Analysing Internet Policy As a Field of Struggle

This essay proposes an analytical approach that conceptualises internet policy as a field of struggle which emerges through processes of discursive institutionalisation. By combining field theory and selected Science and Technology Studies (STS) concepts, the essay highlights the performative function of discourses in the field. It does so by showing how actors entered the policy field through the creation of expertise and regulatory competences and uncovers key conflicts that have shaped internet policy. Drawing on interviews and document analysis, the essay illustrates the proposed research approach via three selected examples which demonstrate how international discourses materialised in internet-related organisational structures and regulatory competences in German ministries.


INTRODUCTION
Over the last decade, it has become increasingly obvious that the internet and digital technology are changing modern societies in fundamental ways. They are undergoing an open-ended process of transformation that is gradually affecting all areas of social life, calling into question societies' normative and institutional underpinnings. On the national and international levels, public authorities started to respond to the new political challenges by developing new expertise and competences to understand the technical change and to adapt their regulatory repertoire.
Likewise, in most parts of the world, the digital transformation results in a dynamic public discourse and an expanding network of non-governmental actors seeking to shape and assess the socio-technical changes and discuss corresponding policy options.
This short research essay proposes to analyse these new constellations of actors, issues and policies as an emerging policy field related to the internet. Rather than assessing them as loosely connected elements that form a fragmented mosaic, the policy field perspective allows us to view actors, issues, discourses, policies and regulatory competences that emerge around internet issues as interrelated. But what is the relation between the various actors in the field and how do these actors, through their interactions, link issues with institutional structures, expertise and regulatory responsibilities? Which are the discursive and institutional processes that contribute to the emergence and shape of the policy field? In order to answer these research questions, the essay proposes an analytical approach that draws on different relational and constructivist theories from the social sciences. Since processes in the field are often conflict-laden, this approach is based on the conceptualisation of policy fields as fields of struggle and aims to retrace how conflicting discourses and power struggles materialise in form of the institutional and regulatory structures of the policy field.
By combining field theory with conceptual tools from discourse and institutional theory as well as Science and Technology Studies (STS), the proposed qualitative and historical approach builds on and simultaneously informs research interested in the emergence of policy fields (Bergemann et al., 2016;Bernhard, 2011;Haunss, 2015;Knoke, 2004;Lynggaard, 2007;Massey & Huitema, 2013)1, the role of discourse for institutional change (Schmidt, 2008; hand, the ever growing amount of internet policy related documents, institutions and decisionmaking structures indicates that public authorities and non-state actors persistently seek to create competences and influence the new societal and technical developments. As a result, we can currently witness the emergence of a multifaceted ensemble of policies which, thanks to the shared reference to the internet, link issues that previously were not interrelated. This ensemble also involves a relatively fixed number of individual and collective actors and institutions which are interconnected through their 'common concerns' (Lynggaard, 2007, p.293), that is, their shared interest in internet-related questions. Collectively these policies, issues, actors and institutions form an emergent policy field related to the internet.
Although still in emergence, the internet policy field has already acquired a high degree of complexity and diversity. This is not least due to the tension between the inherently global nature of the internet infrastructure and the attempts to regulate public policy aspects on the national or regional level (Johnson & Post, 1996). Since the 1990s, the technical coordination of the global internet infrastructures has become gradually institutionalised in transnational settings that often apply a multi-stakeholder governance approach, such as the Internet Corporation for Assigned Names and Numbers (ICANN) (Mueller, 2010). In contrast, the regulation of public policy issues related to the internet-such as data protection, competition, security, access and content control-is primarily coordinated and implemented by national or regional authorities (Drake, Cerf, & Kleinwächter, 2016, pp.31ff).3 Yet internet policies and related responsibilities differ quite significantly from one country to another. Some countries, such as the United States, tend to leave internet rule-making primarily to the 'free market', while others, such as France, assign regulation more strictly to public authorities. While many developing countries still focus on basic access problems, others' policy approaches are shaped by discourses on cyber threats, counter-terrorism or child protection, for instance in the UK. Moreover, countries put forward different governance instruments, such as national commissions (e.g. the German Parliamentary Enquete Commission on Internet and Society) or charters of internet principles (e.g. the Brazilian Civil Rights Framework for the Internet).4 As a result, national differences regarding internet policymaking appear to be strongly influenced by the concrete constellations of actors, existing and newly created institutional structures and the influence of dominant discourses. To analyse the emergent internet policy field of one particular country or on the global level, we propose to study the interrelation of these different elements constituting a social field. combines the field approach with the concept of discursive institutionalisation.

