Making Broadband Wireless Services: An Actor- Network Study of the US Wireless Industry Standard Adoption

Transcription

1 Association for Information Systems AIS Electronic Library (AISeL) All Sprouts Content Sprouts Making Broadband Wireless Services: An Actor- Network Study of the US Wireless Industry Standard Adoption David Tilson Case Western Reserve University Kalle Lyytinen Case Western Reserve University, Follow this and additional works at: Recommended Citation Tilson, David and Lyytinen, Kalle, " Making Broadband Wireless Services: An Actor-Network Study of the US Wireless Industry Standard Adoption" (2008). All Sprouts Content This material is brought to you by the Sprouts at AIS Electronic Library (AISeL). It has been accepted for inclusion in All Sprouts Content by an authorized administrator of AIS Electronic Library (AISeL). For more information, please contact

2 Working Papers on Information Systems ISSN Making Broadband Wireless Services: An Actor-Network Study of the US Wireless Industry Standard Adoption David Tilson Case Western Reserve University, USA Kalle Lyytinen Case Western Reserve University, USA Abstract We adopt actor-network theory to examine how technical and human actors interact to reach agreement on the creation and adoption of wireless services and standards. We present a model in which actors formulate standardization strategies based on their perceptions of existing and future actor-network configurations in light of their interests. The model is used to explore changes in the US wireless industry in the midst of its transition to the third generation (3G) technologies. We interview top executives from key players in the US wireless industry and analyze a major US wireless network operatorâs standardization strategy using the model. The new technical potential offered by 3G has instigated a transformation of the industryâs actor-network and stimulated renewed interest in standards setting and adoption. Operators have been among the most influential actors in initiating and enabling this transformation even though the technical potential has largely been brought to light and standardized by others. New relationships need to be established with, and among, diverse industry actors through specific standards exploitation plans. As a result new technological interfaces and inter-organizational arrangements are being forged to exploit the emerging technological potential. These strategies are, for the moment, ambiguous and open due to the uncertain nature of standards creation and adoption choices as well as rapid change in the market place (including the impact of ongoing mergers). Keywords: Standards, Actor-Network Theory, Wireless Industry, 3G Transition Permanent URL: Copyright: Creative Commons Attribution-Noncommercial-No Derivative Works License Reference: Tilson, D., Lyytinen, K. (2005). "Making Broadband Wireless Services: An Actor-Network Study of the US Wireless Industry Standard Adoption," Case Western Reserve University, USA. Sprouts: Working Papers on Information Systems, 5(21).

3 Introduction Standards are a set of technical specifications adhered to by a producer, either tacitly or as a result of a formal agreement (David 1990). Standardization is the social process through which participants reach agreement on the content of technical specifications. Much of the economics literature on standards has dealt with the competition between two firms and how network externalities lead to winner-take-all outcomes (e.g. Farrell and Saloner 1985; Besen and Farrell 1994). The stylized abstractions of one-dimensional market benefits pursued by actors neglect the significance of actors on-going concerns about their relationships with others. Consequently, the identification of the standard needs and the social process involved in their creation has not been addressed (Schmidt and Werle 1998; Mansell and Steinmueller 2002). In this paper we deploy Actor-Network Theory (ANT) as a theoretical lens to examine the socio-technical means through which agreements are reached during standards making and adoption. ANT offers a vocabulary for describing technical and social mechanisms that go into the negotiations preceding agreements. The existing ANT literature on standards, while insightful, does not address the actual creation of standards in a systematic way. Most studies explore established standards as irreversible networks and resulting configurations of the actornetworks once the translations have been successful. In this paper we unpack the translation processes actors undertake with one another in standards making and adoption by focusing on how actors formulate diverse standardization strategies to pursue their own interests and how they relate to other actors to make that possible. We examine these ongoing translations in the context of the US mobile wireless industry facing the shift to 3G services. Studying standards during the 3G transition provides a unique opportunity to examine the dynamics caused by a disruptive technological shift and the consequent reordering of relationships in the formerly stable actor-network. ANT has not been previously used in such exploring technological change in such a large scale, dynamic and global setting. Examining the provisioning of public computing services entered the IS research mainstream primarily when the Internet spawned in the mid 90 s and e-commerce and related services were invented. Currently new computing services are now being designed for users to be rendered via wireless platforms. This is already happening on a wide scale in Japan and Korea- and many are expecting similar growth in the US during the coming decade. Classic IS adoption studies (e.g. TAM or diffusion of innovation theory) have primarily looked at how individuals adopt services within and by organizations. Yet, the provisioning of mobile services requires the building and integration of extremely complex heterogeneous systems regulated by standards. Therefore before analyzing how individuals can adopt such services we need to examine how such services can become possible through industry wide cooperation and design. Thus design and adoption of such services is about actors strategies and the politics of design in very large scale information systems. In what follows we will introduce an ANT based model for such strategy formulation and provide a brief overview of the wireless industry and its 3G transition. We explain our research questions and the cross-sectional case study methodology used. Finally we present and discuss the results of the study and draw our conclusions Sprouts 5(4), pp , 138

