Abstract
While we witness a growing belief in transparency as an ideal solution to a wide range of societal problems, we know less about the practical workings of transparency as it guides conduct in organizational and regulatory settings. This article argues that transparency efforts involve much more than the provision of information and other forms of ‘sunlight’, and are rather a matter of managing visibilities than providing insight and clarity. Building on actor-network theory and Foucauldian governmentality studies, it calls for careful attention to the ways in which transparency ideals are translated into more situated practices and become associated with specific organizational and regulatory concerns. The article conceptualizes transparency as a force that shapes conduct in organizational and socio-political domains. In the second section, this conceptualization of transparency as a form of ‘ordering’ is substantiated further by using illustrations of the effects of transparency efforts in the internet domain.
The role of transparency as an effective form of governance is largely taken for granted. The belief in transparency and the workings of ‘sunlight as a disinfectant’ (Brandeis, 1913) relies on a simple and deceptive formula that equates information with transparency and considers the clarity obtained through information to be a direct path to accountability and good governance. That is, if information is shared, we can see things as they really are, and if everything is visible, no bad behavior takes place. But this conception conflates and simplifies what is often a much more complex relationship between transparency and governance. This article seeks to contribute to theorizations of transparency by conceptualizing it as a form of ordering and to point to some ways in which developments in the internet domain can be used to illustrate the value and facets of this conceptualization.
Transparency is intimately tied to developments in governance. Organizations, corporations and regulatory bodies are expected to share information that will give insight into their operations, spur public debate and contribute to accountability (Hood and Heald, 2006; Fung et al., 2007). In broader terms, transparency has emerged as a widely acknowledged governance mechanism driven by voluntary and interactive forms of engagement (Braithwaite and Drahos, 2000). But extant conceptualizations of both transparency and governance have limitations that stand in the way of understanding the role of vision, ‘sunlight’ and information in the context of organizational control and societal steering. Much of the literature seeks to account for the degree to which particular actors and policy processes have institutionalized and reacted to transparency requirements and expectations (Braithwaite and Drahos, 2000). Alternatively, it explains the origins and spread of ‘governance by transparency’ (Fung et al., 2007) by referring to a set of core characteristics of transparency projects and the qualities of the information they make available. Transparency, then, is understood as a matter of disclosing ‘timely’, ‘clear’, ‘accurate’, ‘standardized’, and ‘comparable’ types of information about products or practices in order to enhance decision-making and accountability (Fung et al., 2007; Schnackenberg and Tomlinson, 2014). Furthermore, the literature on the intersection of transparency and governance rarely probes deeper into concrete governance practices. The majority of this literature starts from an understanding of transparency as a regulatory norm seeking to ensure accountability through (often mandatory) requirements about the timely and public disclosure of information (Schneiberg and Bartley, 2008). While transparency can certainly be considered and researched as an administrative, organizational and political norm, it can also be viewed as a form of governance in itself. Similar arguments have been made by Fung et al. (2007) in their work on ‘targeted transparency’, by Majone (1997) who talks about ‘regulation by information’, and by Florini (2003) in her work on ‘regulation by revelation’. But we still know relatively little about the concrete ways in which socio-political phenomena are governed by – and in the name of – transparency.
To explore alternative paths for conceptualizations of the role of transparency in processes of governance, this article elaborates on insights from Foucauldian governmentality studies and actor-network theory (ANT). Such approaches focus on ‘modes of governance irreducible to state, sovereign, or law’ (Otter, 2008: 16), and consider governance to be distributed, without a center, material and relational, and often carried out through technology and other proxies. These insights help us unpack how transparency can work as a form of disciplinary control and governmental rationality (Foucault, 1982; Burchell et al., 1991), or what I refer to as ‘ordering’. The primary value of ANT and governmentality studies is that they can be used to foreground the dynamics, circulations, and contestations of transparency efforts in the context of governance. Also, they problematize the normative starting point that transparency is an inherently valuable solution to governance problems and instead invite us to explore how transparency is constituted, mobilized and put to work in efforts to govern organizations, politics and societies.
The article first outlines what a focus on ordering implies in the context of studies of transparency and conceptualizes transparency in terms of translations and associations that condition particular forms of conduct. The second section uses illustrations from the internet domain to discuss the multiple ways in which transparency orders organizational, material and regulatory dimensions of this emergent issue area in global politics.
