, economic consumer agency and demonstrate little faith that even this agency can be productively applied to change corporate misconduct.

The documentary primarily assigned responsibility to chocolate companies, especially Nestlé, but also to civil society organizations (CSOs) such as the 'International Labour Organization'. The documentary took its point of departure in the failure of the chocolate industry to follow the 2001 Cocoa Protocol commitments to "eradicate the use of child slavery and achieve full slave-free certification by 2005". Mirroring this explicit agenda, the documentary gives substantial attention to the industry as such as well as to specific industry actors, in particular Nestlé. Several minutes are dedicated to Miki Mistrati protesting outside of the Nestlé headquarters. The agreement signed by the chocolate manufacturers' foundation was to be implemented in collaboration with a series of CSOs such as 'Free the Slaves' and 'International Labour Organization'. As a consequence, these CSOs also figure prominently in the film, with for instance 3 min of interview with International Labour Organization.

In spite of the pronounced proposal of the documentary towards holding industry accountable and despite its naming and shaming of specific named chocolate manufacturers, this articulation is not appropriated by the online publics. The online debates had very little to do with protesting against irresponsible behaviour in industry.

Only a total of 24 comments reference chocolate manufacturers in general. Only nine of these assign responsibility to manufacturers and just two mention any of the specific chocolate companies named in the documentary (Nestlé, Mars, Cargil, Barry Callebaut). The rest assign responsibility to multinational chocolate companies without mentioning specific companies. Instead, they articulate profit-driven logics as drivers of corporate irresponsibility and accountability schemes as mechanisms that serve to shift responsibility from companies. This is reflected in the following two comments: I wonder if it's because the industry doesn't care, otherwise they wouldn't make a system of intermediaries so that they could deride responsibility (comment no. 56, Ekstrabladet 16.03.2010). How is action orientation and assignment of responsibility legitimated? Constructive legitimation -Legitimates the position by building a positive argument Systemic rationalization Articulates child slavery in general socio-economic terms Agentic rationalization

Addresses the feasibility of actors to function as agents of change Destructive legitimation -Legitimates the position by undermining opposing positions Deauthorization

Questions the authenticity, sincerity or relevance of communication about child labour and slavery, attempting to weaken the authority of key texts and arguments Demoralization

Refusal of morality is used as a devise for legitimation, either by refusing the claim to morality of (imagined) opponents or one's own

There's such big money at stake for the big corporations that not a damned thing will happen, I am afraid (comment no. 119, Punditokraterne 17.03.2010).

Rather than assigning responsibility to companies, this very limited contestation is articulated as broad systemic critiques. However, this broad systemic critique not only does not concern itself much with industry, it also does not attribute responsibility to other societal actors. The articulations of responsibility in the comments rarely assign responsibility to civil society or the political system. In the entire dataset, only four comments mention political actors, and civil society organizations are mentioned in just two entries. In this way, political and civil society actors hardly figure into the debate. Neither of these actors, held to be the primary pillars of democracy, are articulated by these publics as having any conspicuous responsibility with respect to the circumstances around child slavery. Thus, in spite of the pronounced framing of the documentary towards holding both industry and civil society actors accountable, and despite its naming and shaming of specific chocolate manufacturers, this discourse is not appropriated by these online publics. The online debates had little to do with the denunciation and protest, which the documentary invites. Intrinsically related to these articulations of responsibility is the question of agency. The next section analyses commenters' articulations of various societal actors' responsibility and possibilities for acting on modern slavery.

Overall, 293 comments articulate an orientation towards action, while 569 comments articulate an inaction orientation, that is, do not take up the proposal to act against child slavery that is articulated by the documentary. In other words, the vast majority of comments refuse action, as in for example "I wouldn't want to deprive them of their livelihood" (comment no. 441, Jubii Debatforum, 18.03.2010) or "If I wasnt trying to lose weight I'd eat their slave chocolate every day (comment no. 20, Ekstrabladet 19.03.2010)".

Among action-oriented comments, as indicated above, comments that mention companies as responsible for acting against modern slavery are scarce. Only nine comments attribute companies with such responsibility. When users do talk about demanding action from industry, they do so in terms of consumption, as illustrated in the following comment:

we'd have to pay a fair price for chocolate, but I do believe so much of the trading margin goes to the producers, who would just have to turn down profits a little. The result could be that producers and consumers shared the added cost (comment no. 783, Jubii Debatforum 19.03.2010) .

