DRAFT – AUGUST 2002. Rolland, K.H. 2002a. In the Business of Risk: Global Information th Infrastructures and Shifting Identity. (An earlier version of this paper was presented at the 25 Information Systems Research Seminar in Scandinavia (IRIS 25) at Bautahøj, Denamrk, August 10 – 13 2002.) THIS PAPER IS SUBMITTED TO SJIS. In the Business of Risk: global information infrastructures and shifting identity Knut H. Rolland Department of Informatics, University of Oslo PO BOX 1080 Blindern 0316 OSLO, NORWAY Phone: +47 22 85 29 34 Fax: +47 22 85 24 01 knutr@ifi.uio.no Abstract. This paper aims at exploring the linkages between standardizing aspects of global information infrastructures and professional identity. In doing so, this paper adopts Beck and Giddens’ theory of ‘reflexive modernization’ to analyze how the introduction of a global information infrastructure (the GIS) in a Maritime Classification Company (MCC) transformed work practices and professional identity of ship surveyors. The analysis of the MCC case suggests that tensions between the professional identity of surveyors and the ways in which the GIS infrastructure standardized local practices introduced new risks and ‘radical doubt’ in surveyors’ work. As a consequence, surveyors strive to re-construct their professional identity and establish new systems of trust to accommodate risk and the standardizing aspects of the GIS infrastructure. While these unanticipated transformations lead to innovative new ways of working through users’ improvisations and reskilling, it also undermined some of the initially envisioned organizational transformations. Thus, risks do not stem from poor design or lack of strategic alignment, but are rather reflexively produced and diffused through use and management of the technology itself. The paper concludes by giving some implications for theorizing information infrastructures and for the practice of implementing such technologies. 1 Introduction The potential for advanced information and communication technologies to enable globally distributed work and global business strategies as well as improving control and coordination, is widely acknowledged in the Information Systems literature (e.g. Earl and Fenny, 1996; Ives and Jarvenpaa, 1991; Weill and Broadbent, 1998). However, recent empirical studies of large-scale information systems and information infrastructures, indicate that implementation and management of these technologies are more complex than first anticipated (e.g. Ciborra et al., 2000; Davenport, 1998; Star and Ruhleder, 1996; Walsham, 2001). More often than expected, information infrastructures tend to have unintended organizational consequences, produce sideeffects, and ‘drift’ away from intended plans and envisioned scenarios of usage (Ciborra, 1996; Ciborra et al., 2000). Drifting in this sense, is not inherently ‘bad’, but captures how information infrastructures imply both new risks and opportunities as users’ improvise and explore technologies’ potential in specific situations and use contexts. This paper aims at examining how drifting in the introduction of (global) corporate-wide information infrastructures is intrinsically related to the dialectical relationship between they way in which these technologies tend to standardize work and transformations of individuals’ professional identity. In doing so, this paper draws on the sociologists Ulrich Beck and Anthony Giddens conceptualisations of social transformations in contemporary society (e.g. Beck, 1992; Giddens, 1990, 1991). This body of work identifies globalization as a key tenet of modernity and emphasises that these processes of globalization are closely linked to changes in self-identity of individuals (e.g. Giddens, 1991). Information technologies are increasingly an integral part of these processes in the way that they are both influencing and influenced by processes of globalization (Bradley, Hausman and Nolan, 1993). Several contributions in IS research have attempted to conceptualise these dynamics by perceiving information and communication technologies as an instance of modernity that tend to increase the separation of time and space, serve as disembedding mechanisms, and enable institutional reflexivity (Barrett and Walsham, 1999; Hanseth and Braa, 2000). Particularly, large-scale information systems and information infrastructures can be perceived to encapsulate a ‘radicalisation’ of modernity (Beck, 1994) in a nutshell – as they tend to accelerate uncontrollability and unpredictability (Hanseth and Braa, 2000; Hanseth, Ciborra, and Braa, 2002). On the other hand, consistent with Giddens’ (1991) argument, the introduction and use of these technologies in organizational work also have deep-seated implications for how professional groups and individuals perceive themselves, their work and their role in the larger organization (Walsham, 1998, Walsham, 2001). This paper aims at extending this analysis by looking at how specific features of information infrastructures transform individuals’ professional identities in a global context of usage. Furthermore, this paper explores what kind of risks and opportunities different transformations produce for different individuals and how these risks are compensated for. The issue of professional identity is a relatively under-researched topic (Walsham, 2000) that could be considered especially important in relation to information infrastructures. Information infrastructures are implemented across diverse contexts of usage and typically involve multiple professional groups where a strong professional identity is closely linked to practice (cf. Star and Ruhleder, 1996). Introducing a common information infrastructure that re-organizes and standardise practices and make them more interdependent with other user communities’ practices is likely to cause individuals’ to question their work and role in relation to other individuals and groups. Empirically this paper draws from the findings from an explorative field study