A focus on ordering
We may think of ordering as ways of acting on the world, that is, ‘all practices concerned with the control and management of things’ (Kendall and Wickham, 2001: 28). Clearly, much social science is driven by an interest in the ‘problem of order’ (Wrong, 1994), and often seeks to account for the structured and habitual workings of organizations, societies and technologies. But still questions about ordering (as a verb) seem to slip between the cracks of scholarly disciplines. Put bluntly, anthropologists study cultural habits and traditions, lawyers study formal regulation, sociologists study social structures and computer scientists study technological formations. While all these phenomena involve processes of ordering, they are often studied separately and cast in the analytical vocabularies of different fields of inquiry (Busch, 2011). In contrast, more flexible analytical approaches seek to remedy such disciplinary myopia by relying on a single vantage point, such as ‘standardization’ (Higgins and Larner, 2010; Busch, 2011) or ‘ordering’ (Kendall and Wickham, 2001; Law, 2003), and allow us to explore the intersections between phenomena and dynamics that are normally treated as separate.
Similarly, a basic tenet of an ordering approach is to give attention to multiplicity, incompletion and heterogeneity, and consider these to be fundamental features of social life rather than anomalies or topics to be dealt with by other disciplines. That is, ‘perhaps there is ordering, but there is certainly no order’ (Law, 1994: 1, emphases in the original). This point can be unfolded by pointing to three characteristics of ordering approaches. First, they consider ordering to be fragile because attempts to shape conduct are always contested, and ‘more or less precarious and partial accomplishments that may be overturned’ (Law, 1994: 1–2). In the context of governance and transparency, this implies a focus on ordering efforts that may be embryonic or only partially successful. Second, such approaches give primary attention to practices of ordering. That is, how are transparency projects produced and ‘worked into’ particular settings? And how do they shape organizations, policy issues or relationships between different actors by creating particular visibilities, relations or boundaries? The result of this attention to practices is a more nuanced understanding of the many workings and meanings of transparency as it shapes conduct across social and political domains. Finally, studies of ordering focus on multiplicity. In the context of studies of governance, this suggests that we need to grasp the various shapes and orientations of attempts to govern. As Law (2003: 3) puts it, ‘Discipline is…about bodies. It is about architectures. It is about time. It is about texts. It is about sight. It is about furniture. And finally, it is about the soul.’ To sum up, a conceptualization of transparency as a form of ordering moves these questions about fragility, practices, and multiplicity front and center.
Some studies of ordering build on insights emanating from the field of institutional theory (Powell and DiMaggio, 1991; Scott, 2008). Others draw on concepts from Foucault’s work on practices, technologies of government and political rationalities (Foucault, 1982; Miller and Rose, 1990). But also the Science and Technology Studies literature, particularly ANT, has contributed significantly to discussions of ordering as a starting point for studies of the entanglement of technology, social formations and regulation (Law, 1994; Bowker and Star, 1999; Latour, 2005; Hackett et al., 2008). The focus on ordering can be identified in and across a wide range of sociological disciplines. Studies of ordering may take the shape of situated, fine-grained accounts in sociology (Clarke, 2005), studies of ‘organizing’ (Czarniawska, 2008) and the ‘organization of organizations’ (Brunsson et al., 2007), as well as work on the making and governance of international spaces that draws on the literature on ‘global governmentality’ (Larner and Walters, 2004) and networked forms of governance (Barry, 2001; Kendall, 2004). When it comes to the conception of ordering developed here, the two most important pillars are the Foucauldian literature on power, governmentality and subjectification (Burchell et al., 1991; Miller and Rose, 1990; Rose, 1999), and research based on insights from ANT (Law, 1994, 2003).
While there are multiple differences between these research traditions, they share a focus on conceptualizing social phenomena in terms of associations and relations, rather than treating them as static entities (Emirbayer, 1997). That is, in these and related traditions, what we normally refer to as ‘society’, ‘the economy’ or ‘transparency’ are understood as accomplished, normalized and contested assemblages of heterogeneous elements that set conditions for human conduct (Miller and Rose, 1990; Latour, 2005; Flyverbom, 2011). Thus, governmentality studies and ANT can be understood as complementary approaches with a focus on how ordering occurs.