4 Actor-Network Theory Actor-network theory (ANT) adopts a socio-technical perspective into design and analysis of technological systems that views the world as networks of technical and social actors (elements). For Latour (1998) there is nothing but networks. He (1992) describes modern societies as having a fibrous, thread-like character and argues that actors are defined solely by their ties to other actors. ANT does not distinguish between macro and micro actors. Actors can be an individual, a group, an organization or a governmental institution. They can also be a technological artifacts ranging from the smallest component to the largest system 1. Law (1992) describes the building of actor-networks as the process of overcoming the resistance of all sorts of human and technical actors and weaving them into networks with other actors. The challenge is to explore how such underlying actor-networks come to generate effects like organizations, power, innovations, or technical standards. Latour (1986) argues that the actor-network based view of the spread of innovation applies to anything from goods and artifacts to claims and ideas. In this paper the view is used to explain the creation and diffusion of technical standards that define wireless services. The core of ANT analysis is to examine the process of translation (Callon 1986; Latour 1987) where actors align the interests of others with their own. Translation follows three phases. During the first phase, problematization, a focal actor frames the problem and defines the identities and interests of other actors that are consistent with its own interests. The focal-actor renders itself indispensable by defining a process under its control that must occur for all actors to achieve their interests. Callon (1986) calls this process an Obligatory Passage Point (OPP). The OPP is typically in the direct path of the focal actor in the pursuit of its interests. Other actors may have to overcome some obstacles to pass through the OPP (Callon 1986; Sidorova and Sarker 2000). For example, in the computing industry the Windows API can be thought of an OPP for software developers. Microsoft has considerable power because of its control of this OPP and as a result a huge actor-network is aligned with its interests. The definition of the interests of others and of the OPP can be thought of as part of an actor s strategy for aligning the interests of others with its own. Other elements of the strategy might include creating incentives to encourage other actors to overcome the obstacles in the way of passing through the OPP. During the second phase of translation, interessement, the focal actor executes these strategies to convince other actors to accept its definition of their interests. The final phase of translation, enrollment, is the moment when another actor accepts the interests defined by the focal actor. Enrollment also includes the definition of roles of each actor in the newly created actor-network. Actor-networks with strong, stable ties among actors can become taken for granted and used as packages or resources in the continued construction of actor-networks (Latour 1987). These stable networks- black-boxes- can include agents, devices, texts, relatively standardized sets of organizational relations, social technologies, boundary protocols or organizational forms (Law 1992). For example, Bowker et al. (1996) found that a classification scheme of nursing work acted as a black-boxed political actor. Boland and Schultze (1996) explained how activity-based costing became black boxed through the enrollment of allies. However, black-boxes are always precarious (they continue to face the resistance of actors) (Law (1992). Black-boxing thus resembles punctualization, which is maintained by being 1 As actors can be human or non-human we purposely use the pronoun it rather than him or her Sprouts 5(4), pp , 139