of a global information infrastructure (the GIS) in a globally distributed organization (MCC). The GIS is a state-of-the-art component-based client/server system based on Microsoft’s COM architecture, a product model, and Microsoft’s SQL-based database technology, and it is interconnected with many other databases and applications including a Web-based system used by customers. The GIS was implemented worldwide in 2000 after 6 years spending approximately 130 million EURO on inhouse development. After initial disappointments and organizational turbulence, the GIS infrastructure is at the time of writing regarded as relatively successful, and by management perceived as an important tool for increasing uniformity and quality of survey work across the world. Nevertheless, the findings described in this paper indicate that attempts to control and standardize work also imply new ‘modernization risks’. This is exemplified by the gap between the standardizing aspects of the GIS (e.g. how the GIS structures and formalizes survey work) and how surveyors’ perceive themselves, their role in surveyor work, and their experience on how quality was achieved. This had multifaceted implications across different contexts as this implicated both new innovative ways of working and reskilling, but also introduced ‘radical doubt’ in the work of surveyors and in some cases decreased the quality of reports and services. The reminder of this paper is structured as follows. The following section briefly describes some key aspects of Beck and Giddens’ theory on ‘reflexive modernization’ that are argued to be of relevance for understanding global information infrastructures. Next, a short synopsis of the GIS project, MCC, and examples of GIS in use are given along with an outline of the research approach and data analysis. Then, some of the key findings from the MCC case are discussed using the framework of reflexive modernization, and finally the paper concludes by discussing some theoretical implications for understanding global information infrastructures as well as some practical considerations possibly relevant for other cases. 2 ‘Reflexive Modernization’ as a Theoretical Perspective on Global Information Infrastructures In studying global information infrastructures reflexive modernization can be argued as especially relevant since this theory links transformations on a macro-level with issues on a micro-level, and thus enables a multi-level analysis (Walsham, 1998). More specifically, Giddens argues that globalizing tendencies are inherent in the dynamism of modernity and that this dynamism develops in a dialectical fashion. In this way, Giddens underscores that the dynamism of modernity not only has consequences on a global – macro scale, but also shapes individuals’ self-identity. An important aspect here is that the standardizing tendencies of abstract systems might lead to ‘existential anxiety’ and ‘personal meaninglessness’ (Giddens, 1991). In these circumstances the self becomes reflexive in the sense that individuals pursue to re-construct their identities in the light of the new abstract systems. A key tenet of reflexive modernization that underscores the globalizing tendencies of modernity is what Giddens refers to as ‘disembedding mechanisms’ (Giddens, 1990, 1991, 1994). The metaphor of disembedding attributes to “the ’lifting out’ of social relations from local contexts and their rearticulation across indefinite tracts of timespace” (Giddens, 1991: p. 18). Giddens mentions two specific types of disembedding mechanisms which he denotes ‘symbolic tokens’ and ‘expert systems’. Taken together, these are described as ‘abstract systems’. Information infrastructures then, can be perceived as abstract systems that disembed local work practices, separate social interaction from the particularities of local contexts, and make work practices interdependent across different contexts. However, adoption of abstract systems depend upon their re-embedding in local contexts, which requires new systems of trust to be established. According to Giddens (1991), trust is required when social interaction involves absence in time and space as well as ignorance to the specifics of ‘expert systems’. A further important aspect of the dialectics of the local and the global is that “new risks and dangers are created through the disembedding mechanisms themselves” and due to interdependencies “these may be local or global” (Giddens, 1991: pp. 19-20). Similarly, in Beck’s writings production and diffusion of risks are perceived as a key characteristic of reflexive modernization (Beck, 1992). Risk is by Beck (1992: p. 21) defined as a “systematic way of dealing with hazards and insecurities induced and introduced by modernization itself”. According to Beck (1992, 1994), complex interdependencies and formerly ‘latent side effects’ created through technical and economic development are producing new ‘modernization risks’ that are different from risks known in traditional societies. A primary difference is that risks are aggregated and diffused globally, so that local events potentially have global consequences. The reflexive nature of risks is by Beck captured in the notion of ‘boomerang effects’ in the sense that local attempts to control backfires as lack of control (Beck, 1992). Whilst Giddens and Beck do not explicitly mention information and communication technologies in their theories of social transformation, others have drawn from and extended their work with an ‘IT dimension’ (Barrett and Walsham, 1999; Ciborra et al., 2000; Hanseth, Ciborra, and Braa, 2002; Walsham, 1998; Walsham, 2001). In a study of the introduction of an electronic trading system across the London Insurance Market, Barrett and Walsham (1999) explore the linkages between the globalizing tendencies of the information system and changes in self-identity of brokers and underwriters. Here information technology is conceptualised as an instance of modernity that facilitates the globalizing