Actor-network theory
Emerging out of Science and Technology Studies as well as critiques of mainstream sociology, actor-network theory (ANT) has challenged taken-for-granted categories and distinctions in research such as the social and the technical, the human and the non-human. With ANT, the analytical goal is rather to capture how associations between such entities are created and made durable, as well as the practical ways in which human and non-human elements become linked (Latour, 2005) through ‘translations’ and ‘associations’ that stabilize meanings and relations in momentary configurations (Callon, 1986; Latour, 1986). Translations are central to ordering because they constitute moments where ‘the identity of actors, the possibility of action and the margins of maneuver are negotiated and delimited’ (Callon, 1986: 203). ANT accounts refrain from treating, for instance, an organization as an entity with a set of definable characteristics, and focus instead on processes of ordering – i.e. how a configuration of elements comes together and apart over time and through constant negotiations (Czarniawska, 2008). To fully appreciate the concept of ordering, we need to look at the starting point of actor-network theory, namely, to replace ‘sociologies of the social’ with a ‘sociology of associations’, because the former ‘have simply confused what they should explain with the explanation. They begin with society or other social aggregates, whereas one should end with them’ (Latour, 2005: 8). In contrast to such ‘sociologies of the social’, which consider the world to consist of structures, layers and actors, and rely on ‘society’ or ‘culture’ as explanations, a ‘sociology of associations’ stresses that ‘modern societies cannot be described without recognizing them as having a fibrous, thread-like, wiry, stringy, ropy, capillary character that is never captured by the notions of levels, layers, territories, spheres, categories, structure, systems’ (Latour, 1997: 2). Instead ANT proposes that we consider subjects, order or power as possible outcomes of associations. So ‘connections create actors, not the other way around’ (Czarniawska, 2008: 20–1) and power is ‘something that has to be obtained by enrolling many actors’ (Latour, 1986: 271). Another important contribution of ANT to conceptions of ordering lies in the acceptance of multiplicity and instability as natural conditions, which implies that ordering is always precarious and in-the-making. The attention to translations and associations is central to the conceptualization of transparency and ordering advanced in this article.
Foucauldian governmentality studies
Studies of governmentality focus on the subtle operations of steering, power and discipline. Such approaches employ and extend Foucault’s concerns with ‘conditions of possibility’, ‘power/knowledge’ and the practical ‘art of government’ (Burchell et al., 1991). The better-known Foucauldian works on the normalization and institutionalization of sexuality and madness hint at the usefulness of this literature for studies of transparency. All social settings and moments in time are characterized by particular world-views, aspirations and regularized practices, and governmentality studies seek to show how these constitute a horizon of possibilities that shape and guide our actions, i.e. what Foucault (1982: 341) termed the ‘conduct of conduct’. By singling out for analysis the normalizations that particular forms of steering rely on – such as the one attained by transparency in our times – governmentality approaches study how ways of acting and thinking emerge as taken-for-granted ideas with consequences for the objects and subjects involved. Also, work in this area focuses on the ability to ‘govern at a distance’ and reflect on the ‘governmental’ effects of such efforts for political subjects. The ambition of the governmentality literature – to account for the origins and normalization of regimes involved in the steering of human conduct and ponder on their disciplinary effects – is central to a conception of transparency as a particular form of ordering.
There are important differences between ANT and governmentality studies. Where ANT relies on concepts such as translations and associations, governmentality studies focus on normalized configurations of technologies and rationalities at work in attempts to shape conduct. While ANT has a stronger focus on materiality and the making of associations, the governmentality perspective gives more attention to the role of language in the formation of problematizations, political rationalities and normalizations beyond the situated. But they converge around the question of how order(ing) is accomplished and the focus on stabilizations of meanings and associations. These theoretical insights underpin the conception of transparency and its ordering capacities and suggest how we may study the ways in which ideals and practices related to transparency condition and shape the ‘conduct of conduct’.
Transparency as form of ordering
These insights from ANT and governmentality studies help us to approach transparency as a matter of ordering. While concepts like management, control and governance may be better known in the literature, the concept of ordering allows us to understand transparency and its role in the shaping of socio-political domains in new ways. Transparency is best understood as an ambiguous, partial and impermanent ‘script’ (Sahlin-Andersson, 1996) that is circulated, edited and translated. By implication, transparency is not a unified project or established set of guidelines, but rather a matter of interpretation, editing and association in concrete settings. The central question is how particular actors make sense of, negotiate and engage with transparency ideals and practices in attempts to shape organizational, technical, and political formations.