5 performed and reproduced. No organization, innovation or standard is ever complete as actors can always leave the network (Callon 1986). Black-boxes exhibit the property of irreversibility- the extent to which it is subsequently impossible to go back to a point where that translation was only one amongst others; and the extent to which it shapes and determines subsequent translations. (Callon (1991 p.150) Irreversibility not only makes it difficult to undo previous translations, but also constraints future possibilities 2. IS field has faced in recent years a growing number of ANT based research on standards. Hanseth et al. (1996) applied ANT to intangible technologies i.e. standardized Internet protocols and showed how ANT provides a fruitful way of thinking about how standards get stabilized and become irreversible. Monteiro & Hanseth (1996) argued that infrastructure standards inscribe organizational behaviors deeply within their technical details, and that such standards are different in that they require the coordination of the many actors, as well as mobilization of institutional arrangements. More lately, Hanseth & Monteiro (1997) used ANT to describe the historical evolution of EDI. In most parts the past ANT-based literature of standards has focused on existing stable actor-networks. ANT is generally used as a framework for historical description and narrative. There is a distinct lack of attention to the dynamics of problematization and interessement phases. Moreover, there is a paucity of the IS research that uses ANT at the industry level (Walsham 1997; Howcroft, Mitev et al. 2004; McLean and Hassard 2004). ANT and Standardization We view standardization as a cascading stream of translations which can either reinforce or change the existing actor-network. A minor upgrade to an existing standard for example, may improve the efficiency or effectiveness of products built upon the standard (e.g. a new voice compression scheme that improves both call quality and the call carrying capacity of a wireless system). In such a case the alignment of interests by a focal actor is relatively easy. Improved performance attracts additional actors that strengthen the network while causing little disruption to the existing structure. In contrast, a major technological change brings with it the potential to radically reshape the existing network. A focal actor undertaking the problematization of a major change is likely to face stiff resistance to enrollment because of the strong ties among the actors in the existing actor-network. However if problematization and subsequent enrollment is successful, the existing actor-network will change deeply and widely (e.g. the 3G transition). One could view this as the whole-sale demise of the old network and the creation of a new one. We view it here as a major reconfiguration of the existing actor-network since the old configuration forms the starting point and its irreversibility influences ongoing translations. Standards making entails a series of translations that include but are not limited to: the initiation of a standardization forum or effort, the creation of agreed objectives, the creation of an agreed specification, its adoption by industry participants and finally the market diffusion of products and services based on the standard. Standards have many characteristics (e.g. scope, flexibility, procedure for definition, legal status, and procedure for enforcement) that are subject 2 Irreversibility bears a close resemblance to what David (2000;, 1985) referred to as path dependence in which accidental or serendipitous historical choices limit subsequent economic decisions. Arthur (1989) added that the even the order of small events can have a significant effect on outcomes. One of the key conclusions from path dependency is that while one can t predict system behavior it is possible in retrospect to trace the reasons for why it behaved as it did Sprouts 5(4), pp , 140

6 to negotiation during the translation. The focal actors and OPPs may change as standards making progresses. The reconfiguration of the network through a series of translations can be thought of as on-going process of negotiating and discovering the OPPs and focal actors that make the standard. Problematization A focal actor identifies potential standards during a problematization stage. Preliminary decisions concerning what will be standardized (and what will be left proprietary) as well as how it should be standardized are made at this stage. The focal actor s interests, which underlie such decisions, are based on the benefits it hopes to realize from the standard s implementation. Once a standard is in place problematization revolves around convincing other actors to adopt it. In either case the focal-actor identifies the other actors it wants to enroll. We define the collection of strategies created to enroll other actors as a standardization strategy. In formulating its standardization strategy the focal actor hypothesizes or imagines alternative actor-network configurations (Callon 1986; Latour 1995). Its perception of the existing actor-network provides an understanding of the translations required to reach his or her alternatives. The focal actor s interests shape its preferences for alternative configurations. Thus we can conceive of a model of strategy formulation in which the focal actor s interests are mediated by how it imagines future actor-networks configurations and the translations required to reach them (Figure 1). Influences Influences preferences Influences Perception of existing actornetwork Actor s perceived interests Imagined future actornetwork configurations Standardization Strategy Constrains preferences Alters Interactions among actors Influences Figure 1. Standardization strategy formulation for actors in the problematization and interessement phases Interessement The focal actor strives to convince other actors to accept its problematization by enacting its standardization strategy during the interessement phase. Other actors become aware of the problematization and ensuing negotiations may eventually lead to the creation of a standard. According to Latour (1986) innovations like standards are not transmitted unaltered. A standard, or a proposed standard, moves through time and space in the hands of actors that react to it in different ways (modify, deflect, betray, augment, appropriate or drop) [omitted for anonymity] Sprouts 5(4), pp , 141

7 Standard creation and adoption is thus determined by the actions of all the actors that come into contact with it. If they do nothing the standardization process or the diffusion of the standard ceases. The creation and adoption of a standard depends on its transformation by actors to suit their own needs and often entails some loss of control by the initiating actor. In standards forums the interessement phase takes-on the character of a multi-way negotiation. Actors participating in standards making formulate standardization strategies to pursue their own interests and preferred imagined futures. Thus the model in Figure 1 applies to strategy formulation by all actors during interessement as well as by the focal actor s during problematization. The execution of standardization strategies alters the actor-network configuration, the interests of participating actors, and their preferences for alternative future configurations. Thus standardization strategy formulation is an on-going dynamic process at the heart of aligning the diverse interests and imagined futures of a wide range of actors. Enrollment In a successful standardization effort the interactions of actors pursuing their own standardization strategies has resulted in an agreement on the scope and content of a standard. In the process the different interests of a range of actors have been translated into the agreement. The standard becomes inscribed in documents and technical artifacts or software, which bolster its durability. The standard may be framed as the solution (OPP) to certain industry problems and the actor-network is widened in time and space as more actors are persuaded to adopt the standard. If the standard builds a strong enough network of actors (hardware and software artifacts, community of adopters/users, implementation and ratification by regulators) it may become black-boxed and irreversible i.e. the way of doing things it inscribes becomes taken for granted. As with all actor-networks a standard remains precarious as actors can leave at any time. The Wireless Industry The wireless industry has been offering telephony services to corporate customers and consumers since the early 1980s (Bekkers 2001; Funk 2002). The automated systems that make wireless services possible are made up of many components including: wireless handsets, antenna towers and base stations to support the radio links to handsets, mobile switching centers to provide mobility management and interconnect with the public telephone network, and backend systems for provisioning, customer service and billing. Standards play a vital role in the industry by facilitating the interoperation of these components. Wireless services have been critically dependent upon the creation and implementation of standards (Funk 2001; Funk and Methe 2001). Standards play a particularly important role in shaping the relationships and overall structure of the wireless industry. A series of studies (published in a special issue of Telecommunications Policy) on the evolution of first (1G) and second generation (2G) wireless services highlighted the importance of the relationships among industry participants and the central role of standards in the diffusion of wireless services. Synthesizing the implications of these studies [omitted for anonymity] conjectured that (a) the evolution of wireless services is critically dependent upon the creation and implementation of standards, (b) many of the critical industry relationships were organized around standards, and (c) the diffusion of the services is enabled and shaped by the dynamics of the relationships among three analytically distinct domains: 2005 Sprouts 5(4), pp , 142