dynamism of modernity as described by Giddens (1991). The case further illustrates that the transformations of work occurring after implementation of the technology were closely linked to systems of trust and issues of self-identity. Brokers and underwriters tended to resist the information system, as they perceived that new systems of trust required by the information system would replace traditional trust between persons in face-to-face contact. Moreover, the globalizing tendencies of the new information system created tensions of identity at the individual level, especially regarding concerns for deskilling and existential anxiety. Hanseth, Ciborra, and Braa (2002) emphasize different aspects of reflexive modernization when analyzing the consequences of an ERP implementation in a global company. Because of the interdependencies between the installed ERP system, the underlying operating system, desktop applications, and Lotus Notes groupware applications the decision to outsource the operation of the ERP system spurred side effects that further generated new side effects. Eventually, this lead to a re-invention of their entire information infrastructure including a purchase of 40 000 Microsoft Office licenses. Thus the introduced technology had side effects that spurred more side effects in a domino-like fashion. The case also illustrates that integrated information systems are reflexive in the sense that side-effects reflexively ‘strike back’ and affect the very phenomena that created them. Consequently, the authors conclude, “ERP installations in global organizations conform well to Giddens’ (1999) image of the modern world as a juggernaut”. 3 Case Description 3.1 Research setting and method CASE ORGANIZATION AND THE GIS PROJECT MCC is a maritime classification company headquartered in Scandinavia, employing over 5500 highly qualified engineers in 300 offices in over 100 countries worldwide. The company’s main products and services are quality and risk assessments concerning various types of ships and maritime installations. The international classification business is highly competitive and is subject to major structural changes through transformations, acquisitions and mergers. MCC’s main strategy for meeting global competition has been heavy investments in ICT. The GIS project was initiated in 1994 and implemented in local offices in 1999 and 2000 and at the time of writing the GIS is used by 1500 users in over 150 offices worldwide, and it is now, at least by managers – viewed as a tremendous success. The primary objective was to develop an integrated information system based on product model technology to improve information management and decision making involved throughout the entire ship survey process. The idea of an integrated information system was also motivated by a perceived lack of structure in the way that surveys were conducted and reported in local offices as well as a growing dissatisfaction with the existing Technical Ship Database (TSD) from the 1970s running on an old IBM mainframe computer. Additional details on the case study are described in Rolland and Monteiro (2002). Figure 1: Example of a GIS screen. Work tasks are shown to the left and example of a generated report on the right. Overview of the survey job and technical information are located in the middle. DATA COLLECTION AND ANALYSIS This research can be characterized as an in-depth case study grounded in the interpretive approach to information systems research as outlined in Klein and Myers (1999). The actual fieldwork took place at 5 different offices in Europe during three phases between February 1999 and November 2001, aiming at gaining insights in how the GIS infrastructure had transformed practices in different local contexts. Observations and in-depth interviews with surveyors using the GIS infrastructure in their daily work were conduced at three major MCC offices. Furthermore, additional interviews of local managers were conducted at one medium sized office as well as the HQ. These sites were mainly selected based on practical considerations, such as access and travel expenses. In order to supplement the evidence from these sites all located in Western Europe, MCC’s own evaluation reports from 3 offices in Asia were used as a secondary data source. During the first phase of the fieldwork (February – July 1999), 11 in-depth interviews lasting more than 2 hours with software developers and managers in the GIS project were conducted. Since the GIS infrastructure was not yet widely used throughout the organization, the fieldwork was at this stage concentrated at HQ. In July 1999 I visited a local office and conducted 6 in-depth interviews with 1 superuser and 5 surveyors. Taking place only a few months after introducing GIS, the first round of fieldwork highlighted the ongoing appropriation processes and the surveyors’ first impressions of the system. In the second phase of the field work (Mars – October 2000), a second round of interviews were conducted at HQ in order to capture the transition process from a paper-based reporting system and the TSD database to fully digital reporting based on the GIS infrastructure. During this stage, three offices were visited and over 40 indepth interviews were conducted with district managers, superusers, local IT facilitators, engineers/support personnel and managers from the GIS implementation team. From Mars until August 2000, I also spent 2 days a week at the HQ working together with the department at HQ responsible for planning further improvements as well as managing the implementation process of the GIS. I was given a desk in an open office environment and became engaged in informal discussions around the coffee machine and during lunch breaks. I also attended meetings, brainstorming sessions, and one field trip together with the department. During this phase more extensive observations of GIS in use were conducted at two different offices. Four surveyors were observed using the GIS over 5 days