Discussions about transparency often start from a normative ambition to increase the amount of transparency in a given domain, as when Weber (2008) calls for more transparency in the area of internet governance or when the organization Transparency International links an increase in transparency directly to better governance. However, Heald (2006) has offered a more useful starting point for understanding these complex relationships. Instead of a focus on the quantity of transparency, we can conceive of transparency in terms of its ‘directions’ and ‘varieties’. Transparency can either be directed downwards, meaning that the superiors can observe the conduct of subordinates below, or upwards, so the subordinated can observe their superiors. In horizontal terms, transparency may be directed outwards or inwards, as when an agent can observe what is happening outside the organization, or when those outside can observe what is going on inside organizations. This conceptualization not only suggests that information flows in different directions, but also that transparency can be symmetrical, so that all parties in principle have access to the same amounts of information about each other (Heald, 2006: 26–9). But to appreciate the workings of transparency as a form of ordering, we need an even more elaborate understanding of the dynamics involved. Transparency revolves around the production and management of visibilities, and these efforts are important features of power and governance. As Otter (2008: 1) points out: ‘Who [can] see what, whom, when, where, and how…remains an integral dimension of the everyday operation and experience of power.’ Along the same lines, Brighenti (2010: 148) reminds us that: ‘The management of visibilities lies at the core of all forms of social control, whether formal or informal. More precisely…control consists of a purposeful and contextual asymmetrisation and hierarchisation of visibilities.’ In this perspective, transparency efforts can be understood as attempts to act on the world by managing possibilities for seeing, knowing and governing. That is, transparency always involves decisions about what to disclose and to whom, but also questions about flows of information and directions of visibility in concrete settings and projects. By differentiating between ‘upwards’, ‘downwards’, ‘inwards’ and ‘outwards’ forms of transparency, and by adding that these forms of transparency are ultimately about the management of visibilities, we are able to see that transparency always involves choices and divisions – who can observe whom, which activities are opened up and kept closed, and which objects and processes are subjected to transparency standards and which are not? If transparency is always managed and multi-directional, we need to conceptualize it in ways that are more attuned to these dynamics. To this end, the concept of translations is central.
Translations and associations
If we accept the argument that transparency does not have a single essence or a direct relation to governance, then we need to account for the ways in which transparency is given meaning and emerges as a resource for attempts to govern. By conceptualizing transparency in terms of translations and associations, we get a better grasp of processes of meaning-making that are often overlooked in the transparency literature, just as we can get a glimpse of how transparency ideals and practices tend to intersect with alternative, even conflicting, ideals about how to govern and organize in particular domains. Fuzzy transparency ideals circulate widely, but can be picked up and put to work in specific settings. Here, they both get entangled with other concerns and may be used to position organizations as participatory or innovative, or energize particular policy agendas or aspirations. These interpretations, enactments and entanglements are what the focus on translations and associations helps us account for. Transparency is ordered through translations and associations, but is also itself a force in the ordering of organizational and political activity.
Ordering by transparency
According to ANT and governmentality studies, translations and normalizations are what produce ordering. As Law (2003: 5) puts it, ordering occurs by ‘delegation into more durable materials’, such as when fuzzy transparency ideals are turned into organizational arrangements and political projects and other normalizations. As we will see below, transparency ideals shape not only organizational and procedural arrangements in the digital domain, but also the development of products and services, as well as corporate advocacy and policy deliberations. Across these very different dimensions of the internet domain, transparency ideals are translated into more durable normalizations, e.g. a set of principles for managing organizations, standards for user control over products, or templates for socio-political re-engineering. These processes of translation are important because they ensure ordering, that is, they guide employees’ and stakeholders’ conduct, they create (a)symmetries of visibility so that only selected parts of organizational processes are made visible, and they recast policy issues and ethical concerns in new forms, for instance, when political or ethical problems are re-articulated as techno-administrative matters.
Ordering by transparency in the internet domain
Cyberspace constitutes an exemplary site to flesh out how translations of transparency contribute to ordering. The governance of the internet is an emergent and increasingly visible and salient issue area involving both governmental and non-governmental actors. Whether in the shape of Chinese filtering efforts, Egyptian decisions to block the internet, United Nations efforts to develop international principles or public protests against trade agreements like SOPA, PIPA and ACTA, questions about the governance of the internet have entered public discourse and global politics. As shown by Lessig’s (1999) influential work and scholarship on internet governance (Mueller, 2002, 2010; Goldsmith and Wu, 2006; DeNardis, 2009; Flyverbom, 2011), it is increasingly clear that the internet is not a regulatory void, but is subject to many forms of regulation, such as laws, norms, markets, and codes, and often in surprising configurations. However, it remains unclear how cyberspace is governed across particular organizational settings or policy issues, and how ideals like transparency shape this emergent issue area.