8 The Innovation system is the interlinked network of sites, competencies, ideas and resources, which is capable over time of developing novel technologies and solutions based on research and development activity. Exploitation of these innovations and technologies in wider systems often requires the creation of standards. The marketplace is a set of actors that produce some telecommunications services or technologies (within a value network) exploiting the technological potential defined within telecommunications standards The regulatory regime is any type of authority (industrial, national, international), which can influence, direct, limit or prohibit any activity in the innovation system, the marketplace or the regulatory regime itself 3. These domains form the institutional environment in which standards are created and adopted is illustrated in Figure 2. [omitted for anonymity] treated the three institutional domains as constellations of actors in their actor-network based description of wireless service diffusion in Korea. Fomin et al. (2004) also adopted the framework for their on-going study of the Danish wireless industry. We use the framework as a way organizing our examination of how translation plays out in the building of actor-networks in the US industry. The transition from analog 1G to digital 2G systems was primarily motivated by 2G s more efficient use of radio spectrum and increasing market demand for wireless telephony. Although digital, 2G standards remained voice-centric and were based on the then dominant ISDN circuit-switched technology. Innovation System Marketplace Standards Regulatory Regime Figure 2. Institutional environment for standardization [omitted for anonymity] The 2G transition resulted in few changes to the overall structure of the industry s actornetworks. The transition brought new network operators and the first successful wireless data service, text messaging, brought a few new players (e.g. banks and airlines), albeit in a peripheral way. The main industry participants remained the network operators, national and regional regulators, and the manufacturers of infrastructure, handsets and semiconductors (Funk 2002). The flow of products and services among them is illustrated in 3 Definitions adapted from [omitted for anonymity] 2005 Sprouts 5(4), pp , 143

9 Figure 3.[Omitted for anonymity] highlighted two key interfaces in 1G and 2G evolution. The first was the air interface. By specifying how mobile devices operate within the wireless infrastructure, air interfaces played an important role in defining the relationship between infrastructure and device manufacturers. The second key interface was the licensing and pricing policies established by national or regional regulators that influenced the relationship between network operators and their customers (e.g. it has been argued that whether the caller or recipient pays for calls to wireless devices has effected the diffusion of wireless telephony). The creation of standards in the wireless industry entails a series of many translations. While few can be known in advance, industry participants at least know that certain interfaces are required (e.g. the air interface). They also know that key suppliers, like semiconductor manufactures, have to be convinced that the market for products will be large enough to justify the necessary investments in R&D and participation in standards making. The wireless industry s actor-networks have exhibited properties of irreversibility and the development of wireless telephony has followed divergent paths in different national contexts Fomin et al., (2004). The different technological trajectories and patterns of use have been shaped by regulation (West 2000; West and Fomin 2001), market structure (Lera 2000), and socio-cultural settings (Katz 2003) among other factors. These differences influenced the emergence of focal actors and OPPs and constrained the range of future actor-network configurations imagined by industry actors. The 3G Transition The huge popularity of the Internet during the late 1990s sparked the interest in mobile wireless data services. As the wireless telephony market reached saturation in many countries, and strong price competition resulted in falling voice revenues, data services were seen as a way of continuing revenue growth. The perceived need for broadband packet-switched data transport capabilities drove the creation of a new generation of air interfaces. Major battles were fought around the selection of these 3G standards (Bekkers 2001; Funk 2002) but the standards are now in place 4. In Japan and Korea network operators launched packet-switching based 3G networks that supported a wide range of data services. Today thousands of Internet-like services are offered to consumers on Japanese and Korean 3G networks (e.g. DoCoMo in Japan). In Europe and the U.S. initial offerings were based on circuit-switched 2G technologies. Subsequent 2.5G 5 and 3G based services have not reached any where near Japanese or Korean adoption rates. The technological potential of the 3G broadband data capabilities and improvements in key device technologies (e.g. processors, displays, storage and to a lesser extent batteries) enabled the creation of new services. Many of these new services were content based (e.g. audio and video entertainment, browsing of news, financial and other information and downloading or games and ring tones). This contrasts with the person-to-person communications services that dominated 1G and 2G. The development of new wireless data services is necessitates the integration of new types of actors in the industry. A first attempt at mapping the relationships among the actors in 2004 is depicted in Figure 3 (the new actors are shown in dashed boxes). While this is by no means complete it illustrates important industry changes in terms of relationships, the scope of the industry and the nature complexity of the network. 4 The air interfaces may not be the dominant OPPs that they were in 2G. The prospect of software defined radio may make the selection of one air interface standard or another largely irrelevant G networks are upgraded 2G networks that offer packet-based data services at near 3G bandwidth 2005 Sprouts 5(4), pp , 144