In fact, the majority of action-oriented comments (77%) assign responsibility to act to consumers. Action statements construe action as a matter of consumption, with proposed modes of action as boycott, critical consumption or stopping chocolate consumption altogether, as exemplified below.

A boycott for a month or two certainly helps. Although the sales turn back to normal.. the producers do get sweaty palms by being the bad news of the day (comment no. 803, Punditokraterne, 20.03.2010) . I won't stop eating chocolate, but maybe I will go harder for fair-trade producers instead (comment no. 640, Jubii Debatforum, 18.03.2010) .

A significant amount (66%) of comments articulate a refusal of action. The action that is refused is consistently consumer action, i.e. critical consumption, boycott, etc.

In sum, only a minority of comments are action oriented, whereas most (66%) refuse action. In both the action and inaction-oriented comments, the imagined action is overwhelmingly consumer action, whereas civil society organizations and political actors are not to any significant degree construed as potential agents of change. In the following section, in order to assess the extent to which these comments reflect a cosmopolitan disposition in commentators, we explore the basic articulations by which action and inaction orientations are legitimized.

This section analyses commenters' arguments to legitimate their (in)action orientation. In the entire data set, there is not a single instance where either action or inaction is legitimated on moral grounds. Instead, articulations of legitimation either work constructively by providing a rationalization or destructively, by questioning the legitimacy of opposing views and orientations. These two types of articulations will be analysed in turn below.

A total of 318 comments (34%) articulate legitimation as a justification that works constructively. All of them do so by means of rationalization.

In legitimating their action orientation, commenters use two different types of rationalizations, systemic rationalizations and agency rationalizations. Systemic rationalizations (205 comments) articulate child labour and slavery in general socio-economic terms. Reflecting the minimal assignment of responsibility and agency to specific societal actors-corporations, civil society organizations or policy makers-shown in the previous sections, comments rarely provide sources, material or immaterial, as they have been shown to do on corporations' social media sites (Etter et al. 2019 ). As such, they are not authority based. Rather, they show attempts at independently analysing and diagnosing the problem at stake. When it comes to rationalizing action, this is consistently done on the basis on an economic logic. The below quotes exemplify this:

Prices are kept artificially low because 7-8 year olds are used as labour. There would still be a marketprices would merely go up (comment no. 643 Jubii Debatforum, 18.03.2010) . The consequence would be that actual jobs were established in the countries in question, for adults who would then be able to provide for their children and send them to school. The increased purchasing power in the relevant countries would increase growth and create more jobs (comment no. 723, Ekstrabladet 20.03.2010).

Conversely, rationalizations of inaction consistently articulate child labour and slavery as a natural and static state of affairs, which exists in isolation, independently of other actors. Commenters present these articulations as if they were rational substantiation (Joutsenvirta and Vaara 2009) , as illustrated in the following examples:

Child labour is common in most places in the third world and it is better than being a child soldier (comment no. 14, Jubii Debatforum, 16.03.2010) . The world will never be perfect and there will always be losers. We are just fortunate not to live in a place where the world is like that (comment no. 562, Ekstrabladet 20.03.2010).

The comments exemplify a multitude of commentary which articulates child slavery as simply unavoidable rather than "situating and relativizing [their] own form of life within other horizons of possibility" and seeing themselves "from the perspective of cultural others" (Beck 2006, p. 89) .

At the extreme end of this debate, numerous comments state that child labour still exists in Denmark as a constructive way of socializing children into economic reality. This is captured in the following comment:

…and that was a good way of introducing children to the tough realities of life. A bit of work every day never did anyone any harm (comment no. 331, Jubii Debatforum 18.03.2010).

Agentic rationalizations (113 comments), in contrast to systemic rationalizations, are oriented around the self and based on a perceived lack of agency in Western publics. Rationalizations, which question the ability of Western publics to function as agents of change, do not only problematize the effectiveness of action but also problematize the adequacy of information available (uncertainty) and the intolerability of awareness, that being conscious of the ultimate consequences of our consumption involves an unbearable burden of guilt. These types of rationalizations are illustrated in below quotes: I can't go around feeling guilty 24/7 about the things I buy (comment no. 47, Jubii Debatforum, 16.03.2010) .