The quest for transparency figures prominently in current discussions about internet governance and the internet industry. For instance, there are growing demands that policy and negotiation processes must be accessible and accountable to a wide array of stakeholders. In the corporate sector, internet giants like Google and Facebook stress their commitment to principles of transparency, and we can understand transparency as a foundational ideal that many internet companies subscribe to and seek to advance. 1 At the same time, users and critical voices in media and government call for increased transparency when it comes to the governance of the internet, and articulate concerns about data aggregation, commercial uses of digital traces, and the lack of user control and privacy. Finally, whistleblowing and leakages, such as the work of WikiLeaks and Edward Snowden, and professional transparency organizations, like Transparency International and the Sunlight Foundation, contribute to the circulation of transparency ideals in the internet domain and beyond. In the following I analyze how transparency is translated and normalized in ways that contribute to particular forms of ordering in cyberspace, namely, in relation to: (1) organizational procedures; (2) designs of product and services; and (3) advocacy efforts. Prior and ongoing empirical research carried out by the author suggests that transparency ideals play a central role in these areas, and points to the diversity and dynamics of transparency as a form of ordering (Flyverbom, 2011).
Transparency as organizational ordering
Both UN-driven, inter-governmental initiatives to address internet governance and corporations like Google and Facebook have an explicit focus on transparency. The translation of this ideal into organizational and procedural principles is an important feature of the internet domain, and one way in which transparency has ordering effects.
In response to widespread concerns about the opaque and inaccessible nature of internet governance, a number of UN-facilitated working groups and so-called multi-stakeholder organizations have emerged. In these contexts, broad transparency ideals are translated into specific procedures and information-sharing practices that allow for participation, particularly by technical groups and other non-state actors that normally play a marginal role in an intergovernmental institution like the UN. Calls for transparency come not only from governments concerned about the fact that many parts of the internet infrastructure are controlled by technical and private organizations with little oversight. Also, the technical and private organizations currently in charge of internet governance voice fears about the negative effects of governmental and intergovernmental interventions, and call for transparency when it comes to intergovernmental attempts to govern the internet (Chadwick, 2006; DeNardis, 2009; Mueller, 2010; Flyverbom, 2011). Responding to these tensions, the UN-initiated Internet Governance Forum (IGF) has become a pillar in attempts to facilitate dialogues between stakeholders in this domain. The yearly national and global meetings of the IGF can be seen as ambitious attempts to translate transparency ideals into regularized organizational and procedural arrangements. Such efforts include virtual attendance via web-conferencing software, transcripts from meetings, and attempts to engage an increasing number of stakeholders in dialogues, collaborative sessions and other procedural experiments that rarely occur in other UN processes. In these processes of translation, transparency is connected to a range of existing practices, and such associations are, as I have argued, the foundation of ordering. In the case of UN internet governance arrangements, the ordering effects take a number of forms. For instance, once transparency has been translated and normalized in organizational settings, such procedural experiments can be used to help position IGF as innovative and attractive, particularly in contrast to established conceptions of the UN system as bureaucratic and slow-moving. In this respect, transparency contributes to ordering by positioning organizations and marking differences between different types of organizations. Ordering effects also take the shape of particular ways of managing visibilities: these efforts correspond to what Heald (2006) calls inwards transparency, such as opportunities for outsiders to look into and for stakeholders to participate in policy-shaping activities. Finally, translations of transparency ideals into procedural practices contribute to the ordering of internet governance as an important socio-political issue area in global politics that a wide range of actors have a stake in and are able to voice their opinions about (Flyverbom, 2011).