10 Innovation System Market Place Semiconductor Manufacturers Device Manufacturers Network Operators Customers Corporate Consumer OS and middleware vendors Content Providers Infrastructure Manufacturers Service Providers System Integrators/ Solution Providers Regulatory System Industrial policy Operator licenses Spectrum allocation Regulatory framework Government / Regulators (New types of actors associated with data services are shown in dashed boxes) Figure 3. Major organizational actors in the wireless industry This additional network complexity has led to the problematization of many new interoperability standards. For example, from the network operator s perspective, standards are needed to handle the management and aggregation of data flowing from content and service providers. As a result, both proprietary and open content delivery platforms have emerged. The role of the terminals operating systems and middleware on both terminals and backend systems has become more important. Open higher layer interface definitions for 3G, which mostly influence interactions with new industry participants, are being designed and negotiated in multiple industry-wide for a like the Open Mobile Alliance 6 (OMA). As wireless devices have adopted the same characteristics as a computer there has been also increased attention to modularization and standardization of their internal interfaces (Smith 2003). The scope of industry s standardization efforts has been greatly complicated due to 3G transition. Currently, there are over 100 standard bodies with participants from the computing, data networking and content industries. Standards-making has become global (Steinbock 2003). In ANT terms the size and complexity of the actor-network has greatly increased and actors are faced with the challenge of trying to figure out how to make connections, and how to reconfigure their existing actor-networks Sprouts 5(4), pp , 145

11 Research Goals and Methods Diverse actors have different ways of interpreting their interests and of imagining the actor-networks in which they might enroll. A key aspect of understanding how actors reach agreement on standards making or adoption is the understanding of the three phases of translation in these contexts. Question 1. How do problematization, interessement and enrollment play out in the standards making and adoption as well as the transformation of the industry and services? In formulating their standardization strategies (Figure 1) actors draw upon perceptions of their existing actor-network and imagine future network configurations. How human actors relate to one another can effect how they relate to technical actors. Likewise how human actors relate to a technical actor can influence how they relate to one another. Building an appreciation of this dynamic is important to understanding standards making and adoption. Question 2. How do actors relate to others? How do they build relationships with other human actors, artifacts and with standards? To carry-out a complete actor-network based study of standards making during the 3G transition would require unprecedented access to information about the activities in the standardization forums and further a field in the industry. As we don t have such access we performed a cross-sectional case study of the industry. We conducted 9 in-depth interviews with key industry decision makers. Interviews are a suitable means of data collection as the research goals deal primarily with perceptions and rationales. We started with interviewees from network operators and manufacturers of wireless infrastructure and mobile devices. We then followed the actor-network approach (Latour 1987) by asking interviewees who else we should interview. We used this snowballing strategy to discover the range of actors that had to be enrolled to enable the delivery of broadband wireless data services. Using this method we identified actors in each of the institutional domains ( Figure 2). The interviewees included executive level employees of a network operator, an infrastructure manufacturer, two device manufacturers, two semiconductor manufacturers, a middleware vendor, a system integrator and an industry consortia involved in wireless standards making. Each interview explored how and why the actors participated in standards making. A representation of their actor-network was built up by encouraging them to talk about how they relate to other actors, how and why they enroll in actor-networks, and how they build them. We also asked open questions about technology and services to explore how they went about problematizing. The interviewee s narratives are used to understand the ways in which different actors built their standardization strategies. The interview guide was based on one used in a study of the Korean wireless industry [omitted for anonymity]. Two interviews were conducted faceto-face and the rest by phone. The interviews took place between October 2003 and May 2004 and typically lasted two hours. Transcriptions of the interviews were produced by a professional audio typist. The transcriptions (~450 pages) we carefully corrected by the authors and we listened to the 2005 Sprouts 5(4), pp , 146