The sense of inefficacy articulated in these types of comments reflects the absence of "Self-Conscious Political Affirmation" characteristic of a cosmopolitan outlook (Beck 2012; Kyriakidou 2017) . At the same time, they point to a logic of complicity (Chouliaraki 2013) , which is enabled by commenters' tendency to ascribe responsibility for child slavery in the chocolate industry to themselves as consumers rather than to companies, political actors or CSOs. In this way, rationalizations exclude "the moral question of 'why'" (Madianou 2013, p. 255 ).

Destructive legitimation functions by destruction, that is, by delegitimizing opposing perspectives or even, as we shall see, delegitimizing opposition to opposition. This is done either by deauthorizing the opposing view or by opposing moralizing arguments through ridicule. In total, 423 comments (44%) articulate legitimation as a justification that works destructively.

Deauthorization articulations (278 comments) question communication concerning the problem of child labour and slavery. By bringing into question the authenticity, the sincerity, or the relevance of communication, the authority of key texts and arguments concerning child labour and slavery is weakened.

Questioning authenticity involves challenging the reliability of the call for action by raising doubts as to whether what is communicated is the real, unmanipulated and unbiased truth. A total of 71 authenticity questions are raised with respect to the documentary itself, the news reporting which followed the documentary as well as arguments raised by action-oriented members of the public. This is illustrated by the following example from a commenter, who questions the authenticity of the documentary:

That they [children interviewed in plantations] say they'd rather live on the streets -a journalist can make them say anything. Look pitiful and you will be on TV (comment no. 270, Ekstrabladet, 17.03.2010) .

Relevance questions (a total of 24 comments) raise doubts as to the appropriateness of the documentary's targeted audience as well as its scope. Relevance questions are raised not with respect to the fundamental significance of the problem of child labour and slavery to Western publics, but as to characteristics of communication and journalism, as in the following example:

What else is new? This is not news. We cannot grow cocoa in Denmark and we cannot transfer out morals to Africa, so news? Where? (comment no. 897, Jubii Debatforum, 22.03.2010) . If you didn't already know that cocoa fruits are picked by child slaves, then you are frankly a blind idiot with no ability for reasoning (comment no. 356, Zeitgeistmovement, 18.03.2010) .

Sincerity questions (a total of 14 comments) concern the sincerity of intentions behind calls for action. Put under suspicion here is the genuineness, consistency and moral purity of those who argue for action with the more or less implicit premise that in order for the call for action to be legitimate, there can be no secondary motives involved-the proponents must be selflessly engaged, and there can be no deviationsall actions of proponents must obey the same moral logic. Questions as to sincerity are raised with respect to the sincerity of journalism, including news reporters, documentary producers as well as the contesting public.

It is the people who write articles about this that profit the most from child labour (comment no. 323, Ekstrabladet, 17.03.2010) .

Questioning certainty, finally, involves challenging the sufficiency of information available to enable informed judgments and appropriate action. A total of 36 comments articulate deauthorization in this way, which suspends judgment to the benefit of doubt.

I am left with a lot of questions, which they do not even try to answer with the material laid out in the media. And that certainly is not conducive for change (comment no. 678, Punditokraterne, 21.03.2010) .

While articulations of deauthorization do not directly legitimate inaction, they delegitimize the call for action by undermining its authority. In van Leeuwen's terms it removes the opponent's answer to the question 'why'. At the same time, it points to the fundamental challenge of mediated appeals, that they are always associated with a degree of uncertainty, which makes them vulnerable to deauthorization.

As already mentioned, moralization is not articulated in terms of legitimation in the debates studied here. Rather the opposite is the case. In the case of commenters refusing action, a simultaneous refusal of morality is a common devise used for legitimation. Such comments (37) ridicule the perceived moral imperative in the documentary as illustrated by the comment below:

Political correctness my ass. The alternative is that CHILD SLAVES sit at home SNIFFING GLUE (comment no. 89, Ekstrabladet 16.03.2010) .

The most common form of legitimation by demoralization (71), however, insists on the right to fulfil consumption desires as taking precedence over moral/ethical obligation. By insisting on the primacy of desire, these comments ridicule a morality that sets aside immediate, personal desire for a distant good. By overstating the refusal of action, these statements involve an element of self-irony, ridiculing their own priorities, while not apologizing for them. As expressions of cynicism, of a generalized disbelief in the sincerity or goodness of human motives, these comments ridicule the ideological faith and hope of others as unrealistic and naïve and, while rejecting any feeling of guilt, simultaneously ridicule their own apathy and disillusionment.