Transparency ideals are also translated and made durable by private internet companies. Both Google and Facebook claim to be transparent organizations, describe their work as focused on making information available and accessible, and strive to contribute to transparency and openness in societies and politics. Thus, they subscribe to and push for a ‘radical social premise – that an inevitable enveloping transparency will overtake modern life’ (Kirkpatrick, 2010: 200). In these contexts, transparency is translated into procedural and organizational arrangements, such as architecture, information-sharing and opportunities for employee participation. As one employee explains: ‘Transparency has been of core importance to Facebook since its earliest days. There are no private offices or cubicles. We tore down those unnecessary walls so that everyone could sit out in the open with their teams. Every conference room features a glass wall or panel so that you can quickly see what’s going on inside’ (interview, Facebook, 2012). Similarly, Google invests considerable resources in facilitating a ‘culture of openness’. In both companies, every Friday afternoon revolves around so-called TGIFs with founders and top-level management, and the opportunity to ask questions about current developments, strategies and challenges in the company. While other companies also celebrate the coming of Friday, Google and Facebook explicitly link these meetings to their ambition to remain organizations marked by transparency, voice and openness (Google newsletter, 2011; interview, Facebook, 2012). Such organizational practices associate transparency with a particular style of management and show how ordering by transparency operates in organizational settings.
These translations of transparency into organizational culture and aspirations for novel forms of management, however, do not occur without friction. Transparency is always entangled with other concerns. For instance, one does not get past the reception desk in either company without an invitation and a signed non-disclosure agreement stating that no confidential information obtained on the premises may be disclosed. While non-disclosure agreements are widespread in such companies, they take on new meanings in the context of an explicit commitment to transparency. As noted, transparency rarely means full disclosure or openness, so the management of visibilities (Brighenti, 2010) – what to make transparent and highlight, and what to keep opaque and out of sight – becomes a central issue. Such translations and the sorts of visibility they produce point to the ordering effects of transparency. In the cases of Google and Facebook, there is an unstated but important division between those who can see into the workings of the companies, and those who are left to marvel at a distance. In Heald’s (2006) terms, we can think of this as strong forms of vertical transparency, where employees and employers can observe each other very easily, but also as very little horizontal transparency because outsiders have very few opportunities to scrutinize the insides of these companies. Thus, transparency orders by creating a hierarchy and (a)symmetry of visibility in and around organizations. Such decisions about what to make transparent and what to keep opaque can be understood as very strategic attempts at positioning organizations vis-à-vis others. Within organizations, vertical transparency may enthuse and motivate employees, and give them a sense of working in a superior type of company. At the same time, the lack of vertical, particularly inwards, transparency contributes to the creation of organizational mythologies about innovation – enticing outsiders to think that the companies may be building the ‘next big thing’. The careful attention to managed visibilities and their ordering effects is central to the conceptualization of transparency that I seek to advance.
Transparency as material ordering
In the internet domain, transparency ideals are also translated into decisions about design features and other material arrangements. In the face of widespread concerns about the uses of personal data for purposes of profiling and marketing, for example, internet companies increasingly seek to address privacy questions at the level of design and services. Thus, rather than a regulatory issue to be solved by policy-makers, privacy is cast as a matter of transparency and ‘user control’. For instance, Google and Facebook provide ways for users to see and alter the information and data that the companies can store and reuse, such as various opt-out mechanisms, privacy settings and ‘dashboard’ functions giving users an overview of their subscriptions, history and relations with the companies. Google’s Privacy Director explained this at a Senate Committee hearing in 2010: ‘As we work to bring more relevant ads to our users, we continually seek to preserve transparency and user control over the information used in our ad system’ (Whitten, 2010). To support and institutionalize such features across Google’s various services and products, the company has created the so-called Data Liberation Front, a team of engineers tasked with making it easier for users to move data in and out of Google products and services (interviews, Google, 2011). Similarly, Facebook seeks to improve user control over data, sharing and ‘digital histories’ (interview, Facebook, 2012). Finally, transparency ideals are also translated into relationships between users and the services provided by the two companies, for instance, in the requirement that everyone uses their real identity when signing up. This demand for authenticity is described as a way to increase security and trust online, but clearly also ties in with need to have reliable data to mine and aggregate for purposes of profiling users and targeting advertisements. Such attempts to balance user trust and the data aggregation that makes these companies’ business models possible are emblematic of the translation of fuzzy transparency commitments into material arrangements in this area.