12 recordings at least two more times. On the first pass we identified the themes that made up the interviews and created summaries for each. On the second pass we distinguished between the stages of translation for each theme. We noted the interests expressed and the descriptions of the relationships among industry participants. We also noted where historical, current or imagined future actor-network configurations were described, and where standardization and other strategies were discussed. The notes for each theme and narrative were analyzed for relevance to each research question. The findings were synthesized for each question to gain an understanding of the ongoing dynamic interactions among the actors during the 3G transition in the U.S. We distinguish between the human 7 (mostly organizational) actors that participate directly in standards making and the other actors that don t. Only the active participants engage in imagining future actor-networks and formulate standardization strategies. The actors that don t participate directly include both human actors (e.g. end-users) and non-human actors (e.g. technological artifacts and inscriptions of existing standards). Discussion of Findings Translations In this section we discuss the findings pertaining to how translation plays out in the transformation of an actor-network. Tables are used to organize the discussion: themes and narratives are summarized on the left of the tables and the implications are presented on the right. The first theme reveals that in several instances problematization is preceded by some sort of disruption to actors imagined futures. These disruptions may arrive from any of the institutional domains ( Figure 2) or from outside the existing industry actor-network. Theme: Source of problematization There were several internal sources of industry change Availability of digital wireless networks capable of supporting data services Possible upgrade of network capacity Need to compensate for low growth in voice revenue Number portability Some came from outside the industry Growth in Internet usage New technologies (Wi-Fi, Wi-Max & VoIP) Implications Sources of disruption to existing actor-networks can come from Within the existing actor-network - Technical potential (technology push) - Commercial necessity (organizational actor push) Outside the existing actor-network - Explosive growth in use of Internet services (perceived customer pull) - Emergence of alternative wireless transport technologies (competitor push) The innovation system, marketplace and the regulatory system 7 ANT s implied symmetry between human and non-human network elements has been the subject of considerable controversy (e.g. Callon & Latour, 1992; Collins & Yearly, 1992a,, 1992b; Latour, 1998). For Walsham (1997) the removal of the rigid separation of the humans and non-humans is valuable in an age of hybrids and blurred and negotiable boundaries, but this does not imply an acceptance of the extreme position of symmetry. Technological actors and inscriptions, while influential, do not have conscious interests to pursue, cannot rationalize, and certainly do not strategize Sprouts 5(4), pp , 147

13 Standardization strategy is but part of actors wider strategies The first narrative can be described in actor-network terms: The technical potential of the digital wireless network helped create new imagined futures which included the provision of profitable mass-market wireless data services. The imagined futures required the enrollment of existing and new actors. Most actors shared compatible imagined futures and were easily enrolled. However, one category of actors (customers) did not share these imagined futures and did not enroll. Narrative: The launch of the network operators initial data service Technical potential of digital wireless network to provide new wireless data services The increasing popularity of the Internet Existing wireless industry players recognized the opportunities. As did companies in the computing and content industries Technological interfaces were put in place (e.g. WAP) and new commercial relationships established - Device and infrastructure manufacturers, middleware vendors worked with one another to deliver a data service platform - Content providers vied with one another for opportunity to work with operator - Operator worked with content providers to bring new data services to market Few customers adopted the services Possible generalizations Some actors enroll contingent on others joining later Broad consensus about an imagined future does not ensure that it will happen Actors play mutually dependent roles in the industry actor-network While the enrollment of some actors may be necessary for an imagined future it may not be enough if some critical actors (the customers in this case) are not enrolled. The actor s standardization strategies are part of overall strategies to construct their preferred imagined future actor-network configuration. Standards are but one inscription used by actors to shape the future and the relationships among actors (others include contracts and industry practices). Theme: Standardization The need for a technical standard can be identified by a desire to realize the potential on an technical innovation Figuring out how to fulfill a customer s needs can lead to for a new interface and the Creation of a proprietary interface Adoption of an existing standard Creation of a new standard Possible generalizations The problematization of a new standard can come from - Technology push (from innovation system) - Customer pull (from the marketplace) - Regulator Each source of problematization requires the formulation of imagined futures The rationales for participating in standardization efforts included economic reasons. Strategies were aimed at shaping the industry in favored ways, and at preventing others shaping it in undesirable ways. This implies that the actors create and evaluate both positive and negative imagined futures. Rationales also include identifying possible OPPs, or architectural control points, that an actor would not want others to possess Sprouts 5(4), pp , 148