throughout the entire survey process: from the moment the survey was assigned by a secretary, to the survey report was electronically submitted using the GIS infrastructure. These observations and additional in-depth interviews took place a year after the GIS had been introduced, and illuminated how the GIS infrastructure had been appropriated in different local offices and situated practices. During the third phase (February – October 2001), a few additional informants were interviewed in a fifth office and some follow-up interviews at the HQ were conducted. All phases together, 61 in-depth interviews were conducted of 41 informants. The interviews lasted between 1 and 3 hours, and around 30 of them were tape-recorded and transcribed. Most of the interviews were unstructured in the sense that I only had a small list of a few basic key words to guide my interviewing. Other data sources included observation notes from studying use of the GIS, notes from participating in meetings, MCC’s intranet, and various types of printed documents including news bulletins, technical reports, evaluation reports, design documents, and requirements specifications. I was also given detailed demonstrations of the different features of the GIS at different stages in the process as well as prototypes of future releases. Type of informant Number of informants Surveyors 13 Support personnel 6 General managers 10 Local managers 3 GIS development team members 7 Superusers 2 Total 41 informants Table 1: Categories of informants in the case study Theory on information infrastructures was used as an initial guide for data collection (e.g. Ciborra et al., 2000; Star and Ruhleder, 1996). The collected data were analyzed in an iterative fashion using a subset of concepts and the theoretical framework of ‘reflexive modernization’ as outlined above. 3.2 Standardization as strategy for achieving quality In MCC, quality is strongly linked to standards. Apart from having issued around 10 percent of the world market of ISO 9000 and ISO 14000 certificates, standards and standardization is increasingly considered the crux for MCC’s internal production and organizing of ship surveys. For sure, in ship classification – standardization is absolutely vital to ensure that the end result of a ship survey is the same, independently of where the survey was conducted and which surveyor conducted it. Traditionally, uniformity in survey practice has been attempted inscribed through Classification Rules, survey checklists, and common courses for surveyors and managers. However, lack of uniformity in local practices – and particularly the differences between the way survey reports have been written, was frequently emphasised by managers as an up front problem and barrier for implementing the perceived necessary organizational changes: There have been so many variants out there. The local offices have had their own way of working…The reporting has been done in a very individual way – based on region, history and culture. (Manager) Thus, the design of the GIS followed a strong focus on standardization with the primary goal of increasing uniformity in how ship surveys are conducted and reported. A second organizational change expected enabled by the GIS, was a distributed survey process. Distributing the survey process would increase the flexibility for ship owners who are for obvious reasons not interested in having their ships docked simply because it takes MCC several days to conduct a survey. Thus, the GIS was designed to support one surveyor to start a survey on a ship in one port, passing an electronic “survey job” on to a second surveyor, before a third surveyor finishes the survey in a third port. Subsequently, to allow this to be conceived as ‘one survey’ – from the point of view of the customer, an agreed upon standard for how surveys must be conducted and reported across all 300 offices worldwide must be established. 3.3 Improving practice? uniformity at the lowest level A main inscription to ensure uniformity in the survey process was to include a socalled product model as the main information source (i.e. data model) for the GIS. As explained by a software developer in the GIS project, the product model was not simply understood as a common database for the various applications comprising a global information infrastructure, but rather perceived as a standard for surveyors’ ‘knowledge’: “Like TCP/IP is the standard on the Internet, the product model is the standard for the domain knowledge of MCC” (Software developer) Prior to the GIS project MCC had been participating in several international R&D projects related to product models and the development of the ISO 10303 standard referred to as STEP (STandards for the Exchange of Product Model Data). Product models are sophisticated and highly detailed data models especially designed for describing advanced mass-produced products like cars and aeroplanes. Such models have traditionally been used in relation to advanced CAD/CAM environments enabling 3D views of technical components and possibilities for developing statistics and trend analysis of individual components included in a product. In this way product models are designed for finding weak spots in complex constructions and to trace the performance of components over time. Managers vividly illustrated the different advantages of a product model based information system: “For example, whenever a user is registering a specific diesel engine – the user can just check the system and get access to relevant historical information. The representation in the product model will be totally consistent – for example an engine having the ID ‘MAN-341’ in the product model could never be exchanged for something else. Thus, this will imply improved quality…” (Manager) As underscored in the above quote, one of the key features of the GIS, was that the product model allowed for more detailed, accurate, and updated – as well as historical information on all ships. In contrast to