Also commentators who promote action articulate demoralization, but in this case to distance themselves from moralization (37). What is ridiculed here is the sincerity of the commenters themselves and their acceptance of the role of critical consumer, as a form of humility by which they insist that they are not or do not consider themselves morally superior to those that refuse action. This manoeuvre would appear to be a defensive response to latent questions as to the commenter's sincerity, as described above. These comments attempt to reinstall reliability, by denying a position that is morally invested and inscribing it instead into a rational, morally disengaged discourse. This is captured in the following two comments:

I am not talking about a boycott but about choosing other products which are more…. socially sustainable (omg, just threw up a little bit there) but.. well.. yes (comment no. 23, Jubii Debatforum, 16.03.2010 ). I am not happy about the way this is done, but I must admit that chocolate is right under alcohol on my list of things that make me forget normal moral standards (comment no. 466, Zeitgeistmovement, 18.03.2010) .

In these cases, the distancing from moralization serves to neutralize an opposition, which is based on demoralization. Demoralization, then, is articulated in terms of legitimation both as a justification for action and inaction orientations. Also, in the second comment, chocolate is articulated as a luxury good along with alcohol rather than a life staple such as bread, invoking notions of chocolate as a sinful indulgence and privileging consumption desires, which further distances the commenter from moralization. Certainly, in both cases, a distantiation from morality appears to strengthen the legitimacy of one's position.

This study contributes to the discursive legitimation literature (e.g. Glozer et al. 2018; Castelló et al. 2013; Colleoni 2013; Schultz et al. 2013 ) by focusing on the potential of ordinary citizens to shape discursive constructions of corporate legitimacy. The role of ordinary citizens in contributing to the discursive construction of corporate legitimacy has been examined against the backdrop of the increasing popularization of social media platforms (e.g. Castelló et al. 2016; Joutsenvirta and Vaara 2009; Nyberg et al. 2013; Author A 2015) . We argue that this focus is crucial because of current hopes invested in social media publics for holding powerful multinational corporations accountable for their conduct. As Caruana and Crane (2008) have argued, this potential rests on assumptions that citizens are concerned about issues such as climate change, social justice and human rights (Caruana and Crane 2008) . This paper contributes to these debates by focusing on the ways in which social media publics ascribe responsibility and action in relation to responsibility. To assess this (emancipatory) potential we propose to employ the concept of cosmopolitanism. This concept opens up texts to an analysis of the extent to which citizens have the basic disposition necessary to engage critically with corporate conduct in the Global South.

The analysis indicated an absence of such potentiality in two ways. First, in relation to responsibility and action orientation, the analysis showed that these online debates over multinational corporations' implication in practices leading to the worst forms of child labour barely mention industry actors as potential agents of change although the documentary, which was the starting point of debate, was very explicit and animate in holding the chocolate industry accountable for its complicity and demanding action on formal commitments already made by industry. At the same time, debates did not consider the possible role that civil society organizations or political institutions might play in causing, sustaining, aggravating or eradicating it, although the documentary featured interviews with experts from various such bodies.

Instead of collective conceptions of agency vis-à-vis corporate power, the analysis shows how commenters articulate an understanding of possibilities to act as almost solely consumer-based. Indicative of the absence of articulations of collective or corporate responsibility, the analysis shows that the comments propelled by the documentary employ systemic and agentic rationalizations to translate collective and corporate responsibility into a question concerning individual agency. Rather than being concerned with the nature of responsibility or who should be assigned with it, debates simply consider whether or not to change consumption behaviour. The notion that agency can only be individual and economic is taken for granted.