14 Theme: Rationales for standardization Several rationales for standardization strategy and participation in particular standardization efforts were raised in the interviews Minimized cost of infrastructure since it is not seen as a differentiator in the marketplace Avoid lock-in to infrastructure and device manufacturers Avoid one industry player from obtaining control of important interface(s) i.e. playing not to loose Possible generalizations Rationales for standardization can be economic Standardization strategy can be based on trying to - Shape the industry in the image of attractive imagined futures - Avoid the industry s transformation into unattractive imagined futures Imagined futures may have many elements and the uncertainty associated with each of them can vary. Actors try out a variety of combinations of the elements to identify which they consider plausible future actor-network configurations. Narrative: Assessment of network upgrade Infrastructure manufacturers presented network upgrade options with increase data capacity Operator tried to create business case for enhanced data services but were skeptical because of poor uptake of prior services Upgrade was justified by increased voice capacity Possible generalizations Assessing a range of imagined futures can be thought of as trialing different actor-network configurations before problematizing the issue to enroll other actors Imagined future can have several dimensions and elements Different levels of uncertainty can be associated with different elements of the imagined futures A perspective that limits analysis to a single focal actor with a single OPP is too simplistic for this sort of industrial context. Many actors can be regarded as focal actors. The same sources of problematization were instrumental in prompting multiple actors to attempt enrolling others. Where imagined futures are shared, enrollment is mutual. Theme: Industry transformation Wireless data services created important roles for actors from other industries Computing (e.g. middleware) Content (traditional and web) IT Services Operator adopted Java middleware solution from computer industry rather than a solution from existing supplier Implications Focal actors from multiple industries attempted to enroll one another. Network operator used the technical potential of its digital network and space on the all important home deck acting as OPPs Content providers had their own imagined futures and their most attractive content may have acted as OPPs Java adoption could be viewed as - Enrollment into the computing industry s actornetwork built upon the larger wired communications networks - Weakening of ties with existing wireless technology provider New service opportunities attracted actors from outside the industry. New connections were formed among actors, and the changes to these relationships were at the heart of the transformation of the industry s structure Sprouts 5(4), pp , 149

15 Relationships among actors The relationships with the infrastructure and device manufacturers (Table 1) existed long before the advent of wireless data services. They ways in which the network operator coordinated its activities with these actors had remained largely the same. In actor-network terminology the relationships had become irreversible. The relationship with customers (Table 2) also exhibits irreversibility with the operator s historical focus on the consumer market continuing into the data services. Infrastructure manufacturer Aspects of commercial relationship Works with more than one infrastructure manufacturer across its many markets Uses competition between manufacturers to identify best solutions to problems Maintains close, longer-term relationships (due to longevity of wireless infrastructure) Relationship around standardization Works with manufacturer to have preferred solutions adopted as standards (to avoid being locked-in to proprietary solutions) Does not see infrastructure as a differentiator - uses standardization as way of minimizing costs Handset manufacturer Has been characterized as a more arms-length relationship than with infrastructure manufacturer. Network operator expects to specify device requirements and for manufacturers to meet the specification The network operator works with several handset manufacturers. Handset features seen as a differentiator less standardization desirable The handset manufacturer is required to provide the data capabilities required (e.g. WAP) through collaboration with specified middleware providers The device manufacturers have more recently developed direct relationships with middleware providers without the mediation of operators Table 1. Overview of a network operator s relationships with key actors in the innovation system However, the industry changes have not left these relationships completely unaffected. The advent of wireless data services added considerably to system complexity and introduced new technical interfaces. The network operator and its infrastructure manufacturers had to coordinate their stances to an increasing number of interface specifications and standards forums. The establishment of new relationships can affect existing ones. The network operator s adoption of Java based middleware, for example, strengthened its ties with actors from the computing industry. The rejection of the middleware solution proposed by a wireless industry vendor represented a slight weakening of one of the operator s most longstanding relationships. There are many ways in which the relationships with new actors could be configured. While the actors were striving to align their imagined futures with one another and to enroll one another the relationships remain more fluid that those among the traditional wireless industry actors. The more dynamic standards development at the top of the stack contributed to a change in the distribution of power between the operator and content providers (Table 2). The operator s digital network initially acted as an OPP for enrolling content providers and gave the operator power in these relationships. Increasing standardization led to the weakening of the OPP and the operator s power Sprouts 5(4), pp , 150