THE INSTITUTIONALISATION OF DISCOURSE IN THE POLICY FIELD AND ITS MATERIALITY
To identify the struggles that shape a policy field in an observable way, it is necessary to understand how actors build links between themselves and the policies, perceptions, expertise and regulatory competences that, through their interrelation, form the field and its structures.
For this purpose, the analytical approach proposed in this essay combines the meso-level analysis inspired by field-theoretical approaches with a micro-level analysis focussing on discourses and their role for the emergence of the policy field. Inspired by interpretative approaches that emphasise the importance of discourse for policymaking and institution building, we understand discourses as the ensemble of ideas, narratives and definitions that attribute meaning to objects and phenomena and, by doing so, create and reproduce a certain worldview.11 Hence, discourses are not simply the result of meaning making and an expression of problem perceptions, they also actively influence them due to their productive and reproductive function (Milliken, 1999, p. 235).
The performative effect of a discourse is particularly strong if it affects the formal and informal institutions that define the context in and modes through which actors interact in a field (Schmidt & Radaelli, 2004, p. 192).12 Institutions serve as external structures that set the rules of the field. They are constructed by and, at the same time, structure the discourses and interactions in the field.13 But not all discourses are eventually institutionalised and develop a performative effect on the field and its actors (Phillips et al., 2004, p.638). In a policy field as a contested space, the process of discourse institutionalisation involves struggles between competing actors and their discourses. As our empirical research will show, in the end, it is either one of the competing discourses or the conflict itself that is inscribed in and, thereby, (re)shapes the policy field.
Thus, to better understand how actors are linked to institutions through discourses, it is necessary to analyse which discourses become institutionalised and, thereby, produce and change the institutional structure of the field: 'If a discourse solidifies in particular institutional arrangements [...] then we speak of discourse institutionalization' (Hajer, 2005, p.303). In order to retrace, on the micro-level, these moments of discursive institutionalisation and assess their performative effects on the actors, perceptions and practices on the field-level, we borrow the concept of 'inscription' from STS.14 There, the term refers to a process through which engineers, inventors and designers embed their visions, ideas and discourses in the material structure of objects or technical artefacts (Akrich, 1994), like the internet's technical designers inscribed the principles of openness and de-centrality into the materiality of the network's structure.
Similarly, these visions can also be inscribed into institutional structures that constitute the material foundations of a field, such as policy programmes, organisational units and regulatory competences, and are equally artefacts as they are produced and shaped by the field's actors. By inscribing discourses into the materiality of the policy field, actors link issues with perceptions, institutions and competences and, that way, fundamentally add to the processes of social ordering in the field.15 Through their inscription, discourses not only become part of the material world but also of actors' social practices; as a consequence, they are conducive to their own reproduction: Objectified beliefs often become embedded in routines, forms, and documents, e.g. the types of classifications employed, and artifacts -tools, hardware, and machinery. We organize our material world in accordance with our mental categories, and the two become self-reinforcing. (Scott, 2014, p.149) In this sense, materialisations are not pure representations of reality, they also transform and reproduce the discourses and conflicts embedded in them. They have a performative character and contribute to the exclusion of alternative worldviews. Accordingly, the inscription of discourses into documents, technologies or organisational structures can be seen as ' […] enactments of reality; they are means by which some things are made present and others absent, so that specific ontologies are performed into being and others made invisible' (Nimmo, 2011, p.114). Moreover, through their inscription, discourses reach a more stable form and develop more permanent effects on actors, their perceptions and interactions. That way, discursive institutionalisations contribute to the development of a field-specific logic (Bourdieu, 1996, pp.227f;Thornton, Ocasio, & Lounsbury, 2012, p.148), just as the visions inscribed in the internet's technical characteristics determine its inherent logic.

RETRACING THE INSTITUTIONALISATION OF DISCOURSES AND COMPETENCES IN THE GERMAN INTERNET POLICY FIELD
As outlined above, policy fields contain a heterogeneous population of state actors (i.e. political parties, governments, ministries) and non-state actors (i.e. NGOs, private companies). Despite this wide range of actors, this essay takes a closer look at selected processes of institutionalisation of discourses and competences within German ministries, which for two main reasons are particularly useful for illustrating our analytical ideas.