paper-based reporting, which relied up on each surveyor’s ability to track down previous reports archived somewhere in the global organization, the GIS offers a range of possibilities for viewing detailed technical items and components, and associate CCs and MEMOs to these items electronically. The obvious advantage is that it becomes possible to develop advanced statistics based on surveyors’ reporting and that updated technical information is available on the tip of a keystroke independently of time and place. In this way, the product model inscribed uniformity in reporting by restricting the entering of collected data at the most detailed level in a complex hierarchical structure (i.e. an electronic checklist). In a product model technical information on a ship is modeled down to the very nitty-gritty detail, and henceforth, the entering of information into the system must conform to this level of detail – otherwise the product model will not be fully updated – as underscored in the GIS user guide: All recordings must be made at the lowest possible level (e.g.: make recordings on the “P/V valves” of a tanker, instead of the general category of “deck equipment”)” (my emphasis). In practice, from the point of view of surveyors this is experienced as quite confusing – and even ‘chaotic’: “I‘m always forced to enter information on the lowest and most detailed level. This is extremely time-consuming – and the work becomes very fragmented […] it's chaotic, [and] I miss the ability to have a view of the whole while I‘m working on a specific detail. (Surveyor) In contrast to the good intentions, this level of detail enforced by the product model seem to contradict surveyors’ practice in the sense that this level of detail is seldom needed in order to meet customers expectations and to deliver what surveyors’ perceive as a high quality report: “The quality can be judged by the report, […] you know, a high quality, professional looking document. And now it looks something that has been thrown together by a dyslectic word processor – and it does not have that – carry that sort of safe feel of conviction.” (Local manager) This was consistently reported by surveyors across cultures and departments in both European and Asian offices. 4 Analysis 4.1 Reskilling The GIS can be understood as a specific form of ‘disembedding mechanism’ in the way that non-local knowledge and abstract systems to an increasing extent structure the way that surveys are conducted and reported locally. Prior to the GIS, surveyor work also involved using abstract systems – like for instance checklists and company-wide ‘instructions for surveyors’, however the GIS to a larger extent imposes a specific sequence for how work is to be conducted. Thus, an interesting question is whether the GIS implies deskilling and loss of pre-existing forms local control. Giddens denies that disembedding mechanisms ultimately lead to the “steel hard” iron cage that determines human actions. Rather, abstract systems and human actions represent a duality in the sense that abstract systems get re-embedded in various local contexts: “The counterpart of displacement is reembedding. The disembedding mechanisms lift social relations and the exchange of information out of specific time-space contexts, but at the same time provide new opportunities for their reinsertion.” (Giddens, 1990: p. 141; my emphasis) In line with this, and certainly different from Braverman’s sombre ‘deskilling hypothesis’, the standardizing aspects of the GIS tends to transform surveyor work in different unexpected ways. An important element in these local transformations, and as a consequence of local re-embedding and appropriation (Ciborra, 1996) of the GIS in different contexts, surveyors are reskilled rather than deskilled. Reskilling is here related to ‘smart improvisations’ and juggling adopted by surveyors in order to align with the diversity across local offices, as for instance differences between customers: “[It] all depends upon the local offices’ relationships to the larger shipping companies. The larger shipping companies value written reports more than others. We have had examples of survey reports that have only been 3 lines long. The local fisherman will be happy with that – but not the larger shipping companies…” (Local Manager) The consequences of this is that surveyors have to juggle with the reports and apply various smart improvisations and work arounds in order to meet varying expectations and needs of the customers: “They juggle with it [the reports] by for instance importing it into Word there they modify it – so that we get one version in [GIS] and one paper-based version of the report.” (Superuser) To some extent, then, the quality of the survey reports is just as much related to each individual surveyor’s ability and skills in modifying generated reports, as the technical functionality of the GIS. Quality is in this sense not defined by the GIS, but produced through different heterogeneous elements that constitute surveyors’ work practice (cf. Berg, 1997). The quality is an effect of the alignment or particular configuration of survey reports, customers, managers, and the knowledge and skills of the surveyor. According to Giddens (1990: p. 80) “all disembedding mechanisms interact with reembedded contexts of action”. Likewise, even though the GIS have profound standardizing effects – surveyor work can never be carried out independently from local contexts. This is illustrated by the fact that surveyor work involves a large amount of ‘interactive service work’. In a similar way as described by Leidner (1998) in the case of McDonald’s, surveying involves ongoing negotiations and coordination with customers as well as managers and other surveyors. Independently of the survey type, surveyors are depend on co-operation with members of the crew to get the survey smoothly done and therefore they ‘juggle’ with the inscribed sequence of work: “It never happens that we follow this order of working [Surveyor points on the screen similar to the one shown on figure 1]. Most of the things we do – depends upon how the crew organizes things [...] Prior to conducting a survey, we always discuss how to organize things with the crew.” (Surveyor) In order to produce a service of high quality, surveyors deal with contingencies in their work by negotiating and organizing each survey job. Subsequently, surveyor practice necessarily deviate from the inscribed standard as surveyors negotiate and organize the scope and order of the practical survey job with the customer at the local site. 4.2 Shifting identity The globalizing dynamism of reflexive modernization can be directly linked to shifting identity of individuals. Both Beck and Giddens argue that it is not merely the institutions of modernity that becomes reflexive – but reflexive modernization also implies that biographies become self-reflexive (Beck, 1992; Giddens, 1991). To quote Giddens (1991: p. 32), “[Transformations] in self-identity and globalisation, I want to propose, are the two poles of the dialectic of the local and the global in conditions of high modernity”. Disembedding mechanisms and abstract systems, then, are involved not only in the shaping of a new institutional order, but also in individuals’ ongoing reconstruction of self-identity. However, this ‘individualization’ and the reflexive nature of knowledge implies that knowledge and expert systems become unstable and malleable, introduces a ‘radical doubt’ that creates existential anxiety and personal meaninglessness (Giddens, 1991). The introduction of the GIS in the work place – and especially its standardizing aspects provoked what Giddens refers to as ‘existential anxiety’ for some individual surveyors. The professional identity of the surveyors is in multiple ways linked to the survey report. ‘What it means to be a surveyor’ is by the surveyors themselves associated with being morally and professionally committed to writing high quality reports that are valued by customers and colleagues. Thus, for surveyors, survey reports are not merely a medium through which information is exchanged – it also reflects their professional identity and a way of signaling to others that they take their job seriously and do it well. Surveyors experienced the product model and the report generator that inscribed uniformity in local practices as quite dramatic changes to what it meant to be a surveyor: “The consequences will be that a surveyor’s job will change from having the role of a ‘verificator’ to be a ‘data collector’ for MCC … We have 130 years of experience and tradition to support the former – and suddenly everything is changing.” (Surveyor) Many of the surveyors were quite frustrated and distressed by this, in particular because they experienced that the system produced survey reports that contradicted their professional view on how surveys should be produced in order to meet the requirements of the customers: “[The] main frustration seems to be in defect reporting and basically getting in all information at the lowest level. […] [The] way we used to write reports, which was more a top down approach as opposed to bottom up did not fit in. What you issued was a condition of class [on an] international load line certificate. […] It was always linked to certificate, because until they were dealt with they couldn’t have a full time certificate. And now it’s linked to the lowest possible element, so you might – in some ways you can end up with being stupid.” (Local manager) In addition, MCC’s official policies and the user guidelines prohibited modifications to survey reports produced by the GIS. One implication of this unfortunate situation is that surveyors are caught up in a moral dilemma: should they modify survey reports in order to deliver what they perceive as high quality and professional engineering work – or should they blindly follow the inscribed procedure in the GIS? This illustrates well the new risks that are unintentionally introduced by global information infrastructures as tensions are created between their inherent ‘standardizing tendencies’ and the way that work is conceived and traditionally represented. Contrary to the intention, this actually threatens the quality of survey reports because the standardizing tendency of the GIS makes reports ‘anonymous’ and therefore also introduces a gap between the professional identity of the surveyor and the GIS. 4.3 New regimes of trust and risk Giddens identifies trust as the main mechanism that re-embed abstract systems in local contexts (Giddens, 1990). Important in Giddens conceptualization of trust is that trust can be understood as a form of “faith” in abstract systems, which is not necessarily linked to a cognitive understanding of those abstract systems (Giddens, 1990: p. 27). The GIS enabled splitting up surveys, so that different parts of a survey job could be conducted in different ports. This was expected to become a popular feature because it gave customers increased flexibility for when and where to conduct annual surveys. Especially, in more complicated surveys, which can take up to several weeks this can be invaluable for ship owners. In MCC this feature was often referred to as ‘commenced surveys’, and prior to the GIS they were literally impossible to conduct – or very difficult to conduct, because of the time lag and the variations in the paperbased reporting. In practice, however, even though GIS supported commenced surveys, they were seldom carried out due to accountability. Highly institutionalized practice in at least some of the regions of MCC, was that a survey report was always supposed to be signed-off by one single surveyor. Hence, the surveyors seemed to be reluctant to – and in many cases not willing to sign off a survey job that had partly been undertaken by another surveyor: “This got to do with accountability. We generally try not to hand unfinished things over, because at the end of the day there is somebody’s signature on it