16 Aspects of commercial relationship Consumer Consumer focus for telephony services led to similar focus for data services Competitive market and market saturation limiting telephony revenue growth Relationship around standardization Consumer Consumer generally not involved in standardization arena although regulator may act on their behalf Customers Content Providers Corporate customers Some applications are cross industry (e.g. , PIM, and Intranet access) Other applications require custom integration with existing back-end systems Operator uses wireless offering as a door opener for their wired telecom offerings Consumer focus led to desire to offer content based services bringing content providers into the wireless industry - Web Portals providing news/entertainment/information - Music (e.g. ringtones) - TV broadcast and video on demand Prior to the initial launch of wireless data services the network had the power in this relationship - content providers vied with one another for space on the network operators home deck. Corporate customers Standards based solutions sometime speed time to market Willing to adopt proprietary solutions to gain competitive advantage over other operators The battles in the standardization forums now revolves around the delivery of content - New actors (content providers and actors from the computing and networking industries) into wireless standardization forums - Wireless industry actors increasing participation in computer and networking industry standardization forums Increasing use of standards for the delivery of data services were perceived as weakening the network operators power in this relationship Table 2. Overview of a network operator s relationships with key actors in the marketplace The connections with, and among, the new actors remain contingent on the ability of the new configuration to enroll other key actors (customers in this instance). These relationships remain fluid and are far from being black-boxed (many elements of the imagined futures are open). While the relationships among the traditional actors are more stable and exhibit considerable irreversibility, they are also subject to change (albeit more slowly) as the industry s transformation progresses. It is relatively straightforward for commercial actors to connect with one another and with the technology as they interact to align their imagined futures. Figuring out how to make those connections with customer is not at all straightforward. While market research is one way attempting to figure out how to make the connections the customers may not be able to formulate imagined futures or explicit strategies. Enrollment of customers may have to be conceptualized differently. The combination of wide area broadband wireless communications combined with portable computers (most often in the telephone handset form factor) has brought the world a new medium for making content and computing services available to users. The resulting arrival of actors from the computing and content industries has been at the root of a disruption to the previously stable wireless actor-network configuration. Our interview with a major US network operator revealed the strenuous efforts of the innovation system and marketplace actors to deliver new services. The industry s actor-network 2005 Sprouts 5(4), pp , 151

17 has been reconfigured to accommodate the new actors. However, at the time of writing the actors have yet to find a configuration to enroll key actor (i.e. U.S. consumers) in large numbers. While we have not been able to find the definitive explanation why enrolling customers has proven so difficult, it is interesting to note some contrasts between how translations differed in contexts where data services have been successful and where they have not. In Europe, consumers made the connection with the text messaging capability inscribed in GSM handsets. The consumers new connections with a technical actor and the enrollment of the text messaging service to serve their own interests created new interpretations of the technology that had not existed before. Only then did commercial organizational actors create and reflect on imaged futures that included services to utilize consumers new connection with the technical actor (i.e the handset). The hugely successful DoCoMo platform in Japan provided an open platform for content providers to make connections with the platform and consumers. In contrast the U.S. models have been largely top-down, with large traditional actors attempting to reuse content and services from other media to enroll consumers. Why U.S. operators have not been able to establish a successful platform enabling the same sort of connectivity achieved by DoCoMo an outstanding question. Conclusions Few studies in the actor-network genre have attempted to study the process of translation with multiple-actors at the industry level (Walsham 1997; Howcroft, Mitev et al. 2004; McLean and Hassard 2004). This study looked at the on-going reconfiguration of an industry s actornetwork. Our main finding is that the traditional conception of a single focal actor possessing an obligatory passage point (OPP) that allows it to translate the interests of other actors (Callon 1986) is limited in this sort of context. Translation is considerably more dynamic and includes mutual interessement and adjustments to align sufficient elements of actors imagined futures for connections to be established. The problematizations associated with this industry transformation were prompted by disruptions (e.g. the emergence of a mobile medium for the delivery of computing services). Such disruptions bring together powerful actors (with their own extensive actor-networks) that all act as focal-actors. Such disruptions are increasingly likely as the integration of organizational activities within and across industries become more widespread. The connections among actors may become inscribed in contracts, partnerships, industry practices, technical standards, or customers use of technology. Technical standards have a special place in industries built upon complex interconnected information systems and actors standardization strategies are an important part of their overall strategies. There is evidence that a common rationale for participating in standardization it to prevent other actors obtaining control of an OPP that would lead to a translation process with a single (powerful) focal-actor. This rationale may be particularly important during a transition where the increased fluidity implies opportunities for actors to grab power positions before the actor-network configuration stabilizes again. Establishing connections among industry level actors and technical artifacts is an extremely complex feat of heterogeneous engineering. However, the realization of imagined futures is reliant on establishing the most difficult connections consumers in this case. Focalactors should strive to understand the interests of the most challenging actors and to test strategies for their enrollment. The key limitations of this work are that it is based on the actor Sprouts 5(4), pp , 152