First, federal and state ministries in Germany are important sites of production for political programmes, professional expertise and regulatory competences (Derlien, 1995, p.80).
Organised according to the principle of departmental ministers (Ressortprinzip), ministries both shape and reflect particular perceptions of problems and the logics in the policy field for which the ministry is responsible. This is not only the case for a ministry as a whole (e.g. the Ministry of the Interior following the logics of security) but also for the different divisions within the ministries (e.g. the IT Division following a distinct logic of IT security). Second, although ministries produce discourses that shape the policy field through their competences and programmes, they are simultaneously responsive to public, professional and political debates.
Thus, we often see that a certain discourse is taken up by a ministry, where it materialises and solidifies into organisational structures and regulatory competencies. It is through both these functions-the production of discourses and their inscription-that ministries link issues with competences and institutional structures and, hence, contribute to the emergence and shape of a policy field.
To develop a historical perspective on the emergence of the internet policy field in Germany, the empirical examples in this essay illustrate instances of institutionalisation of international discourses in ministerial structures and competences. They took place at different moments in time and in two different ministries, namely the Federal Ministry of the Interior and the Federal Ministry for Economic Affairs.16 Both of them are in charge of implementing the German Digital Agenda, a strategic policy programme adopted by the current government coalition in August 2014.17 To retrace the moments in which conflicts and discourses materialised in these two ministries, we draw on organisational charts and semi-structured interviews with senior officials of federal ministries. The analysis of organisational charts has allowed us to understand how ministries take up and frame new issues and structurally integrate them in light of their specific policy traditions and remit.18 The qualitative interviews help us to detect narratives that explain the findings gained from the organisational charts in more detail.19 Each empirical example addresses in a first step the inscription of a discourse, followed by the creation of material structures and their contribution to social ordering and, lastly, the inscription's further implications for the policy field.

(A) THE INSTITUTIONALISATION OF THE INFORMATION SOCIETY DISCOURSE
Like most other countries, Germany began to develop a political response to the internet and its societal impact in the early 1990s, when the discussion on the information society started to unfold on both the national and international level. The concept of 'information society' was commonly used as a metaphor to capture the importance of information for economic and societal progress. Accordingly, many countries and international organisations started to adopt information society policies designed to pave the way for the transition towards an information economy driven by digital technology.20 Despite their similarity, the various initiatives differed in their nuances, revealing a key conflict inherent in the policy debate on the information society-a conflict that has accompanied internet policy discussions ever since. It consisted in the struggle between those who wished to accomplish the transition towards an information society through market liberalisation and self-regulation of the private sector and those who argued in favour of a more regulatory approach, which would strengthen the role of public authorities in planning and coordinating this transition (see also Kubicek, Dutton, & Williams, 1997;Moore, 1998).
The German government perceived the transition to the information society as a matter of necessity resulting from international pressures and from what the government considered an inevitable external development (Thorein, 1997, pp. 69f). Therefore, it decided to build up regulatory capacities to shape the transition, which led to the institutionalisation of the discourse on the information society. Indeed, the discourse was inscribed in the bureaucratic structure of the German public administration in form of a 'Working Group on the Information Society', created within the Federal Ministry for Economic Affairs in 1995.21 The ministry's Division for Industry Policy developed the idea for this working group in the context of an internal debate reflecting the core conflict related to this early period of internet policies: as interviewees pointed out, there was a sense of anxiety in the ministry that too much regulation might be a barrier to the expected positive effects of the information society. Hence, many traditionalist actors in the ministry claimed that the existing competition policy was a sufficient regulatory frame for the digital age, whereas others argued in favour of a dedicated strategy for the information society in form of an industry policy.22 As the then Minister for Economic Affairs, Roman Herzog, saw both positions as equally important, there was high-level support for both sides. Thus, the constellation within the ministry was favourable for the institutionalisation of the information society discourse.