In summary, product design is a second area in which transparency ideals are translated and normalized, and this material dimension contributes to the unpacking of how ordering by transparency works. Material forms of transparency seek to increase a sense of symmetry when it comes to visibility and insight, so that users do not only feel watched and have their data sourced and aggregated, but are given opportunities to control the digital traces they leave. But these efforts do not only seek to create information symmetries. The engagement with transparency ideals contributes to the ordering of the internet domain by recasting privacy concerns and policies in terms of user controls and technical features. This also implies an ordering of roles and responsibilities, whereby data management is cast as an individual responsibility, rather than a question of regulation and politics (see also Birchall, this issue). These illustrations point to some of the ways in which transparency gets built into material infrastructures that come to shape the ‘conduct of conduct’. By offering ‘user control’ and a sense of vertical transparency, they shape users’ engagement with technology and data management. Such orderings not only condition particular forms of behavior, but also contribute to attempts at positioning the internet domain as a regulatory space marked by individual responsibilities, technical rather than political solutions, and a strategic reliance on transparency measures that do not undermine the business models and interests of internet companies. These translations and the associations to material, regulatory and individual dimensions they create point to the ordering effects of transparency.
Transparency as political and societal ordering
Transparency ideals are not only translated and normalized in the context of organizational arrangements and product designs, but also in attempts to shape the world around organizations in the internet domain. IGF, Google and Facebook seek to advance transparency as a form of political and societal ordering, and engage calls for transparency in their attempts to do advocacy and enhance mobilization and participation around the politics of the internet.
The two companies stress their vision to support the free flow of information and advance transparency in and beyond the internet domain. As the Policy Director of Google points out, increased transparency ‘would be valuable for us, if more made information available and were transparent; it would be easier for us to realize our mission, and this goes for governments also’ (interview, Google, 2011). The vision of societal transparency ties in with the Obama Administration’s notion of ‘open government’, and Google’s various activities to strengthen what they term ‘open data’ are central when it comes to advocating transparency. The idea of transparency as a societal and political aspiration also serves as the foundation of more elaborate initiatives, such as Google’s Transparency Reports. These reports disclose how and when governments around the world make requests for user data and other types of information, when and where the internet has been blocked, and when outages have happened or particular services have been made inaccessible by governments, for instance, when Egypt turned off the internet in 2011 (Gazzar et al., 2011). These examples show how transparency ideals are translated into advocacy projects that fit the aspirations of Google when it comes to acting on the world and positioning itself as a key player in societal transformations. In the same way, Facebook stresses the role it has played as a platform for mobilization and advocacy in relation to the Arab Spring (interview, Facebook, 2012). Such outwards-directed forms of transparency must also be understood in relation to other concerns they become entangled with. Most importantly, attempts to order societies and politics in the name of transparency occur against the backdrop of a growing discomfort with the surveillance and tracking made possible by digital technologies, and can be seen as attempts to shift the attention away from negative portrayals of the nature and effects of the growing importance of the internet as a societal infrastructure and foreground a socio-political vision focused on the benefits of transparency. Once again, we see how transparency gets entangled with a suite of concerns and aspirations and operates as a form of ordering that can be used to recast contested issues in more attractive ways.
In the context of UN activities to address the issue of global internet governance, we see similar uses of transparency as a way to remedy the widespread discomfort with the lack of transparency and possibilities for meaningful participation in the organizations managing this infrastructure. For instance, IGF translates transparency into procedural arrangements by casting global internet governance as a ‘multi-stakeholder’ process and contributing to the pressure on institutional and governmental bodies that do not live up to this ideal. While such ideals are not entirely new in the internet domain, they have certainly gained traction beyond IGF, and a wide range of organizations involved in internet governance now consistently describe themselves as multi-stakeholder forums and increasingly disclose information about their meetings and activities (Flyverbom, 2011).
In contrast to the previously mentioned workings of transparency, these activities involve outwards transparency flows, where organizational approaches and aspirations are transposed from organizations onto the world outside. Such efforts have multiple ordering effects. Transparency Reports not only compile and circulate information about the internet infrastructure, but also influence policy-making in this area by naming and shaming restrictive governments. Furthermore, they constitute one pillar in Google’s ambition to shape policy-making through data provision – to make regulation and politics more data-driven and fact-based and promote what is termed ‘policy by numbers’ rather than ‘policy by emotions’ (interviews, Google, 2011). Transparency Reports are thus intended as ways to show policy-makers the consequences of legal and regulatory initiatives, and to ‘illuminate possible futures’ (interviews, Google, 2011). By offering information about socio-political phenomena, such efforts contribute to the production of geo-political scenarios and to the positioning of transparency as a key component of ‘good governance’. Furthermore, we can think of such forms of what may be termed ‘transparency evangelism’ as attempts to position organizations as political players in the shaping of internet governance and the future of the internet. In similar ways, the translations of transparency into participatory stakeholder models exemplify the ambition to advance transparency outside the realm of individual organizations – that is, to shape the world around them in the name of transparency. In this context, transparency is translated into recipes for the ordering of societies and politics along the lines of information-sharing, participatory processes and a reliance on technologies that allow for new ways of governance through data (see also Birchall and Hansen, this issue).