Second, the struggle over whether or not to change consumption behaviour is characterized by high degrees of reflexivity and associated uncertainties. As such, it appears as a clear reflection of a disposition of self-consciousness and suspicion towards all claims of truth. Rather than positions based on moral judgments and a priori beliefs, what we can observe are attempts at analysing the economic dynamics around child labour and slavery and rationalizing action and inaction on the basis of these systemic and agentic rationalizations. Contrary to judgments made from fundamental moral convictions, assessing the impact of actions in such a global context is an endeavour filled with uncertainties. The knowledge on the basis of which such assessments are made is by necessity mediated, by experts and by media, and this gives rise to deauthorization through the articulation of suspicions and questions concerning authenticity, relevance, sincerity and certainty as those used in the analysed debates to delegitimize the call for action. Testifying to this same disposition, morality surfaces in debates not as a basis of legitimation, but as a barrier to legitimation. This means that while media visibility is a prerequisite for critical engagement with corporate conduct in the Global South, media visibility always comes from an expert position. The incommensurability of competing expert positions, which is an intrinsic part of reflexive modernization (Beck 2012) , not only increases uncertainty but also always enables the suspension of judgment to the benefit of doubt.

Commentators' uses of demoralization as a devise used for legitimation reflect what Beck (2006) calls passive cosmopolitanism rather than a cosmopolitan outlook that is based on dialogical imagination. It is a crucial point here that the high degree of reflexivity observed in debates takes the form of self-reflexivity and self-conscious political affirmation, not of a reflexivity that is oriented towards distant others (Author A 2014). In this way, dispositions exposed in the debates look less like Beck's cosmopolitan outlook as part of a bottom-up process (Beck 2012 ) and more like a cosmopolitanism, whose solidarity is driven by consumption, utilitarianism and personal fulfilment rather than ethically, politically motivated cosmopolitanism (Chouliaraki 2013) .

Taken together, the two ways in which the absence of a cosmopolitan outlook is expressed by commenters gesture towards impediments to the potential of Western publics to hold companies accountable for their uses of child labour in parts of their value chains located outside the immediate realm of the Western spectator. The absence of a cosmopolitan outlook impedes inclinations to assign responsibility to corporations and inclinations to engage in political consumption, even in its most individualized form. This implies that information about corporate irresponsibility such as uses of child slaves is not sufficient to mobilize social media publics to assign responsibility and thus challenge corporate responsibility, even when it is presented as a call for them to do so.

The main limitations of this study relate to the fact that the implications of this rather dire result must be seen as highly contextual. It is crucial to keep in mind the specifics of the Danish context of this study and the role of the documentary in mediating information about child slavery and the complicity of specific MNCs. The particular circumstances of the debate, especially as concerns the nature of the documentary, which was its starting point, may limit generalizability. It is possible that a differently framed or authored visibility would have led to a different, more corporate-critical, response from online publics. In addition, our study does not consider the chronology of comments and thus we cannot eliminate the possibility that early comments have set the tone and focus of later commentary. However, previous research indicates that it is in fact the tone of the source, which is the greatest source of influence on the tone of commentary (Hartley and Eberholst 2016). Finally, it is likely that online comments reflect more polarized positions than face-to-face debate. However, whether or not the positions articulated in this online debate can be understood as reflecting wider public dispositions, they certainly appear to be characteristic of the social media debate, with which high emancipatory hopes are vested.

Online debates in the wake of the broadcast of a highly publicized documentary on child labour and slavery in cocoa plantations have provided a window into struggles concerning global responsibility and cosmopolitan agency between 'ordinary' (non-expert and non-activist) members of the Danish public.

The analysis showed that, while the main focus of the documentary was on holding industry accountable for its complicity in child labour and slavery, the role of industry is almost entirely absent from subsequent online comments from ordinary citizens. Debates over responsibility are overwhelmingly concerned with the agency of the Western individual, of the individual agency of the speakers themselves. Participants in debates understand their agency strictly as economic consumer agency and demonstrate little faith that even this agency can be productively applied to change corporate misconduct. Considerations of whether or not to engage in political consumption-the only possibility to act envisioned in the comments-are grounded in attempts at a rational economic calculus, whereas moral and emotional factors do not figure into opinion making in any explicit way. Perceptions of the citizen's individual agency are entirely market-based. In addition, corporations are consistently seen as strictly economic agents. In this way, our analysis contributes to research on the discursive legitimation struggles between different societal agents, and particularly the ways in which they are played out in social media commentary (Author A 2015; Author B 2016; Etter et al. 2019; Joutsenvirta 2011) by illustrating the ways in which the absence of a cosmopolitan outlook impedes Western citizens' inclinations to assign responsibility to corporations, even when it is presented as a call for them to do so.