Among other tasks, the new working group was responsible for the development of an influential strategy paper on the information society, called 'Info 2000' (BMWi, 1996;Thorein, Analysing internet policy as a field of struggle Internet Policy Review | http://policyreview.info 8 July 2016 | Volume 5 | Issue 3 1997, pp.58ff)23, which institutionalised both the dominant discourse on economic progress and the core conflict between market liberalisation and a stronger regulatory approach. It expressed the idea that governance instruments like data protection and labour law should be adapted to the requirements of the information society (Thorein, 1997, p.70) and the view that laws concerning the protection of consumers, data, youth, and others are not values per se but simply means to increase the acceptance and usage of the internet (Scholz, 2004, p.71). That way, the inscription of the information society discourse in the ministry led not only to its materialisation in form of a newly created organisational structure; it also materialised in form of an important policy text and thus contributed to the ministry's and its actors' effort to shape Germany's response to the new technological developments.
The performative effect of the materialisation of the information society discourse did not remain limited to the Ministry for Economic Affairs alone. Indeed, the 'Info 2000' paper served as one of the major reference points for the 'Multimedia Act', which was adopted in 1997

(B) THE INSTITUTIONALISATION OF THE IT SECURITY DISCOURSE
Some years after the initial institutionalisation of the information society discourse within the German administration, we can observe the inscription of a second discourse that eventually shaped regulatory competences for internet policy, this time created by the Ministry of the Interior. At the turn of the millennium, two major IT-related security issues raised public attention for the risks of technological development. The first issue was related to the change of date from '99 to '00 that was feared to cause dramatic failures in IT systems (known as 'Y2K bug'), potentially affecting almost every aspect of social, economic and political life, including healthcare, financial markets and nuclear power (Special Committee 2000, 1999.24 The second security issue consisted of the computer virus named 'ILOVEYOU', a bug hidden in a fake love letter that spread via e-mail, affecting millions of computers within a few days.
The Y2K bug and the ILOVEYOU virus were not simply technical problems that had to be dealt with. Rather, they were discursive events which, in the ministry and beyond, led to a new level of awareness of how much the state's capacity to act was dependent on global IT infrastructure.
According to some interviewees, both these events were major triggers for the decision of the Thus, through the early inscription, the ministry not only linked internet and security issues in a stable way. It also institutionalised its own role and competence in the emerging policy field.
One particular decision concerning the IT Staff's design in 2002 illustrates how the way in which the security discourse was institutionalised shaped the further development of internet policy within the Ministry of the Interior. As an interviewee explained, the IT Staff's composition was guided by the idea to pool 'genuine IT topics' within a single unit. Yet, instead of regrouping all existing IT-related technical and regulatory competences, the ministerial officials decided that the IT competence of the security agencies (police, internal intelligence agencies) should remain within the Division for Public Security. This organisational setting, characterised by a clear thematic distinction between general IT issues and the IT competence of the security agencies, contributed to institutionalising a conflict in the ministry which remains central to the internet policy field until today: the conflict between the protection of IT systems, on the one hand, and the possibility for security agencies to intrude into these systems, on the other hand.
The conflict between a preventive and repressive side of cybersecurity, as it was characterised by one interviewee, is in fact a variation of what is commonly described as the 'crypto debate'.27 For more than ten years, this conflict has been the subject of ongoing internal debates between two divisions of the ministry, the Division for IT and Cyber Security and the Division for Public Security. At the same time, the controversy over the protection (e.g. by encryption) and intrusion of internet communication (e.g. through spyware) has linked bureaucrats, security officials, politicians, and activists who are jointly but antagonistically engaged in the internet policy field.28 Thus, the inscription of the IT-security discourse in the ministry not only structured its own perception of internet policy but had a long-lasting influence on the larger debate and, hence, on the processes of social ordering in the field in general. Innovation Policy, which among other tasks is in charge of 'digital sovereignty'. With this move, the ministry not only inscribed the sovereignty discourse into its structure and thereby created a material and more permanent link between the conflict embedded in the discourse and its own competences. It also influenced the perception of policy problems and the production of expertise beyond its own institution as it linked economic, academic and administrative actors

(C) THE INSTITUTIONALISATION OF THE POLITICAL DISCOURSE FOLLOWING THE SNOWDEN DISCLOSURES
and key topics of digital sovereignty by developing a policy paper in cooperation with a group of experts (BMWi, 2015).30 It thereby reinforced and shaped the debate on technological and economic dependencies in Germany.