saying this is what I have done. It easily becomes messy if you say he did that and I did this…[…] It’s not that we don’t trust each other at all – it just not our way of doing it.” (Surveyor) The surveyors considered it less risky to draw upon past reports and memo’s linked to a particular ship electronically retrievable from the GIS. However, in such cases the surveyors often checked where the focal report or memo was written, because they tended to perceive certain offices and regions as more reliable and more committed to quality than others. In this way, over time, surveyors constructed different ‘taxonomies of trust’ in order to reduce risks related to use of non-local knowledge. Thus, increased amount of codified and standardized knowledge does not necessarily reduce the perceived risk in surveyors’ decision making. Initial goal Increase uniformity and quality of survey reporting Standardizing aspects of the GIS • • • Automatic generation of reports Recording at lowest level MCC policies prohibiting modification of reports Transformation of work and professional identity • • • • Distributed survey jobs Enable knowledge sharing across local offices • • Splitting up survey jobs Recording according to a predefined set of checklist items • Access to MEMOs, reports, and ship status Standardized way of recording and reporting • • • • Reskilling ‘Radical doubt’ in surveyors’ work: What are the basic principles in producing high quality reports? Surveyors trying to ‘reconstruct’ their selfidentity; What is the role of a surveyor? Introducing risks of producing survey reports of low quality Problem of accountability Lack of systems of trust to support distribution Surveyors’ establishing various informal ‘taxonomies of trust’ to reduce risks of non-local knowledge Table 2: The relationship between the intentions, the standardizing aspects of the GIS infrastructure, and transformation of work and professional identity of surveyors. 5 Some implications for the practice of global information infrastructures The evidence from the case of MCC suggests that implementation of global information infrastructures involve different kinds of complexities and challenges than usually stressed in the literature. Contemporary literature on IT infrastructure often takes a ‘strategic alignment’ perspective and focuses on various up-front barriers and problems related to company-wide infrastructures (e.g. Weill and Broadbent, 1998). Thus implicitly taking it for granted that as long as these barriers are overcome, all the anticipated changes would occur. In contrast, in the case of MCC, most of the unexpected transformations occurred over time after the initial implementation the GIS. A considerable amount of time and resources was spent in order to get the GIS to work effectively in local offices after the GIS infrastructure was technically installed. Since the GIS was introduced across different communities-of-practice in different parts of the world, the way that it represented and organized surveyor work did not align equally well with the variety of situations and contexts of usage. This is by no means special for the MCC case. Rather, this can be perceived as an inherent tension in information infrastructures, which need to be customisable and flexible in local use on the one hand, and standardized on the other (Hanseth, Monteiro, and Hatling, 1996; Star and Ruhleder, 1996). This underscores that global information infrastructures need to be, not only ‘implemented’ in the sense of user training and configuration of infrastructure, but made sense of and re-embedded in local settings. Re-embedding requires practical problem solving as well as incremental learning and reskilling. As a consequence, and in a similar way to groupware, global information infrastructures seem to drift (Ciborra, 1996). However, ‘drifting’ in this case is closely linked to individuals’ striving to re-construct their self-identity in light of the new standardizing aspects inscribed in the GIS. Furthermore, certain uses of the GIS required new systems of trust to be established in order to reduce the risk for each individual surveyor. A crucial point to be emphasised here is that while the standardizing aspects of the GIS are necessary for electronic reporting, interaction and knowledge sharing across distance, this also implies tensions in professional identity of surveyors. In the case of MCC this resulted in ‘radical doubt’ concerning how work should be conducted, and thus potentially undermining initial intentions of increasing quality and efficiency in the production of surveys. What can be concluded from this, is that standardizing tendencies of global information infrastructures seem to be likely to transform individuals’ identity in unexpected ways that might undermine anticipated organizational changes expected from the implementation of a global information infrastructure. Thus, global information infrastructures seem to be reflexive in the sense that that risks and unexpected transformations are inherently produced, re-produced, and diffused through use and management of the technology itself. Global information infrastructures like the GIS, tend to produce new risks which originate in the interplay between the local and the global (e.g. individual surveyor’s identity and standardizing aspects of GIS). Global information infrastructures that contradicts deep-seated professional identity of users through ‘disembedding’ of local interaction and knowledge, might lead to ‘personal meaninglessness’ and ‘existential anxiety’ in individuals. This should be considered important for management as this might strike back as poor decision making and lack of commitment. This illustrates what Beck refers to as the boomerang effect, which captures one of the key characteristics of reflexive modernity that “latent side effects strike back even at the centers of their production” (Beck, 1992: p. 37). Thus, a salient point here is that different risks are related to design and