With this study we have aimed to contribute to a contextualized grounding of research into citizen evaluations of corporate conduct, countering a tendency to explain "sustainable consumerism…by generalized theories of reflexive modernization" (Isenhour 2010, p. 514; Crotty and Crane 2004) . By contributing to debates on the role of citizens in the discursive construction of corporate legitimacy, it is our hope that more scholars will go beyond the current focus on those publics that do engage in critical constructions of corporate legitimacy and also pay attention to those multitudes that do not, in order to examine the dispositions that underpin evaluations of corporate legitimacy in relation to modern slavery and beyond.

Extreme-right responses to the European economic crisis in Denmark and Sweden: The discursive construction of scapegoats and lodestars

Tools of legitimacy: The case of the Petrobras corporate blog

Simulacra and simulation

Whatever happened to compassion?

The cosmopolitan society and its enemies. Theory

Cosmopolitan vision

Redefining the sociological project: The cosmopolitan challenge

Cosmopolitanism: Europe's way out of crisis

Solidary individualism

The logic of connective action: Digital media and the personalization of contentious politics

Dialogism in corporate social responsibility communications: Conceptualising verbal interaction between organisations and their audiences

(Im) possibilities of Autonomy: Social Movements in and beyond Capital, the State and Development

The class consciousness of frequent travelers: Towards a critique of actually existing cosmopolitanism

Constructing consumer responsibility: Exploring the role of corporate communications

Communicative dynamics and the polyphony of corporate social responsibility in the network society

Strategies of legitimacy through social media: The networked strategy

Communication power

Post-humanitarianism: Humanitarian communication: Beyond a politics of pity

The ironic spectator: Solidarity in the age of post-humanitarianism

Discursive closure and discursive openings in sustainability

CSR communication strategies for organizational legitimacy in social media

Media, society, world: Social theory and digital media practice

Modern slavery as a management practice: Exploring the conditions and capabilities for human exploitation

Transitions in environmental risk in a transitional economy: Management capability and community trust in Russia

Protest Inc.: The corporatization of activism

Online activism, CSR and institutional change

Socializing and self-representation online: Exploring Facebook

Measuring organizational legitimacy in social media: Assessing citizens' judgments with sentiment analysis

Social media and the formation of organizational reputation

Left out? Digital media, radical politics and social change

The consequences of modernity

The never-ending story: Discursive legitimation in social media dialogue. Organization Studies

The postnational constellation

Cosmopolitans and locals in a world culture

Democratic accountability and political effectiveness from a cosmopolitan perspective

Beyond control: Crisis strategies and stakeholder media in the Danone boycott of

Human rights as politics and idolatry

On conflicted Swedish consumers, the effort to stop shopping and neoliberal environmental governance

Setting boundaries for corporate social responsibility: Firm-NGO relationship as discursive legitimation struggle

Discursive (de) legitimation of a contested Finnish Greenfield Investment Project in Latin America

Legitimacy struggles and political corporate social responsibility in international settings: A comparative discursive analysis of a contested investment in Latin America

The field behind the screen: Using netnography for marketing research in online communities

Remembering global disasters and the construction of cosmopolitan memory. Communication

Remembering global disasters and the construction of cosmopolitan memory. Communication

Humanitarian campaigns in social media: Network architectures and polymedia events

Chapter 14: Research ethics in context

A case in case study methodology

Compassion: The basic social emotion

Ethnography in/of the world system: The emergence of multi-sited ethnography

Incorporating citizens: Corporate political engagement with climate change in Australia

Corporate legitimacy as deliberation: A communicative framework

Special issue: Socially mediated publicness

Forces of Consumption: From the symbolic to the psychotic. Media

Volume and tone of environmental disclosure: A comparative analysis of a corporation and its stakeholders

The construction of corporate social responsibility in network societies: A communication view

Media and morality on the rise of the Mediapolis

Regarding the pain of others

Publikation: It-anvendelse i befolkningen 2018-Danmarks Statistik

Managing legitimacy: Strategic and institutional approaches

Compassion, morality and the media

Humanitarianism and modern culture

The media and modernity: A social theory of the media

Globalization and culture

Algorithmic harms beyond Facebook and Google: Emergent challenges of computational agency

Civic engagement and social media: Political participation beyond protest

Pulp and paper fiction: On the discursive legitimation of global industrial restructuring

Legitimation in discourse and communication

Corporations and citizenship arenas in the age of social media