CONCLUSION
This essay proposes an analytical approach which conceptualises internet policy as a field of struggle that emerges through processes of discursive institutionalisation. It identifies key conflicts and discourses that, thanks to their materialisation in the policy field, are able to link actors, issues and regulatory competences and, thereby, shape the field's structure in a temporary yet stable manner. For this purpose, we combine field theory and approaches focusing on discourse and institutions with STS tools, in particular the concept of 'inscription'.
While the field approach serves to emphasise linkages and structures and to address their temporal stability on the field-level, the focus on discursive institutionalisations allows us to scrutinise the material traces left by inscriptions, for instance in form of organisational structures or policy papers. Hence, the conceptual combination makes it possible to evaluate the performative effect that the identified discourses and conflicts had on the policy field as a whole.
We illustrate this conceptual approach via three selected examples that show how international discourses related to the internet materialised in organisational structures in German ministries. The discourses on the information society of the early 1990s and IT security around the turn of the millennium were not transient debates that passed without leaving any traces.
Instead, they created cognitive effects that translated into permanent structures within the issues in a solid manner to the German internet policy field, it also created links between different actors in the field, as the ministries we investigated continuously engage with other state and non-state actors in the struggle over these core conflicts.
By combining the observation of the emergent policy field with the micro-level processes of discursive inscriptions, we gain insights into the materiality of the field that links actors, issues, expertise and regulatory competences. In addition, we are able to observe how actors seek to contribute to the processes of social ordering in the field that, ultimately, impact the autonomy and stability of the newly emerging policy field. In order to paint the larger picture and look at the policy field in its entirety, the proposed analysis would need to extend to other actors involved in the field, including non-state actors. Even more importantly, it needs to address the relation between these actors in a systematic way by retracing their linkages via shared or contested discourses and responsibilities. The identification of relevant conflicts and the initial analysis of relevant actors in this essay may provide a useful starting point in this respect. In addition, since in other countries the core conflicts might differ and materialise in other institutions, our future research will also assess processes of discursive institutionalisation in different countries to uncover the many interrelated yet often competing discourses, actors, issues and regulatory responsibilities that configure the internet policy field as a whole.
scrutinising the internet policy field is STS's enlarged understanding of actors that grants agency not only to humans but also to material objects and semantics (Bijker & Law, 1994, p. 449;Latour, 2005, p. 465). This makes it possible to account for the influence of digital technology in a non-deterministic way. However, due to the space restrictions, this short essay does not allow us to elaborate on these aspects in more detail.
16. During our period of investigation  18. We used a historiographical approach for the analysis of the organisational charts that visualises the development of specific organisational units within a ministry as a chronological sequence. These timelines are based on public data available in 'Bund transparent' (1995'Bund transparent' ( -2015 and 'Staatshandbuch Bund' (1995. Besides giving important indications about the inscription of discourses and the institutionalisation of competences, the findings gained from the analysis of the charts also served to prepare the interviews.  (Graham, 1994, p.3).
23. The 'Info 2000' strategy paper was a result of a coordinated effort between five ministries, with the Federal Ministry for Economic Affairs and its working group taking the lead (Thorein, 1997, pp. 58ff). In addition, the paper was influenced by ideas developed in the context of the Petersberger Kreis and the Council for Research, Technology and Innovation (BMWi,1996), both policy forums in which the Federal Ministry for Economic Affairs was involved.
24. The global crisis scenarios around the Y2Y bug led to the establishment of temporary task forces and crisis committees at the national and supranational levels (e.g. de Borchgrave & Lanz, 1999), including in the German Ministry of the Interior.
25. German ministries are generally organised along line organisational structures. A division (Abteilung), led by the division's director, is composed by subdivisions (Unterabteilungen) with several sections (Referate) as the smallest unit (Hustedt and Tiessen, 2006, p.25). In contrast to the hierarchical line organisation, in few cases competencies are organised as a staff (Stab) which directly reports to the state secretaries and the minister.
26. The relationship between public discourse and changes within the public administration is rarely studied. While some authors study the media framing of governmental IT system failures and their impact on the organisation of the government's operational IT management (Pelizza & Hoppe, 2015), we focus on role of public discourses for the emergence of internet-related responsibilities within ministries.
27. The 'crypto debate'-or 'crypto war' as it is often called-refers to the struggle between governments and intelligence agencies on the one hand and civil society and IT companies on the other hand, over the question whether the export and individual use of cryptography should be limited in order to allow for decryption for the purpose of national security. See, for instance, Giacomello (2008, pp.26ff