use of global information infrastructures at different interconnected levels as summarised in the table below: Actors Examples of risks related to design and use of global information infrastructures Individual users • Personal meaninglessness and existential anxiety • Doubt in decision making; risks become individualized Designers and managers of the technology • Difficult to assess the quality of work carried out • Difficult to control and foresee the effect of change initiatives and new infrastructural components Organization as a whole • Implementation of infrastructural technologies with ‘decreasing returns’ (e.g. skyrocketing maintenance costs, continuos re-designs, nonworking technologies) • Because of globalization it becomes increasingly important to change, however it also becomes increasingly difficult and risky • Varying quality in products and services provided Customers and other stakeholders Table 3: Different risks for different actors. However, it is important to emphasize that global information infrastructures also implies various opportunities for individuals and organizations, and that better design strategies and careful implementations are important. The point is, however, that information infrastructures carry with them new kinds of risks that are not entirely related to design of these technologies as such, but also related to the way that they integrate and standardize work practices. Exactly because they are inherently reflexive and transform work and identities in uncontrollable and unexpected ways, traditional ISD practices and techniques cannot tackle, let alone regain control over these problems. Surely, information infrastructures enable new innovative ways of working, however, to invoke Beck, “Along with the growing capacity of technical options [Zweckrationalität] grows the incalculability of their consequences” (Beck, 1992: p.22). 6 Acknowledgements I’m grateful to Edoardo Iacucci and Margunn Aanestad for their thoughtful comments and suggestions for improving various drafts and versions of this paper. 7 References Barrett, M. and Walsham, G. 1999. Electronic Trading and Work Transformation in the London Market. Information Systems Research, Vol. 10, No.1, pp. 1-22. Beck, U. 1992. Risk Society – Towards a New Modernity. London: Sage. Beck, U. 1994. The Reinvention of Politics: Towards a Theory of Reflexive Modernization. In: Beck, U., Giddens, A. and Lash, S. Reflexive Modernization – Politics, Tradition and Aesthetics in the Modern Social Order. Cambridge: Polity Press, pp. 1-55. Bradley, S.P., Hausman, J.A., and Nolan, R.L. (eds.). 1993. Globalization, technology, and competition: the fusion of computers and telecommunications in the 1990s. Boston, Mass.: Harvard Business School Press. Ciborra, C.U. 1996. Chapter 1. Introduction: What does Groupware Mean for the Organizations Hosting it?, In: Ciborra, C.U. (ed.). Groupware and Teamwork: invisible aid or technical hindrance? Chichester, U.K.: John Wiley & Sons. Ciborra, C.U., Braa, K., Cordella, A., Dahlbom, B., Failla, A., Hanseth, O., Hepsø, V., Ljungberg, J., Monteiro, E. and Simon, K.A. 2000. From Control to Drift. Oxford: Oxford University Press. Davenport, T.H. 1998. Putting the enterprise into the enterprise system. Harvard Business Review, July-August, pp. 121-131. Earl, M.J. and Fenny, D.F. 1996. Information Systems in Global Business: Evidence from European Multinationals. In: Earl, M.J. (ed.). Information Management. Oxford: Oxford University Press. Giddens, A. 1990. The Consequences of Modernity. Cambridge, UK: Polity Press. Giddens, A. 1991. Modernity and self-identity. Cambridge: Polity Press. Giddens, A. 1994. Living in a Post-Traditional Society. In: Beck, U., Giddens, A. and Lash, S. Reflexive Modernization – Politics, Tradition and Aesthetics in the Modern Social Order. Cambridge: Polity Press, pp. 56-109. Giddens, A. 1999. Runaway World. London: Profile Books. Hanseth, O. and Braa, K. 2000. Globalization and ‘Risk Society’. In: Ciborra, C.U., Braa, K., Cordella, A., Dahlbom, B., Failla, A., Hanseth, O., Hepsø, V., Ljungberg, J., Monteiro, E. and Simon, K.A. 2000. From Control to Drift. Oxford: Oxford University Press, pp. 41-55. Hanseth, O., Ciborra, C.U, and Braa, K. 2002. The Control Devolution: ERP and the Side Effects of Globalization. The DATA BASE for Advances in Information Systems – Fall 2001, Vol. 32, No. 4, pp. 34-46. Hanseth, O., Monteiro, E. and Hatling, M. 1996. Developing Information Infrastructure: The Tension Between Standardization and Flexibility. Science, Technology, & Human Values, Vol. 21, No. 4., pp. 407-426. Ives, B. and Jarvenpaa, S.L. 1991. Applications of Global Information Technology: Key Issues for Management. MIS Quarterly, 15(March): pp. 32-49 Klein, H.K. and Myers, M.D. 1999. A Set of Principles for Conducting and Evaluating Interpretive Field Studies in Information Systems. MIS Quarterly, vol. 23, no.1, pp. 67-93. Leidner, R. 1993. Fast Food, Fast Talk. Berkely: University of California Press. Rolland, K.H. and Monteiro, E. 2002. Balancing the Local and the Global in Infrastructural Information Systems. The Information Society Journal, Vol. 18, No. 2., pp. 87-100. Star, S.L. and Ruhleder, K. 1996. Steps Toward an Ecology of Infrastructure: Design and Access for Large Information Spaces. Information Systems Research, Vol. 7, No. 1, pp. 111-134. Walsham, G. 1998. IT and Changing Professional Identity: Micro-Studies and MacroTheory. Journal og the American Society for Information Science, Vol. 49, No. 12, pp. 1081-1089. Walsham, G. 2000. Globalization and IT: Agenda for Research. In: Baskerville, R., Stage, J., and DeGross, J.I. (eds.) 2000. Organizational and Social Perspectives on Information Technology. Boston: Kluwer Academic Publishers, pp. 195-210. Walsham, G. 2001. Making a world of difference: IT in a global context. Chichester: Wiley. Weill, P and Broadbent, M. 1998. Leveraging the New Infrastructure: How Market Leaders Capitalize on Information. Boston: Harvard Business School Press.