Conclusion
This article proposes that transparency is best understood as a matter of translating, establishing associations and managing visibilities in ways that contribute to the ordering of specific organizational efforts and regulatory concerns. The conceptualization of transparency as a form of ordering and illustrations from the internet domain support the argument that we need to push beyond questions about amounts or degrees of transparency, i.e. how much clarity exists or can be gained. This approach also suggests that it may be useful to momentarily bracket normative questions about the value or positive effects of transparency and shift the attention to empirical and conceptual investigations of the limits and potentials of transparency. This implies a stronger focus on the ways in which transparency ideals are transformed into more durable and concrete ordering efforts, such as managing organizations, designing products and shaping socio-political developments. Transparency, I have argued, is not only a matter of providing insight and clarity, but also a matter of managing visibilities in ways that contribute to organizational control and societal governance.
The article seeks to contribute to studies of transparency in two ways: The first is to articulate what happens when transparency ideals are connected to organizational and political realities. Using insights from ANT and governmentality studies, the article has focused on translations that turn fuzzy transparency ideals into more tangible, operational and material concerns, and has pointed out how ordering occurs by such ‘delegation into more durable materials’ (Law, 2003: 5). That is, transparency shapes the ‘conduct of conduct’ when such ideals become intimately connected to organizational realities and political aspirations. Examples from cyberspace have shown us how transparency takes on multiple meanings in a concrete domain. Here, translation processes involve entanglements with different, often competing aspirations, like the value of secrecy in relation to organizational mythologies, and growing concerns over surveillance and the end of privacy in times of ubiquitous ‘datafication’ (Mayer-Schönberger and Cukier, 2013). While these illustrations are not exhaustive of the internet domain as a whole, they do point to some dynamics of transparency and their ordering effects, and give texture to the conceptualization of transparency.
The second contribution is the creation of a linkage between studies of transparency and the concept of ordering. The focus on ordering opens up questions about the role of transparency in the shaping of conduct. Transparency ideals create particular visibilities and possibilities for observation, i.e. make parts of organizational and socio-political life knowable and governable while keeping others inaccessible and out of sight. The ordering produced by translations and normalizations of transparency involves visibilities that produce particular hierarchies and (a)symmetries of seeing, knowing and governing. For instance, the two internet companies focus more on making the world knowable and legible than advancing transparency about what goes on inside their organizations. Carefully managed transparency efforts also affect the ways in which organizations are positioned, and may be used to create enthusiasm about the workplace, to cast an organization as more agile and inclusive than others, or to entice outsiders and build excitement about innovations. Such ordering effects point to the ways in which transparency is mobilized in organizational and other settings. There is more to transparency than information provision for purposes of accountability. The ordering effects of transparency also take the shape of more wide-reaching attempts to frame policy issues and regulatory challenges in new ways. As we have seen, regulatory issues like privacy can be recast as technical matters of user control, and transparency efforts may be used to energize calls for new forms of global governance in an issue area like the global politics of the internet. These reconfigurations operating in the name of transparency may best be understood in terms of what Garsten and Jacobsson (2013) refer to as ‘post-political forms of governance’. Such forms of governance re-articulate politics and conflicts using other registers, such as when heated political controversies are translated into administrative or technological matters. In the internet domain, transparency projects are part and parcel of such efforts to shape politics and regulation through re-articulations that suppress conflicts or shift roles and responsibilities in subtle but significant ways. The focus on dynamics of transparency, the management of visibilities and ordering thus contributes to attempts to make sense of governance arrangements and forms of governance in scattered and emergent issue areas in global politics, such as cyberspace.
The trust in ‘sunlight’ as a ‘disinfectant’ is certainly not limited to the internet domain, but is a widespread development in many areas of regulation and politics (Braithwaite and Drahos, 2000). In fact, it underpins many business commitments to corporate social responsibility and drives multiple forms of activism. These aspirations operating in the name of transparency deserve further scrutiny, and may help us understand new facets of the entanglements and visibilities produced by transparency ideals and practices more broadly.